diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,297152 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9999941101981918,
+  "eval_steps": 500,
+  "global_step": 42446,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.663822369435805,
+      "learning_rate": 1.569858712715856e-08,
+      "loss": 0.5153,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9327842301287814,
+      "learning_rate": 3.139717425431712e-08,
+      "loss": 0.8556,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6594968950585425,
+      "learning_rate": 4.709576138147567e-08,
+      "loss": 0.7564,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.648598021617063,
+      "learning_rate": 6.279434850863424e-08,
+      "loss": 0.9297,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7383865036769468,
+      "learning_rate": 7.849293563579278e-08,
+      "loss": 0.7813,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7991074305010986,
+      "learning_rate": 9.419152276295134e-08,
+      "loss": 0.9154,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8821798137769885,
+      "learning_rate": 1.098901098901099e-07,
+      "loss": 0.7036,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9221142632736166,
+      "learning_rate": 1.2558869701726847e-07,
+      "loss": 0.7322,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6620792992117104,
+      "learning_rate": 1.4128728414442702e-07,
+      "loss": 0.8058,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7591411662010912,
+      "learning_rate": 1.5698587127158556e-07,
+      "loss": 0.9155,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6343979201572483,
+      "learning_rate": 1.726844583987441e-07,
+      "loss": 0.8402,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6589410717383055,
+      "learning_rate": 1.8838304552590267e-07,
+      "loss": 0.885,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8109692517587606,
+      "learning_rate": 2.0408163265306121e-07,
+      "loss": 0.8047,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9095480158091243,
+      "learning_rate": 2.197802197802198e-07,
+      "loss": 0.8298,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6015168694502875,
+      "learning_rate": 2.3547880690737838e-07,
+      "loss": 0.82,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8644450347608426,
+      "learning_rate": 2.5117739403453695e-07,
+      "loss": 0.7735,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.669389148507826,
+      "learning_rate": 2.6687598116169546e-07,
+      "loss": 0.8306,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.819166462371553,
+      "learning_rate": 2.8257456828885403e-07,
+      "loss": 0.8664,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5840739224125565,
+      "learning_rate": 2.982731554160126e-07,
+      "loss": 0.5106,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6059807916530795,
+      "learning_rate": 3.139717425431711e-07,
+      "loss": 0.8243,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.502549481758672,
+      "learning_rate": 3.296703296703297e-07,
+      "loss": 0.7027,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.0292661470814246,
+      "learning_rate": 3.453689167974882e-07,
+      "loss": 0.8922,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.806543342767738,
+      "learning_rate": 3.610675039246468e-07,
+      "loss": 0.9201,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8222261938648925,
+      "learning_rate": 3.7676609105180534e-07,
+      "loss": 0.8439,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8138026697344025,
+      "learning_rate": 3.924646781789639e-07,
+      "loss": 1.0019,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.923688919593931,
+      "learning_rate": 4.0816326530612243e-07,
+      "loss": 0.9369,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.668344532455656,
+      "learning_rate": 4.2386185243328105e-07,
+      "loss": 0.7455,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9425454307178411,
+      "learning_rate": 4.395604395604396e-07,
+      "loss": 0.7784,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.784718668115782,
+      "learning_rate": 4.5525902668759813e-07,
+      "loss": 0.8556,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6073502069646306,
+      "learning_rate": 4.7095761381475676e-07,
+      "loss": 0.8237,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5734848061296014,
+      "learning_rate": 4.866562009419152e-07,
+      "loss": 0.9655,
+      "step": 31
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9964005380234628,
+      "learning_rate": 5.023547880690739e-07,
+      "loss": 0.7722,
+      "step": 32
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7663800465822717,
+      "learning_rate": 5.180533751962324e-07,
+      "loss": 0.9779,
+      "step": 33
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6472557228507425,
+      "learning_rate": 5.337519623233909e-07,
+      "loss": 0.8311,
+      "step": 34
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.988877163818641,
+      "learning_rate": 5.494505494505495e-07,
+      "loss": 0.8583,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8650349024769093,
+      "learning_rate": 5.651491365777081e-07,
+      "loss": 0.7628,
+      "step": 36
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.689778379114679,
+      "learning_rate": 5.808477237048666e-07,
+      "loss": 0.8175,
+      "step": 37
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9291559869005865,
+      "learning_rate": 5.965463108320252e-07,
+      "loss": 0.5413,
+      "step": 38
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.818327202356103,
+      "learning_rate": 6.122448979591837e-07,
+      "loss": 0.9568,
+      "step": 39
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6235380494373423,
+      "learning_rate": 6.279434850863422e-07,
+      "loss": 0.5068,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6012400417818065,
+      "learning_rate": 6.436420722135009e-07,
+      "loss": 0.5236,
+      "step": 41
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6465211630282854,
+      "learning_rate": 6.593406593406594e-07,
+      "loss": 0.8648,
+      "step": 42
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.106837139927634,
+      "learning_rate": 6.750392464678179e-07,
+      "loss": 0.9021,
+      "step": 43
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8191257671255665,
+      "learning_rate": 6.907378335949764e-07,
+      "loss": 0.7146,
+      "step": 44
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6720867811077527,
+      "learning_rate": 7.064364207221351e-07,
+      "loss": 0.7677,
+      "step": 45
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.4752284253625976,
+      "learning_rate": 7.221350078492937e-07,
+      "loss": 0.8696,
+      "step": 46
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.57868923188503,
+      "learning_rate": 7.378335949764521e-07,
+      "loss": 0.4962,
+      "step": 47
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9074697748394578,
+      "learning_rate": 7.535321821036107e-07,
+      "loss": 0.8604,
+      "step": 48
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4443148182189924,
+      "learning_rate": 7.692307692307694e-07,
+      "loss": 0.7358,
+      "step": 49
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6911921528091178,
+      "learning_rate": 7.849293563579278e-07,
+      "loss": 0.8794,
+      "step": 50
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7304639271635032,
+      "learning_rate": 8.006279434850864e-07,
+      "loss": 0.7793,
+      "step": 51
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6719737063121578,
+      "learning_rate": 8.163265306122449e-07,
+      "loss": 0.8134,
+      "step": 52
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5663850351363169,
+      "learning_rate": 8.320251177394035e-07,
+      "loss": 0.9659,
+      "step": 53
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9913088213847177,
+      "learning_rate": 8.477237048665621e-07,
+      "loss": 0.8074,
+      "step": 54
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6089788205755233,
+      "learning_rate": 8.634222919937206e-07,
+      "loss": 0.767,
+      "step": 55
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.441796534977343,
+      "learning_rate": 8.791208791208792e-07,
+      "loss": 0.7595,
+      "step": 56
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.382924108034673,
+      "learning_rate": 8.948194662480378e-07,
+      "loss": 0.6623,
+      "step": 57
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.728859114563259,
+      "learning_rate": 9.105180533751963e-07,
+      "loss": 0.7529,
+      "step": 58
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5182005306949917,
+      "learning_rate": 9.262166405023548e-07,
+      "loss": 0.6615,
+      "step": 59
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6811511967021056,
+      "learning_rate": 9.419152276295135e-07,
+      "loss": 0.8834,
+      "step": 60
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.0166731528948643,
+      "learning_rate": 9.57613814756672e-07,
+      "loss": 0.7845,
+      "step": 61
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5969628749866691,
+      "learning_rate": 9.733124018838304e-07,
+      "loss": 0.7672,
+      "step": 62
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.470462612267645,
+      "learning_rate": 9.890109890109891e-07,
+      "loss": 0.8996,
+      "step": 63
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6003949222359775,
+      "learning_rate": 1.0047095761381478e-06,
+      "loss": 0.9674,
+      "step": 64
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6304535599654086,
+      "learning_rate": 1.0204081632653063e-06,
+      "loss": 0.7809,
+      "step": 65
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5895518800125619,
+      "learning_rate": 1.0361067503924647e-06,
+      "loss": 0.8839,
+      "step": 66
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3981918867533745,
+      "learning_rate": 1.0518053375196232e-06,
+      "loss": 0.738,
+      "step": 67
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3466809199194771,
+      "learning_rate": 1.0675039246467819e-06,
+      "loss": 0.66,
+      "step": 68
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4526371553364266,
+      "learning_rate": 1.0832025117739405e-06,
+      "loss": 0.832,
+      "step": 69
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4310893875978292,
+      "learning_rate": 1.098901098901099e-06,
+      "loss": 0.8051,
+      "step": 70
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5354289750273207,
+      "learning_rate": 1.1145996860282575e-06,
+      "loss": 0.8622,
+      "step": 71
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5345966840734842,
+      "learning_rate": 1.1302982731554161e-06,
+      "loss": 0.7406,
+      "step": 72
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6007458315524439,
+      "learning_rate": 1.1459968602825746e-06,
+      "loss": 0.4684,
+      "step": 73
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3710834129490967,
+      "learning_rate": 1.1616954474097333e-06,
+      "loss": 0.755,
+      "step": 74
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7361246349709085,
+      "learning_rate": 1.1773940345368917e-06,
+      "loss": 0.8371,
+      "step": 75
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.54880490473918,
+      "learning_rate": 1.1930926216640504e-06,
+      "loss": 0.7344,
+      "step": 76
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3478222246238343,
+      "learning_rate": 1.2087912087912089e-06,
+      "loss": 0.6257,
+      "step": 77
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4433187172240418,
+      "learning_rate": 1.2244897959183673e-06,
+      "loss": 0.7954,
+      "step": 78
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5769917589548887,
+      "learning_rate": 1.240188383045526e-06,
+      "loss": 0.7886,
+      "step": 79
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5937341106399976,
+      "learning_rate": 1.2558869701726845e-06,
+      "loss": 0.517,
+      "step": 80
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3477580966897102,
+      "learning_rate": 1.2715855572998431e-06,
+      "loss": 0.6881,
+      "step": 81
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4360741414449292,
+      "learning_rate": 1.2872841444270018e-06,
+      "loss": 0.8533,
+      "step": 82
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3747340560114572,
+      "learning_rate": 1.30298273155416e-06,
+      "loss": 0.9223,
+      "step": 83
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8845243067790682,
+      "learning_rate": 1.3186813186813187e-06,
+      "loss": 0.7442,
+      "step": 84
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6226662958821981,
+      "learning_rate": 1.3343799058084774e-06,
+      "loss": 0.5094,
+      "step": 85
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4787296677776682,
+      "learning_rate": 1.3500784929356359e-06,
+      "loss": 0.7759,
+      "step": 86
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5739355602668312,
+      "learning_rate": 1.3657770800627946e-06,
+      "loss": 0.5192,
+      "step": 87
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.211454656185043,
+      "learning_rate": 1.3814756671899528e-06,
+      "loss": 0.5866,
+      "step": 88
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3355673997897535,
+      "learning_rate": 1.3971742543171115e-06,
+      "loss": 0.756,
+      "step": 89
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.1924352294103557,
+      "learning_rate": 1.4128728414442702e-06,
+      "loss": 0.6864,
+      "step": 90
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5027946883416932,
+      "learning_rate": 1.4285714285714286e-06,
+      "loss": 0.8637,
+      "step": 91
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3166417484868087,
+      "learning_rate": 1.4442700156985873e-06,
+      "loss": 0.8807,
+      "step": 92
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3800386875062236,
+      "learning_rate": 1.459968602825746e-06,
+      "loss": 0.7754,
+      "step": 93
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3641369975594158,
+      "learning_rate": 1.4756671899529042e-06,
+      "loss": 0.7467,
+      "step": 94
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5852860552234955,
+      "learning_rate": 1.491365777080063e-06,
+      "loss": 0.4831,
+      "step": 95
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6220255019003587,
+      "learning_rate": 1.5070643642072214e-06,
+      "loss": 0.656,
+      "step": 96
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5167226156810827,
+      "learning_rate": 1.52276295133438e-06,
+      "loss": 0.745,
+      "step": 97
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5553144622074345,
+      "learning_rate": 1.5384615384615387e-06,
+      "loss": 0.5126,
+      "step": 98
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5180367128293153,
+      "learning_rate": 1.554160125588697e-06,
+      "loss": 0.9081,
+      "step": 99
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3700839240402132,
+      "learning_rate": 1.5698587127158556e-06,
+      "loss": 0.7766,
+      "step": 100
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.424529275892127,
+      "learning_rate": 1.5855572998430143e-06,
+      "loss": 0.7862,
+      "step": 101
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.317864558677452,
+      "learning_rate": 1.6012558869701728e-06,
+      "loss": 0.8974,
+      "step": 102
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.526486607970249,
+      "learning_rate": 1.6169544740973315e-06,
+      "loss": 0.7954,
+      "step": 103
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.2990713108637604,
+      "learning_rate": 1.6326530612244897e-06,
+      "loss": 0.7618,
+      "step": 104
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5973690776397802,
+      "learning_rate": 1.6483516483516484e-06,
+      "loss": 0.6394,
+      "step": 105
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5116575602331392,
+      "learning_rate": 1.664050235478807e-06,
+      "loss": 0.8206,
+      "step": 106
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5308089105466773,
+      "learning_rate": 1.6797488226059655e-06,
+      "loss": 0.7696,
+      "step": 107
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.2937132971846526,
+      "learning_rate": 1.6954474097331242e-06,
+      "loss": 0.4683,
+      "step": 108
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3854937554074511,
+      "learning_rate": 1.7111459968602829e-06,
+      "loss": 0.6561,
+      "step": 109
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.326835957375491,
+      "learning_rate": 1.7268445839874411e-06,
+      "loss": 0.7058,
+      "step": 110
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.376736722991603,
+      "learning_rate": 1.7425431711145998e-06,
+      "loss": 0.6636,
+      "step": 111
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5785092224915072,
+      "learning_rate": 1.7582417582417585e-06,
+      "loss": 0.5333,
+      "step": 112
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4917924524751216,
+      "learning_rate": 1.773940345368917e-06,
+      "loss": 0.728,
+      "step": 113
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4431000946653934,
+      "learning_rate": 1.7896389324960756e-06,
+      "loss": 0.8649,
+      "step": 114
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5701092115984455,
+      "learning_rate": 1.8053375196232339e-06,
+      "loss": 0.5152,
+      "step": 115
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6487623058990981,
+      "learning_rate": 1.8210361067503925e-06,
+      "loss": 0.7627,
+      "step": 116
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3684918222179747,
+      "learning_rate": 1.8367346938775512e-06,
+      "loss": 0.7557,
+      "step": 117
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4013998376198706,
+      "learning_rate": 1.8524332810047097e-06,
+      "loss": 0.8569,
+      "step": 118
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4840718319244108,
+      "learning_rate": 1.8681318681318684e-06,
+      "loss": 0.7246,
+      "step": 119
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5381844009849792,
+      "learning_rate": 1.883830455259027e-06,
+      "loss": 0.8148,
+      "step": 120
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3688001290603382,
+      "learning_rate": 1.8995290423861853e-06,
+      "loss": 0.7931,
+      "step": 121
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.326137683416932,
+      "learning_rate": 1.915227629513344e-06,
+      "loss": 0.8245,
+      "step": 122
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.2928332628155081,
+      "learning_rate": 1.9309262166405024e-06,
+      "loss": 0.7642,
+      "step": 123
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4824641220170909,
+      "learning_rate": 1.946624803767661e-06,
+      "loss": 0.6017,
+      "step": 124
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9896936678123105,
+      "learning_rate": 1.9623233908948198e-06,
+      "loss": 0.8897,
+      "step": 125
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.2939934389818097,
+      "learning_rate": 1.9780219780219782e-06,
+      "loss": 0.8072,
+      "step": 126
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.2819536002023468,
+      "learning_rate": 1.9937205651491367e-06,
+      "loss": 0.7428,
+      "step": 127
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4588569475612156,
+      "learning_rate": 2.0094191522762956e-06,
+      "loss": 0.7975,
+      "step": 128
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5195375264021,
+      "learning_rate": 2.0251177394034536e-06,
+      "loss": 0.8239,
+      "step": 129
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.405284256001685,
+      "learning_rate": 2.0408163265306125e-06,
+      "loss": 0.9551,
+      "step": 130
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.553033185915086,
+      "learning_rate": 2.056514913657771e-06,
+      "loss": 0.7051,
+      "step": 131
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3084650490585061,
+      "learning_rate": 2.0722135007849294e-06,
+      "loss": 0.66,
+      "step": 132
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.698945069984713,
+      "learning_rate": 2.0879120879120883e-06,
+      "loss": 0.7209,
+      "step": 133
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.096634053109896,
+      "learning_rate": 2.1036106750392464e-06,
+      "loss": 0.8244,
+      "step": 134
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5791221538905906,
+      "learning_rate": 2.1193092621664052e-06,
+      "loss": 0.8137,
+      "step": 135
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4222042747318173,
+      "learning_rate": 2.1350078492935637e-06,
+      "loss": 0.698,
+      "step": 136
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3581046433589263,
+      "learning_rate": 2.150706436420722e-06,
+      "loss": 0.7105,
+      "step": 137
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3130664176884104,
+      "learning_rate": 2.166405023547881e-06,
+      "loss": 0.8047,
+      "step": 138
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5454352964843012,
+      "learning_rate": 2.1821036106750395e-06,
+      "loss": 0.7588,
+      "step": 139
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4031617363644104,
+      "learning_rate": 2.197802197802198e-06,
+      "loss": 0.72,
+      "step": 140
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4615689205299716,
+      "learning_rate": 2.2135007849293564e-06,
+      "loss": 0.6849,
+      "step": 141
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.323876672771436,
+      "learning_rate": 2.229199372056515e-06,
+      "loss": 0.7473,
+      "step": 142
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4706717157773799,
+      "learning_rate": 2.244897959183674e-06,
+      "loss": 0.7378,
+      "step": 143
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.460688887535409,
+      "learning_rate": 2.2605965463108323e-06,
+      "loss": 0.8495,
+      "step": 144
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.2305064834072263,
+      "learning_rate": 2.2762951334379907e-06,
+      "loss": 0.7673,
+      "step": 145
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3797692183922952,
+      "learning_rate": 2.291993720565149e-06,
+      "loss": 0.8751,
+      "step": 146
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6935128013532186,
+      "learning_rate": 2.307692307692308e-06,
+      "loss": 0.8182,
+      "step": 147
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3684092571644122,
+      "learning_rate": 2.3233908948194665e-06,
+      "loss": 0.5974,
+      "step": 148
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.908554574484203,
+      "learning_rate": 2.339089481946625e-06,
+      "loss": 0.8409,
+      "step": 149
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.633383474446539,
+      "learning_rate": 2.3547880690737835e-06,
+      "loss": 0.5194,
+      "step": 150
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3331432654589586,
+      "learning_rate": 2.370486656200942e-06,
+      "loss": 0.6852,
+      "step": 151
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.430936836234601,
+      "learning_rate": 2.386185243328101e-06,
+      "loss": 0.8188,
+      "step": 152
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5974059938156016,
+      "learning_rate": 2.4018838304552593e-06,
+      "loss": 0.6886,
+      "step": 153
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6247738929087859,
+      "learning_rate": 2.4175824175824177e-06,
+      "loss": 0.5337,
+      "step": 154
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6875582149869328,
+      "learning_rate": 2.4332810047095766e-06,
+      "loss": 0.8204,
+      "step": 155
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5918871014089198,
+      "learning_rate": 2.4489795918367347e-06,
+      "loss": 0.5037,
+      "step": 156
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6450493762757357,
+      "learning_rate": 2.4646781789638936e-06,
+      "loss": 0.7108,
+      "step": 157
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3652758797735693,
+      "learning_rate": 2.480376766091052e-06,
+      "loss": 0.7552,
+      "step": 158
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3695942928110418,
+      "learning_rate": 2.4960753532182105e-06,
+      "loss": 0.6935,
+      "step": 159
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3647625753903685,
+      "learning_rate": 2.511773940345369e-06,
+      "loss": 0.7506,
+      "step": 160
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.35637621724195,
+      "learning_rate": 2.5274725274725274e-06,
+      "loss": 0.675,
+      "step": 161
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7085709763848371,
+      "learning_rate": 2.5431711145996863e-06,
+      "loss": 0.7904,
+      "step": 162
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3776378344900218,
+      "learning_rate": 2.5588697017268448e-06,
+      "loss": 0.8072,
+      "step": 163
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5763882580253433,
+      "learning_rate": 2.5745682888540036e-06,
+      "loss": 0.5027,
+      "step": 164
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3892700372779936,
+      "learning_rate": 2.5902668759811617e-06,
+      "loss": 0.7479,
+      "step": 165
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5002797705296935,
+      "learning_rate": 2.60596546310832e-06,
+      "loss": 0.7157,
+      "step": 166
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6361600773431728,
+      "learning_rate": 2.621664050235479e-06,
+      "loss": 0.6501,
+      "step": 167
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3881580639527598,
+      "learning_rate": 2.6373626373626375e-06,
+      "loss": 0.6899,
+      "step": 168
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.392806570886587,
+      "learning_rate": 2.6530612244897964e-06,
+      "loss": 0.7585,
+      "step": 169
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5887904570954188,
+      "learning_rate": 2.668759811616955e-06,
+      "loss": 0.7117,
+      "step": 170
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.337772555856093,
+      "learning_rate": 2.684458398744113e-06,
+      "loss": 0.7488,
+      "step": 171
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6999759754012629,
+      "learning_rate": 2.7001569858712718e-06,
+      "loss": 0.6882,
+      "step": 172
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5103144657926646,
+      "learning_rate": 2.7158555729984302e-06,
+      "loss": 0.775,
+      "step": 173
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5100000152180575,
+      "learning_rate": 2.731554160125589e-06,
+      "loss": 0.7131,
+      "step": 174
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.922109717723067,
+      "learning_rate": 2.7472527472527476e-06,
+      "loss": 0.7481,
+      "step": 175
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6677725292328816,
+      "learning_rate": 2.7629513343799056e-06,
+      "loss": 0.7375,
+      "step": 176
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.893748805382321,
+      "learning_rate": 2.7786499215070645e-06,
+      "loss": 0.7813,
+      "step": 177
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5742301751996294,
+      "learning_rate": 2.794348508634223e-06,
+      "loss": 0.4918,
+      "step": 178
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4477261320082637,
+      "learning_rate": 2.810047095761382e-06,
+      "loss": 0.7379,
+      "step": 179
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.042971536315801,
+      "learning_rate": 2.8257456828885403e-06,
+      "loss": 0.7657,
+      "step": 180
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7115528015508341,
+      "learning_rate": 2.8414442700156984e-06,
+      "loss": 0.8552,
+      "step": 181
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4533374528263592,
+      "learning_rate": 2.8571428571428573e-06,
+      "loss": 0.9294,
+      "step": 182
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8176608135157872,
+      "learning_rate": 2.8728414442700157e-06,
+      "loss": 0.5986,
+      "step": 183
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.407336864855272,
+      "learning_rate": 2.8885400313971746e-06,
+      "loss": 0.7699,
+      "step": 184
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7517049933182047,
+      "learning_rate": 2.904238618524333e-06,
+      "loss": 0.7849,
+      "step": 185
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6745161342734352,
+      "learning_rate": 2.919937205651492e-06,
+      "loss": 0.8078,
+      "step": 186
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.44726786109749,
+      "learning_rate": 2.93563579277865e-06,
+      "loss": 0.7406,
+      "step": 187
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4819302466950177,
+      "learning_rate": 2.9513343799058085e-06,
+      "loss": 0.7818,
+      "step": 188
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6100649477317823,
+      "learning_rate": 2.9670329670329673e-06,
+      "loss": 0.6655,
+      "step": 189
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4766621057486817,
+      "learning_rate": 2.982731554160126e-06,
+      "loss": 0.7638,
+      "step": 190
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.1949243184143126,
+      "learning_rate": 2.9984301412872847e-06,
+      "loss": 0.6929,
+      "step": 191
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4496332995335866,
+      "learning_rate": 3.0141287284144427e-06,
+      "loss": 0.839,
+      "step": 192
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.0426138787105006,
+      "learning_rate": 3.029827315541601e-06,
+      "loss": 0.7982,
+      "step": 193
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3321231184883422,
+      "learning_rate": 3.04552590266876e-06,
+      "loss": 0.683,
+      "step": 194
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.318946033213783,
+      "learning_rate": 3.0612244897959185e-06,
+      "loss": 0.7874,
+      "step": 195
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6241078648928259,
+      "learning_rate": 3.0769230769230774e-06,
+      "loss": 0.4836,
+      "step": 196
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6170847087710375,
+      "learning_rate": 3.092621664050236e-06,
+      "loss": 0.5223,
+      "step": 197
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.2958473738078766,
+      "learning_rate": 3.108320251177394e-06,
+      "loss": 0.7692,
+      "step": 198
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6210274679960512,
+      "learning_rate": 3.124018838304553e-06,
+      "loss": 0.4804,
+      "step": 199
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6709771130560036,
+      "learning_rate": 3.1397174254317113e-06,
+      "loss": 0.7233,
+      "step": 200
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3360662272497894,
+      "learning_rate": 3.15541601255887e-06,
+      "loss": 0.7439,
+      "step": 201
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6136323970807174,
+      "learning_rate": 3.1711145996860286e-06,
+      "loss": 0.4829,
+      "step": 202
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3091796330964218,
+      "learning_rate": 3.1868131868131867e-06,
+      "loss": 0.7406,
+      "step": 203
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3138188172689351,
+      "learning_rate": 3.2025117739403456e-06,
+      "loss": 0.8224,
+      "step": 204
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4968227357814052,
+      "learning_rate": 3.218210361067504e-06,
+      "loss": 0.7219,
+      "step": 205
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.51347838411381,
+      "learning_rate": 3.233908948194663e-06,
+      "loss": 0.6944,
+      "step": 206
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.256596376928321,
+      "learning_rate": 3.2496075353218214e-06,
+      "loss": 0.779,
+      "step": 207
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8712801698565569,
+      "learning_rate": 3.2653061224489794e-06,
+      "loss": 0.7563,
+      "step": 208
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5773593966267709,
+      "learning_rate": 3.2810047095761383e-06,
+      "loss": 0.4809,
+      "step": 209
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.162303864257998,
+      "learning_rate": 3.2967032967032968e-06,
+      "loss": 0.8469,
+      "step": 210
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5831477240840228,
+      "learning_rate": 3.3124018838304557e-06,
+      "loss": 0.8245,
+      "step": 211
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5773906052939746,
+      "learning_rate": 3.328100470957614e-06,
+      "loss": 0.7658,
+      "step": 212
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3179350694576664,
+      "learning_rate": 3.343799058084773e-06,
+      "loss": 0.6879,
+      "step": 213
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9775212684172223,
+      "learning_rate": 3.359497645211931e-06,
+      "loss": 0.643,
+      "step": 214
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.66803088415483,
+      "learning_rate": 3.3751962323390895e-06,
+      "loss": 0.7433,
+      "step": 215
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3804957703151652,
+      "learning_rate": 3.3908948194662484e-06,
+      "loss": 0.8484,
+      "step": 216
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7396494081306202,
+      "learning_rate": 3.406593406593407e-06,
+      "loss": 0.7758,
+      "step": 217
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.397485037488961,
+      "learning_rate": 3.4222919937205657e-06,
+      "loss": 0.7273,
+      "step": 218
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8029957774106913,
+      "learning_rate": 3.4379905808477238e-06,
+      "loss": 0.6783,
+      "step": 219
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3016644823098293,
+      "learning_rate": 3.4536891679748822e-06,
+      "loss": 0.7429,
+      "step": 220
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4182013803868005,
+      "learning_rate": 3.469387755102041e-06,
+      "loss": 0.7635,
+      "step": 221
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3935403135590243,
+      "learning_rate": 3.4850863422291996e-06,
+      "loss": 0.6284,
+      "step": 222
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.58059061578339,
+      "learning_rate": 3.5007849293563585e-06,
+      "loss": 0.5211,
+      "step": 223
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8857446482117959,
+      "learning_rate": 3.516483516483517e-06,
+      "loss": 0.8515,
+      "step": 224
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3778887075356618,
+      "learning_rate": 3.532182103610675e-06,
+      "loss": 0.7308,
+      "step": 225
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5125413183513572,
+      "learning_rate": 3.547880690737834e-06,
+      "loss": 0.7026,
+      "step": 226
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3064885390463534,
+      "learning_rate": 3.5635792778649923e-06,
+      "loss": 0.7507,
+      "step": 227
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3166840865794984,
+      "learning_rate": 3.5792778649921512e-06,
+      "loss": 0.761,
+      "step": 228
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4043777680173295,
+      "learning_rate": 3.5949764521193097e-06,
+      "loss": 0.6983,
+      "step": 229
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3913983460377004,
+      "learning_rate": 3.6106750392464677e-06,
+      "loss": 0.7943,
+      "step": 230
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3672805069363358,
+      "learning_rate": 3.6263736263736266e-06,
+      "loss": 0.7171,
+      "step": 231
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5966348651234366,
+      "learning_rate": 3.642072213500785e-06,
+      "loss": 0.5111,
+      "step": 232
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9459407268464375,
+      "learning_rate": 3.657770800627944e-06,
+      "loss": 0.7452,
+      "step": 233
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6184188033307815,
+      "learning_rate": 3.6734693877551024e-06,
+      "loss": 0.4663,
+      "step": 234
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.333474393543286,
+      "learning_rate": 3.6891679748822605e-06,
+      "loss": 0.8208,
+      "step": 235
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4065841550256437,
+      "learning_rate": 3.7048665620094194e-06,
+      "loss": 0.862,
+      "step": 236
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4642320648531548,
+      "learning_rate": 3.720565149136578e-06,
+      "loss": 0.68,
+      "step": 237
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6302090266101072,
+      "learning_rate": 3.7362637362637367e-06,
+      "loss": 0.8615,
+      "step": 238
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5827213397967247,
+      "learning_rate": 3.751962323390895e-06,
+      "loss": 0.4918,
+      "step": 239
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2244634708214235,
+      "learning_rate": 3.767660910518054e-06,
+      "loss": 0.6084,
+      "step": 240
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5636951193997081,
+      "learning_rate": 3.783359497645212e-06,
+      "loss": 0.5088,
+      "step": 241
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4311084009897879,
+      "learning_rate": 3.7990580847723706e-06,
+      "loss": 0.7411,
+      "step": 242
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4495098430043087,
+      "learning_rate": 3.814756671899529e-06,
+      "loss": 0.6817,
+      "step": 243
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3600301388283287,
+      "learning_rate": 3.830455259026688e-06,
+      "loss": 0.8914,
+      "step": 244
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6258355384981082,
+      "learning_rate": 3.846153846153847e-06,
+      "loss": 0.7425,
+      "step": 245
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7970248604707053,
+      "learning_rate": 3.861852433281005e-06,
+      "loss": 0.7631,
+      "step": 246
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3909408480292424,
+      "learning_rate": 3.877551020408164e-06,
+      "loss": 0.7055,
+      "step": 247
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5408103731524083,
+      "learning_rate": 3.893249607535322e-06,
+      "loss": 0.7419,
+      "step": 248
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.357969727867511,
+      "learning_rate": 3.908948194662481e-06,
+      "loss": 0.8326,
+      "step": 249
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4864031030400162,
+      "learning_rate": 3.9246467817896395e-06,
+      "loss": 0.7871,
+      "step": 250
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3692489367356895,
+      "learning_rate": 3.940345368916798e-06,
+      "loss": 0.7139,
+      "step": 251
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5800878905996112,
+      "learning_rate": 3.9560439560439565e-06,
+      "loss": 0.6793,
+      "step": 252
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3001340983526242,
+      "learning_rate": 3.9717425431711145e-06,
+      "loss": 0.8675,
+      "step": 253
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.482076340470562,
+      "learning_rate": 3.987441130298273e-06,
+      "loss": 0.6584,
+      "step": 254
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2694262061195556,
+      "learning_rate": 4.003139717425432e-06,
+      "loss": 0.6751,
+      "step": 255
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3199155178900221,
+      "learning_rate": 4.018838304552591e-06,
+      "loss": 0.6485,
+      "step": 256
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7939776570728652,
+      "learning_rate": 4.034536891679749e-06,
+      "loss": 0.7254,
+      "step": 257
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.1216884536550635,
+      "learning_rate": 4.050235478806907e-06,
+      "loss": 0.6687,
+      "step": 258
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2477671455329507,
+      "learning_rate": 4.065934065934066e-06,
+      "loss": 0.5892,
+      "step": 259
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4094214819632813,
+      "learning_rate": 4.081632653061225e-06,
+      "loss": 0.7291,
+      "step": 260
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3060168842068707,
+      "learning_rate": 4.097331240188384e-06,
+      "loss": 0.5954,
+      "step": 261
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5893162720346189,
+      "learning_rate": 4.113029827315542e-06,
+      "loss": 0.4973,
+      "step": 262
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3935099990155442,
+      "learning_rate": 4.1287284144427e-06,
+      "loss": 0.7643,
+      "step": 263
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3125567744247333,
+      "learning_rate": 4.144427001569859e-06,
+      "loss": 0.6964,
+      "step": 264
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2702517755089013,
+      "learning_rate": 4.160125588697018e-06,
+      "loss": 0.7505,
+      "step": 265
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5406990801911564,
+      "learning_rate": 4.175824175824177e-06,
+      "loss": 0.6989,
+      "step": 266
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4983995060781221,
+      "learning_rate": 4.191522762951335e-06,
+      "loss": 0.885,
+      "step": 267
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4635801019584218,
+      "learning_rate": 4.207221350078493e-06,
+      "loss": 0.7973,
+      "step": 268
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.394595544398799,
+      "learning_rate": 4.222919937205652e-06,
+      "loss": 0.7257,
+      "step": 269
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6033403702258744,
+      "learning_rate": 4.2386185243328105e-06,
+      "loss": 0.7494,
+      "step": 270
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.581183610234299,
+      "learning_rate": 4.254317111459969e-06,
+      "loss": 0.49,
+      "step": 271
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.463161687730172,
+      "learning_rate": 4.270015698587127e-06,
+      "loss": 0.7412,
+      "step": 272
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5913018733898403,
+      "learning_rate": 4.2857142857142855e-06,
+      "loss": 0.7266,
+      "step": 273
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4607652869113514,
+      "learning_rate": 4.301412872841444e-06,
+      "loss": 0.7648,
+      "step": 274
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.591442838085182,
+      "learning_rate": 4.317111459968603e-06,
+      "loss": 0.7847,
+      "step": 275
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.468063003992475,
+      "learning_rate": 4.332810047095762e-06,
+      "loss": 0.7663,
+      "step": 276
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4627748356496244,
+      "learning_rate": 4.34850863422292e-06,
+      "loss": 0.7303,
+      "step": 277
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.278694369046216,
+      "learning_rate": 4.364207221350079e-06,
+      "loss": 0.8051,
+      "step": 278
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4934657370736628,
+      "learning_rate": 4.379905808477237e-06,
+      "loss": 0.6903,
+      "step": 279
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3360980956290103,
+      "learning_rate": 4.395604395604396e-06,
+      "loss": 0.7855,
+      "step": 280
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5267797420777707,
+      "learning_rate": 4.411302982731555e-06,
+      "loss": 0.7471,
+      "step": 281
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4308405691704238,
+      "learning_rate": 4.427001569858713e-06,
+      "loss": 0.7622,
+      "step": 282
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6311770632971372,
+      "learning_rate": 4.442700156985872e-06,
+      "loss": 0.7529,
+      "step": 283
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6166933436673672,
+      "learning_rate": 4.45839874411303e-06,
+      "loss": 0.6592,
+      "step": 284
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5478425612511326,
+      "learning_rate": 4.474097331240189e-06,
+      "loss": 0.8656,
+      "step": 285
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4853557055950537,
+      "learning_rate": 4.489795918367348e-06,
+      "loss": 0.8406,
+      "step": 286
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1446437885013547,
+      "learning_rate": 4.505494505494506e-06,
+      "loss": 0.7468,
+      "step": 287
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5497165510478337,
+      "learning_rate": 4.5211930926216645e-06,
+      "loss": 0.7155,
+      "step": 288
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8865010784958618,
+      "learning_rate": 4.5368916797488226e-06,
+      "loss": 0.7428,
+      "step": 289
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3327046349128082,
+      "learning_rate": 4.5525902668759815e-06,
+      "loss": 0.6737,
+      "step": 290
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5659597070571598,
+      "learning_rate": 4.56828885400314e-06,
+      "loss": 0.6505,
+      "step": 291
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3705587646467305,
+      "learning_rate": 4.583987441130298e-06,
+      "loss": 0.6012,
+      "step": 292
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5608986862894751,
+      "learning_rate": 4.599686028257457e-06,
+      "loss": 0.8217,
+      "step": 293
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5324626332881557,
+      "learning_rate": 4.615384615384616e-06,
+      "loss": 0.7785,
+      "step": 294
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6144882022667797,
+      "learning_rate": 4.631083202511774e-06,
+      "loss": 0.6793,
+      "step": 295
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5991737381746979,
+      "learning_rate": 4.646781789638933e-06,
+      "loss": 0.4948,
+      "step": 296
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4340732087105679,
+      "learning_rate": 4.662480376766091e-06,
+      "loss": 0.6833,
+      "step": 297
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3793506694462103,
+      "learning_rate": 4.67817896389325e-06,
+      "loss": 0.8578,
+      "step": 298
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3700084214871846,
+      "learning_rate": 4.693877551020409e-06,
+      "loss": 0.7537,
+      "step": 299
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4045054504000132,
+      "learning_rate": 4.709576138147567e-06,
+      "loss": 0.8491,
+      "step": 300
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4212239221805538,
+      "learning_rate": 4.725274725274726e-06,
+      "loss": 0.6379,
+      "step": 301
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3012967494934762,
+      "learning_rate": 4.740973312401884e-06,
+      "loss": 0.7771,
+      "step": 302
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3849506276661407,
+      "learning_rate": 4.756671899529043e-06,
+      "loss": 0.6632,
+      "step": 303
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3758912004945794,
+      "learning_rate": 4.772370486656202e-06,
+      "loss": 0.6901,
+      "step": 304
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3857424077935412,
+      "learning_rate": 4.7880690737833605e-06,
+      "loss": 0.6743,
+      "step": 305
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6917073159653575,
+      "learning_rate": 4.8037676609105186e-06,
+      "loss": 0.7351,
+      "step": 306
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4634184927648721,
+      "learning_rate": 4.819466248037677e-06,
+      "loss": 0.7774,
+      "step": 307
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5268349805499721,
+      "learning_rate": 4.8351648351648355e-06,
+      "loss": 0.6668,
+      "step": 308
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5960134738059415,
+      "learning_rate": 4.850863422291994e-06,
+      "loss": 0.8515,
+      "step": 309
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5133453030417656,
+      "learning_rate": 4.866562009419153e-06,
+      "loss": 0.71,
+      "step": 310
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5118204944770437,
+      "learning_rate": 4.882260596546311e-06,
+      "loss": 0.6612,
+      "step": 311
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2992212928847016,
+      "learning_rate": 4.897959183673469e-06,
+      "loss": 0.7761,
+      "step": 312
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4201340688155375,
+      "learning_rate": 4.913657770800628e-06,
+      "loss": 0.6556,
+      "step": 313
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6074265024517377,
+      "learning_rate": 4.929356357927787e-06,
+      "loss": 0.5162,
+      "step": 314
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7482021665806093,
+      "learning_rate": 4.945054945054946e-06,
+      "loss": 0.6815,
+      "step": 315
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2184514006388623,
+      "learning_rate": 4.960753532182104e-06,
+      "loss": 0.5674,
+      "step": 316
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2841123091065387,
+      "learning_rate": 4.976452119309262e-06,
+      "loss": 0.6079,
+      "step": 317
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3099978585414616,
+      "learning_rate": 4.992150706436421e-06,
+      "loss": 0.6381,
+      "step": 318
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4934276116139582,
+      "learning_rate": 5.007849293563579e-06,
+      "loss": 0.8213,
+      "step": 319
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6443728811687974,
+      "learning_rate": 5.023547880690738e-06,
+      "loss": 0.8281,
+      "step": 320
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6180725846083994,
+      "learning_rate": 5.039246467817897e-06,
+      "loss": 0.783,
+      "step": 321
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5441158594341295,
+      "learning_rate": 5.054945054945055e-06,
+      "loss": 0.7633,
+      "step": 322
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5458681275820665,
+      "learning_rate": 5.070643642072214e-06,
+      "loss": 0.6846,
+      "step": 323
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2233337540081743,
+      "learning_rate": 5.086342229199373e-06,
+      "loss": 0.6679,
+      "step": 324
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5951872676517118,
+      "learning_rate": 5.1020408163265315e-06,
+      "loss": 0.608,
+      "step": 325
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1202765177970626,
+      "learning_rate": 5.1177394034536895e-06,
+      "loss": 0.8925,
+      "step": 326
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3268939867338525,
+      "learning_rate": 5.133437990580848e-06,
+      "loss": 0.7353,
+      "step": 327
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6826312628981057,
+      "learning_rate": 5.149136577708007e-06,
+      "loss": 0.7559,
+      "step": 328
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5898822981184014,
+      "learning_rate": 5.164835164835166e-06,
+      "loss": 0.4785,
+      "step": 329
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6338446238080242,
+      "learning_rate": 5.180533751962323e-06,
+      "loss": 0.8925,
+      "step": 330
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2397360400559987,
+      "learning_rate": 5.196232339089482e-06,
+      "loss": 0.6715,
+      "step": 331
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5105491933645772,
+      "learning_rate": 5.21193092621664e-06,
+      "loss": 0.784,
+      "step": 332
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3055824471140511,
+      "learning_rate": 5.227629513343799e-06,
+      "loss": 0.756,
+      "step": 333
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7587827351295868,
+      "learning_rate": 5.243328100470958e-06,
+      "loss": 0.8117,
+      "step": 334
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5491619437793325,
+      "learning_rate": 5.259026687598117e-06,
+      "loss": 0.5111,
+      "step": 335
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4337375780604769,
+      "learning_rate": 5.274725274725275e-06,
+      "loss": 0.6715,
+      "step": 336
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4681398534731025,
+      "learning_rate": 5.290423861852434e-06,
+      "loss": 0.8116,
+      "step": 337
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6225683133100537,
+      "learning_rate": 5.306122448979593e-06,
+      "loss": 0.5165,
+      "step": 338
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.707814709185016,
+      "learning_rate": 5.321821036106752e-06,
+      "loss": 0.9109,
+      "step": 339
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6574452205797714,
+      "learning_rate": 5.33751962323391e-06,
+      "loss": 0.717,
+      "step": 340
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.625122248451332,
+      "learning_rate": 5.353218210361068e-06,
+      "loss": 0.4986,
+      "step": 341
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6243474243612691,
+      "learning_rate": 5.368916797488226e-06,
+      "loss": 0.4796,
+      "step": 342
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5924629606825436,
+      "learning_rate": 5.384615384615385e-06,
+      "loss": 0.5185,
+      "step": 343
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.507778519391294,
+      "learning_rate": 5.4003139717425436e-06,
+      "loss": 0.7558,
+      "step": 344
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3979605803213695,
+      "learning_rate": 5.4160125588697024e-06,
+      "loss": 0.6631,
+      "step": 345
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6261270910631063,
+      "learning_rate": 5.4317111459968605e-06,
+      "loss": 0.7465,
+      "step": 346
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6404637163441387,
+      "learning_rate": 5.447409733124019e-06,
+      "loss": 0.4904,
+      "step": 347
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.586399650140207,
+      "learning_rate": 5.463108320251178e-06,
+      "loss": 0.6049,
+      "step": 348
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5251626180275886,
+      "learning_rate": 5.478806907378337e-06,
+      "loss": 0.8425,
+      "step": 349
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.79535009340198,
+      "learning_rate": 5.494505494505495e-06,
+      "loss": 0.7457,
+      "step": 350
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.323064747693181,
+      "learning_rate": 5.510204081632653e-06,
+      "loss": 0.6882,
+      "step": 351
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6589810996366274,
+      "learning_rate": 5.525902668759811e-06,
+      "loss": 0.7352,
+      "step": 352
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5460049000233458,
+      "learning_rate": 5.54160125588697e-06,
+      "loss": 0.5759,
+      "step": 353
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4803617756234206,
+      "learning_rate": 5.557299843014129e-06,
+      "loss": 0.7125,
+      "step": 354
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6962879022936486,
+      "learning_rate": 5.572998430141288e-06,
+      "loss": 0.775,
+      "step": 355
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4876460623583259,
+      "learning_rate": 5.588697017268446e-06,
+      "loss": 0.7079,
+      "step": 356
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4602293098579255,
+      "learning_rate": 5.604395604395605e-06,
+      "loss": 0.7273,
+      "step": 357
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8304618426144326,
+      "learning_rate": 5.620094191522764e-06,
+      "loss": 0.6795,
+      "step": 358
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.436765814677204,
+      "learning_rate": 5.635792778649923e-06,
+      "loss": 0.6536,
+      "step": 359
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.431406691484073,
+      "learning_rate": 5.651491365777081e-06,
+      "loss": 0.6163,
+      "step": 360
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5160972309107603,
+      "learning_rate": 5.6671899529042395e-06,
+      "loss": 0.8334,
+      "step": 361
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.346411440768391,
+      "learning_rate": 5.682888540031397e-06,
+      "loss": 0.8033,
+      "step": 362
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.293527208516528,
+      "learning_rate": 5.698587127158556e-06,
+      "loss": 0.7031,
+      "step": 363
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6188551040364607,
+      "learning_rate": 5.7142857142857145e-06,
+      "loss": 0.5148,
+      "step": 364
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4102466473188031,
+      "learning_rate": 5.729984301412873e-06,
+      "loss": 0.7193,
+      "step": 365
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5192644332158642,
+      "learning_rate": 5.7456828885400314e-06,
+      "loss": 0.7021,
+      "step": 366
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2738278569814994,
+      "learning_rate": 5.76138147566719e-06,
+      "loss": 0.6661,
+      "step": 367
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4564016874205514,
+      "learning_rate": 5.777080062794349e-06,
+      "loss": 0.6406,
+      "step": 368
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6272705207453599,
+      "learning_rate": 5.792778649921508e-06,
+      "loss": 0.5269,
+      "step": 369
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4182233224563257,
+      "learning_rate": 5.808477237048666e-06,
+      "loss": 0.744,
+      "step": 370
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5227056887929553,
+      "learning_rate": 5.824175824175825e-06,
+      "loss": 0.7969,
+      "step": 371
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5548381168988135,
+      "learning_rate": 5.839874411302984e-06,
+      "loss": 0.6915,
+      "step": 372
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3410115156358502,
+      "learning_rate": 5.855572998430141e-06,
+      "loss": 0.7,
+      "step": 373
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.570684600802234,
+      "learning_rate": 5.8712715855573e-06,
+      "loss": 0.4709,
+      "step": 374
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2759997729506802,
+      "learning_rate": 5.886970172684459e-06,
+      "loss": 0.6052,
+      "step": 375
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.420191979875886,
+      "learning_rate": 5.902668759811617e-06,
+      "loss": 0.7741,
+      "step": 376
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5751323529005518,
+      "learning_rate": 5.918367346938776e-06,
+      "loss": 0.5018,
+      "step": 377
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4910314837671068,
+      "learning_rate": 5.934065934065935e-06,
+      "loss": 0.6553,
+      "step": 378
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4385664460150596,
+      "learning_rate": 5.949764521193094e-06,
+      "loss": 0.7672,
+      "step": 379
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4003528291493716,
+      "learning_rate": 5.965463108320252e-06,
+      "loss": 0.7671,
+      "step": 380
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4032280313238512,
+      "learning_rate": 5.9811616954474105e-06,
+      "loss": 0.6803,
+      "step": 381
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5247492283063429,
+      "learning_rate": 5.996860282574569e-06,
+      "loss": 0.5731,
+      "step": 382
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4553803160415442,
+      "learning_rate": 6.012558869701728e-06,
+      "loss": 0.7711,
+      "step": 383
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.494309605296534,
+      "learning_rate": 6.0282574568288855e-06,
+      "loss": 0.7079,
+      "step": 384
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3390123833672822,
+      "learning_rate": 6.043956043956044e-06,
+      "loss": 0.7015,
+      "step": 385
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3222731668699705,
+      "learning_rate": 6.059654631083202e-06,
+      "loss": 0.5293,
+      "step": 386
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3633817075898587,
+      "learning_rate": 6.075353218210361e-06,
+      "loss": 0.8264,
+      "step": 387
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2203334166832602,
+      "learning_rate": 6.09105180533752e-06,
+      "loss": 0.7594,
+      "step": 388
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.446656761696967,
+      "learning_rate": 6.106750392464679e-06,
+      "loss": 0.6238,
+      "step": 389
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4089091136764185,
+      "learning_rate": 6.122448979591837e-06,
+      "loss": 0.7032,
+      "step": 390
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6018441639280597,
+      "learning_rate": 6.138147566718996e-06,
+      "loss": 0.5794,
+      "step": 391
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6385865036684752,
+      "learning_rate": 6.153846153846155e-06,
+      "loss": 0.6794,
+      "step": 392
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.427837685568755,
+      "learning_rate": 6.169544740973314e-06,
+      "loss": 0.7437,
+      "step": 393
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4047233647840254,
+      "learning_rate": 6.185243328100472e-06,
+      "loss": 0.689,
+      "step": 394
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.444547500983624,
+      "learning_rate": 6.20094191522763e-06,
+      "loss": 0.73,
+      "step": 395
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3921001979017327,
+      "learning_rate": 6.216640502354788e-06,
+      "loss": 0.6097,
+      "step": 396
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5609739171398098,
+      "learning_rate": 6.232339089481947e-06,
+      "loss": 0.5799,
+      "step": 397
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.8337964737083827,
+      "learning_rate": 6.248037676609106e-06,
+      "loss": 0.8105,
+      "step": 398
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7194518783574344,
+      "learning_rate": 6.2637362637362645e-06,
+      "loss": 0.7115,
+      "step": 399
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7765676692544439,
+      "learning_rate": 6.279434850863423e-06,
+      "loss": 0.6143,
+      "step": 400
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5202566035536575,
+      "learning_rate": 6.2951334379905815e-06,
+      "loss": 0.7034,
+      "step": 401
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5344718256385852,
+      "learning_rate": 6.31083202511774e-06,
+      "loss": 0.7654,
+      "step": 402
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5340752937544244,
+      "learning_rate": 6.326530612244899e-06,
+      "loss": 0.6961,
+      "step": 403
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7311045183835425,
+      "learning_rate": 6.342229199372057e-06,
+      "loss": 0.7623,
+      "step": 404
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.671570808867082,
+      "learning_rate": 6.357927786499215e-06,
+      "loss": 0.6954,
+      "step": 405
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5028202746024577,
+      "learning_rate": 6.373626373626373e-06,
+      "loss": 0.8119,
+      "step": 406
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3065160963430429,
+      "learning_rate": 6.389324960753532e-06,
+      "loss": 0.8006,
+      "step": 407
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.563702870327949,
+      "learning_rate": 6.405023547880691e-06,
+      "loss": 0.7165,
+      "step": 408
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4107108869069787,
+      "learning_rate": 6.42072213500785e-06,
+      "loss": 0.761,
+      "step": 409
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6445554106396105,
+      "learning_rate": 6.436420722135008e-06,
+      "loss": 0.8272,
+      "step": 410
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3266457101619271,
+      "learning_rate": 6.452119309262167e-06,
+      "loss": 0.5821,
+      "step": 411
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.460810584325398,
+      "learning_rate": 6.467817896389326e-06,
+      "loss": 0.7432,
+      "step": 412
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4019170947366153,
+      "learning_rate": 6.483516483516485e-06,
+      "loss": 0.7907,
+      "step": 413
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5297461601323117,
+      "learning_rate": 6.499215070643643e-06,
+      "loss": 0.7861,
+      "step": 414
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5879478487755313,
+      "learning_rate": 6.514913657770802e-06,
+      "loss": 0.5201,
+      "step": 415
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.385664883727152,
+      "learning_rate": 6.530612244897959e-06,
+      "loss": 0.7214,
+      "step": 416
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.404972726951191,
+      "learning_rate": 6.546310832025118e-06,
+      "loss": 0.877,
+      "step": 417
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7794939792028788,
+      "learning_rate": 6.562009419152277e-06,
+      "loss": 0.7873,
+      "step": 418
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3802890261300755,
+      "learning_rate": 6.5777080062794355e-06,
+      "loss": 0.7613,
+      "step": 419
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3950768498377344,
+      "learning_rate": 6.5934065934065935e-06,
+      "loss": 0.8399,
+      "step": 420
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6382257237224086,
+      "learning_rate": 6.609105180533752e-06,
+      "loss": 0.4902,
+      "step": 421
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0539940904150185,
+      "learning_rate": 6.624803767660911e-06,
+      "loss": 0.7918,
+      "step": 422
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.704710895759628,
+      "learning_rate": 6.64050235478807e-06,
+      "loss": 0.587,
+      "step": 423
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4912213134439773,
+      "learning_rate": 6.656200941915228e-06,
+      "loss": 0.6977,
+      "step": 424
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3698947959071486,
+      "learning_rate": 6.671899529042387e-06,
+      "loss": 0.6825,
+      "step": 425
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.499212935471376,
+      "learning_rate": 6.687598116169546e-06,
+      "loss": 0.6756,
+      "step": 426
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4382578882626782,
+      "learning_rate": 6.703296703296703e-06,
+      "loss": 0.6891,
+      "step": 427
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3937719765533267,
+      "learning_rate": 6.718995290423862e-06,
+      "loss": 0.7893,
+      "step": 428
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3841825263535357,
+      "learning_rate": 6.734693877551021e-06,
+      "loss": 0.6514,
+      "step": 429
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6345994958857413,
+      "learning_rate": 6.750392464678179e-06,
+      "loss": 0.6923,
+      "step": 430
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3239352756933285,
+      "learning_rate": 6.766091051805338e-06,
+      "loss": 0.7226,
+      "step": 431
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0033610459413826,
+      "learning_rate": 6.781789638932497e-06,
+      "loss": 0.6738,
+      "step": 432
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7428864634854861,
+      "learning_rate": 6.797488226059656e-06,
+      "loss": 0.7156,
+      "step": 433
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.538713641675999,
+      "learning_rate": 6.813186813186814e-06,
+      "loss": 0.6153,
+      "step": 434
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5588289249030896,
+      "learning_rate": 6.828885400313973e-06,
+      "loss": 0.6671,
+      "step": 435
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5443430066588016,
+      "learning_rate": 6.8445839874411315e-06,
+      "loss": 0.4591,
+      "step": 436
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6069963425688337,
+      "learning_rate": 6.8602825745682895e-06,
+      "loss": 0.4737,
+      "step": 437
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7688101072787645,
+      "learning_rate": 6.8759811616954476e-06,
+      "loss": 0.8503,
+      "step": 438
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.706121735744233,
+      "learning_rate": 6.8916797488226065e-06,
+      "loss": 0.6713,
+      "step": 439
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5649814175647554,
+      "learning_rate": 6.9073783359497645e-06,
+      "loss": 0.8384,
+      "step": 440
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.639013666362064,
+      "learning_rate": 6.923076923076923e-06,
+      "loss": 0.6985,
+      "step": 441
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5113437541374828,
+      "learning_rate": 6.938775510204082e-06,
+      "loss": 0.9012,
+      "step": 442
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7948649822269938,
+      "learning_rate": 6.954474097331241e-06,
+      "loss": 0.7154,
+      "step": 443
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9490446145383427,
+      "learning_rate": 6.970172684458399e-06,
+      "loss": 0.6477,
+      "step": 444
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.546045594759205,
+      "learning_rate": 6.985871271585558e-06,
+      "loss": 0.812,
+      "step": 445
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3124032481391819,
+      "learning_rate": 7.001569858712717e-06,
+      "loss": 0.6307,
+      "step": 446
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3966632010659359,
+      "learning_rate": 7.017268445839875e-06,
+      "loss": 0.7587,
+      "step": 447
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.209203816715923,
+      "learning_rate": 7.032967032967034e-06,
+      "loss": 0.783,
+      "step": 448
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1463480193697007,
+      "learning_rate": 7.048665620094192e-06,
+      "loss": 0.6489,
+      "step": 449
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.682868793268095,
+      "learning_rate": 7.06436420722135e-06,
+      "loss": 0.8387,
+      "step": 450
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4261213648026474,
+      "learning_rate": 7.080062794348509e-06,
+      "loss": 0.6999,
+      "step": 451
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8335790922202417,
+      "learning_rate": 7.095761381475668e-06,
+      "loss": 0.828,
+      "step": 452
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.364364086331788,
+      "learning_rate": 7.111459968602827e-06,
+      "loss": 0.8167,
+      "step": 453
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2838025804448812,
+      "learning_rate": 7.127158555729985e-06,
+      "loss": 0.8218,
+      "step": 454
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3909557719756545,
+      "learning_rate": 7.1428571428571436e-06,
+      "loss": 0.7627,
+      "step": 455
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5270263174114695,
+      "learning_rate": 7.1585557299843024e-06,
+      "loss": 0.7867,
+      "step": 456
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7660891370675948,
+      "learning_rate": 7.174254317111461e-06,
+      "loss": 0.7138,
+      "step": 457
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.625068483661218,
+      "learning_rate": 7.189952904238619e-06,
+      "loss": 0.4694,
+      "step": 458
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3579498671942682,
+      "learning_rate": 7.205651491365777e-06,
+      "loss": 0.6845,
+      "step": 459
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5392943533596821,
+      "learning_rate": 7.2213500784929355e-06,
+      "loss": 0.8039,
+      "step": 460
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6094585505993532,
+      "learning_rate": 7.237048665620094e-06,
+      "loss": 0.5142,
+      "step": 461
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6374990100756082,
+      "learning_rate": 7.252747252747253e-06,
+      "loss": 0.7224,
+      "step": 462
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3026441276045968,
+      "learning_rate": 7.268445839874412e-06,
+      "loss": 0.725,
+      "step": 463
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4492808976152063,
+      "learning_rate": 7.28414442700157e-06,
+      "loss": 0.664,
+      "step": 464
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4698478462745255,
+      "learning_rate": 7.299843014128729e-06,
+      "loss": 0.7818,
+      "step": 465
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7769573282437576,
+      "learning_rate": 7.315541601255888e-06,
+      "loss": 0.5004,
+      "step": 466
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3981246925136057,
+      "learning_rate": 7.331240188383047e-06,
+      "loss": 0.712,
+      "step": 467
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3786774395790709,
+      "learning_rate": 7.346938775510205e-06,
+      "loss": 0.7031,
+      "step": 468
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5874975547219647,
+      "learning_rate": 7.362637362637364e-06,
+      "loss": 0.7208,
+      "step": 469
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.434754075138524,
+      "learning_rate": 7.378335949764521e-06,
+      "loss": 0.6879,
+      "step": 470
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4736278210567901,
+      "learning_rate": 7.39403453689168e-06,
+      "loss": 0.7623,
+      "step": 471
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.744156040108044,
+      "learning_rate": 7.409733124018839e-06,
+      "loss": 0.7872,
+      "step": 472
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3833356771202272,
+      "learning_rate": 7.425431711145998e-06,
+      "loss": 0.6674,
+      "step": 473
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5117229391649578,
+      "learning_rate": 7.441130298273156e-06,
+      "loss": 0.7881,
+      "step": 474
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4073256133921137,
+      "learning_rate": 7.4568288854003145e-06,
+      "loss": 0.5933,
+      "step": 475
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5887341066619534,
+      "learning_rate": 7.472527472527473e-06,
+      "loss": 0.5067,
+      "step": 476
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2982672933994366,
+      "learning_rate": 7.488226059654632e-06,
+      "loss": 0.7663,
+      "step": 477
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.588546440455139,
+      "learning_rate": 7.50392464678179e-06,
+      "loss": 0.8652,
+      "step": 478
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.761384652392467,
+      "learning_rate": 7.519623233908949e-06,
+      "loss": 0.6453,
+      "step": 479
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5675888215342624,
+      "learning_rate": 7.535321821036108e-06,
+      "loss": 0.7446,
+      "step": 480
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2921316816671502,
+      "learning_rate": 7.551020408163265e-06,
+      "loss": 0.7551,
+      "step": 481
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4771978455030348,
+      "learning_rate": 7.566718995290424e-06,
+      "loss": 0.5925,
+      "step": 482
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.548668946835046,
+      "learning_rate": 7.582417582417583e-06,
+      "loss": 0.7923,
+      "step": 483
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4998289014632722,
+      "learning_rate": 7.598116169544741e-06,
+      "loss": 0.6832,
+      "step": 484
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5872007273182067,
+      "learning_rate": 7.6138147566719e-06,
+      "loss": 0.4853,
+      "step": 485
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5018089926386806,
+      "learning_rate": 7.629513343799058e-06,
+      "loss": 0.9259,
+      "step": 486
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.467346451682239,
+      "learning_rate": 7.645211930926217e-06,
+      "loss": 0.6623,
+      "step": 487
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.493825241907333,
+      "learning_rate": 7.660910518053376e-06,
+      "loss": 0.7566,
+      "step": 488
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5968878413765075,
+      "learning_rate": 7.676609105180535e-06,
+      "loss": 0.7434,
+      "step": 489
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5982605501159104,
+      "learning_rate": 7.692307692307694e-06,
+      "loss": 0.4806,
+      "step": 490
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5904279564877255,
+      "learning_rate": 7.708006279434852e-06,
+      "loss": 0.7882,
+      "step": 491
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1295558725055987,
+      "learning_rate": 7.72370486656201e-06,
+      "loss": 0.7688,
+      "step": 492
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4830898714986405,
+      "learning_rate": 7.739403453689169e-06,
+      "loss": 0.779,
+      "step": 493
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.28038109913459,
+      "learning_rate": 7.755102040816327e-06,
+      "loss": 0.7455,
+      "step": 494
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5763589709931619,
+      "learning_rate": 7.770800627943486e-06,
+      "loss": 0.4606,
+      "step": 495
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4337187172424812,
+      "learning_rate": 7.786499215070644e-06,
+      "loss": 0.7541,
+      "step": 496
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4100100506787097,
+      "learning_rate": 7.802197802197802e-06,
+      "loss": 0.5217,
+      "step": 497
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4049571658012456,
+      "learning_rate": 7.817896389324961e-06,
+      "loss": 0.808,
+      "step": 498
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.546298675214665,
+      "learning_rate": 7.83359497645212e-06,
+      "loss": 0.7418,
+      "step": 499
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.37269795433611,
+      "learning_rate": 7.849293563579279e-06,
+      "loss": 0.7992,
+      "step": 500
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8682720956458854,
+      "learning_rate": 7.864992150706438e-06,
+      "loss": 0.8746,
+      "step": 501
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.364370589426429,
+      "learning_rate": 7.880690737833597e-06,
+      "loss": 0.6395,
+      "step": 502
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.405124115147336,
+      "learning_rate": 7.896389324960754e-06,
+      "loss": 0.8137,
+      "step": 503
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3899565299248122,
+      "learning_rate": 7.912087912087913e-06,
+      "loss": 0.7515,
+      "step": 504
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5986177032975952,
+      "learning_rate": 7.927786499215072e-06,
+      "loss": 0.4765,
+      "step": 505
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.612182394131532,
+      "learning_rate": 7.943485086342229e-06,
+      "loss": 0.6966,
+      "step": 506
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7179169152973948,
+      "learning_rate": 7.959183673469388e-06,
+      "loss": 0.7335,
+      "step": 507
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3446269788541416,
+      "learning_rate": 7.974882260596547e-06,
+      "loss": 0.7677,
+      "step": 508
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3800523369423587,
+      "learning_rate": 7.990580847723706e-06,
+      "loss": 0.6946,
+      "step": 509
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4644406762682114,
+      "learning_rate": 8.006279434850865e-06,
+      "loss": 0.7873,
+      "step": 510
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4861338598176306,
+      "learning_rate": 8.021978021978023e-06,
+      "loss": 0.7556,
+      "step": 511
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6692196663005122,
+      "learning_rate": 8.037676609105182e-06,
+      "loss": 0.4972,
+      "step": 512
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3516054811777023,
+      "learning_rate": 8.05337519623234e-06,
+      "loss": 0.7276,
+      "step": 513
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4009905899225714,
+      "learning_rate": 8.069073783359498e-06,
+      "loss": 0.8371,
+      "step": 514
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6120129709855258,
+      "learning_rate": 8.084772370486657e-06,
+      "loss": 0.5329,
+      "step": 515
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9493446917999768,
+      "learning_rate": 8.100470957613814e-06,
+      "loss": 0.683,
+      "step": 516
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5930539400876236,
+      "learning_rate": 8.116169544740973e-06,
+      "loss": 0.9308,
+      "step": 517
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6699395992446184,
+      "learning_rate": 8.131868131868132e-06,
+      "loss": 0.843,
+      "step": 518
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4191782159695099,
+      "learning_rate": 8.147566718995291e-06,
+      "loss": 0.616,
+      "step": 519
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5813533868664047,
+      "learning_rate": 8.16326530612245e-06,
+      "loss": 0.6618,
+      "step": 520
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5880243221389252,
+      "learning_rate": 8.178963893249609e-06,
+      "loss": 0.774,
+      "step": 521
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.531607275259759,
+      "learning_rate": 8.194662480376768e-06,
+      "loss": 0.6649,
+      "step": 522
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.387144840014408,
+      "learning_rate": 8.210361067503925e-06,
+      "loss": 0.6636,
+      "step": 523
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6101254765440656,
+      "learning_rate": 8.226059654631084e-06,
+      "loss": 0.4771,
+      "step": 524
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5836564686573629,
+      "learning_rate": 8.241758241758243e-06,
+      "loss": 0.4761,
+      "step": 525
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5773786653133859,
+      "learning_rate": 8.2574568288854e-06,
+      "loss": 0.4646,
+      "step": 526
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.478882121183704,
+      "learning_rate": 8.273155416012559e-06,
+      "loss": 0.7832,
+      "step": 527
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5950762778317649,
+      "learning_rate": 8.288854003139718e-06,
+      "loss": 0.627,
+      "step": 528
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.390104825931837,
+      "learning_rate": 8.304552590266877e-06,
+      "loss": 0.6842,
+      "step": 529
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4335376752044084,
+      "learning_rate": 8.320251177394036e-06,
+      "loss": 0.6143,
+      "step": 530
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6862250319487997,
+      "learning_rate": 8.335949764521194e-06,
+      "loss": 0.7605,
+      "step": 531
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4607772034590654,
+      "learning_rate": 8.351648351648353e-06,
+      "loss": 0.8464,
+      "step": 532
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7198589143608858,
+      "learning_rate": 8.36734693877551e-06,
+      "loss": 0.874,
+      "step": 533
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.631892187192339,
+      "learning_rate": 8.38304552590267e-06,
+      "loss": 0.8342,
+      "step": 534
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3223470971640876,
+      "learning_rate": 8.398744113029828e-06,
+      "loss": 0.7095,
+      "step": 535
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7944404699151646,
+      "learning_rate": 8.414442700156985e-06,
+      "loss": 0.7373,
+      "step": 536
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6192004752525948,
+      "learning_rate": 8.430141287284144e-06,
+      "loss": 0.8171,
+      "step": 537
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5778545107571622,
+      "learning_rate": 8.445839874411303e-06,
+      "loss": 0.782,
+      "step": 538
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4475378891132902,
+      "learning_rate": 8.461538461538462e-06,
+      "loss": 0.7744,
+      "step": 539
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4310896175667542,
+      "learning_rate": 8.477237048665621e-06,
+      "loss": 0.6879,
+      "step": 540
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8234917353970488,
+      "learning_rate": 8.49293563579278e-06,
+      "loss": 0.7632,
+      "step": 541
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.503436905001224,
+      "learning_rate": 8.508634222919939e-06,
+      "loss": 0.768,
+      "step": 542
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.771223293824161,
+      "learning_rate": 8.524332810047096e-06,
+      "loss": 0.6855,
+      "step": 543
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4084955976948748,
+      "learning_rate": 8.540031397174255e-06,
+      "loss": 0.8338,
+      "step": 544
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9004568644690436,
+      "learning_rate": 8.555729984301414e-06,
+      "loss": 0.7825,
+      "step": 545
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6858716866436014,
+      "learning_rate": 8.571428571428571e-06,
+      "loss": 0.7878,
+      "step": 546
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.39796453788212,
+      "learning_rate": 8.58712715855573e-06,
+      "loss": 0.6819,
+      "step": 547
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3198034873508024,
+      "learning_rate": 8.602825745682889e-06,
+      "loss": 0.8546,
+      "step": 548
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2341109748871744,
+      "learning_rate": 8.618524332810048e-06,
+      "loss": 0.7665,
+      "step": 549
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4710606453063302,
+      "learning_rate": 8.634222919937206e-06,
+      "loss": 0.7605,
+      "step": 550
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7635886031341594,
+      "learning_rate": 8.649921507064365e-06,
+      "loss": 0.697,
+      "step": 551
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5925636565215178,
+      "learning_rate": 8.665620094191524e-06,
+      "loss": 0.7008,
+      "step": 552
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6053565367298845,
+      "learning_rate": 8.681318681318681e-06,
+      "loss": 0.5081,
+      "step": 553
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.363218245527281,
+      "learning_rate": 8.69701726844584e-06,
+      "loss": 0.7796,
+      "step": 554
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.583300881571005,
+      "learning_rate": 8.712715855573e-06,
+      "loss": 0.5195,
+      "step": 555
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5152207129084543,
+      "learning_rate": 8.728414442700158e-06,
+      "loss": 0.7517,
+      "step": 556
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9496704637187383,
+      "learning_rate": 8.744113029827315e-06,
+      "loss": 0.812,
+      "step": 557
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3518518164877886,
+      "learning_rate": 8.759811616954474e-06,
+      "loss": 0.8147,
+      "step": 558
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5906932039660896,
+      "learning_rate": 8.775510204081633e-06,
+      "loss": 0.4905,
+      "step": 559
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3754041846081124,
+      "learning_rate": 8.791208791208792e-06,
+      "loss": 0.5685,
+      "step": 560
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4161086920663595,
+      "learning_rate": 8.80690737833595e-06,
+      "loss": 0.5456,
+      "step": 561
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4782869757920254,
+      "learning_rate": 8.82260596546311e-06,
+      "loss": 0.7386,
+      "step": 562
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.36851617625765,
+      "learning_rate": 8.838304552590269e-06,
+      "loss": 0.7447,
+      "step": 563
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4748950966539243,
+      "learning_rate": 8.854003139717426e-06,
+      "loss": 0.7617,
+      "step": 564
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4319035462292558,
+      "learning_rate": 8.869701726844585e-06,
+      "loss": 0.7014,
+      "step": 565
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5695730166553448,
+      "learning_rate": 8.885400313971744e-06,
+      "loss": 0.4962,
+      "step": 566
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5007869840532235,
+      "learning_rate": 8.9010989010989e-06,
+      "loss": 0.6203,
+      "step": 567
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.561847381981409,
+      "learning_rate": 8.91679748822606e-06,
+      "loss": 0.7201,
+      "step": 568
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.860370659323419,
+      "learning_rate": 8.932496075353219e-06,
+      "loss": 0.719,
+      "step": 569
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.412416881359316,
+      "learning_rate": 8.948194662480377e-06,
+      "loss": 0.6936,
+      "step": 570
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6181859828023304,
+      "learning_rate": 8.963893249607536e-06,
+      "loss": 0.5284,
+      "step": 571
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9437381931979651,
+      "learning_rate": 8.979591836734695e-06,
+      "loss": 0.7862,
+      "step": 572
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9662926397627805,
+      "learning_rate": 8.995290423861854e-06,
+      "loss": 0.6986,
+      "step": 573
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0051178702869574,
+      "learning_rate": 9.010989010989011e-06,
+      "loss": 0.764,
+      "step": 574
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.487160289803928,
+      "learning_rate": 9.02668759811617e-06,
+      "loss": 0.7682,
+      "step": 575
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6228293028180725,
+      "learning_rate": 9.042386185243329e-06,
+      "loss": 0.5614,
+      "step": 576
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6298765827224033,
+      "learning_rate": 9.058084772370488e-06,
+      "loss": 0.7944,
+      "step": 577
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7819189580243247,
+      "learning_rate": 9.073783359497645e-06,
+      "loss": 0.8021,
+      "step": 578
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4565372331054272,
+      "learning_rate": 9.089481946624804e-06,
+      "loss": 0.7629,
+      "step": 579
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6297141741146104,
+      "learning_rate": 9.105180533751963e-06,
+      "loss": 0.5298,
+      "step": 580
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3853750866468535,
+      "learning_rate": 9.120879120879122e-06,
+      "loss": 0.7703,
+      "step": 581
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.004912614920245,
+      "learning_rate": 9.13657770800628e-06,
+      "loss": 0.7153,
+      "step": 582
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4185868062510212,
+      "learning_rate": 9.15227629513344e-06,
+      "loss": 0.7777,
+      "step": 583
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3423848514417038,
+      "learning_rate": 9.167974882260597e-06,
+      "loss": 0.6873,
+      "step": 584
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4435422068886716,
+      "learning_rate": 9.183673469387756e-06,
+      "loss": 0.6581,
+      "step": 585
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4299344093837512,
+      "learning_rate": 9.199372056514915e-06,
+      "loss": 0.7318,
+      "step": 586
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.534193885528169,
+      "learning_rate": 9.215070643642073e-06,
+      "loss": 0.7172,
+      "step": 587
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5158282501544085,
+      "learning_rate": 9.230769230769232e-06,
+      "loss": 0.7583,
+      "step": 588
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7302586402815683,
+      "learning_rate": 9.24646781789639e-06,
+      "loss": 0.7625,
+      "step": 589
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4802308906570232,
+      "learning_rate": 9.262166405023548e-06,
+      "loss": 0.6878,
+      "step": 590
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1985124155216242,
+      "learning_rate": 9.277864992150707e-06,
+      "loss": 0.6,
+      "step": 591
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6058913741411875,
+      "learning_rate": 9.293563579277866e-06,
+      "loss": 0.9002,
+      "step": 592
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.357080729640305,
+      "learning_rate": 9.309262166405025e-06,
+      "loss": 0.7194,
+      "step": 593
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2664088470449746,
+      "learning_rate": 9.324960753532182e-06,
+      "loss": 0.6197,
+      "step": 594
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.669829612181856,
+      "learning_rate": 9.340659340659341e-06,
+      "loss": 0.8881,
+      "step": 595
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3505369190737055,
+      "learning_rate": 9.3563579277865e-06,
+      "loss": 0.6539,
+      "step": 596
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5737994437103422,
+      "learning_rate": 9.372056514913659e-06,
+      "loss": 0.7053,
+      "step": 597
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.571872662027918,
+      "learning_rate": 9.387755102040818e-06,
+      "loss": 0.5462,
+      "step": 598
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5459043477154975,
+      "learning_rate": 9.403453689167977e-06,
+      "loss": 0.6878,
+      "step": 599
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5460735701616346,
+      "learning_rate": 9.419152276295134e-06,
+      "loss": 0.7057,
+      "step": 600
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.588473862259256,
+      "learning_rate": 9.434850863422293e-06,
+      "loss": 0.7297,
+      "step": 601
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4068978853659566,
+      "learning_rate": 9.450549450549452e-06,
+      "loss": 0.7025,
+      "step": 602
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1308497475783894,
+      "learning_rate": 9.46624803767661e-06,
+      "loss": 0.56,
+      "step": 603
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2582388744055157,
+      "learning_rate": 9.481946624803768e-06,
+      "loss": 0.6599,
+      "step": 604
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5006873797134157,
+      "learning_rate": 9.497645211930927e-06,
+      "loss": 0.6946,
+      "step": 605
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.40148231116752,
+      "learning_rate": 9.513343799058085e-06,
+      "loss": 0.6951,
+      "step": 606
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3188322751717942,
+      "learning_rate": 9.529042386185244e-06,
+      "loss": 0.7682,
+      "step": 607
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5654675767622632,
+      "learning_rate": 9.544740973312403e-06,
+      "loss": 0.6762,
+      "step": 608
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.245958830446841,
+      "learning_rate": 9.560439560439562e-06,
+      "loss": 0.7566,
+      "step": 609
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4662339678678624,
+      "learning_rate": 9.576138147566721e-06,
+      "loss": 0.8209,
+      "step": 610
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1612318673924267,
+      "learning_rate": 9.591836734693878e-06,
+      "loss": 0.68,
+      "step": 611
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3779562744212304,
+      "learning_rate": 9.607535321821037e-06,
+      "loss": 0.685,
+      "step": 612
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5038385469273416,
+      "learning_rate": 9.623233908948196e-06,
+      "loss": 0.7725,
+      "step": 613
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.839701688843084,
+      "learning_rate": 9.638932496075353e-06,
+      "loss": 0.6465,
+      "step": 614
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5887600781073012,
+      "learning_rate": 9.654631083202512e-06,
+      "loss": 0.7417,
+      "step": 615
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6338792993203965,
+      "learning_rate": 9.670329670329671e-06,
+      "loss": 0.9778,
+      "step": 616
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.500819500968634,
+      "learning_rate": 9.68602825745683e-06,
+      "loss": 0.6472,
+      "step": 617
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3582056028834208,
+      "learning_rate": 9.701726844583989e-06,
+      "loss": 0.8422,
+      "step": 618
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7477709594236868,
+      "learning_rate": 9.717425431711148e-06,
+      "loss": 0.6713,
+      "step": 619
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3930343907398957,
+      "learning_rate": 9.733124018838307e-06,
+      "loss": 0.7484,
+      "step": 620
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.59129933978578,
+      "learning_rate": 9.748822605965464e-06,
+      "loss": 0.5061,
+      "step": 621
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.092383522612859,
+      "learning_rate": 9.764521193092623e-06,
+      "loss": 0.7348,
+      "step": 622
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4738652550474947,
+      "learning_rate": 9.780219780219781e-06,
+      "loss": 0.7504,
+      "step": 623
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.487985723435953,
+      "learning_rate": 9.795918367346939e-06,
+      "loss": 0.7318,
+      "step": 624
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4473545002772297,
+      "learning_rate": 9.811616954474098e-06,
+      "loss": 0.7452,
+      "step": 625
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6440972723580376,
+      "learning_rate": 9.827315541601256e-06,
+      "loss": 0.7977,
+      "step": 626
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5527466886147594,
+      "learning_rate": 9.843014128728415e-06,
+      "loss": 0.6154,
+      "step": 627
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.480404036105871,
+      "learning_rate": 9.858712715855574e-06,
+      "loss": 0.627,
+      "step": 628
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8665496241341555,
+      "learning_rate": 9.874411302982733e-06,
+      "loss": 0.6936,
+      "step": 629
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5519185496121868,
+      "learning_rate": 9.890109890109892e-06,
+      "loss": 0.6416,
+      "step": 630
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5206034305511353,
+      "learning_rate": 9.90580847723705e-06,
+      "loss": 0.7583,
+      "step": 631
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4617323112449663,
+      "learning_rate": 9.921507064364208e-06,
+      "loss": 0.6849,
+      "step": 632
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.068853885158118,
+      "learning_rate": 9.937205651491367e-06,
+      "loss": 0.6801,
+      "step": 633
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4073890331490504,
+      "learning_rate": 9.952904238618524e-06,
+      "loss": 0.659,
+      "step": 634
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0165545957227495,
+      "learning_rate": 9.968602825745683e-06,
+      "loss": 0.6706,
+      "step": 635
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3601666233841225,
+      "learning_rate": 9.984301412872842e-06,
+      "loss": 0.7179,
+      "step": 636
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3501068624416255,
+      "learning_rate": 1e-05,
+      "loss": 0.7399,
+      "step": 637
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6833155776488424,
+      "learning_rate": 1.0015698587127158e-05,
+      "loss": 0.6117,
+      "step": 638
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5201897794923616,
+      "learning_rate": 1.0031397174254319e-05,
+      "loss": 0.7223,
+      "step": 639
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4877770749212784,
+      "learning_rate": 1.0047095761381476e-05,
+      "loss": 0.6125,
+      "step": 640
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5893391602861668,
+      "learning_rate": 1.0062794348508635e-05,
+      "loss": 0.7442,
+      "step": 641
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6110562852560002,
+      "learning_rate": 1.0078492935635794e-05,
+      "loss": 0.7277,
+      "step": 642
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.696470541300494,
+      "learning_rate": 1.0094191522762952e-05,
+      "loss": 0.7401,
+      "step": 643
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4728869680088343,
+      "learning_rate": 1.010989010989011e-05,
+      "loss": 0.6905,
+      "step": 644
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7237235650379021,
+      "learning_rate": 1.012558869701727e-05,
+      "loss": 0.7101,
+      "step": 645
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.572224516601439,
+      "learning_rate": 1.0141287284144427e-05,
+      "loss": 0.7406,
+      "step": 646
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4777919571829308,
+      "learning_rate": 1.0156985871271588e-05,
+      "loss": 0.864,
+      "step": 647
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.384845687496834,
+      "learning_rate": 1.0172684458398745e-05,
+      "loss": 0.6748,
+      "step": 648
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5966843849770684,
+      "learning_rate": 1.0188383045525902e-05,
+      "loss": 0.7136,
+      "step": 649
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4502381282022532,
+      "learning_rate": 1.0204081632653063e-05,
+      "loss": 0.7442,
+      "step": 650
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4463424055604293,
+      "learning_rate": 1.021978021978022e-05,
+      "loss": 0.8023,
+      "step": 651
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5922774616117388,
+      "learning_rate": 1.0235478806907379e-05,
+      "loss": 0.7442,
+      "step": 652
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4787805771138676,
+      "learning_rate": 1.0251177394034538e-05,
+      "loss": 0.8077,
+      "step": 653
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5632449049696915,
+      "learning_rate": 1.0266875981161697e-05,
+      "loss": 0.4769,
+      "step": 654
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3677491567277063,
+      "learning_rate": 1.0282574568288854e-05,
+      "loss": 0.8618,
+      "step": 655
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4275589637364698,
+      "learning_rate": 1.0298273155416015e-05,
+      "loss": 0.7709,
+      "step": 656
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3623536058307233,
+      "learning_rate": 1.0313971742543172e-05,
+      "loss": 0.6904,
+      "step": 657
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.938809319786773,
+      "learning_rate": 1.0329670329670332e-05,
+      "loss": 0.7434,
+      "step": 658
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6675020946497332,
+      "learning_rate": 1.034536891679749e-05,
+      "loss": 0.7879,
+      "step": 659
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4234461562982548,
+      "learning_rate": 1.0361067503924647e-05,
+      "loss": 0.7461,
+      "step": 660
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5867036432732031,
+      "learning_rate": 1.0376766091051806e-05,
+      "loss": 0.6129,
+      "step": 661
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5809753253475236,
+      "learning_rate": 1.0392464678178965e-05,
+      "loss": 0.8071,
+      "step": 662
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4309389815301488,
+      "learning_rate": 1.0408163265306123e-05,
+      "loss": 0.6716,
+      "step": 663
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2769738438369738,
+      "learning_rate": 1.042386185243328e-05,
+      "loss": 0.7421,
+      "step": 664
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.573809657668073,
+      "learning_rate": 1.0439560439560441e-05,
+      "loss": 0.7322,
+      "step": 665
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4876431797452085,
+      "learning_rate": 1.0455259026687598e-05,
+      "loss": 0.6433,
+      "step": 666
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.401928963637588,
+      "learning_rate": 1.0470957613814759e-05,
+      "loss": 0.7165,
+      "step": 667
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.862256175376017,
+      "learning_rate": 1.0486656200941916e-05,
+      "loss": 0.6886,
+      "step": 668
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.334042421919565,
+      "learning_rate": 1.0502354788069075e-05,
+      "loss": 0.7089,
+      "step": 669
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.766472270483698,
+      "learning_rate": 1.0518053375196234e-05,
+      "loss": 0.8299,
+      "step": 670
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4918161831192414,
+      "learning_rate": 1.0533751962323391e-05,
+      "loss": 0.6913,
+      "step": 671
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4592974338899336,
+      "learning_rate": 1.054945054945055e-05,
+      "loss": 0.7449,
+      "step": 672
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4757758763830062,
+      "learning_rate": 1.0565149136577709e-05,
+      "loss": 0.724,
+      "step": 673
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7317559720595441,
+      "learning_rate": 1.0580847723704868e-05,
+      "loss": 0.6517,
+      "step": 674
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5912046049101805,
+      "learning_rate": 1.0596546310832025e-05,
+      "loss": 0.5731,
+      "step": 675
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6329037009521989,
+      "learning_rate": 1.0612244897959186e-05,
+      "loss": 0.5436,
+      "step": 676
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.618390230350077,
+      "learning_rate": 1.0627943485086343e-05,
+      "loss": 0.7752,
+      "step": 677
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7860312438433057,
+      "learning_rate": 1.0643642072213503e-05,
+      "loss": 0.7993,
+      "step": 678
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6030621054174021,
+      "learning_rate": 1.065934065934066e-05,
+      "loss": 0.7975,
+      "step": 679
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9810156061177364,
+      "learning_rate": 1.067503924646782e-05,
+      "loss": 0.7141,
+      "step": 680
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4524127909924567,
+      "learning_rate": 1.0690737833594977e-05,
+      "loss": 0.8573,
+      "step": 681
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2535711761898387,
+      "learning_rate": 1.0706436420722135e-05,
+      "loss": 0.6777,
+      "step": 682
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.466976762345103,
+      "learning_rate": 1.0722135007849294e-05,
+      "loss": 0.7399,
+      "step": 683
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4134791756296177,
+      "learning_rate": 1.0737833594976452e-05,
+      "loss": 0.7813,
+      "step": 684
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2691160702048063,
+      "learning_rate": 1.0753532182103612e-05,
+      "loss": 0.7623,
+      "step": 685
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7266128159453598,
+      "learning_rate": 1.076923076923077e-05,
+      "loss": 0.7468,
+      "step": 686
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6882546933505551,
+      "learning_rate": 1.078492935635793e-05,
+      "loss": 0.5068,
+      "step": 687
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.336353040392601,
+      "learning_rate": 1.0800627943485087e-05,
+      "loss": 0.8015,
+      "step": 688
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2201362999980405,
+      "learning_rate": 1.0816326530612246e-05,
+      "loss": 0.7983,
+      "step": 689
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4447222108237616,
+      "learning_rate": 1.0832025117739405e-05,
+      "loss": 0.7008,
+      "step": 690
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7964766793641949,
+      "learning_rate": 1.0847723704866562e-05,
+      "loss": 0.6922,
+      "step": 691
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.550225637637075,
+      "learning_rate": 1.0863422291993721e-05,
+      "loss": 0.932,
+      "step": 692
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.634777176114583,
+      "learning_rate": 1.087912087912088e-05,
+      "loss": 0.5219,
+      "step": 693
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4586639654499831,
+      "learning_rate": 1.0894819466248039e-05,
+      "loss": 0.8428,
+      "step": 694
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.456771291542947,
+      "learning_rate": 1.0910518053375196e-05,
+      "loss": 0.7626,
+      "step": 695
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7365656710159199,
+      "learning_rate": 1.0926216640502357e-05,
+      "loss": 0.5501,
+      "step": 696
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1495787017852495,
+      "learning_rate": 1.0941915227629514e-05,
+      "loss": 0.7724,
+      "step": 697
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8404022275143628,
+      "learning_rate": 1.0957613814756674e-05,
+      "loss": 0.7226,
+      "step": 698
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4473830294129695,
+      "learning_rate": 1.0973312401883831e-05,
+      "loss": 0.6748,
+      "step": 699
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5957551288452677,
+      "learning_rate": 1.098901098901099e-05,
+      "loss": 0.4907,
+      "step": 700
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.495751403885016,
+      "learning_rate": 1.1004709576138148e-05,
+      "loss": 0.7637,
+      "step": 701
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5165841951307701,
+      "learning_rate": 1.1020408163265306e-05,
+      "loss": 0.8888,
+      "step": 702
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6626581546952304,
+      "learning_rate": 1.1036106750392465e-05,
+      "loss": 0.6089,
+      "step": 703
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3586794788434031,
+      "learning_rate": 1.1051805337519623e-05,
+      "loss": 0.7378,
+      "step": 704
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5861717155235726,
+      "learning_rate": 1.1067503924646783e-05,
+      "loss": 0.4673,
+      "step": 705
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.457257831444212,
+      "learning_rate": 1.108320251177394e-05,
+      "loss": 0.7533,
+      "step": 706
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.385408456894965,
+      "learning_rate": 1.1098901098901101e-05,
+      "loss": 0.7737,
+      "step": 707
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4027266549419262,
+      "learning_rate": 1.1114599686028258e-05,
+      "loss": 0.719,
+      "step": 708
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4145067289355508,
+      "learning_rate": 1.1130298273155417e-05,
+      "loss": 0.7576,
+      "step": 709
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.798582493708091,
+      "learning_rate": 1.1145996860282576e-05,
+      "loss": 0.6479,
+      "step": 710
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5614796353743003,
+      "learning_rate": 1.1161695447409735e-05,
+      "loss": 0.6538,
+      "step": 711
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5457173648718395,
+      "learning_rate": 1.1177394034536892e-05,
+      "loss": 0.4715,
+      "step": 712
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8691639439728631,
+      "learning_rate": 1.119309262166405e-05,
+      "loss": 0.7266,
+      "step": 713
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4858682995243078,
+      "learning_rate": 1.120879120879121e-05,
+      "loss": 0.8205,
+      "step": 714
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.330108421655368,
+      "learning_rate": 1.1224489795918367e-05,
+      "loss": 0.8097,
+      "step": 715
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6215142239312068,
+      "learning_rate": 1.1240188383045527e-05,
+      "loss": 0.6807,
+      "step": 716
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6051898187139553,
+      "learning_rate": 1.1255886970172685e-05,
+      "loss": 0.4725,
+      "step": 717
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3854933582449238,
+      "learning_rate": 1.1271585557299845e-05,
+      "loss": 0.5599,
+      "step": 718
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5675160524642869,
+      "learning_rate": 1.1287284144427002e-05,
+      "loss": 0.4437,
+      "step": 719
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5686756933116481,
+      "learning_rate": 1.1302982731554161e-05,
+      "loss": 0.4493,
+      "step": 720
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5209896354821375,
+      "learning_rate": 1.131868131868132e-05,
+      "loss": 0.8075,
+      "step": 721
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.663286861300298,
+      "learning_rate": 1.1334379905808479e-05,
+      "loss": 0.731,
+      "step": 722
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5870984813391148,
+      "learning_rate": 1.1350078492935636e-05,
+      "loss": 0.4546,
+      "step": 723
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7782878178030708,
+      "learning_rate": 1.1365777080062793e-05,
+      "loss": 0.6347,
+      "step": 724
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6130077425123194,
+      "learning_rate": 1.1381475667189954e-05,
+      "loss": 0.4605,
+      "step": 725
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4195952568630923,
+      "learning_rate": 1.1397174254317111e-05,
+      "loss": 0.7804,
+      "step": 726
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8377573700190437,
+      "learning_rate": 1.1412872841444272e-05,
+      "loss": 0.5677,
+      "step": 727
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3726944598738222,
+      "learning_rate": 1.1428571428571429e-05,
+      "loss": 0.8306,
+      "step": 728
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2638297826655789,
+      "learning_rate": 1.1444270015698588e-05,
+      "loss": 0.7368,
+      "step": 729
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 11.233557361510204,
+      "learning_rate": 1.1459968602825747e-05,
+      "loss": 0.6472,
+      "step": 730
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4062355834562206,
+      "learning_rate": 1.1475667189952906e-05,
+      "loss": 0.7553,
+      "step": 731
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9272577169604597,
+      "learning_rate": 1.1491365777080063e-05,
+      "loss": 0.8126,
+      "step": 732
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2618195902475935,
+      "learning_rate": 1.1507064364207223e-05,
+      "loss": 0.6581,
+      "step": 733
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5200560456737235,
+      "learning_rate": 1.152276295133438e-05,
+      "loss": 0.7415,
+      "step": 734
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.544467480842421,
+      "learning_rate": 1.1538461538461538e-05,
+      "loss": 0.676,
+      "step": 735
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5222153776146146,
+      "learning_rate": 1.1554160125588698e-05,
+      "loss": 0.7349,
+      "step": 736
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4402292363260012,
+      "learning_rate": 1.1569858712715856e-05,
+      "loss": 0.8364,
+      "step": 737
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6105642963177388,
+      "learning_rate": 1.1585557299843016e-05,
+      "loss": 0.4639,
+      "step": 738
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9613641776135564,
+      "learning_rate": 1.1601255886970173e-05,
+      "loss": 0.7012,
+      "step": 739
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.508254380877017,
+      "learning_rate": 1.1616954474097332e-05,
+      "loss": 0.6033,
+      "step": 740
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5304618082307742,
+      "learning_rate": 1.1632653061224491e-05,
+      "loss": 0.7369,
+      "step": 741
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5390099846511227,
+      "learning_rate": 1.164835164835165e-05,
+      "loss": 0.7877,
+      "step": 742
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.523288314620273,
+      "learning_rate": 1.1664050235478807e-05,
+      "loss": 0.8279,
+      "step": 743
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.647565281577536,
+      "learning_rate": 1.1679748822605968e-05,
+      "loss": 0.7526,
+      "step": 744
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.530290995693306,
+      "learning_rate": 1.1695447409733125e-05,
+      "loss": 0.8365,
+      "step": 745
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5813515647672063,
+      "learning_rate": 1.1711145996860282e-05,
+      "loss": 0.6745,
+      "step": 746
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8253429636090412,
+      "learning_rate": 1.1726844583987443e-05,
+      "loss": 0.6771,
+      "step": 747
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4346771578097357,
+      "learning_rate": 1.17425431711146e-05,
+      "loss": 0.7635,
+      "step": 748
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7639159689984885,
+      "learning_rate": 1.1758241758241759e-05,
+      "loss": 0.8574,
+      "step": 749
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5952344275447993,
+      "learning_rate": 1.1773940345368918e-05,
+      "loss": 0.4753,
+      "step": 750
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5650195340867494,
+      "learning_rate": 1.1789638932496077e-05,
+      "loss": 0.8324,
+      "step": 751
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5590405565619367,
+      "learning_rate": 1.1805337519623234e-05,
+      "loss": 0.6536,
+      "step": 752
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5974451661807119,
+      "learning_rate": 1.1821036106750394e-05,
+      "loss": 0.6357,
+      "step": 753
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5873518993964444,
+      "learning_rate": 1.1836734693877552e-05,
+      "loss": 0.6916,
+      "step": 754
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8297031418951908,
+      "learning_rate": 1.1852433281004712e-05,
+      "loss": 0.7272,
+      "step": 755
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5444987702139867,
+      "learning_rate": 1.186813186813187e-05,
+      "loss": 0.6621,
+      "step": 756
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4796371647411966,
+      "learning_rate": 1.1883830455259027e-05,
+      "loss": 0.6025,
+      "step": 757
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4854760563866134,
+      "learning_rate": 1.1899529042386187e-05,
+      "loss": 0.8386,
+      "step": 758
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4325501362007622,
+      "learning_rate": 1.1915227629513344e-05,
+      "loss": 0.6768,
+      "step": 759
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3458040325399596,
+      "learning_rate": 1.1930926216640503e-05,
+      "loss": 0.9307,
+      "step": 760
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.775938579723207,
+      "learning_rate": 1.1946624803767662e-05,
+      "loss": 0.8131,
+      "step": 761
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7125079734089077,
+      "learning_rate": 1.1962323390894821e-05,
+      "loss": 0.7454,
+      "step": 762
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5823278333638489,
+      "learning_rate": 1.1978021978021978e-05,
+      "loss": 0.6137,
+      "step": 763
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5424537390522466,
+      "learning_rate": 1.1993720565149139e-05,
+      "loss": 0.8354,
+      "step": 764
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1845268021281603,
+      "learning_rate": 1.2009419152276296e-05,
+      "loss": 0.6212,
+      "step": 765
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8404111692794953,
+      "learning_rate": 1.2025117739403457e-05,
+      "loss": 0.8172,
+      "step": 766
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5366249376671055,
+      "learning_rate": 1.2040816326530614e-05,
+      "loss": 0.8574,
+      "step": 767
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5574571235584196,
+      "learning_rate": 1.2056514913657771e-05,
+      "loss": 0.7697,
+      "step": 768
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9223811114494203,
+      "learning_rate": 1.207221350078493e-05,
+      "loss": 0.6186,
+      "step": 769
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3920777830919737,
+      "learning_rate": 1.2087912087912089e-05,
+      "loss": 0.5838,
+      "step": 770
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4142059763521868,
+      "learning_rate": 1.2103610675039248e-05,
+      "loss": 0.8126,
+      "step": 771
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.83566113914711,
+      "learning_rate": 1.2119309262166405e-05,
+      "loss": 0.6763,
+      "step": 772
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.585405405646765,
+      "learning_rate": 1.2135007849293565e-05,
+      "loss": 0.6805,
+      "step": 773
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4924183663102852,
+      "learning_rate": 1.2150706436420723e-05,
+      "loss": 0.716,
+      "step": 774
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1112427738070108,
+      "learning_rate": 1.2166405023547883e-05,
+      "loss": 0.6905,
+      "step": 775
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.520992286807098,
+      "learning_rate": 1.218210361067504e-05,
+      "loss": 0.7732,
+      "step": 776
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4974048559765296,
+      "learning_rate": 1.21978021978022e-05,
+      "loss": 0.8304,
+      "step": 777
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5000900466150076,
+      "learning_rate": 1.2213500784929358e-05,
+      "loss": 0.7207,
+      "step": 778
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6216961377163315,
+      "learning_rate": 1.2229199372056515e-05,
+      "loss": 0.7322,
+      "step": 779
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.940531603494104,
+      "learning_rate": 1.2244897959183674e-05,
+      "loss": 0.6548,
+      "step": 780
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3236067486471885,
+      "learning_rate": 1.2260596546310833e-05,
+      "loss": 0.7807,
+      "step": 781
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3504830758127702,
+      "learning_rate": 1.2276295133437992e-05,
+      "loss": 0.7587,
+      "step": 782
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6156470678666277,
+      "learning_rate": 1.229199372056515e-05,
+      "loss": 0.7607,
+      "step": 783
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5084412530638853,
+      "learning_rate": 1.230769230769231e-05,
+      "loss": 0.6718,
+      "step": 784
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5288103972510865,
+      "learning_rate": 1.2323390894819467e-05,
+      "loss": 0.6531,
+      "step": 785
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5690888831539056,
+      "learning_rate": 1.2339089481946628e-05,
+      "loss": 0.8814,
+      "step": 786
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.283970437865121,
+      "learning_rate": 1.2354788069073785e-05,
+      "loss": 0.7311,
+      "step": 787
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3983938055331497,
+      "learning_rate": 1.2370486656200944e-05,
+      "loss": 0.6704,
+      "step": 788
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4901842767371525,
+      "learning_rate": 1.23861852433281e-05,
+      "loss": 0.7206,
+      "step": 789
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5374465009636724,
+      "learning_rate": 1.240188383045526e-05,
+      "loss": 0.7361,
+      "step": 790
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.119517986676318,
+      "learning_rate": 1.2417582417582419e-05,
+      "loss": 0.8407,
+      "step": 791
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7286691121171351,
+      "learning_rate": 1.2433281004709576e-05,
+      "loss": 0.8776,
+      "step": 792
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5702281887829106,
+      "learning_rate": 1.2448979591836736e-05,
+      "loss": 0.7683,
+      "step": 793
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.583930012796994,
+      "learning_rate": 1.2464678178963894e-05,
+      "loss": 0.4958,
+      "step": 794
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2058065422182378,
+      "learning_rate": 1.2480376766091054e-05,
+      "loss": 0.705,
+      "step": 795
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5702222679560465,
+      "learning_rate": 1.2496075353218211e-05,
+      "loss": 0.7791,
+      "step": 796
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7102976595435775,
+      "learning_rate": 1.251177394034537e-05,
+      "loss": 0.8026,
+      "step": 797
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.463623346101223,
+      "learning_rate": 1.2527472527472529e-05,
+      "loss": 0.8124,
+      "step": 798
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7873500518396406,
+      "learning_rate": 1.2543171114599686e-05,
+      "loss": 0.8068,
+      "step": 799
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3168473625779105,
+      "learning_rate": 1.2558869701726845e-05,
+      "loss": 0.7897,
+      "step": 800
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0955556662812547,
+      "learning_rate": 1.2574568288854004e-05,
+      "loss": 0.719,
+      "step": 801
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4566721645498386,
+      "learning_rate": 1.2590266875981163e-05,
+      "loss": 0.8006,
+      "step": 802
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5045461076856967,
+      "learning_rate": 1.260596546310832e-05,
+      "loss": 0.8106,
+      "step": 803
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.465760198990363,
+      "learning_rate": 1.262166405023548e-05,
+      "loss": 0.8324,
+      "step": 804
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6057549369871276,
+      "learning_rate": 1.2637362637362638e-05,
+      "loss": 0.8481,
+      "step": 805
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4445405545673706,
+      "learning_rate": 1.2653061224489798e-05,
+      "loss": 0.6608,
+      "step": 806
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.279393607849924,
+      "learning_rate": 1.2668759811616956e-05,
+      "loss": 0.6695,
+      "step": 807
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5610218273211431,
+      "learning_rate": 1.2684458398744115e-05,
+      "loss": 0.7317,
+      "step": 808
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7218370964162106,
+      "learning_rate": 1.2700156985871272e-05,
+      "loss": 0.7721,
+      "step": 809
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6278317768737762,
+      "learning_rate": 1.271585557299843e-05,
+      "loss": 0.6693,
+      "step": 810
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3541517510678265,
+      "learning_rate": 1.273155416012559e-05,
+      "loss": 0.8448,
+      "step": 811
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.56045780423589,
+      "learning_rate": 1.2747252747252747e-05,
+      "loss": 0.6778,
+      "step": 812
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9983791036683427,
+      "learning_rate": 1.2762951334379907e-05,
+      "loss": 0.8244,
+      "step": 813
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4554700133899645,
+      "learning_rate": 1.2778649921507064e-05,
+      "loss": 0.6279,
+      "step": 814
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4633204094364558,
+      "learning_rate": 1.2794348508634225e-05,
+      "loss": 0.8315,
+      "step": 815
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4108310898076994,
+      "learning_rate": 1.2810047095761382e-05,
+      "loss": 0.651,
+      "step": 816
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5887185638405534,
+      "learning_rate": 1.2825745682888541e-05,
+      "loss": 0.6075,
+      "step": 817
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4087087804526148,
+      "learning_rate": 1.28414442700157e-05,
+      "loss": 0.751,
+      "step": 818
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4213187634077689,
+      "learning_rate": 1.2857142857142859e-05,
+      "loss": 0.5663,
+      "step": 819
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.535453036309214,
+      "learning_rate": 1.2872841444270016e-05,
+      "loss": 0.6877,
+      "step": 820
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4918317497507205,
+      "learning_rate": 1.2888540031397175e-05,
+      "loss": 0.7965,
+      "step": 821
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.588153046965294,
+      "learning_rate": 1.2904238618524334e-05,
+      "loss": 0.6421,
+      "step": 822
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8087813756492699,
+      "learning_rate": 1.2919937205651491e-05,
+      "loss": 0.6837,
+      "step": 823
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7347075668322087,
+      "learning_rate": 1.2935635792778652e-05,
+      "loss": 0.7129,
+      "step": 824
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3784488998881521,
+      "learning_rate": 1.2951334379905809e-05,
+      "loss": 0.663,
+      "step": 825
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6003478478761772,
+      "learning_rate": 1.296703296703297e-05,
+      "loss": 0.8473,
+      "step": 826
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3833806301497964,
+      "learning_rate": 1.2982731554160127e-05,
+      "loss": 0.6617,
+      "step": 827
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7969711471783805,
+      "learning_rate": 1.2998430141287286e-05,
+      "loss": 0.77,
+      "step": 828
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5026143986559866,
+      "learning_rate": 1.3014128728414443e-05,
+      "loss": 0.7479,
+      "step": 829
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8378293518923168,
+      "learning_rate": 1.3029827315541603e-05,
+      "loss": 0.7049,
+      "step": 830
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6477799063966947,
+      "learning_rate": 1.304552590266876e-05,
+      "loss": 0.799,
+      "step": 831
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3321679669168995,
+      "learning_rate": 1.3061224489795918e-05,
+      "loss": 0.6885,
+      "step": 832
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5947934033830808,
+      "learning_rate": 1.3076923076923078e-05,
+      "loss": 0.6454,
+      "step": 833
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4246478093861443,
+      "learning_rate": 1.3092621664050235e-05,
+      "loss": 0.8763,
+      "step": 834
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4740711017424173,
+      "learning_rate": 1.3108320251177396e-05,
+      "loss": 0.7913,
+      "step": 835
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5456801641988114,
+      "learning_rate": 1.3124018838304553e-05,
+      "loss": 0.8524,
+      "step": 836
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3810517927444137,
+      "learning_rate": 1.3139717425431712e-05,
+      "loss": 0.6164,
+      "step": 837
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3560685777097017,
+      "learning_rate": 1.3155416012558871e-05,
+      "loss": 0.6945,
+      "step": 838
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5913609245971403,
+      "learning_rate": 1.317111459968603e-05,
+      "loss": 0.6394,
+      "step": 839
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.415794255355489,
+      "learning_rate": 1.3186813186813187e-05,
+      "loss": 0.7329,
+      "step": 840
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6711193923320524,
+      "learning_rate": 1.3202511773940348e-05,
+      "loss": 0.7333,
+      "step": 841
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5817518524219405,
+      "learning_rate": 1.3218210361067505e-05,
+      "loss": 0.5025,
+      "step": 842
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5583159378963674,
+      "learning_rate": 1.3233908948194662e-05,
+      "loss": 0.701,
+      "step": 843
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4976364805751405,
+      "learning_rate": 1.3249607535321823e-05,
+      "loss": 0.8128,
+      "step": 844
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3579130766386927,
+      "learning_rate": 1.326530612244898e-05,
+      "loss": 0.8128,
+      "step": 845
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.596745622929281,
+      "learning_rate": 1.328100470957614e-05,
+      "loss": 0.6846,
+      "step": 846
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.549457648026892,
+      "learning_rate": 1.3296703296703298e-05,
+      "loss": 0.7502,
+      "step": 847
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.560981024086419,
+      "learning_rate": 1.3312401883830456e-05,
+      "loss": 0.6189,
+      "step": 848
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4732976064375602,
+      "learning_rate": 1.3328100470957614e-05,
+      "loss": 0.6761,
+      "step": 849
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4257086206865617,
+      "learning_rate": 1.3343799058084774e-05,
+      "loss": 0.6882,
+      "step": 850
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.57200276198082,
+      "learning_rate": 1.3359497645211931e-05,
+      "loss": 0.8091,
+      "step": 851
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.623474145788031,
+      "learning_rate": 1.3375196232339092e-05,
+      "loss": 0.7814,
+      "step": 852
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5585613649213723,
+      "learning_rate": 1.339089481946625e-05,
+      "loss": 0.746,
+      "step": 853
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6446087389643358,
+      "learning_rate": 1.3406593406593406e-05,
+      "loss": 0.6051,
+      "step": 854
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3880715801005372,
+      "learning_rate": 1.3422291993720567e-05,
+      "loss": 0.752,
+      "step": 855
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.089002314954026,
+      "learning_rate": 1.3437990580847724e-05,
+      "loss": 0.7532,
+      "step": 856
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4035018001222752,
+      "learning_rate": 1.3453689167974883e-05,
+      "loss": 0.7873,
+      "step": 857
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3643392169922335,
+      "learning_rate": 1.3469387755102042e-05,
+      "loss": 0.6596,
+      "step": 858
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2737606641412351,
+      "learning_rate": 1.34850863422292e-05,
+      "loss": 0.6788,
+      "step": 859
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3014372157655734,
+      "learning_rate": 1.3500784929356358e-05,
+      "loss": 0.6309,
+      "step": 860
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1499768866689313,
+      "learning_rate": 1.3516483516483519e-05,
+      "loss": 0.8153,
+      "step": 861
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6802225734799204,
+      "learning_rate": 1.3532182103610676e-05,
+      "loss": 0.495,
+      "step": 862
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4326360131397198,
+      "learning_rate": 1.3547880690737836e-05,
+      "loss": 0.7187,
+      "step": 863
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4493569009871627,
+      "learning_rate": 1.3563579277864994e-05,
+      "loss": 0.5837,
+      "step": 864
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6806366763669602,
+      "learning_rate": 1.357927786499215e-05,
+      "loss": 0.6569,
+      "step": 865
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5033091761350774,
+      "learning_rate": 1.3594976452119311e-05,
+      "loss": 0.6426,
+      "step": 866
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6991684362979063,
+      "learning_rate": 1.3610675039246469e-05,
+      "loss": 0.723,
+      "step": 867
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4923624319531608,
+      "learning_rate": 1.3626373626373627e-05,
+      "loss": 0.601,
+      "step": 868
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6974887503356335,
+      "learning_rate": 1.3642072213500786e-05,
+      "loss": 0.8607,
+      "step": 869
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5689016779794964,
+      "learning_rate": 1.3657770800627945e-05,
+      "loss": 0.6365,
+      "step": 870
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3538799972853173,
+      "learning_rate": 1.3673469387755102e-05,
+      "loss": 0.7051,
+      "step": 871
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6192964397182663,
+      "learning_rate": 1.3689167974882263e-05,
+      "loss": 0.6734,
+      "step": 872
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.629685433454437,
+      "learning_rate": 1.370486656200942e-05,
+      "loss": 0.7307,
+      "step": 873
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.572469596967176,
+      "learning_rate": 1.3720565149136579e-05,
+      "loss": 0.8507,
+      "step": 874
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.404848884973224,
+      "learning_rate": 1.3736263736263738e-05,
+      "loss": 0.7957,
+      "step": 875
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5499869402665092,
+      "learning_rate": 1.3751962323390895e-05,
+      "loss": 0.7711,
+      "step": 876
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5527642819610383,
+      "learning_rate": 1.3767660910518054e-05,
+      "loss": 0.7926,
+      "step": 877
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4212352157041956,
+      "learning_rate": 1.3783359497645213e-05,
+      "loss": 0.7901,
+      "step": 878
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4163562202566053,
+      "learning_rate": 1.3799058084772372e-05,
+      "loss": 0.7176,
+      "step": 879
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.356656329528207,
+      "learning_rate": 1.3814756671899529e-05,
+      "loss": 0.7457,
+      "step": 880
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5270591504449533,
+      "learning_rate": 1.383045525902669e-05,
+      "loss": 0.6205,
+      "step": 881
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.416561310441446,
+      "learning_rate": 1.3846153846153847e-05,
+      "loss": 0.7311,
+      "step": 882
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5907108417778573,
+      "learning_rate": 1.3861852433281007e-05,
+      "loss": 0.4691,
+      "step": 883
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.407300819686516,
+      "learning_rate": 1.3877551020408165e-05,
+      "loss": 0.8299,
+      "step": 884
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.538330633570945,
+      "learning_rate": 1.3893249607535323e-05,
+      "loss": 0.7636,
+      "step": 885
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5150973623272534,
+      "learning_rate": 1.3908948194662482e-05,
+      "loss": 0.6649,
+      "step": 886
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2159963899555903,
+      "learning_rate": 1.392464678178964e-05,
+      "loss": 0.8037,
+      "step": 887
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4680569112513224,
+      "learning_rate": 1.3940345368916798e-05,
+      "loss": 0.7658,
+      "step": 888
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6245571338805405,
+      "learning_rate": 1.3956043956043957e-05,
+      "loss": 0.6997,
+      "step": 889
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3689806504982411,
+      "learning_rate": 1.3971742543171116e-05,
+      "loss": 0.7351,
+      "step": 890
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5519532234016695,
+      "learning_rate": 1.3987441130298273e-05,
+      "loss": 0.757,
+      "step": 891
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4965588816187163,
+      "learning_rate": 1.4003139717425434e-05,
+      "loss": 0.6823,
+      "step": 892
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5776481887745177,
+      "learning_rate": 1.4018838304552591e-05,
+      "loss": 0.8067,
+      "step": 893
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3909710453499988,
+      "learning_rate": 1.403453689167975e-05,
+      "loss": 0.7196,
+      "step": 894
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3193426835615965,
+      "learning_rate": 1.4050235478806909e-05,
+      "loss": 0.7125,
+      "step": 895
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.647239135099113,
+      "learning_rate": 1.4065934065934068e-05,
+      "loss": 0.6515,
+      "step": 896
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3976297524406844,
+      "learning_rate": 1.4081632653061225e-05,
+      "loss": 0.6992,
+      "step": 897
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5775543607833065,
+      "learning_rate": 1.4097331240188384e-05,
+      "loss": 0.6636,
+      "step": 898
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7235177337192076,
+      "learning_rate": 1.4113029827315543e-05,
+      "loss": 0.8471,
+      "step": 899
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3958503605421928,
+      "learning_rate": 1.41287284144427e-05,
+      "loss": 0.8439,
+      "step": 900
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5354127786265954,
+      "learning_rate": 1.414442700156986e-05,
+      "loss": 0.7663,
+      "step": 901
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5479325125819456,
+      "learning_rate": 1.4160125588697018e-05,
+      "loss": 0.7179,
+      "step": 902
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.58420040569103,
+      "learning_rate": 1.4175824175824178e-05,
+      "loss": 0.7373,
+      "step": 903
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5782246841009513,
+      "learning_rate": 1.4191522762951335e-05,
+      "loss": 0.5347,
+      "step": 904
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3193078684212198,
+      "learning_rate": 1.4207221350078494e-05,
+      "loss": 0.8321,
+      "step": 905
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4142029483450547,
+      "learning_rate": 1.4222919937205653e-05,
+      "loss": 0.6748,
+      "step": 906
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4891481622798228,
+      "learning_rate": 1.423861852433281e-05,
+      "loss": 0.7784,
+      "step": 907
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4798716856067258,
+      "learning_rate": 1.425431711145997e-05,
+      "loss": 0.8133,
+      "step": 908
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3368176401613698,
+      "learning_rate": 1.4270015698587128e-05,
+      "loss": 0.6781,
+      "step": 909
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4699450165951962,
+      "learning_rate": 1.4285714285714287e-05,
+      "loss": 0.762,
+      "step": 910
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.374591422734443,
+      "learning_rate": 1.4301412872841444e-05,
+      "loss": 0.6661,
+      "step": 911
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3728434305893833,
+      "learning_rate": 1.4317111459968605e-05,
+      "loss": 0.6097,
+      "step": 912
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4675309622542065,
+      "learning_rate": 1.4332810047095762e-05,
+      "loss": 0.7688,
+      "step": 913
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7291992829147844,
+      "learning_rate": 1.4348508634222923e-05,
+      "loss": 0.571,
+      "step": 914
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2757657869884296,
+      "learning_rate": 1.436420722135008e-05,
+      "loss": 0.692,
+      "step": 915
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3527691846337015,
+      "learning_rate": 1.4379905808477239e-05,
+      "loss": 0.6702,
+      "step": 916
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4064413723083722,
+      "learning_rate": 1.4395604395604396e-05,
+      "loss": 0.6399,
+      "step": 917
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9061746383685723,
+      "learning_rate": 1.4411302982731555e-05,
+      "loss": 0.7959,
+      "step": 918
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.380904865440847,
+      "learning_rate": 1.4427001569858714e-05,
+      "loss": 0.7259,
+      "step": 919
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5717200853631934,
+      "learning_rate": 1.4442700156985871e-05,
+      "loss": 0.7727,
+      "step": 920
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4198052362996205,
+      "learning_rate": 1.4458398744113031e-05,
+      "loss": 0.8249,
+      "step": 921
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.404700693711668,
+      "learning_rate": 1.4474097331240189e-05,
+      "loss": 0.7307,
+      "step": 922
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.319546826750454,
+      "learning_rate": 1.448979591836735e-05,
+      "loss": 0.7718,
+      "step": 923
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6099771491489008,
+      "learning_rate": 1.4505494505494506e-05,
+      "loss": 0.8022,
+      "step": 924
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7486109346733392,
+      "learning_rate": 1.4521193092621665e-05,
+      "loss": 0.7576,
+      "step": 925
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.687398553743063,
+      "learning_rate": 1.4536891679748824e-05,
+      "loss": 0.7816,
+      "step": 926
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6010747532245152,
+      "learning_rate": 1.4552590266875983e-05,
+      "loss": 0.5191,
+      "step": 927
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5734238068542423,
+      "learning_rate": 1.456828885400314e-05,
+      "loss": 0.7306,
+      "step": 928
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5848262735770351,
+      "learning_rate": 1.45839874411303e-05,
+      "loss": 0.4672,
+      "step": 929
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4909270574983147,
+      "learning_rate": 1.4599686028257458e-05,
+      "loss": 0.6682,
+      "step": 930
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5969476187212084,
+      "learning_rate": 1.4615384615384615e-05,
+      "loss": 0.7526,
+      "step": 931
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.773542318242813,
+      "learning_rate": 1.4631083202511776e-05,
+      "loss": 0.7827,
+      "step": 932
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.361199074741736,
+      "learning_rate": 1.4646781789638933e-05,
+      "loss": 0.7197,
+      "step": 933
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4336874581881478,
+      "learning_rate": 1.4662480376766094e-05,
+      "loss": 0.8191,
+      "step": 934
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.598529828182746,
+      "learning_rate": 1.467817896389325e-05,
+      "loss": 0.4689,
+      "step": 935
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4812748356860472,
+      "learning_rate": 1.469387755102041e-05,
+      "loss": 0.7344,
+      "step": 936
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8474732146650168,
+      "learning_rate": 1.4709576138147567e-05,
+      "loss": 0.7895,
+      "step": 937
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6067070265024698,
+      "learning_rate": 1.4725274725274727e-05,
+      "loss": 0.543,
+      "step": 938
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.915271512760823,
+      "learning_rate": 1.4740973312401885e-05,
+      "loss": 0.737,
+      "step": 939
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.624239103778881,
+      "learning_rate": 1.4756671899529042e-05,
+      "loss": 0.6797,
+      "step": 940
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6427877466251736,
+      "learning_rate": 1.4772370486656202e-05,
+      "loss": 0.5816,
+      "step": 941
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1877855018998087,
+      "learning_rate": 1.478806907378336e-05,
+      "loss": 0.766,
+      "step": 942
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6376627899320052,
+      "learning_rate": 1.480376766091052e-05,
+      "loss": 0.4271,
+      "step": 943
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4111335592903154,
+      "learning_rate": 1.4819466248037677e-05,
+      "loss": 0.771,
+      "step": 944
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.570505427456985,
+      "learning_rate": 1.4835164835164836e-05,
+      "loss": 0.664,
+      "step": 945
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9208404590022001,
+      "learning_rate": 1.4850863422291995e-05,
+      "loss": 0.6921,
+      "step": 946
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5773092660488461,
+      "learning_rate": 1.4866562009419154e-05,
+      "loss": 0.7163,
+      "step": 947
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6208863653892802,
+      "learning_rate": 1.4882260596546311e-05,
+      "loss": 0.8295,
+      "step": 948
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.644456323714995,
+      "learning_rate": 1.4897959183673472e-05,
+      "loss": 0.687,
+      "step": 949
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5452051776594233,
+      "learning_rate": 1.4913657770800629e-05,
+      "loss": 0.6404,
+      "step": 950
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.605350724460326,
+      "learning_rate": 1.4929356357927786e-05,
+      "loss": 0.5431,
+      "step": 951
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5595139800920443,
+      "learning_rate": 1.4945054945054947e-05,
+      "loss": 0.721,
+      "step": 952
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5659228733477792,
+      "learning_rate": 1.4960753532182104e-05,
+      "loss": 0.7446,
+      "step": 953
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6546854507858375,
+      "learning_rate": 1.4976452119309265e-05,
+      "loss": 0.8042,
+      "step": 954
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5726816878467278,
+      "learning_rate": 1.4992150706436422e-05,
+      "loss": 0.7272,
+      "step": 955
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4073067710243046,
+      "learning_rate": 1.500784929356358e-05,
+      "loss": 0.6821,
+      "step": 956
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5547507579331423,
+      "learning_rate": 1.5023547880690738e-05,
+      "loss": 0.687,
+      "step": 957
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4479036868640347,
+      "learning_rate": 1.5039246467817898e-05,
+      "loss": 0.78,
+      "step": 958
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4445819428818532,
+      "learning_rate": 1.5054945054945056e-05,
+      "loss": 0.7851,
+      "step": 959
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.151434281918363,
+      "learning_rate": 1.5070643642072216e-05,
+      "loss": 0.8466,
+      "step": 960
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4158856749818947,
+      "learning_rate": 1.5086342229199373e-05,
+      "loss": 0.8152,
+      "step": 961
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6252755381271295,
+      "learning_rate": 1.510204081632653e-05,
+      "loss": 0.712,
+      "step": 962
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5324680675851907,
+      "learning_rate": 1.5117739403453691e-05,
+      "loss": 0.8475,
+      "step": 963
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.039069863529676,
+      "learning_rate": 1.5133437990580848e-05,
+      "loss": 0.6861,
+      "step": 964
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6068377181962217,
+      "learning_rate": 1.5149136577708007e-05,
+      "loss": 0.4986,
+      "step": 965
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.590920175115581,
+      "learning_rate": 1.5164835164835166e-05,
+      "loss": 0.4874,
+      "step": 966
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5336658352552062,
+      "learning_rate": 1.5180533751962325e-05,
+      "loss": 0.6441,
+      "step": 967
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7563112393566962,
+      "learning_rate": 1.5196232339089482e-05,
+      "loss": 0.7317,
+      "step": 968
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.844890311215925,
+      "learning_rate": 1.5211930926216643e-05,
+      "loss": 0.6411,
+      "step": 969
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3377037315564786,
+      "learning_rate": 1.52276295133438e-05,
+      "loss": 0.6063,
+      "step": 970
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6147851936861354,
+      "learning_rate": 1.524332810047096e-05,
+      "loss": 0.5154,
+      "step": 971
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4087297667502394,
+      "learning_rate": 1.5259026687598116e-05,
+      "loss": 0.697,
+      "step": 972
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2640948036046074,
+      "learning_rate": 1.5274725274725277e-05,
+      "loss": 0.7164,
+      "step": 973
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4965362396554505,
+      "learning_rate": 1.5290423861852434e-05,
+      "loss": 0.8137,
+      "step": 974
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.431188717901761,
+      "learning_rate": 1.530612244897959e-05,
+      "loss": 0.6626,
+      "step": 975
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.532437444701721,
+      "learning_rate": 1.532182103610675e-05,
+      "loss": 0.7102,
+      "step": 976
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.544318389164944,
+      "learning_rate": 1.533751962323391e-05,
+      "loss": 0.6195,
+      "step": 977
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6201372050917974,
+      "learning_rate": 1.535321821036107e-05,
+      "loss": 0.5547,
+      "step": 978
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6596414453742387,
+      "learning_rate": 1.5368916797488227e-05,
+      "loss": 0.6875,
+      "step": 979
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5937531270129133,
+      "learning_rate": 1.5384615384615387e-05,
+      "loss": 0.8019,
+      "step": 980
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8776806696184392,
+      "learning_rate": 1.5400313971742544e-05,
+      "loss": 0.7903,
+      "step": 981
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7934547202419049,
+      "learning_rate": 1.5416012558869705e-05,
+      "loss": 0.7033,
+      "step": 982
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.526251089768118,
+      "learning_rate": 1.5431711145996862e-05,
+      "loss": 0.7164,
+      "step": 983
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6147051034669789,
+      "learning_rate": 1.544740973312402e-05,
+      "loss": 0.4947,
+      "step": 984
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4060214623540803,
+      "learning_rate": 1.546310832025118e-05,
+      "loss": 0.8308,
+      "step": 985
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.502855753613786,
+      "learning_rate": 1.5478806907378337e-05,
+      "loss": 0.7795,
+      "step": 986
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6895721092179399,
+      "learning_rate": 1.5494505494505498e-05,
+      "loss": 0.7529,
+      "step": 987
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8007546873933444,
+      "learning_rate": 1.5510204081632655e-05,
+      "loss": 0.6505,
+      "step": 988
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.283394843930391,
+      "learning_rate": 1.5525902668759812e-05,
+      "loss": 0.7459,
+      "step": 989
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5285277299585365,
+      "learning_rate": 1.5541601255886973e-05,
+      "loss": 0.6758,
+      "step": 990
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8543678250420117,
+      "learning_rate": 1.555729984301413e-05,
+      "loss": 0.7218,
+      "step": 991
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.584825114700285,
+      "learning_rate": 1.5572998430141287e-05,
+      "loss": 0.4843,
+      "step": 992
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.616070139152509,
+      "learning_rate": 1.5588697017268448e-05,
+      "loss": 0.618,
+      "step": 993
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4653131151042968,
+      "learning_rate": 1.5604395604395605e-05,
+      "loss": 0.6604,
+      "step": 994
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4273514501198175,
+      "learning_rate": 1.5620094191522762e-05,
+      "loss": 0.6527,
+      "step": 995
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4760673966740687,
+      "learning_rate": 1.5635792778649923e-05,
+      "loss": 0.7443,
+      "step": 996
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6583787870996447,
+      "learning_rate": 1.565149136577708e-05,
+      "loss": 0.7278,
+      "step": 997
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9108235713071573,
+      "learning_rate": 1.566718995290424e-05,
+      "loss": 0.746,
+      "step": 998
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6150111792301183,
+      "learning_rate": 1.5682888540031398e-05,
+      "loss": 0.5492,
+      "step": 999
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6281168556967098,
+      "learning_rate": 1.5698587127158558e-05,
+      "loss": 0.7564,
+      "step": 1000
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5315494508077223,
+      "learning_rate": 1.5714285714285715e-05,
+      "loss": 0.7455,
+      "step": 1001
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.501032107203606,
+      "learning_rate": 1.5729984301412876e-05,
+      "loss": 0.7376,
+      "step": 1002
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5748429672279617,
+      "learning_rate": 1.5745682888540033e-05,
+      "loss": 0.479,
+      "step": 1003
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6237551198466287,
+      "learning_rate": 1.5761381475667194e-05,
+      "loss": 0.6898,
+      "step": 1004
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6018126935369136,
+      "learning_rate": 1.577708006279435e-05,
+      "loss": 0.5183,
+      "step": 1005
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.817406976473913,
+      "learning_rate": 1.5792778649921508e-05,
+      "loss": 0.7205,
+      "step": 1006
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.360003014549453,
+      "learning_rate": 1.580847723704867e-05,
+      "loss": 0.6834,
+      "step": 1007
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6580197523339075,
+      "learning_rate": 1.5824175824175826e-05,
+      "loss": 0.8667,
+      "step": 1008
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4313332151929319,
+      "learning_rate": 1.5839874411302983e-05,
+      "loss": 0.6974,
+      "step": 1009
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8203647960411502,
+      "learning_rate": 1.5855572998430144e-05,
+      "loss": 0.801,
+      "step": 1010
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.5971772786189,
+      "learning_rate": 1.58712715855573e-05,
+      "loss": 0.8432,
+      "step": 1011
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5602585800358102,
+      "learning_rate": 1.5886970172684458e-05,
+      "loss": 0.6915,
+      "step": 1012
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4723577612897523,
+      "learning_rate": 1.590266875981162e-05,
+      "loss": 0.7312,
+      "step": 1013
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5753895308623693,
+      "learning_rate": 1.5918367346938776e-05,
+      "loss": 0.7326,
+      "step": 1014
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9571547643258476,
+      "learning_rate": 1.5934065934065933e-05,
+      "loss": 0.7114,
+      "step": 1015
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7046141774414554,
+      "learning_rate": 1.5949764521193094e-05,
+      "loss": 0.6661,
+      "step": 1016
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4841669989499338,
+      "learning_rate": 1.596546310832025e-05,
+      "loss": 0.7794,
+      "step": 1017
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4681244319921118,
+      "learning_rate": 1.598116169544741e-05,
+      "loss": 0.7779,
+      "step": 1018
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5566535324183173,
+      "learning_rate": 1.599686028257457e-05,
+      "loss": 0.8776,
+      "step": 1019
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.46758457142252,
+      "learning_rate": 1.601255886970173e-05,
+      "loss": 0.7964,
+      "step": 1020
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4726952556839237,
+      "learning_rate": 1.6028257456828886e-05,
+      "loss": 0.7469,
+      "step": 1021
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3143552527990638,
+      "learning_rate": 1.6043956043956047e-05,
+      "loss": 0.6988,
+      "step": 1022
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4619903457353507,
+      "learning_rate": 1.6059654631083204e-05,
+      "loss": 0.7672,
+      "step": 1023
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.424437062722666,
+      "learning_rate": 1.6075353218210365e-05,
+      "loss": 0.7802,
+      "step": 1024
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5522972873052088,
+      "learning_rate": 1.6091051805337522e-05,
+      "loss": 0.7821,
+      "step": 1025
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.622713444447618,
+      "learning_rate": 1.610675039246468e-05,
+      "loss": 0.8207,
+      "step": 1026
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3979004990671828,
+      "learning_rate": 1.612244897959184e-05,
+      "loss": 0.8259,
+      "step": 1027
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3853033219891495,
+      "learning_rate": 1.6138147566718997e-05,
+      "loss": 0.7784,
+      "step": 1028
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3890182859599631,
+      "learning_rate": 1.6153846153846154e-05,
+      "loss": 0.7537,
+      "step": 1029
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.392610259484936,
+      "learning_rate": 1.6169544740973315e-05,
+      "loss": 0.7252,
+      "step": 1030
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.489281759147729,
+      "learning_rate": 1.6185243328100472e-05,
+      "loss": 0.7574,
+      "step": 1031
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6572627871606254,
+      "learning_rate": 1.620094191522763e-05,
+      "loss": 0.7929,
+      "step": 1032
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6114440285114104,
+      "learning_rate": 1.621664050235479e-05,
+      "loss": 0.7477,
+      "step": 1033
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.525150868873773,
+      "learning_rate": 1.6232339089481947e-05,
+      "loss": 0.758,
+      "step": 1034
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6690019158097437,
+      "learning_rate": 1.6248037676609107e-05,
+      "loss": 0.4975,
+      "step": 1035
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7730965973956834,
+      "learning_rate": 1.6263736263736265e-05,
+      "loss": 0.6612,
+      "step": 1036
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.42334850849071,
+      "learning_rate": 1.6279434850863422e-05,
+      "loss": 0.663,
+      "step": 1037
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9924860473028263,
+      "learning_rate": 1.6295133437990582e-05,
+      "loss": 0.6966,
+      "step": 1038
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.58422548492478,
+      "learning_rate": 1.631083202511774e-05,
+      "loss": 0.737,
+      "step": 1039
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6115735012041617,
+      "learning_rate": 1.63265306122449e-05,
+      "loss": 0.734,
+      "step": 1040
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4122577583670308,
+      "learning_rate": 1.6342229199372057e-05,
+      "loss": 0.7624,
+      "step": 1041
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.505525477020911,
+      "learning_rate": 1.6357927786499218e-05,
+      "loss": 0.7049,
+      "step": 1042
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8137604545345505,
+      "learning_rate": 1.6373626373626375e-05,
+      "loss": 0.7886,
+      "step": 1043
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5271305267297786,
+      "learning_rate": 1.6389324960753536e-05,
+      "loss": 0.7668,
+      "step": 1044
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5034309804321726,
+      "learning_rate": 1.6405023547880693e-05,
+      "loss": 0.8032,
+      "step": 1045
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4796430704792087,
+      "learning_rate": 1.642072213500785e-05,
+      "loss": 0.7869,
+      "step": 1046
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6136220424833059,
+      "learning_rate": 1.643642072213501e-05,
+      "loss": 0.8121,
+      "step": 1047
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.606768122682228,
+      "learning_rate": 1.6452119309262168e-05,
+      "loss": 0.7345,
+      "step": 1048
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.481239404710713,
+      "learning_rate": 1.6467817896389325e-05,
+      "loss": 0.7022,
+      "step": 1049
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3449410799394892,
+      "learning_rate": 1.6483516483516486e-05,
+      "loss": 0.7987,
+      "step": 1050
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4126120455991684,
+      "learning_rate": 1.6499215070643643e-05,
+      "loss": 0.7996,
+      "step": 1051
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5128260711853516,
+      "learning_rate": 1.65149136577708e-05,
+      "loss": 0.8301,
+      "step": 1052
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.071370736051152,
+      "learning_rate": 1.653061224489796e-05,
+      "loss": 0.6316,
+      "step": 1053
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5124458004815302,
+      "learning_rate": 1.6546310832025118e-05,
+      "loss": 0.6801,
+      "step": 1054
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.457346763926452,
+      "learning_rate": 1.6562009419152278e-05,
+      "loss": 0.7834,
+      "step": 1055
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.478628904592789,
+      "learning_rate": 1.6577708006279435e-05,
+      "loss": 0.6922,
+      "step": 1056
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.584060506349512,
+      "learning_rate": 1.6593406593406596e-05,
+      "loss": 0.5047,
+      "step": 1057
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4029261586703214,
+      "learning_rate": 1.6609105180533753e-05,
+      "loss": 0.78,
+      "step": 1058
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3187273851435524,
+      "learning_rate": 1.662480376766091e-05,
+      "loss": 0.7166,
+      "step": 1059
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3976618624756951,
+      "learning_rate": 1.664050235478807e-05,
+      "loss": 0.7414,
+      "step": 1060
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6505999453985396,
+      "learning_rate": 1.6656200941915228e-05,
+      "loss": 0.7583,
+      "step": 1061
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3353506866867424,
+      "learning_rate": 1.667189952904239e-05,
+      "loss": 0.7886,
+      "step": 1062
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5021827852398113,
+      "learning_rate": 1.6687598116169546e-05,
+      "loss": 0.7687,
+      "step": 1063
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.398991115152886,
+      "learning_rate": 1.6703296703296707e-05,
+      "loss": 0.8817,
+      "step": 1064
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.593095297108636,
+      "learning_rate": 1.6718995290423864e-05,
+      "loss": 0.5293,
+      "step": 1065
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5214179503136924,
+      "learning_rate": 1.673469387755102e-05,
+      "loss": 0.6341,
+      "step": 1066
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7195833731086478,
+      "learning_rate": 1.675039246467818e-05,
+      "loss": 0.6729,
+      "step": 1067
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9927782660481383,
+      "learning_rate": 1.676609105180534e-05,
+      "loss": 0.5933,
+      "step": 1068
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6425098992083326,
+      "learning_rate": 1.6781789638932496e-05,
+      "loss": 0.541,
+      "step": 1069
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5534670275373705,
+      "learning_rate": 1.6797488226059656e-05,
+      "loss": 0.6642,
+      "step": 1070
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.832007657457541,
+      "learning_rate": 1.6813186813186814e-05,
+      "loss": 0.7248,
+      "step": 1071
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0121874306494645,
+      "learning_rate": 1.682888540031397e-05,
+      "loss": 0.7428,
+      "step": 1072
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1085370515027924,
+      "learning_rate": 1.684458398744113e-05,
+      "loss": 0.6857,
+      "step": 1073
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8072074855755627,
+      "learning_rate": 1.686028257456829e-05,
+      "loss": 0.7939,
+      "step": 1074
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5691567480948798,
+      "learning_rate": 1.687598116169545e-05,
+      "loss": 0.7215,
+      "step": 1075
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.749584868771914,
+      "learning_rate": 1.6891679748822606e-05,
+      "loss": 0.6192,
+      "step": 1076
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3658140252494326,
+      "learning_rate": 1.6907378335949767e-05,
+      "loss": 0.6941,
+      "step": 1077
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.420048084532095,
+      "learning_rate": 1.6923076923076924e-05,
+      "loss": 0.6273,
+      "step": 1078
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3454004766069028,
+      "learning_rate": 1.6938775510204085e-05,
+      "loss": 0.672,
+      "step": 1079
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5242690892731705,
+      "learning_rate": 1.6954474097331242e-05,
+      "loss": 0.7457,
+      "step": 1080
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4347980549993973,
+      "learning_rate": 1.69701726844584e-05,
+      "loss": 0.8198,
+      "step": 1081
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6651665185062567,
+      "learning_rate": 1.698587127158556e-05,
+      "loss": 0.5509,
+      "step": 1082
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5154931162394685,
+      "learning_rate": 1.7001569858712717e-05,
+      "loss": 0.828,
+      "step": 1083
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.40924728947278,
+      "learning_rate": 1.7017268445839878e-05,
+      "loss": 0.6997,
+      "step": 1084
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5212026069734088,
+      "learning_rate": 1.7032967032967035e-05,
+      "loss": 0.6616,
+      "step": 1085
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5988444657077807,
+      "learning_rate": 1.7048665620094192e-05,
+      "loss": 0.5045,
+      "step": 1086
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5425879866403476,
+      "learning_rate": 1.7064364207221352e-05,
+      "loss": 0.6269,
+      "step": 1087
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7203182580670262,
+      "learning_rate": 1.708006279434851e-05,
+      "loss": 0.764,
+      "step": 1088
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5203589075232602,
+      "learning_rate": 1.7095761381475667e-05,
+      "loss": 0.6353,
+      "step": 1089
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5081698174351417,
+      "learning_rate": 1.7111459968602827e-05,
+      "loss": 0.6593,
+      "step": 1090
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5961877626133592,
+      "learning_rate": 1.7127158555729985e-05,
+      "loss": 0.7344,
+      "step": 1091
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6760594203033743,
+      "learning_rate": 1.7142857142857142e-05,
+      "loss": 0.7588,
+      "step": 1092
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4247199374899888,
+      "learning_rate": 1.7158555729984302e-05,
+      "loss": 0.7476,
+      "step": 1093
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8034020363751804,
+      "learning_rate": 1.717425431711146e-05,
+      "loss": 0.7132,
+      "step": 1094
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5836609170475697,
+      "learning_rate": 1.718995290423862e-05,
+      "loss": 0.7661,
+      "step": 1095
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6380910387929262,
+      "learning_rate": 1.7205651491365777e-05,
+      "loss": 0.8658,
+      "step": 1096
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.529386913538966,
+      "learning_rate": 1.7221350078492938e-05,
+      "loss": 0.7327,
+      "step": 1097
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.122833280833028,
+      "learning_rate": 1.7237048665620095e-05,
+      "loss": 0.7581,
+      "step": 1098
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.548998343570299,
+      "learning_rate": 1.7252747252747256e-05,
+      "loss": 0.8021,
+      "step": 1099
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4294725459700095,
+      "learning_rate": 1.7268445839874413e-05,
+      "loss": 0.9197,
+      "step": 1100
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9112532702771938,
+      "learning_rate": 1.7284144427001574e-05,
+      "loss": 0.7443,
+      "step": 1101
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2706468096630033,
+      "learning_rate": 1.729984301412873e-05,
+      "loss": 0.7423,
+      "step": 1102
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4648307558317821,
+      "learning_rate": 1.7315541601255888e-05,
+      "loss": 0.654,
+      "step": 1103
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6156476048198098,
+      "learning_rate": 1.733124018838305e-05,
+      "loss": 0.918,
+      "step": 1104
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6826015997407895,
+      "learning_rate": 1.7346938775510206e-05,
+      "loss": 0.7782,
+      "step": 1105
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5798882426800521,
+      "learning_rate": 1.7362637362637363e-05,
+      "loss": 0.4645,
+      "step": 1106
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7071253697971647,
+      "learning_rate": 1.7378335949764523e-05,
+      "loss": 0.789,
+      "step": 1107
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4368831724245843,
+      "learning_rate": 1.739403453689168e-05,
+      "loss": 0.7697,
+      "step": 1108
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6614949457583643,
+      "learning_rate": 1.7409733124018838e-05,
+      "loss": 0.8165,
+      "step": 1109
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4945641940958025,
+      "learning_rate": 1.7425431711146e-05,
+      "loss": 0.7675,
+      "step": 1110
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4538076072829849,
+      "learning_rate": 1.7441130298273156e-05,
+      "loss": 0.7672,
+      "step": 1111
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3857600157583374,
+      "learning_rate": 1.7456828885400316e-05,
+      "loss": 0.6624,
+      "step": 1112
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9752862101316093,
+      "learning_rate": 1.7472527472527473e-05,
+      "loss": 0.7634,
+      "step": 1113
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6005182874441986,
+      "learning_rate": 1.748822605965463e-05,
+      "loss": 0.8144,
+      "step": 1114
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6247064225168865,
+      "learning_rate": 1.750392464678179e-05,
+      "loss": 0.5339,
+      "step": 1115
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4290689765547129,
+      "learning_rate": 1.751962323390895e-05,
+      "loss": 0.8616,
+      "step": 1116
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.821939601734514,
+      "learning_rate": 1.753532182103611e-05,
+      "loss": 0.7362,
+      "step": 1117
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.894059434050511,
+      "learning_rate": 1.7551020408163266e-05,
+      "loss": 0.5805,
+      "step": 1118
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.947746456424682,
+      "learning_rate": 1.7566718995290427e-05,
+      "loss": 0.7701,
+      "step": 1119
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.515813174507109,
+      "learning_rate": 1.7582417582417584e-05,
+      "loss": 0.6384,
+      "step": 1120
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4249663702851598,
+      "learning_rate": 1.7598116169544744e-05,
+      "loss": 0.8271,
+      "step": 1121
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4985998277294188,
+      "learning_rate": 1.76138147566719e-05,
+      "loss": 0.7036,
+      "step": 1122
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6169508615185952,
+      "learning_rate": 1.762951334379906e-05,
+      "loss": 0.4937,
+      "step": 1123
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5007368768477265,
+      "learning_rate": 1.764521193092622e-05,
+      "loss": 0.7833,
+      "step": 1124
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.552605784398588,
+      "learning_rate": 1.7660910518053377e-05,
+      "loss": 0.5161,
+      "step": 1125
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7689733677379529,
+      "learning_rate": 1.7676609105180537e-05,
+      "loss": 0.8189,
+      "step": 1126
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6310579654691697,
+      "learning_rate": 1.7692307692307694e-05,
+      "loss": 0.7954,
+      "step": 1127
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5601084655471488,
+      "learning_rate": 1.770800627943485e-05,
+      "loss": 0.492,
+      "step": 1128
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5495600517910595,
+      "learning_rate": 1.772370486656201e-05,
+      "loss": 0.6616,
+      "step": 1129
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3905977231883133,
+      "learning_rate": 1.773940345368917e-05,
+      "loss": 0.7497,
+      "step": 1130
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5527908604134906,
+      "learning_rate": 1.7755102040816327e-05,
+      "loss": 0.7659,
+      "step": 1131
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.490507725793383,
+      "learning_rate": 1.7770800627943487e-05,
+      "loss": 0.6346,
+      "step": 1132
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3198778651036185,
+      "learning_rate": 1.7786499215070644e-05,
+      "loss": 0.7638,
+      "step": 1133
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7760609582080311,
+      "learning_rate": 1.78021978021978e-05,
+      "loss": 0.6407,
+      "step": 1134
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.607162719321703,
+      "learning_rate": 1.7817896389324962e-05,
+      "loss": 0.8205,
+      "step": 1135
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6068226734324331,
+      "learning_rate": 1.783359497645212e-05,
+      "loss": 0.4956,
+      "step": 1136
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3863040780490747,
+      "learning_rate": 1.784929356357928e-05,
+      "loss": 0.8791,
+      "step": 1137
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.317687407884226,
+      "learning_rate": 1.7864992150706437e-05,
+      "loss": 0.7407,
+      "step": 1138
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4813499402401509,
+      "learning_rate": 1.7880690737833598e-05,
+      "loss": 0.6415,
+      "step": 1139
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5892416009475914,
+      "learning_rate": 1.7896389324960755e-05,
+      "loss": 0.7523,
+      "step": 1140
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.121931933103499,
+      "learning_rate": 1.7912087912087915e-05,
+      "loss": 0.7446,
+      "step": 1141
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5739630593947364,
+      "learning_rate": 1.7927786499215073e-05,
+      "loss": 0.4707,
+      "step": 1142
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.911279730535706,
+      "learning_rate": 1.7943485086342233e-05,
+      "loss": 0.6997,
+      "step": 1143
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4361734168456906,
+      "learning_rate": 1.795918367346939e-05,
+      "loss": 0.6165,
+      "step": 1144
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5286542339805502,
+      "learning_rate": 1.7974882260596548e-05,
+      "loss": 0.6711,
+      "step": 1145
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7829903587907292,
+      "learning_rate": 1.7990580847723708e-05,
+      "loss": 0.6629,
+      "step": 1146
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4586189206008298,
+      "learning_rate": 1.8006279434850865e-05,
+      "loss": 0.7637,
+      "step": 1147
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7335437996788186,
+      "learning_rate": 1.8021978021978023e-05,
+      "loss": 0.8582,
+      "step": 1148
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5215200008395566,
+      "learning_rate": 1.803767660910518e-05,
+      "loss": 0.8172,
+      "step": 1149
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7930150977097785,
+      "learning_rate": 1.805337519623234e-05,
+      "loss": 0.7991,
+      "step": 1150
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5725526783756025,
+      "learning_rate": 1.8069073783359498e-05,
+      "loss": 0.6957,
+      "step": 1151
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3552523378249306,
+      "learning_rate": 1.8084772370486658e-05,
+      "loss": 0.6972,
+      "step": 1152
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8127491924909074,
+      "learning_rate": 1.8100470957613815e-05,
+      "loss": 0.9389,
+      "step": 1153
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4847352320939526,
+      "learning_rate": 1.8116169544740976e-05,
+      "loss": 0.6325,
+      "step": 1154
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7177913506185438,
+      "learning_rate": 1.8131868131868133e-05,
+      "loss": 0.7684,
+      "step": 1155
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3919260592521636,
+      "learning_rate": 1.814756671899529e-05,
+      "loss": 0.67,
+      "step": 1156
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4887031601843081,
+      "learning_rate": 1.816326530612245e-05,
+      "loss": 0.8577,
+      "step": 1157
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6689821066537214,
+      "learning_rate": 1.8178963893249608e-05,
+      "loss": 0.7605,
+      "step": 1158
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4650361085397756,
+      "learning_rate": 1.819466248037677e-05,
+      "loss": 0.895,
+      "step": 1159
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1327438651092154,
+      "learning_rate": 1.8210361067503926e-05,
+      "loss": 0.97,
+      "step": 1160
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.596185219835646,
+      "learning_rate": 1.8226059654631086e-05,
+      "loss": 0.6596,
+      "step": 1161
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4945135625862183,
+      "learning_rate": 1.8241758241758244e-05,
+      "loss": 0.7999,
+      "step": 1162
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5849447504778251,
+      "learning_rate": 1.8257456828885404e-05,
+      "loss": 0.7848,
+      "step": 1163
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4155353818422105,
+      "learning_rate": 1.827315541601256e-05,
+      "loss": 0.7374,
+      "step": 1164
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.433995307339946,
+      "learning_rate": 1.828885400313972e-05,
+      "loss": 0.7361,
+      "step": 1165
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5096462703183637,
+      "learning_rate": 1.830455259026688e-05,
+      "loss": 0.8186,
+      "step": 1166
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4084771486736793,
+      "learning_rate": 1.8320251177394036e-05,
+      "loss": 0.7443,
+      "step": 1167
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8188814491134573,
+      "learning_rate": 1.8335949764521194e-05,
+      "loss": 0.6592,
+      "step": 1168
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4818139079356238,
+      "learning_rate": 1.835164835164835e-05,
+      "loss": 0.7408,
+      "step": 1169
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4955781998045323,
+      "learning_rate": 1.836734693877551e-05,
+      "loss": 0.7613,
+      "step": 1170
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7734684659180668,
+      "learning_rate": 1.838304552590267e-05,
+      "loss": 0.6181,
+      "step": 1171
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7689166689911617,
+      "learning_rate": 1.839874411302983e-05,
+      "loss": 0.7221,
+      "step": 1172
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6896056329225384,
+      "learning_rate": 1.8414442700156986e-05,
+      "loss": 0.7282,
+      "step": 1173
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.25371504319865,
+      "learning_rate": 1.8430141287284147e-05,
+      "loss": 0.5522,
+      "step": 1174
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.116514314054544,
+      "learning_rate": 1.8445839874411304e-05,
+      "loss": 0.6956,
+      "step": 1175
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4659177143523827,
+      "learning_rate": 1.8461538461538465e-05,
+      "loss": 0.6381,
+      "step": 1176
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4320542289906124,
+      "learning_rate": 1.8477237048665622e-05,
+      "loss": 0.7744,
+      "step": 1177
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9998460098375548,
+      "learning_rate": 1.849293563579278e-05,
+      "loss": 0.6987,
+      "step": 1178
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3757927428013066,
+      "learning_rate": 1.850863422291994e-05,
+      "loss": 0.774,
+      "step": 1179
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5048501877316356,
+      "learning_rate": 1.8524332810047097e-05,
+      "loss": 0.6881,
+      "step": 1180
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7163243576054352,
+      "learning_rate": 1.8540031397174257e-05,
+      "loss": 0.6617,
+      "step": 1181
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.470859609688909,
+      "learning_rate": 1.8555729984301415e-05,
+      "loss": 0.68,
+      "step": 1182
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7570784504462051,
+      "learning_rate": 1.8571428571428575e-05,
+      "loss": 0.6748,
+      "step": 1183
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.830135882205963,
+      "learning_rate": 1.8587127158555732e-05,
+      "loss": 0.8467,
+      "step": 1184
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3952220544569824,
+      "learning_rate": 1.860282574568289e-05,
+      "loss": 0.8516,
+      "step": 1185
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6392279135233876,
+      "learning_rate": 1.861852433281005e-05,
+      "loss": 0.692,
+      "step": 1186
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4390109616144025,
+      "learning_rate": 1.8634222919937207e-05,
+      "loss": 0.7146,
+      "step": 1187
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5640201299422138,
+      "learning_rate": 1.8649921507064364e-05,
+      "loss": 0.6659,
+      "step": 1188
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3596558116072102,
+      "learning_rate": 1.866562009419152e-05,
+      "loss": 0.7376,
+      "step": 1189
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5919492790955243,
+      "learning_rate": 1.8681318681318682e-05,
+      "loss": 0.4808,
+      "step": 1190
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5651552641684678,
+      "learning_rate": 1.869701726844584e-05,
+      "loss": 0.6892,
+      "step": 1191
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5327648643897205,
+      "learning_rate": 1.8712715855573e-05,
+      "loss": 0.7712,
+      "step": 1192
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5770538371757781,
+      "learning_rate": 1.8728414442700157e-05,
+      "loss": 0.5118,
+      "step": 1193
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.533868793143932,
+      "learning_rate": 1.8744113029827318e-05,
+      "loss": 0.7166,
+      "step": 1194
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9197666405768055,
+      "learning_rate": 1.8759811616954475e-05,
+      "loss": 0.7362,
+      "step": 1195
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4740240815996397,
+      "learning_rate": 1.8775510204081636e-05,
+      "loss": 0.7345,
+      "step": 1196
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6342561712440614,
+      "learning_rate": 1.8791208791208793e-05,
+      "loss": 0.5749,
+      "step": 1197
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5585243178280155,
+      "learning_rate": 1.8806907378335953e-05,
+      "loss": 0.786,
+      "step": 1198
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.764630155931772,
+      "learning_rate": 1.882260596546311e-05,
+      "loss": 0.7271,
+      "step": 1199
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4123759613091642,
+      "learning_rate": 1.8838304552590268e-05,
+      "loss": 0.6995,
+      "step": 1200
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5940476504927952,
+      "learning_rate": 1.885400313971743e-05,
+      "loss": 0.6867,
+      "step": 1201
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8160694847202488,
+      "learning_rate": 1.8869701726844586e-05,
+      "loss": 0.8251,
+      "step": 1202
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4650564793456808,
+      "learning_rate": 1.8885400313971746e-05,
+      "loss": 0.7538,
+      "step": 1203
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8023055244467596,
+      "learning_rate": 1.8901098901098903e-05,
+      "loss": 0.885,
+      "step": 1204
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1997673908482502,
+      "learning_rate": 1.891679748822606e-05,
+      "loss": 0.6386,
+      "step": 1205
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4499300503400732,
+      "learning_rate": 1.893249607535322e-05,
+      "loss": 0.6856,
+      "step": 1206
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3262478477548474,
+      "learning_rate": 1.8948194662480378e-05,
+      "loss": 0.6361,
+      "step": 1207
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9746594258258972,
+      "learning_rate": 1.8963893249607535e-05,
+      "loss": 0.8672,
+      "step": 1208
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6211214085543995,
+      "learning_rate": 1.8979591836734696e-05,
+      "loss": 0.8023,
+      "step": 1209
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5285654005919906,
+      "learning_rate": 1.8995290423861853e-05,
+      "loss": 0.7409,
+      "step": 1210
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5657061080524908,
+      "learning_rate": 1.901098901098901e-05,
+      "loss": 0.8466,
+      "step": 1211
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3735347046488613,
+      "learning_rate": 1.902668759811617e-05,
+      "loss": 0.8455,
+      "step": 1212
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3503873920958882,
+      "learning_rate": 1.9042386185243328e-05,
+      "loss": 0.7467,
+      "step": 1213
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5043471402840207,
+      "learning_rate": 1.905808477237049e-05,
+      "loss": 0.7413,
+      "step": 1214
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.497311936481248,
+      "learning_rate": 1.9073783359497646e-05,
+      "loss": 0.7808,
+      "step": 1215
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7235278071420492,
+      "learning_rate": 1.9089481946624807e-05,
+      "loss": 0.7213,
+      "step": 1216
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.469634265314476,
+      "learning_rate": 1.9105180533751964e-05,
+      "loss": 0.7628,
+      "step": 1217
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5266169855418505,
+      "learning_rate": 1.9120879120879124e-05,
+      "loss": 0.7972,
+      "step": 1218
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4188727918127113,
+      "learning_rate": 1.913657770800628e-05,
+      "loss": 0.6972,
+      "step": 1219
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.537930815051496,
+      "learning_rate": 1.9152276295133442e-05,
+      "loss": 0.9613,
+      "step": 1220
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6263903223869274,
+      "learning_rate": 1.91679748822606e-05,
+      "loss": 0.664,
+      "step": 1221
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4541058625844745,
+      "learning_rate": 1.9183673469387756e-05,
+      "loss": 0.8032,
+      "step": 1222
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7976400742462697,
+      "learning_rate": 1.9199372056514917e-05,
+      "loss": 0.7795,
+      "step": 1223
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6456471015983516,
+      "learning_rate": 1.9215070643642074e-05,
+      "loss": 0.8086,
+      "step": 1224
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.596556639204615,
+      "learning_rate": 1.923076923076923e-05,
+      "loss": 0.5137,
+      "step": 1225
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6313448581386827,
+      "learning_rate": 1.9246467817896392e-05,
+      "loss": 0.926,
+      "step": 1226
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5799441556066895,
+      "learning_rate": 1.926216640502355e-05,
+      "loss": 0.5685,
+      "step": 1227
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.46333192771827,
+      "learning_rate": 1.9277864992150706e-05,
+      "loss": 0.712,
+      "step": 1228
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6258747834792128,
+      "learning_rate": 1.9293563579277867e-05,
+      "loss": 0.5097,
+      "step": 1229
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6457042639346515,
+      "learning_rate": 1.9309262166405024e-05,
+      "loss": 0.7042,
+      "step": 1230
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6764661024491467,
+      "learning_rate": 1.932496075353218e-05,
+      "loss": 0.7468,
+      "step": 1231
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5622947702817063,
+      "learning_rate": 1.9340659340659342e-05,
+      "loss": 0.7571,
+      "step": 1232
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0939247909283196,
+      "learning_rate": 1.93563579277865e-05,
+      "loss": 0.7398,
+      "step": 1233
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7434442734863185,
+      "learning_rate": 1.937205651491366e-05,
+      "loss": 0.7013,
+      "step": 1234
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3415893491890754,
+      "learning_rate": 1.9387755102040817e-05,
+      "loss": 0.7374,
+      "step": 1235
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4511923400964408,
+      "learning_rate": 1.9403453689167977e-05,
+      "loss": 0.7307,
+      "step": 1236
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4676596150233527,
+      "learning_rate": 1.9419152276295135e-05,
+      "loss": 0.8492,
+      "step": 1237
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4785717341397075,
+      "learning_rate": 1.9434850863422295e-05,
+      "loss": 0.8019,
+      "step": 1238
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6108839159356105,
+      "learning_rate": 1.9450549450549452e-05,
+      "loss": 0.4655,
+      "step": 1239
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0119213310976862,
+      "learning_rate": 1.9466248037676613e-05,
+      "loss": 0.736,
+      "step": 1240
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.655444952557443,
+      "learning_rate": 1.948194662480377e-05,
+      "loss": 0.8396,
+      "step": 1241
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4201860616004327,
+      "learning_rate": 1.9497645211930927e-05,
+      "loss": 0.7253,
+      "step": 1242
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4190331543209167,
+      "learning_rate": 1.9513343799058088e-05,
+      "loss": 0.7227,
+      "step": 1243
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5974718884300849,
+      "learning_rate": 1.9529042386185245e-05,
+      "loss": 0.5193,
+      "step": 1244
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4997916406070853,
+      "learning_rate": 1.9544740973312402e-05,
+      "loss": 0.8278,
+      "step": 1245
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5118156788113464,
+      "learning_rate": 1.9560439560439563e-05,
+      "loss": 0.722,
+      "step": 1246
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3950245221860362,
+      "learning_rate": 1.957613814756672e-05,
+      "loss": 0.6687,
+      "step": 1247
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4294354189860692,
+      "learning_rate": 1.9591836734693877e-05,
+      "loss": 0.8318,
+      "step": 1248
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.650628463627366,
+      "learning_rate": 1.9607535321821038e-05,
+      "loss": 0.7867,
+      "step": 1249
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.52434194309722,
+      "learning_rate": 1.9623233908948195e-05,
+      "loss": 0.6401,
+      "step": 1250
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5654350414953256,
+      "learning_rate": 1.9638932496075356e-05,
+      "loss": 0.4887,
+      "step": 1251
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8312435969337766,
+      "learning_rate": 1.9654631083202513e-05,
+      "loss": 0.8137,
+      "step": 1252
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5103172321402412,
+      "learning_rate": 1.967032967032967e-05,
+      "loss": 0.7125,
+      "step": 1253
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.468502859192207,
+      "learning_rate": 1.968602825745683e-05,
+      "loss": 0.6349,
+      "step": 1254
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8693064652313627,
+      "learning_rate": 1.9701726844583988e-05,
+      "loss": 0.7937,
+      "step": 1255
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9434973015359123,
+      "learning_rate": 1.971742543171115e-05,
+      "loss": 0.6777,
+      "step": 1256
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4315951214466018,
+      "learning_rate": 1.9733124018838306e-05,
+      "loss": 0.7194,
+      "step": 1257
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7522424054340857,
+      "learning_rate": 1.9748822605965466e-05,
+      "loss": 0.7553,
+      "step": 1258
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.705462494186174,
+      "learning_rate": 1.9764521193092623e-05,
+      "loss": 0.7121,
+      "step": 1259
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6383889730097838,
+      "learning_rate": 1.9780219780219784e-05,
+      "loss": 0.8902,
+      "step": 1260
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7388030362529945,
+      "learning_rate": 1.979591836734694e-05,
+      "loss": 0.7369,
+      "step": 1261
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6100563731671413,
+      "learning_rate": 1.98116169544741e-05,
+      "loss": 0.5525,
+      "step": 1262
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.578153502178434,
+      "learning_rate": 1.982731554160126e-05,
+      "loss": 0.7972,
+      "step": 1263
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.081380426283973,
+      "learning_rate": 1.9843014128728416e-05,
+      "loss": 0.801,
+      "step": 1264
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4727211873149324,
+      "learning_rate": 1.9858712715855573e-05,
+      "loss": 0.7045,
+      "step": 1265
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4477628275602836,
+      "learning_rate": 1.9874411302982734e-05,
+      "loss": 0.7939,
+      "step": 1266
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.613469546897564,
+      "learning_rate": 1.989010989010989e-05,
+      "loss": 0.7825,
+      "step": 1267
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.624804221140527,
+      "learning_rate": 1.990580847723705e-05,
+      "loss": 0.4981,
+      "step": 1268
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.564729257498145,
+      "learning_rate": 1.992150706436421e-05,
+      "loss": 0.7811,
+      "step": 1269
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.340373784760729,
+      "learning_rate": 1.9937205651491366e-05,
+      "loss": 0.829,
+      "step": 1270
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4246317721444077,
+      "learning_rate": 1.9952904238618527e-05,
+      "loss": 0.7147,
+      "step": 1271
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5761481519892807,
+      "learning_rate": 1.9968602825745684e-05,
+      "loss": 0.5238,
+      "step": 1272
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3633490409246165,
+      "learning_rate": 1.9984301412872844e-05,
+      "loss": 0.7291,
+      "step": 1273
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.394214384165405,
+      "learning_rate": 2e-05,
+      "loss": 0.7729,
+      "step": 1274
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6345124697953026,
+      "learning_rate": 1.9999999970888418e-05,
+      "loss": 0.6905,
+      "step": 1275
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5064613480996045,
+      "learning_rate": 1.999999988355367e-05,
+      "loss": 0.8214,
+      "step": 1276
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7385805912917622,
+      "learning_rate": 1.9999999737995757e-05,
+      "loss": 0.7024,
+      "step": 1277
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6360395569180893,
+      "learning_rate": 1.9999999534214676e-05,
+      "loss": 0.5292,
+      "step": 1278
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6718070478617602,
+      "learning_rate": 1.9999999272210435e-05,
+      "loss": 0.6813,
+      "step": 1279
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7302487178686952,
+      "learning_rate": 1.9999998951983033e-05,
+      "loss": 0.6901,
+      "step": 1280
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.435338669959867,
+      "learning_rate": 1.999999857353247e-05,
+      "loss": 0.7477,
+      "step": 1281
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.85008383135011,
+      "learning_rate": 1.9999998136858748e-05,
+      "loss": 0.7373,
+      "step": 1282
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8062807448807854,
+      "learning_rate": 1.9999997641961873e-05,
+      "loss": 0.7186,
+      "step": 1283
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6408824329745244,
+      "learning_rate": 1.9999997088841844e-05,
+      "loss": 0.6814,
+      "step": 1284
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7261408902569224,
+      "learning_rate": 1.9999996477498667e-05,
+      "loss": 0.842,
+      "step": 1285
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5999403707329185,
+      "learning_rate": 1.9999995807932346e-05,
+      "loss": 0.7009,
+      "step": 1286
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4731921387671445,
+      "learning_rate": 1.999999508014288e-05,
+      "loss": 0.8152,
+      "step": 1287
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4604582716332202,
+      "learning_rate": 1.999999429413028e-05,
+      "loss": 0.7281,
+      "step": 1288
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.830738425503431,
+      "learning_rate": 1.9999993449894542e-05,
+      "loss": 0.5902,
+      "step": 1289
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5283181748817047,
+      "learning_rate": 1.9999992547435683e-05,
+      "loss": 0.8869,
+      "step": 1290
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5787687246816396,
+      "learning_rate": 1.99999915867537e-05,
+      "loss": 0.4671,
+      "step": 1291
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.8116100633567522,
+      "learning_rate": 1.9999990567848595e-05,
+      "loss": 0.7411,
+      "step": 1292
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5519817695971225,
+      "learning_rate": 1.9999989490720384e-05,
+      "loss": 0.8688,
+      "step": 1293
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.079552202747287,
+      "learning_rate": 1.9999988355369067e-05,
+      "loss": 0.6957,
+      "step": 1294
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.418567384728434,
+      "learning_rate": 1.999998716179465e-05,
+      "loss": 0.7797,
+      "step": 1295
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4205994347541666,
+      "learning_rate": 1.9999985909997145e-05,
+      "loss": 0.7588,
+      "step": 1296
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4232919200829204,
+      "learning_rate": 1.9999984599976555e-05,
+      "loss": 0.7091,
+      "step": 1297
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2990015524173117,
+      "learning_rate": 1.9999983231732888e-05,
+      "loss": 0.7678,
+      "step": 1298
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.847727321202081,
+      "learning_rate": 1.9999981805266152e-05,
+      "loss": 0.8811,
+      "step": 1299
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4394499664052816,
+      "learning_rate": 1.9999980320576356e-05,
+      "loss": 0.744,
+      "step": 1300
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6405769580770888,
+      "learning_rate": 1.999997877766351e-05,
+      "loss": 0.8807,
+      "step": 1301
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4061819930776978,
+      "learning_rate": 1.9999977176527625e-05,
+      "loss": 0.8679,
+      "step": 1302
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6005279470305736,
+      "learning_rate": 1.99999755171687e-05,
+      "loss": 0.9252,
+      "step": 1303
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.453441477957383,
+      "learning_rate": 1.9999973799586754e-05,
+      "loss": 0.6303,
+      "step": 1304
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5147798190582713,
+      "learning_rate": 1.9999972023781797e-05,
+      "loss": 0.7518,
+      "step": 1305
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7723227623079656,
+      "learning_rate": 1.999997018975383e-05,
+      "loss": 0.8964,
+      "step": 1306
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.109906776418713,
+      "learning_rate": 1.9999968297502878e-05,
+      "loss": 0.6934,
+      "step": 1307
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.477720669163886,
+      "learning_rate": 1.9999966347028943e-05,
+      "loss": 0.6665,
+      "step": 1308
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.414860593489007,
+      "learning_rate": 1.9999964338332037e-05,
+      "loss": 0.7746,
+      "step": 1309
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7429392451862808,
+      "learning_rate": 1.9999962271412173e-05,
+      "loss": 0.7,
+      "step": 1310
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5070986708454794,
+      "learning_rate": 1.9999960146269365e-05,
+      "loss": 0.8917,
+      "step": 1311
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5919225403446966,
+      "learning_rate": 1.999995796290362e-05,
+      "loss": 0.6568,
+      "step": 1312
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.811845431499546,
+      "learning_rate": 1.9999955721314954e-05,
+      "loss": 0.8796,
+      "step": 1313
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.646545352674897,
+      "learning_rate": 1.999995342150338e-05,
+      "loss": 0.6799,
+      "step": 1314
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5195651391912646,
+      "learning_rate": 1.999995106346891e-05,
+      "loss": 0.6941,
+      "step": 1315
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7279358894131314,
+      "learning_rate": 1.9999948647211562e-05,
+      "loss": 0.8013,
+      "step": 1316
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6228163024266681,
+      "learning_rate": 1.9999946172731347e-05,
+      "loss": 0.5415,
+      "step": 1317
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.423002783884237,
+      "learning_rate": 1.999994364002828e-05,
+      "loss": 0.7715,
+      "step": 1318
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5275754159420603,
+      "learning_rate": 1.999994104910237e-05,
+      "loss": 0.6695,
+      "step": 1319
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4418374150629094,
+      "learning_rate": 1.9999938399953645e-05,
+      "loss": 0.7801,
+      "step": 1320
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.015498339081336,
+      "learning_rate": 1.9999935692582107e-05,
+      "loss": 0.6429,
+      "step": 1321
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3333981998182727,
+      "learning_rate": 1.9999932926987778e-05,
+      "loss": 0.8167,
+      "step": 1322
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5717530097385755,
+      "learning_rate": 1.9999930103170676e-05,
+      "loss": 0.858,
+      "step": 1323
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6068828657756468,
+      "learning_rate": 1.9999927221130814e-05,
+      "loss": 0.5066,
+      "step": 1324
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6171652213777932,
+      "learning_rate": 1.9999924280868208e-05,
+      "loss": 0.7706,
+      "step": 1325
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.265890082097722,
+      "learning_rate": 1.999992128238288e-05,
+      "loss": 0.6781,
+      "step": 1326
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0507541538900784,
+      "learning_rate": 1.9999918225674843e-05,
+      "loss": 0.6874,
+      "step": 1327
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5946732090629099,
+      "learning_rate": 1.9999915110744116e-05,
+      "loss": 0.7165,
+      "step": 1328
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4243366182268398,
+      "learning_rate": 1.9999911937590715e-05,
+      "loss": 0.7751,
+      "step": 1329
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.763448668007695,
+      "learning_rate": 1.999990870621466e-05,
+      "loss": 0.8467,
+      "step": 1330
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.594867426016617,
+      "learning_rate": 1.9999905416615976e-05,
+      "loss": 0.7931,
+      "step": 1331
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4638314202302976,
+      "learning_rate": 1.9999902068794674e-05,
+      "loss": 0.5836,
+      "step": 1332
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.325853712855251,
+      "learning_rate": 1.9999898662750778e-05,
+      "loss": 0.7355,
+      "step": 1333
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5154613292801755,
+      "learning_rate": 1.9999895198484303e-05,
+      "loss": 0.7602,
+      "step": 1334
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.723369493883066,
+      "learning_rate": 1.9999891675995274e-05,
+      "loss": 0.6639,
+      "step": 1335
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4081828754887238,
+      "learning_rate": 1.9999888095283707e-05,
+      "loss": 0.697,
+      "step": 1336
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.418916100570024,
+      "learning_rate": 1.999988445634963e-05,
+      "loss": 0.8194,
+      "step": 1337
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.605314977063946,
+      "learning_rate": 1.999988075919306e-05,
+      "loss": 0.7928,
+      "step": 1338
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7299877732560378,
+      "learning_rate": 1.9999877003814015e-05,
+      "loss": 0.6794,
+      "step": 1339
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5710376742711573,
+      "learning_rate": 1.999987319021252e-05,
+      "loss": 0.4836,
+      "step": 1340
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.925823911315577,
+      "learning_rate": 1.99998693183886e-05,
+      "loss": 0.6052,
+      "step": 1341
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4571907892868063,
+      "learning_rate": 1.9999865388342274e-05,
+      "loss": 0.8497,
+      "step": 1342
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4244645654027737,
+      "learning_rate": 1.9999861400073564e-05,
+      "loss": 0.9191,
+      "step": 1343
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.409776863530714,
+      "learning_rate": 1.9999857353582496e-05,
+      "loss": 0.6539,
+      "step": 1344
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6716646895213633,
+      "learning_rate": 1.9999853248869094e-05,
+      "loss": 0.5445,
+      "step": 1345
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7060766513887413,
+      "learning_rate": 1.999984908593338e-05,
+      "loss": 0.6744,
+      "step": 1346
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.104261065200595,
+      "learning_rate": 1.999984486477538e-05,
+      "loss": 0.7755,
+      "step": 1347
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.524653405072395,
+      "learning_rate": 1.9999840585395113e-05,
+      "loss": 0.787,
+      "step": 1348
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6037475263556393,
+      "learning_rate": 1.9999836247792614e-05,
+      "loss": 0.7947,
+      "step": 1349
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.363425399201087,
+      "learning_rate": 1.99998318519679e-05,
+      "loss": 0.7278,
+      "step": 1350
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.520062789724974,
+      "learning_rate": 1.9999827397920993e-05,
+      "loss": 0.7907,
+      "step": 1351
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6382207892820753,
+      "learning_rate": 1.9999822885651935e-05,
+      "loss": 0.809,
+      "step": 1352
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6927601661811926,
+      "learning_rate": 1.9999818315160736e-05,
+      "loss": 0.7357,
+      "step": 1353
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7156727146110786,
+      "learning_rate": 1.999981368644743e-05,
+      "loss": 0.7066,
+      "step": 1354
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.024454441649188,
+      "learning_rate": 1.9999808999512044e-05,
+      "loss": 0.7027,
+      "step": 1355
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4453356256418994,
+      "learning_rate": 1.99998042543546e-05,
+      "loss": 0.8294,
+      "step": 1356
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7373472978488376,
+      "learning_rate": 1.9999799450975136e-05,
+      "loss": 0.6363,
+      "step": 1357
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.490202018154036,
+      "learning_rate": 1.999979458937367e-05,
+      "loss": 0.6064,
+      "step": 1358
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7548968322060685,
+      "learning_rate": 1.9999789669550235e-05,
+      "loss": 0.6727,
+      "step": 1359
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2242719658766954,
+      "learning_rate": 1.999978469150486e-05,
+      "loss": 0.684,
+      "step": 1360
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5859597977977211,
+      "learning_rate": 1.999977965523757e-05,
+      "loss": 0.7392,
+      "step": 1361
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9869643985438699,
+      "learning_rate": 1.9999774560748398e-05,
+      "loss": 0.7455,
+      "step": 1362
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5802426631962805,
+      "learning_rate": 1.9999769408037377e-05,
+      "loss": 0.7937,
+      "step": 1363
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7659239165593634,
+      "learning_rate": 1.9999764197104526e-05,
+      "loss": 0.8395,
+      "step": 1364
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5601642362158619,
+      "learning_rate": 1.9999758927949882e-05,
+      "loss": 0.7215,
+      "step": 1365
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8827359741652778,
+      "learning_rate": 1.999975360057348e-05,
+      "loss": 0.8422,
+      "step": 1366
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7294080773691072,
+      "learning_rate": 1.9999748214975345e-05,
+      "loss": 0.8098,
+      "step": 1367
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7643652991381633,
+      "learning_rate": 1.999974277115551e-05,
+      "loss": 0.7908,
+      "step": 1368
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7886120563806327,
+      "learning_rate": 1.9999737269114008e-05,
+      "loss": 0.7627,
+      "step": 1369
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7669684455965835,
+      "learning_rate": 1.9999731708850868e-05,
+      "loss": 0.7789,
+      "step": 1370
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.407671996062286,
+      "learning_rate": 1.9999726090366122e-05,
+      "loss": 0.7711,
+      "step": 1371
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.443283132465402,
+      "learning_rate": 1.9999720413659808e-05,
+      "loss": 0.7354,
+      "step": 1372
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6948152467933364,
+      "learning_rate": 1.9999714678731955e-05,
+      "loss": 0.7693,
+      "step": 1373
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.526448306989496,
+      "learning_rate": 1.99997088855826e-05,
+      "loss": 0.7011,
+      "step": 1374
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5999644958299654,
+      "learning_rate": 1.9999703034211772e-05,
+      "loss": 0.4838,
+      "step": 1375
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8471214960246687,
+      "learning_rate": 1.9999697124619505e-05,
+      "loss": 0.6266,
+      "step": 1376
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6784045200411974,
+      "learning_rate": 1.9999691156805836e-05,
+      "loss": 0.827,
+      "step": 1377
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7159657972585305,
+      "learning_rate": 1.99996851307708e-05,
+      "loss": 0.7523,
+      "step": 1378
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.610394074671609,
+      "learning_rate": 1.999967904651443e-05,
+      "loss": 0.7858,
+      "step": 1379
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5702122133543055,
+      "learning_rate": 1.9999672904036768e-05,
+      "loss": 0.8049,
+      "step": 1380
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5976263898159644,
+      "learning_rate": 1.9999666703337838e-05,
+      "loss": 0.5793,
+      "step": 1381
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5406770888630354,
+      "learning_rate": 1.9999660444417687e-05,
+      "loss": 0.7738,
+      "step": 1382
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6020926921531595,
+      "learning_rate": 1.9999654127276345e-05,
+      "loss": 0.5034,
+      "step": 1383
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.467228563612939,
+      "learning_rate": 1.999964775191385e-05,
+      "loss": 0.747,
+      "step": 1384
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5491919321931709,
+      "learning_rate": 1.999964131833024e-05,
+      "loss": 0.8958,
+      "step": 1385
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4574552432575523,
+      "learning_rate": 1.9999634826525555e-05,
+      "loss": 0.849,
+      "step": 1386
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5413502345592187,
+      "learning_rate": 1.999962827649983e-05,
+      "loss": 0.6137,
+      "step": 1387
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8487020481225849,
+      "learning_rate": 1.9999621668253103e-05,
+      "loss": 0.6903,
+      "step": 1388
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.695249438057142,
+      "learning_rate": 1.9999615001785412e-05,
+      "loss": 0.6493,
+      "step": 1389
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7047392458146797,
+      "learning_rate": 1.9999608277096796e-05,
+      "loss": 0.8208,
+      "step": 1390
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4282084297629078,
+      "learning_rate": 1.99996014941873e-05,
+      "loss": 0.8122,
+      "step": 1391
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6723097493660413,
+      "learning_rate": 1.9999594653056952e-05,
+      "loss": 0.7316,
+      "step": 1392
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.680575118682573,
+      "learning_rate": 1.99995877537058e-05,
+      "loss": 0.7441,
+      "step": 1393
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.396425095140688,
+      "learning_rate": 1.999958079613388e-05,
+      "loss": 0.6835,
+      "step": 1394
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5394551044854656,
+      "learning_rate": 1.9999573780341236e-05,
+      "loss": 0.7851,
+      "step": 1395
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.404220362996764,
+      "learning_rate": 1.9999566706327908e-05,
+      "loss": 0.6238,
+      "step": 1396
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4363590422362296,
+      "learning_rate": 1.999955957409394e-05,
+      "loss": 0.6373,
+      "step": 1397
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6051238635761058,
+      "learning_rate": 1.9999552383639368e-05,
+      "loss": 0.724,
+      "step": 1398
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8031108689623463,
+      "learning_rate": 1.9999545134964235e-05,
+      "loss": 0.857,
+      "step": 1399
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7688557788086587,
+      "learning_rate": 1.9999537828068586e-05,
+      "loss": 0.6928,
+      "step": 1400
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9460939693889285,
+      "learning_rate": 1.999953046295246e-05,
+      "loss": 0.7642,
+      "step": 1401
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4084186882502778,
+      "learning_rate": 1.9999523039615904e-05,
+      "loss": 0.6886,
+      "step": 1402
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.776848965280057,
+      "learning_rate": 1.9999515558058958e-05,
+      "loss": 0.7665,
+      "step": 1403
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.710858786870861,
+      "learning_rate": 1.9999508018281667e-05,
+      "loss": 0.7066,
+      "step": 1404
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8135598116274787,
+      "learning_rate": 1.999950042028407e-05,
+      "loss": 0.6449,
+      "step": 1405
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.090182288674023,
+      "learning_rate": 1.9999492764066225e-05,
+      "loss": 0.7451,
+      "step": 1406
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3946723313569678,
+      "learning_rate": 1.999948504962816e-05,
+      "loss": 0.7727,
+      "step": 1407
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8175888273846923,
+      "learning_rate": 1.9999477276969932e-05,
+      "loss": 0.5951,
+      "step": 1408
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.602666311602884,
+      "learning_rate": 1.9999469446091578e-05,
+      "loss": 0.683,
+      "step": 1409
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.549014970373278,
+      "learning_rate": 1.999946155699315e-05,
+      "loss": 0.7638,
+      "step": 1410
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9507768536028756,
+      "learning_rate": 1.999945360967469e-05,
+      "loss": 0.6847,
+      "step": 1411
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.066783493854184,
+      "learning_rate": 1.999944560413624e-05,
+      "loss": 0.8664,
+      "step": 1412
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4322204060081123,
+      "learning_rate": 1.999943754037786e-05,
+      "loss": 0.8219,
+      "step": 1413
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8557502280229188,
+      "learning_rate": 1.9999429418399585e-05,
+      "loss": 0.7111,
+      "step": 1414
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6793842152760428,
+      "learning_rate": 1.9999421238201465e-05,
+      "loss": 0.8255,
+      "step": 1415
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6067688248793484,
+      "learning_rate": 1.9999412999783555e-05,
+      "loss": 0.9574,
+      "step": 1416
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5970159798238106,
+      "learning_rate": 1.999940470314589e-05,
+      "loss": 0.6554,
+      "step": 1417
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4548205667764569,
+      "learning_rate": 1.9999396348288526e-05,
+      "loss": 0.815,
+      "step": 1418
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.499561730958424,
+      "learning_rate": 1.9999387935211515e-05,
+      "loss": 0.7375,
+      "step": 1419
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.930774817852928,
+      "learning_rate": 1.99993794639149e-05,
+      "loss": 0.8072,
+      "step": 1420
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5642408622552555,
+      "learning_rate": 1.999937093439873e-05,
+      "loss": 0.6838,
+      "step": 1421
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2691514662377978,
+      "learning_rate": 1.9999362346663056e-05,
+      "loss": 0.6482,
+      "step": 1422
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5394257977417785,
+      "learning_rate": 1.999935370070793e-05,
+      "loss": 0.8643,
+      "step": 1423
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5993064972817645,
+      "learning_rate": 1.9999344996533402e-05,
+      "loss": 0.9172,
+      "step": 1424
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4340980600102426,
+      "learning_rate": 1.9999336234139522e-05,
+      "loss": 0.7969,
+      "step": 1425
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3629004482371474,
+      "learning_rate": 1.9999327413526337e-05,
+      "loss": 0.6091,
+      "step": 1426
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5157424423890278,
+      "learning_rate": 1.9999318534693904e-05,
+      "loss": 0.7121,
+      "step": 1427
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6322746334402534,
+      "learning_rate": 1.9999309597642275e-05,
+      "loss": 0.725,
+      "step": 1428
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7373742690932352,
+      "learning_rate": 1.9999300602371495e-05,
+      "loss": 0.8597,
+      "step": 1429
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3181936374767496,
+      "learning_rate": 1.9999291548881628e-05,
+      "loss": 0.6691,
+      "step": 1430
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6785707705545698,
+      "learning_rate": 1.9999282437172716e-05,
+      "loss": 0.7175,
+      "step": 1431
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.003923781422666,
+      "learning_rate": 1.9999273267244813e-05,
+      "loss": 0.7693,
+      "step": 1432
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3987623033383654,
+      "learning_rate": 1.9999264039097982e-05,
+      "loss": 0.6686,
+      "step": 1433
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7521692413681749,
+      "learning_rate": 1.9999254752732265e-05,
+      "loss": 0.8386,
+      "step": 1434
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9737532065753824,
+      "learning_rate": 1.999924540814772e-05,
+      "loss": 0.7272,
+      "step": 1435
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5178641865737166,
+      "learning_rate": 1.999923600534441e-05,
+      "loss": 0.8729,
+      "step": 1436
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.538283561349982,
+      "learning_rate": 1.9999226544322377e-05,
+      "loss": 0.6951,
+      "step": 1437
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7627003920549953,
+      "learning_rate": 1.999921702508168e-05,
+      "loss": 0.7332,
+      "step": 1438
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.704944926849184,
+      "learning_rate": 1.9999207447622376e-05,
+      "loss": 0.8924,
+      "step": 1439
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7544561026219523,
+      "learning_rate": 1.999919781194452e-05,
+      "loss": 0.8262,
+      "step": 1440
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7922142596150152,
+      "learning_rate": 1.9999188118048173e-05,
+      "loss": 0.8031,
+      "step": 1441
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.705940685560223,
+      "learning_rate": 1.9999178365933388e-05,
+      "loss": 0.7739,
+      "step": 1442
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.116636588480111,
+      "learning_rate": 1.9999168555600215e-05,
+      "loss": 0.8215,
+      "step": 1443
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3512842199015593,
+      "learning_rate": 1.999915868704872e-05,
+      "loss": 0.6724,
+      "step": 1444
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5888314921568352,
+      "learning_rate": 1.9999148760278956e-05,
+      "loss": 0.5052,
+      "step": 1445
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4031885911693431,
+      "learning_rate": 1.9999138775290985e-05,
+      "loss": 0.7217,
+      "step": 1446
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6003368334598667,
+      "learning_rate": 1.999912873208486e-05,
+      "loss": 0.776,
+      "step": 1447
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5699489605681203,
+      "learning_rate": 1.9999118630660643e-05,
+      "loss": 0.7755,
+      "step": 1448
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.422456982056862,
+      "learning_rate": 1.999910847101839e-05,
+      "loss": 0.8489,
+      "step": 1449
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6103506369106904,
+      "learning_rate": 1.9999098253158166e-05,
+      "loss": 0.56,
+      "step": 1450
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.881797702795658,
+      "learning_rate": 1.999908797708002e-05,
+      "loss": 0.7211,
+      "step": 1451
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5346133576917775,
+      "learning_rate": 1.9999077642784024e-05,
+      "loss": 0.8741,
+      "step": 1452
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.45914778818988,
+      "learning_rate": 1.999906725027023e-05,
+      "loss": 0.7562,
+      "step": 1453
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6357781499852222,
+      "learning_rate": 1.9999056799538698e-05,
+      "loss": 0.7389,
+      "step": 1454
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7843881937981714,
+      "learning_rate": 1.9999046290589495e-05,
+      "loss": 0.8136,
+      "step": 1455
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9188005385192362,
+      "learning_rate": 1.9999035723422677e-05,
+      "loss": 0.8198,
+      "step": 1456
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.203227848941037,
+      "learning_rate": 1.999902509803831e-05,
+      "loss": 0.9486,
+      "step": 1457
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4778485784127,
+      "learning_rate": 1.9999014414436454e-05,
+      "loss": 0.8198,
+      "step": 1458
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7050848612503433,
+      "learning_rate": 1.9999003672617165e-05,
+      "loss": 0.7542,
+      "step": 1459
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.474234751536556,
+      "learning_rate": 1.9998992872580516e-05,
+      "loss": 0.7423,
+      "step": 1460
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4858632231696813,
+      "learning_rate": 1.9998982014326563e-05,
+      "loss": 0.6784,
+      "step": 1461
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6921790742089242,
+      "learning_rate": 1.9998971097855372e-05,
+      "loss": 0.6732,
+      "step": 1462
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3925901668626164,
+      "learning_rate": 1.9998960123167003e-05,
+      "loss": 0.6734,
+      "step": 1463
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.71726392476464,
+      "learning_rate": 1.9998949090261525e-05,
+      "loss": 0.8046,
+      "step": 1464
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.586992916825652,
+      "learning_rate": 1.9998937999138998e-05,
+      "loss": 0.77,
+      "step": 1465
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4099903233516458,
+      "learning_rate": 1.999892684979949e-05,
+      "loss": 0.7926,
+      "step": 1466
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.787668032792554,
+      "learning_rate": 1.9998915642243064e-05,
+      "loss": 0.7315,
+      "step": 1467
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6305645483996156,
+      "learning_rate": 1.9998904376469784e-05,
+      "loss": 0.6897,
+      "step": 1468
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7680154333629299,
+      "learning_rate": 1.999889305247972e-05,
+      "loss": 0.756,
+      "step": 1469
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.380806704797967,
+      "learning_rate": 1.9998881670272933e-05,
+      "loss": 0.6996,
+      "step": 1470
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6283551156928772,
+      "learning_rate": 1.9998870229849492e-05,
+      "loss": 0.5513,
+      "step": 1471
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4078110911792638,
+      "learning_rate": 1.9998858731209463e-05,
+      "loss": 0.6488,
+      "step": 1472
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.278664763499909,
+      "learning_rate": 1.999884717435291e-05,
+      "loss": 0.7407,
+      "step": 1473
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0115168293636887,
+      "learning_rate": 1.999883555927991e-05,
+      "loss": 0.7489,
+      "step": 1474
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.3461664322237796,
+      "learning_rate": 1.999882388599052e-05,
+      "loss": 0.8462,
+      "step": 1475
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4392251342816946,
+      "learning_rate": 1.999881215448481e-05,
+      "loss": 0.8649,
+      "step": 1476
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5752974249633416,
+      "learning_rate": 1.999880036476285e-05,
+      "loss": 0.8588,
+      "step": 1477
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.520127626901398,
+      "learning_rate": 1.999878851682471e-05,
+      "loss": 0.6769,
+      "step": 1478
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4623953091411608,
+      "learning_rate": 1.999877661067046e-05,
+      "loss": 0.8146,
+      "step": 1479
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4942153763635273,
+      "learning_rate": 1.9998764646300163e-05,
+      "loss": 0.7696,
+      "step": 1480
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.047355435910422,
+      "learning_rate": 1.9998752623713894e-05,
+      "loss": 0.8126,
+      "step": 1481
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5618087620002103,
+      "learning_rate": 1.9998740542911722e-05,
+      "loss": 0.5066,
+      "step": 1482
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6065532891790906,
+      "learning_rate": 1.9998728403893717e-05,
+      "loss": 0.5198,
+      "step": 1483
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.405263518458893,
+      "learning_rate": 1.9998716206659948e-05,
+      "loss": 0.7493,
+      "step": 1484
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4141418192805675,
+      "learning_rate": 1.9998703951210492e-05,
+      "loss": 0.8617,
+      "step": 1485
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5750179254804068,
+      "learning_rate": 1.9998691637545417e-05,
+      "loss": 0.5198,
+      "step": 1486
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7570821872767826,
+      "learning_rate": 1.9998679265664787e-05,
+      "loss": 0.7562,
+      "step": 1487
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4291293179487314,
+      "learning_rate": 1.9998666835568684e-05,
+      "loss": 0.6023,
+      "step": 1488
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.070757819755257,
+      "learning_rate": 1.999865434725718e-05,
+      "loss": 0.7422,
+      "step": 1489
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6064351322819613,
+      "learning_rate": 1.999864180073034e-05,
+      "loss": 0.5051,
+      "step": 1490
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5856472490440465,
+      "learning_rate": 1.9998629195988243e-05,
+      "loss": 0.4778,
+      "step": 1491
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7042597644825124,
+      "learning_rate": 1.9998616533030965e-05,
+      "loss": 0.9081,
+      "step": 1492
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4096750897631338,
+      "learning_rate": 1.9998603811858573e-05,
+      "loss": 0.6293,
+      "step": 1493
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.63762412041065,
+      "learning_rate": 1.9998591032471144e-05,
+      "loss": 0.697,
+      "step": 1494
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.640671569108502,
+      "learning_rate": 1.9998578194868754e-05,
+      "loss": 0.7135,
+      "step": 1495
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9696463641993156,
+      "learning_rate": 1.9998565299051473e-05,
+      "loss": 0.6324,
+      "step": 1496
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5557140270590406,
+      "learning_rate": 1.9998552345019383e-05,
+      "loss": 0.7609,
+      "step": 1497
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7575353834148242,
+      "learning_rate": 1.9998539332772555e-05,
+      "loss": 0.8105,
+      "step": 1498
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6642322859600263,
+      "learning_rate": 1.9998526262311063e-05,
+      "loss": 0.7752,
+      "step": 1499
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7983633311443246,
+      "learning_rate": 1.999851313363499e-05,
+      "loss": 0.6811,
+      "step": 1500
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.35696004430268,
+      "learning_rate": 1.9998499946744404e-05,
+      "loss": 0.7334,
+      "step": 1501
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8363006611091648,
+      "learning_rate": 1.999848670163939e-05,
+      "loss": 0.7665,
+      "step": 1502
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4898240125850373,
+      "learning_rate": 1.9998473398320016e-05,
+      "loss": 0.8081,
+      "step": 1503
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4801244765061872,
+      "learning_rate": 1.9998460036786365e-05,
+      "loss": 0.7104,
+      "step": 1504
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4281834011613095,
+      "learning_rate": 1.999844661703852e-05,
+      "loss": 0.7132,
+      "step": 1505
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.480802188849405,
+      "learning_rate": 1.9998433139076547e-05,
+      "loss": 0.8119,
+      "step": 1506
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4170563805564629,
+      "learning_rate": 1.9998419602900535e-05,
+      "loss": 0.6727,
+      "step": 1507
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5315646096712305,
+      "learning_rate": 1.9998406008510555e-05,
+      "loss": 0.6114,
+      "step": 1508
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.507246391098961,
+      "learning_rate": 1.999839235590669e-05,
+      "loss": 0.7353,
+      "step": 1509
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5040786632141852,
+      "learning_rate": 1.999837864508902e-05,
+      "loss": 0.8191,
+      "step": 1510
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.7495757448566747,
+      "learning_rate": 1.999836487605763e-05,
+      "loss": 0.8234,
+      "step": 1511
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6551440745503994,
+      "learning_rate": 1.9998351048812585e-05,
+      "loss": 0.8,
+      "step": 1512
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4489891215821293,
+      "learning_rate": 1.999833716335398e-05,
+      "loss": 0.7089,
+      "step": 1513
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.379247202923135,
+      "learning_rate": 1.999832321968189e-05,
+      "loss": 0.7228,
+      "step": 1514
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.525622514947043,
+      "learning_rate": 1.9998309217796395e-05,
+      "loss": 0.6777,
+      "step": 1515
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.4583844504034755,
+      "learning_rate": 1.9998295157697583e-05,
+      "loss": 0.6289,
+      "step": 1516
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6914266568192717,
+      "learning_rate": 1.9998281039385527e-05,
+      "loss": 0.7315,
+      "step": 1517
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9039055551202662,
+      "learning_rate": 1.9998266862860315e-05,
+      "loss": 0.7147,
+      "step": 1518
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7309375390090551,
+      "learning_rate": 1.9998252628122025e-05,
+      "loss": 0.7457,
+      "step": 1519
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5499343224247941,
+      "learning_rate": 1.9998238335170746e-05,
+      "loss": 0.7241,
+      "step": 1520
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7048276244886436,
+      "learning_rate": 1.9998223984006558e-05,
+      "loss": 0.7622,
+      "step": 1521
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5458238353661813,
+      "learning_rate": 1.9998209574629542e-05,
+      "loss": 0.7645,
+      "step": 1522
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.503762134439106,
+      "learning_rate": 1.9998195107039785e-05,
+      "loss": 0.7461,
+      "step": 1523
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.963040112634932,
+      "learning_rate": 1.9998180581237372e-05,
+      "loss": 0.8976,
+      "step": 1524
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7166302772406414,
+      "learning_rate": 1.9998165997222387e-05,
+      "loss": 0.6921,
+      "step": 1525
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5487627296386377,
+      "learning_rate": 1.9998151354994914e-05,
+      "loss": 0.6894,
+      "step": 1526
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4936706195641012,
+      "learning_rate": 1.999813665455504e-05,
+      "loss": 0.7332,
+      "step": 1527
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4108885258492854,
+      "learning_rate": 1.9998121895902843e-05,
+      "loss": 0.7067,
+      "step": 1528
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.537621797853352,
+      "learning_rate": 1.999810707903842e-05,
+      "loss": 0.7455,
+      "step": 1529
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4156145985803898,
+      "learning_rate": 1.999809220396185e-05,
+      "loss": 0.7368,
+      "step": 1530
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5058570883271405,
+      "learning_rate": 1.9998077270673222e-05,
+      "loss": 0.6826,
+      "step": 1531
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4189065315493312,
+      "learning_rate": 1.9998062279172624e-05,
+      "loss": 0.6348,
+      "step": 1532
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6174964217672352,
+      "learning_rate": 1.9998047229460144e-05,
+      "loss": 0.7914,
+      "step": 1533
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6690937128008034,
+      "learning_rate": 1.9998032121535863e-05,
+      "loss": 0.758,
+      "step": 1534
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6453904602295297,
+      "learning_rate": 1.9998016955399877e-05,
+      "loss": 0.7018,
+      "step": 1535
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8839251571117948,
+      "learning_rate": 1.999800173105227e-05,
+      "loss": 0.7088,
+      "step": 1536
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.736050971939349,
+      "learning_rate": 1.9997986448493133e-05,
+      "loss": 0.7372,
+      "step": 1537
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4740225857278981,
+      "learning_rate": 1.999797110772255e-05,
+      "loss": 0.76,
+      "step": 1538
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.630977372093908,
+      "learning_rate": 1.9997955708740616e-05,
+      "loss": 0.7462,
+      "step": 1539
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4425985809342177,
+      "learning_rate": 1.999794025154742e-05,
+      "loss": 0.8003,
+      "step": 1540
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.6981006384229245,
+      "learning_rate": 1.999792473614305e-05,
+      "loss": 0.6988,
+      "step": 1541
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4881497147468543,
+      "learning_rate": 1.9997909162527596e-05,
+      "loss": 0.7907,
+      "step": 1542
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5346485564772812,
+      "learning_rate": 1.999789353070115e-05,
+      "loss": 0.701,
+      "step": 1543
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5510357593779445,
+      "learning_rate": 1.99978778406638e-05,
+      "loss": 0.8233,
+      "step": 1544
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8689867172051555,
+      "learning_rate": 1.9997862092415645e-05,
+      "loss": 0.6562,
+      "step": 1545
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6066050183140972,
+      "learning_rate": 1.9997846285956767e-05,
+      "loss": 0.7613,
+      "step": 1546
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4629533910448662,
+      "learning_rate": 1.9997830421287265e-05,
+      "loss": 0.784,
+      "step": 1547
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5989623763309755,
+      "learning_rate": 1.999781449840723e-05,
+      "loss": 0.8879,
+      "step": 1548
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8351021534715182,
+      "learning_rate": 1.999779851731675e-05,
+      "loss": 0.7476,
+      "step": 1549
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.558157977764073,
+      "learning_rate": 1.9997782478015923e-05,
+      "loss": 0.8762,
+      "step": 1550
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3758681089812836,
+      "learning_rate": 1.9997766380504842e-05,
+      "loss": 0.6763,
+      "step": 1551
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.48825475214577,
+      "learning_rate": 1.9997750224783605e-05,
+      "loss": 0.8165,
+      "step": 1552
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7464227655183007,
+      "learning_rate": 1.9997734010852294e-05,
+      "loss": 0.7252,
+      "step": 1553
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7969000034967126,
+      "learning_rate": 1.9997717738711013e-05,
+      "loss": 0.6436,
+      "step": 1554
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5157830845595819,
+      "learning_rate": 1.9997701408359852e-05,
+      "loss": 0.8211,
+      "step": 1555
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.484405808909415,
+      "learning_rate": 1.9997685019798913e-05,
+      "loss": 0.7878,
+      "step": 1556
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3896423634048205,
+      "learning_rate": 1.999766857302828e-05,
+      "loss": 0.6852,
+      "step": 1557
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4964097667904628,
+      "learning_rate": 1.9997652068048057e-05,
+      "loss": 0.9078,
+      "step": 1558
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.823220619230586,
+      "learning_rate": 1.999763550485834e-05,
+      "loss": 0.6919,
+      "step": 1559
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5232475606524527,
+      "learning_rate": 1.9997618883459225e-05,
+      "loss": 0.8499,
+      "step": 1560
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3860176776927418,
+      "learning_rate": 1.9997602203850807e-05,
+      "loss": 0.9266,
+      "step": 1561
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6123720132410682,
+      "learning_rate": 1.9997585466033186e-05,
+      "loss": 0.8094,
+      "step": 1562
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8101999990564586,
+      "learning_rate": 1.999756867000645e-05,
+      "loss": 0.7268,
+      "step": 1563
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4634580430486654,
+      "learning_rate": 1.9997551815770712e-05,
+      "loss": 0.6763,
+      "step": 1564
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7485476352476177,
+      "learning_rate": 1.9997534903326058e-05,
+      "loss": 0.6979,
+      "step": 1565
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5993352661463418,
+      "learning_rate": 1.9997517932672592e-05,
+      "loss": 0.4895,
+      "step": 1566
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7278812951583873,
+      "learning_rate": 1.9997500903810412e-05,
+      "loss": 0.7304,
+      "step": 1567
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0198854004903946,
+      "learning_rate": 1.9997483816739613e-05,
+      "loss": 0.7714,
+      "step": 1568
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9249744110624414,
+      "learning_rate": 1.99974666714603e-05,
+      "loss": 0.6928,
+      "step": 1569
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4940812049804633,
+      "learning_rate": 1.9997449467972574e-05,
+      "loss": 0.9281,
+      "step": 1570
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5038736607783412,
+      "learning_rate": 1.9997432206276528e-05,
+      "loss": 0.7308,
+      "step": 1571
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4199838935492837,
+      "learning_rate": 1.9997414886372272e-05,
+      "loss": 0.8787,
+      "step": 1572
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9963629788036654,
+      "learning_rate": 1.9997397508259895e-05,
+      "loss": 0.7568,
+      "step": 1573
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2791460588629975,
+      "learning_rate": 1.9997380071939506e-05,
+      "loss": 0.8676,
+      "step": 1574
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.655458663207651,
+      "learning_rate": 1.999736257741121e-05,
+      "loss": 0.7367,
+      "step": 1575
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.51503314611773,
+      "learning_rate": 1.9997345024675095e-05,
+      "loss": 0.831,
+      "step": 1576
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6760094836716085,
+      "learning_rate": 1.999732741373128e-05,
+      "loss": 0.4984,
+      "step": 1577
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4887302540762513,
+      "learning_rate": 1.9997309744579857e-05,
+      "loss": 0.8421,
+      "step": 1578
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.04629656156013,
+      "learning_rate": 1.9997292017220933e-05,
+      "loss": 0.8931,
+      "step": 1579
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5799291064623335,
+      "learning_rate": 1.9997274231654605e-05,
+      "loss": 0.9105,
+      "step": 1580
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2961756773775146,
+      "learning_rate": 1.9997256387880987e-05,
+      "loss": 0.7308,
+      "step": 1581
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5780573838518585,
+      "learning_rate": 1.9997238485900175e-05,
+      "loss": 0.4696,
+      "step": 1582
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5483383042727252,
+      "learning_rate": 1.9997220525712276e-05,
+      "loss": 0.6674,
+      "step": 1583
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4923449732573202,
+      "learning_rate": 1.999720250731739e-05,
+      "loss": 0.822,
+      "step": 1584
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.597286396541263,
+      "learning_rate": 1.9997184430715632e-05,
+      "loss": 0.5171,
+      "step": 1585
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6740942974846624,
+      "learning_rate": 1.9997166295907097e-05,
+      "loss": 0.7635,
+      "step": 1586
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3049761876624495,
+      "learning_rate": 1.9997148102891897e-05,
+      "loss": 0.7204,
+      "step": 1587
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3305303914878701,
+      "learning_rate": 1.9997129851670136e-05,
+      "loss": 0.7104,
+      "step": 1588
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3080944881333427,
+      "learning_rate": 1.999711154224192e-05,
+      "loss": 0.8318,
+      "step": 1589
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6891054893815418,
+      "learning_rate": 1.9997093174607354e-05,
+      "loss": 0.6535,
+      "step": 1590
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6287036330008766,
+      "learning_rate": 1.999707474876655e-05,
+      "loss": 0.6941,
+      "step": 1591
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4233403110891887,
+      "learning_rate": 1.9997056264719607e-05,
+      "loss": 0.9132,
+      "step": 1592
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4918815206007332,
+      "learning_rate": 1.999703772246664e-05,
+      "loss": 0.7695,
+      "step": 1593
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6292496663577776,
+      "learning_rate": 1.999701912200775e-05,
+      "loss": 0.8106,
+      "step": 1594
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6878262753013245,
+      "learning_rate": 1.9997000463343057e-05,
+      "loss": 0.741,
+      "step": 1595
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4945732990863774,
+      "learning_rate": 1.9996981746472657e-05,
+      "loss": 0.8877,
+      "step": 1596
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6227518406078787,
+      "learning_rate": 1.9996962971396664e-05,
+      "loss": 0.7123,
+      "step": 1597
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3490263384848005,
+      "learning_rate": 1.9996944138115192e-05,
+      "loss": 0.7483,
+      "step": 1598
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5808697459282133,
+      "learning_rate": 1.9996925246628344e-05,
+      "loss": 0.6361,
+      "step": 1599
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4708099285589198,
+      "learning_rate": 1.999690629693623e-05,
+      "loss": 0.777,
+      "step": 1600
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8453543457581434,
+      "learning_rate": 1.9996887289038968e-05,
+      "loss": 0.6323,
+      "step": 1601
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9673856370699565,
+      "learning_rate": 1.999686822293666e-05,
+      "loss": 0.673,
+      "step": 1602
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5562409543806022,
+      "learning_rate": 1.999684909862942e-05,
+      "loss": 0.4872,
+      "step": 1603
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.668163168497084,
+      "learning_rate": 1.999682991611736e-05,
+      "loss": 0.8477,
+      "step": 1604
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5049261638306168,
+      "learning_rate": 1.9996810675400588e-05,
+      "loss": 0.7225,
+      "step": 1605
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5418143873239347,
+      "learning_rate": 1.9996791376479223e-05,
+      "loss": 0.662,
+      "step": 1606
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6001598861450665,
+      "learning_rate": 1.9996772019353373e-05,
+      "loss": 0.7137,
+      "step": 1607
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.452164638769783,
+      "learning_rate": 1.9996752604023152e-05,
+      "loss": 0.762,
+      "step": 1608
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3710074112803088,
+      "learning_rate": 1.999673313048867e-05,
+      "loss": 0.6829,
+      "step": 1609
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.944758934930072,
+      "learning_rate": 1.9996713598750047e-05,
+      "loss": 0.6845,
+      "step": 1610
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8727405453479997,
+      "learning_rate": 1.999669400880739e-05,
+      "loss": 0.7643,
+      "step": 1611
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4260237976973769,
+      "learning_rate": 1.9996674360660815e-05,
+      "loss": 0.7229,
+      "step": 1612
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.418621136324294,
+      "learning_rate": 1.999665465431044e-05,
+      "loss": 0.6698,
+      "step": 1613
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.589098181606434,
+      "learning_rate": 1.999663488975637e-05,
+      "loss": 0.8028,
+      "step": 1614
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3828066112716542,
+      "learning_rate": 1.9996615066998733e-05,
+      "loss": 0.6125,
+      "step": 1615
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4845495096909878,
+      "learning_rate": 1.9996595186037635e-05,
+      "loss": 0.8377,
+      "step": 1616
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8167664504581353,
+      "learning_rate": 1.9996575246873196e-05,
+      "loss": 0.8558,
+      "step": 1617
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5871029674444754,
+      "learning_rate": 1.9996555249505533e-05,
+      "loss": 0.7948,
+      "step": 1618
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4127246819334833,
+      "learning_rate": 1.999653519393476e-05,
+      "loss": 0.6759,
+      "step": 1619
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7522501494861074,
+      "learning_rate": 1.999651508016099e-05,
+      "loss": 0.8088,
+      "step": 1620
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5819977583658829,
+      "learning_rate": 1.999649490818435e-05,
+      "loss": 0.8412,
+      "step": 1621
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5390158610036533,
+      "learning_rate": 1.999647467800495e-05,
+      "loss": 0.7406,
+      "step": 1622
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5711967365362645,
+      "learning_rate": 1.999645438962291e-05,
+      "loss": 0.8009,
+      "step": 1623
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6249959603833115,
+      "learning_rate": 1.9996434043038344e-05,
+      "loss": 0.7959,
+      "step": 1624
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5903447388472742,
+      "learning_rate": 1.9996413638251377e-05,
+      "loss": 0.9371,
+      "step": 1625
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.448341545597247,
+      "learning_rate": 1.9996393175262123e-05,
+      "loss": 0.874,
+      "step": 1626
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4945920061089084,
+      "learning_rate": 1.9996372654070708e-05,
+      "loss": 0.7602,
+      "step": 1627
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4800555017016954,
+      "learning_rate": 1.999635207467724e-05,
+      "loss": 0.7353,
+      "step": 1628
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8351716231134454,
+      "learning_rate": 1.999633143708185e-05,
+      "loss": 0.7691,
+      "step": 1629
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3608706820139882,
+      "learning_rate": 1.9996310741284652e-05,
+      "loss": 0.8327,
+      "step": 1630
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5940433469248402,
+      "learning_rate": 1.999628998728577e-05,
+      "loss": 0.5265,
+      "step": 1631
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.540499602383907,
+      "learning_rate": 1.999626917508532e-05,
+      "loss": 0.776,
+      "step": 1632
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6015135216992425,
+      "learning_rate": 1.999624830468343e-05,
+      "loss": 0.4975,
+      "step": 1633
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.641678100933532,
+      "learning_rate": 1.9996227376080213e-05,
+      "loss": 0.8109,
+      "step": 1634
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5939934096530795,
+      "learning_rate": 1.99962063892758e-05,
+      "loss": 0.5235,
+      "step": 1635
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6690096748934091,
+      "learning_rate": 1.9996185344270306e-05,
+      "loss": 1.036,
+      "step": 1636
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.801246823425655,
+      "learning_rate": 1.9996164241063857e-05,
+      "loss": 0.8678,
+      "step": 1637
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5404751342123264,
+      "learning_rate": 1.9996143079656575e-05,
+      "loss": 0.6789,
+      "step": 1638
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7111549712462322,
+      "learning_rate": 1.999612186004858e-05,
+      "loss": 0.7758,
+      "step": 1639
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8749298904410086,
+      "learning_rate": 1.9996100582240002e-05,
+      "loss": 0.8938,
+      "step": 1640
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9702205411481197,
+      "learning_rate": 1.9996079246230962e-05,
+      "loss": 0.7299,
+      "step": 1641
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4901922864371555,
+      "learning_rate": 1.9996057852021584e-05,
+      "loss": 0.8647,
+      "step": 1642
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5198689287577625,
+      "learning_rate": 1.9996036399611992e-05,
+      "loss": 0.7236,
+      "step": 1643
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6273261056733498,
+      "learning_rate": 1.999601488900231e-05,
+      "loss": 0.497,
+      "step": 1644
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4497066944783876,
+      "learning_rate": 1.9995993320192668e-05,
+      "loss": 0.7176,
+      "step": 1645
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.626175518105043,
+      "learning_rate": 1.9995971693183185e-05,
+      "loss": 0.7168,
+      "step": 1646
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.948726765611906,
+      "learning_rate": 1.999595000797399e-05,
+      "loss": 0.7017,
+      "step": 1647
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6528621853257017,
+      "learning_rate": 1.999592826456521e-05,
+      "loss": 0.7506,
+      "step": 1648
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.895418738773665,
+      "learning_rate": 1.9995906462956973e-05,
+      "loss": 0.6552,
+      "step": 1649
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4263209675291182,
+      "learning_rate": 1.9995884603149403e-05,
+      "loss": 0.7331,
+      "step": 1650
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6324182805982268,
+      "learning_rate": 1.999586268514263e-05,
+      "loss": 0.4766,
+      "step": 1651
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.612363098644215,
+      "learning_rate": 1.9995840708936773e-05,
+      "loss": 0.522,
+      "step": 1652
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.452904512561089,
+      "learning_rate": 1.9995818674531972e-05,
+      "loss": 0.7707,
+      "step": 1653
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4725819218733205,
+      "learning_rate": 1.999579658192835e-05,
+      "loss": 0.771,
+      "step": 1654
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.466578487542734,
+      "learning_rate": 1.9995774431126036e-05,
+      "loss": 0.7429,
+      "step": 1655
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.202056225219215,
+      "learning_rate": 1.999575222212516e-05,
+      "loss": 0.7478,
+      "step": 1656
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5433785066667034,
+      "learning_rate": 1.9995729954925844e-05,
+      "loss": 0.7964,
+      "step": 1657
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8274652428886202,
+      "learning_rate": 1.9995707629528227e-05,
+      "loss": 0.7974,
+      "step": 1658
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3852073208756355,
+      "learning_rate": 1.9995685245932436e-05,
+      "loss": 0.6921,
+      "step": 1659
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7793516969832377,
+      "learning_rate": 1.99956628041386e-05,
+      "loss": 0.7205,
+      "step": 1660
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6871922837130515,
+      "learning_rate": 1.999564030414685e-05,
+      "loss": 0.799,
+      "step": 1661
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4768285542396442,
+      "learning_rate": 1.999561774595732e-05,
+      "loss": 0.8349,
+      "step": 1662
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8013482942789076,
+      "learning_rate": 1.999559512957014e-05,
+      "loss": 0.7192,
+      "step": 1663
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6183898318963301,
+      "learning_rate": 1.9995572454985437e-05,
+      "loss": 0.7143,
+      "step": 1664
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4018267150270711,
+      "learning_rate": 1.999554972220335e-05,
+      "loss": 0.787,
+      "step": 1665
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.528321767452495,
+      "learning_rate": 1.9995526931224004e-05,
+      "loss": 0.7552,
+      "step": 1666
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.552530795710425,
+      "learning_rate": 1.9995504082047538e-05,
+      "loss": 0.7688,
+      "step": 1667
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6483393818389768,
+      "learning_rate": 1.9995481174674084e-05,
+      "loss": 0.56,
+      "step": 1668
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5008841061588898,
+      "learning_rate": 1.999545820910377e-05,
+      "loss": 0.7629,
+      "step": 1669
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6027521546513177,
+      "learning_rate": 1.9995435185336735e-05,
+      "loss": 0.755,
+      "step": 1670
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.477266965071843,
+      "learning_rate": 1.9995412103373115e-05,
+      "loss": 0.7113,
+      "step": 1671
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8739784374252382,
+      "learning_rate": 1.999538896321304e-05,
+      "loss": 0.8702,
+      "step": 1672
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6383057484368645,
+      "learning_rate": 1.9995365764856646e-05,
+      "loss": 0.5448,
+      "step": 1673
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5747555329371947,
+      "learning_rate": 1.9995342508304065e-05,
+      "loss": 0.853,
+      "step": 1674
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.339161311375608,
+      "learning_rate": 1.999531919355544e-05,
+      "loss": 0.7538,
+      "step": 1675
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.719844581087974,
+      "learning_rate": 1.9995295820610897e-05,
+      "loss": 0.7579,
+      "step": 1676
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7646691603580484,
+      "learning_rate": 1.9995272389470582e-05,
+      "loss": 0.7883,
+      "step": 1677
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6096172816829121,
+      "learning_rate": 1.9995248900134624e-05,
+      "loss": 0.5308,
+      "step": 1678
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6368959607434557,
+      "learning_rate": 1.9995225352603166e-05,
+      "loss": 0.6236,
+      "step": 1679
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5319964553942547,
+      "learning_rate": 1.9995201746876337e-05,
+      "loss": 0.7644,
+      "step": 1680
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8232591026578355,
+      "learning_rate": 1.999517808295428e-05,
+      "loss": 0.7959,
+      "step": 1681
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6577449516787737,
+      "learning_rate": 1.9995154360837135e-05,
+      "loss": 0.8128,
+      "step": 1682
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5759967859835562,
+      "learning_rate": 1.999513058052503e-05,
+      "loss": 0.7725,
+      "step": 1683
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.611996005596002,
+      "learning_rate": 1.999510674201812e-05,
+      "loss": 0.7766,
+      "step": 1684
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.1290628506950586,
+      "learning_rate": 1.9995082845316528e-05,
+      "loss": 0.8193,
+      "step": 1685
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.031046232361495,
+      "learning_rate": 1.99950588904204e-05,
+      "loss": 0.7393,
+      "step": 1686
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.941864615213049,
+      "learning_rate": 1.9995034877329876e-05,
+      "loss": 0.7938,
+      "step": 1687
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6899037847718,
+      "learning_rate": 1.9995010806045096e-05,
+      "loss": 0.79,
+      "step": 1688
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.503376523708327,
+      "learning_rate": 1.9994986676566194e-05,
+      "loss": 0.832,
+      "step": 1689
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8387098921076266,
+      "learning_rate": 1.9994962488893316e-05,
+      "loss": 0.5839,
+      "step": 1690
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4902961177662548,
+      "learning_rate": 1.9994938243026605e-05,
+      "loss": 0.82,
+      "step": 1691
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6355809200812615,
+      "learning_rate": 1.99949139389662e-05,
+      "loss": 0.5044,
+      "step": 1692
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4543360246872914,
+      "learning_rate": 1.999488957671224e-05,
+      "loss": 0.6442,
+      "step": 1693
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5077316114498265,
+      "learning_rate": 1.9994865156264868e-05,
+      "loss": 0.7266,
+      "step": 1694
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.782403412072403,
+      "learning_rate": 1.9994840677624226e-05,
+      "loss": 0.6008,
+      "step": 1695
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8121218539930957,
+      "learning_rate": 1.999481614079046e-05,
+      "loss": 0.8018,
+      "step": 1696
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5882438645010173,
+      "learning_rate": 1.9994791545763707e-05,
+      "loss": 0.886,
+      "step": 1697
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.257986942727932,
+      "learning_rate": 1.9994766892544113e-05,
+      "loss": 0.7802,
+      "step": 1698
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5806633022305314,
+      "learning_rate": 1.9994742181131824e-05,
+      "loss": 0.6899,
+      "step": 1699
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.634773392581232,
+      "learning_rate": 1.999471741152698e-05,
+      "loss": 0.7156,
+      "step": 1700
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5391068411796598,
+      "learning_rate": 1.9994692583729728e-05,
+      "loss": 0.7927,
+      "step": 1701
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6437455973142983,
+      "learning_rate": 1.9994667697740212e-05,
+      "loss": 0.7875,
+      "step": 1702
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6947000560611567,
+      "learning_rate": 1.9994642753558577e-05,
+      "loss": 0.8047,
+      "step": 1703
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7321202769387534,
+      "learning_rate": 1.999461775118497e-05,
+      "loss": 0.7964,
+      "step": 1704
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.775335819478242,
+      "learning_rate": 1.9994592690619526e-05,
+      "loss": 0.7792,
+      "step": 1705
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5401557221732967,
+      "learning_rate": 1.9994567571862405e-05,
+      "loss": 0.8491,
+      "step": 1706
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8120550377669173,
+      "learning_rate": 1.9994542394913748e-05,
+      "loss": 0.8261,
+      "step": 1707
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7732744712593778,
+      "learning_rate": 1.99945171597737e-05,
+      "loss": 0.8159,
+      "step": 1708
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3635966473378958,
+      "learning_rate": 1.9994491866442408e-05,
+      "loss": 0.8328,
+      "step": 1709
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6152724931655301,
+      "learning_rate": 1.9994466514920023e-05,
+      "loss": 0.5319,
+      "step": 1710
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7181473042993403,
+      "learning_rate": 1.9994441105206684e-05,
+      "loss": 0.7482,
+      "step": 1711
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5782051095231404,
+      "learning_rate": 1.9994415637302545e-05,
+      "loss": 0.7227,
+      "step": 1712
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4568224155941778,
+      "learning_rate": 1.999439011120776e-05,
+      "loss": 0.8206,
+      "step": 1713
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5268833327331808,
+      "learning_rate": 1.9994364526922467e-05,
+      "loss": 0.8264,
+      "step": 1714
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.600977271914005,
+      "learning_rate": 1.9994338884446823e-05,
+      "loss": 0.7929,
+      "step": 1715
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6188065434360711,
+      "learning_rate": 1.999431318378097e-05,
+      "loss": 0.5153,
+      "step": 1716
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9277303612842633,
+      "learning_rate": 1.9994287424925063e-05,
+      "loss": 0.6262,
+      "step": 1717
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7595317912921329,
+      "learning_rate": 1.9994261607879252e-05,
+      "loss": 0.7904,
+      "step": 1718
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.9176659638948257,
+      "learning_rate": 1.9994235732643687e-05,
+      "loss": 0.7997,
+      "step": 1719
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5243056556951224,
+      "learning_rate": 1.9994209799218514e-05,
+      "loss": 0.7679,
+      "step": 1720
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.632322397945654,
+      "learning_rate": 1.9994183807603885e-05,
+      "loss": 0.7961,
+      "step": 1721
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9949228655453262,
+      "learning_rate": 1.999415775779996e-05,
+      "loss": 0.7502,
+      "step": 1722
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6109904009887841,
+      "learning_rate": 1.999413164980688e-05,
+      "loss": 0.5528,
+      "step": 1723
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5565343725794707,
+      "learning_rate": 1.9994105483624807e-05,
+      "loss": 0.6997,
+      "step": 1724
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5456119102267696,
+      "learning_rate": 1.9994079259253883e-05,
+      "loss": 0.7415,
+      "step": 1725
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.031804408805569,
+      "learning_rate": 1.9994052976694265e-05,
+      "loss": 0.6588,
+      "step": 1726
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8471233521729382,
+      "learning_rate": 1.999402663594611e-05,
+      "loss": 0.7669,
+      "step": 1727
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.2806108787295347,
+      "learning_rate": 1.9994000237009565e-05,
+      "loss": 0.773,
+      "step": 1728
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9695601585406128,
+      "learning_rate": 1.9993973779884785e-05,
+      "loss": 0.9192,
+      "step": 1729
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4869112893930776,
+      "learning_rate": 1.999394726457193e-05,
+      "loss": 0.7429,
+      "step": 1730
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4453571505290634,
+      "learning_rate": 1.9993920691071146e-05,
+      "loss": 0.7207,
+      "step": 1731
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6642114514972612,
+      "learning_rate": 1.9993894059382595e-05,
+      "loss": 0.6462,
+      "step": 1732
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4616318568197988,
+      "learning_rate": 1.999386736950643e-05,
+      "loss": 0.6883,
+      "step": 1733
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.625664965212474,
+      "learning_rate": 1.9993840621442802e-05,
+      "loss": 0.7863,
+      "step": 1734
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5801970588682375,
+      "learning_rate": 1.9993813815191868e-05,
+      "loss": 0.4849,
+      "step": 1735
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.529969288210875,
+      "learning_rate": 1.9993786950753792e-05,
+      "loss": 0.7545,
+      "step": 1736
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6171003060601683,
+      "learning_rate": 1.999376002812872e-05,
+      "loss": 0.662,
+      "step": 1737
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5406140679846212,
+      "learning_rate": 1.9993733047316814e-05,
+      "loss": 0.877,
+      "step": 1738
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5879381748853956,
+      "learning_rate": 1.999370600831823e-05,
+      "loss": 0.5457,
+      "step": 1739
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.4205099372711705,
+      "learning_rate": 1.9993678911133126e-05,
+      "loss": 0.7857,
+      "step": 1740
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.553528113416292,
+      "learning_rate": 1.9993651755761658e-05,
+      "loss": 0.8581,
+      "step": 1741
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.379176010096718,
+      "learning_rate": 1.9993624542203987e-05,
+      "loss": 0.8221,
+      "step": 1742
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.400688366899524,
+      "learning_rate": 1.9993597270460273e-05,
+      "loss": 0.7979,
+      "step": 1743
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5182227235568182,
+      "learning_rate": 1.999356994053067e-05,
+      "loss": 0.9401,
+      "step": 1744
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0101776749375455,
+      "learning_rate": 1.9993542552415334e-05,
+      "loss": 0.7582,
+      "step": 1745
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.929483555211011,
+      "learning_rate": 1.9993515106114433e-05,
+      "loss": 0.8458,
+      "step": 1746
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5482926875906813,
+      "learning_rate": 1.9993487601628124e-05,
+      "loss": 0.6358,
+      "step": 1747
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5531593357234565,
+      "learning_rate": 1.9993460038956565e-05,
+      "loss": 0.8079,
+      "step": 1748
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1183369730511923,
+      "learning_rate": 1.999343241809992e-05,
+      "loss": 0.7192,
+      "step": 1749
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7109024134273265,
+      "learning_rate": 1.9993404739058346e-05,
+      "loss": 0.7559,
+      "step": 1750
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5876191419343793,
+      "learning_rate": 1.9993377001832005e-05,
+      "loss": 0.5035,
+      "step": 1751
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5466972880918415,
+      "learning_rate": 1.999334920642106e-05,
+      "loss": 0.6871,
+      "step": 1752
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0561493719163155,
+      "learning_rate": 1.999332135282567e-05,
+      "loss": 0.8108,
+      "step": 1753
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.453669506743669,
+      "learning_rate": 1.9993293441046e-05,
+      "loss": 0.6956,
+      "step": 1754
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.547269669117197,
+      "learning_rate": 1.9993265471082213e-05,
+      "loss": 0.8453,
+      "step": 1755
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5402233285334572,
+      "learning_rate": 1.999323744293447e-05,
+      "loss": 0.8062,
+      "step": 1756
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9076015817226841,
+      "learning_rate": 1.999320935660294e-05,
+      "loss": 0.853,
+      "step": 1757
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5014559275858623,
+      "learning_rate": 1.9993181212087775e-05,
+      "loss": 0.5757,
+      "step": 1758
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9609905245160602,
+      "learning_rate": 1.9993153009389145e-05,
+      "loss": 0.7015,
+      "step": 1759
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5295861901601482,
+      "learning_rate": 1.999312474850722e-05,
+      "loss": 0.7916,
+      "step": 1760
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.853136194136239,
+      "learning_rate": 1.9993096429442153e-05,
+      "loss": 0.7785,
+      "step": 1761
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.423596396029853,
+      "learning_rate": 1.9993068052194115e-05,
+      "loss": 0.5968,
+      "step": 1762
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4158933708632486,
+      "learning_rate": 1.9993039616763276e-05,
+      "loss": 0.6898,
+      "step": 1763
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5750110794008647,
+      "learning_rate": 1.999301112314979e-05,
+      "loss": 0.4849,
+      "step": 1764
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4785729027927714,
+      "learning_rate": 1.9992982571353837e-05,
+      "loss": 0.5847,
+      "step": 1765
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6322906243806233,
+      "learning_rate": 1.9992953961375573e-05,
+      "loss": 0.5619,
+      "step": 1766
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6298528887443122,
+      "learning_rate": 1.9992925293215164e-05,
+      "loss": 0.5564,
+      "step": 1767
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1791266235975777,
+      "learning_rate": 1.999289656687278e-05,
+      "loss": 0.7464,
+      "step": 1768
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5885110750308159,
+      "learning_rate": 1.9992867782348595e-05,
+      "loss": 0.4752,
+      "step": 1769
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3983224837064046,
+      "learning_rate": 1.9992838939642766e-05,
+      "loss": 0.7816,
+      "step": 1770
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5919186137641999,
+      "learning_rate": 1.9992810038755464e-05,
+      "loss": 0.7974,
+      "step": 1771
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4823777733954435,
+      "learning_rate": 1.999278107968686e-05,
+      "loss": 0.6486,
+      "step": 1772
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8135400822613168,
+      "learning_rate": 1.999275206243712e-05,
+      "loss": 0.6054,
+      "step": 1773
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.638291242694972,
+      "learning_rate": 1.9992722987006412e-05,
+      "loss": 0.7708,
+      "step": 1774
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.622953210743848,
+      "learning_rate": 1.999269385339491e-05,
+      "loss": 0.6433,
+      "step": 1775
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5575792709918075,
+      "learning_rate": 1.9992664661602782e-05,
+      "loss": 0.6938,
+      "step": 1776
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6171699832360287,
+      "learning_rate": 1.9992635411630194e-05,
+      "loss": 0.5066,
+      "step": 1777
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3619011278985746,
+      "learning_rate": 1.9992606103477318e-05,
+      "loss": 0.6926,
+      "step": 1778
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5754862494285784,
+      "learning_rate": 1.999257673714433e-05,
+      "loss": 0.6896,
+      "step": 1779
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5465262029961255,
+      "learning_rate": 1.999254731263139e-05,
+      "loss": 0.7739,
+      "step": 1780
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8348918031683092,
+      "learning_rate": 1.9992517829938683e-05,
+      "loss": 0.707,
+      "step": 1781
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6676108703757009,
+      "learning_rate": 1.999248828906637e-05,
+      "loss": 0.7946,
+      "step": 1782
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7076188527267047,
+      "learning_rate": 1.9992458690014625e-05,
+      "loss": 0.6781,
+      "step": 1783
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6046078337623264,
+      "learning_rate": 1.9992429032783625e-05,
+      "loss": 0.5421,
+      "step": 1784
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5089495673614695,
+      "learning_rate": 1.9992399317373537e-05,
+      "loss": 0.7663,
+      "step": 1785
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.723706455327448,
+      "learning_rate": 1.999236954378454e-05,
+      "loss": 0.7275,
+      "step": 1786
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7869189561564633,
+      "learning_rate": 1.9992339712016805e-05,
+      "loss": 0.6974,
+      "step": 1787
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7228154046368995,
+      "learning_rate": 1.9992309822070503e-05,
+      "loss": 0.8326,
+      "step": 1788
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8851542578727114,
+      "learning_rate": 1.9992279873945807e-05,
+      "loss": 0.9026,
+      "step": 1789
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9101845640685648,
+      "learning_rate": 1.9992249867642894e-05,
+      "loss": 0.6669,
+      "step": 1790
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4171473152134106,
+      "learning_rate": 1.999221980316194e-05,
+      "loss": 0.8343,
+      "step": 1791
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7869653261834262,
+      "learning_rate": 1.9992189680503124e-05,
+      "loss": 0.7608,
+      "step": 1792
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5750036232460687,
+      "learning_rate": 1.999215949966661e-05,
+      "loss": 0.7619,
+      "step": 1793
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5488395744612486,
+      "learning_rate": 1.999212926065258e-05,
+      "loss": 0.7039,
+      "step": 1794
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5668178064176632,
+      "learning_rate": 1.9992098963461213e-05,
+      "loss": 0.716,
+      "step": 1795
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2752559713300777,
+      "learning_rate": 1.999206860809268e-05,
+      "loss": 0.7629,
+      "step": 1796
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9693069252435855,
+      "learning_rate": 1.9992038194547163e-05,
+      "loss": 0.7793,
+      "step": 1797
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4205687296333838,
+      "learning_rate": 1.9992007722824832e-05,
+      "loss": 0.6765,
+      "step": 1798
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.694769989899447,
+      "learning_rate": 1.999197719292587e-05,
+      "loss": 0.6982,
+      "step": 1799
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0157244617140493,
+      "learning_rate": 1.9991946604850456e-05,
+      "loss": 0.767,
+      "step": 1800
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.571637118984663,
+      "learning_rate": 1.9991915958598766e-05,
+      "loss": 0.6968,
+      "step": 1801
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4351240358052593,
+      "learning_rate": 1.9991885254170972e-05,
+      "loss": 0.8287,
+      "step": 1802
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8010068071712177,
+      "learning_rate": 1.9991854491567265e-05,
+      "loss": 0.8846,
+      "step": 1803
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7754089734967409,
+      "learning_rate": 1.9991823670787817e-05,
+      "loss": 0.8324,
+      "step": 1804
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6276257405713739,
+      "learning_rate": 1.9991792791832805e-05,
+      "loss": 0.544,
+      "step": 1805
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8237922464577754,
+      "learning_rate": 1.999176185470241e-05,
+      "loss": 0.6377,
+      "step": 1806
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4991185245697844,
+      "learning_rate": 1.9991730859396822e-05,
+      "loss": 0.7642,
+      "step": 1807
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6192333688978775,
+      "learning_rate": 1.9991699805916207e-05,
+      "loss": 0.498,
+      "step": 1808
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.505194356357355,
+      "learning_rate": 1.9991668694260756e-05,
+      "loss": 0.8616,
+      "step": 1809
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5327601280469467,
+      "learning_rate": 1.9991637524430642e-05,
+      "loss": 0.6692,
+      "step": 1810
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.579410434309972,
+      "learning_rate": 1.9991606296426054e-05,
+      "loss": 0.5348,
+      "step": 1811
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.16532235876847,
+      "learning_rate": 1.9991575010247173e-05,
+      "loss": 0.7241,
+      "step": 1812
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4327706293076192,
+      "learning_rate": 1.9991543665894177e-05,
+      "loss": 0.7451,
+      "step": 1813
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.64200694728455,
+      "learning_rate": 1.999151226336725e-05,
+      "loss": 0.5512,
+      "step": 1814
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6174234887934408,
+      "learning_rate": 1.9991480802666573e-05,
+      "loss": 0.4801,
+      "step": 1815
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3788589235009066,
+      "learning_rate": 1.9991449283792338e-05,
+      "loss": 0.8095,
+      "step": 1816
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5010894758301627,
+      "learning_rate": 1.9991417706744718e-05,
+      "loss": 0.7542,
+      "step": 1817
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.579946452309982,
+      "learning_rate": 1.99913860715239e-05,
+      "loss": 0.7675,
+      "step": 1818
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.480835377604912,
+      "learning_rate": 1.9991354378130074e-05,
+      "loss": 0.7369,
+      "step": 1819
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6780847751749202,
+      "learning_rate": 1.9991322626563413e-05,
+      "loss": 0.7551,
+      "step": 1820
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4998459597574014,
+      "learning_rate": 1.9991290816824112e-05,
+      "loss": 0.8474,
+      "step": 1821
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5275200819254078,
+      "learning_rate": 1.999125894891235e-05,
+      "loss": 0.8109,
+      "step": 1822
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6436072307575393,
+      "learning_rate": 1.999122702282832e-05,
+      "loss": 0.959,
+      "step": 1823
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3751258461888094,
+      "learning_rate": 1.99911950385722e-05,
+      "loss": 0.7258,
+      "step": 1824
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3539218326805558,
+      "learning_rate": 1.999116299614418e-05,
+      "loss": 0.7585,
+      "step": 1825
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6916437168873313,
+      "learning_rate": 1.9991130895544443e-05,
+      "loss": 0.7616,
+      "step": 1826
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.456332393896532,
+      "learning_rate": 1.999109873677318e-05,
+      "loss": 0.7455,
+      "step": 1827
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5450431757121292,
+      "learning_rate": 1.9991066519830584e-05,
+      "loss": 0.6886,
+      "step": 1828
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4519363003578105,
+      "learning_rate": 1.9991034244716826e-05,
+      "loss": 0.6712,
+      "step": 1829
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6136703714112371,
+      "learning_rate": 1.9991001911432108e-05,
+      "loss": 0.5417,
+      "step": 1830
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.464292961133937,
+      "learning_rate": 1.9990969519976614e-05,
+      "loss": 0.687,
+      "step": 1831
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3352421255420834,
+      "learning_rate": 1.999093707035053e-05,
+      "loss": 0.7041,
+      "step": 1832
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4814583257438065,
+      "learning_rate": 1.9990904562554048e-05,
+      "loss": 0.8023,
+      "step": 1833
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.984697990365097,
+      "learning_rate": 1.999087199658736e-05,
+      "loss": 0.7964,
+      "step": 1834
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4653469184125925,
+      "learning_rate": 1.999083937245065e-05,
+      "loss": 0.7844,
+      "step": 1835
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.897752164267034,
+      "learning_rate": 1.999080669014411e-05,
+      "loss": 0.6625,
+      "step": 1836
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6250793008807576,
+      "learning_rate": 1.999077394966793e-05,
+      "loss": 0.8149,
+      "step": 1837
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8957085140824923,
+      "learning_rate": 1.9990741151022302e-05,
+      "loss": 0.8637,
+      "step": 1838
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2969565230931668,
+      "learning_rate": 1.9990708294207414e-05,
+      "loss": 0.6494,
+      "step": 1839
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3580906369795664,
+      "learning_rate": 1.999067537922346e-05,
+      "loss": 0.7375,
+      "step": 1840
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3998206828801458,
+      "learning_rate": 1.9990642406070634e-05,
+      "loss": 0.6244,
+      "step": 1841
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6006422548524553,
+      "learning_rate": 1.999060937474912e-05,
+      "loss": 0.779,
+      "step": 1842
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5382180879792537,
+      "learning_rate": 1.999057628525912e-05,
+      "loss": 0.7883,
+      "step": 1843
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6101038428719419,
+      "learning_rate": 1.9990543137600816e-05,
+      "loss": 0.7453,
+      "step": 1844
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6200628846387966,
+      "learning_rate": 1.9990509931774412e-05,
+      "loss": 0.7371,
+      "step": 1845
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.650118778591318,
+      "learning_rate": 1.9990476667780097e-05,
+      "loss": 0.8395,
+      "step": 1846
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5097199848100307,
+      "learning_rate": 1.999044334561806e-05,
+      "loss": 0.7885,
+      "step": 1847
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4389806132585943,
+      "learning_rate": 1.9990409965288503e-05,
+      "loss": 0.7566,
+      "step": 1848
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6045709664543664,
+      "learning_rate": 1.999037652679161e-05,
+      "loss": 0.8175,
+      "step": 1849
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.58680696284663,
+      "learning_rate": 1.9990343030127588e-05,
+      "loss": 0.7608,
+      "step": 1850
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6328054310382111,
+      "learning_rate": 1.9990309475296623e-05,
+      "loss": 0.517,
+      "step": 1851
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5950033262025969,
+      "learning_rate": 1.9990275862298916e-05,
+      "loss": 0.6641,
+      "step": 1852
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6194651820928336,
+      "learning_rate": 1.9990242191134657e-05,
+      "loss": 0.6796,
+      "step": 1853
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5856385214377717,
+      "learning_rate": 1.999020846180405e-05,
+      "loss": 0.7972,
+      "step": 1854
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6338279416964971,
+      "learning_rate": 1.999017467430728e-05,
+      "loss": 0.7092,
+      "step": 1855
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5766902788649604,
+      "learning_rate": 1.9990140828644553e-05,
+      "loss": 0.7769,
+      "step": 1856
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.636564016887599,
+      "learning_rate": 1.999010692481606e-05,
+      "loss": 0.7603,
+      "step": 1857
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.6812613309847593,
+      "learning_rate": 1.999007296282201e-05,
+      "loss": 0.7373,
+      "step": 1858
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.62055079689061,
+      "learning_rate": 1.9990038942662585e-05,
+      "loss": 0.7506,
+      "step": 1859
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3530156118014405,
+      "learning_rate": 1.9990004864337994e-05,
+      "loss": 0.769,
+      "step": 1860
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.70213936808028,
+      "learning_rate": 1.998997072784843e-05,
+      "loss": 0.6612,
+      "step": 1861
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.528168664856734,
+      "learning_rate": 1.9989936533194093e-05,
+      "loss": 0.6921,
+      "step": 1862
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5617121382058592,
+      "learning_rate": 1.9989902280375185e-05,
+      "loss": 0.8136,
+      "step": 1863
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4831703276121084,
+      "learning_rate": 1.9989867969391904e-05,
+      "loss": 0.8329,
+      "step": 1864
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.8971839460059248,
+      "learning_rate": 1.9989833600244446e-05,
+      "loss": 0.7152,
+      "step": 1865
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7100839999578825,
+      "learning_rate": 1.9989799172933018e-05,
+      "loss": 0.8003,
+      "step": 1866
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.38373376143951,
+      "learning_rate": 1.9989764687457813e-05,
+      "loss": 0.9066,
+      "step": 1867
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4943785022650458,
+      "learning_rate": 1.9989730143819033e-05,
+      "loss": 0.8267,
+      "step": 1868
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7484980367264074,
+      "learning_rate": 1.9989695542016887e-05,
+      "loss": 0.8432,
+      "step": 1869
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6554221745368266,
+      "learning_rate": 1.998966088205157e-05,
+      "loss": 0.5592,
+      "step": 1870
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5605332331137978,
+      "learning_rate": 1.998962616392328e-05,
+      "loss": 0.7383,
+      "step": 1871
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4757767552113101,
+      "learning_rate": 1.998959138763223e-05,
+      "loss": 0.781,
+      "step": 1872
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3860941991998408,
+      "learning_rate": 1.998955655317861e-05,
+      "loss": 0.8989,
+      "step": 1873
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.709770976017035,
+      "learning_rate": 1.998952166056263e-05,
+      "loss": 0.7905,
+      "step": 1874
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.529774562586392,
+      "learning_rate": 1.99894867097845e-05,
+      "loss": 0.8321,
+      "step": 1875
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5417590262191225,
+      "learning_rate": 1.998945170084441e-05,
+      "loss": 0.6981,
+      "step": 1876
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5113049692145888,
+      "learning_rate": 1.998941663374257e-05,
+      "loss": 0.8382,
+      "step": 1877
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6911312545043589,
+      "learning_rate": 1.998938150847918e-05,
+      "loss": 0.7333,
+      "step": 1878
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6578599639770807,
+      "learning_rate": 1.9989346325054454e-05,
+      "loss": 0.6122,
+      "step": 1879
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5745172108890305,
+      "learning_rate": 1.998931108346859e-05,
+      "loss": 0.5084,
+      "step": 1880
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.586500298624718,
+      "learning_rate": 1.998927578372179e-05,
+      "loss": 0.8402,
+      "step": 1881
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5066311820199505,
+      "learning_rate": 1.998924042581427e-05,
+      "loss": 0.808,
+      "step": 1882
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4794562815016845,
+      "learning_rate": 1.9989205009746227e-05,
+      "loss": 0.7932,
+      "step": 1883
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6239094427191065,
+      "learning_rate": 1.998916953551787e-05,
+      "loss": 0.5436,
+      "step": 1884
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6881839703267043,
+      "learning_rate": 1.9989134003129405e-05,
+      "loss": 0.6468,
+      "step": 1885
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.604151533974231,
+      "learning_rate": 1.998909841258104e-05,
+      "loss": 0.8405,
+      "step": 1886
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.090632434157726,
+      "learning_rate": 1.9989062763872978e-05,
+      "loss": 0.8081,
+      "step": 1887
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6467030824757274,
+      "learning_rate": 1.9989027057005438e-05,
+      "loss": 0.6094,
+      "step": 1888
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6169776578873105,
+      "learning_rate": 1.9988991291978613e-05,
+      "loss": 0.5488,
+      "step": 1889
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5065247555638386,
+      "learning_rate": 1.998895546879272e-05,
+      "loss": 0.7524,
+      "step": 1890
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.378966163819176,
+      "learning_rate": 1.9988919587447967e-05,
+      "loss": 0.6826,
+      "step": 1891
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6023244467814768,
+      "learning_rate": 1.9988883647944563e-05,
+      "loss": 0.8046,
+      "step": 1892
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3567930053096147,
+      "learning_rate": 1.998884765028271e-05,
+      "loss": 0.7907,
+      "step": 1893
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6319606165580092,
+      "learning_rate": 1.9988811594462628e-05,
+      "loss": 0.8327,
+      "step": 1894
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7157712657082589,
+      "learning_rate": 1.9988775480484524e-05,
+      "loss": 0.737,
+      "step": 1895
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7732318543498513,
+      "learning_rate": 1.9988739308348603e-05,
+      "loss": 0.862,
+      "step": 1896
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.552368153188648,
+      "learning_rate": 1.998870307805508e-05,
+      "loss": 0.7547,
+      "step": 1897
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6427869472921417,
+      "learning_rate": 1.998866678960417e-05,
+      "loss": 0.8843,
+      "step": 1898
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6966422799611423,
+      "learning_rate": 1.9988630442996072e-05,
+      "loss": 0.7207,
+      "step": 1899
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4471057190447887,
+      "learning_rate": 1.9988594038231012e-05,
+      "loss": 0.8777,
+      "step": 1900
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.2263633313171205,
+      "learning_rate": 1.9988557575309192e-05,
+      "loss": 0.6997,
+      "step": 1901
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8905572947460674,
+      "learning_rate": 1.9988521054230828e-05,
+      "loss": 0.7716,
+      "step": 1902
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6510194213513844,
+      "learning_rate": 1.9988484474996134e-05,
+      "loss": 0.5583,
+      "step": 1903
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7181369999819411,
+      "learning_rate": 1.9988447837605317e-05,
+      "loss": 0.9555,
+      "step": 1904
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4484232108258837,
+      "learning_rate": 1.9988411142058596e-05,
+      "loss": 0.6657,
+      "step": 1905
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.866608943968396,
+      "learning_rate": 1.9988374388356186e-05,
+      "loss": 0.7152,
+      "step": 1906
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8382214528960894,
+      "learning_rate": 1.9988337576498296e-05,
+      "loss": 0.6974,
+      "step": 1907
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3761971268170585,
+      "learning_rate": 1.998830070648514e-05,
+      "loss": 0.6892,
+      "step": 1908
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7508904361246818,
+      "learning_rate": 1.9988263778316937e-05,
+      "loss": 0.8433,
+      "step": 1909
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5283123214957617,
+      "learning_rate": 1.99882267919939e-05,
+      "loss": 0.6074,
+      "step": 1910
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0907086047740226,
+      "learning_rate": 1.9988189747516245e-05,
+      "loss": 0.7889,
+      "step": 1911
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5105306490981567,
+      "learning_rate": 1.9988152644884186e-05,
+      "loss": 0.8893,
+      "step": 1912
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9245422812273403,
+      "learning_rate": 1.998811548409794e-05,
+      "loss": 0.7607,
+      "step": 1913
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8469978946727668,
+      "learning_rate": 1.9988078265157727e-05,
+      "loss": 0.6812,
+      "step": 1914
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.655011682786635,
+      "learning_rate": 1.9988040988063755e-05,
+      "loss": 0.7868,
+      "step": 1915
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5389066735621126,
+      "learning_rate": 1.998800365281625e-05,
+      "loss": 0.8039,
+      "step": 1916
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5056692480495482,
+      "learning_rate": 1.9987966259415423e-05,
+      "loss": 0.7828,
+      "step": 1917
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6067579797731811,
+      "learning_rate": 1.9987928807861494e-05,
+      "loss": 0.5374,
+      "step": 1918
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.342744752731957,
+      "learning_rate": 1.9987891298154684e-05,
+      "loss": 0.7644,
+      "step": 1919
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9402390483000256,
+      "learning_rate": 1.998785373029521e-05,
+      "loss": 0.8034,
+      "step": 1920
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7171291302843448,
+      "learning_rate": 1.9987816104283284e-05,
+      "loss": 0.7302,
+      "step": 1921
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7690118215611352,
+      "learning_rate": 1.998777842011913e-05,
+      "loss": 0.7236,
+      "step": 1922
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5201456402181475,
+      "learning_rate": 1.9987740677802976e-05,
+      "loss": 0.8224,
+      "step": 1923
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.883675556232612,
+      "learning_rate": 1.998770287733503e-05,
+      "loss": 0.9269,
+      "step": 1924
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.498321421027852,
+      "learning_rate": 1.9987665018715514e-05,
+      "loss": 0.752,
+      "step": 1925
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7655926534613402,
+      "learning_rate": 1.9987627101944652e-05,
+      "loss": 0.6956,
+      "step": 1926
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7120339763342185,
+      "learning_rate": 1.998758912702266e-05,
+      "loss": 0.6865,
+      "step": 1927
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5253812205257347,
+      "learning_rate": 1.9987551093949765e-05,
+      "loss": 0.7752,
+      "step": 1928
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5169537337759247,
+      "learning_rate": 1.9987513002726186e-05,
+      "loss": 0.7043,
+      "step": 1929
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6265386988845936,
+      "learning_rate": 1.9987474853352142e-05,
+      "loss": 0.7605,
+      "step": 1930
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6861587207212174,
+      "learning_rate": 1.998743664582786e-05,
+      "loss": 0.8019,
+      "step": 1931
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6120737535781204,
+      "learning_rate": 1.9987398380153556e-05,
+      "loss": 0.7037,
+      "step": 1932
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.633757788796115,
+      "learning_rate": 1.9987360056329458e-05,
+      "loss": 0.8688,
+      "step": 1933
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.934639489148063,
+      "learning_rate": 1.9987321674355792e-05,
+      "loss": 0.7368,
+      "step": 1934
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6896790443584253,
+      "learning_rate": 1.9987283234232777e-05,
+      "loss": 0.7874,
+      "step": 1935
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5094270746254905,
+      "learning_rate": 1.998724473596063e-05,
+      "loss": 0.7679,
+      "step": 1936
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.624260147474185,
+      "learning_rate": 1.9987206179539587e-05,
+      "loss": 0.8149,
+      "step": 1937
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.576743769778444,
+      "learning_rate": 1.9987167564969868e-05,
+      "loss": 0.8346,
+      "step": 1938
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6191874962656374,
+      "learning_rate": 1.9987128892251696e-05,
+      "loss": 0.6662,
+      "step": 1939
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5502714243090199,
+      "learning_rate": 1.99870901613853e-05,
+      "loss": 0.829,
+      "step": 1940
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7070888719541983,
+      "learning_rate": 1.99870513723709e-05,
+      "loss": 0.7513,
+      "step": 1941
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5680341504164754,
+      "learning_rate": 1.9987012525208728e-05,
+      "loss": 0.5209,
+      "step": 1942
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4571428061989813,
+      "learning_rate": 1.9986973619899006e-05,
+      "loss": 0.8734,
+      "step": 1943
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3883671838434442,
+      "learning_rate": 1.9986934656441962e-05,
+      "loss": 0.8062,
+      "step": 1944
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5076060308738146,
+      "learning_rate": 1.998689563483782e-05,
+      "loss": 0.7776,
+      "step": 1945
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7854544664603373,
+      "learning_rate": 1.9986856555086812e-05,
+      "loss": 0.804,
+      "step": 1946
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4922444753939956,
+      "learning_rate": 1.9986817417189168e-05,
+      "loss": 0.761,
+      "step": 1947
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7408736413919388,
+      "learning_rate": 1.9986778221145103e-05,
+      "loss": 0.8149,
+      "step": 1948
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.666645095307326,
+      "learning_rate": 1.9986738966954856e-05,
+      "loss": 0.7111,
+      "step": 1949
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4003795131371515,
+      "learning_rate": 1.9986699654618656e-05,
+      "loss": 0.6757,
+      "step": 1950
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7342696286746295,
+      "learning_rate": 1.9986660284136728e-05,
+      "loss": 0.6816,
+      "step": 1951
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.650881573877433,
+      "learning_rate": 1.99866208555093e-05,
+      "loss": 0.7549,
+      "step": 1952
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7048330490179937,
+      "learning_rate": 1.9986581368736607e-05,
+      "loss": 0.6244,
+      "step": 1953
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5944502101015112,
+      "learning_rate": 1.9986541823818872e-05,
+      "loss": 0.5044,
+      "step": 1954
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8720574290453988,
+      "learning_rate": 1.9986502220756333e-05,
+      "loss": 0.7592,
+      "step": 1955
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3438852951910893,
+      "learning_rate": 1.998646255954921e-05,
+      "loss": 0.65,
+      "step": 1956
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6308233334576205,
+      "learning_rate": 1.9986422840197746e-05,
+      "loss": 0.7169,
+      "step": 1957
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.576745387573051,
+      "learning_rate": 1.9986383062702165e-05,
+      "loss": 0.8176,
+      "step": 1958
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8689935125650672,
+      "learning_rate": 1.99863432270627e-05,
+      "loss": 0.7969,
+      "step": 1959
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.958692928666211,
+      "learning_rate": 1.998630333327958e-05,
+      "loss": 0.6929,
+      "step": 1960
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4849172369186778,
+      "learning_rate": 1.9986263381353045e-05,
+      "loss": 0.6392,
+      "step": 1961
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5867428865877214,
+      "learning_rate": 1.998622337128332e-05,
+      "loss": 0.511,
+      "step": 1962
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6451232340567084,
+      "learning_rate": 1.9986183303070644e-05,
+      "loss": 0.833,
+      "step": 1963
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7255702985466135,
+      "learning_rate": 1.9986143176715246e-05,
+      "loss": 0.815,
+      "step": 1964
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5832492945122674,
+      "learning_rate": 1.9986102992217356e-05,
+      "loss": 0.4517,
+      "step": 1965
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6465410778383038,
+      "learning_rate": 1.998606274957722e-05,
+      "loss": 0.8176,
+      "step": 1966
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5610879786816352,
+      "learning_rate": 1.9986022448795062e-05,
+      "loss": 0.924,
+      "step": 1967
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.0930486587447574,
+      "learning_rate": 1.998598208987112e-05,
+      "loss": 0.6462,
+      "step": 1968
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3299329727239149,
+      "learning_rate": 1.998594167280563e-05,
+      "loss": 0.6705,
+      "step": 1969
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6744771566753271,
+      "learning_rate": 1.998590119759882e-05,
+      "loss": 0.79,
+      "step": 1970
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7473299477238522,
+      "learning_rate": 1.998586066425094e-05,
+      "loss": 0.7629,
+      "step": 1971
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4506992397143526,
+      "learning_rate": 1.9985820072762213e-05,
+      "loss": 0.7275,
+      "step": 1972
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7157148690590198,
+      "learning_rate": 1.9985779423132883e-05,
+      "loss": 0.869,
+      "step": 1973
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6630076165525434,
+      "learning_rate": 1.9985738715363185e-05,
+      "loss": 0.7544,
+      "step": 1974
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6024191424197132,
+      "learning_rate": 1.9985697949453348e-05,
+      "loss": 0.5378,
+      "step": 1975
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.513994547941403,
+      "learning_rate": 1.998565712540362e-05,
+      "loss": 0.646,
+      "step": 1976
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3670179454726608,
+      "learning_rate": 1.9985616243214236e-05,
+      "loss": 0.7695,
+      "step": 1977
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.584807625809501,
+      "learning_rate": 1.9985575302885435e-05,
+      "loss": 0.6614,
+      "step": 1978
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6063048506056158,
+      "learning_rate": 1.9985534304417448e-05,
+      "loss": 0.5397,
+      "step": 1979
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9721131043093951,
+      "learning_rate": 1.998549324781052e-05,
+      "loss": 0.7614,
+      "step": 1980
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6516626530325156,
+      "learning_rate": 1.998545213306489e-05,
+      "loss": 0.7236,
+      "step": 1981
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.798319951670053,
+      "learning_rate": 1.9985410960180795e-05,
+      "loss": 0.6247,
+      "step": 1982
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.600489707517704,
+      "learning_rate": 1.9985369729158476e-05,
+      "loss": 0.7769,
+      "step": 1983
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5001084858824598,
+      "learning_rate": 1.9985328439998176e-05,
+      "loss": 0.7418,
+      "step": 1984
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7926236804313602,
+      "learning_rate": 1.998528709270013e-05,
+      "loss": 0.7812,
+      "step": 1985
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6151492194480146,
+      "learning_rate": 1.9985245687264583e-05,
+      "loss": 0.6967,
+      "step": 1986
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5640343192632787,
+      "learning_rate": 1.9985204223691773e-05,
+      "loss": 0.5089,
+      "step": 1987
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5694506898930595,
+      "learning_rate": 1.9985162701981942e-05,
+      "loss": 0.5149,
+      "step": 1988
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4282710310902749,
+      "learning_rate": 1.9985121122135334e-05,
+      "loss": 0.7835,
+      "step": 1989
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4636635947814878,
+      "learning_rate": 1.998507948415219e-05,
+      "loss": 0.9748,
+      "step": 1990
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.474012466027782,
+      "learning_rate": 1.9985037788032753e-05,
+      "loss": 0.6563,
+      "step": 1991
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8852169201862266,
+      "learning_rate": 1.9984996033777263e-05,
+      "loss": 0.9194,
+      "step": 1992
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3890282472338633,
+      "learning_rate": 1.9984954221385962e-05,
+      "loss": 0.8776,
+      "step": 1993
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5138924899281898,
+      "learning_rate": 1.99849123508591e-05,
+      "loss": 0.8171,
+      "step": 1994
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8388305711997237,
+      "learning_rate": 1.998487042219692e-05,
+      "loss": 0.6861,
+      "step": 1995
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.7323901482198787,
+      "learning_rate": 1.9984828435399657e-05,
+      "loss": 0.7778,
+      "step": 1996
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.837827444471185,
+      "learning_rate": 1.9984786390467566e-05,
+      "loss": 0.6779,
+      "step": 1997
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5922922661123076,
+      "learning_rate": 1.9984744287400883e-05,
+      "loss": 0.8863,
+      "step": 1998
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.508228163581887,
+      "learning_rate": 1.9984702126199863e-05,
+      "loss": 0.7384,
+      "step": 1999
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.529158657029751,
+      "learning_rate": 1.9984659906864742e-05,
+      "loss": 0.8808,
+      "step": 2000
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1171895731986417,
+      "learning_rate": 1.998461762939577e-05,
+      "loss": 0.75,
+      "step": 2001
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4156608893672034,
+      "learning_rate": 1.9984575293793198e-05,
+      "loss": 0.7401,
+      "step": 2002
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8946300155479987,
+      "learning_rate": 1.9984532900057263e-05,
+      "loss": 0.7982,
+      "step": 2003
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.862163394775941,
+      "learning_rate": 1.998449044818822e-05,
+      "loss": 0.7788,
+      "step": 2004
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.678293495536697,
+      "learning_rate": 1.998444793818631e-05,
+      "loss": 0.7514,
+      "step": 2005
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6502490730533441,
+      "learning_rate": 1.9984405370051783e-05,
+      "loss": 0.555,
+      "step": 2006
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6130261362010592,
+      "learning_rate": 1.9984362743784887e-05,
+      "loss": 0.6693,
+      "step": 2007
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5343758132196552,
+      "learning_rate": 1.998432005938587e-05,
+      "loss": 0.7779,
+      "step": 2008
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.773101489225166,
+      "learning_rate": 1.998427731685498e-05,
+      "loss": 0.754,
+      "step": 2009
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5688384819775405,
+      "learning_rate": 1.998423451619247e-05,
+      "loss": 0.7651,
+      "step": 2010
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4799730020055946,
+      "learning_rate": 1.998419165739858e-05,
+      "loss": 0.6716,
+      "step": 2011
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4405312239493668,
+      "learning_rate": 1.998414874047357e-05,
+      "loss": 0.8077,
+      "step": 2012
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4037487872797676,
+      "learning_rate": 1.998410576541768e-05,
+      "loss": 0.75,
+      "step": 2013
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6018500048290798,
+      "learning_rate": 1.9984062732231174e-05,
+      "loss": 0.8499,
+      "step": 2014
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.542303873274101,
+      "learning_rate": 1.9984019640914285e-05,
+      "loss": 0.6596,
+      "step": 2015
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4628647786180557,
+      "learning_rate": 1.9983976491467278e-05,
+      "loss": 0.7442,
+      "step": 2016
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4026543339437103,
+      "learning_rate": 1.9983933283890398e-05,
+      "loss": 0.7908,
+      "step": 2017
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5915528526882496,
+      "learning_rate": 1.9983890018183896e-05,
+      "loss": 0.6825,
+      "step": 2018
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.619704408831142,
+      "learning_rate": 1.9983846694348027e-05,
+      "loss": 0.8264,
+      "step": 2019
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5519866480169433,
+      "learning_rate": 1.9983803312383042e-05,
+      "loss": 0.8744,
+      "step": 2020
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5870766832690429,
+      "learning_rate": 1.998375987228919e-05,
+      "loss": 0.5229,
+      "step": 2021
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3897421685488665,
+      "learning_rate": 1.998371637406673e-05,
+      "loss": 0.7522,
+      "step": 2022
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.745684965497594,
+      "learning_rate": 1.9983672817715914e-05,
+      "loss": 0.7097,
+      "step": 2023
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6721198890116666,
+      "learning_rate": 1.9983629203236993e-05,
+      "loss": 0.867,
+      "step": 2024
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5199379711437302,
+      "learning_rate": 1.998358553063022e-05,
+      "loss": 0.6987,
+      "step": 2025
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.8384241833478576,
+      "learning_rate": 1.998354179989585e-05,
+      "loss": 0.7648,
+      "step": 2026
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.687186415792143,
+      "learning_rate": 1.9983498011034143e-05,
+      "loss": 0.7941,
+      "step": 2027
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5972791604518628,
+      "learning_rate": 1.9983454164045352e-05,
+      "loss": 0.5319,
+      "step": 2028
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6898490474915944,
+      "learning_rate": 1.998341025892972e-05,
+      "loss": 0.6951,
+      "step": 2029
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5114098274136145,
+      "learning_rate": 1.9983366295687522e-05,
+      "loss": 0.7496,
+      "step": 2030
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.386160963572272,
+      "learning_rate": 1.9983322274319005e-05,
+      "loss": 0.7639,
+      "step": 2031
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.005279652160791,
+      "learning_rate": 1.998327819482442e-05,
+      "loss": 0.8378,
+      "step": 2032
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5457960347688433,
+      "learning_rate": 1.998323405720403e-05,
+      "loss": 0.7264,
+      "step": 2033
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4909087423237868,
+      "learning_rate": 1.9983189861458093e-05,
+      "loss": 0.8755,
+      "step": 2034
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6352833659185285,
+      "learning_rate": 1.9983145607586862e-05,
+      "loss": 0.5515,
+      "step": 2035
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.677142573771933,
+      "learning_rate": 1.9983101295590597e-05,
+      "loss": 0.7706,
+      "step": 2036
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6937130032492975,
+      "learning_rate": 1.9983056925469558e-05,
+      "loss": 0.8071,
+      "step": 2037
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.432029023166436,
+      "learning_rate": 1.9983012497223994e-05,
+      "loss": 0.895,
+      "step": 2038
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6281559881309613,
+      "learning_rate": 1.998296801085418e-05,
+      "loss": 0.7345,
+      "step": 2039
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3786421631121168,
+      "learning_rate": 1.9982923466360357e-05,
+      "loss": 0.81,
+      "step": 2040
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.734232936721938,
+      "learning_rate": 1.9982878863742798e-05,
+      "loss": 0.6943,
+      "step": 2041
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6993696946962245,
+      "learning_rate": 1.9982834203001757e-05,
+      "loss": 0.6811,
+      "step": 2042
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7408925664379127,
+      "learning_rate": 1.9982789484137495e-05,
+      "loss": 0.7673,
+      "step": 2043
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8938119067235568,
+      "learning_rate": 1.9982744707150272e-05,
+      "loss": 0.6983,
+      "step": 2044
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6670486225234677,
+      "learning_rate": 1.9982699872040346e-05,
+      "loss": 0.8055,
+      "step": 2045
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.63905229934099,
+      "learning_rate": 1.9982654978807986e-05,
+      "loss": 0.5324,
+      "step": 2046
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6495240007150702,
+      "learning_rate": 1.9982610027453447e-05,
+      "loss": 0.6671,
+      "step": 2047
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5677319467785247,
+      "learning_rate": 1.9982565017976992e-05,
+      "loss": 0.7513,
+      "step": 2048
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6476573425296017,
+      "learning_rate": 1.998251995037888e-05,
+      "loss": 0.8246,
+      "step": 2049
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.012686993933071,
+      "learning_rate": 1.9982474824659378e-05,
+      "loss": 0.6987,
+      "step": 2050
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7669111195394354,
+      "learning_rate": 1.9982429640818747e-05,
+      "loss": 0.7108,
+      "step": 2051
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6284316546402032,
+      "learning_rate": 1.998238439885725e-05,
+      "loss": 0.5784,
+      "step": 2052
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.901253414855411,
+      "learning_rate": 1.998233909877515e-05,
+      "loss": 0.8215,
+      "step": 2053
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4316361583038948,
+      "learning_rate": 1.9982293740572714e-05,
+      "loss": 0.7055,
+      "step": 2054
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8320109420273665,
+      "learning_rate": 1.99822483242502e-05,
+      "loss": 0.8458,
+      "step": 2055
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5716757078971908,
+      "learning_rate": 1.998220284980788e-05,
+      "loss": 0.8371,
+      "step": 2056
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6710319991426892,
+      "learning_rate": 1.9982157317246015e-05,
+      "loss": 0.719,
+      "step": 2057
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.489769811872799,
+      "learning_rate": 1.9982111726564865e-05,
+      "loss": 0.791,
+      "step": 2058
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.1559474525718847,
+      "learning_rate": 1.9982066077764702e-05,
+      "loss": 0.7813,
+      "step": 2059
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6106929761062652,
+      "learning_rate": 1.998202037084579e-05,
+      "loss": 0.8505,
+      "step": 2060
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6853437013246035,
+      "learning_rate": 1.9981974605808398e-05,
+      "loss": 0.8057,
+      "step": 2061
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7120322520761144,
+      "learning_rate": 1.9981928782652788e-05,
+      "loss": 0.7582,
+      "step": 2062
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4770403732374475,
+      "learning_rate": 1.9981882901379228e-05,
+      "loss": 0.7683,
+      "step": 2063
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5403041082376305,
+      "learning_rate": 1.9981836961987987e-05,
+      "loss": 0.7454,
+      "step": 2064
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.580815449593149,
+      "learning_rate": 1.998179096447933e-05,
+      "loss": 0.534,
+      "step": 2065
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0961214067984253,
+      "learning_rate": 1.9981744908853524e-05,
+      "loss": 0.813,
+      "step": 2066
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.475171141126231,
+      "learning_rate": 1.998169879511084e-05,
+      "loss": 0.6404,
+      "step": 2067
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.659196472700511,
+      "learning_rate": 1.9981652623251546e-05,
+      "loss": 0.8539,
+      "step": 2068
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5205272440355473,
+      "learning_rate": 1.998160639327591e-05,
+      "loss": 0.72,
+      "step": 2069
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.829678610737411,
+      "learning_rate": 1.9981560105184205e-05,
+      "loss": 0.7607,
+      "step": 2070
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8626035637606344,
+      "learning_rate": 1.998151375897669e-05,
+      "loss": 0.8296,
+      "step": 2071
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.487694570443898,
+      "learning_rate": 1.9981467354653645e-05,
+      "loss": 0.6453,
+      "step": 2072
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5338822880856295,
+      "learning_rate": 1.998142089221534e-05,
+      "loss": 0.7938,
+      "step": 2073
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5556231979467539,
+      "learning_rate": 1.9981374371662035e-05,
+      "loss": 0.7175,
+      "step": 2074
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4782769584731597,
+      "learning_rate": 1.9981327792994016e-05,
+      "loss": 0.8122,
+      "step": 2075
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4677051414603304,
+      "learning_rate": 1.9981281156211543e-05,
+      "loss": 0.6299,
+      "step": 2076
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6906142962902266,
+      "learning_rate": 1.9981234461314893e-05,
+      "loss": 0.7271,
+      "step": 2077
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7662611387188705,
+      "learning_rate": 1.9981187708304335e-05,
+      "loss": 0.7416,
+      "step": 2078
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4013546926596827,
+      "learning_rate": 1.9981140897180144e-05,
+      "loss": 0.6924,
+      "step": 2079
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8014838419225663,
+      "learning_rate": 1.9981094027942585e-05,
+      "loss": 0.866,
+      "step": 2080
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5883720624658562,
+      "learning_rate": 1.998104710059194e-05,
+      "loss": 0.8435,
+      "step": 2081
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0852453084993954,
+      "learning_rate": 1.998100011512848e-05,
+      "loss": 0.8209,
+      "step": 2082
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5726383694128314,
+      "learning_rate": 1.9980953071552477e-05,
+      "loss": 0.6751,
+      "step": 2083
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5978089976347536,
+      "learning_rate": 1.9980905969864204e-05,
+      "loss": 0.7477,
+      "step": 2084
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8842085733928682,
+      "learning_rate": 1.998085881006394e-05,
+      "loss": 0.8088,
+      "step": 2085
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5681054157805367,
+      "learning_rate": 1.998081159215195e-05,
+      "loss": 0.6491,
+      "step": 2086
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.781981552671008,
+      "learning_rate": 1.998076431612852e-05,
+      "loss": 0.7809,
+      "step": 2087
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4435387263515955,
+      "learning_rate": 1.998071698199392e-05,
+      "loss": 0.7797,
+      "step": 2088
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5267337959161236,
+      "learning_rate": 1.9980669589748426e-05,
+      "loss": 0.753,
+      "step": 2089
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8576538646978416,
+      "learning_rate": 1.998062213939231e-05,
+      "loss": 0.8501,
+      "step": 2090
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4294074165588067,
+      "learning_rate": 1.9980574630925856e-05,
+      "loss": 0.7481,
+      "step": 2091
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.524904551754085,
+      "learning_rate": 1.9980527064349334e-05,
+      "loss": 0.8551,
+      "step": 2092
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5828346021849895,
+      "learning_rate": 1.9980479439663024e-05,
+      "loss": 0.7352,
+      "step": 2093
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.654611337382673,
+      "learning_rate": 1.9980431756867206e-05,
+      "loss": 0.8327,
+      "step": 2094
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5009173878820352,
+      "learning_rate": 1.9980384015962152e-05,
+      "loss": 0.8225,
+      "step": 2095
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6271503072904079,
+      "learning_rate": 1.9980336216948144e-05,
+      "loss": 0.7973,
+      "step": 2096
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.580239247295924,
+      "learning_rate": 1.9980288359825457e-05,
+      "loss": 0.8556,
+      "step": 2097
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5584913323545875,
+      "learning_rate": 1.9980240444594372e-05,
+      "loss": 0.7852,
+      "step": 2098
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9655458345498562,
+      "learning_rate": 1.998019247125517e-05,
+      "loss": 0.8253,
+      "step": 2099
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0758167218671177,
+      "learning_rate": 1.998014443980812e-05,
+      "loss": 0.7059,
+      "step": 2100
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6852886018411932,
+      "learning_rate": 1.998009635025352e-05,
+      "loss": 0.9138,
+      "step": 2101
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5830833913066131,
+      "learning_rate": 1.9980048202591632e-05,
+      "loss": 0.8228,
+      "step": 2102
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5957649887802854,
+      "learning_rate": 1.9979999996822744e-05,
+      "loss": 0.6827,
+      "step": 2103
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5571811693944966,
+      "learning_rate": 1.997995173294714e-05,
+      "loss": 0.8365,
+      "step": 2104
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6844759078628595,
+      "learning_rate": 1.9979903410965095e-05,
+      "loss": 0.7666,
+      "step": 2105
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5625544973830952,
+      "learning_rate": 1.9979855030876894e-05,
+      "loss": 0.7971,
+      "step": 2106
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.787893819325571,
+      "learning_rate": 1.997980659268282e-05,
+      "loss": 0.6565,
+      "step": 2107
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.631190501096755,
+      "learning_rate": 1.9979758096383147e-05,
+      "loss": 0.8655,
+      "step": 2108
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5423054112011516,
+      "learning_rate": 1.9979709541978164e-05,
+      "loss": 0.5968,
+      "step": 2109
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5497266340314766,
+      "learning_rate": 1.9979660929468157e-05,
+      "loss": 0.8248,
+      "step": 2110
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7119933521481148,
+      "learning_rate": 1.99796122588534e-05,
+      "loss": 0.8264,
+      "step": 2111
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.565588590939699,
+      "learning_rate": 1.997956353013418e-05,
+      "loss": 0.7697,
+      "step": 2112
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5695717773065639,
+      "learning_rate": 1.9979514743310787e-05,
+      "loss": 0.4819,
+      "step": 2113
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.436856922861511,
+      "learning_rate": 1.9979465898383498e-05,
+      "loss": 0.6647,
+      "step": 2114
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6180292259868896,
+      "learning_rate": 1.9979416995352595e-05,
+      "loss": 0.7976,
+      "step": 2115
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4900157306174397,
+      "learning_rate": 1.997936803421837e-05,
+      "loss": 0.7947,
+      "step": 2116
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0493200358955592,
+      "learning_rate": 1.9979319014981105e-05,
+      "loss": 0.7642,
+      "step": 2117
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.514525497450417,
+      "learning_rate": 1.9979269937641087e-05,
+      "loss": 0.7763,
+      "step": 2118
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9198014400984824,
+      "learning_rate": 1.9979220802198597e-05,
+      "loss": 0.7215,
+      "step": 2119
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7000030757882885,
+      "learning_rate": 1.9979171608653926e-05,
+      "loss": 0.8574,
+      "step": 2120
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5284537474209228,
+      "learning_rate": 1.9979122357007355e-05,
+      "loss": 0.7319,
+      "step": 2121
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6049620137620264,
+      "learning_rate": 1.9979073047259176e-05,
+      "loss": 0.7641,
+      "step": 2122
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5235281545441808,
+      "learning_rate": 1.9979023679409676e-05,
+      "loss": 0.8594,
+      "step": 2123
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.501794036311955,
+      "learning_rate": 1.997897425345914e-05,
+      "loss": 0.6643,
+      "step": 2124
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8478005040066872,
+      "learning_rate": 1.9978924769407858e-05,
+      "loss": 0.7203,
+      "step": 2125
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.59577476079778,
+      "learning_rate": 1.9978875227256115e-05,
+      "loss": 0.5501,
+      "step": 2126
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5898834698233252,
+      "learning_rate": 1.99788256270042e-05,
+      "loss": 0.5373,
+      "step": 2127
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5484338813002756,
+      "learning_rate": 1.9978775968652405e-05,
+      "loss": 0.686,
+      "step": 2128
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.37519225973396,
+      "learning_rate": 1.9978726252201018e-05,
+      "loss": 0.7226,
+      "step": 2129
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4640186220968885,
+      "learning_rate": 1.9978676477650326e-05,
+      "loss": 0.904,
+      "step": 2130
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.530337911012872,
+      "learning_rate": 1.9978626645000617e-05,
+      "loss": 0.8123,
+      "step": 2131
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3055648819594086,
+      "learning_rate": 1.997857675425219e-05,
+      "loss": 0.7809,
+      "step": 2132
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4399353055531268,
+      "learning_rate": 1.997852680540533e-05,
+      "loss": 0.8721,
+      "step": 2133
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.591915415758052,
+      "learning_rate": 1.9978476798460324e-05,
+      "loss": 0.5129,
+      "step": 2134
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8936964841921893,
+      "learning_rate": 1.997842673341747e-05,
+      "loss": 0.674,
+      "step": 2135
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5762001017045164,
+      "learning_rate": 1.9978376610277053e-05,
+      "loss": 0.4354,
+      "step": 2136
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6828026402542757,
+      "learning_rate": 1.997832642903937e-05,
+      "loss": 0.7824,
+      "step": 2137
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5771219805570322,
+      "learning_rate": 1.9978276189704713e-05,
+      "loss": 0.8336,
+      "step": 2138
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8246803478003477,
+      "learning_rate": 1.997822589227337e-05,
+      "loss": 0.7611,
+      "step": 2139
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6035611925738362,
+      "learning_rate": 1.9978175536745642e-05,
+      "loss": 0.5393,
+      "step": 2140
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4906201240665482,
+      "learning_rate": 1.9978125123121812e-05,
+      "loss": 0.7937,
+      "step": 2141
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.432746810382094,
+      "learning_rate": 1.997807465140218e-05,
+      "loss": 0.6985,
+      "step": 2142
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5854223716854231,
+      "learning_rate": 1.997802412158704e-05,
+      "loss": 0.5312,
+      "step": 2143
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5114498700967223,
+      "learning_rate": 1.997797353367668e-05,
+      "loss": 0.8725,
+      "step": 2144
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8023209611031907,
+      "learning_rate": 1.9977922887671404e-05,
+      "loss": 0.6825,
+      "step": 2145
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6480005156515858,
+      "learning_rate": 1.9977872183571502e-05,
+      "loss": 0.8618,
+      "step": 2146
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5165403402033373,
+      "learning_rate": 1.9977821421377266e-05,
+      "loss": 0.705,
+      "step": 2147
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4611953604543253,
+      "learning_rate": 1.9977770601088994e-05,
+      "loss": 0.8311,
+      "step": 2148
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6356988255425602,
+      "learning_rate": 1.9977719722706985e-05,
+      "loss": 0.6242,
+      "step": 2149
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5801701058830484,
+      "learning_rate": 1.9977668786231536e-05,
+      "loss": 0.8154,
+      "step": 2150
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.607156742546809,
+      "learning_rate": 1.9977617791662935e-05,
+      "loss": 0.8606,
+      "step": 2151
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6164718660108238,
+      "learning_rate": 1.9977566739001485e-05,
+      "loss": 0.5226,
+      "step": 2152
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5386396979298398,
+      "learning_rate": 1.9977515628247482e-05,
+      "loss": 0.7672,
+      "step": 2153
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9501200401783019,
+      "learning_rate": 1.997746445940123e-05,
+      "loss": 0.9225,
+      "step": 2154
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6723528742379397,
+      "learning_rate": 1.9977413232463017e-05,
+      "loss": 0.7756,
+      "step": 2155
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.775223083429506,
+      "learning_rate": 1.9977361947433144e-05,
+      "loss": 0.734,
+      "step": 2156
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8451677707401788,
+      "learning_rate": 1.9977310604311912e-05,
+      "loss": 0.6413,
+      "step": 2157
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9562771989801875,
+      "learning_rate": 1.9977259203099623e-05,
+      "loss": 0.7318,
+      "step": 2158
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6159564586643842,
+      "learning_rate": 1.9977207743796566e-05,
+      "loss": 0.8046,
+      "step": 2159
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4303265819539617,
+      "learning_rate": 1.9977156226403052e-05,
+      "loss": 0.7235,
+      "step": 2160
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8673825562913402,
+      "learning_rate": 1.997710465091937e-05,
+      "loss": 0.6924,
+      "step": 2161
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7328302365363335,
+      "learning_rate": 1.9977053017345832e-05,
+      "loss": 0.7999,
+      "step": 2162
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.461038129574019,
+      "learning_rate": 1.997700132568273e-05,
+      "loss": 0.7351,
+      "step": 2163
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.538256340871592,
+      "learning_rate": 1.997694957593037e-05,
+      "loss": 0.7576,
+      "step": 2164
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.625667945141398,
+      "learning_rate": 1.9976897768089047e-05,
+      "loss": 0.526,
+      "step": 2165
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.553809565155561,
+      "learning_rate": 1.9976845902159066e-05,
+      "loss": 0.8386,
+      "step": 2166
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.664049001229578,
+      "learning_rate": 1.997679397814073e-05,
+      "loss": 0.5236,
+      "step": 2167
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4179132609000074,
+      "learning_rate": 1.9976741996034344e-05,
+      "loss": 0.8533,
+      "step": 2168
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6880156028187232,
+      "learning_rate": 1.9976689955840206e-05,
+      "loss": 0.7784,
+      "step": 2169
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6165572540426524,
+      "learning_rate": 1.997663785755862e-05,
+      "loss": 0.5328,
+      "step": 2170
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5191595851007926,
+      "learning_rate": 1.997658570118989e-05,
+      "loss": 0.8546,
+      "step": 2171
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4002473892512506,
+      "learning_rate": 1.997653348673432e-05,
+      "loss": 0.7526,
+      "step": 2172
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4604083893861974,
+      "learning_rate": 1.9976481214192215e-05,
+      "loss": 0.7717,
+      "step": 2173
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6636201962576351,
+      "learning_rate": 1.9976428883563876e-05,
+      "loss": 0.7295,
+      "step": 2174
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7028369585055942,
+      "learning_rate": 1.997637649484961e-05,
+      "loss": 0.7676,
+      "step": 2175
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.609686744216427,
+      "learning_rate": 1.997632404804972e-05,
+      "loss": 0.9286,
+      "step": 2176
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7555611707786813,
+      "learning_rate": 1.9976271543164514e-05,
+      "loss": 0.8138,
+      "step": 2177
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4176062191582262,
+      "learning_rate": 1.99762189801943e-05,
+      "loss": 0.8154,
+      "step": 2178
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5264232897889465,
+      "learning_rate": 1.9976166359139375e-05,
+      "loss": 0.7209,
+      "step": 2179
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.457636544018925,
+      "learning_rate": 1.9976113680000055e-05,
+      "loss": 0.7184,
+      "step": 2180
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4283443425783342,
+      "learning_rate": 1.997606094277664e-05,
+      "loss": 0.7471,
+      "step": 2181
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7154389094491176,
+      "learning_rate": 1.997600814746944e-05,
+      "loss": 0.7106,
+      "step": 2182
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0075730694918867,
+      "learning_rate": 1.9975955294078764e-05,
+      "loss": 0.6213,
+      "step": 2183
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5017996222127423,
+      "learning_rate": 1.9975902382604916e-05,
+      "loss": 0.7048,
+      "step": 2184
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7396068929171828,
+      "learning_rate": 1.997584941304821e-05,
+      "loss": 0.7366,
+      "step": 2185
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.5529939697287065,
+      "learning_rate": 1.9975796385408948e-05,
+      "loss": 0.6759,
+      "step": 2186
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6213614582612943,
+      "learning_rate": 1.997574329968744e-05,
+      "loss": 0.778,
+      "step": 2187
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6170195892420364,
+      "learning_rate": 1.9975690155883998e-05,
+      "loss": 0.4857,
+      "step": 2188
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.016567203607724,
+      "learning_rate": 1.997563695399893e-05,
+      "loss": 0.6577,
+      "step": 2189
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.786947483903031,
+      "learning_rate": 1.997558369403254e-05,
+      "loss": 0.8334,
+      "step": 2190
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4538192710587192,
+      "learning_rate": 1.997553037598515e-05,
+      "loss": 0.7404,
+      "step": 2191
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6264276204958467,
+      "learning_rate": 1.997547699985706e-05,
+      "loss": 0.5602,
+      "step": 2192
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6181180588555767,
+      "learning_rate": 1.997542356564859e-05,
+      "loss": 0.5175,
+      "step": 2193
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.170815516672121,
+      "learning_rate": 1.997537007336004e-05,
+      "loss": 0.8218,
+      "step": 2194
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4411092668843062,
+      "learning_rate": 1.997531652299173e-05,
+      "loss": 0.8169,
+      "step": 2195
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.253297646634068,
+      "learning_rate": 1.9975262914543968e-05,
+      "loss": 0.8175,
+      "step": 2196
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.452834207630021,
+      "learning_rate": 1.997520924801707e-05,
+      "loss": 0.6811,
+      "step": 2197
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5090810041936964,
+      "learning_rate": 1.9975155523411343e-05,
+      "loss": 0.7504,
+      "step": 2198
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4366265042330302,
+      "learning_rate": 1.9975101740727104e-05,
+      "loss": 0.7578,
+      "step": 2199
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4266116806702016,
+      "learning_rate": 1.9975047899964663e-05,
+      "loss": 0.7505,
+      "step": 2200
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3996665735134195,
+      "learning_rate": 1.9974994001124336e-05,
+      "loss": 0.8338,
+      "step": 2201
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6020572882848441,
+      "learning_rate": 1.9974940044206433e-05,
+      "loss": 0.528,
+      "step": 2202
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7131884659265584,
+      "learning_rate": 1.9974886029211275e-05,
+      "loss": 0.786,
+      "step": 2203
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6304960451817302,
+      "learning_rate": 1.9974831956139172e-05,
+      "loss": 0.5355,
+      "step": 2204
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5675487200998146,
+      "learning_rate": 1.997477782499044e-05,
+      "loss": 0.8601,
+      "step": 2205
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.535236198562971,
+      "learning_rate": 1.997472363576539e-05,
+      "loss": 0.836,
+      "step": 2206
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3970634441323224,
+      "learning_rate": 1.997466938846434e-05,
+      "loss": 0.7313,
+      "step": 2207
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.451954309792968,
+      "learning_rate": 1.9974615083087614e-05,
+      "loss": 0.6252,
+      "step": 2208
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5892258249348947,
+      "learning_rate": 1.9974560719635517e-05,
+      "loss": 0.6977,
+      "step": 2209
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4894693337962477,
+      "learning_rate": 1.997450629810837e-05,
+      "loss": 0.8211,
+      "step": 2210
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8613495788974599,
+      "learning_rate": 1.997445181850649e-05,
+      "loss": 0.7382,
+      "step": 2211
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7101407606193515,
+      "learning_rate": 1.997439728083019e-05,
+      "loss": 0.7463,
+      "step": 2212
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5904775097614878,
+      "learning_rate": 1.9974342685079795e-05,
+      "loss": 0.7036,
+      "step": 2213
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9412430548430208,
+      "learning_rate": 1.997428803125562e-05,
+      "loss": 0.6973,
+      "step": 2214
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3173290733098595,
+      "learning_rate": 1.9974233319357977e-05,
+      "loss": 0.7291,
+      "step": 2215
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5964930059582012,
+      "learning_rate": 1.9974178549387195e-05,
+      "loss": 0.871,
+      "step": 2216
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.575530100056009,
+      "learning_rate": 1.9974123721343587e-05,
+      "loss": 0.8123,
+      "step": 2217
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.598702008444967,
+      "learning_rate": 1.9974068835227468e-05,
+      "loss": 0.6914,
+      "step": 2218
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7231725797447188,
+      "learning_rate": 1.997401389103917e-05,
+      "loss": 0.837,
+      "step": 2219
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5966835100306096,
+      "learning_rate": 1.9973958888779e-05,
+      "loss": 0.8868,
+      "step": 2220
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7749689875425863,
+      "learning_rate": 1.9973903828447286e-05,
+      "loss": 0.891,
+      "step": 2221
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6457103729442135,
+      "learning_rate": 1.9973848710044345e-05,
+      "loss": 0.7177,
+      "step": 2222
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.475119001187712,
+      "learning_rate": 1.99737935335705e-05,
+      "loss": 0.6956,
+      "step": 2223
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.405017334487759,
+      "learning_rate": 1.997373829902607e-05,
+      "loss": 0.7928,
+      "step": 2224
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4997162783606799,
+      "learning_rate": 1.997368300641138e-05,
+      "loss": 0.7843,
+      "step": 2225
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5269827703867715,
+      "learning_rate": 1.997362765572675e-05,
+      "loss": 0.8385,
+      "step": 2226
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8290895029735625,
+      "learning_rate": 1.99735722469725e-05,
+      "loss": 0.6408,
+      "step": 2227
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6159102591289003,
+      "learning_rate": 1.9973516780148957e-05,
+      "loss": 0.7999,
+      "step": 2228
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.620606403136494,
+      "learning_rate": 1.9973461255256443e-05,
+      "loss": 0.8855,
+      "step": 2229
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3878502056730873,
+      "learning_rate": 1.9973405672295276e-05,
+      "loss": 0.6928,
+      "step": 2230
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8305167248023266,
+      "learning_rate": 1.9973350031265786e-05,
+      "loss": 0.8621,
+      "step": 2231
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4447843089309165,
+      "learning_rate": 1.9973294332168294e-05,
+      "loss": 0.8432,
+      "step": 2232
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.483286774482018,
+      "learning_rate": 1.9973238575003123e-05,
+      "loss": 0.8662,
+      "step": 2233
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5708310179116973,
+      "learning_rate": 1.9973182759770605e-05,
+      "loss": 0.8374,
+      "step": 2234
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4839636499555082,
+      "learning_rate": 1.9973126886471053e-05,
+      "loss": 0.7423,
+      "step": 2235
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8747925668454382,
+      "learning_rate": 1.9973070955104802e-05,
+      "loss": 0.8448,
+      "step": 2236
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6435392299295375,
+      "learning_rate": 1.9973014965672175e-05,
+      "loss": 0.6739,
+      "step": 2237
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6066877034083412,
+      "learning_rate": 1.9972958918173498e-05,
+      "loss": 0.5259,
+      "step": 2238
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.544167757381344,
+      "learning_rate": 1.9972902812609093e-05,
+      "loss": 0.7011,
+      "step": 2239
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6151544395527294,
+      "learning_rate": 1.9972846648979292e-05,
+      "loss": 0.5426,
+      "step": 2240
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4900451229620117,
+      "learning_rate": 1.9972790427284424e-05,
+      "loss": 0.7038,
+      "step": 2241
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5242855559605968,
+      "learning_rate": 1.997273414752481e-05,
+      "loss": 0.6085,
+      "step": 2242
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5572869859766911,
+      "learning_rate": 1.997267780970078e-05,
+      "loss": 0.7894,
+      "step": 2243
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.601849701699288,
+      "learning_rate": 1.9972621413812664e-05,
+      "loss": 0.5405,
+      "step": 2244
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1781395400956884,
+      "learning_rate": 1.997256495986079e-05,
+      "loss": 0.6199,
+      "step": 2245
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6202682183157813,
+      "learning_rate": 1.9972508447845483e-05,
+      "loss": 0.8357,
+      "step": 2246
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.523828580286749,
+      "learning_rate": 1.9972451877767075e-05,
+      "loss": 0.7859,
+      "step": 2247
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6029552763744568,
+      "learning_rate": 1.9972395249625897e-05,
+      "loss": 0.6978,
+      "step": 2248
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.494018362339661,
+      "learning_rate": 1.9972338563422273e-05,
+      "loss": 0.8149,
+      "step": 2249
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.763475585222528,
+      "learning_rate": 1.997228181915654e-05,
+      "loss": 0.8887,
+      "step": 2250
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9921380621254827,
+      "learning_rate": 1.9972225016829023e-05,
+      "loss": 0.7392,
+      "step": 2251
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5730859266907211,
+      "learning_rate": 1.9972168156440054e-05,
+      "loss": 0.8067,
+      "step": 2252
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.348605025629185,
+      "learning_rate": 1.9972111237989968e-05,
+      "loss": 0.8022,
+      "step": 2253
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7575782592075082,
+      "learning_rate": 1.997205426147909e-05,
+      "loss": 0.7194,
+      "step": 2254
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7356498375349096,
+      "learning_rate": 1.9971997226907756e-05,
+      "loss": 0.7303,
+      "step": 2255
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4462815824197073,
+      "learning_rate": 1.9971940134276298e-05,
+      "loss": 0.7427,
+      "step": 2256
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.746925448729011,
+      "learning_rate": 1.9971882983585048e-05,
+      "loss": 0.7748,
+      "step": 2257
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8358331563300343,
+      "learning_rate": 1.9971825774834338e-05,
+      "loss": 0.7812,
+      "step": 2258
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0027356741753515,
+      "learning_rate": 1.99717685080245e-05,
+      "loss": 0.9347,
+      "step": 2259
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4388874275512933,
+      "learning_rate": 1.997171118315587e-05,
+      "loss": 0.7329,
+      "step": 2260
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6130798463502306,
+      "learning_rate": 1.9971653800228782e-05,
+      "loss": 0.53,
+      "step": 2261
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.377480575885844,
+      "learning_rate": 1.9971596359243562e-05,
+      "loss": 0.7163,
+      "step": 2262
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8909500116474103,
+      "learning_rate": 1.9971538860200556e-05,
+      "loss": 0.6734,
+      "step": 2263
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.643070567357461,
+      "learning_rate": 1.9971481303100093e-05,
+      "loss": 0.7729,
+      "step": 2264
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7441683316119267,
+      "learning_rate": 1.997142368794251e-05,
+      "loss": 0.7319,
+      "step": 2265
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4218176658178545,
+      "learning_rate": 1.997136601472814e-05,
+      "loss": 0.788,
+      "step": 2266
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4912897787770738,
+      "learning_rate": 1.997130828345732e-05,
+      "loss": 0.778,
+      "step": 2267
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5526753153642754,
+      "learning_rate": 1.997125049413039e-05,
+      "loss": 0.7443,
+      "step": 2268
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.704922761163243,
+      "learning_rate": 1.9971192646747678e-05,
+      "loss": 0.6741,
+      "step": 2269
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9008555368598368,
+      "learning_rate": 1.997113474130953e-05,
+      "loss": 0.6848,
+      "step": 2270
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6633928114911394,
+      "learning_rate": 1.9971076777816275e-05,
+      "loss": 0.6976,
+      "step": 2271
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5735962767822524,
+      "learning_rate": 1.9971018756268256e-05,
+      "loss": 0.6877,
+      "step": 2272
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4484257188690037,
+      "learning_rate": 1.997096067666581e-05,
+      "loss": 0.7669,
+      "step": 2273
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7111824508703988,
+      "learning_rate": 1.997090253900927e-05,
+      "loss": 0.6532,
+      "step": 2274
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6268544237995709,
+      "learning_rate": 1.997084434329898e-05,
+      "loss": 0.5271,
+      "step": 2275
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.524624418608965,
+      "learning_rate": 1.9970786089535283e-05,
+      "loss": 0.7174,
+      "step": 2276
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6051577319614883,
+      "learning_rate": 1.9970727777718505e-05,
+      "loss": 0.5713,
+      "step": 2277
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5037070514668474,
+      "learning_rate": 1.9970669407848997e-05,
+      "loss": 0.8078,
+      "step": 2278
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.864182719485609,
+      "learning_rate": 1.9970610979927096e-05,
+      "loss": 0.7821,
+      "step": 2279
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7228528011615774,
+      "learning_rate": 1.997055249395314e-05,
+      "loss": 0.6683,
+      "step": 2280
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7243824522759397,
+      "learning_rate": 1.9970493949927473e-05,
+      "loss": 0.7728,
+      "step": 2281
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8203224839904455,
+      "learning_rate": 1.9970435347850428e-05,
+      "loss": 0.5962,
+      "step": 2282
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5063023392002397,
+      "learning_rate": 1.9970376687722356e-05,
+      "loss": 0.7023,
+      "step": 2283
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.905186607791761,
+      "learning_rate": 1.9970317969543592e-05,
+      "loss": 0.8452,
+      "step": 2284
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3154538586650046,
+      "learning_rate": 1.9970259193314483e-05,
+      "loss": 0.9285,
+      "step": 2285
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0487161247117536,
+      "learning_rate": 1.9970200359035366e-05,
+      "loss": 0.7065,
+      "step": 2286
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5103743907014024,
+      "learning_rate": 1.9970141466706588e-05,
+      "loss": 0.7684,
+      "step": 2287
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.056351316168378,
+      "learning_rate": 1.997008251632849e-05,
+      "loss": 0.7201,
+      "step": 2288
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5297885655079744,
+      "learning_rate": 1.9970023507901416e-05,
+      "loss": 0.5225,
+      "step": 2289
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.413282842389577,
+      "learning_rate": 1.9969964441425706e-05,
+      "loss": 0.8171,
+      "step": 2290
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.604284880832777,
+      "learning_rate": 1.9969905316901704e-05,
+      "loss": 0.5237,
+      "step": 2291
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5096305307923232,
+      "learning_rate": 1.9969846134329765e-05,
+      "loss": 0.8105,
+      "step": 2292
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.607123276066594,
+      "learning_rate": 1.996978689371022e-05,
+      "loss": 0.7434,
+      "step": 2293
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5543067445342638,
+      "learning_rate": 1.9969727595043417e-05,
+      "loss": 0.6949,
+      "step": 2294
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0077126779591676,
+      "learning_rate": 1.996966823832971e-05,
+      "loss": 0.8096,
+      "step": 2295
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3720633899439423,
+      "learning_rate": 1.9969608823569435e-05,
+      "loss": 0.7089,
+      "step": 2296
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5697976116877832,
+      "learning_rate": 1.996954935076294e-05,
+      "loss": 0.7275,
+      "step": 2297
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3990302028388246,
+      "learning_rate": 1.9969489819910572e-05,
+      "loss": 0.7749,
+      "step": 2298
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6623723900564964,
+      "learning_rate": 1.9969430231012682e-05,
+      "loss": 0.7041,
+      "step": 2299
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.667817171755104,
+      "learning_rate": 1.9969370584069607e-05,
+      "loss": 0.6729,
+      "step": 2300
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.596479571212065,
+      "learning_rate": 1.9969310879081705e-05,
+      "loss": 0.7566,
+      "step": 2301
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.544776553463573,
+      "learning_rate": 1.996925111604932e-05,
+      "loss": 0.7686,
+      "step": 2302
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.544558581648929,
+      "learning_rate": 1.9969191294972794e-05,
+      "loss": 0.6121,
+      "step": 2303
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.67356826858668,
+      "learning_rate": 1.9969131415852482e-05,
+      "loss": 0.7704,
+      "step": 2304
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.6953453435806156,
+      "learning_rate": 1.996907147868873e-05,
+      "loss": 0.6839,
+      "step": 2305
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4788069987716705,
+      "learning_rate": 1.9969011483481888e-05,
+      "loss": 0.8685,
+      "step": 2306
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5956907778445935,
+      "learning_rate": 1.9968951430232307e-05,
+      "loss": 0.8233,
+      "step": 2307
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4641079306463667,
+      "learning_rate": 1.9968891318940332e-05,
+      "loss": 0.7261,
+      "step": 2308
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1355307050029952,
+      "learning_rate": 1.996883114960632e-05,
+      "loss": 0.627,
+      "step": 2309
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7914138204913292,
+      "learning_rate": 1.996877092223061e-05,
+      "loss": 0.8056,
+      "step": 2310
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.2938850827060344,
+      "learning_rate": 1.9968710636813564e-05,
+      "loss": 0.6599,
+      "step": 2311
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0265776301962797,
+      "learning_rate": 1.996865029335553e-05,
+      "loss": 0.7986,
+      "step": 2312
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7217419047670905,
+      "learning_rate": 1.9968589891856854e-05,
+      "loss": 0.8189,
+      "step": 2313
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9611009088386493,
+      "learning_rate": 1.9968529432317894e-05,
+      "loss": 0.7094,
+      "step": 2314
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4492696814647739,
+      "learning_rate": 1.9968468914738997e-05,
+      "loss": 0.818,
+      "step": 2315
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3553661081608013,
+      "learning_rate": 1.9968408339120522e-05,
+      "loss": 0.7722,
+      "step": 2316
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4122441985880574,
+      "learning_rate": 1.9968347705462817e-05,
+      "loss": 0.6801,
+      "step": 2317
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.380383931575679,
+      "learning_rate": 1.9968287013766235e-05,
+      "loss": 0.909,
+      "step": 2318
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.624363639837058,
+      "learning_rate": 1.996822626403113e-05,
+      "loss": 0.8849,
+      "step": 2319
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6101812698211926,
+      "learning_rate": 1.9968165456257855e-05,
+      "loss": 0.8107,
+      "step": 2320
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7016125720166462,
+      "learning_rate": 1.9968104590446766e-05,
+      "loss": 0.667,
+      "step": 2321
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.343860726443333,
+      "learning_rate": 1.9968043666598216e-05,
+      "loss": 0.6729,
+      "step": 2322
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8692437472062238,
+      "learning_rate": 1.996798268471256e-05,
+      "loss": 0.6546,
+      "step": 2323
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.544004586283403,
+      "learning_rate": 1.9967921644790153e-05,
+      "loss": 0.8053,
+      "step": 2324
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7089027177682312,
+      "learning_rate": 1.9967860546831348e-05,
+      "loss": 0.6128,
+      "step": 2325
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5303979419545277,
+      "learning_rate": 1.9967799390836505e-05,
+      "loss": 0.6924,
+      "step": 2326
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6342229746834853,
+      "learning_rate": 1.996773817680598e-05,
+      "loss": 0.7199,
+      "step": 2327
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.706746135460145,
+      "learning_rate": 1.9967676904740126e-05,
+      "loss": 0.7242,
+      "step": 2328
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.562078863303349,
+      "learning_rate": 1.99676155746393e-05,
+      "loss": 0.7702,
+      "step": 2329
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6135566339414983,
+      "learning_rate": 1.9967554186503864e-05,
+      "loss": 0.5199,
+      "step": 2330
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5237552100528804,
+      "learning_rate": 1.9967492740334168e-05,
+      "loss": 0.7491,
+      "step": 2331
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3979689838061113,
+      "learning_rate": 1.9967431236130576e-05,
+      "loss": 0.8008,
+      "step": 2332
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8860323402122123,
+      "learning_rate": 1.9967369673893443e-05,
+      "loss": 0.8765,
+      "step": 2333
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6114476165573135,
+      "learning_rate": 1.9967308053623128e-05,
+      "loss": 0.6927,
+      "step": 2334
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6341013329929353,
+      "learning_rate": 1.996724637531999e-05,
+      "loss": 0.8588,
+      "step": 2335
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.362864727404066,
+      "learning_rate": 1.996718463898439e-05,
+      "loss": 0.7248,
+      "step": 2336
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8399231415631878,
+      "learning_rate": 1.996712284461668e-05,
+      "loss": 0.796,
+      "step": 2337
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8874422283507188,
+      "learning_rate": 1.996706099221723e-05,
+      "loss": 0.6885,
+      "step": 2338
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.59206780215296,
+      "learning_rate": 1.9966999081786396e-05,
+      "loss": 0.6839,
+      "step": 2339
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4668897557094367,
+      "learning_rate": 1.9966937113324536e-05,
+      "loss": 0.8131,
+      "step": 2340
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2531839861311713,
+      "learning_rate": 1.9966875086832012e-05,
+      "loss": 0.8164,
+      "step": 2341
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8009181906181164,
+      "learning_rate": 1.9966813002309187e-05,
+      "loss": 0.7679,
+      "step": 2342
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9127649036564927,
+      "learning_rate": 1.9966750859756422e-05,
+      "loss": 0.7786,
+      "step": 2343
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9560006195967017,
+      "learning_rate": 1.9966688659174074e-05,
+      "loss": 0.7617,
+      "step": 2344
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.486824169827692,
+      "learning_rate": 1.9966626400562515e-05,
+      "loss": 0.5628,
+      "step": 2345
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6462055616264644,
+      "learning_rate": 1.9966564083922098e-05,
+      "loss": 0.8971,
+      "step": 2346
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5959597169931043,
+      "learning_rate": 1.996650170925319e-05,
+      "loss": 0.7629,
+      "step": 2347
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9444697023141595,
+      "learning_rate": 1.9966439276556155e-05,
+      "loss": 0.723,
+      "step": 2348
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5553523366356286,
+      "learning_rate": 1.9966376785831354e-05,
+      "loss": 0.7839,
+      "step": 2349
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.669564407936835,
+      "learning_rate": 1.996631423707915e-05,
+      "loss": 0.7267,
+      "step": 2350
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7591011961258698,
+      "learning_rate": 1.9966251630299915e-05,
+      "loss": 0.7898,
+      "step": 2351
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4596164132500526,
+      "learning_rate": 1.9966188965494e-05,
+      "loss": 0.6746,
+      "step": 2352
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.352428200149181,
+      "learning_rate": 1.9966126242661783e-05,
+      "loss": 0.7409,
+      "step": 2353
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4328303773096156,
+      "learning_rate": 1.996606346180362e-05,
+      "loss": 0.5898,
+      "step": 2354
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6181499760929184,
+      "learning_rate": 1.9966000622919886e-05,
+      "loss": 0.5794,
+      "step": 2355
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.922606450928337,
+      "learning_rate": 1.996593772601094e-05,
+      "loss": 0.6873,
+      "step": 2356
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4546244525381706,
+      "learning_rate": 1.996587477107715e-05,
+      "loss": 0.7714,
+      "step": 2357
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4285086683663824,
+      "learning_rate": 1.9965811758118873e-05,
+      "loss": 0.5864,
+      "step": 2358
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0386124447847735,
+      "learning_rate": 1.9965748687136494e-05,
+      "loss": 0.648,
+      "step": 2359
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.2724990107903627,
+      "learning_rate": 1.9965685558130368e-05,
+      "loss": 0.7855,
+      "step": 2360
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.643756796199742,
+      "learning_rate": 1.9965622371100866e-05,
+      "loss": 0.5619,
+      "step": 2361
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4976726452754796,
+      "learning_rate": 1.9965559126048354e-05,
+      "loss": 0.7493,
+      "step": 2362
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3885557798110342,
+      "learning_rate": 1.9965495822973204e-05,
+      "loss": 0.779,
+      "step": 2363
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.789853423914373,
+      "learning_rate": 1.996543246187578e-05,
+      "loss": 0.6716,
+      "step": 2364
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4756291935957266,
+      "learning_rate": 1.9965369042756453e-05,
+      "loss": 0.8425,
+      "step": 2365
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6427245634255816,
+      "learning_rate": 1.99653055656156e-05,
+      "loss": 0.5817,
+      "step": 2366
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6734754032057721,
+      "learning_rate": 1.9965242030453577e-05,
+      "loss": 0.729,
+      "step": 2367
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2870160933976311,
+      "learning_rate": 1.9965178437270757e-05,
+      "loss": 0.6386,
+      "step": 2368
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4976408376904826,
+      "learning_rate": 1.9965114786067515e-05,
+      "loss": 0.7646,
+      "step": 2369
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6853126251422845,
+      "learning_rate": 1.9965051076844224e-05,
+      "loss": 0.6897,
+      "step": 2370
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5854278534523738,
+      "learning_rate": 1.9964987309601246e-05,
+      "loss": 0.7851,
+      "step": 2371
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5310383783961634,
+      "learning_rate": 1.996492348433896e-05,
+      "loss": 0.6576,
+      "step": 2372
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5798347996028389,
+      "learning_rate": 1.9964859601057734e-05,
+      "loss": 0.5035,
+      "step": 2373
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6154184983636727,
+      "learning_rate": 1.996479565975794e-05,
+      "loss": 0.5548,
+      "step": 2374
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6430613749460357,
+      "learning_rate": 1.996473166043995e-05,
+      "loss": 0.6859,
+      "step": 2375
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6104435026182565,
+      "learning_rate": 1.996466760310414e-05,
+      "loss": 0.5406,
+      "step": 2376
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9076997241336278,
+      "learning_rate": 1.996460348775088e-05,
+      "loss": 0.6785,
+      "step": 2377
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3402903104761328,
+      "learning_rate": 1.9964539314380542e-05,
+      "loss": 0.7331,
+      "step": 2378
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.1530243325301046,
+      "learning_rate": 1.99644750829935e-05,
+      "loss": 0.72,
+      "step": 2379
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5366597562227122,
+      "learning_rate": 1.996441079359013e-05,
+      "loss": 0.6569,
+      "step": 2380
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6217445880942198,
+      "learning_rate": 1.996434644617081e-05,
+      "loss": 0.7838,
+      "step": 2381
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3501423359823528,
+      "learning_rate": 1.9964282040735907e-05,
+      "loss": 0.769,
+      "step": 2382
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9114870156160628,
+      "learning_rate": 1.99642175772858e-05,
+      "loss": 0.8165,
+      "step": 2383
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6315001114029122,
+      "learning_rate": 1.9964153055820863e-05,
+      "loss": 0.7911,
+      "step": 2384
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6422709254374535,
+      "learning_rate": 1.996408847634147e-05,
+      "loss": 0.7631,
+      "step": 2385
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6279896187688268,
+      "learning_rate": 1.9964023838848004e-05,
+      "loss": 0.7479,
+      "step": 2386
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8480355530539092,
+      "learning_rate": 1.9963959143340833e-05,
+      "loss": 0.8514,
+      "step": 2387
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8276023232340546,
+      "learning_rate": 1.996389438982034e-05,
+      "loss": 0.8017,
+      "step": 2388
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9896540492363333,
+      "learning_rate": 1.99638295782869e-05,
+      "loss": 0.7982,
+      "step": 2389
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.746675890954931,
+      "learning_rate": 1.996376470874089e-05,
+      "loss": 0.6589,
+      "step": 2390
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6012784394140143,
+      "learning_rate": 1.9963699781182685e-05,
+      "loss": 0.7523,
+      "step": 2391
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2881224580224662,
+      "learning_rate": 1.9963634795612666e-05,
+      "loss": 0.7722,
+      "step": 2392
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5502295735681146,
+      "learning_rate": 1.9963569752031207e-05,
+      "loss": 0.771,
+      "step": 2393
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7142932474764003,
+      "learning_rate": 1.9963504650438698e-05,
+      "loss": 0.7749,
+      "step": 2394
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5651441238135406,
+      "learning_rate": 1.9963439490835504e-05,
+      "loss": 0.6927,
+      "step": 2395
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.978253870302656,
+      "learning_rate": 1.9963374273222015e-05,
+      "loss": 0.8187,
+      "step": 2396
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6316762304151349,
+      "learning_rate": 1.9963308997598604e-05,
+      "loss": 0.5106,
+      "step": 2397
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5885483984148268,
+      "learning_rate": 1.9963243663965658e-05,
+      "loss": 0.6195,
+      "step": 2398
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7796928852466507,
+      "learning_rate": 1.9963178272323548e-05,
+      "loss": 0.7485,
+      "step": 2399
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4822657981860243,
+      "learning_rate": 1.996311282267266e-05,
+      "loss": 0.8925,
+      "step": 2400
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0112773016246246,
+      "learning_rate": 1.9963047315013376e-05,
+      "loss": 0.8488,
+      "step": 2401
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.57793023048649,
+      "learning_rate": 1.996298174934608e-05,
+      "loss": 0.6349,
+      "step": 2402
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.884072978145781,
+      "learning_rate": 1.9962916125671146e-05,
+      "loss": 0.7782,
+      "step": 2403
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6595832079916077,
+      "learning_rate": 1.996285044398896e-05,
+      "loss": 0.8149,
+      "step": 2404
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.207906471746926,
+      "learning_rate": 1.9962784704299907e-05,
+      "loss": 0.7565,
+      "step": 2405
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5879264880926661,
+      "learning_rate": 1.9962718906604365e-05,
+      "loss": 0.5584,
+      "step": 2406
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4098111491109355,
+      "learning_rate": 1.9962653050902722e-05,
+      "loss": 0.6886,
+      "step": 2407
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7086622242085205,
+      "learning_rate": 1.996258713719536e-05,
+      "loss": 0.7633,
+      "step": 2408
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.900406003571686,
+      "learning_rate": 1.9962521165482657e-05,
+      "loss": 0.8104,
+      "step": 2409
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5714194596607147,
+      "learning_rate": 1.9962455135765003e-05,
+      "loss": 0.7255,
+      "step": 2410
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.816515245927938,
+      "learning_rate": 1.996238904804278e-05,
+      "loss": 0.5869,
+      "step": 2411
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8319146053504325,
+      "learning_rate": 1.9962322902316376e-05,
+      "loss": 0.7262,
+      "step": 2412
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5942779022759165,
+      "learning_rate": 1.9962256698586174e-05,
+      "loss": 0.5607,
+      "step": 2413
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5890836941311448,
+      "learning_rate": 1.996219043685256e-05,
+      "loss": 0.8625,
+      "step": 2414
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.463521610034693,
+      "learning_rate": 1.9962124117115914e-05,
+      "loss": 0.8691,
+      "step": 2415
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.781793976411732,
+      "learning_rate": 1.9962057739376634e-05,
+      "loss": 0.7814,
+      "step": 2416
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5253383075663731,
+      "learning_rate": 1.9961991303635096e-05,
+      "loss": 0.7231,
+      "step": 2417
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.656901851693396,
+      "learning_rate": 1.9961924809891693e-05,
+      "loss": 0.7327,
+      "step": 2418
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9842762159103504,
+      "learning_rate": 1.996185825814681e-05,
+      "loss": 0.6745,
+      "step": 2419
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.169028489247297,
+      "learning_rate": 1.9961791648400836e-05,
+      "loss": 0.7638,
+      "step": 2420
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5727270490275614,
+      "learning_rate": 1.9961724980654154e-05,
+      "loss": 0.5053,
+      "step": 2421
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.149428210811486,
+      "learning_rate": 1.9961658254907153e-05,
+      "loss": 0.6925,
+      "step": 2422
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3369321326006705,
+      "learning_rate": 1.9961591471160228e-05,
+      "loss": 0.8436,
+      "step": 2423
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5919563942461296,
+      "learning_rate": 1.9961524629413765e-05,
+      "loss": 0.5443,
+      "step": 2424
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.701250587947097,
+      "learning_rate": 1.9961457729668148e-05,
+      "loss": 0.7704,
+      "step": 2425
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6037635135541513,
+      "learning_rate": 1.9961390771923772e-05,
+      "loss": 0.8069,
+      "step": 2426
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5067926562287768,
+      "learning_rate": 1.9961323756181024e-05,
+      "loss": 0.7293,
+      "step": 2427
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8328498804034217,
+      "learning_rate": 1.9961256682440298e-05,
+      "loss": 0.7139,
+      "step": 2428
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.506339078460287,
+      "learning_rate": 1.996118955070198e-05,
+      "loss": 0.78,
+      "step": 2429
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4921060388331586,
+      "learning_rate": 1.9961122360966463e-05,
+      "loss": 0.771,
+      "step": 2430
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6609642170684502,
+      "learning_rate": 1.9961055113234137e-05,
+      "loss": 0.828,
+      "step": 2431
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8897673905728662,
+      "learning_rate": 1.9960987807505398e-05,
+      "loss": 0.8212,
+      "step": 2432
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.040331717590839,
+      "learning_rate": 1.996092044378063e-05,
+      "loss": 0.6782,
+      "step": 2433
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.720317077625553,
+      "learning_rate": 1.996085302206023e-05,
+      "loss": 0.9134,
+      "step": 2434
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0398258804637788,
+      "learning_rate": 1.9960785542344594e-05,
+      "loss": 0.7099,
+      "step": 2435
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5915782330908046,
+      "learning_rate": 1.996071800463411e-05,
+      "loss": 0.7386,
+      "step": 2436
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7664509098094507,
+      "learning_rate": 1.996065040892917e-05,
+      "loss": 0.6036,
+      "step": 2437
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4011865065000708,
+      "learning_rate": 1.9960582755230172e-05,
+      "loss": 0.6992,
+      "step": 2438
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4700227220221953,
+      "learning_rate": 1.9960515043537504e-05,
+      "loss": 0.8007,
+      "step": 2439
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7362792405245644,
+      "learning_rate": 1.9960447273851567e-05,
+      "loss": 0.761,
+      "step": 2440
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.639645089510075,
+      "learning_rate": 1.9960379446172754e-05,
+      "loss": 0.7837,
+      "step": 2441
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6286768537514993,
+      "learning_rate": 1.9960311560501457e-05,
+      "loss": 0.768,
+      "step": 2442
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1815343702095302,
+      "learning_rate": 1.996024361683807e-05,
+      "loss": 0.7257,
+      "step": 2443
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.579919262807439,
+      "learning_rate": 1.996017561518299e-05,
+      "loss": 0.74,
+      "step": 2444
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7560368239372095,
+      "learning_rate": 1.996010755553662e-05,
+      "loss": 0.8441,
+      "step": 2445
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4179117170086653,
+      "learning_rate": 1.996003943789935e-05,
+      "loss": 0.6314,
+      "step": 2446
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6845172472297767,
+      "learning_rate": 1.9959971262271574e-05,
+      "loss": 0.7581,
+      "step": 2447
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.877996259509275,
+      "learning_rate": 1.9959903028653695e-05,
+      "loss": 0.6688,
+      "step": 2448
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9135290163288052,
+      "learning_rate": 1.9959834737046107e-05,
+      "loss": 0.8206,
+      "step": 2449
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7485229162826674,
+      "learning_rate": 1.9959766387449203e-05,
+      "loss": 0.7824,
+      "step": 2450
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6153821269043973,
+      "learning_rate": 1.995969797986339e-05,
+      "loss": 0.8814,
+      "step": 2451
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.3514510808110956,
+      "learning_rate": 1.9959629514289065e-05,
+      "loss": 0.6683,
+      "step": 2452
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6675114654428904,
+      "learning_rate": 1.995956099072662e-05,
+      "loss": 0.7653,
+      "step": 2453
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.563139159606656,
+      "learning_rate": 1.995949240917646e-05,
+      "loss": 0.7059,
+      "step": 2454
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5262357074238027,
+      "learning_rate": 1.9959423769638978e-05,
+      "loss": 0.6702,
+      "step": 2455
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5579836946627024,
+      "learning_rate": 1.995935507211458e-05,
+      "loss": 0.8981,
+      "step": 2456
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.711946815245001,
+      "learning_rate": 1.9959286316603662e-05,
+      "loss": 0.7951,
+      "step": 2457
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4855306277386084,
+      "learning_rate": 1.995921750310663e-05,
+      "loss": 0.651,
+      "step": 2458
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.066704790861894,
+      "learning_rate": 1.9959148631623878e-05,
+      "loss": 0.775,
+      "step": 2459
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6136559112388744,
+      "learning_rate": 1.995907970215581e-05,
+      "loss": 0.8057,
+      "step": 2460
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6037336350770224,
+      "learning_rate": 1.9959010714702827e-05,
+      "loss": 0.877,
+      "step": 2461
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6860853157719555,
+      "learning_rate": 1.995894166926533e-05,
+      "loss": 0.6981,
+      "step": 2462
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9548627344457534,
+      "learning_rate": 1.9958872565843724e-05,
+      "loss": 0.7109,
+      "step": 2463
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4712872086851876,
+      "learning_rate": 1.9958803404438406e-05,
+      "loss": 0.7125,
+      "step": 2464
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.790043390920336,
+      "learning_rate": 1.9958734185049782e-05,
+      "loss": 0.8127,
+      "step": 2465
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8518459140272425,
+      "learning_rate": 1.9958664907678257e-05,
+      "loss": 0.742,
+      "step": 2466
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4596846412754296,
+      "learning_rate": 1.995859557232423e-05,
+      "loss": 0.6995,
+      "step": 2467
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7162638023522392,
+      "learning_rate": 1.9958526178988107e-05,
+      "loss": 0.9389,
+      "step": 2468
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1280761812576405,
+      "learning_rate": 1.9958456727670296e-05,
+      "loss": 0.7907,
+      "step": 2469
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7019045486663427,
+      "learning_rate": 1.9958387218371192e-05,
+      "loss": 0.8013,
+      "step": 2470
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5267398917699921,
+      "learning_rate": 1.9958317651091206e-05,
+      "loss": 0.8768,
+      "step": 2471
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.118853256130706,
+      "learning_rate": 1.9958248025830743e-05,
+      "loss": 0.6723,
+      "step": 2472
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7062130641305153,
+      "learning_rate": 1.9958178342590205e-05,
+      "loss": 0.8137,
+      "step": 2473
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6917276972441557,
+      "learning_rate": 1.995810860137e-05,
+      "loss": 0.8185,
+      "step": 2474
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8277727106449888,
+      "learning_rate": 1.9958038802170537e-05,
+      "loss": 0.8002,
+      "step": 2475
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8621032291570607,
+      "learning_rate": 1.995796894499222e-05,
+      "loss": 0.8245,
+      "step": 2476
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7218634205087149,
+      "learning_rate": 1.9957899029835453e-05,
+      "loss": 0.7074,
+      "step": 2477
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2826441784437512,
+      "learning_rate": 1.9957829056700643e-05,
+      "loss": 0.8626,
+      "step": 2478
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4484308668751666,
+      "learning_rate": 1.9957759025588203e-05,
+      "loss": 0.7171,
+      "step": 2479
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3316500060185574,
+      "learning_rate": 1.9957688936498537e-05,
+      "loss": 0.7693,
+      "step": 2480
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.776122323507968,
+      "learning_rate": 1.9957618789432053e-05,
+      "loss": 0.6992,
+      "step": 2481
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5613563725767547,
+      "learning_rate": 1.995754858438916e-05,
+      "loss": 0.8558,
+      "step": 2482
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5807682360966613,
+      "learning_rate": 1.9957478321370266e-05,
+      "loss": 0.8861,
+      "step": 2483
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.815592586819593,
+      "learning_rate": 1.995740800037578e-05,
+      "loss": 0.6919,
+      "step": 2484
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4533446779439971,
+      "learning_rate": 1.9957337621406112e-05,
+      "loss": 0.7532,
+      "step": 2485
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8385489717916483,
+      "learning_rate": 1.9957267184461674e-05,
+      "loss": 0.7503,
+      "step": 2486
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0900299054391396,
+      "learning_rate": 1.9957196689542873e-05,
+      "loss": 0.7212,
+      "step": 2487
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5817361631610796,
+      "learning_rate": 1.995712613665012e-05,
+      "loss": 0.7629,
+      "step": 2488
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5138977287787792,
+      "learning_rate": 1.995705552578383e-05,
+      "loss": 0.846,
+      "step": 2489
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5414033525401198,
+      "learning_rate": 1.9956984856944403e-05,
+      "loss": 0.8554,
+      "step": 2490
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.391729788498506,
+      "learning_rate": 1.9956914130132264e-05,
+      "loss": 0.8071,
+      "step": 2491
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.406163979464045,
+      "learning_rate": 1.9956843345347817e-05,
+      "loss": 0.8023,
+      "step": 2492
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4274212485736408,
+      "learning_rate": 1.9956772502591473e-05,
+      "loss": 0.6841,
+      "step": 2493
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.715107963366491,
+      "learning_rate": 1.9956701601863653e-05,
+      "loss": 0.8682,
+      "step": 2494
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.665618324636142,
+      "learning_rate": 1.9956630643164758e-05,
+      "loss": 0.8623,
+      "step": 2495
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6305407818333777,
+      "learning_rate": 1.995655962649521e-05,
+      "loss": 0.7531,
+      "step": 2496
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.599984590356335,
+      "learning_rate": 1.995648855185542e-05,
+      "loss": 0.881,
+      "step": 2497
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6296951206035302,
+      "learning_rate": 1.9956417419245802e-05,
+      "loss": 0.7251,
+      "step": 2498
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7889905988744512,
+      "learning_rate": 1.995634622866677e-05,
+      "loss": 0.6963,
+      "step": 2499
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6036413332580739,
+      "learning_rate": 1.9956274980118736e-05,
+      "loss": 0.4829,
+      "step": 2500
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8020167952422328,
+      "learning_rate": 1.9956203673602116e-05,
+      "loss": 0.5515,
+      "step": 2501
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.51330150604023,
+      "learning_rate": 1.995613230911733e-05,
+      "loss": 0.672,
+      "step": 2502
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6210285369520168,
+      "learning_rate": 1.9956060886664786e-05,
+      "loss": 0.7829,
+      "step": 2503
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6158015315342955,
+      "learning_rate": 1.9955989406244906e-05,
+      "loss": 0.813,
+      "step": 2504
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.403229984488162,
+      "learning_rate": 1.9955917867858102e-05,
+      "loss": 0.8121,
+      "step": 2505
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8181370475031537,
+      "learning_rate": 1.9955846271504792e-05,
+      "loss": 0.6357,
+      "step": 2506
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5124375653482303,
+      "learning_rate": 1.9955774617185395e-05,
+      "loss": 0.7787,
+      "step": 2507
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.730633327820172,
+      "learning_rate": 1.9955702904900325e-05,
+      "loss": 0.7734,
+      "step": 2508
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4601237831797653,
+      "learning_rate": 1.995563113465e-05,
+      "loss": 0.7726,
+      "step": 2509
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2095085342248402,
+      "learning_rate": 1.9955559306434843e-05,
+      "loss": 0.6815,
+      "step": 2510
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8494322508450254,
+      "learning_rate": 1.995548742025526e-05,
+      "loss": 0.8655,
+      "step": 2511
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5337808105543873,
+      "learning_rate": 1.9955415476111683e-05,
+      "loss": 0.6214,
+      "step": 2512
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8097692522509816,
+      "learning_rate": 1.9955343474004526e-05,
+      "loss": 0.6434,
+      "step": 2513
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7793084732998241,
+      "learning_rate": 1.9955271413934203e-05,
+      "loss": 0.8192,
+      "step": 2514
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6053137696093536,
+      "learning_rate": 1.9955199295901144e-05,
+      "loss": 0.7035,
+      "step": 2515
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4504391203582798,
+      "learning_rate": 1.9955127119905756e-05,
+      "loss": 0.916,
+      "step": 2516
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5474438101925978,
+      "learning_rate": 1.995505488594847e-05,
+      "loss": 0.7636,
+      "step": 2517
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4503989348746116,
+      "learning_rate": 1.9954982594029703e-05,
+      "loss": 0.7052,
+      "step": 2518
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5956154664637398,
+      "learning_rate": 1.9954910244149873e-05,
+      "loss": 0.497,
+      "step": 2519
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5580718774814715,
+      "learning_rate": 1.9954837836309407e-05,
+      "loss": 0.7496,
+      "step": 2520
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5857298423205732,
+      "learning_rate": 1.9954765370508723e-05,
+      "loss": 0.7062,
+      "step": 2521
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7456551941378269,
+      "learning_rate": 1.995469284674824e-05,
+      "loss": 0.7263,
+      "step": 2522
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.565450119489368,
+      "learning_rate": 1.995462026502839e-05,
+      "loss": 0.7684,
+      "step": 2523
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3937343150810249,
+      "learning_rate": 1.9954547625349584e-05,
+      "loss": 0.6324,
+      "step": 2524
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6436617023229665,
+      "learning_rate": 1.9954474927712252e-05,
+      "loss": 0.7477,
+      "step": 2525
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4782578938431987,
+      "learning_rate": 1.9954402172116816e-05,
+      "loss": 0.7419,
+      "step": 2526
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9778771243652367,
+      "learning_rate": 1.99543293585637e-05,
+      "loss": 0.7751,
+      "step": 2527
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.531661537131277,
+      "learning_rate": 1.9954256487053326e-05,
+      "loss": 0.7857,
+      "step": 2528
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7034002977817626,
+      "learning_rate": 1.995418355758612e-05,
+      "loss": 0.8365,
+      "step": 2529
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3928833596937578,
+      "learning_rate": 1.99541105701625e-05,
+      "loss": 0.6534,
+      "step": 2530
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6776588908887655,
+      "learning_rate": 1.99540375247829e-05,
+      "loss": 0.7946,
+      "step": 2531
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4298014944073674,
+      "learning_rate": 1.9953964421447745e-05,
+      "loss": 0.8142,
+      "step": 2532
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.882690135741469,
+      "learning_rate": 1.995389126015746e-05,
+      "loss": 0.8091,
+      "step": 2533
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.362138977705478,
+      "learning_rate": 1.995381804091246e-05,
+      "loss": 0.8613,
+      "step": 2534
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6047227890900064,
+      "learning_rate": 1.9953744763713188e-05,
+      "loss": 0.8254,
+      "step": 2535
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5272869172000516,
+      "learning_rate": 1.995367142856006e-05,
+      "loss": 0.7606,
+      "step": 2536
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.533892696684252,
+      "learning_rate": 1.9953598035453505e-05,
+      "loss": 0.7353,
+      "step": 2537
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2746247256423446,
+      "learning_rate": 1.995352458439395e-05,
+      "loss": 0.6982,
+      "step": 2538
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2929113433245196,
+      "learning_rate": 1.9953451075381826e-05,
+      "loss": 0.8122,
+      "step": 2539
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6379280635875142,
+      "learning_rate": 1.995337750841756e-05,
+      "loss": 0.8797,
+      "step": 2540
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8154041249270767,
+      "learning_rate": 1.9953303883501577e-05,
+      "loss": 0.8354,
+      "step": 2541
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.454709745976317,
+      "learning_rate": 1.9953230200634305e-05,
+      "loss": 0.767,
+      "step": 2542
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5573305536497302,
+      "learning_rate": 1.995315645981618e-05,
+      "loss": 0.8224,
+      "step": 2543
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3183794012027765,
+      "learning_rate": 1.9953082661047624e-05,
+      "loss": 0.736,
+      "step": 2544
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.50302342827547,
+      "learning_rate": 1.995300880432907e-05,
+      "loss": 0.7054,
+      "step": 2545
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.7456891390768567,
+      "learning_rate": 1.9952934889660947e-05,
+      "loss": 0.6863,
+      "step": 2546
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6431529135075995,
+      "learning_rate": 1.9952860917043687e-05,
+      "loss": 0.8588,
+      "step": 2547
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5263598586661276,
+      "learning_rate": 1.9952786886477718e-05,
+      "loss": 0.8079,
+      "step": 2548
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0946992078047626,
+      "learning_rate": 1.9952712797963477e-05,
+      "loss": 0.8624,
+      "step": 2549
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4019748798261276,
+      "learning_rate": 1.995263865150139e-05,
+      "loss": 0.8754,
+      "step": 2550
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3388833312503303,
+      "learning_rate": 1.995256444709189e-05,
+      "loss": 0.8261,
+      "step": 2551
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.922167544498609,
+      "learning_rate": 1.9952490184735404e-05,
+      "loss": 0.8036,
+      "step": 2552
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5831274418191053,
+      "learning_rate": 1.9952415864432373e-05,
+      "loss": 0.828,
+      "step": 2553
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5263519625844848,
+      "learning_rate": 1.9952341486183228e-05,
+      "loss": 0.8008,
+      "step": 2554
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6192160630494157,
+      "learning_rate": 1.9952267049988393e-05,
+      "loss": 0.7069,
+      "step": 2555
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5947105453833834,
+      "learning_rate": 1.9952192555848312e-05,
+      "loss": 0.4863,
+      "step": 2556
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7512801086700183,
+      "learning_rate": 1.995211800376342e-05,
+      "loss": 0.7891,
+      "step": 2557
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5153020083392323,
+      "learning_rate": 1.995204339373414e-05,
+      "loss": 0.7682,
+      "step": 2558
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4330634141215235,
+      "learning_rate": 1.9951968725760916e-05,
+      "loss": 0.7332,
+      "step": 2559
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5838339508142707,
+      "learning_rate": 1.9951893999844178e-05,
+      "loss": 0.7688,
+      "step": 2560
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5278264169069415,
+      "learning_rate": 1.995181921598436e-05,
+      "loss": 0.7791,
+      "step": 2561
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4191719682362194,
+      "learning_rate": 1.9951744374181905e-05,
+      "loss": 0.7721,
+      "step": 2562
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5777986370658637,
+      "learning_rate": 1.9951669474437244e-05,
+      "loss": 0.7666,
+      "step": 2563
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5043277926098344,
+      "learning_rate": 1.9951594516750807e-05,
+      "loss": 0.9435,
+      "step": 2564
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.461952390541416,
+      "learning_rate": 1.995151950112304e-05,
+      "loss": 0.7356,
+      "step": 2565
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5180623080560858,
+      "learning_rate": 1.9951444427554375e-05,
+      "loss": 0.7607,
+      "step": 2566
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7664230940651058,
+      "learning_rate": 1.9951369296045247e-05,
+      "loss": 0.7713,
+      "step": 2567
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5085082470549285,
+      "learning_rate": 1.9951294106596098e-05,
+      "loss": 0.7758,
+      "step": 2568
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6160780806692563,
+      "learning_rate": 1.9951218859207368e-05,
+      "loss": 0.7478,
+      "step": 2569
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5094085532273074,
+      "learning_rate": 1.995114355387949e-05,
+      "loss": 0.7831,
+      "step": 2570
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7493176975094678,
+      "learning_rate": 1.99510681906129e-05,
+      "loss": 0.7112,
+      "step": 2571
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5717766921816627,
+      "learning_rate": 1.9950992769408043e-05,
+      "loss": 0.8609,
+      "step": 2572
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6638649573101545,
+      "learning_rate": 1.9950917290265356e-05,
+      "loss": 0.8395,
+      "step": 2573
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7404241392030648,
+      "learning_rate": 1.9950841753185276e-05,
+      "loss": 0.8098,
+      "step": 2574
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2949471537416384,
+      "learning_rate": 1.995076615816825e-05,
+      "loss": 0.7473,
+      "step": 2575
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5352458525225234,
+      "learning_rate": 1.9950690505214707e-05,
+      "loss": 0.5985,
+      "step": 2576
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9578389197357315,
+      "learning_rate": 1.9950614794325096e-05,
+      "loss": 0.7007,
+      "step": 2577
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7302030510892898,
+      "learning_rate": 1.9950539025499857e-05,
+      "loss": 0.6751,
+      "step": 2578
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4691623963971663,
+      "learning_rate": 1.995046319873943e-05,
+      "loss": 0.7926,
+      "step": 2579
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.1607437140971224,
+      "learning_rate": 1.9950387314044254e-05,
+      "loss": 0.877,
+      "step": 2580
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4108839705675709,
+      "learning_rate": 1.995031137141477e-05,
+      "loss": 0.6968,
+      "step": 2581
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2883431037241106,
+      "learning_rate": 1.9950235370851428e-05,
+      "loss": 0.843,
+      "step": 2582
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7443019570829363,
+      "learning_rate": 1.9950159312354664e-05,
+      "loss": 0.9057,
+      "step": 2583
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5252060878360267,
+      "learning_rate": 1.9950083195924922e-05,
+      "loss": 0.7535,
+      "step": 2584
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3993099695566866,
+      "learning_rate": 1.9950007021562643e-05,
+      "loss": 0.899,
+      "step": 2585
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5453596969870724,
+      "learning_rate": 1.9949930789268274e-05,
+      "loss": 0.7968,
+      "step": 2586
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8137963728884434,
+      "learning_rate": 1.994985449904226e-05,
+      "loss": 0.8,
+      "step": 2587
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5641461101746983,
+      "learning_rate": 1.994977815088504e-05,
+      "loss": 0.7442,
+      "step": 2588
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7036260780492505,
+      "learning_rate": 1.994970174479707e-05,
+      "loss": 0.8538,
+      "step": 2589
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4393870990588038,
+      "learning_rate": 1.994962528077878e-05,
+      "loss": 0.8453,
+      "step": 2590
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5839469236488894,
+      "learning_rate": 1.9949548758830618e-05,
+      "loss": 0.7494,
+      "step": 2591
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.62127903703812,
+      "learning_rate": 1.9949472178953036e-05,
+      "loss": 0.8254,
+      "step": 2592
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4466801312024762,
+      "learning_rate": 1.9949395541146475e-05,
+      "loss": 0.7716,
+      "step": 2593
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.638933685807484,
+      "learning_rate": 1.994931884541139e-05,
+      "loss": 0.8572,
+      "step": 2594
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.474492071916425,
+      "learning_rate": 1.9949242091748217e-05,
+      "loss": 0.7266,
+      "step": 2595
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6132622823436721,
+      "learning_rate": 1.9949165280157404e-05,
+      "loss": 0.5699,
+      "step": 2596
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1155358277103176,
+      "learning_rate": 1.9949088410639407e-05,
+      "loss": 0.7276,
+      "step": 2597
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.880751009519961,
+      "learning_rate": 1.9949011483194663e-05,
+      "loss": 0.7537,
+      "step": 2598
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7118019571471965,
+      "learning_rate": 1.9948934497823626e-05,
+      "loss": 0.8573,
+      "step": 2599
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.677292580685652,
+      "learning_rate": 1.994885745452674e-05,
+      "loss": 0.665,
+      "step": 2600
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.679966870108861,
+      "learning_rate": 1.994878035330446e-05,
+      "loss": 0.6493,
+      "step": 2601
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6826613290727346,
+      "learning_rate": 1.9948703194157227e-05,
+      "loss": 0.7482,
+      "step": 2602
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4907211645029383,
+      "learning_rate": 1.9948625977085497e-05,
+      "loss": 0.7297,
+      "step": 2603
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4501227649691104,
+      "learning_rate": 1.9948548702089715e-05,
+      "loss": 0.6686,
+      "step": 2604
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6659783178831442,
+      "learning_rate": 1.9948471369170336e-05,
+      "loss": 0.8637,
+      "step": 2605
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6018524474704479,
+      "learning_rate": 1.9948393978327802e-05,
+      "loss": 0.6498,
+      "step": 2606
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4501873453239298,
+      "learning_rate": 1.9948316529562573e-05,
+      "loss": 0.8725,
+      "step": 2607
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8634092950194105,
+      "learning_rate": 1.9948239022875094e-05,
+      "loss": 0.7278,
+      "step": 2608
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.65857504708531,
+      "learning_rate": 1.9948161458265817e-05,
+      "loss": 0.6513,
+      "step": 2609
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5466462906912175,
+      "learning_rate": 1.9948083835735198e-05,
+      "loss": 0.771,
+      "step": 2610
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7064439695898377,
+      "learning_rate": 1.994800615528368e-05,
+      "loss": 0.8086,
+      "step": 2611
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2059470488856356,
+      "learning_rate": 1.9947928416911723e-05,
+      "loss": 0.7349,
+      "step": 2612
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.2011356995675,
+      "learning_rate": 1.994785062061978e-05,
+      "loss": 0.8029,
+      "step": 2613
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4008895862231405,
+      "learning_rate": 1.9947772766408298e-05,
+      "loss": 0.5257,
+      "step": 2614
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9360289511935938,
+      "learning_rate": 1.9947694854277734e-05,
+      "loss": 0.805,
+      "step": 2615
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5666866386909617,
+      "learning_rate": 1.9947616884228545e-05,
+      "loss": 0.6277,
+      "step": 2616
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5685052819700431,
+      "learning_rate": 1.9947538856261177e-05,
+      "loss": 0.7847,
+      "step": 2617
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5532916647161448,
+      "learning_rate": 1.994746077037609e-05,
+      "loss": 0.798,
+      "step": 2618
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7119252936460083,
+      "learning_rate": 1.9947382626573736e-05,
+      "loss": 0.8176,
+      "step": 2619
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4664951692384451,
+      "learning_rate": 1.994730442485457e-05,
+      "loss": 0.6839,
+      "step": 2620
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6309099501667,
+      "learning_rate": 1.994722616521905e-05,
+      "loss": 0.8542,
+      "step": 2621
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5879685924578383,
+      "learning_rate": 1.994714784766763e-05,
+      "loss": 0.4936,
+      "step": 2622
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7630118329328783,
+      "learning_rate": 1.9947069472200766e-05,
+      "loss": 0.7226,
+      "step": 2623
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.586656799776863,
+      "learning_rate": 1.9946991038818916e-05,
+      "loss": 0.5178,
+      "step": 2624
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.708440060955908,
+      "learning_rate": 1.994691254752253e-05,
+      "loss": 0.925,
+      "step": 2625
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8318146636490884,
+      "learning_rate": 1.9946833998312077e-05,
+      "loss": 0.8138,
+      "step": 2626
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6220833428088879,
+      "learning_rate": 1.9946755391188e-05,
+      "loss": 0.8307,
+      "step": 2627
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6142577198748123,
+      "learning_rate": 1.994667672615077e-05,
+      "loss": 0.6786,
+      "step": 2628
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6511210493558375,
+      "learning_rate": 1.9946598003200837e-05,
+      "loss": 0.7467,
+      "step": 2629
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5217606495621352,
+      "learning_rate": 1.994651922233866e-05,
+      "loss": 0.9055,
+      "step": 2630
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.479141193283916,
+      "learning_rate": 1.9946440383564703e-05,
+      "loss": 0.8449,
+      "step": 2631
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.596176396840092,
+      "learning_rate": 1.9946361486879416e-05,
+      "loss": 0.7918,
+      "step": 2632
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4826509072084246,
+      "learning_rate": 1.9946282532283266e-05,
+      "loss": 0.8622,
+      "step": 2633
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.435107624709241,
+      "learning_rate": 1.994620351977671e-05,
+      "loss": 0.7438,
+      "step": 2634
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.569109720140646,
+      "learning_rate": 1.9946124449360205e-05,
+      "loss": 0.8109,
+      "step": 2635
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8757449036341427,
+      "learning_rate": 1.9946045321034223e-05,
+      "loss": 0.7809,
+      "step": 2636
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5723213183993185,
+      "learning_rate": 1.9945966134799207e-05,
+      "loss": 0.7212,
+      "step": 2637
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.341080690121896,
+      "learning_rate": 1.9945886890655632e-05,
+      "loss": 0.5587,
+      "step": 2638
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4384910607675576,
+      "learning_rate": 1.9945807588603954e-05,
+      "loss": 0.8422,
+      "step": 2639
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.528447935302827,
+      "learning_rate": 1.9945728228644636e-05,
+      "loss": 0.8154,
+      "step": 2640
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5614848554474792,
+      "learning_rate": 1.994564881077814e-05,
+      "loss": 0.6961,
+      "step": 2641
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.595322587356611,
+      "learning_rate": 1.9945569335004925e-05,
+      "loss": 0.714,
+      "step": 2642
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.533487814315676,
+      "learning_rate": 1.994548980132546e-05,
+      "loss": 0.8079,
+      "step": 2643
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.363624179925382,
+      "learning_rate": 1.9945410209740202e-05,
+      "loss": 0.7703,
+      "step": 2644
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3645214762739692,
+      "learning_rate": 1.994533056024962e-05,
+      "loss": 0.6895,
+      "step": 2645
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.429734637254975,
+      "learning_rate": 1.994525085285417e-05,
+      "loss": 0.798,
+      "step": 2646
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6236564535168263,
+      "learning_rate": 1.9945171087554324e-05,
+      "loss": 0.8159,
+      "step": 2647
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6245085156824508,
+      "learning_rate": 1.9945091264350545e-05,
+      "loss": 0.652,
+      "step": 2648
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6976701641844263,
+      "learning_rate": 1.9945011383243296e-05,
+      "loss": 0.8404,
+      "step": 2649
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6108997274813188,
+      "learning_rate": 1.994493144423304e-05,
+      "loss": 0.7673,
+      "step": 2650
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.7089448937096923,
+      "learning_rate": 1.9944851447320243e-05,
+      "loss": 0.8696,
+      "step": 2651
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5112562536534233,
+      "learning_rate": 1.9944771392505374e-05,
+      "loss": 0.7606,
+      "step": 2652
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7514068749140557,
+      "learning_rate": 1.99446912797889e-05,
+      "loss": 0.9769,
+      "step": 2653
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4237923923847036,
+      "learning_rate": 1.9944611109171284e-05,
+      "loss": 0.7334,
+      "step": 2654
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6191099024686144,
+      "learning_rate": 1.9944530880652993e-05,
+      "loss": 0.8067,
+      "step": 2655
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7155665051907354,
+      "learning_rate": 1.9944450594234492e-05,
+      "loss": 0.7268,
+      "step": 2656
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6645480392767216,
+      "learning_rate": 1.9944370249916254e-05,
+      "loss": 0.7988,
+      "step": 2657
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7202095853450818,
+      "learning_rate": 1.9944289847698744e-05,
+      "loss": 0.7204,
+      "step": 2658
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5967613784244616,
+      "learning_rate": 1.994420938758243e-05,
+      "loss": 0.6463,
+      "step": 2659
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9246257059885232,
+      "learning_rate": 1.9944128869567782e-05,
+      "loss": 0.7395,
+      "step": 2660
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5252935260195537,
+      "learning_rate": 1.9944048293655266e-05,
+      "loss": 0.8099,
+      "step": 2661
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5257100050525534,
+      "learning_rate": 1.994396765984535e-05,
+      "loss": 0.8959,
+      "step": 2662
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4997947183637659,
+      "learning_rate": 1.9943886968138512e-05,
+      "loss": 0.7378,
+      "step": 2663
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9096747146701634,
+      "learning_rate": 1.9943806218535212e-05,
+      "loss": 0.6616,
+      "step": 2664
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5811065019737378,
+      "learning_rate": 1.9943725411035926e-05,
+      "loss": 0.855,
+      "step": 2665
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3539987877639996,
+      "learning_rate": 1.994364454564112e-05,
+      "loss": 0.8005,
+      "step": 2666
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5987661366140533,
+      "learning_rate": 1.9943563622351267e-05,
+      "loss": 0.6712,
+      "step": 2667
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.704036758824762,
+      "learning_rate": 1.994348264116684e-05,
+      "loss": 0.6625,
+      "step": 2668
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.012209499142572,
+      "learning_rate": 1.994340160208831e-05,
+      "loss": 0.8523,
+      "step": 2669
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5802973165793912,
+      "learning_rate": 1.994332050511615e-05,
+      "loss": 0.8151,
+      "step": 2670
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6113206255756715,
+      "learning_rate": 1.9943239350250823e-05,
+      "loss": 0.4824,
+      "step": 2671
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5913133241277924,
+      "learning_rate": 1.9943158137492815e-05,
+      "loss": 0.8447,
+      "step": 2672
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.489725835994004,
+      "learning_rate": 1.994307686684259e-05,
+      "loss": 0.7654,
+      "step": 2673
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8070890216702193,
+      "learning_rate": 1.9942995538300624e-05,
+      "loss": 0.6833,
+      "step": 2674
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7658802497899464,
+      "learning_rate": 1.994291415186739e-05,
+      "loss": 0.74,
+      "step": 2675
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.197171988233852,
+      "learning_rate": 1.994283270754336e-05,
+      "loss": 0.8785,
+      "step": 2676
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6081269777609841,
+      "learning_rate": 1.994275120532901e-05,
+      "loss": 0.5391,
+      "step": 2677
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.302335040782696,
+      "learning_rate": 1.994266964522482e-05,
+      "loss": 0.702,
+      "step": 2678
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6068854314059753,
+      "learning_rate": 1.9942588027231254e-05,
+      "loss": 0.4724,
+      "step": 2679
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9290200122376833,
+      "learning_rate": 1.9942506351348796e-05,
+      "loss": 0.8271,
+      "step": 2680
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4289589534603833,
+      "learning_rate": 1.9942424617577915e-05,
+      "loss": 0.7505,
+      "step": 2681
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5317193890071694,
+      "learning_rate": 1.994234282591909e-05,
+      "loss": 0.7817,
+      "step": 2682
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4938224620663687,
+      "learning_rate": 1.9942260976372802e-05,
+      "loss": 0.6338,
+      "step": 2683
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6164820566131244,
+      "learning_rate": 1.994217906893952e-05,
+      "loss": 0.7402,
+      "step": 2684
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6307920375555713,
+      "learning_rate": 1.9942097103619728e-05,
+      "loss": 0.6142,
+      "step": 2685
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7103715931603323,
+      "learning_rate": 1.9942015080413894e-05,
+      "loss": 0.6714,
+      "step": 2686
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.45866559681623,
+      "learning_rate": 1.99419329993225e-05,
+      "loss": 0.6596,
+      "step": 2687
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4285050903136711,
+      "learning_rate": 1.9941850860346024e-05,
+      "loss": 0.7187,
+      "step": 2688
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5894755497556636,
+      "learning_rate": 1.9941768663484948e-05,
+      "loss": 0.7906,
+      "step": 2689
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.508895504462178,
+      "learning_rate": 1.9941686408739748e-05,
+      "loss": 0.8379,
+      "step": 2690
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4511827044536976,
+      "learning_rate": 1.99416040961109e-05,
+      "loss": 0.7843,
+      "step": 2691
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5359441671644838,
+      "learning_rate": 1.9941521725598888e-05,
+      "loss": 0.6819,
+      "step": 2692
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.304842331857128,
+      "learning_rate": 1.9941439297204185e-05,
+      "loss": 0.7791,
+      "step": 2693
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4607540880328924,
+      "learning_rate": 1.994135681092728e-05,
+      "loss": 0.677,
+      "step": 2694
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6781226185969027,
+      "learning_rate": 1.9941274266768643e-05,
+      "loss": 0.6833,
+      "step": 2695
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.576419600124443,
+      "learning_rate": 1.9941191664728763e-05,
+      "loss": 0.7802,
+      "step": 2696
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7056755137037989,
+      "learning_rate": 1.9941109004808117e-05,
+      "loss": 0.7188,
+      "step": 2697
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1078738017768455,
+      "learning_rate": 1.9941026287007186e-05,
+      "loss": 0.7533,
+      "step": 2698
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6165136568163498,
+      "learning_rate": 1.9940943511326453e-05,
+      "loss": 0.7209,
+      "step": 2699
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4629116741015413,
+      "learning_rate": 1.9940860677766404e-05,
+      "loss": 0.8252,
+      "step": 2700
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4513137494009094,
+      "learning_rate": 1.9940777786327513e-05,
+      "loss": 0.8639,
+      "step": 2701
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.510143444406966,
+      "learning_rate": 1.9940694837010265e-05,
+      "loss": 0.7644,
+      "step": 2702
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.504570078392677,
+      "learning_rate": 1.994061182981515e-05,
+      "loss": 0.7009,
+      "step": 2703
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.028601281890227,
+      "learning_rate": 1.994052876474264e-05,
+      "loss": 0.8027,
+      "step": 2704
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6243007197471534,
+      "learning_rate": 1.9940445641793228e-05,
+      "loss": 0.5489,
+      "step": 2705
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5239174338936106,
+      "learning_rate": 1.994036246096739e-05,
+      "loss": 0.7149,
+      "step": 2706
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.444016828489519,
+      "learning_rate": 1.9940279222265618e-05,
+      "loss": 0.6677,
+      "step": 2707
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0863988112642216,
+      "learning_rate": 1.9940195925688395e-05,
+      "loss": 0.6151,
+      "step": 2708
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6307572354326556,
+      "learning_rate": 1.9940112571236203e-05,
+      "loss": 0.6766,
+      "step": 2709
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5708711383995937,
+      "learning_rate": 1.9940029158909525e-05,
+      "loss": 0.7751,
+      "step": 2710
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6480253963965776,
+      "learning_rate": 1.9939945688708854e-05,
+      "loss": 0.7013,
+      "step": 2711
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4571456349730958,
+      "learning_rate": 1.9939862160634674e-05,
+      "loss": 0.7659,
+      "step": 2712
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.509196765905573,
+      "learning_rate": 1.9939778574687465e-05,
+      "loss": 0.7152,
+      "step": 2713
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5810480051047082,
+      "learning_rate": 1.993969493086772e-05,
+      "loss": 0.5554,
+      "step": 2714
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6454983511401244,
+      "learning_rate": 1.9939611229175926e-05,
+      "loss": 0.5698,
+      "step": 2715
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6551087166490799,
+      "learning_rate": 1.9939527469612567e-05,
+      "loss": 0.5497,
+      "step": 2716
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.61363034890741,
+      "learning_rate": 1.9939443652178136e-05,
+      "loss": 0.7634,
+      "step": 2717
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5470911403161087,
+      "learning_rate": 1.9939359776873112e-05,
+      "loss": 0.8314,
+      "step": 2718
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6899906198624122,
+      "learning_rate": 1.9939275843697988e-05,
+      "loss": 0.7109,
+      "step": 2719
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4225350189957768,
+      "learning_rate": 1.9939191852653257e-05,
+      "loss": 0.8802,
+      "step": 2720
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6720696992466977,
+      "learning_rate": 1.9939107803739402e-05,
+      "loss": 0.7723,
+      "step": 2721
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6132023711020682,
+      "learning_rate": 1.9939023696956916e-05,
+      "loss": 0.8275,
+      "step": 2722
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4194900704058184,
+      "learning_rate": 1.993893953230629e-05,
+      "loss": 0.7672,
+      "step": 2723
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3642527323785607,
+      "learning_rate": 1.993885530978801e-05,
+      "loss": 0.7271,
+      "step": 2724
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8864908462675782,
+      "learning_rate": 1.9938771029402564e-05,
+      "loss": 0.7598,
+      "step": 2725
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5726742780913365,
+      "learning_rate": 1.993868669115045e-05,
+      "loss": 0.862,
+      "step": 2726
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4045208955823394,
+      "learning_rate": 1.993860229503215e-05,
+      "loss": 0.6854,
+      "step": 2727
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6499090476889815,
+      "learning_rate": 1.9938517841048167e-05,
+      "loss": 0.8016,
+      "step": 2728
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5910028267603058,
+      "learning_rate": 1.9938433329198985e-05,
+      "loss": 0.482,
+      "step": 2729
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5622808478100356,
+      "learning_rate": 1.99383487594851e-05,
+      "loss": 0.7759,
+      "step": 2730
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.372133377675313,
+      "learning_rate": 1.9938264131906998e-05,
+      "loss": 0.7799,
+      "step": 2731
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0126459501461578,
+      "learning_rate": 1.993817944646518e-05,
+      "loss": 0.7671,
+      "step": 2732
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6187015690536759,
+      "learning_rate": 1.993809470316013e-05,
+      "loss": 0.5573,
+      "step": 2733
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7099514151293882,
+      "learning_rate": 1.993800990199235e-05,
+      "loss": 0.7842,
+      "step": 2734
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.8737241066215207,
+      "learning_rate": 1.993792504296233e-05,
+      "loss": 0.6982,
+      "step": 2735
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.458161609702887,
+      "learning_rate": 1.993784012607056e-05,
+      "loss": 0.7266,
+      "step": 2736
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3581275365151886,
+      "learning_rate": 1.993775515131754e-05,
+      "loss": 0.6532,
+      "step": 2737
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.678571271306529,
+      "learning_rate": 1.9937670118703768e-05,
+      "loss": 0.6882,
+      "step": 2738
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6036897550429012,
+      "learning_rate": 1.993758502822973e-05,
+      "loss": 0.517,
+      "step": 2739
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4083565129609266,
+      "learning_rate": 1.993749987989593e-05,
+      "loss": 0.8424,
+      "step": 2740
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4150895490267594,
+      "learning_rate": 1.9937414673702852e-05,
+      "loss": 0.7689,
+      "step": 2741
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.472738271128021,
+      "learning_rate": 1.9937329409651007e-05,
+      "loss": 0.7747,
+      "step": 2742
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.432491982990521,
+      "learning_rate": 1.993724408774088e-05,
+      "loss": 0.7877,
+      "step": 2743
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5276153965729866,
+      "learning_rate": 1.9937158707972975e-05,
+      "loss": 0.64,
+      "step": 2744
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7484285230853316,
+      "learning_rate": 1.9937073270347782e-05,
+      "loss": 0.6613,
+      "step": 2745
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.530549610708333,
+      "learning_rate": 1.9936987774865805e-05,
+      "loss": 0.7583,
+      "step": 2746
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5833087087555867,
+      "learning_rate": 1.9936902221527536e-05,
+      "loss": 0.4706,
+      "step": 2747
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.520136723077473,
+      "learning_rate": 1.993681661033348e-05,
+      "loss": 0.6564,
+      "step": 2748
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1285630827217155,
+      "learning_rate": 1.993673094128413e-05,
+      "loss": 0.7216,
+      "step": 2749
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6341282613883623,
+      "learning_rate": 1.9936645214379988e-05,
+      "loss": 0.7655,
+      "step": 2750
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5762352496069766,
+      "learning_rate": 1.993655942962155e-05,
+      "loss": 0.5058,
+      "step": 2751
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5145410597500464,
+      "learning_rate": 1.993647358700932e-05,
+      "loss": 0.7448,
+      "step": 2752
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5839894686533524,
+      "learning_rate": 1.993638768654379e-05,
+      "loss": 0.7757,
+      "step": 2753
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7886402415215958,
+      "learning_rate": 1.993630172822547e-05,
+      "loss": 0.8586,
+      "step": 2754
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6772554722586908,
+      "learning_rate": 1.9936215712054854e-05,
+      "loss": 0.6933,
+      "step": 2755
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.554711907891719,
+      "learning_rate": 1.9936129638032443e-05,
+      "loss": 0.7515,
+      "step": 2756
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6379252705163996,
+      "learning_rate": 1.9936043506158742e-05,
+      "loss": 0.7263,
+      "step": 2757
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7122318825016734,
+      "learning_rate": 1.9935957316434246e-05,
+      "loss": 0.6285,
+      "step": 2758
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5420186683441748,
+      "learning_rate": 1.9935871068859463e-05,
+      "loss": 0.8428,
+      "step": 2759
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4392080937630323,
+      "learning_rate": 1.9935784763434894e-05,
+      "loss": 0.934,
+      "step": 2760
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5175387644243057,
+      "learning_rate": 1.993569840016104e-05,
+      "loss": 0.5765,
+      "step": 2761
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0163462073587213,
+      "learning_rate": 1.9935611979038403e-05,
+      "loss": 0.6878,
+      "step": 2762
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.486353516831727,
+      "learning_rate": 1.993552550006749e-05,
+      "loss": 0.6554,
+      "step": 2763
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4930094812546473,
+      "learning_rate": 1.99354389632488e-05,
+      "loss": 0.7599,
+      "step": 2764
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4588208288010958,
+      "learning_rate": 1.9935352368582837e-05,
+      "loss": 0.6971,
+      "step": 2765
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6035129001542878,
+      "learning_rate": 1.9935265716070113e-05,
+      "loss": 0.7179,
+      "step": 2766
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4690731901760707,
+      "learning_rate": 1.9935179005711124e-05,
+      "loss": 0.8948,
+      "step": 2767
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.374605760546045,
+      "learning_rate": 1.993509223750637e-05,
+      "loss": 0.7981,
+      "step": 2768
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.242623978628688,
+      "learning_rate": 1.9935005411456373e-05,
+      "loss": 0.728,
+      "step": 2769
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1159437449848912,
+      "learning_rate": 1.9934918527561627e-05,
+      "loss": 0.7961,
+      "step": 2770
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6199999334799248,
+      "learning_rate": 1.993483158582264e-05,
+      "loss": 0.5279,
+      "step": 2771
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5858565317489102,
+      "learning_rate": 1.9934744586239917e-05,
+      "loss": 0.843,
+      "step": 2772
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7991665348530377,
+      "learning_rate": 1.9934657528813965e-05,
+      "loss": 0.7267,
+      "step": 2773
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.57799780198292,
+      "learning_rate": 1.993457041354529e-05,
+      "loss": 0.8294,
+      "step": 2774
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7535827377500253,
+      "learning_rate": 1.9934483240434406e-05,
+      "loss": 0.7871,
+      "step": 2775
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.379460588145792,
+      "learning_rate": 1.9934396009481815e-05,
+      "loss": 0.8084,
+      "step": 2776
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.542316877514223,
+      "learning_rate": 1.9934308720688024e-05,
+      "loss": 0.7706,
+      "step": 2777
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4854685938726406,
+      "learning_rate": 1.9934221374053538e-05,
+      "loss": 0.7428,
+      "step": 2778
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5965073183821271,
+      "learning_rate": 1.9934133969578875e-05,
+      "loss": 0.5502,
+      "step": 2779
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.809678975211616,
+      "learning_rate": 1.9934046507264537e-05,
+      "loss": 0.7699,
+      "step": 2780
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.878002462810969,
+      "learning_rate": 1.9933958987111035e-05,
+      "loss": 0.6254,
+      "step": 2781
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4073878709094563,
+      "learning_rate": 1.993387140911888e-05,
+      "loss": 0.7315,
+      "step": 2782
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5126782664182967,
+      "learning_rate": 1.993378377328858e-05,
+      "loss": 0.6939,
+      "step": 2783
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.672293613489945,
+      "learning_rate": 1.9933696079620647e-05,
+      "loss": 0.8405,
+      "step": 2784
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5727541042906297,
+      "learning_rate": 1.993360832811559e-05,
+      "loss": 0.7401,
+      "step": 2785
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.944478720654882,
+      "learning_rate": 1.993352051877392e-05,
+      "loss": 0.8156,
+      "step": 2786
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4534184932739833,
+      "learning_rate": 1.9933432651596146e-05,
+      "loss": 0.7214,
+      "step": 2787
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6547948854880932,
+      "learning_rate": 1.9933344726582787e-05,
+      "loss": 0.7748,
+      "step": 2788
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.539643579406803,
+      "learning_rate": 1.9933256743734346e-05,
+      "loss": 0.7371,
+      "step": 2789
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7995897516686588,
+      "learning_rate": 1.9933168703051343e-05,
+      "loss": 0.7687,
+      "step": 2790
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5684182010275385,
+      "learning_rate": 1.9933080604534285e-05,
+      "loss": 0.7487,
+      "step": 2791
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3765452949618704,
+      "learning_rate": 1.9932992448183688e-05,
+      "loss": 0.7532,
+      "step": 2792
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6394010502604863,
+      "learning_rate": 1.9932904234000063e-05,
+      "loss": 0.7884,
+      "step": 2793
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3397410540168277,
+      "learning_rate": 1.9932815961983925e-05,
+      "loss": 0.7174,
+      "step": 2794
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6256067511757316,
+      "learning_rate": 1.993272763213579e-05,
+      "loss": 0.5029,
+      "step": 2795
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.636437156555059,
+      "learning_rate": 1.9932639244456165e-05,
+      "loss": 0.6873,
+      "step": 2796
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.646638523990953,
+      "learning_rate": 1.9932550798945575e-05,
+      "loss": 0.7741,
+      "step": 2797
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2355762082481023,
+      "learning_rate": 1.9932462295604526e-05,
+      "loss": 0.6362,
+      "step": 2798
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8096720940817983,
+      "learning_rate": 1.993237373443354e-05,
+      "loss": 0.7512,
+      "step": 2799
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8889148778284577,
+      "learning_rate": 1.9932285115433126e-05,
+      "loss": 0.8108,
+      "step": 2800
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0135289174349724,
+      "learning_rate": 1.9932196438603805e-05,
+      "loss": 0.7449,
+      "step": 2801
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.005995202059418,
+      "learning_rate": 1.993210770394609e-05,
+      "loss": 0.9414,
+      "step": 2802
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6669337732612153,
+      "learning_rate": 1.9932018911460503e-05,
+      "loss": 0.7241,
+      "step": 2803
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.666106618316386,
+      "learning_rate": 1.9931930061147556e-05,
+      "loss": 0.7892,
+      "step": 2804
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7391940934236978,
+      "learning_rate": 1.993184115300777e-05,
+      "loss": 0.8691,
+      "step": 2805
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0715649250290187,
+      "learning_rate": 1.993175218704166e-05,
+      "loss": 0.795,
+      "step": 2806
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5668619368195973,
+      "learning_rate": 1.9931663163249744e-05,
+      "loss": 0.6741,
+      "step": 2807
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6244783733018484,
+      "learning_rate": 1.993157408163254e-05,
+      "loss": 0.5007,
+      "step": 2808
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4797853647647572,
+      "learning_rate": 1.9931484942190567e-05,
+      "loss": 0.685,
+      "step": 2809
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.399358358219951,
+      "learning_rate": 1.9931395744924345e-05,
+      "loss": 0.8127,
+      "step": 2810
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5605279505037857,
+      "learning_rate": 1.993130648983439e-05,
+      "loss": 0.8259,
+      "step": 2811
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6505832960839522,
+      "learning_rate": 1.9931217176921228e-05,
+      "loss": 0.8495,
+      "step": 2812
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6093093628438745,
+      "learning_rate": 1.9931127806185376e-05,
+      "loss": 0.4799,
+      "step": 2813
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5791904374128101,
+      "learning_rate": 1.993103837762735e-05,
+      "loss": 0.5151,
+      "step": 2814
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.612141939092521,
+      "learning_rate": 1.9930948891247678e-05,
+      "loss": 0.6291,
+      "step": 2815
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6823828561223124,
+      "learning_rate": 1.9930859347046875e-05,
+      "loss": 0.8318,
+      "step": 2816
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.767470623939097,
+      "learning_rate": 1.9930769745025468e-05,
+      "loss": 0.784,
+      "step": 2817
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.480700198146825,
+      "learning_rate": 1.993068008518397e-05,
+      "loss": 0.8915,
+      "step": 2818
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0669110028920437,
+      "learning_rate": 1.993059036752291e-05,
+      "loss": 0.7685,
+      "step": 2819
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.497669492568427,
+      "learning_rate": 1.9930500592042815e-05,
+      "loss": 0.8106,
+      "step": 2820
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0323294229921065,
+      "learning_rate": 1.9930410758744195e-05,
+      "loss": 0.7097,
+      "step": 2821
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4284906543746023,
+      "learning_rate": 1.9930320867627584e-05,
+      "loss": 0.7174,
+      "step": 2822
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.477035919126504,
+      "learning_rate": 1.99302309186935e-05,
+      "loss": 0.7969,
+      "step": 2823
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4744625380414855,
+      "learning_rate": 1.9930140911942465e-05,
+      "loss": 0.759,
+      "step": 2824
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.397098350082738,
+      "learning_rate": 1.993005084737501e-05,
+      "loss": 0.691,
+      "step": 2825
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5450881991566663,
+      "learning_rate": 1.9929960724991652e-05,
+      "loss": 0.7253,
+      "step": 2826
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6299066856715432,
+      "learning_rate": 1.9929870544792918e-05,
+      "loss": 0.5554,
+      "step": 2827
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6471372643280645,
+      "learning_rate": 1.9929780306779337e-05,
+      "loss": 0.5373,
+      "step": 2828
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4541931579136567,
+      "learning_rate": 1.9929690010951427e-05,
+      "loss": 0.6973,
+      "step": 2829
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6501745633824805,
+      "learning_rate": 1.992959965730972e-05,
+      "loss": 0.7176,
+      "step": 2830
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1190381870940342,
+      "learning_rate": 1.9929509245854743e-05,
+      "loss": 0.7568,
+      "step": 2831
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1769285459097616,
+      "learning_rate": 1.9929418776587018e-05,
+      "loss": 0.8028,
+      "step": 2832
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.830518125159428,
+      "learning_rate": 1.9929328249507072e-05,
+      "loss": 0.8207,
+      "step": 2833
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5761674338660636,
+      "learning_rate": 1.9929237664615434e-05,
+      "loss": 0.6528,
+      "step": 2834
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5086523098412126,
+      "learning_rate": 1.992914702191263e-05,
+      "loss": 0.7327,
+      "step": 2835
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8951707857759694,
+      "learning_rate": 1.9929056321399193e-05,
+      "loss": 0.7427,
+      "step": 2836
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9624851278723447,
+      "learning_rate": 1.992896556307564e-05,
+      "loss": 0.7422,
+      "step": 2837
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4437773844734514,
+      "learning_rate": 1.992887474694251e-05,
+      "loss": 0.6512,
+      "step": 2838
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9933050349476011,
+      "learning_rate": 1.992878387300033e-05,
+      "loss": 0.7156,
+      "step": 2839
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8148984179394088,
+      "learning_rate": 1.992869294124962e-05,
+      "loss": 0.8686,
+      "step": 2840
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5002501918955469,
+      "learning_rate": 1.9928601951690922e-05,
+      "loss": 0.8243,
+      "step": 2841
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6349177975839475,
+      "learning_rate": 1.992851090432476e-05,
+      "loss": 0.7444,
+      "step": 2842
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2640726957927044,
+      "learning_rate": 1.9928419799151658e-05,
+      "loss": 0.6835,
+      "step": 2843
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5750862860055618,
+      "learning_rate": 1.992832863617216e-05,
+      "loss": 0.5329,
+      "step": 2844
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5258112102655224,
+      "learning_rate": 1.9928237415386785e-05,
+      "loss": 0.745,
+      "step": 2845
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5063711673649907,
+      "learning_rate": 1.992814613679607e-05,
+      "loss": 0.8796,
+      "step": 2846
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.595367486405249,
+      "learning_rate": 1.992805480040055e-05,
+      "loss": 0.7323,
+      "step": 2847
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8741360385349692,
+      "learning_rate": 1.9927963406200743e-05,
+      "loss": 0.9277,
+      "step": 2848
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4044187196592957,
+      "learning_rate": 1.9927871954197194e-05,
+      "loss": 0.779,
+      "step": 2849
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6663002846231265,
+      "learning_rate": 1.9927780444390433e-05,
+      "loss": 0.7274,
+      "step": 2850
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4880468134391531,
+      "learning_rate": 1.9927688876780988e-05,
+      "loss": 0.7303,
+      "step": 2851
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6281756138144703,
+      "learning_rate": 1.9927597251369398e-05,
+      "loss": 0.7465,
+      "step": 2852
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5101627800239268,
+      "learning_rate": 1.992750556815619e-05,
+      "loss": 0.717,
+      "step": 2853
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1226801521254144,
+      "learning_rate": 1.9927413827141904e-05,
+      "loss": 0.7165,
+      "step": 2854
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3892944703112706,
+      "learning_rate": 1.9927322028327074e-05,
+      "loss": 0.7964,
+      "step": 2855
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1769557601515155,
+      "learning_rate": 1.992723017171223e-05,
+      "loss": 0.7271,
+      "step": 2856
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.952523694374646,
+      "learning_rate": 1.9927138257297908e-05,
+      "loss": 0.8372,
+      "step": 2857
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4368012065606246,
+      "learning_rate": 1.9927046285084642e-05,
+      "loss": 0.8149,
+      "step": 2858
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4956509546336776,
+      "learning_rate": 1.9926954255072976e-05,
+      "loss": 0.6458,
+      "step": 2859
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9086115040175509,
+      "learning_rate": 1.9926862167263437e-05,
+      "loss": 0.6865,
+      "step": 2860
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7996554405067025,
+      "learning_rate": 1.9926770021656564e-05,
+      "loss": 0.7841,
+      "step": 2861
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4514215756098077,
+      "learning_rate": 1.992667781825289e-05,
+      "loss": 0.7364,
+      "step": 2862
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.039526146986154,
+      "learning_rate": 1.9926585557052958e-05,
+      "loss": 0.6873,
+      "step": 2863
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4943695879715293,
+      "learning_rate": 1.99264932380573e-05,
+      "loss": 0.8692,
+      "step": 2864
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.674164265973802,
+      "learning_rate": 1.9926400861266458e-05,
+      "loss": 0.7894,
+      "step": 2865
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5524475029930611,
+      "learning_rate": 1.9926308426680966e-05,
+      "loss": 0.787,
+      "step": 2866
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9132712304068138,
+      "learning_rate": 1.9926215934301366e-05,
+      "loss": 0.7545,
+      "step": 2867
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8617163222966082,
+      "learning_rate": 1.992612338412819e-05,
+      "loss": 0.8458,
+      "step": 2868
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3325904166400466,
+      "learning_rate": 1.9926030776161986e-05,
+      "loss": 0.7673,
+      "step": 2869
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5894999579411626,
+      "learning_rate": 1.9925938110403286e-05,
+      "loss": 0.7601,
+      "step": 2870
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6196863412161182,
+      "learning_rate": 1.992584538685263e-05,
+      "loss": 0.829,
+      "step": 2871
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5915791851589716,
+      "learning_rate": 1.9925752605510566e-05,
+      "loss": 0.5708,
+      "step": 2872
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.564393053601096,
+      "learning_rate": 1.9925659766377622e-05,
+      "loss": 0.7487,
+      "step": 2873
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5284780321056968,
+      "learning_rate": 1.9925566869454345e-05,
+      "loss": 0.6986,
+      "step": 2874
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6224111073712268,
+      "learning_rate": 1.9925473914741275e-05,
+      "loss": 0.9121,
+      "step": 2875
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.528023303173711,
+      "learning_rate": 1.9925380902238956e-05,
+      "loss": 0.6794,
+      "step": 2876
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5889098717104546,
+      "learning_rate": 1.9925287831947922e-05,
+      "loss": 0.7616,
+      "step": 2877
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5903851443877843,
+      "learning_rate": 1.9925194703868727e-05,
+      "loss": 0.7544,
+      "step": 2878
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5563844190017677,
+      "learning_rate": 1.99251015180019e-05,
+      "loss": 0.6239,
+      "step": 2879
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8111157488876575,
+      "learning_rate": 1.9925008274347993e-05,
+      "loss": 0.6359,
+      "step": 2880
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.716303350025485,
+      "learning_rate": 1.9924914972907547e-05,
+      "loss": 0.7869,
+      "step": 2881
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3834940878029673,
+      "learning_rate": 1.9924821613681103e-05,
+      "loss": 0.8285,
+      "step": 2882
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8481925468367955,
+      "learning_rate": 1.9924728196669205e-05,
+      "loss": 0.7328,
+      "step": 2883
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3947067375303925,
+      "learning_rate": 1.9924634721872396e-05,
+      "loss": 0.7488,
+      "step": 2884
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6327689452766934,
+      "learning_rate": 1.992454118929122e-05,
+      "loss": 0.7433,
+      "step": 2885
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.663024359170714,
+      "learning_rate": 1.9924447598926227e-05,
+      "loss": 0.7066,
+      "step": 2886
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5141891470489195,
+      "learning_rate": 1.9924353950777956e-05,
+      "loss": 0.9452,
+      "step": 2887
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6491603414792324,
+      "learning_rate": 1.9924260244846954e-05,
+      "loss": 0.8069,
+      "step": 2888
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7085930929734727,
+      "learning_rate": 1.9924166481133767e-05,
+      "loss": 0.7916,
+      "step": 2889
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.683635242463519,
+      "learning_rate": 1.9924072659638944e-05,
+      "loss": 0.787,
+      "step": 2890
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6707470372903186,
+      "learning_rate": 1.9923978780363025e-05,
+      "loss": 0.6273,
+      "step": 2891
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6099980644974647,
+      "learning_rate": 1.992388484330656e-05,
+      "loss": 0.8011,
+      "step": 2892
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.861151012692106,
+      "learning_rate": 1.9923790848470098e-05,
+      "loss": 0.8051,
+      "step": 2893
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5785280829514214,
+      "learning_rate": 1.992369679585418e-05,
+      "loss": 0.7242,
+      "step": 2894
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5030736820059023,
+      "learning_rate": 1.992360268545936e-05,
+      "loss": 0.8249,
+      "step": 2895
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7426363373589537,
+      "learning_rate": 1.992350851728618e-05,
+      "loss": 0.7535,
+      "step": 2896
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6930343884175718,
+      "learning_rate": 1.9923414291335198e-05,
+      "loss": 0.8296,
+      "step": 2897
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4493364249740868,
+      "learning_rate": 1.992332000760695e-05,
+      "loss": 0.7009,
+      "step": 2898
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5176558509183464,
+      "learning_rate": 1.9923225666101994e-05,
+      "loss": 0.8594,
+      "step": 2899
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.246110780783542,
+      "learning_rate": 1.9923131266820878e-05,
+      "loss": 0.7119,
+      "step": 2900
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.389101349953166,
+      "learning_rate": 1.9923036809764145e-05,
+      "loss": 0.7939,
+      "step": 2901
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.588975246606803,
+      "learning_rate": 1.9922942294932356e-05,
+      "loss": 0.816,
+      "step": 2902
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6763425890399224,
+      "learning_rate": 1.9922847722326052e-05,
+      "loss": 0.9216,
+      "step": 2903
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6461649734682509,
+      "learning_rate": 1.992275309194579e-05,
+      "loss": 0.8688,
+      "step": 2904
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8252367454219902,
+      "learning_rate": 1.9922658403792115e-05,
+      "loss": 0.7773,
+      "step": 2905
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5988597354447511,
+      "learning_rate": 1.9922563657865584e-05,
+      "loss": 0.8564,
+      "step": 2906
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5490488057449314,
+      "learning_rate": 1.9922468854166744e-05,
+      "loss": 0.6796,
+      "step": 2907
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3951481298438608,
+      "learning_rate": 1.9922373992696146e-05,
+      "loss": 0.8345,
+      "step": 2908
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.326867171577964,
+      "learning_rate": 1.992227907345435e-05,
+      "loss": 0.7943,
+      "step": 2909
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4387878893824841,
+      "learning_rate": 1.9922184096441904e-05,
+      "loss": 0.746,
+      "step": 2910
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6130184875424519,
+      "learning_rate": 1.992208906165936e-05,
+      "loss": 0.5406,
+      "step": 2911
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3930988680488303,
+      "learning_rate": 1.992199396910727e-05,
+      "loss": 0.8014,
+      "step": 2912
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6003514173812838,
+      "learning_rate": 1.992189881878619e-05,
+      "loss": 0.7794,
+      "step": 2913
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.727767972138208,
+      "learning_rate": 1.9921803610696675e-05,
+      "loss": 0.8771,
+      "step": 2914
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6162515211448245,
+      "learning_rate": 1.992170834483928e-05,
+      "loss": 0.6611,
+      "step": 2915
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8833666794683195,
+      "learning_rate": 1.9921613021214555e-05,
+      "loss": 0.7814,
+      "step": 2916
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8339013212277244,
+      "learning_rate": 1.992151763982306e-05,
+      "loss": 0.703,
+      "step": 2917
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4290166256168502,
+      "learning_rate": 1.9921422200665347e-05,
+      "loss": 0.8395,
+      "step": 2918
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0754110896742892,
+      "learning_rate": 1.9921326703741976e-05,
+      "loss": 0.787,
+      "step": 2919
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.867282479939369,
+      "learning_rate": 1.9921231149053495e-05,
+      "loss": 0.6727,
+      "step": 2920
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5503227702696873,
+      "learning_rate": 1.9921135536600465e-05,
+      "loss": 0.7238,
+      "step": 2921
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4921589547388157,
+      "learning_rate": 1.9921039866383447e-05,
+      "loss": 0.7017,
+      "step": 2922
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1816981191455658,
+      "learning_rate": 1.9920944138402993e-05,
+      "loss": 0.5937,
+      "step": 2923
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5406932656821037,
+      "learning_rate": 1.9920848352659662e-05,
+      "loss": 0.7471,
+      "step": 2924
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4495764884299585,
+      "learning_rate": 1.9920752509154008e-05,
+      "loss": 0.6996,
+      "step": 2925
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5354595739111827,
+      "learning_rate": 1.992065660788659e-05,
+      "loss": 0.6785,
+      "step": 2926
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.838007943165036,
+      "learning_rate": 1.9920560648857976e-05,
+      "loss": 0.7813,
+      "step": 2927
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3463662249426271,
+      "learning_rate": 1.9920464632068713e-05,
+      "loss": 0.77,
+      "step": 2928
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.596776234062118,
+      "learning_rate": 1.9920368557519362e-05,
+      "loss": 0.7491,
+      "step": 2929
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7547181682231863,
+      "learning_rate": 1.992027242521049e-05,
+      "loss": 0.6145,
+      "step": 2930
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6027646419677177,
+      "learning_rate": 1.9920176235142645e-05,
+      "loss": 0.7485,
+      "step": 2931
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6779418725128672,
+      "learning_rate": 1.9920079987316398e-05,
+      "loss": 0.7078,
+      "step": 2932
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.801914106415762,
+      "learning_rate": 1.9919983681732302e-05,
+      "loss": 0.931,
+      "step": 2933
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8379655594484345,
+      "learning_rate": 1.9919887318390923e-05,
+      "loss": 0.8082,
+      "step": 2934
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5192708570754925,
+      "learning_rate": 1.9919790897292817e-05,
+      "loss": 0.7745,
+      "step": 2935
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5699432318428354,
+      "learning_rate": 1.9919694418438547e-05,
+      "loss": 0.5034,
+      "step": 2936
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6291614032001858,
+      "learning_rate": 1.9919597881828678e-05,
+      "loss": 0.7815,
+      "step": 2937
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7958875243636228,
+      "learning_rate": 1.991950128746377e-05,
+      "loss": 0.7964,
+      "step": 2938
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5156205385358454,
+      "learning_rate": 1.9919404635344383e-05,
+      "loss": 0.6974,
+      "step": 2939
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1498901840093403,
+      "learning_rate": 1.991930792547108e-05,
+      "loss": 0.8464,
+      "step": 2940
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.839944123279269,
+      "learning_rate": 1.991921115784443e-05,
+      "loss": 0.7114,
+      "step": 2941
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5860684325784277,
+      "learning_rate": 1.9919114332464994e-05,
+      "loss": 0.538,
+      "step": 2942
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.882855241385785,
+      "learning_rate": 1.991901744933333e-05,
+      "loss": 0.746,
+      "step": 2943
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.39122424093412,
+      "learning_rate": 1.9918920508450003e-05,
+      "loss": 0.7966,
+      "step": 2944
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.410390855758358,
+      "learning_rate": 1.9918823509815585e-05,
+      "loss": 0.8232,
+      "step": 2945
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5775918422986008,
+      "learning_rate": 1.9918726453430635e-05,
+      "loss": 0.7695,
+      "step": 2946
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2747848877978076,
+      "learning_rate": 1.9918629339295722e-05,
+      "loss": 0.8253,
+      "step": 2947
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4761262575478316,
+      "learning_rate": 1.9918532167411407e-05,
+      "loss": 0.6649,
+      "step": 2948
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6754893182225026,
+      "learning_rate": 1.991843493777826e-05,
+      "loss": 0.813,
+      "step": 2949
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.775174154297924,
+      "learning_rate": 1.991833765039684e-05,
+      "loss": 0.7803,
+      "step": 2950
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6498468674391125,
+      "learning_rate": 1.991824030526772e-05,
+      "loss": 0.6948,
+      "step": 2951
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7435705221788373,
+      "learning_rate": 1.9918142902391463e-05,
+      "loss": 0.8228,
+      "step": 2952
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.479603620865376,
+      "learning_rate": 1.9918045441768642e-05,
+      "loss": 0.7591,
+      "step": 2953
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4019687367703926,
+      "learning_rate": 1.9917947923399818e-05,
+      "loss": 0.7966,
+      "step": 2954
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7013183507782321,
+      "learning_rate": 1.9917850347285563e-05,
+      "loss": 0.8119,
+      "step": 2955
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7590511719846786,
+      "learning_rate": 1.9917752713426442e-05,
+      "loss": 0.7404,
+      "step": 2956
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6329574556452899,
+      "learning_rate": 1.9917655021823022e-05,
+      "loss": 0.5749,
+      "step": 2957
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4571141297190553,
+      "learning_rate": 1.991755727247588e-05,
+      "loss": 0.7933,
+      "step": 2958
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3899540343795083,
+      "learning_rate": 1.9917459465385578e-05,
+      "loss": 0.6119,
+      "step": 2959
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6013259663847981,
+      "learning_rate": 1.9917361600552685e-05,
+      "loss": 0.5089,
+      "step": 2960
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5694770214816547,
+      "learning_rate": 1.9917263677977773e-05,
+      "loss": 0.8806,
+      "step": 2961
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4484379584244265,
+      "learning_rate": 1.9917165697661415e-05,
+      "loss": 0.7634,
+      "step": 2962
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9072738441529402,
+      "learning_rate": 1.9917067659604176e-05,
+      "loss": 0.8846,
+      "step": 2963
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5267363487287433,
+      "learning_rate": 1.991696956380663e-05,
+      "loss": 0.7443,
+      "step": 2964
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3260044161835873,
+      "learning_rate": 1.991687141026935e-05,
+      "loss": 0.9249,
+      "step": 2965
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6216916143391,
+      "learning_rate": 1.99167731989929e-05,
+      "loss": 0.8168,
+      "step": 2966
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.772262750121927,
+      "learning_rate": 1.991667492997786e-05,
+      "loss": 0.8395,
+      "step": 2967
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6700889314048284,
+      "learning_rate": 1.99165766032248e-05,
+      "loss": 0.8246,
+      "step": 2968
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6046995282868597,
+      "learning_rate": 1.991647821873429e-05,
+      "loss": 0.7918,
+      "step": 2969
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5501798606897055,
+      "learning_rate": 1.9916379776506908e-05,
+      "loss": 0.7879,
+      "step": 2970
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.72973950013238,
+      "learning_rate": 1.9916281276543215e-05,
+      "loss": 0.734,
+      "step": 2971
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.937214997229587,
+      "learning_rate": 1.99161827188438e-05,
+      "loss": 0.7522,
+      "step": 2972
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7082642599873932,
+      "learning_rate": 1.9916084103409228e-05,
+      "loss": 0.6959,
+      "step": 2973
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8831484750940997,
+      "learning_rate": 1.9915985430240075e-05,
+      "loss": 0.7105,
+      "step": 2974
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.706988003801727,
+      "learning_rate": 1.9915886699336915e-05,
+      "loss": 0.7572,
+      "step": 2975
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7004390507401144,
+      "learning_rate": 1.9915787910700322e-05,
+      "loss": 0.8822,
+      "step": 2976
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.56761787422672,
+      "learning_rate": 1.9915689064330875e-05,
+      "loss": 0.5672,
+      "step": 2977
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1869451890403555,
+      "learning_rate": 1.991559016022915e-05,
+      "loss": 0.777,
+      "step": 2978
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.6466179447364997,
+      "learning_rate": 1.9915491198395714e-05,
+      "loss": 0.7676,
+      "step": 2979
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9212591944778952,
+      "learning_rate": 1.991539217883115e-05,
+      "loss": 0.8472,
+      "step": 2980
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9889805559259146,
+      "learning_rate": 1.991529310153603e-05,
+      "loss": 0.8572,
+      "step": 2981
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8533189333049167,
+      "learning_rate": 1.9915193966510945e-05,
+      "loss": 0.7703,
+      "step": 2982
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.5837776060086117,
+      "learning_rate": 1.9915094773756454e-05,
+      "loss": 0.8014,
+      "step": 2983
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.53137808466354,
+      "learning_rate": 1.9914995523273144e-05,
+      "loss": 0.8521,
+      "step": 2984
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1370344813320994,
+      "learning_rate": 1.991489621506159e-05,
+      "loss": 0.8014,
+      "step": 2985
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4118485071021354,
+      "learning_rate": 1.9914796849122374e-05,
+      "loss": 0.7749,
+      "step": 2986
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4929188460180847,
+      "learning_rate": 1.991469742545607e-05,
+      "loss": 0.8413,
+      "step": 2987
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.410998641077798,
+      "learning_rate": 1.9914597944063257e-05,
+      "loss": 0.6499,
+      "step": 2988
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7354811857027899,
+      "learning_rate": 1.991449840494452e-05,
+      "loss": 0.694,
+      "step": 2989
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.584609462905564,
+      "learning_rate": 1.991439880810043e-05,
+      "loss": 0.9527,
+      "step": 2990
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7728448831746388,
+      "learning_rate": 1.991429915353157e-05,
+      "loss": 0.7342,
+      "step": 2991
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.579492847211323,
+      "learning_rate": 1.9914199441238528e-05,
+      "loss": 0.7138,
+      "step": 2992
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4803048980555338,
+      "learning_rate": 1.9914099671221874e-05,
+      "loss": 0.7674,
+      "step": 2993
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.725808422708521,
+      "learning_rate": 1.9913999843482195e-05,
+      "loss": 0.7412,
+      "step": 2994
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8307831196308133,
+      "learning_rate": 1.9913899958020067e-05,
+      "loss": 0.8226,
+      "step": 2995
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4925757459174394,
+      "learning_rate": 1.9913800014836076e-05,
+      "loss": 0.6979,
+      "step": 2996
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.631512620988522,
+      "learning_rate": 1.9913700013930804e-05,
+      "loss": 0.8235,
+      "step": 2997
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.87945944931201,
+      "learning_rate": 1.991359995530483e-05,
+      "loss": 0.9056,
+      "step": 2998
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4995595031158115,
+      "learning_rate": 1.991349983895874e-05,
+      "loss": 0.8465,
+      "step": 2999
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7980668294806483,
+      "learning_rate": 1.9913399664893113e-05,
+      "loss": 0.7067,
+      "step": 3000
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.507168164109152,
+      "learning_rate": 1.991329943310854e-05,
+      "loss": 0.7612,
+      "step": 3001
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6274457922195966,
+      "learning_rate": 1.9913199143605595e-05,
+      "loss": 0.8017,
+      "step": 3002
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5318331775216507,
+      "learning_rate": 1.9913098796384865e-05,
+      "loss": 0.7906,
+      "step": 3003
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6874892820815672,
+      "learning_rate": 1.991299839144694e-05,
+      "loss": 0.8408,
+      "step": 3004
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5759429405597436,
+      "learning_rate": 1.9912897928792395e-05,
+      "loss": 0.5568,
+      "step": 3005
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6086561860186197,
+      "learning_rate": 1.9912797408421823e-05,
+      "loss": 0.8636,
+      "step": 3006
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7204055562407568,
+      "learning_rate": 1.9912696830335806e-05,
+      "loss": 0.7956,
+      "step": 3007
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.544721370764685,
+      "learning_rate": 1.991259619453493e-05,
+      "loss": 0.887,
+      "step": 3008
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4927730476748102,
+      "learning_rate": 1.9912495501019777e-05,
+      "loss": 0.7978,
+      "step": 3009
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5795057053705845,
+      "learning_rate": 1.9912394749790942e-05,
+      "loss": 0.7491,
+      "step": 3010
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7760610221408548,
+      "learning_rate": 1.9912293940849002e-05,
+      "loss": 0.869,
+      "step": 3011
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9695149092638766,
+      "learning_rate": 1.991219307419455e-05,
+      "loss": 0.7352,
+      "step": 3012
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6482298147829881,
+      "learning_rate": 1.9912092149828175e-05,
+      "loss": 0.7089,
+      "step": 3013
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9610641891865574,
+      "learning_rate": 1.9911991167750458e-05,
+      "loss": 0.6858,
+      "step": 3014
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5880007929846589,
+      "learning_rate": 1.9911890127961993e-05,
+      "loss": 0.4945,
+      "step": 3015
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6273453590520575,
+      "learning_rate": 1.991178903046336e-05,
+      "loss": 0.8369,
+      "step": 3016
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.235870562801822,
+      "learning_rate": 1.991168787525516e-05,
+      "loss": 0.7005,
+      "step": 3017
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4147923901924475,
+      "learning_rate": 1.991158666233797e-05,
+      "loss": 0.6691,
+      "step": 3018
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4983472392420514,
+      "learning_rate": 1.991148539171239e-05,
+      "loss": 0.8668,
+      "step": 3019
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6072265343219946,
+      "learning_rate": 1.9911384063378996e-05,
+      "loss": 0.5821,
+      "step": 3020
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5615454557720896,
+      "learning_rate": 1.9911282677338395e-05,
+      "loss": 0.7319,
+      "step": 3021
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.658172767344634,
+      "learning_rate": 1.9911181233591163e-05,
+      "loss": 0.7203,
+      "step": 3022
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6299687710145523,
+      "learning_rate": 1.9911079732137897e-05,
+      "loss": 0.7767,
+      "step": 3023
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8406937949538418,
+      "learning_rate": 1.9910978172979187e-05,
+      "loss": 0.924,
+      "step": 3024
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3768181861964668,
+      "learning_rate": 1.9910876556115626e-05,
+      "loss": 0.7572,
+      "step": 3025
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.149899381019423,
+      "learning_rate": 1.9910774881547803e-05,
+      "loss": 0.7635,
+      "step": 3026
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6684583742378714,
+      "learning_rate": 1.9910673149276308e-05,
+      "loss": 0.8361,
+      "step": 3027
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2461650734996275,
+      "learning_rate": 1.991057135930174e-05,
+      "loss": 0.7253,
+      "step": 3028
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6133377102024424,
+      "learning_rate": 1.991046951162469e-05,
+      "loss": 0.5287,
+      "step": 3029
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3522121417794506,
+      "learning_rate": 1.9910367606245744e-05,
+      "loss": 0.6615,
+      "step": 3030
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9827009537235711,
+      "learning_rate": 1.9910265643165502e-05,
+      "loss": 0.8665,
+      "step": 3031
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5052920366204583,
+      "learning_rate": 1.9910163622384556e-05,
+      "loss": 0.6996,
+      "step": 3032
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.592789740849199,
+      "learning_rate": 1.9910061543903502e-05,
+      "loss": 0.7565,
+      "step": 3033
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5987439577903617,
+      "learning_rate": 1.9909959407722926e-05,
+      "loss": 0.7577,
+      "step": 3034
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9278753686470251,
+      "learning_rate": 1.990985721384343e-05,
+      "loss": 0.7131,
+      "step": 3035
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5978410105822551,
+      "learning_rate": 1.9909754962265615e-05,
+      "loss": 0.8479,
+      "step": 3036
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4321290427940128,
+      "learning_rate": 1.9909652652990065e-05,
+      "loss": 0.8223,
+      "step": 3037
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6054276488710795,
+      "learning_rate": 1.990955028601738e-05,
+      "loss": 0.5855,
+      "step": 3038
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5911701415275492,
+      "learning_rate": 1.9909447861348154e-05,
+      "loss": 0.8636,
+      "step": 3039
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6186814632878292,
+      "learning_rate": 1.9909345378982987e-05,
+      "loss": 0.8724,
+      "step": 3040
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4738501091073086,
+      "learning_rate": 1.9909242838922472e-05,
+      "loss": 0.7446,
+      "step": 3041
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4872593478294631,
+      "learning_rate": 1.9909140241167212e-05,
+      "loss": 0.8193,
+      "step": 3042
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.740244184175976,
+      "learning_rate": 1.99090375857178e-05,
+      "loss": 0.7099,
+      "step": 3043
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6277967667816977,
+      "learning_rate": 1.990893487257483e-05,
+      "loss": 0.7715,
+      "step": 3044
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6676811455745868,
+      "learning_rate": 1.990883210173891e-05,
+      "loss": 0.6633,
+      "step": 3045
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8280529729110138,
+      "learning_rate": 1.9908729273210627e-05,
+      "loss": 0.7809,
+      "step": 3046
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7679173468240652,
+      "learning_rate": 1.9908626386990586e-05,
+      "loss": 0.6291,
+      "step": 3047
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.421229661334622,
+      "learning_rate": 1.9908523443079384e-05,
+      "loss": 0.8833,
+      "step": 3048
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5584225182053201,
+      "learning_rate": 1.9908420441477626e-05,
+      "loss": 0.526,
+      "step": 3049
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5062075737399363,
+      "learning_rate": 1.9908317382185904e-05,
+      "loss": 0.8155,
+      "step": 3050
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4355892406873292,
+      "learning_rate": 1.9908214265204825e-05,
+      "loss": 0.6864,
+      "step": 3051
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4051742344002935,
+      "learning_rate": 1.990811109053498e-05,
+      "loss": 0.7169,
+      "step": 3052
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.471685271548772,
+      "learning_rate": 1.990800785817698e-05,
+      "loss": 0.8838,
+      "step": 3053
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6293232933006498,
+      "learning_rate": 1.990790456813142e-05,
+      "loss": 0.655,
+      "step": 3054
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4781039495747113,
+      "learning_rate": 1.99078012203989e-05,
+      "loss": 0.7855,
+      "step": 3055
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6366486501469373,
+      "learning_rate": 1.9907697814980025e-05,
+      "loss": 0.7717,
+      "step": 3056
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6758594767868968,
+      "learning_rate": 1.9907594351875402e-05,
+      "loss": 0.7417,
+      "step": 3057
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5997826385930278,
+      "learning_rate": 1.9907490831085626e-05,
+      "loss": 0.8546,
+      "step": 3058
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.046042949419614,
+      "learning_rate": 1.99073872526113e-05,
+      "loss": 0.8158,
+      "step": 3059
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.89247477348302,
+      "learning_rate": 1.9907283616453027e-05,
+      "loss": 0.7211,
+      "step": 3060
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4822629898924597,
+      "learning_rate": 1.9907179922611414e-05,
+      "loss": 0.7337,
+      "step": 3061
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5463629562827534,
+      "learning_rate": 1.9907076171087064e-05,
+      "loss": 0.7202,
+      "step": 3062
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.978953995083792,
+      "learning_rate": 1.990697236188058e-05,
+      "loss": 0.81,
+      "step": 3063
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.69711213315841,
+      "learning_rate": 1.9906868494992566e-05,
+      "loss": 0.7695,
+      "step": 3064
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6016770043930613,
+      "learning_rate": 1.9906764570423624e-05,
+      "loss": 0.5363,
+      "step": 3065
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5368154677623287,
+      "learning_rate": 1.9906660588174367e-05,
+      "loss": 0.7071,
+      "step": 3066
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8112077379041465,
+      "learning_rate": 1.9906556548245395e-05,
+      "loss": 0.7762,
+      "step": 3067
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.54300103176426,
+      "learning_rate": 1.9906452450637312e-05,
+      "loss": 0.7591,
+      "step": 3068
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2400052271937552,
+      "learning_rate": 1.9906348295350727e-05,
+      "loss": 0.7701,
+      "step": 3069
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.52678202236704,
+      "learning_rate": 1.990624408238625e-05,
+      "loss": 0.8333,
+      "step": 3070
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3769547400735622,
+      "learning_rate": 1.990613981174448e-05,
+      "loss": 0.908,
+      "step": 3071
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4666945405728382,
+      "learning_rate": 1.9906035483426024e-05,
+      "loss": 0.7744,
+      "step": 3072
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5672916683319271,
+      "learning_rate": 1.99059310974315e-05,
+      "loss": 0.6421,
+      "step": 3073
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6371636719060257,
+      "learning_rate": 1.9905826653761506e-05,
+      "loss": 0.5125,
+      "step": 3074
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.227102586362284,
+      "learning_rate": 1.9905722152416652e-05,
+      "loss": 0.9382,
+      "step": 3075
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4251695619041824,
+      "learning_rate": 1.990561759339755e-05,
+      "loss": 0.6795,
+      "step": 3076
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.7380318353287953,
+      "learning_rate": 1.9905512976704805e-05,
+      "loss": 0.7431,
+      "step": 3077
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3899405496552337,
+      "learning_rate": 1.9905408302339025e-05,
+      "loss": 0.7991,
+      "step": 3078
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.054731484881636,
+      "learning_rate": 1.9905303570300828e-05,
+      "loss": 0.8402,
+      "step": 3079
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5776873338851025,
+      "learning_rate": 1.990519878059081e-05,
+      "loss": 0.8213,
+      "step": 3080
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.744425248521529,
+      "learning_rate": 1.9905093933209593e-05,
+      "loss": 0.7833,
+      "step": 3081
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5416404293727588,
+      "learning_rate": 1.9904989028157784e-05,
+      "loss": 0.8287,
+      "step": 3082
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.0130311943706567,
+      "learning_rate": 1.990488406543599e-05,
+      "loss": 0.7697,
+      "step": 3083
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7397020835693848,
+      "learning_rate": 1.9904779045044828e-05,
+      "loss": 0.7784,
+      "step": 3084
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4053700971178225,
+      "learning_rate": 1.9904673966984907e-05,
+      "loss": 0.782,
+      "step": 3085
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7582907113135424,
+      "learning_rate": 1.9904568831256838e-05,
+      "loss": 0.7537,
+      "step": 3086
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7054529498726927,
+      "learning_rate": 1.990446363786123e-05,
+      "loss": 0.7848,
+      "step": 3087
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.569542882976719,
+      "learning_rate": 1.9904358386798704e-05,
+      "loss": 0.8137,
+      "step": 3088
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7047034362933173,
+      "learning_rate": 1.9904253078069865e-05,
+      "loss": 0.739,
+      "step": 3089
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6956746264246416,
+      "learning_rate": 1.990414771167533e-05,
+      "loss": 0.7471,
+      "step": 3090
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5476060921896786,
+      "learning_rate": 1.9904042287615712e-05,
+      "loss": 0.7143,
+      "step": 3091
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4151631827007654,
+      "learning_rate": 1.990393680589162e-05,
+      "loss": 0.6555,
+      "step": 3092
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.658799917860578,
+      "learning_rate": 1.9903831266503676e-05,
+      "loss": 0.6587,
+      "step": 3093
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4335721212683001,
+      "learning_rate": 1.9903725669452492e-05,
+      "loss": 0.85,
+      "step": 3094
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3188609438480672,
+      "learning_rate": 1.990362001473868e-05,
+      "loss": 0.6808,
+      "step": 3095
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.964024950671446,
+      "learning_rate": 1.9903514302362856e-05,
+      "loss": 0.7252,
+      "step": 3096
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.769533990627339,
+      "learning_rate": 1.9903408532325637e-05,
+      "loss": 0.8681,
+      "step": 3097
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5482211873516825,
+      "learning_rate": 1.990330270462764e-05,
+      "loss": 0.9415,
+      "step": 3098
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.674910243009035,
+      "learning_rate": 1.9903196819269476e-05,
+      "loss": 0.6021,
+      "step": 3099
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5575594554825587,
+      "learning_rate": 1.9903090876251767e-05,
+      "loss": 0.7671,
+      "step": 3100
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8398797679053867,
+      "learning_rate": 1.9902984875575127e-05,
+      "loss": 0.754,
+      "step": 3101
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6302665425957226,
+      "learning_rate": 1.9902878817240175e-05,
+      "loss": 0.5271,
+      "step": 3102
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6232299170607918,
+      "learning_rate": 1.9902772701247527e-05,
+      "loss": 0.8741,
+      "step": 3103
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5815930803345799,
+      "learning_rate": 1.9902666527597797e-05,
+      "loss": 0.5247,
+      "step": 3104
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.524398315814657,
+      "learning_rate": 1.9902560296291612e-05,
+      "loss": 0.6921,
+      "step": 3105
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8528787704445644,
+      "learning_rate": 1.9902454007329585e-05,
+      "loss": 0.7019,
+      "step": 3106
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6578437092677407,
+      "learning_rate": 1.9902347660712336e-05,
+      "loss": 0.7526,
+      "step": 3107
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8253158873981974,
+      "learning_rate": 1.990224125644048e-05,
+      "loss": 0.8119,
+      "step": 3108
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4740802452068609,
+      "learning_rate": 1.9902134794514648e-05,
+      "loss": 0.7247,
+      "step": 3109
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7714818658611846,
+      "learning_rate": 1.9902028274935446e-05,
+      "loss": 0.8858,
+      "step": 3110
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7239415990295657,
+      "learning_rate": 1.9901921697703503e-05,
+      "loss": 0.6228,
+      "step": 3111
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9149795304554664,
+      "learning_rate": 1.9901815062819437e-05,
+      "loss": 0.6668,
+      "step": 3112
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4186190413541364,
+      "learning_rate": 1.9901708370283866e-05,
+      "loss": 0.7345,
+      "step": 3113
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.076728983648862,
+      "learning_rate": 1.9901601620097417e-05,
+      "loss": 0.8146,
+      "step": 3114
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.708562015513445,
+      "learning_rate": 1.990149481226071e-05,
+      "loss": 0.8361,
+      "step": 3115
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6077534862700149,
+      "learning_rate": 1.9901387946774363e-05,
+      "loss": 0.5541,
+      "step": 3116
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.902004217227346,
+      "learning_rate": 1.9901281023639002e-05,
+      "loss": 0.6724,
+      "step": 3117
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4967954675558164,
+      "learning_rate": 1.9901174042855243e-05,
+      "loss": 0.7057,
+      "step": 3118
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5410493074225182,
+      "learning_rate": 1.9901067004423722e-05,
+      "loss": 0.7417,
+      "step": 3119
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5936319995566469,
+      "learning_rate": 1.990095990834505e-05,
+      "loss": 0.5382,
+      "step": 3120
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6053155512486155,
+      "learning_rate": 1.9900852754619853e-05,
+      "loss": 0.8187,
+      "step": 3121
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4916064931266784,
+      "learning_rate": 1.9900745543248762e-05,
+      "loss": 0.8033,
+      "step": 3122
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6199840348087198,
+      "learning_rate": 1.990063827423239e-05,
+      "loss": 0.7499,
+      "step": 3123
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.123118393787619,
+      "learning_rate": 1.990053094757137e-05,
+      "loss": 0.8615,
+      "step": 3124
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6435930377541175,
+      "learning_rate": 1.9900423563266325e-05,
+      "loss": 0.7627,
+      "step": 3125
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.58178058738618,
+      "learning_rate": 1.9900316121317883e-05,
+      "loss": 0.7846,
+      "step": 3126
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7230477136294458,
+      "learning_rate": 1.990020862172666e-05,
+      "loss": 0.8677,
+      "step": 3127
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4831324583474805,
+      "learning_rate": 1.990010106449329e-05,
+      "loss": 0.9177,
+      "step": 3128
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.562620328548427,
+      "learning_rate": 1.9899993449618397e-05,
+      "loss": 0.7831,
+      "step": 3129
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4960229260516313,
+      "learning_rate": 1.989988577710261e-05,
+      "loss": 0.7458,
+      "step": 3130
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8452966437835243,
+      "learning_rate": 1.9899778046946552e-05,
+      "loss": 0.7478,
+      "step": 3131
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8633276194312343,
+      "learning_rate": 1.9899670259150855e-05,
+      "loss": 0.7951,
+      "step": 3132
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5586786958813041,
+      "learning_rate": 1.9899562413716137e-05,
+      "loss": 0.6437,
+      "step": 3133
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6915031312201345,
+      "learning_rate": 1.989945451064304e-05,
+      "loss": 0.7945,
+      "step": 3134
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3997663118777763,
+      "learning_rate": 1.989934654993218e-05,
+      "loss": 0.7758,
+      "step": 3135
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8558377989366077,
+      "learning_rate": 1.989923853158419e-05,
+      "loss": 0.8207,
+      "step": 3136
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.647050411302329,
+      "learning_rate": 1.9899130455599703e-05,
+      "loss": 0.5333,
+      "step": 3137
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5332559928389515,
+      "learning_rate": 1.989902232197934e-05,
+      "loss": 0.8131,
+      "step": 3138
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.530673735906271,
+      "learning_rate": 1.989891413072374e-05,
+      "loss": 0.8315,
+      "step": 3139
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4531084735468784,
+      "learning_rate": 1.9898805881833523e-05,
+      "loss": 0.7481,
+      "step": 3140
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8986794676255683,
+      "learning_rate": 1.9898697575309326e-05,
+      "loss": 0.6796,
+      "step": 3141
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.415584013556386,
+      "learning_rate": 1.989858921115178e-05,
+      "loss": 0.7336,
+      "step": 3142
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.640406996739398,
+      "learning_rate": 1.989848078936151e-05,
+      "loss": 0.8124,
+      "step": 3143
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5166100416879849,
+      "learning_rate": 1.9898372309939153e-05,
+      "loss": 0.6931,
+      "step": 3144
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7560412743542302,
+      "learning_rate": 1.989826377288534e-05,
+      "loss": 0.8006,
+      "step": 3145
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5970742059389387,
+      "learning_rate": 1.98981551782007e-05,
+      "loss": 0.5401,
+      "step": 3146
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.556021821366498,
+      "learning_rate": 1.9898046525885868e-05,
+      "loss": 0.8659,
+      "step": 3147
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6169406573555103,
+      "learning_rate": 1.9897937815941474e-05,
+      "loss": 0.7553,
+      "step": 3148
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0091620206618193,
+      "learning_rate": 1.9897829048368153e-05,
+      "loss": 0.8912,
+      "step": 3149
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6090200644747453,
+      "learning_rate": 1.989772022316654e-05,
+      "loss": 0.5264,
+      "step": 3150
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6501879623900173,
+      "learning_rate": 1.9897611340337263e-05,
+      "loss": 0.7694,
+      "step": 3151
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7725259215780955,
+      "learning_rate": 1.989750239988096e-05,
+      "loss": 0.6885,
+      "step": 3152
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.584939518299469,
+      "learning_rate": 1.9897393401798265e-05,
+      "loss": 0.8113,
+      "step": 3153
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.464445686375496,
+      "learning_rate": 1.989728434608981e-05,
+      "loss": 0.7811,
+      "step": 3154
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5345480398885072,
+      "learning_rate": 1.9897175232756236e-05,
+      "loss": 0.8452,
+      "step": 3155
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.428390576504756,
+      "learning_rate": 1.9897066061798172e-05,
+      "loss": 0.8769,
+      "step": 3156
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6524130325713265,
+      "learning_rate": 1.9896956833216257e-05,
+      "loss": 0.5613,
+      "step": 3157
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8705742248264892,
+      "learning_rate": 1.9896847547011124e-05,
+      "loss": 0.724,
+      "step": 3158
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5312538410735854,
+      "learning_rate": 1.9896738203183413e-05,
+      "loss": 0.7761,
+      "step": 3159
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5654384133837314,
+      "learning_rate": 1.9896628801733763e-05,
+      "loss": 0.8982,
+      "step": 3160
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6018265900600024,
+      "learning_rate": 1.9896519342662798e-05,
+      "loss": 0.9214,
+      "step": 3161
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6702469786827103,
+      "learning_rate": 1.989640982597117e-05,
+      "loss": 0.552,
+      "step": 3162
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8556119911137876,
+      "learning_rate": 1.989630025165951e-05,
+      "loss": 0.7054,
+      "step": 3163
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.603816363282586,
+      "learning_rate": 1.9896190619728458e-05,
+      "loss": 0.5138,
+      "step": 3164
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9196795855899162,
+      "learning_rate": 1.9896080930178646e-05,
+      "loss": 0.739,
+      "step": 3165
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4831750765176828,
+      "learning_rate": 1.9895971183010726e-05,
+      "loss": 0.7575,
+      "step": 3166
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3437627842691438,
+      "learning_rate": 1.9895861378225324e-05,
+      "loss": 0.747,
+      "step": 3167
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.67132409496585,
+      "learning_rate": 1.989575151582308e-05,
+      "loss": 0.6892,
+      "step": 3168
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.54935993533513,
+      "learning_rate": 1.9895641595804644e-05,
+      "loss": 0.869,
+      "step": 3169
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4987862623494514,
+      "learning_rate": 1.9895531618170646e-05,
+      "loss": 0.7959,
+      "step": 3170
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.447416260238682,
+      "learning_rate": 1.9895421582921735e-05,
+      "loss": 0.6947,
+      "step": 3171
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.151301318987612,
+      "learning_rate": 1.989531149005854e-05,
+      "loss": 0.6699,
+      "step": 3172
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6411577198979926,
+      "learning_rate": 1.9895201339581717e-05,
+      "loss": 0.546,
+      "step": 3173
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.19962312796588,
+      "learning_rate": 1.989509113149189e-05,
+      "loss": 0.8986,
+      "step": 3174
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3976787966751867,
+      "learning_rate": 1.9894980865789716e-05,
+      "loss": 0.6771,
+      "step": 3175
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4695998697927466,
+      "learning_rate": 1.9894870542475827e-05,
+      "loss": 0.8011,
+      "step": 3176
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4186790868436452,
+      "learning_rate": 1.989476016155087e-05,
+      "loss": 0.8085,
+      "step": 3177
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6020681788113131,
+      "learning_rate": 1.9894649723015487e-05,
+      "loss": 0.5187,
+      "step": 3178
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.400392874585141,
+      "learning_rate": 1.9894539226870325e-05,
+      "loss": 0.7769,
+      "step": 3179
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6819442603215682,
+      "learning_rate": 1.9894428673116016e-05,
+      "loss": 0.8225,
+      "step": 3180
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.381773460889481,
+      "learning_rate": 1.9894318061753214e-05,
+      "loss": 0.8199,
+      "step": 3181
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4991074839738112,
+      "learning_rate": 1.989420739278256e-05,
+      "loss": 0.707,
+      "step": 3182
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4048546834718763,
+      "learning_rate": 1.98940966662047e-05,
+      "loss": 0.7332,
+      "step": 3183
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6719745587438155,
+      "learning_rate": 1.9893985882020272e-05,
+      "loss": 0.8783,
+      "step": 3184
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7230326854917286,
+      "learning_rate": 1.989387504022993e-05,
+      "loss": 0.8121,
+      "step": 3185
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.783623586573801,
+      "learning_rate": 1.9893764140834312e-05,
+      "loss": 0.744,
+      "step": 3186
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6834235339074795,
+      "learning_rate": 1.9893653183834068e-05,
+      "loss": 0.5431,
+      "step": 3187
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5426906934310851,
+      "learning_rate": 1.9893542169229844e-05,
+      "loss": 0.7895,
+      "step": 3188
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5136547686971527,
+      "learning_rate": 1.9893431097022284e-05,
+      "loss": 0.658,
+      "step": 3189
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6051636256296326,
+      "learning_rate": 1.989331996721204e-05,
+      "loss": 0.5131,
+      "step": 3190
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5077189697678872,
+      "learning_rate": 1.9893208779799747e-05,
+      "loss": 0.7686,
+      "step": 3191
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5445682813339652,
+      "learning_rate": 1.989309753478607e-05,
+      "loss": 0.7441,
+      "step": 3192
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6332761187046282,
+      "learning_rate": 1.989298623217164e-05,
+      "loss": 0.5305,
+      "step": 3193
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5275748937659626,
+      "learning_rate": 1.989287487195711e-05,
+      "loss": 0.7479,
+      "step": 3194
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.0088452942316533,
+      "learning_rate": 1.9892763454143134e-05,
+      "loss": 0.8245,
+      "step": 3195
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.553128384792828,
+      "learning_rate": 1.989265197873036e-05,
+      "loss": 0.8259,
+      "step": 3196
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8435431067787467,
+      "learning_rate": 1.989254044571943e-05,
+      "loss": 0.7326,
+      "step": 3197
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7666538899161461,
+      "learning_rate": 1.9892428855111e-05,
+      "loss": 0.8267,
+      "step": 3198
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8008835367947962,
+      "learning_rate": 1.989231720690572e-05,
+      "loss": 0.7139,
+      "step": 3199
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.019564023641987,
+      "learning_rate": 1.989220550110423e-05,
+      "loss": 0.7633,
+      "step": 3200
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6414843927129992,
+      "learning_rate": 1.989209373770719e-05,
+      "loss": 0.8652,
+      "step": 3201
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5566073141167005,
+      "learning_rate": 1.9891981916715254e-05,
+      "loss": 0.7466,
+      "step": 3202
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4973572897183218,
+      "learning_rate": 1.9891870038129066e-05,
+      "loss": 0.7532,
+      "step": 3203
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6606547271005558,
+      "learning_rate": 1.9891758101949275e-05,
+      "loss": 0.7683,
+      "step": 3204
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5783246457989324,
+      "learning_rate": 1.9891646108176536e-05,
+      "loss": 0.7539,
+      "step": 3205
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4863193053270223,
+      "learning_rate": 1.9891534056811507e-05,
+      "loss": 0.818,
+      "step": 3206
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.132117195183337,
+      "learning_rate": 1.9891421947854833e-05,
+      "loss": 0.8872,
+      "step": 3207
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5595661331525636,
+      "learning_rate": 1.9891309781307167e-05,
+      "loss": 0.7979,
+      "step": 3208
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7971349671877266,
+      "learning_rate": 1.9891197557169164e-05,
+      "loss": 0.7559,
+      "step": 3209
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5717837186473285,
+      "learning_rate": 1.989108527544148e-05,
+      "loss": 0.7183,
+      "step": 3210
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6358654465515945,
+      "learning_rate": 1.9890972936124764e-05,
+      "loss": 0.5578,
+      "step": 3211
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5784013486918662,
+      "learning_rate": 1.989086053921967e-05,
+      "loss": 0.7812,
+      "step": 3212
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6392218225219767,
+      "learning_rate": 1.989074808472686e-05,
+      "loss": 0.5514,
+      "step": 3213
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5702995292464578,
+      "learning_rate": 1.989063557264698e-05,
+      "loss": 0.5204,
+      "step": 3214
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5690696631636533,
+      "learning_rate": 1.9890523002980687e-05,
+      "loss": 0.4993,
+      "step": 3215
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6827335388873703,
+      "learning_rate": 1.989041037572864e-05,
+      "loss": 0.7264,
+      "step": 3216
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5149685409905933,
+      "learning_rate": 1.989029769089149e-05,
+      "loss": 0.7071,
+      "step": 3217
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5888803222285675,
+      "learning_rate": 1.98901849484699e-05,
+      "loss": 0.7874,
+      "step": 3218
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5092006137585043,
+      "learning_rate": 1.9890072148464518e-05,
+      "loss": 0.7221,
+      "step": 3219
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6435638416748481,
+      "learning_rate": 1.988995929087601e-05,
+      "loss": 0.8454,
+      "step": 3220
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6254987463344948,
+      "learning_rate": 1.988984637570502e-05,
+      "loss": 0.7404,
+      "step": 3221
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.573345799592635,
+      "learning_rate": 1.9889733402952222e-05,
+      "loss": 0.5419,
+      "step": 3222
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5929435788886297,
+      "learning_rate": 1.988962037261826e-05,
+      "loss": 0.8328,
+      "step": 3223
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7038857991717875,
+      "learning_rate": 1.98895072847038e-05,
+      "loss": 0.6529,
+      "step": 3224
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.584571926874539,
+      "learning_rate": 1.9889394139209492e-05,
+      "loss": 0.491,
+      "step": 3225
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7095794447485173,
+      "learning_rate": 1.9889280936136007e-05,
+      "loss": 0.8393,
+      "step": 3226
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4896703512783567,
+      "learning_rate": 1.9889167675483994e-05,
+      "loss": 0.7155,
+      "step": 3227
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9085967078028758,
+      "learning_rate": 1.9889054357254113e-05,
+      "loss": 0.7214,
+      "step": 3228
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5040812042460345,
+      "learning_rate": 1.988894098144703e-05,
+      "loss": 0.7627,
+      "step": 3229
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3898712099730413,
+      "learning_rate": 1.9888827548063404e-05,
+      "loss": 0.7804,
+      "step": 3230
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7744959233104098,
+      "learning_rate": 1.988871405710389e-05,
+      "loss": 0.7797,
+      "step": 3231
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3792527097823482,
+      "learning_rate": 1.9888600508569157e-05,
+      "loss": 0.6835,
+      "step": 3232
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8162367642887383,
+      "learning_rate": 1.9888486902459852e-05,
+      "loss": 0.7364,
+      "step": 3233
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7326489950432915,
+      "learning_rate": 1.9888373238776654e-05,
+      "loss": 0.8326,
+      "step": 3234
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.800043684203661,
+      "learning_rate": 1.9888259517520212e-05,
+      "loss": 0.6743,
+      "step": 3235
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.59785574497489,
+      "learning_rate": 1.9888145738691196e-05,
+      "loss": 0.6531,
+      "step": 3236
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6170221509258411,
+      "learning_rate": 1.9888031902290265e-05,
+      "loss": 0.7474,
+      "step": 3237
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.739617197995899,
+      "learning_rate": 1.9887918008318076e-05,
+      "loss": 0.7621,
+      "step": 3238
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3557406289024396,
+      "learning_rate": 1.98878040567753e-05,
+      "loss": 0.7328,
+      "step": 3239
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7139568403900585,
+      "learning_rate": 1.9887690047662604e-05,
+      "loss": 0.7262,
+      "step": 3240
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.765758405394224,
+      "learning_rate": 1.988757598098064e-05,
+      "loss": 0.8868,
+      "step": 3241
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6110258853730541,
+      "learning_rate": 1.988746185673008e-05,
+      "loss": 0.5451,
+      "step": 3242
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5885434859549092,
+      "learning_rate": 1.9887347674911587e-05,
+      "loss": 0.6625,
+      "step": 3243
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.971564154352719,
+      "learning_rate": 1.9887233435525827e-05,
+      "loss": 0.7538,
+      "step": 3244
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.549368263536402,
+      "learning_rate": 1.9887119138573462e-05,
+      "loss": 0.5972,
+      "step": 3245
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5315845074750603,
+      "learning_rate": 1.988700478405516e-05,
+      "loss": 0.6736,
+      "step": 3246
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.634899875193465,
+      "learning_rate": 1.9886890371971585e-05,
+      "loss": 0.84,
+      "step": 3247
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6124827275835016,
+      "learning_rate": 1.9886775902323405e-05,
+      "loss": 0.8572,
+      "step": 3248
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9012081025456102,
+      "learning_rate": 1.9886661375111286e-05,
+      "loss": 0.7563,
+      "step": 3249
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5886615674691702,
+      "learning_rate": 1.9886546790335895e-05,
+      "loss": 0.7827,
+      "step": 3250
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5289486862974482,
+      "learning_rate": 1.9886432147997897e-05,
+      "loss": 0.7893,
+      "step": 3251
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6192589908874535,
+      "learning_rate": 1.988631744809796e-05,
+      "loss": 0.7812,
+      "step": 3252
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5303409852265986,
+      "learning_rate": 1.9886202690636755e-05,
+      "loss": 0.8049,
+      "step": 3253
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.142859361558791,
+      "learning_rate": 1.9886087875614946e-05,
+      "loss": 0.7809,
+      "step": 3254
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4295963831117202,
+      "learning_rate": 1.9885973003033207e-05,
+      "loss": 0.9268,
+      "step": 3255
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.2624374901783972,
+      "learning_rate": 1.98858580728922e-05,
+      "loss": 0.7768,
+      "step": 3256
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4617994880882272,
+      "learning_rate": 1.9885743085192598e-05,
+      "loss": 0.844,
+      "step": 3257
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5668236110967649,
+      "learning_rate": 1.988562803993507e-05,
+      "loss": 0.5263,
+      "step": 3258
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.591899717357993,
+      "learning_rate": 1.988551293712029e-05,
+      "loss": 0.5092,
+      "step": 3259
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4047281807611458,
+      "learning_rate": 1.9885397776748918e-05,
+      "loss": 0.6369,
+      "step": 3260
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.544127408315355,
+      "learning_rate": 1.9885282558821633e-05,
+      "loss": 0.8208,
+      "step": 3261
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5860659033928358,
+      "learning_rate": 1.98851672833391e-05,
+      "loss": 0.5337,
+      "step": 3262
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4768272235942095,
+      "learning_rate": 1.9885051950301998e-05,
+      "loss": 0.8016,
+      "step": 3263
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5250546808127898,
+      "learning_rate": 1.9884936559710995e-05,
+      "loss": 0.5848,
+      "step": 3264
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3399964987114374,
+      "learning_rate": 1.988482111156676e-05,
+      "loss": 0.7785,
+      "step": 3265
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4991980184053266,
+      "learning_rate": 1.9884705605869965e-05,
+      "loss": 0.8014,
+      "step": 3266
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8139821045312317,
+      "learning_rate": 1.9884590042621284e-05,
+      "loss": 0.8282,
+      "step": 3267
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.536223110901778,
+      "learning_rate": 1.9884474421821393e-05,
+      "loss": 0.9,
+      "step": 3268
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7324776709703977,
+      "learning_rate": 1.9884358743470963e-05,
+      "loss": 0.7601,
+      "step": 3269
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4954506393542553,
+      "learning_rate": 1.9884243007570665e-05,
+      "loss": 0.8143,
+      "step": 3270
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6481221521381022,
+      "learning_rate": 1.9884127214121176e-05,
+      "loss": 0.73,
+      "step": 3271
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7143819863474938,
+      "learning_rate": 1.9884011363123168e-05,
+      "loss": 0.8312,
+      "step": 3272
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5276151936521993,
+      "learning_rate": 1.988389545457732e-05,
+      "loss": 0.7928,
+      "step": 3273
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.501118850327565,
+      "learning_rate": 1.9883779488484297e-05,
+      "loss": 0.8032,
+      "step": 3274
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5545307778146011,
+      "learning_rate": 1.9883663464844786e-05,
+      "loss": 0.8197,
+      "step": 3275
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6633541719448908,
+      "learning_rate": 1.9883547383659456e-05,
+      "loss": 0.7246,
+      "step": 3276
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.518114868653846,
+      "learning_rate": 1.9883431244928984e-05,
+      "loss": 0.6768,
+      "step": 3277
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7897697053308128,
+      "learning_rate": 1.9883315048654048e-05,
+      "loss": 0.6852,
+      "step": 3278
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5974999179797791,
+      "learning_rate": 1.988319879483532e-05,
+      "loss": 0.5104,
+      "step": 3279
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1189823505739827,
+      "learning_rate": 1.988308248347348e-05,
+      "loss": 0.8028,
+      "step": 3280
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4183576736906027,
+      "learning_rate": 1.9882966114569205e-05,
+      "loss": 0.7299,
+      "step": 3281
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5026853682063142,
+      "learning_rate": 1.9882849688123172e-05,
+      "loss": 0.8694,
+      "step": 3282
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6545962037507311,
+      "learning_rate": 1.988273320413606e-05,
+      "loss": 0.7658,
+      "step": 3283
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5846574828257516,
+      "learning_rate": 1.9882616662608547e-05,
+      "loss": 0.7112,
+      "step": 3284
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3833568393345084,
+      "learning_rate": 1.988250006354131e-05,
+      "loss": 0.7402,
+      "step": 3285
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5195678247568514,
+      "learning_rate": 1.988238340693503e-05,
+      "loss": 0.7724,
+      "step": 3286
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5489421722876386,
+      "learning_rate": 1.9882266692790387e-05,
+      "loss": 0.8621,
+      "step": 3287
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3457752716528435,
+      "learning_rate": 1.9882149921108053e-05,
+      "loss": 0.7554,
+      "step": 3288
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4587731679486702,
+      "learning_rate": 1.988203309188872e-05,
+      "loss": 0.7306,
+      "step": 3289
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8464184581292906,
+      "learning_rate": 1.9881916205133057e-05,
+      "loss": 0.6884,
+      "step": 3290
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.0567743754202787,
+      "learning_rate": 1.988179926084175e-05,
+      "loss": 0.7876,
+      "step": 3291
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7260552115840553,
+      "learning_rate": 1.988168225901548e-05,
+      "loss": 0.8827,
+      "step": 3292
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5245392536414992,
+      "learning_rate": 1.9881565199654926e-05,
+      "loss": 0.6399,
+      "step": 3293
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.600582336033481,
+      "learning_rate": 1.9881448082760774e-05,
+      "loss": 0.6667,
+      "step": 3294
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6619149133289424,
+      "learning_rate": 1.9881330908333702e-05,
+      "loss": 0.7261,
+      "step": 3295
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.73858160014401,
+      "learning_rate": 1.988121367637439e-05,
+      "loss": 0.8041,
+      "step": 3296
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5654864117442069,
+      "learning_rate": 1.9881096386883526e-05,
+      "loss": 0.7735,
+      "step": 3297
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4692504868300702,
+      "learning_rate": 1.988097903986179e-05,
+      "loss": 0.9262,
+      "step": 3298
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5522040011441216,
+      "learning_rate": 1.9880861635309867e-05,
+      "loss": 0.7546,
+      "step": 3299
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3727440930702206,
+      "learning_rate": 1.9880744173228437e-05,
+      "loss": 0.7076,
+      "step": 3300
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.489317683820206,
+      "learning_rate": 1.988062665361819e-05,
+      "loss": 0.7369,
+      "step": 3301
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.849089943111829,
+      "learning_rate": 1.9880509076479803e-05,
+      "loss": 0.8489,
+      "step": 3302
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6130745005712428,
+      "learning_rate": 1.9880391441813963e-05,
+      "loss": 0.5364,
+      "step": 3303
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.732707510013991,
+      "learning_rate": 1.988027374962136e-05,
+      "loss": 0.7973,
+      "step": 3304
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5857585770690696,
+      "learning_rate": 1.9880155999902672e-05,
+      "loss": 0.7713,
+      "step": 3305
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8446839191198465,
+      "learning_rate": 1.988003819265859e-05,
+      "loss": 0.7541,
+      "step": 3306
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.459681583393877,
+      "learning_rate": 1.98799203278898e-05,
+      "loss": 0.8823,
+      "step": 3307
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6018861228595083,
+      "learning_rate": 1.987980240559698e-05,
+      "loss": 0.7409,
+      "step": 3308
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5726883096400469,
+      "learning_rate": 1.9879684425780825e-05,
+      "loss": 0.5165,
+      "step": 3309
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4385646832853947,
+      "learning_rate": 1.987956638844202e-05,
+      "loss": 0.8165,
+      "step": 3310
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.352915410194381,
+      "learning_rate": 1.9879448293581253e-05,
+      "loss": 0.8005,
+      "step": 3311
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8142343955797562,
+      "learning_rate": 1.9879330141199206e-05,
+      "loss": 0.7227,
+      "step": 3312
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2651218945210017,
+      "learning_rate": 1.9879211931296572e-05,
+      "loss": 0.8226,
+      "step": 3313
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5400878619998035,
+      "learning_rate": 1.9879093663874044e-05,
+      "loss": 0.8247,
+      "step": 3314
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5415102074411702,
+      "learning_rate": 1.98789753389323e-05,
+      "loss": 0.9277,
+      "step": 3315
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9992116506662982,
+      "learning_rate": 1.9878856956472033e-05,
+      "loss": 0.7713,
+      "step": 3316
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6681058509790143,
+      "learning_rate": 1.987873851649394e-05,
+      "loss": 0.7794,
+      "step": 3317
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.140514535011797,
+      "learning_rate": 1.9878620018998696e-05,
+      "loss": 0.746,
+      "step": 3318
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4420228994381816,
+      "learning_rate": 1.9878501463987002e-05,
+      "loss": 0.7986,
+      "step": 3319
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5874353035388211,
+      "learning_rate": 1.9878382851459545e-05,
+      "loss": 0.707,
+      "step": 3320
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0067109726745076,
+      "learning_rate": 1.9878264181417016e-05,
+      "loss": 0.9171,
+      "step": 3321
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4287825462656973,
+      "learning_rate": 1.9878145453860106e-05,
+      "loss": 0.8068,
+      "step": 3322
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.5145886112475875,
+      "learning_rate": 1.9878026668789503e-05,
+      "loss": 0.9149,
+      "step": 3323
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5778322403492412,
+      "learning_rate": 1.9877907826205902e-05,
+      "loss": 0.5714,
+      "step": 3324
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5184481066780797,
+      "learning_rate": 1.9877788926109997e-05,
+      "loss": 0.8293,
+      "step": 3325
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4082335751213464,
+      "learning_rate": 1.9877669968502476e-05,
+      "loss": 0.7541,
+      "step": 3326
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6411538224418079,
+      "learning_rate": 1.9877550953384037e-05,
+      "loss": 0.8432,
+      "step": 3327
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5239799517900463,
+      "learning_rate": 1.9877431880755364e-05,
+      "loss": 0.7977,
+      "step": 3328
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6207553637371466,
+      "learning_rate": 1.9877312750617155e-05,
+      "loss": 0.6809,
+      "step": 3329
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6791526596581938,
+      "learning_rate": 1.9877193562970105e-05,
+      "loss": 0.7391,
+      "step": 3330
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.868853577890783,
+      "learning_rate": 1.987707431781491e-05,
+      "loss": 0.7591,
+      "step": 3331
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6742613219173874,
+      "learning_rate": 1.987695501515226e-05,
+      "loss": 0.7984,
+      "step": 3332
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4567224116764226,
+      "learning_rate": 1.9876835654982848e-05,
+      "loss": 0.7436,
+      "step": 3333
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6166549350247834,
+      "learning_rate": 1.9876716237307376e-05,
+      "loss": 0.7092,
+      "step": 3334
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4053581031506641,
+      "learning_rate": 1.9876596762126532e-05,
+      "loss": 0.8032,
+      "step": 3335
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.041027700665676,
+      "learning_rate": 1.9876477229441014e-05,
+      "loss": 0.7296,
+      "step": 3336
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7017578325123293,
+      "learning_rate": 1.987635763925152e-05,
+      "loss": 0.9876,
+      "step": 3337
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6963904153175446,
+      "learning_rate": 1.9876237991558746e-05,
+      "loss": 0.9264,
+      "step": 3338
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.565904464606207,
+      "learning_rate": 1.987611828636339e-05,
+      "loss": 0.8372,
+      "step": 3339
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9549402373633056,
+      "learning_rate": 1.987599852366614e-05,
+      "loss": 0.8521,
+      "step": 3340
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7634299177207513,
+      "learning_rate": 1.9875878703467707e-05,
+      "loss": 0.7387,
+      "step": 3341
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5434427207608747,
+      "learning_rate": 1.987575882576878e-05,
+      "loss": 0.7331,
+      "step": 3342
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.663042414269876,
+      "learning_rate": 1.987563889057006e-05,
+      "loss": 0.7173,
+      "step": 3343
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5625314090893256,
+      "learning_rate": 1.9875518897872236e-05,
+      "loss": 0.5184,
+      "step": 3344
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7240904702624522,
+      "learning_rate": 1.987539884767602e-05,
+      "loss": 0.8602,
+      "step": 3345
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4409970003015986,
+      "learning_rate": 1.9875278739982104e-05,
+      "loss": 0.7672,
+      "step": 3346
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6338263015826846,
+      "learning_rate": 1.9875158574791192e-05,
+      "loss": 0.7718,
+      "step": 3347
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9147777837187936,
+      "learning_rate": 1.9875038352103975e-05,
+      "loss": 0.8869,
+      "step": 3348
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6674742487013658,
+      "learning_rate": 1.9874918071921162e-05,
+      "loss": 0.6152,
+      "step": 3349
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5127396754057363,
+      "learning_rate": 1.9874797734243446e-05,
+      "loss": 0.8354,
+      "step": 3350
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.846848691238793,
+      "learning_rate": 1.9874677339071536e-05,
+      "loss": 0.6786,
+      "step": 3351
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.788620703357662,
+      "learning_rate": 1.9874556886406126e-05,
+      "loss": 0.7169,
+      "step": 3352
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6509790724043438,
+      "learning_rate": 1.987443637624792e-05,
+      "loss": 0.713,
+      "step": 3353
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6118321235805375,
+      "learning_rate": 1.987431580859762e-05,
+      "loss": 0.7275,
+      "step": 3354
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6226806911480839,
+      "learning_rate": 1.9874195183455925e-05,
+      "loss": 0.5344,
+      "step": 3355
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.797586634046015,
+      "learning_rate": 1.987407450082354e-05,
+      "loss": 0.7516,
+      "step": 3356
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7612105101034063,
+      "learning_rate": 1.987395376070117e-05,
+      "loss": 0.7577,
+      "step": 3357
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.516737608231185,
+      "learning_rate": 1.9873832963089512e-05,
+      "loss": 0.9166,
+      "step": 3358
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6748159952101713,
+      "learning_rate": 1.9873712107989273e-05,
+      "loss": 0.8241,
+      "step": 3359
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.684264705025116,
+      "learning_rate": 1.9873591195401157e-05,
+      "loss": 0.7758,
+      "step": 3360
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8373901428937245,
+      "learning_rate": 1.9873470225325868e-05,
+      "loss": 0.8215,
+      "step": 3361
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5400560876865117,
+      "learning_rate": 1.987334919776411e-05,
+      "loss": 0.672,
+      "step": 3362
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5231167447962066,
+      "learning_rate": 1.9873228112716583e-05,
+      "loss": 0.7571,
+      "step": 3363
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4254059242749486,
+      "learning_rate": 1.9873106970184e-05,
+      "loss": 0.7731,
+      "step": 3364
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0447634355604705,
+      "learning_rate": 1.987298577016706e-05,
+      "loss": 0.7524,
+      "step": 3365
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5076794163342824,
+      "learning_rate": 1.9872864512666473e-05,
+      "loss": 0.7008,
+      "step": 3366
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.521827187107359,
+      "learning_rate": 1.987274319768294e-05,
+      "loss": 0.6809,
+      "step": 3367
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4083960951462742,
+      "learning_rate": 1.9872621825217174e-05,
+      "loss": 0.7605,
+      "step": 3368
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5363624589081168,
+      "learning_rate": 1.9872500395269876e-05,
+      "loss": 0.8172,
+      "step": 3369
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.676801223776473,
+      "learning_rate": 1.9872378907841757e-05,
+      "loss": 0.6235,
+      "step": 3370
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3941451017116362,
+      "learning_rate": 1.987225736293352e-05,
+      "loss": 0.7791,
+      "step": 3371
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4835891185043226,
+      "learning_rate": 1.9872135760545877e-05,
+      "loss": 0.7945,
+      "step": 3372
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6231321308215312,
+      "learning_rate": 1.9872014100679533e-05,
+      "loss": 0.7623,
+      "step": 3373
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4759681372635831,
+      "learning_rate": 1.98718923833352e-05,
+      "loss": 0.7496,
+      "step": 3374
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5069638863495618,
+      "learning_rate": 1.987177060851358e-05,
+      "loss": 0.7731,
+      "step": 3375
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7126957240795726,
+      "learning_rate": 1.987164877621539e-05,
+      "loss": 0.6667,
+      "step": 3376
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6302308003844211,
+      "learning_rate": 1.9871526886441334e-05,
+      "loss": 0.5506,
+      "step": 3377
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9166384163808678,
+      "learning_rate": 1.9871404939192123e-05,
+      "loss": 0.6752,
+      "step": 3378
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.565365087950792,
+      "learning_rate": 1.9871282934468467e-05,
+      "loss": 0.7459,
+      "step": 3379
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4636840537562636,
+      "learning_rate": 1.9871160872271078e-05,
+      "loss": 0.8697,
+      "step": 3380
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6173397648839866,
+      "learning_rate": 1.9871038752600663e-05,
+      "loss": 0.53,
+      "step": 3381
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6367514559746805,
+      "learning_rate": 1.987091657545794e-05,
+      "loss": 0.5807,
+      "step": 3382
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8634024706737682,
+      "learning_rate": 1.9870794340843608e-05,
+      "loss": 0.891,
+      "step": 3383
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6462755440846253,
+      "learning_rate": 1.987067204875839e-05,
+      "loss": 0.5413,
+      "step": 3384
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7763053323173459,
+      "learning_rate": 1.9870549699202994e-05,
+      "loss": 0.7816,
+      "step": 3385
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.513032234369285,
+      "learning_rate": 1.9870427292178136e-05,
+      "loss": 0.8742,
+      "step": 3386
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.677752976810892,
+      "learning_rate": 1.9870304827684522e-05,
+      "loss": 0.7632,
+      "step": 3387
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4611498510855174,
+      "learning_rate": 1.9870182305722867e-05,
+      "loss": 0.73,
+      "step": 3388
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4804187364743986,
+      "learning_rate": 1.9870059726293892e-05,
+      "loss": 0.826,
+      "step": 3389
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7168598288021848,
+      "learning_rate": 1.9869937089398296e-05,
+      "loss": 0.7661,
+      "step": 3390
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.454264423772036,
+      "learning_rate": 1.986981439503681e-05,
+      "loss": 0.7909,
+      "step": 3391
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3735647966758862,
+      "learning_rate": 1.9869691643210134e-05,
+      "loss": 0.6458,
+      "step": 3392
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7134609840670283,
+      "learning_rate": 1.986956883391899e-05,
+      "loss": 0.7145,
+      "step": 3393
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6275550137002046,
+      "learning_rate": 1.9869445967164093e-05,
+      "loss": 0.8868,
+      "step": 3394
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5377131237902133,
+      "learning_rate": 1.9869323042946155e-05,
+      "loss": 0.8248,
+      "step": 3395
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3105126169159014,
+      "learning_rate": 1.9869200061265893e-05,
+      "loss": 0.6316,
+      "step": 3396
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.467160655777897,
+      "learning_rate": 1.9869077022124026e-05,
+      "loss": 0.8424,
+      "step": 3397
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.52154161935446,
+      "learning_rate": 1.9868953925521267e-05,
+      "loss": 0.6347,
+      "step": 3398
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7589116697795726,
+      "learning_rate": 1.9868830771458337e-05,
+      "loss": 0.6367,
+      "step": 3399
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8953594148937096,
+      "learning_rate": 1.9868707559935945e-05,
+      "loss": 0.8141,
+      "step": 3400
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4276833828152133,
+      "learning_rate": 1.9868584290954812e-05,
+      "loss": 0.6775,
+      "step": 3401
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4436125975983927,
+      "learning_rate": 1.9868460964515663e-05,
+      "loss": 0.7719,
+      "step": 3402
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6090605582800408,
+      "learning_rate": 1.9868337580619207e-05,
+      "loss": 0.6816,
+      "step": 3403
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6268296120192047,
+      "learning_rate": 1.9868214139266168e-05,
+      "loss": 0.7304,
+      "step": 3404
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6414677629071621,
+      "learning_rate": 1.9868090640457258e-05,
+      "loss": 0.6785,
+      "step": 3405
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4774970385201802,
+      "learning_rate": 1.98679670841932e-05,
+      "loss": 0.9016,
+      "step": 3406
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6947037967157155,
+      "learning_rate": 1.9867843470474714e-05,
+      "loss": 0.7553,
+      "step": 3407
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6641902405395228,
+      "learning_rate": 1.9867719799302524e-05,
+      "loss": 0.8351,
+      "step": 3408
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.394650030946087,
+      "learning_rate": 1.986759607067734e-05,
+      "loss": 0.7066,
+      "step": 3409
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5565846994419594,
+      "learning_rate": 1.9867472284599888e-05,
+      "loss": 0.8358,
+      "step": 3410
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6351053992438134,
+      "learning_rate": 1.986734844107089e-05,
+      "loss": 0.5267,
+      "step": 3411
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.4659425726539075,
+      "learning_rate": 1.9867224540091064e-05,
+      "loss": 0.8032,
+      "step": 3412
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5985706936164296,
+      "learning_rate": 1.9867100581661135e-05,
+      "loss": 0.5495,
+      "step": 3413
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7413592198284682,
+      "learning_rate": 1.9866976565781823e-05,
+      "loss": 0.7399,
+      "step": 3414
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5552373703052544,
+      "learning_rate": 1.9866852492453846e-05,
+      "loss": 0.7533,
+      "step": 3415
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6035304188868196,
+      "learning_rate": 1.986672836167793e-05,
+      "loss": 0.5458,
+      "step": 3416
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3956469044147424,
+      "learning_rate": 1.9866604173454802e-05,
+      "loss": 0.6885,
+      "step": 3417
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2831149058684843,
+      "learning_rate": 1.986647992778518e-05,
+      "loss": 0.8611,
+      "step": 3418
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6103948749212578,
+      "learning_rate": 1.986635562466978e-05,
+      "loss": 0.8504,
+      "step": 3419
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7479394164850983,
+      "learning_rate": 1.9866231264109343e-05,
+      "loss": 0.7126,
+      "step": 3420
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3903755074641686,
+      "learning_rate": 1.986610684610458e-05,
+      "loss": 0.8396,
+      "step": 3421
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.617136537153728,
+      "learning_rate": 1.9865982370656224e-05,
+      "loss": 0.7678,
+      "step": 3422
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8120050306981856,
+      "learning_rate": 1.9865857837764992e-05,
+      "loss": 0.876,
+      "step": 3423
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4132408423360394,
+      "learning_rate": 1.986573324743161e-05,
+      "loss": 0.7226,
+      "step": 3424
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1359693674398343,
+      "learning_rate": 1.986560859965681e-05,
+      "loss": 0.7136,
+      "step": 3425
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.495209533635916,
+      "learning_rate": 1.9865483894441312e-05,
+      "loss": 0.7738,
+      "step": 3426
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.722426091672363,
+      "learning_rate": 1.986535913178584e-05,
+      "loss": 0.7871,
+      "step": 3427
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6723580869712356,
+      "learning_rate": 1.9865234311691127e-05,
+      "loss": 0.7058,
+      "step": 3428
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6400111543016935,
+      "learning_rate": 1.9865109434157896e-05,
+      "loss": 0.7095,
+      "step": 3429
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6103202902525915,
+      "learning_rate": 1.9864984499186873e-05,
+      "loss": 0.6986,
+      "step": 3430
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5954779215292982,
+      "learning_rate": 1.9864859506778787e-05,
+      "loss": 0.6955,
+      "step": 3431
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7302149135384004,
+      "learning_rate": 1.9864734456934372e-05,
+      "loss": 0.7097,
+      "step": 3432
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3612183248290834,
+      "learning_rate": 1.986460934965434e-05,
+      "loss": 0.7551,
+      "step": 3433
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5175319836967338,
+      "learning_rate": 1.9864484184939438e-05,
+      "loss": 0.7002,
+      "step": 3434
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.413744810033997,
+      "learning_rate": 1.986435896279038e-05,
+      "loss": 0.7708,
+      "step": 3435
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6863774219311622,
+      "learning_rate": 1.9864233683207907e-05,
+      "loss": 0.7862,
+      "step": 3436
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5469846350394996,
+      "learning_rate": 1.986410834619274e-05,
+      "loss": 0.7643,
+      "step": 3437
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4721193552392684,
+      "learning_rate": 1.986398295174561e-05,
+      "loss": 0.7502,
+      "step": 3438
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6370216035163285,
+      "learning_rate": 1.9863857499867248e-05,
+      "loss": 0.709,
+      "step": 3439
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8683387155432924,
+      "learning_rate": 1.9863731990558386e-05,
+      "loss": 0.7748,
+      "step": 3440
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4158831980458289,
+      "learning_rate": 1.9863606423819753e-05,
+      "loss": 0.7971,
+      "step": 3441
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4788001597226383,
+      "learning_rate": 1.9863480799652078e-05,
+      "loss": 0.7317,
+      "step": 3442
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4854937493051275,
+      "learning_rate": 1.9863355118056102e-05,
+      "loss": 0.9044,
+      "step": 3443
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9250453673635808,
+      "learning_rate": 1.9863229379032542e-05,
+      "loss": 0.7663,
+      "step": 3444
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.482691258924289,
+      "learning_rate": 1.9863103582582144e-05,
+      "loss": 0.6673,
+      "step": 3445
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6501794136770063,
+      "learning_rate": 1.9862977728705633e-05,
+      "loss": 0.8325,
+      "step": 3446
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9918957593726876,
+      "learning_rate": 1.986285181740374e-05,
+      "loss": 0.6382,
+      "step": 3447
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8784901516401975,
+      "learning_rate": 1.98627258486772e-05,
+      "loss": 0.6567,
+      "step": 3448
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8907798263925923,
+      "learning_rate": 1.9862599822526755e-05,
+      "loss": 0.7206,
+      "step": 3449
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.604534943594097,
+      "learning_rate": 1.9862473738953127e-05,
+      "loss": 0.5436,
+      "step": 3450
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.560305790408745,
+      "learning_rate": 1.9862347597957054e-05,
+      "loss": 0.8589,
+      "step": 3451
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8047615183876362,
+      "learning_rate": 1.9862221399539273e-05,
+      "loss": 0.6915,
+      "step": 3452
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8775991241341599,
+      "learning_rate": 1.9862095143700512e-05,
+      "loss": 0.9164,
+      "step": 3453
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5533158543944372,
+      "learning_rate": 1.986196883044152e-05,
+      "loss": 0.824,
+      "step": 3454
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7790904558055773,
+      "learning_rate": 1.9861842459763014e-05,
+      "loss": 0.7433,
+      "step": 3455
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5823668108499365,
+      "learning_rate": 1.9861716031665744e-05,
+      "loss": 0.8171,
+      "step": 3456
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6749838162023992,
+      "learning_rate": 1.986158954615044e-05,
+      "loss": 0.5538,
+      "step": 3457
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.152858959308292,
+      "learning_rate": 1.986146300321784e-05,
+      "loss": 0.7878,
+      "step": 3458
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.671787522968839,
+      "learning_rate": 1.9861336402868678e-05,
+      "loss": 0.8022,
+      "step": 3459
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.551087515162292,
+      "learning_rate": 1.9861209745103693e-05,
+      "loss": 0.945,
+      "step": 3460
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7686082389941242,
+      "learning_rate": 1.9861083029923623e-05,
+      "loss": 0.7986,
+      "step": 3461
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.812508547385462,
+      "learning_rate": 1.986095625732921e-05,
+      "loss": 0.6265,
+      "step": 3462
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.56073279439471,
+      "learning_rate": 1.9860829427321184e-05,
+      "loss": 0.7523,
+      "step": 3463
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6581753952604608,
+      "learning_rate": 1.9860702539900288e-05,
+      "loss": 0.7315,
+      "step": 3464
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5344170498053358,
+      "learning_rate": 1.986057559506726e-05,
+      "loss": 0.7986,
+      "step": 3465
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4699730460479772,
+      "learning_rate": 1.9860448592822837e-05,
+      "loss": 0.7638,
+      "step": 3466
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5513992387141795,
+      "learning_rate": 1.9860321533167763e-05,
+      "loss": 0.6157,
+      "step": 3467
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4509172373001527,
+      "learning_rate": 1.9860194416102777e-05,
+      "loss": 0.7548,
+      "step": 3468
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.791715631571533,
+      "learning_rate": 1.9860067241628616e-05,
+      "loss": 0.6957,
+      "step": 3469
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6134536374032244,
+      "learning_rate": 1.985994000974602e-05,
+      "loss": 0.6148,
+      "step": 3470
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.068096037185549,
+      "learning_rate": 1.985981272045573e-05,
+      "loss": 0.8535,
+      "step": 3471
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5341951895236563,
+      "learning_rate": 1.9859685373758493e-05,
+      "loss": 0.8008,
+      "step": 3472
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0710375673036334,
+      "learning_rate": 1.9859557969655045e-05,
+      "loss": 0.8452,
+      "step": 3473
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6237556231055282,
+      "learning_rate": 1.9859430508146128e-05,
+      "loss": 0.5312,
+      "step": 3474
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.775579945961631,
+      "learning_rate": 1.9859302989232486e-05,
+      "loss": 0.8497,
+      "step": 3475
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6413791945590568,
+      "learning_rate": 1.9859175412914863e-05,
+      "loss": 0.8688,
+      "step": 3476
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6239240393044871,
+      "learning_rate": 1.9859047779194e-05,
+      "loss": 0.7984,
+      "step": 3477
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3972911832412616,
+      "learning_rate": 1.9858920088070636e-05,
+      "loss": 0.6392,
+      "step": 3478
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9081000604171674,
+      "learning_rate": 1.9858792339545515e-05,
+      "loss": 0.7339,
+      "step": 3479
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5884940835620883,
+      "learning_rate": 1.9858664533619388e-05,
+      "loss": 0.4965,
+      "step": 3480
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.376309157572116,
+      "learning_rate": 1.9858536670292995e-05,
+      "loss": 0.808,
+      "step": 3481
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4766181601098278,
+      "learning_rate": 1.985840874956708e-05,
+      "loss": 0.8071,
+      "step": 3482
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.781982741407804,
+      "learning_rate": 1.9858280771442387e-05,
+      "loss": 0.8783,
+      "step": 3483
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4789286366558116,
+      "learning_rate": 1.985815273591966e-05,
+      "loss": 0.752,
+      "step": 3484
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6430001910303806,
+      "learning_rate": 1.9858024642999655e-05,
+      "loss": 0.5278,
+      "step": 3485
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6941350021926198,
+      "learning_rate": 1.98578964926831e-05,
+      "loss": 0.7416,
+      "step": 3486
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9504813524049949,
+      "learning_rate": 1.9857768284970754e-05,
+      "loss": 0.7654,
+      "step": 3487
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4057000795398444,
+      "learning_rate": 1.9857640019863364e-05,
+      "loss": 0.7228,
+      "step": 3488
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6297610918061385,
+      "learning_rate": 1.985751169736167e-05,
+      "loss": 0.8878,
+      "step": 3489
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.635483517086121,
+      "learning_rate": 1.985738331746642e-05,
+      "loss": 0.6852,
+      "step": 3490
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6115841906183206,
+      "learning_rate": 1.9857254880178368e-05,
+      "loss": 0.7266,
+      "step": 3491
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.171488548880665,
+      "learning_rate": 1.9857126385498255e-05,
+      "loss": 0.7849,
+      "step": 3492
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3685264559171004,
+      "learning_rate": 1.9856997833426828e-05,
+      "loss": 0.913,
+      "step": 3493
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7511299515094427,
+      "learning_rate": 1.9856869223964844e-05,
+      "loss": 0.815,
+      "step": 3494
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7250822239091586,
+      "learning_rate": 1.9856740557113046e-05,
+      "loss": 0.906,
+      "step": 3495
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6547541937274262,
+      "learning_rate": 1.9856611832872185e-05,
+      "loss": 0.5742,
+      "step": 3496
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5599373444225844,
+      "learning_rate": 1.9856483051243005e-05,
+      "loss": 0.8251,
+      "step": 3497
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5440825483951228,
+      "learning_rate": 1.9856354212226262e-05,
+      "loss": 0.7193,
+      "step": 3498
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5095811328508322,
+      "learning_rate": 1.9856225315822705e-05,
+      "loss": 0.8514,
+      "step": 3499
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6371248597499901,
+      "learning_rate": 1.9856096362033083e-05,
+      "loss": 0.5616,
+      "step": 3500
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8371609293958644,
+      "learning_rate": 1.9855967350858147e-05,
+      "loss": 0.7764,
+      "step": 3501
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7154740899419278,
+      "learning_rate": 1.985583828229865e-05,
+      "loss": 0.8099,
+      "step": 3502
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.314139398845525,
+      "learning_rate": 1.985570915635534e-05,
+      "loss": 0.8376,
+      "step": 3503
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.613533535174664,
+      "learning_rate": 1.985557997302898e-05,
+      "loss": 0.7854,
+      "step": 3504
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.422183933746689,
+      "learning_rate": 1.98554507323203e-05,
+      "loss": 0.8524,
+      "step": 3505
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6124472804423693,
+      "learning_rate": 1.9855321434230075e-05,
+      "loss": 0.7078,
+      "step": 3506
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9066721594805005,
+      "learning_rate": 1.9855192078759045e-05,
+      "loss": 0.6694,
+      "step": 3507
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6019320259751775,
+      "learning_rate": 1.9855062665907967e-05,
+      "loss": 0.7905,
+      "step": 3508
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.405048998334979,
+      "learning_rate": 1.9854933195677596e-05,
+      "loss": 0.719,
+      "step": 3509
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7599992046573132,
+      "learning_rate": 1.9854803668068682e-05,
+      "loss": 0.8601,
+      "step": 3510
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9410716494786644,
+      "learning_rate": 1.985467408308198e-05,
+      "loss": 0.6344,
+      "step": 3511
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3176492256364065,
+      "learning_rate": 1.9854544440718247e-05,
+      "loss": 0.6176,
+      "step": 3512
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6624920156956755,
+      "learning_rate": 1.9854414740978235e-05,
+      "loss": 0.7676,
+      "step": 3513
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.448806619306175,
+      "learning_rate": 1.98542849838627e-05,
+      "loss": 0.8963,
+      "step": 3514
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3332994896057895,
+      "learning_rate": 1.98541551693724e-05,
+      "loss": 0.6545,
+      "step": 3515
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4744825064934521,
+      "learning_rate": 1.985402529750809e-05,
+      "loss": 0.8485,
+      "step": 3516
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5202672906584895,
+      "learning_rate": 1.9853895368270524e-05,
+      "loss": 0.8517,
+      "step": 3517
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8309921969240726,
+      "learning_rate": 1.9853765381660456e-05,
+      "loss": 0.8439,
+      "step": 3518
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.670915519630103,
+      "learning_rate": 1.985363533767865e-05,
+      "loss": 0.7823,
+      "step": 3519
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9122407880713743,
+      "learning_rate": 1.9853505236325857e-05,
+      "loss": 0.8573,
+      "step": 3520
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4979280966613828,
+      "learning_rate": 1.985337507760284e-05,
+      "loss": 0.8409,
+      "step": 3521
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5335120490678367,
+      "learning_rate": 1.985324486151035e-05,
+      "loss": 0.895,
+      "step": 3522
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6497982557102115,
+      "learning_rate": 1.9853114588049154e-05,
+      "loss": 0.7232,
+      "step": 3523
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.948412140453617,
+      "learning_rate": 1.9852984257220004e-05,
+      "loss": 0.6258,
+      "step": 3524
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.527343551292431,
+      "learning_rate": 1.9852853869023656e-05,
+      "loss": 0.749,
+      "step": 3525
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.966560808808369,
+      "learning_rate": 1.9852723423460878e-05,
+      "loss": 0.8818,
+      "step": 3526
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4811126823492653,
+      "learning_rate": 1.9852592920532424e-05,
+      "loss": 0.8075,
+      "step": 3527
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5697998514084069,
+      "learning_rate": 1.9852462360239053e-05,
+      "loss": 0.5611,
+      "step": 3528
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5219500161264052,
+      "learning_rate": 1.9852331742581528e-05,
+      "loss": 0.8369,
+      "step": 3529
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7451356831535347,
+      "learning_rate": 1.9852201067560607e-05,
+      "loss": 0.8979,
+      "step": 3530
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5202126300972791,
+      "learning_rate": 1.9852070335177057e-05,
+      "loss": 0.7924,
+      "step": 3531
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4787286021343675,
+      "learning_rate": 1.985193954543163e-05,
+      "loss": 0.64,
+      "step": 3532
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.685416138865628,
+      "learning_rate": 1.985180869832509e-05,
+      "loss": 0.754,
+      "step": 3533
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5938281431884298,
+      "learning_rate": 1.9851677793858206e-05,
+      "loss": 0.7367,
+      "step": 3534
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5024665718873567,
+      "learning_rate": 1.985154683203173e-05,
+      "loss": 0.8612,
+      "step": 3535
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4930216882639882,
+      "learning_rate": 1.9851415812846432e-05,
+      "loss": 0.7944,
+      "step": 3536
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.823248884406354,
+      "learning_rate": 1.985128473630307e-05,
+      "loss": 0.7821,
+      "step": 3537
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.587248254414328,
+      "learning_rate": 1.9851153602402414e-05,
+      "loss": 0.8405,
+      "step": 3538
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6986508450411617,
+      "learning_rate": 1.9851022411145222e-05,
+      "loss": 0.7981,
+      "step": 3539
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4716302247015838,
+      "learning_rate": 1.9850891162532255e-05,
+      "loss": 0.8345,
+      "step": 3540
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6330503778997452,
+      "learning_rate": 1.9850759856564285e-05,
+      "loss": 0.5723,
+      "step": 3541
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.7043863474888963,
+      "learning_rate": 1.9850628493242067e-05,
+      "loss": 0.6355,
+      "step": 3542
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.282762897444102,
+      "learning_rate": 1.9850497072566377e-05,
+      "loss": 0.8198,
+      "step": 3543
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4707537321040078,
+      "learning_rate": 1.9850365594537972e-05,
+      "loss": 0.7944,
+      "step": 3544
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.46055571085554,
+      "learning_rate": 1.985023405915762e-05,
+      "loss": 0.6127,
+      "step": 3545
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6172990484059462,
+      "learning_rate": 1.9850102466426086e-05,
+      "loss": 0.5414,
+      "step": 3546
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4060004945120423,
+      "learning_rate": 1.984997081634414e-05,
+      "loss": 0.7222,
+      "step": 3547
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.373377260074919,
+      "learning_rate": 1.9849839108912544e-05,
+      "loss": 0.663,
+      "step": 3548
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7546685066197267,
+      "learning_rate": 1.9849707344132065e-05,
+      "loss": 0.7815,
+      "step": 3549
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7137981892686767,
+      "learning_rate": 1.984957552200347e-05,
+      "loss": 0.6648,
+      "step": 3550
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5301534545992235,
+      "learning_rate": 1.984944364252753e-05,
+      "loss": 0.7051,
+      "step": 3551
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6569487432573888,
+      "learning_rate": 1.984931170570501e-05,
+      "loss": 0.6785,
+      "step": 3552
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6323834046791772,
+      "learning_rate": 1.984917971153668e-05,
+      "loss": 0.7646,
+      "step": 3553
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6367033606022816,
+      "learning_rate": 1.984904766002331e-05,
+      "loss": 0.7132,
+      "step": 3554
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6400985959817576,
+      "learning_rate": 1.984891555116566e-05,
+      "loss": 0.7972,
+      "step": 3555
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5716444340207084,
+      "learning_rate": 1.9848783384964513e-05,
+      "loss": 0.5699,
+      "step": 3556
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5123547843695562,
+      "learning_rate": 1.9848651161420625e-05,
+      "loss": 0.6923,
+      "step": 3557
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.476140864229296,
+      "learning_rate": 1.9848518880534775e-05,
+      "loss": 0.8089,
+      "step": 3558
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4034116208380718,
+      "learning_rate": 1.9848386542307728e-05,
+      "loss": 0.6193,
+      "step": 3559
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9218923782038078,
+      "learning_rate": 1.9848254146740257e-05,
+      "loss": 0.8889,
+      "step": 3560
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7319986130553353,
+      "learning_rate": 1.9848121693833135e-05,
+      "loss": 0.7676,
+      "step": 3561
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.178659823606541,
+      "learning_rate": 1.9847989183587127e-05,
+      "loss": 0.7738,
+      "step": 3562
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.639359558758116,
+      "learning_rate": 1.9847856616003013e-05,
+      "loss": 0.5725,
+      "step": 3563
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6105538850041796,
+      "learning_rate": 1.9847723991081555e-05,
+      "loss": 0.7877,
+      "step": 3564
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7602241515913217,
+      "learning_rate": 1.9847591308823535e-05,
+      "loss": 0.8267,
+      "step": 3565
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3218659559871302,
+      "learning_rate": 1.9847458569229716e-05,
+      "loss": 0.7757,
+      "step": 3566
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4377467916780478,
+      "learning_rate": 1.9847325772300878e-05,
+      "loss": 0.8348,
+      "step": 3567
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.5915650229986995,
+      "learning_rate": 1.984719291803779e-05,
+      "loss": 0.8058,
+      "step": 3568
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5606807674072831,
+      "learning_rate": 1.984706000644123e-05,
+      "loss": 0.7662,
+      "step": 3569
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.209733547369635,
+      "learning_rate": 1.984692703751197e-05,
+      "loss": 0.7452,
+      "step": 3570
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6039003138657945,
+      "learning_rate": 1.984679401125078e-05,
+      "loss": 0.5295,
+      "step": 3571
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5741163545423937,
+      "learning_rate": 1.9846660927658437e-05,
+      "loss": 0.8163,
+      "step": 3572
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6403235351076957,
+      "learning_rate": 1.9846527786735722e-05,
+      "loss": 0.8442,
+      "step": 3573
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6308643522300836,
+      "learning_rate": 1.98463945884834e-05,
+      "loss": 0.8883,
+      "step": 3574
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.666551592741426,
+      "learning_rate": 1.984626133290225e-05,
+      "loss": 0.8211,
+      "step": 3575
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.0981683536459865,
+      "learning_rate": 1.9846128019993054e-05,
+      "loss": 0.7482,
+      "step": 3576
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3818149826332076,
+      "learning_rate": 1.984599464975658e-05,
+      "loss": 0.7768,
+      "step": 3577
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3389239892389682,
+      "learning_rate": 1.984586122219361e-05,
+      "loss": 0.6192,
+      "step": 3578
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7992444728996952,
+      "learning_rate": 1.9845727737304917e-05,
+      "loss": 0.6812,
+      "step": 3579
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5597394521845958,
+      "learning_rate": 1.9845594195091283e-05,
+      "loss": 0.6477,
+      "step": 3580
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.303122886035851,
+      "learning_rate": 1.9845460595553482e-05,
+      "loss": 0.7,
+      "step": 3581
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5528550663225162,
+      "learning_rate": 1.9845326938692292e-05,
+      "loss": 0.6475,
+      "step": 3582
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.477958404684111,
+      "learning_rate": 1.984519322450849e-05,
+      "loss": 0.8419,
+      "step": 3583
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1695286302817185,
+      "learning_rate": 1.9845059453002858e-05,
+      "loss": 0.74,
+      "step": 3584
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6012989351465807,
+      "learning_rate": 1.984492562417617e-05,
+      "loss": 0.5569,
+      "step": 3585
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6424837720974717,
+      "learning_rate": 1.984479173802921e-05,
+      "loss": 0.7635,
+      "step": 3586
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5103258811036506,
+      "learning_rate": 1.9844657794562755e-05,
+      "loss": 0.8356,
+      "step": 3587
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.091971391643531,
+      "learning_rate": 1.984452379377759e-05,
+      "loss": 0.8562,
+      "step": 3588
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.637463622311053,
+      "learning_rate": 1.9844389735674484e-05,
+      "loss": 0.7489,
+      "step": 3589
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4514727202107405,
+      "learning_rate": 1.9844255620254227e-05,
+      "loss": 0.705,
+      "step": 3590
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7747645757947028,
+      "learning_rate": 1.98441214475176e-05,
+      "loss": 0.7218,
+      "step": 3591
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6182861926760367,
+      "learning_rate": 1.9843987217465378e-05,
+      "loss": 0.5739,
+      "step": 3592
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3733507972315022,
+      "learning_rate": 1.9843852930098345e-05,
+      "loss": 0.8449,
+      "step": 3593
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6475423203325668,
+      "learning_rate": 1.9843718585417284e-05,
+      "loss": 0.7197,
+      "step": 3594
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7566362791877659,
+      "learning_rate": 1.9843584183422983e-05,
+      "loss": 0.7794,
+      "step": 3595
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.269798668110093,
+      "learning_rate": 1.984344972411621e-05,
+      "loss": 0.7746,
+      "step": 3596
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5042046964513824,
+      "learning_rate": 1.984331520749776e-05,
+      "loss": 0.8554,
+      "step": 3597
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6207013369146204,
+      "learning_rate": 1.9843180633568412e-05,
+      "loss": 0.8583,
+      "step": 3598
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7050606448241827,
+      "learning_rate": 1.9843046002328953e-05,
+      "loss": 0.7905,
+      "step": 3599
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.478555023499585,
+      "learning_rate": 1.984291131378016e-05,
+      "loss": 0.7871,
+      "step": 3600
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.849695253881937,
+      "learning_rate": 1.9842776567922825e-05,
+      "loss": 0.7486,
+      "step": 3601
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4824956031992356,
+      "learning_rate": 1.9842641764757724e-05,
+      "loss": 0.757,
+      "step": 3602
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.478049062020568,
+      "learning_rate": 1.984250690428565e-05,
+      "loss": 0.7399,
+      "step": 3603
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5623527719022096,
+      "learning_rate": 1.9842371986507384e-05,
+      "loss": 0.5084,
+      "step": 3604
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3436091328925723,
+      "learning_rate": 1.9842237011423712e-05,
+      "loss": 0.7621,
+      "step": 3605
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4862439414135964,
+      "learning_rate": 1.9842101979035416e-05,
+      "loss": 0.624,
+      "step": 3606
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.434545371188221,
+      "learning_rate": 1.984196688934329e-05,
+      "loss": 0.7802,
+      "step": 3607
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7628219393817286,
+      "learning_rate": 1.984183174234812e-05,
+      "loss": 0.8678,
+      "step": 3608
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6237831994196783,
+      "learning_rate": 1.9841696538050686e-05,
+      "loss": 0.7221,
+      "step": 3609
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2388955726432846,
+      "learning_rate": 1.984156127645178e-05,
+      "loss": 0.7146,
+      "step": 3610
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5750698555055012,
+      "learning_rate": 1.9841425957552188e-05,
+      "loss": 0.8988,
+      "step": 3611
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5518788653932776,
+      "learning_rate": 1.98412905813527e-05,
+      "loss": 0.7759,
+      "step": 3612
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7885795119047403,
+      "learning_rate": 1.98411551478541e-05,
+      "loss": 0.7315,
+      "step": 3613
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5508630225077877,
+      "learning_rate": 1.984101965705718e-05,
+      "loss": 0.8915,
+      "step": 3614
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5492927207648106,
+      "learning_rate": 1.984088410896273e-05,
+      "loss": 0.6915,
+      "step": 3615
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5147953936544414,
+      "learning_rate": 1.9840748503571535e-05,
+      "loss": 0.8567,
+      "step": 3616
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6145684071714028,
+      "learning_rate": 1.9840612840884386e-05,
+      "loss": 0.5562,
+      "step": 3617
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5755623579115419,
+      "learning_rate": 1.9840477120902076e-05,
+      "loss": 0.7553,
+      "step": 3618
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7156752214851685,
+      "learning_rate": 1.984034134362539e-05,
+      "loss": 0.6875,
+      "step": 3619
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8535941321620895,
+      "learning_rate": 1.9840205509055123e-05,
+      "loss": 0.6279,
+      "step": 3620
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5742215270071366,
+      "learning_rate": 1.9840069617192063e-05,
+      "loss": 0.528,
+      "step": 3621
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.944001066128934,
+      "learning_rate": 1.9839933668037003e-05,
+      "loss": 0.7131,
+      "step": 3622
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5815600198373918,
+      "learning_rate": 1.9839797661590734e-05,
+      "loss": 0.7148,
+      "step": 3623
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6020461730875423,
+      "learning_rate": 1.9839661597854048e-05,
+      "loss": 0.5278,
+      "step": 3624
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4409155985191586,
+      "learning_rate": 1.983952547682774e-05,
+      "loss": 0.6896,
+      "step": 3625
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5992890075438417,
+      "learning_rate": 1.9839389298512593e-05,
+      "loss": 0.7816,
+      "step": 3626
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3095950853461658,
+      "learning_rate": 1.9839253062909408e-05,
+      "loss": 0.8775,
+      "step": 3627
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.552726440758047,
+      "learning_rate": 1.983911677001898e-05,
+      "loss": 0.8225,
+      "step": 3628
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.607000324382684,
+      "learning_rate": 1.98389804198421e-05,
+      "loss": 0.7516,
+      "step": 3629
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4405640374382669,
+      "learning_rate": 1.9838844012379557e-05,
+      "loss": 0.7263,
+      "step": 3630
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5956335054060036,
+      "learning_rate": 1.983870754763215e-05,
+      "loss": 0.5386,
+      "step": 3631
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5691805410926877,
+      "learning_rate": 1.9838571025600674e-05,
+      "loss": 0.554,
+      "step": 3632
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6265012530750445,
+      "learning_rate": 1.983843444628592e-05,
+      "loss": 0.8812,
+      "step": 3633
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.620289525792955,
+      "learning_rate": 1.983829780968869e-05,
+      "loss": 0.7681,
+      "step": 3634
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5748355473705136,
+      "learning_rate": 1.9838161115809773e-05,
+      "loss": 0.7231,
+      "step": 3635
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4079059850605888,
+      "learning_rate": 1.983802436464997e-05,
+      "loss": 0.7215,
+      "step": 3636
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6151627838941912,
+      "learning_rate": 1.983788755621007e-05,
+      "loss": 0.5519,
+      "step": 3637
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1193229625414767,
+      "learning_rate": 1.9837750690490876e-05,
+      "loss": 0.7859,
+      "step": 3638
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.315856945004765,
+      "learning_rate": 1.9837613767493186e-05,
+      "loss": 0.8104,
+      "step": 3639
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.475307670784385,
+      "learning_rate": 1.983747678721779e-05,
+      "loss": 0.7601,
+      "step": 3640
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5688302543891524,
+      "learning_rate": 1.983733974966549e-05,
+      "loss": 0.8735,
+      "step": 3641
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5857119599346035,
+      "learning_rate": 1.9837202654837085e-05,
+      "loss": 0.8172,
+      "step": 3642
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4675714366249328,
+      "learning_rate": 1.983706550273337e-05,
+      "loss": 0.8593,
+      "step": 3643
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5224827200890902,
+      "learning_rate": 1.9836928293355146e-05,
+      "loss": 0.755,
+      "step": 3644
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6086763377568297,
+      "learning_rate": 1.9836791026703213e-05,
+      "loss": 0.5712,
+      "step": 3645
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5465155884277633,
+      "learning_rate": 1.983665370277837e-05,
+      "loss": 0.6873,
+      "step": 3646
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6215127693440987,
+      "learning_rate": 1.9836516321581408e-05,
+      "loss": 0.7553,
+      "step": 3647
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5948429871996168,
+      "learning_rate": 1.9836378883113137e-05,
+      "loss": 0.5271,
+      "step": 3648
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8586527043573706,
+      "learning_rate": 1.9836241387374354e-05,
+      "loss": 0.7667,
+      "step": 3649
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.826281345290947,
+      "learning_rate": 1.9836103834365864e-05,
+      "loss": 0.7262,
+      "step": 3650
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.709413123869902,
+      "learning_rate": 1.983596622408846e-05,
+      "loss": 0.7819,
+      "step": 3651
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.380325596133877,
+      "learning_rate": 1.9835828556542944e-05,
+      "loss": 0.6678,
+      "step": 3652
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4990432390612138,
+      "learning_rate": 1.9835690831730125e-05,
+      "loss": 0.7875,
+      "step": 3653
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5838669890773642,
+      "learning_rate": 1.9835553049650796e-05,
+      "loss": 0.7831,
+      "step": 3654
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7074411835189076,
+      "learning_rate": 1.9835415210305763e-05,
+      "loss": 0.7764,
+      "step": 3655
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5843287745386169,
+      "learning_rate": 1.983527731369583e-05,
+      "loss": 0.5281,
+      "step": 3656
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.493570528785699,
+      "learning_rate": 1.98351393598218e-05,
+      "loss": 0.7478,
+      "step": 3657
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4162478382044674,
+      "learning_rate": 1.9835001348684476e-05,
+      "loss": 0.7446,
+      "step": 3658
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6511382469236673,
+      "learning_rate": 1.9834863280284657e-05,
+      "loss": 0.7448,
+      "step": 3659
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.137863678576031,
+      "learning_rate": 1.983472515462315e-05,
+      "loss": 0.7084,
+      "step": 3660
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7528484328638723,
+      "learning_rate": 1.983458697170076e-05,
+      "loss": 0.651,
+      "step": 3661
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4782605229565131,
+      "learning_rate": 1.9834448731518296e-05,
+      "loss": 0.8116,
+      "step": 3662
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.587089023361082,
+      "learning_rate": 1.9834310434076553e-05,
+      "loss": 0.5313,
+      "step": 3663
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.453888544014575,
+      "learning_rate": 1.9834172079376342e-05,
+      "loss": 0.7407,
+      "step": 3664
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.383073215004426,
+      "learning_rate": 1.983403366741847e-05,
+      "loss": 0.7954,
+      "step": 3665
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4498612969472358,
+      "learning_rate": 1.9833895198203737e-05,
+      "loss": 0.7266,
+      "step": 3666
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5157690473462337,
+      "learning_rate": 1.9833756671732956e-05,
+      "loss": 0.6901,
+      "step": 3667
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6927239403259262,
+      "learning_rate": 1.983361808800693e-05,
+      "loss": 0.7245,
+      "step": 3668
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3489942472367467,
+      "learning_rate": 1.9833479447026464e-05,
+      "loss": 0.8211,
+      "step": 3669
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3754425536133992,
+      "learning_rate": 1.983334074879237e-05,
+      "loss": 0.8117,
+      "step": 3670
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5569155559688588,
+      "learning_rate": 1.9833201993305452e-05,
+      "loss": 0.474,
+      "step": 3671
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7002605316887618,
+      "learning_rate": 1.9833063180566517e-05,
+      "loss": 0.9761,
+      "step": 3672
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.632531864416666,
+      "learning_rate": 1.9832924310576377e-05,
+      "loss": 0.8399,
+      "step": 3673
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6569504576692058,
+      "learning_rate": 1.983278538333584e-05,
+      "loss": 0.6019,
+      "step": 3674
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.49275980856463,
+      "learning_rate": 1.983264639884571e-05,
+      "loss": 0.852,
+      "step": 3675
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0102173761828994,
+      "learning_rate": 1.98325073571068e-05,
+      "loss": 0.7142,
+      "step": 3676
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.457265820578059,
+      "learning_rate": 1.983236825811992e-05,
+      "loss": 0.7254,
+      "step": 3677
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8295742644659543,
+      "learning_rate": 1.9832229101885882e-05,
+      "loss": 0.7299,
+      "step": 3678
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5728352958434468,
+      "learning_rate": 1.983208988840549e-05,
+      "loss": 0.474,
+      "step": 3679
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3437300352196067,
+      "learning_rate": 1.9831950617679562e-05,
+      "loss": 0.7975,
+      "step": 3680
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4595979494469387,
+      "learning_rate": 1.98318112897089e-05,
+      "loss": 0.7134,
+      "step": 3681
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4003524978545332,
+      "learning_rate": 1.983167190449432e-05,
+      "loss": 0.7602,
+      "step": 3682
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5780358305182398,
+      "learning_rate": 1.9831532462036634e-05,
+      "loss": 0.7124,
+      "step": 3683
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4508325941489562,
+      "learning_rate": 1.983139296233666e-05,
+      "loss": 0.8915,
+      "step": 3684
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4347387491823171,
+      "learning_rate": 1.9831253405395196e-05,
+      "loss": 0.8171,
+      "step": 3685
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5056373314533393,
+      "learning_rate": 1.9831113791213062e-05,
+      "loss": 0.6361,
+      "step": 3686
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4789379754118663,
+      "learning_rate": 1.9830974119791075e-05,
+      "loss": 0.7236,
+      "step": 3687
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5152346682973876,
+      "learning_rate": 1.9830834391130043e-05,
+      "loss": 0.708,
+      "step": 3688
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6536866552127298,
+      "learning_rate": 1.983069460523078e-05,
+      "loss": 0.7945,
+      "step": 3689
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9438168139180065,
+      "learning_rate": 1.98305547620941e-05,
+      "loss": 0.7389,
+      "step": 3690
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.743478027520577,
+      "learning_rate": 1.9830414861720818e-05,
+      "loss": 0.7542,
+      "step": 3691
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.59349348059202,
+      "learning_rate": 1.9830274904111754e-05,
+      "loss": 0.5752,
+      "step": 3692
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.626535839214717,
+      "learning_rate": 1.983013488926771e-05,
+      "loss": 0.7302,
+      "step": 3693
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5411129443003475,
+      "learning_rate": 1.9829994817189512e-05,
+      "loss": 0.6504,
+      "step": 3694
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6240214761995566,
+      "learning_rate": 1.982985468787797e-05,
+      "loss": 0.7853,
+      "step": 3695
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5937856163082125,
+      "learning_rate": 1.9829714501333905e-05,
+      "loss": 0.8005,
+      "step": 3696
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5502291621759363,
+      "learning_rate": 1.9829574257558128e-05,
+      "loss": 0.7387,
+      "step": 3697
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4649989324764658,
+      "learning_rate": 1.982943395655146e-05,
+      "loss": 0.8817,
+      "step": 3698
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4519900605648293,
+      "learning_rate": 1.982929359831471e-05,
+      "loss": 0.6853,
+      "step": 3699
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1266036919117592,
+      "learning_rate": 1.9829153182848705e-05,
+      "loss": 0.8477,
+      "step": 3700
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6461450611890158,
+      "learning_rate": 1.9829012710154258e-05,
+      "loss": 0.8735,
+      "step": 3701
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.805832698553379,
+      "learning_rate": 1.9828872180232185e-05,
+      "loss": 0.6646,
+      "step": 3702
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7291391363764126,
+      "learning_rate": 1.982873159308331e-05,
+      "loss": 0.8732,
+      "step": 3703
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.524872095877292,
+      "learning_rate": 1.9828590948708446e-05,
+      "loss": 0.9015,
+      "step": 3704
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.562414152396612,
+      "learning_rate": 1.9828450247108413e-05,
+      "loss": 0.6136,
+      "step": 3705
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3479669765127533,
+      "learning_rate": 1.982830948828403e-05,
+      "loss": 0.7003,
+      "step": 3706
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5092759217803524,
+      "learning_rate": 1.9828168672236125e-05,
+      "loss": 0.7684,
+      "step": 3707
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7587471206008576,
+      "learning_rate": 1.9828027798965502e-05,
+      "loss": 0.8139,
+      "step": 3708
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7426478885163708,
+      "learning_rate": 1.9827886868472993e-05,
+      "loss": 0.8158,
+      "step": 3709
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5890633838301889,
+      "learning_rate": 1.9827745880759414e-05,
+      "loss": 0.7362,
+      "step": 3710
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6172307964459119,
+      "learning_rate": 1.982760483582559e-05,
+      "loss": 0.5523,
+      "step": 3711
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6208013569539979,
+      "learning_rate": 1.9827463733672334e-05,
+      "loss": 0.5855,
+      "step": 3712
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5082808453616845,
+      "learning_rate": 1.9827322574300477e-05,
+      "loss": 0.7722,
+      "step": 3713
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9760120536918797,
+      "learning_rate": 1.9827181357710837e-05,
+      "loss": 0.7983,
+      "step": 3714
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6031633460077964,
+      "learning_rate": 1.9827040083904236e-05,
+      "loss": 0.5102,
+      "step": 3715
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8902233864434688,
+      "learning_rate": 1.9826898752881494e-05,
+      "loss": 0.8921,
+      "step": 3716
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4299988864348787,
+      "learning_rate": 1.9826757364643435e-05,
+      "loss": 0.7915,
+      "step": 3717
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6440940027306347,
+      "learning_rate": 1.9826615919190886e-05,
+      "loss": 0.5723,
+      "step": 3718
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4385677543416657,
+      "learning_rate": 1.9826474416524667e-05,
+      "loss": 0.8027,
+      "step": 3719
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5799792868230134,
+      "learning_rate": 1.9826332856645603e-05,
+      "loss": 0.4893,
+      "step": 3720
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1010741150963703,
+      "learning_rate": 1.9826191239554523e-05,
+      "loss": 0.7705,
+      "step": 3721
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.431278307996874,
+      "learning_rate": 1.9826049565252242e-05,
+      "loss": 0.6854,
+      "step": 3722
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4512808297547524,
+      "learning_rate": 1.982590783373959e-05,
+      "loss": 0.6979,
+      "step": 3723
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5874994502768554,
+      "learning_rate": 1.9825766045017395e-05,
+      "loss": 0.8052,
+      "step": 3724
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5214493283659745,
+      "learning_rate": 1.9825624199086473e-05,
+      "loss": 0.6799,
+      "step": 3725
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5996593229103636,
+      "learning_rate": 1.982548229594766e-05,
+      "loss": 0.5226,
+      "step": 3726
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.705749762695086,
+      "learning_rate": 1.9825340335601776e-05,
+      "loss": 0.7723,
+      "step": 3727
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6316887697781255,
+      "learning_rate": 1.9825198318049656e-05,
+      "loss": 0.7367,
+      "step": 3728
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3381741920431622,
+      "learning_rate": 1.9825056243292112e-05,
+      "loss": 0.7011,
+      "step": 3729
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4138963110078713,
+      "learning_rate": 1.9824914111329988e-05,
+      "loss": 0.7004,
+      "step": 3730
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4537540862792315,
+      "learning_rate": 1.98247719221641e-05,
+      "loss": 0.7077,
+      "step": 3731
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.455680535542723,
+      "learning_rate": 1.982462967579528e-05,
+      "loss": 0.9049,
+      "step": 3732
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5798909361877913,
+      "learning_rate": 1.9824487372224354e-05,
+      "loss": 0.4959,
+      "step": 3733
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.882910332508293,
+      "learning_rate": 1.9824345011452152e-05,
+      "loss": 0.7954,
+      "step": 3734
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.453819640537092,
+      "learning_rate": 1.9824202593479507e-05,
+      "loss": 0.7605,
+      "step": 3735
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4740459733795273,
+      "learning_rate": 1.982406011830724e-05,
+      "loss": 0.8597,
+      "step": 3736
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6274586193598549,
+      "learning_rate": 1.9823917585936185e-05,
+      "loss": 0.5085,
+      "step": 3737
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3540881005344914,
+      "learning_rate": 1.9823774996367177e-05,
+      "loss": 0.7386,
+      "step": 3738
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.10760350806038,
+      "learning_rate": 1.9823632349601033e-05,
+      "loss": 0.785,
+      "step": 3739
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5672355099810096,
+      "learning_rate": 1.9823489645638596e-05,
+      "loss": 0.5094,
+      "step": 3740
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8293898741358725,
+      "learning_rate": 1.9823346884480694e-05,
+      "loss": 0.8265,
+      "step": 3741
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3379632552908602,
+      "learning_rate": 1.982320406612815e-05,
+      "loss": 0.6592,
+      "step": 3742
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6304744943011276,
+      "learning_rate": 1.9823061190581807e-05,
+      "loss": 0.7743,
+      "step": 3743
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6257155013224559,
+      "learning_rate": 1.982291825784249e-05,
+      "loss": 0.4867,
+      "step": 3744
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.494020952273768,
+      "learning_rate": 1.9822775267911033e-05,
+      "loss": 0.7264,
+      "step": 3745
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.12366837662067,
+      "learning_rate": 1.982263222078827e-05,
+      "loss": 0.6959,
+      "step": 3746
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5701720100714218,
+      "learning_rate": 1.982248911647503e-05,
+      "loss": 0.784,
+      "step": 3747
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5454097957030624,
+      "learning_rate": 1.982234595497215e-05,
+      "loss": 0.781,
+      "step": 3748
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.752247683685489,
+      "learning_rate": 1.9822202736280462e-05,
+      "loss": 0.7441,
+      "step": 3749
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6175054113460954,
+      "learning_rate": 1.9822059460400804e-05,
+      "loss": 0.7635,
+      "step": 3750
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4182671895400518,
+      "learning_rate": 1.9821916127334e-05,
+      "loss": 0.6758,
+      "step": 3751
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7933944753520898,
+      "learning_rate": 1.9821772737080892e-05,
+      "loss": 0.7702,
+      "step": 3752
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8683785402132052,
+      "learning_rate": 1.9821629289642316e-05,
+      "loss": 0.7024,
+      "step": 3753
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.526907369560632,
+      "learning_rate": 1.9821485785019103e-05,
+      "loss": 0.7514,
+      "step": 3754
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3630100524945015,
+      "learning_rate": 1.982134222321209e-05,
+      "loss": 0.6229,
+      "step": 3755
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.508105616185708,
+      "learning_rate": 1.9821198604222114e-05,
+      "loss": 0.6743,
+      "step": 3756
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8994201710775602,
+      "learning_rate": 1.9821054928050012e-05,
+      "loss": 0.7654,
+      "step": 3757
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3581676923073065,
+      "learning_rate": 1.9820911194696613e-05,
+      "loss": 0.8202,
+      "step": 3758
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5662093706295737,
+      "learning_rate": 1.9820767404162765e-05,
+      "loss": 0.7591,
+      "step": 3759
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.632722481511032,
+      "learning_rate": 1.98206235564493e-05,
+      "loss": 0.7208,
+      "step": 3760
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7114986120017297,
+      "learning_rate": 1.9820479651557056e-05,
+      "loss": 0.7524,
+      "step": 3761
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.492125034848471,
+      "learning_rate": 1.982033568948687e-05,
+      "loss": 0.8061,
+      "step": 3762
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5005560368955113,
+      "learning_rate": 1.982019167023958e-05,
+      "loss": 0.8486,
+      "step": 3763
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6699102567681967,
+      "learning_rate": 1.9820047593816024e-05,
+      "loss": 0.6822,
+      "step": 3764
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4762449511919016,
+      "learning_rate": 1.981990346021704e-05,
+      "loss": 0.8227,
+      "step": 3765
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.54424430893915,
+      "learning_rate": 1.9819759269443475e-05,
+      "loss": 0.8181,
+      "step": 3766
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5489557390183704,
+      "learning_rate": 1.9819615021496158e-05,
+      "loss": 0.816,
+      "step": 3767
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6842630608040081,
+      "learning_rate": 1.9819470716375933e-05,
+      "loss": 0.7289,
+      "step": 3768
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5271072857097405,
+      "learning_rate": 1.9819326354083644e-05,
+      "loss": 0.6394,
+      "step": 3769
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8131756076223648,
+      "learning_rate": 1.9819181934620128e-05,
+      "loss": 0.7193,
+      "step": 3770
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6679724485722893,
+      "learning_rate": 1.9819037457986224e-05,
+      "loss": 0.8622,
+      "step": 3771
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5945891774116043,
+      "learning_rate": 1.9818892924182774e-05,
+      "loss": 0.7057,
+      "step": 3772
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6165819273638665,
+      "learning_rate": 1.9818748333210624e-05,
+      "loss": 0.5455,
+      "step": 3773
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8055633263333053,
+      "learning_rate": 1.981860368507061e-05,
+      "loss": 0.6942,
+      "step": 3774
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4912380898130388,
+      "learning_rate": 1.981845897976358e-05,
+      "loss": 0.9293,
+      "step": 3775
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4047922747813342,
+      "learning_rate": 1.9818314217290373e-05,
+      "loss": 0.82,
+      "step": 3776
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5470023661194492,
+      "learning_rate": 1.981816939765183e-05,
+      "loss": 0.8001,
+      "step": 3777
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.307937028845218,
+      "learning_rate": 1.9818024520848797e-05,
+      "loss": 0.7668,
+      "step": 3778
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4454508902449372,
+      "learning_rate": 1.9817879586882117e-05,
+      "loss": 0.8654,
+      "step": 3779
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9698813744130135,
+      "learning_rate": 1.9817734595752632e-05,
+      "loss": 0.6822,
+      "step": 3780
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.622279836910931,
+      "learning_rate": 1.9817589547461193e-05,
+      "loss": 0.5529,
+      "step": 3781
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.78113753783555,
+      "learning_rate": 1.9817444442008634e-05,
+      "loss": 0.7406,
+      "step": 3782
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6093508707069637,
+      "learning_rate": 1.981729927939581e-05,
+      "loss": 0.6471,
+      "step": 3783
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2954741265069356,
+      "learning_rate": 1.981715405962356e-05,
+      "loss": 0.8222,
+      "step": 3784
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5974173855843818,
+      "learning_rate": 1.981700878269273e-05,
+      "loss": 0.776,
+      "step": 3785
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5228423657545676,
+      "learning_rate": 1.9816863448604165e-05,
+      "loss": 0.8697,
+      "step": 3786
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.686181315055503,
+      "learning_rate": 1.9816718057358715e-05,
+      "loss": 0.7723,
+      "step": 3787
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.453120404757023,
+      "learning_rate": 1.9816572608957227e-05,
+      "loss": 0.7895,
+      "step": 3788
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4459351568777306,
+      "learning_rate": 1.981642710340054e-05,
+      "loss": 0.7069,
+      "step": 3789
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5309294880186795,
+      "learning_rate": 1.981628154068951e-05,
+      "loss": 0.9729,
+      "step": 3790
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5502771131575763,
+      "learning_rate": 1.981613592082498e-05,
+      "loss": 0.7079,
+      "step": 3791
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4437738039342416,
+      "learning_rate": 1.98159902438078e-05,
+      "loss": 0.6483,
+      "step": 3792
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.577029639095704,
+      "learning_rate": 1.9815844509638817e-05,
+      "loss": 0.6963,
+      "step": 3793
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8104504682939602,
+      "learning_rate": 1.981569871831888e-05,
+      "loss": 0.756,
+      "step": 3794
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8239323702151573,
+      "learning_rate": 1.9815552869848836e-05,
+      "loss": 0.8101,
+      "step": 3795
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6140255132959117,
+      "learning_rate": 1.9815406964229537e-05,
+      "loss": 0.5633,
+      "step": 3796
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.623883017939779,
+      "learning_rate": 1.981526100146183e-05,
+      "loss": 0.742,
+      "step": 3797
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.761840640082053,
+      "learning_rate": 1.9815114981546564e-05,
+      "loss": 0.7587,
+      "step": 3798
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6883720975003742,
+      "learning_rate": 1.9814968904484595e-05,
+      "loss": 0.7791,
+      "step": 3799
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4141188614487445,
+      "learning_rate": 1.981482277027677e-05,
+      "loss": 0.7144,
+      "step": 3800
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7182213515594629,
+      "learning_rate": 1.9814676578923936e-05,
+      "loss": 0.8471,
+      "step": 3801
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7129240077529186,
+      "learning_rate": 1.981453033042695e-05,
+      "loss": 0.8424,
+      "step": 3802
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4694811372106475,
+      "learning_rate": 1.981438402478666e-05,
+      "loss": 0.7901,
+      "step": 3803
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5071128058661127,
+      "learning_rate": 1.981423766200392e-05,
+      "loss": 0.6927,
+      "step": 3804
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4835605519314343,
+      "learning_rate": 1.9814091242079583e-05,
+      "loss": 0.7959,
+      "step": 3805
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7842823502029972,
+      "learning_rate": 1.98139447650145e-05,
+      "loss": 0.8591,
+      "step": 3806
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6783579251786358,
+      "learning_rate": 1.981379823080952e-05,
+      "loss": 0.8131,
+      "step": 3807
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4883675917362345,
+      "learning_rate": 1.98136516394655e-05,
+      "loss": 0.7955,
+      "step": 3808
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3964912420217364,
+      "learning_rate": 1.9813504990983293e-05,
+      "loss": 0.7948,
+      "step": 3809
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6878939011393599,
+      "learning_rate": 1.9813358285363754e-05,
+      "loss": 0.7925,
+      "step": 3810
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5821133294270853,
+      "learning_rate": 1.981321152260774e-05,
+      "loss": 0.6959,
+      "step": 3811
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6163840510450558,
+      "learning_rate": 1.9813064702716094e-05,
+      "loss": 0.7518,
+      "step": 3812
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6195109718048458,
+      "learning_rate": 1.9812917825689683e-05,
+      "loss": 0.5125,
+      "step": 3813
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0339972678968183,
+      "learning_rate": 1.981277089152936e-05,
+      "loss": 0.7878,
+      "step": 3814
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1719029104392975,
+      "learning_rate": 1.981262390023597e-05,
+      "loss": 0.8347,
+      "step": 3815
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5910712122924617,
+      "learning_rate": 1.9812476851810384e-05,
+      "loss": 0.509,
+      "step": 3816
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6056950384584936,
+      "learning_rate": 1.981232974625345e-05,
+      "loss": 0.5755,
+      "step": 3817
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5130184123778665,
+      "learning_rate": 1.9812182583566026e-05,
+      "loss": 0.7904,
+      "step": 3818
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4776290019700598,
+      "learning_rate": 1.981203536374897e-05,
+      "loss": 0.8453,
+      "step": 3819
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6346132668703919,
+      "learning_rate": 1.9811888086803137e-05,
+      "loss": 0.7274,
+      "step": 3820
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.143937688039417,
+      "learning_rate": 1.981174075272938e-05,
+      "loss": 0.7533,
+      "step": 3821
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1062534687777146,
+      "learning_rate": 1.9811593361528566e-05,
+      "loss": 0.809,
+      "step": 3822
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4774668141229226,
+      "learning_rate": 1.981144591320155e-05,
+      "loss": 0.6114,
+      "step": 3823
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.362730817253584,
+      "learning_rate": 1.9811298407749188e-05,
+      "loss": 0.7183,
+      "step": 3824
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.623921873358828,
+      "learning_rate": 1.9811150845172342e-05,
+      "loss": 0.738,
+      "step": 3825
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5778275507749968,
+      "learning_rate": 1.981100322547187e-05,
+      "loss": 0.7667,
+      "step": 3826
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7514657305080328,
+      "learning_rate": 1.9810855548648628e-05,
+      "loss": 0.7454,
+      "step": 3827
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4907463035365833,
+      "learning_rate": 1.981070781470348e-05,
+      "loss": 0.9799,
+      "step": 3828
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.688840655707054,
+      "learning_rate": 1.9810560023637285e-05,
+      "loss": 0.8465,
+      "step": 3829
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.556701990146972,
+      "learning_rate": 1.9810412175450907e-05,
+      "loss": 0.8329,
+      "step": 3830
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.456570051925649,
+      "learning_rate": 1.9810264270145198e-05,
+      "loss": 0.8486,
+      "step": 3831
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6493828556266084,
+      "learning_rate": 1.9810116307721028e-05,
+      "loss": 0.8052,
+      "step": 3832
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5996327161523347,
+      "learning_rate": 1.9809968288179254e-05,
+      "loss": 0.8495,
+      "step": 3833
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.584953404729357,
+      "learning_rate": 1.980982021152074e-05,
+      "loss": 0.6457,
+      "step": 3834
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4491715794459055,
+      "learning_rate": 1.9809672077746347e-05,
+      "loss": 0.7471,
+      "step": 3835
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.553965976587721,
+      "learning_rate": 1.9809523886856937e-05,
+      "loss": 0.8987,
+      "step": 3836
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.0253380082777315,
+      "learning_rate": 1.9809375638853373e-05,
+      "loss": 0.7641,
+      "step": 3837
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.852031757763416,
+      "learning_rate": 1.9809227333736518e-05,
+      "loss": 0.7202,
+      "step": 3838
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5511595872151513,
+      "learning_rate": 1.9809078971507234e-05,
+      "loss": 0.7318,
+      "step": 3839
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3957032579915019,
+      "learning_rate": 1.980893055216639e-05,
+      "loss": 0.698,
+      "step": 3840
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6558054507384032,
+      "learning_rate": 1.980878207571485e-05,
+      "loss": 0.8159,
+      "step": 3841
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4789183335340694,
+      "learning_rate": 1.980863354215347e-05,
+      "loss": 0.8074,
+      "step": 3842
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4051747784241733,
+      "learning_rate": 1.980848495148312e-05,
+      "loss": 0.7972,
+      "step": 3843
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.378717129285666,
+      "learning_rate": 1.9808336303704666e-05,
+      "loss": 0.8035,
+      "step": 3844
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5070578402529762,
+      "learning_rate": 1.9808187598818974e-05,
+      "loss": 0.7991,
+      "step": 3845
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.439282982604751,
+      "learning_rate": 1.9808038836826906e-05,
+      "loss": 0.6325,
+      "step": 3846
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.980239921499924,
+      "learning_rate": 1.980789001772933e-05,
+      "loss": 0.7285,
+      "step": 3847
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7666093941260443,
+      "learning_rate": 1.980774114152712e-05,
+      "loss": 0.8047,
+      "step": 3848
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6041771476024008,
+      "learning_rate": 1.980759220822113e-05,
+      "loss": 0.5438,
+      "step": 3849
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4774836362121713,
+      "learning_rate": 1.9807443217812236e-05,
+      "loss": 0.7401,
+      "step": 3850
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4611582198452984,
+      "learning_rate": 1.9807294170301303e-05,
+      "loss": 0.732,
+      "step": 3851
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4390960360772727,
+      "learning_rate": 1.9807145065689196e-05,
+      "loss": 0.8382,
+      "step": 3852
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.832287418731358,
+      "learning_rate": 1.9806995903976785e-05,
+      "loss": 0.6842,
+      "step": 3853
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5910298255500746,
+      "learning_rate": 1.980684668516494e-05,
+      "loss": 0.7207,
+      "step": 3854
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6959601108476816,
+      "learning_rate": 1.9806697409254532e-05,
+      "loss": 0.6662,
+      "step": 3855
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6090161998820638,
+      "learning_rate": 1.9806548076246425e-05,
+      "loss": 0.796,
+      "step": 3856
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.346580523590969,
+      "learning_rate": 1.9806398686141487e-05,
+      "loss": 0.7239,
+      "step": 3857
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8065641662972622,
+      "learning_rate": 1.9806249238940596e-05,
+      "loss": 0.7688,
+      "step": 3858
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5063822893384131,
+      "learning_rate": 1.9806099734644615e-05,
+      "loss": 0.7948,
+      "step": 3859
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4990146766691606,
+      "learning_rate": 1.980595017325442e-05,
+      "loss": 0.7556,
+      "step": 3860
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5829179078034898,
+      "learning_rate": 1.9805800554770872e-05,
+      "loss": 0.7325,
+      "step": 3861
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.340690066335146,
+      "learning_rate": 1.980565087919485e-05,
+      "loss": 0.7419,
+      "step": 3862
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.342972818770456,
+      "learning_rate": 1.9805501146527228e-05,
+      "loss": 0.7931,
+      "step": 3863
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6564437665110963,
+      "learning_rate": 1.9805351356768872e-05,
+      "loss": 0.8076,
+      "step": 3864
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8735919866848063,
+      "learning_rate": 1.980520150992066e-05,
+      "loss": 0.6964,
+      "step": 3865
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.425105219844916,
+      "learning_rate": 1.9805051605983454e-05,
+      "loss": 0.7183,
+      "step": 3866
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6150267791116735,
+      "learning_rate": 1.9804901644958137e-05,
+      "loss": 0.5184,
+      "step": 3867
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.744420880869139,
+      "learning_rate": 1.9804751626845577e-05,
+      "loss": 0.68,
+      "step": 3868
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5818360376470602,
+      "learning_rate": 1.9804601551646652e-05,
+      "loss": 0.8384,
+      "step": 3869
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6646202347646455,
+      "learning_rate": 1.980445141936223e-05,
+      "loss": 0.8531,
+      "step": 3870
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4878812581854821,
+      "learning_rate": 1.9804301229993185e-05,
+      "loss": 0.8809,
+      "step": 3871
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.900801564791282,
+      "learning_rate": 1.98041509835404e-05,
+      "loss": 0.8192,
+      "step": 3872
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7272300334307664,
+      "learning_rate": 1.9804000680004744e-05,
+      "loss": 0.8006,
+      "step": 3873
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.474255335668639,
+      "learning_rate": 1.9803850319387087e-05,
+      "loss": 0.842,
+      "step": 3874
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5626866334786609,
+      "learning_rate": 1.9803699901688313e-05,
+      "loss": 0.7679,
+      "step": 3875
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.409546171647367,
+      "learning_rate": 1.9803549426909293e-05,
+      "loss": 0.7772,
+      "step": 3876
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5450337200093163,
+      "learning_rate": 1.9803398895050905e-05,
+      "loss": 0.7382,
+      "step": 3877
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5243303019459196,
+      "learning_rate": 1.9803248306114026e-05,
+      "loss": 0.9296,
+      "step": 3878
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.877937198104162,
+      "learning_rate": 1.980309766009953e-05,
+      "loss": 0.7884,
+      "step": 3879
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6676154725177992,
+      "learning_rate": 1.9802946957008296e-05,
+      "loss": 0.6971,
+      "step": 3880
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5789873753162207,
+      "learning_rate": 1.98027961968412e-05,
+      "loss": 0.5422,
+      "step": 3881
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5378581737955763,
+      "learning_rate": 1.9802645379599122e-05,
+      "loss": 0.823,
+      "step": 3882
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4769441624654565,
+      "learning_rate": 1.980249450528294e-05,
+      "loss": 0.8136,
+      "step": 3883
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9015086779910138,
+      "learning_rate": 1.9802343573893536e-05,
+      "loss": 0.7988,
+      "step": 3884
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6537407046459538,
+      "learning_rate": 1.9802192585431778e-05,
+      "loss": 0.8151,
+      "step": 3885
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.453635310825634,
+      "learning_rate": 1.9802041539898553e-05,
+      "loss": 0.7097,
+      "step": 3886
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3377833570403186,
+      "learning_rate": 1.980189043729474e-05,
+      "loss": 0.6036,
+      "step": 3887
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6331633023669163,
+      "learning_rate": 1.9801739277621215e-05,
+      "loss": 0.7762,
+      "step": 3888
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.899458290163758,
+      "learning_rate": 1.980158806087886e-05,
+      "loss": 0.6546,
+      "step": 3889
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.063285888029732,
+      "learning_rate": 1.9801436787068563e-05,
+      "loss": 0.754,
+      "step": 3890
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6805278686112277,
+      "learning_rate": 1.980128545619119e-05,
+      "loss": 0.77,
+      "step": 3891
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4136724911737923,
+      "learning_rate": 1.9801134068247636e-05,
+      "loss": 0.7826,
+      "step": 3892
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6981693267946282,
+      "learning_rate": 1.9800982623238775e-05,
+      "loss": 0.5407,
+      "step": 3893
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.548736862000383,
+      "learning_rate": 1.9800831121165488e-05,
+      "loss": 0.6813,
+      "step": 3894
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3304685195954782,
+      "learning_rate": 1.9800679562028662e-05,
+      "loss": 0.6805,
+      "step": 3895
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5842843772644598,
+      "learning_rate": 1.9800527945829175e-05,
+      "loss": 0.4912,
+      "step": 3896
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5821010865026784,
+      "learning_rate": 1.9800376272567913e-05,
+      "loss": 0.5457,
+      "step": 3897
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3997203375763885,
+      "learning_rate": 1.9800224542245756e-05,
+      "loss": 0.7886,
+      "step": 3898
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3011163279855826,
+      "learning_rate": 1.980007275486359e-05,
+      "loss": 0.8085,
+      "step": 3899
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0453741121898315,
+      "learning_rate": 1.9799920910422296e-05,
+      "loss": 0.8136,
+      "step": 3900
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6525110287968597,
+      "learning_rate": 1.9799769008922762e-05,
+      "loss": 0.5771,
+      "step": 3901
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3534886289550185,
+      "learning_rate": 1.979961705036587e-05,
+      "loss": 0.7254,
+      "step": 3902
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6542021488081493,
+      "learning_rate": 1.9799465034752505e-05,
+      "loss": 0.7825,
+      "step": 3903
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.903717041917611,
+      "learning_rate": 1.979931296208355e-05,
+      "loss": 0.7013,
+      "step": 3904
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.427962170407232,
+      "learning_rate": 1.9799160832359896e-05,
+      "loss": 0.8392,
+      "step": 3905
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.993912275616886,
+      "learning_rate": 1.9799008645582424e-05,
+      "loss": 0.7861,
+      "step": 3906
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7634354714007088,
+      "learning_rate": 1.9798856401752017e-05,
+      "loss": 0.6746,
+      "step": 3907
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9482778737587032,
+      "learning_rate": 1.979870410086957e-05,
+      "loss": 0.744,
+      "step": 3908
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8721645656151897,
+      "learning_rate": 1.979855174293597e-05,
+      "loss": 0.8049,
+      "step": 3909
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6616463207480083,
+      "learning_rate": 1.9798399327952093e-05,
+      "loss": 0.5282,
+      "step": 3910
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5133339393340792,
+      "learning_rate": 1.9798246855918833e-05,
+      "loss": 0.6968,
+      "step": 3911
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.090961812003083,
+      "learning_rate": 1.979809432683708e-05,
+      "loss": 0.7254,
+      "step": 3912
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.551096449339519,
+      "learning_rate": 1.9797941740707718e-05,
+      "loss": 0.9026,
+      "step": 3913
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5305289309945118,
+      "learning_rate": 1.9797789097531638e-05,
+      "loss": 0.7999,
+      "step": 3914
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6337897270937368,
+      "learning_rate": 1.979763639730973e-05,
+      "loss": 0.5098,
+      "step": 3915
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.634808236914318,
+      "learning_rate": 1.979748364004288e-05,
+      "loss": 0.6329,
+      "step": 3916
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.302323432759879,
+      "learning_rate": 1.9797330825731976e-05,
+      "loss": 0.7021,
+      "step": 3917
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.956081351484375,
+      "learning_rate": 1.979717795437791e-05,
+      "loss": 0.7803,
+      "step": 3918
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5476610029843445,
+      "learning_rate": 1.9797025025981576e-05,
+      "loss": 0.9002,
+      "step": 3919
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.316884690921759,
+      "learning_rate": 1.979687204054386e-05,
+      "loss": 0.683,
+      "step": 3920
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5975251387208185,
+      "learning_rate": 1.9796718998065653e-05,
+      "loss": 0.7314,
+      "step": 3921
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.059300925001698,
+      "learning_rate": 1.9796565898547845e-05,
+      "loss": 0.7423,
+      "step": 3922
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.220065769317998,
+      "learning_rate": 1.979641274199133e-05,
+      "loss": 0.8563,
+      "step": 3923
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8205502039834072,
+      "learning_rate": 1.9796259528396998e-05,
+      "loss": 0.8588,
+      "step": 3924
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5227613330428738,
+      "learning_rate": 1.979610625776574e-05,
+      "loss": 0.9776,
+      "step": 3925
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0907363563428767,
+      "learning_rate": 1.9795952930098453e-05,
+      "loss": 0.8061,
+      "step": 3926
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.71062101334774,
+      "learning_rate": 1.9795799545396022e-05,
+      "loss": 0.7193,
+      "step": 3927
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4799574074723028,
+      "learning_rate": 1.9795646103659346e-05,
+      "loss": 0.7273,
+      "step": 3928
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3673048345420609,
+      "learning_rate": 1.9795492604889323e-05,
+      "loss": 0.6898,
+      "step": 3929
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7073074862440227,
+      "learning_rate": 1.9795339049086837e-05,
+      "loss": 0.7577,
+      "step": 3930
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5735070212973739,
+      "learning_rate": 1.9795185436252785e-05,
+      "loss": 0.8604,
+      "step": 3931
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3767025805656037,
+      "learning_rate": 1.9795031766388063e-05,
+      "loss": 0.5895,
+      "step": 3932
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0644801587084127,
+      "learning_rate": 1.9794878039493564e-05,
+      "loss": 0.8743,
+      "step": 3933
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.751158199649681,
+      "learning_rate": 1.9794724255570187e-05,
+      "loss": 0.7593,
+      "step": 3934
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4732954008388432,
+      "learning_rate": 1.9794570414618823e-05,
+      "loss": 0.7191,
+      "step": 3935
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4198334506215735,
+      "learning_rate": 1.9794416516640365e-05,
+      "loss": 0.7384,
+      "step": 3936
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9162814423349244,
+      "learning_rate": 1.9794262561635716e-05,
+      "loss": 0.8876,
+      "step": 3937
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.634032744775637,
+      "learning_rate": 1.9794108549605774e-05,
+      "loss": 0.8106,
+      "step": 3938
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4599805276610542,
+      "learning_rate": 1.979395448055143e-05,
+      "loss": 0.7439,
+      "step": 3939
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5872708353304592,
+      "learning_rate": 1.9793800354473575e-05,
+      "loss": 0.7488,
+      "step": 3940
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8691633563885446,
+      "learning_rate": 1.979364617137312e-05,
+      "loss": 0.8538,
+      "step": 3941
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4244919711072388,
+      "learning_rate": 1.9793491931250955e-05,
+      "loss": 0.858,
+      "step": 3942
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4897760126173758,
+      "learning_rate": 1.9793337634107977e-05,
+      "loss": 0.7324,
+      "step": 3943
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4256104557655385,
+      "learning_rate": 1.9793183279945093e-05,
+      "loss": 0.7734,
+      "step": 3944
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5775380795357757,
+      "learning_rate": 1.979302886876319e-05,
+      "loss": 0.5467,
+      "step": 3945
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5650568366253892,
+      "learning_rate": 1.9792874400563172e-05,
+      "loss": 0.8142,
+      "step": 3946
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.730849681486022,
+      "learning_rate": 1.979271987534594e-05,
+      "loss": 0.8155,
+      "step": 3947
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.62189601963019,
+      "learning_rate": 1.9792565293112396e-05,
+      "loss": 0.8299,
+      "step": 3948
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4884239769206689,
+      "learning_rate": 1.9792410653863436e-05,
+      "loss": 0.7743,
+      "step": 3949
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8466304557711797,
+      "learning_rate": 1.9792255957599956e-05,
+      "loss": 0.7172,
+      "step": 3950
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6379014128905844,
+      "learning_rate": 1.9792101204322866e-05,
+      "loss": 0.7719,
+      "step": 3951
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5261111010505146,
+      "learning_rate": 1.979194639403306e-05,
+      "loss": 0.8257,
+      "step": 3952
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8221788149879308,
+      "learning_rate": 1.9791791526731444e-05,
+      "loss": 0.725,
+      "step": 3953
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6599778889814332,
+      "learning_rate": 1.9791636602418916e-05,
+      "loss": 0.8855,
+      "step": 3954
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.339230921704936,
+      "learning_rate": 1.979148162109638e-05,
+      "loss": 0.6095,
+      "step": 3955
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4557886901085089,
+      "learning_rate": 1.9791326582764742e-05,
+      "loss": 0.6819,
+      "step": 3956
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3434710706298,
+      "learning_rate": 1.9791171487424895e-05,
+      "loss": 0.8244,
+      "step": 3957
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.565619792408031,
+      "learning_rate": 1.979101633507775e-05,
+      "loss": 0.792,
+      "step": 3958
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3339355179866341,
+      "learning_rate": 1.979086112572421e-05,
+      "loss": 0.8334,
+      "step": 3959
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6352733719539398,
+      "learning_rate": 1.9790705859365178e-05,
+      "loss": 0.9403,
+      "step": 3960
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6298520782322836,
+      "learning_rate": 1.9790550536001554e-05,
+      "loss": 0.5141,
+      "step": 3961
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6724644883188682,
+      "learning_rate": 1.9790395155634243e-05,
+      "loss": 0.833,
+      "step": 3962
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.941287495236605,
+      "learning_rate": 1.9790239718264155e-05,
+      "loss": 0.6266,
+      "step": 3963
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6186209333347537,
+      "learning_rate": 1.9790084223892192e-05,
+      "loss": 0.7407,
+      "step": 3964
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7083119745867212,
+      "learning_rate": 1.978992867251926e-05,
+      "loss": 0.8555,
+      "step": 3965
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4688579397481634,
+      "learning_rate": 1.9789773064146257e-05,
+      "loss": 0.8137,
+      "step": 3966
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5835302702856642,
+      "learning_rate": 1.97896173987741e-05,
+      "loss": 0.7226,
+      "step": 3967
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4815745269057516,
+      "learning_rate": 1.9789461676403694e-05,
+      "loss": 0.7462,
+      "step": 3968
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.952104177955424,
+      "learning_rate": 1.9789305897035942e-05,
+      "loss": 0.7423,
+      "step": 3969
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.738853176615611,
+      "learning_rate": 1.978915006067175e-05,
+      "loss": 0.7955,
+      "step": 3970
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9595782761684417,
+      "learning_rate": 1.978899416731203e-05,
+      "loss": 0.7313,
+      "step": 3971
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3884437119280428,
+      "learning_rate": 1.978883821695768e-05,
+      "loss": 0.6424,
+      "step": 3972
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2803031216004936,
+      "learning_rate": 1.978868220960962e-05,
+      "loss": 0.7827,
+      "step": 3973
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5120184083920394,
+      "learning_rate": 1.9788526145268752e-05,
+      "loss": 0.688,
+      "step": 3974
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.6602237182896773,
+      "learning_rate": 1.978837002393599e-05,
+      "loss": 0.8305,
+      "step": 3975
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6298674482322242,
+      "learning_rate": 1.9788213845612233e-05,
+      "loss": 0.6977,
+      "step": 3976
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.634779539981579,
+      "learning_rate": 1.9788057610298396e-05,
+      "loss": 0.7844,
+      "step": 3977
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.087806822519143,
+      "learning_rate": 1.9787901317995393e-05,
+      "loss": 0.7169,
+      "step": 3978
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2866201140596947,
+      "learning_rate": 1.9787744968704125e-05,
+      "loss": 0.8396,
+      "step": 3979
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4118220556277001,
+      "learning_rate": 1.9787588562425514e-05,
+      "loss": 0.8492,
+      "step": 3980
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6599503111732006,
+      "learning_rate": 1.9787432099160457e-05,
+      "loss": 0.8359,
+      "step": 3981
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.448716223772541,
+      "learning_rate": 1.9787275578909873e-05,
+      "loss": 0.6622,
+      "step": 3982
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7051729938892999,
+      "learning_rate": 1.9787119001674677e-05,
+      "loss": 0.7581,
+      "step": 3983
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.423667488155936,
+      "learning_rate": 1.978696236745577e-05,
+      "loss": 0.896,
+      "step": 3984
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5190708427951078,
+      "learning_rate": 1.978680567625407e-05,
+      "loss": 0.7062,
+      "step": 3985
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6375409766681739,
+      "learning_rate": 1.9786648928070493e-05,
+      "loss": 0.5056,
+      "step": 3986
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5600696248237604,
+      "learning_rate": 1.978649212290594e-05,
+      "loss": 0.6873,
+      "step": 3987
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5626514397374724,
+      "learning_rate": 1.978633526076134e-05,
+      "loss": 0.8298,
+      "step": 3988
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6175115064849326,
+      "learning_rate": 1.9786178341637594e-05,
+      "loss": 0.5315,
+      "step": 3989
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6165128666353814,
+      "learning_rate": 1.9786021365535616e-05,
+      "loss": 0.6778,
+      "step": 3990
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6785819447293997,
+      "learning_rate": 1.9785864332456327e-05,
+      "loss": 0.6897,
+      "step": 3991
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4331547560595572,
+      "learning_rate": 1.9785707242400638e-05,
+      "loss": 0.7104,
+      "step": 3992
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.822086117824942,
+      "learning_rate": 1.9785550095369465e-05,
+      "loss": 0.7887,
+      "step": 3993
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5622380978909531,
+      "learning_rate": 1.9785392891363718e-05,
+      "loss": 0.7481,
+      "step": 3994
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.7081505739250393,
+      "learning_rate": 1.9785235630384316e-05,
+      "loss": 0.7158,
+      "step": 3995
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.530812946868181,
+      "learning_rate": 1.9785078312432176e-05,
+      "loss": 0.635,
+      "step": 3996
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1772147885212028,
+      "learning_rate": 1.9784920937508208e-05,
+      "loss": 0.7642,
+      "step": 3997
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4505640222458414,
+      "learning_rate": 1.978476350561334e-05,
+      "loss": 0.8923,
+      "step": 3998
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6929907739098298,
+      "learning_rate": 1.9784606016748474e-05,
+      "loss": 0.8278,
+      "step": 3999
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5140947124436086,
+      "learning_rate": 1.9784448470914538e-05,
+      "loss": 0.6921,
+      "step": 4000
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.688055357401726,
+      "learning_rate": 1.9784290868112442e-05,
+      "loss": 0.7589,
+      "step": 4001
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5286390749479275,
+      "learning_rate": 1.978413320834311e-05,
+      "loss": 0.7855,
+      "step": 4002
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8942288809798993,
+      "learning_rate": 1.9783975491607457e-05,
+      "loss": 0.7708,
+      "step": 4003
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.082279165114556,
+      "learning_rate": 1.9783817717906398e-05,
+      "loss": 0.6536,
+      "step": 4004
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5607806493797394,
+      "learning_rate": 1.9783659887240855e-05,
+      "loss": 0.7879,
+      "step": 4005
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.583954431899245,
+      "learning_rate": 1.978350199961175e-05,
+      "loss": 0.8556,
+      "step": 4006
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.534486104051205,
+      "learning_rate": 1.978334405502e-05,
+      "loss": 0.7459,
+      "step": 4007
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3571275739575077,
+      "learning_rate": 1.9783186053466518e-05,
+      "loss": 0.7873,
+      "step": 4008
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7042182396360135,
+      "learning_rate": 1.9783027994952233e-05,
+      "loss": 0.8141,
+      "step": 4009
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.554898533050877,
+      "learning_rate": 1.9782869879478058e-05,
+      "loss": 0.8249,
+      "step": 4010
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.665839809814368,
+      "learning_rate": 1.9782711707044923e-05,
+      "loss": 0.6415,
+      "step": 4011
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6919328690564175,
+      "learning_rate": 1.9782553477653738e-05,
+      "loss": 0.7603,
+      "step": 4012
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5826929260566366,
+      "learning_rate": 1.9782395191305436e-05,
+      "loss": 0.523,
+      "step": 4013
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5627034795970771,
+      "learning_rate": 1.9782236848000927e-05,
+      "loss": 0.8197,
+      "step": 4014
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6480785840173737,
+      "learning_rate": 1.9782078447741136e-05,
+      "loss": 1.0044,
+      "step": 4015
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7578464114706278,
+      "learning_rate": 1.9781919990526993e-05,
+      "loss": 0.8236,
+      "step": 4016
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6023140999862644,
+      "learning_rate": 1.9781761476359414e-05,
+      "loss": 0.5342,
+      "step": 4017
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.873562680857618,
+      "learning_rate": 1.9781602905239323e-05,
+      "loss": 0.8563,
+      "step": 4018
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.746199257834435,
+      "learning_rate": 1.978144427716764e-05,
+      "loss": 0.7507,
+      "step": 4019
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6921274135512556,
+      "learning_rate": 1.9781285592145292e-05,
+      "loss": 0.7825,
+      "step": 4020
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.011839208104041,
+      "learning_rate": 1.9781126850173204e-05,
+      "loss": 0.877,
+      "step": 4021
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5961222586309729,
+      "learning_rate": 1.97809680512523e-05,
+      "loss": 0.5443,
+      "step": 4022
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3582078850006767,
+      "learning_rate": 1.9780809195383498e-05,
+      "loss": 0.7187,
+      "step": 4023
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.514736558482407,
+      "learning_rate": 1.9780650282567737e-05,
+      "loss": 0.8218,
+      "step": 4024
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5042711861821005,
+      "learning_rate": 1.9780491312805927e-05,
+      "loss": 0.9292,
+      "step": 4025
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4394688910195277,
+      "learning_rate": 1.9780332286099004e-05,
+      "loss": 0.8506,
+      "step": 4026
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7363368224206004,
+      "learning_rate": 1.9780173202447885e-05,
+      "loss": 0.8269,
+      "step": 4027
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6379319978365354,
+      "learning_rate": 1.9780014061853506e-05,
+      "loss": 0.7923,
+      "step": 4028
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.977417911443315,
+      "learning_rate": 1.9779854864316784e-05,
+      "loss": 0.846,
+      "step": 4029
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.6965714885867755,
+      "learning_rate": 1.9779695609838656e-05,
+      "loss": 0.7766,
+      "step": 4030
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5249730125513812,
+      "learning_rate": 1.9779536298420042e-05,
+      "loss": 0.8215,
+      "step": 4031
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4667665046706266,
+      "learning_rate": 1.977937693006187e-05,
+      "loss": 0.7692,
+      "step": 4032
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6312060429726165,
+      "learning_rate": 1.9779217504765072e-05,
+      "loss": 0.9186,
+      "step": 4033
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.982870769003874,
+      "learning_rate": 1.9779058022530572e-05,
+      "loss": 0.7475,
+      "step": 4034
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8722715929776739,
+      "learning_rate": 1.97788984833593e-05,
+      "loss": 0.7743,
+      "step": 4035
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6777523753413623,
+      "learning_rate": 1.9778738887252187e-05,
+      "loss": 0.679,
+      "step": 4036
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6272787351036808,
+      "learning_rate": 1.977857923421016e-05,
+      "loss": 0.5549,
+      "step": 4037
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.471286763242108,
+      "learning_rate": 1.9778419524234148e-05,
+      "loss": 0.7794,
+      "step": 4038
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5870283161238044,
+      "learning_rate": 1.977825975732508e-05,
+      "loss": 0.4968,
+      "step": 4039
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5557051924819034,
+      "learning_rate": 1.977809993348389e-05,
+      "loss": 0.6712,
+      "step": 4040
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0847087552303285,
+      "learning_rate": 1.9777940052711506e-05,
+      "loss": 0.7937,
+      "step": 4041
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.737951766246846,
+      "learning_rate": 1.977778011500886e-05,
+      "loss": 0.8108,
+      "step": 4042
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7244125662290276,
+      "learning_rate": 1.977762012037688e-05,
+      "loss": 0.5796,
+      "step": 4043
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7870207023574596,
+      "learning_rate": 1.9777460068816503e-05,
+      "loss": 0.6878,
+      "step": 4044
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6454188443966327,
+      "learning_rate": 1.9777299960328657e-05,
+      "loss": 0.7821,
+      "step": 4045
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.904760964418331,
+      "learning_rate": 1.9777139794914276e-05,
+      "loss": 0.7454,
+      "step": 4046
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5812639394230171,
+      "learning_rate": 1.9776979572574292e-05,
+      "loss": 0.5626,
+      "step": 4047
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5602678272718224,
+      "learning_rate": 1.9776819293309636e-05,
+      "loss": 0.8017,
+      "step": 4048
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5461900784195954,
+      "learning_rate": 1.9776658957121242e-05,
+      "loss": 0.8066,
+      "step": 4049
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5707060714963739,
+      "learning_rate": 1.9776498564010045e-05,
+      "loss": 0.845,
+      "step": 4050
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9786432988913343,
+      "learning_rate": 1.977633811397698e-05,
+      "loss": 0.6804,
+      "step": 4051
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8171411544632166,
+      "learning_rate": 1.9776177607022976e-05,
+      "loss": 0.6971,
+      "step": 4052
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6333810313725081,
+      "learning_rate": 1.9776017043148975e-05,
+      "loss": 0.5491,
+      "step": 4053
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1129419320153366,
+      "learning_rate": 1.9775856422355904e-05,
+      "loss": 0.8064,
+      "step": 4054
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.690511849567793,
+      "learning_rate": 1.9775695744644706e-05,
+      "loss": 0.6663,
+      "step": 4055
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6822346911591195,
+      "learning_rate": 1.977553501001631e-05,
+      "loss": 0.8667,
+      "step": 4056
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4244436196847232,
+      "learning_rate": 1.9775374218471653e-05,
+      "loss": 0.7371,
+      "step": 4057
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8832087209628188,
+      "learning_rate": 1.9775213370011675e-05,
+      "loss": 0.877,
+      "step": 4058
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5178283519689573,
+      "learning_rate": 1.9775052464637306e-05,
+      "loss": 0.7341,
+      "step": 4059
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5179463135737479,
+      "learning_rate": 1.9774891502349493e-05,
+      "loss": 0.7144,
+      "step": 4060
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.984492301979186,
+      "learning_rate": 1.9774730483149166e-05,
+      "loss": 0.6832,
+      "step": 4061
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.669583521830396,
+      "learning_rate": 1.977456940703726e-05,
+      "loss": 0.8372,
+      "step": 4062
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.791920704253905,
+      "learning_rate": 1.9774408274014715e-05,
+      "loss": 0.7818,
+      "step": 4063
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3528709273456139,
+      "learning_rate": 1.9774247084082474e-05,
+      "loss": 0.5954,
+      "step": 4064
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4091546440947589,
+      "learning_rate": 1.977408583724147e-05,
+      "loss": 0.7321,
+      "step": 4065
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5963681032364302,
+      "learning_rate": 1.9773924533492646e-05,
+      "loss": 0.5445,
+      "step": 4066
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5844218095583529,
+      "learning_rate": 1.9773763172836936e-05,
+      "loss": 0.5809,
+      "step": 4067
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6767922991515505,
+      "learning_rate": 1.9773601755275287e-05,
+      "loss": 0.6497,
+      "step": 4068
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4547659247612696,
+      "learning_rate": 1.9773440280808633e-05,
+      "loss": 0.9048,
+      "step": 4069
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8790422297307188,
+      "learning_rate": 1.9773278749437912e-05,
+      "loss": 0.6312,
+      "step": 4070
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.565790141715809,
+      "learning_rate": 1.977311716116407e-05,
+      "loss": 0.7606,
+      "step": 4071
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8708145468455513,
+      "learning_rate": 1.9772955515988045e-05,
+      "loss": 0.6506,
+      "step": 4072
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5640504994796443,
+      "learning_rate": 1.9772793813910783e-05,
+      "loss": 0.5486,
+      "step": 4073
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3564138824543417,
+      "learning_rate": 1.9772632054933216e-05,
+      "loss": 0.8761,
+      "step": 4074
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.587216082612707,
+      "learning_rate": 1.9772470239056295e-05,
+      "loss": 0.8038,
+      "step": 4075
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5959037345506153,
+      "learning_rate": 1.9772308366280955e-05,
+      "loss": 0.5257,
+      "step": 4076
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6361767637902078,
+      "learning_rate": 1.9772146436608143e-05,
+      "loss": 0.7577,
+      "step": 4077
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5701062320580035,
+      "learning_rate": 1.9771984450038802e-05,
+      "loss": 0.7603,
+      "step": 4078
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.125450089336853,
+      "learning_rate": 1.9771822406573876e-05,
+      "loss": 0.7011,
+      "step": 4079
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5256587841905862,
+      "learning_rate": 1.97716603062143e-05,
+      "loss": 0.6511,
+      "step": 4080
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.613225871409562,
+      "learning_rate": 1.9771498148961027e-05,
+      "loss": 0.8549,
+      "step": 4081
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7155220074895545,
+      "learning_rate": 1.9771335934815e-05,
+      "loss": 0.8442,
+      "step": 4082
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6262999152944595,
+      "learning_rate": 1.977117366377716e-05,
+      "loss": 0.7395,
+      "step": 4083
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6163624647888404,
+      "learning_rate": 1.9771011335848453e-05,
+      "loss": 0.8249,
+      "step": 4084
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4201862732518116,
+      "learning_rate": 1.9770848951029825e-05,
+      "loss": 0.7893,
+      "step": 4085
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2133511257681993,
+      "learning_rate": 1.977068650932222e-05,
+      "loss": 0.7741,
+      "step": 4086
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.601509490822238,
+      "learning_rate": 1.9770524010726587e-05,
+      "loss": 0.7467,
+      "step": 4087
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7963312795480004,
+      "learning_rate": 1.977036145524387e-05,
+      "loss": 0.7152,
+      "step": 4088
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5442190583933264,
+      "learning_rate": 1.9770198842875015e-05,
+      "loss": 0.5116,
+      "step": 4089
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.460490626322781,
+      "learning_rate": 1.9770036173620965e-05,
+      "loss": 0.7882,
+      "step": 4090
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.486558827022148,
+      "learning_rate": 1.9769873447482676e-05,
+      "loss": 0.7626,
+      "step": 4091
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5941823321145854,
+      "learning_rate": 1.976971066446109e-05,
+      "loss": 0.511,
+      "step": 4092
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7821838978582534,
+      "learning_rate": 1.9769547824557155e-05,
+      "loss": 0.8543,
+      "step": 4093
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4621772365187948,
+      "learning_rate": 1.976938492777182e-05,
+      "loss": 0.8026,
+      "step": 4094
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4863525379730294,
+      "learning_rate": 1.976922197410603e-05,
+      "loss": 0.8483,
+      "step": 4095
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4282034705220374,
+      "learning_rate": 1.976905896356074e-05,
+      "loss": 0.8244,
+      "step": 4096
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6867028628378695,
+      "learning_rate": 1.9768895896136893e-05,
+      "loss": 0.8853,
+      "step": 4097
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7367222640019522,
+      "learning_rate": 1.9768732771835446e-05,
+      "loss": 0.7492,
+      "step": 4098
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3787775181930897,
+      "learning_rate": 1.976856959065734e-05,
+      "loss": 0.7923,
+      "step": 4099
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6807171070693563,
+      "learning_rate": 1.976840635260353e-05,
+      "loss": 0.8101,
+      "step": 4100
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.451529096701154,
+      "learning_rate": 1.9768243057674967e-05,
+      "loss": 0.7175,
+      "step": 4101
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6013927608278493,
+      "learning_rate": 1.97680797058726e-05,
+      "loss": 0.6723,
+      "step": 4102
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.585998608698233,
+      "learning_rate": 1.9767916297197378e-05,
+      "loss": 0.5349,
+      "step": 4103
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.987715145441899,
+      "learning_rate": 1.976775283165026e-05,
+      "loss": 0.7358,
+      "step": 4104
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.377065414166651,
+      "learning_rate": 1.976758930923219e-05,
+      "loss": 0.741,
+      "step": 4105
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4461135156744105,
+      "learning_rate": 1.976742572994412e-05,
+      "loss": 0.8016,
+      "step": 4106
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5742808321469455,
+      "learning_rate": 1.9767262093787007e-05,
+      "loss": 0.7889,
+      "step": 4107
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4740038450332034,
+      "learning_rate": 1.97670984007618e-05,
+      "loss": 0.7976,
+      "step": 4108
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6677612704618812,
+      "learning_rate": 1.976693465086946e-05,
+      "loss": 0.7447,
+      "step": 4109
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5790304546496454,
+      "learning_rate": 1.976677084411093e-05,
+      "loss": 0.7643,
+      "step": 4110
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6356057381120674,
+      "learning_rate": 1.9766606980487167e-05,
+      "loss": 0.7739,
+      "step": 4111
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.24849045691058,
+      "learning_rate": 1.9766443059999127e-05,
+      "loss": 0.7697,
+      "step": 4112
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.610308473051681,
+      "learning_rate": 1.9766279082647763e-05,
+      "loss": 0.7844,
+      "step": 4113
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.673456618097301,
+      "learning_rate": 1.976611504843403e-05,
+      "loss": 0.8636,
+      "step": 4114
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5644047110757868,
+      "learning_rate": 1.9765950957358887e-05,
+      "loss": 0.7207,
+      "step": 4115
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5321749290305047,
+      "learning_rate": 1.9765786809423283e-05,
+      "loss": 0.8516,
+      "step": 4116
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0820378294673283,
+      "learning_rate": 1.9765622604628177e-05,
+      "loss": 0.7499,
+      "step": 4117
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7212924826142306,
+      "learning_rate": 1.976545834297452e-05,
+      "loss": 0.7869,
+      "step": 4118
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.603690670593871,
+      "learning_rate": 1.9765294024463276e-05,
+      "loss": 0.7998,
+      "step": 4119
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3827495375615388,
+      "learning_rate": 1.97651296490954e-05,
+      "loss": 0.7829,
+      "step": 4120
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7848008122933359,
+      "learning_rate": 1.9764965216871848e-05,
+      "loss": 0.903,
+      "step": 4121
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5222512019505388,
+      "learning_rate": 1.9764800727793576e-05,
+      "loss": 0.7129,
+      "step": 4122
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9846448112534072,
+      "learning_rate": 1.9764636181861544e-05,
+      "loss": 0.8022,
+      "step": 4123
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.0912790642854495,
+      "learning_rate": 1.9764471579076705e-05,
+      "loss": 0.7035,
+      "step": 4124
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4761212798201044,
+      "learning_rate": 1.9764306919440023e-05,
+      "loss": 0.8354,
+      "step": 4125
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5708608260493323,
+      "learning_rate": 1.9764142202952456e-05,
+      "loss": 0.5437,
+      "step": 4126
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.115254267337495,
+      "learning_rate": 1.976397742961496e-05,
+      "loss": 0.8454,
+      "step": 4127
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6805354345294434,
+      "learning_rate": 1.9763812599428498e-05,
+      "loss": 0.7856,
+      "step": 4128
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8707938519030312,
+      "learning_rate": 1.9763647712394028e-05,
+      "loss": 0.6293,
+      "step": 4129
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4194996478252861,
+      "learning_rate": 1.9763482768512507e-05,
+      "loss": 0.799,
+      "step": 4130
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8184907912583737,
+      "learning_rate": 1.97633177677849e-05,
+      "loss": 0.6233,
+      "step": 4131
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4982375120079412,
+      "learning_rate": 1.976315271021217e-05,
+      "loss": 0.7245,
+      "step": 4132
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8404627748421514,
+      "learning_rate": 1.976298759579527e-05,
+      "loss": 0.7124,
+      "step": 4133
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.592497530699982,
+      "learning_rate": 1.9762822424535163e-05,
+      "loss": 0.6931,
+      "step": 4134
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5824848021518443,
+      "learning_rate": 1.9762657196432814e-05,
+      "loss": 0.6977,
+      "step": 4135
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9779131771749148,
+      "learning_rate": 1.9762491911489184e-05,
+      "loss": 0.6994,
+      "step": 4136
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6006551090803258,
+      "learning_rate": 1.976232656970524e-05,
+      "loss": 0.5672,
+      "step": 4137
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9789725370827493,
+      "learning_rate": 1.9762161171081934e-05,
+      "loss": 0.6532,
+      "step": 4138
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.031697634238476,
+      "learning_rate": 1.976199571562024e-05,
+      "loss": 0.656,
+      "step": 4139
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5860639234218094,
+      "learning_rate": 1.9761830203321114e-05,
+      "loss": 0.518,
+      "step": 4140
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7891356855644287,
+      "learning_rate": 1.9761664634185523e-05,
+      "loss": 0.7998,
+      "step": 4141
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5424742627977521,
+      "learning_rate": 1.976149900821443e-05,
+      "loss": 0.9624,
+      "step": 4142
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8883301259831844,
+      "learning_rate": 1.9761333325408795e-05,
+      "loss": 0.7997,
+      "step": 4143
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7357546704205902,
+      "learning_rate": 1.9761167585769588e-05,
+      "loss": 0.7322,
+      "step": 4144
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5488882709318281,
+      "learning_rate": 1.9761001789297778e-05,
+      "loss": 0.8583,
+      "step": 4145
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9845847226806552,
+      "learning_rate": 1.9760835935994322e-05,
+      "loss": 0.7263,
+      "step": 4146
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.682271813284771,
+      "learning_rate": 1.976067002586019e-05,
+      "loss": 0.6914,
+      "step": 4147
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8253343759765677,
+      "learning_rate": 1.9760504058896348e-05,
+      "loss": 0.8175,
+      "step": 4148
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8816446459093257,
+      "learning_rate": 1.9760338035103758e-05,
+      "loss": 0.7103,
+      "step": 4149
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4859298956869815,
+      "learning_rate": 1.976017195448339e-05,
+      "loss": 0.7288,
+      "step": 4150
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7912785512743847,
+      "learning_rate": 1.9760005817036215e-05,
+      "loss": 0.7724,
+      "step": 4151
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5725337000992994,
+      "learning_rate": 1.9759839622763194e-05,
+      "loss": 0.7537,
+      "step": 4152
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4931945450625732,
+      "learning_rate": 1.9759673371665297e-05,
+      "loss": 0.7559,
+      "step": 4153
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.856209974714612,
+      "learning_rate": 1.975950706374349e-05,
+      "loss": 0.8175,
+      "step": 4154
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6572497078282282,
+      "learning_rate": 1.9759340698998746e-05,
+      "loss": 0.855,
+      "step": 4155
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6182989153483416,
+      "learning_rate": 1.9759174277432026e-05,
+      "loss": 0.7883,
+      "step": 4156
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5843884336679406,
+      "learning_rate": 1.9759007799044307e-05,
+      "loss": 0.7056,
+      "step": 4157
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4560892197970126,
+      "learning_rate": 1.9758841263836553e-05,
+      "loss": 0.7336,
+      "step": 4158
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.714936040752044,
+      "learning_rate": 1.9758674671809742e-05,
+      "loss": 0.837,
+      "step": 4159
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.459780691037746,
+      "learning_rate": 1.975850802296483e-05,
+      "loss": 0.6162,
+      "step": 4160
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5739233476998723,
+      "learning_rate": 1.97583413173028e-05,
+      "loss": 0.5044,
+      "step": 4161
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5756384672159385,
+      "learning_rate": 1.9758174554824614e-05,
+      "loss": 0.6967,
+      "step": 4162
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5359018743086588,
+      "learning_rate": 1.9758007735531247e-05,
+      "loss": 0.7718,
+      "step": 4163
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.549776865937273,
+      "learning_rate": 1.975784085942367e-05,
+      "loss": 0.8277,
+      "step": 4164
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5530577885663857,
+      "learning_rate": 1.9757673926502854e-05,
+      "loss": 0.6853,
+      "step": 4165
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.005401393630069,
+      "learning_rate": 1.975750693676977e-05,
+      "loss": 0.7355,
+      "step": 4166
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7707724893987973,
+      "learning_rate": 1.9757339890225395e-05,
+      "loss": 0.805,
+      "step": 4167
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5026846172553006,
+      "learning_rate": 1.9757172786870695e-05,
+      "loss": 0.7749,
+      "step": 4168
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8704898995530035,
+      "learning_rate": 1.9757005626706648e-05,
+      "loss": 0.8283,
+      "step": 4169
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8223792088353998,
+      "learning_rate": 1.9756838409734222e-05,
+      "loss": 0.8992,
+      "step": 4170
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8254894964493749,
+      "learning_rate": 1.9756671135954396e-05,
+      "loss": 0.7499,
+      "step": 4171
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5778995103171278,
+      "learning_rate": 1.975650380536814e-05,
+      "loss": 0.7538,
+      "step": 4172
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3324694499577772,
+      "learning_rate": 1.9756336417976435e-05,
+      "loss": 0.7448,
+      "step": 4173
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7317726111875353,
+      "learning_rate": 1.9756168973780245e-05,
+      "loss": 0.8236,
+      "step": 4174
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.108832089881786,
+      "learning_rate": 1.975600147278055e-05,
+      "loss": 0.7183,
+      "step": 4175
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5717168755152814,
+      "learning_rate": 1.975583391497833e-05,
+      "loss": 0.5077,
+      "step": 4176
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7179072526784944,
+      "learning_rate": 1.9755666300374554e-05,
+      "loss": 0.7718,
+      "step": 4177
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5820098168983863,
+      "learning_rate": 1.9755498628970202e-05,
+      "loss": 0.693,
+      "step": 4178
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8186636399943648,
+      "learning_rate": 1.9755330900766247e-05,
+      "loss": 0.8754,
+      "step": 4179
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.9707072905035363,
+      "learning_rate": 1.975516311576367e-05,
+      "loss": 0.8547,
+      "step": 4180
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6443369174184052,
+      "learning_rate": 1.975499527396344e-05,
+      "loss": 0.5462,
+      "step": 4181
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.427590373393903,
+      "learning_rate": 1.975482737536654e-05,
+      "loss": 0.7452,
+      "step": 4182
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5890950364287014,
+      "learning_rate": 1.975465941997395e-05,
+      "loss": 0.5007,
+      "step": 4183
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7511644504522765,
+      "learning_rate": 1.9754491407786642e-05,
+      "loss": 0.7949,
+      "step": 4184
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4543564902649664,
+      "learning_rate": 1.9754323338805597e-05,
+      "loss": 0.703,
+      "step": 4185
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5968980836863473,
+      "learning_rate": 1.9754155213031793e-05,
+      "loss": 0.7208,
+      "step": 4186
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6721041764938878,
+      "learning_rate": 1.9753987030466212e-05,
+      "loss": 0.8087,
+      "step": 4187
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.693051687016159,
+      "learning_rate": 1.975381879110983e-05,
+      "loss": 0.685,
+      "step": 4188
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8496525527426924,
+      "learning_rate": 1.975365049496362e-05,
+      "loss": 0.7119,
+      "step": 4189
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.740545103341032,
+      "learning_rate": 1.9753482142028578e-05,
+      "loss": 0.8005,
+      "step": 4190
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6821970736109946,
+      "learning_rate": 1.975331373230567e-05,
+      "loss": 0.8728,
+      "step": 4191
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6596348398242083,
+      "learning_rate": 1.9753145265795885e-05,
+      "loss": 0.5385,
+      "step": 4192
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5554576661823083,
+      "learning_rate": 1.9752976742500198e-05,
+      "loss": 0.7007,
+      "step": 4193
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.600336587651264,
+      "learning_rate": 1.975280816241959e-05,
+      "loss": 0.543,
+      "step": 4194
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.568875702934291,
+      "learning_rate": 1.9752639525555048e-05,
+      "loss": 0.7347,
+      "step": 4195
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7445164965928712,
+      "learning_rate": 1.9752470831907553e-05,
+      "loss": 0.7861,
+      "step": 4196
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5935342867874914,
+      "learning_rate": 1.9752302081478084e-05,
+      "loss": 0.5281,
+      "step": 4197
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5689969635294516,
+      "learning_rate": 1.9752133274267626e-05,
+      "loss": 0.7915,
+      "step": 4198
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6558443218896215,
+      "learning_rate": 1.9751964410277157e-05,
+      "loss": 0.8526,
+      "step": 4199
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5701929024645997,
+      "learning_rate": 1.9751795489507665e-05,
+      "loss": 0.564,
+      "step": 4200
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5810081908951403,
+      "learning_rate": 1.9751626511960134e-05,
+      "loss": 0.7552,
+      "step": 4201
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5379683050044985,
+      "learning_rate": 1.9751457477635544e-05,
+      "loss": 0.7612,
+      "step": 4202
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7447220548797382,
+      "learning_rate": 1.9751288386534885e-05,
+      "loss": 0.8302,
+      "step": 4203
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4083330311548536,
+      "learning_rate": 1.9751119238659132e-05,
+      "loss": 0.7915,
+      "step": 4204
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4411832243736495,
+      "learning_rate": 1.9750950034009282e-05,
+      "loss": 0.7614,
+      "step": 4205
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6009989160929591,
+      "learning_rate": 1.975078077258631e-05,
+      "loss": 0.5351,
+      "step": 4206
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6272241644199252,
+      "learning_rate": 1.9750611454391208e-05,
+      "loss": 0.9236,
+      "step": 4207
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4416148019026884,
+      "learning_rate": 1.9750442079424953e-05,
+      "loss": 0.6162,
+      "step": 4208
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.693280736266034,
+      "learning_rate": 1.9750272647688544e-05,
+      "loss": 0.8445,
+      "step": 4209
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6301323030816641,
+      "learning_rate": 1.9750103159182957e-05,
+      "loss": 0.8631,
+      "step": 4210
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5454427450801982,
+      "learning_rate": 1.974993361390918e-05,
+      "loss": 0.7324,
+      "step": 4211
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3698163639052665,
+      "learning_rate": 1.9749764011868212e-05,
+      "loss": 0.8457,
+      "step": 4212
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4043232749671188,
+      "learning_rate": 1.9749594353061024e-05,
+      "loss": 0.7075,
+      "step": 4213
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8437628130046417,
+      "learning_rate": 1.974942463748861e-05,
+      "loss": 0.6749,
+      "step": 4214
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6549222196677267,
+      "learning_rate": 1.9749254865151966e-05,
+      "loss": 0.7581,
+      "step": 4215
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5377498225585673,
+      "learning_rate": 1.9749085036052066e-05,
+      "loss": 0.7105,
+      "step": 4216
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6256885968416794,
+      "learning_rate": 1.974891515018991e-05,
+      "loss": 0.7606,
+      "step": 4217
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6565065768201708,
+      "learning_rate": 1.9748745207566483e-05,
+      "loss": 0.6155,
+      "step": 4218
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8006396451865672,
+      "learning_rate": 1.9748575208182776e-05,
+      "loss": 0.8652,
+      "step": 4219
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4758511826723455,
+      "learning_rate": 1.9748405152039777e-05,
+      "loss": 0.7324,
+      "step": 4220
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6593351383141384,
+      "learning_rate": 1.974823503913848e-05,
+      "loss": 0.9298,
+      "step": 4221
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6223978362490658,
+      "learning_rate": 1.9748064869479866e-05,
+      "loss": 0.5768,
+      "step": 4222
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6111212786210667,
+      "learning_rate": 1.9747894643064938e-05,
+      "loss": 0.7761,
+      "step": 4223
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4889982258360983,
+      "learning_rate": 1.9747724359894677e-05,
+      "loss": 0.7977,
+      "step": 4224
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9252850850710226,
+      "learning_rate": 1.974755401997008e-05,
+      "loss": 0.7892,
+      "step": 4225
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7113101618834516,
+      "learning_rate": 1.974738362329214e-05,
+      "loss": 0.6658,
+      "step": 4226
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.447511144412779,
+      "learning_rate": 1.9747213169861844e-05,
+      "loss": 0.7353,
+      "step": 4227
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.887411777111182,
+      "learning_rate": 1.974704265968019e-05,
+      "loss": 0.8987,
+      "step": 4228
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.588584520259566,
+      "learning_rate": 1.9746872092748162e-05,
+      "loss": 0.5837,
+      "step": 4229
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.6416998138066057,
+      "learning_rate": 1.9746701469066766e-05,
+      "loss": 0.7509,
+      "step": 4230
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8471183835429674,
+      "learning_rate": 1.974653078863698e-05,
+      "loss": 0.8082,
+      "step": 4231
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.512974138783335,
+      "learning_rate": 1.974636005145981e-05,
+      "loss": 0.7842,
+      "step": 4232
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8012457541555877,
+      "learning_rate": 1.9746189257536252e-05,
+      "loss": 0.815,
+      "step": 4233
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.611469874035382,
+      "learning_rate": 1.9746018406867285e-05,
+      "loss": 0.7055,
+      "step": 4234
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8698578006544562,
+      "learning_rate": 1.9745847499453918e-05,
+      "loss": 0.7861,
+      "step": 4235
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.424289214727475,
+      "learning_rate": 1.974567653529714e-05,
+      "loss": 0.651,
+      "step": 4236
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5276698991660302,
+      "learning_rate": 1.9745505514397952e-05,
+      "loss": 0.7776,
+      "step": 4237
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8061495983892357,
+      "learning_rate": 1.9745334436757342e-05,
+      "loss": 0.8744,
+      "step": 4238
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6039413847125805,
+      "learning_rate": 1.9745163302376306e-05,
+      "loss": 0.7383,
+      "step": 4239
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5976422311542702,
+      "learning_rate": 1.974499211125585e-05,
+      "loss": 0.4792,
+      "step": 4240
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6642717216683918,
+      "learning_rate": 1.9744820863396962e-05,
+      "loss": 0.7429,
+      "step": 4241
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5854099579308036,
+      "learning_rate": 1.9744649558800644e-05,
+      "loss": 0.8127,
+      "step": 4242
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.9455425620187214,
+      "learning_rate": 1.974447819746789e-05,
+      "loss": 0.7935,
+      "step": 4243
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.685175711287032,
+      "learning_rate": 1.9744306779399697e-05,
+      "loss": 0.6886,
+      "step": 4244
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8225267937101715,
+      "learning_rate": 1.9744135304597068e-05,
+      "loss": 0.7085,
+      "step": 4245
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5604906526427103,
+      "learning_rate": 1.9743963773060996e-05,
+      "loss": 0.7929,
+      "step": 4246
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5832357928584408,
+      "learning_rate": 1.974379218479248e-05,
+      "loss": 0.8063,
+      "step": 4247
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.1882827386038,
+      "learning_rate": 1.9743620539792525e-05,
+      "loss": 0.7509,
+      "step": 4248
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4285499370264745,
+      "learning_rate": 1.9743448838062126e-05,
+      "loss": 0.7377,
+      "step": 4249
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.332782488466668,
+      "learning_rate": 1.974327707960228e-05,
+      "loss": 0.7018,
+      "step": 4250
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6967302857363984,
+      "learning_rate": 1.9743105264413994e-05,
+      "loss": 0.6606,
+      "step": 4251
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3737924457852595,
+      "learning_rate": 1.9742933392498263e-05,
+      "loss": 0.7316,
+      "step": 4252
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8062069608323414,
+      "learning_rate": 1.974276146385609e-05,
+      "loss": 0.7114,
+      "step": 4253
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7319440726500805,
+      "learning_rate": 1.9742589478488474e-05,
+      "loss": 0.7704,
+      "step": 4254
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8259147533473534,
+      "learning_rate": 1.974241743639642e-05,
+      "loss": 0.8118,
+      "step": 4255
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6770962413431993,
+      "learning_rate": 1.9742245337580924e-05,
+      "loss": 0.7571,
+      "step": 4256
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.892420170405704,
+      "learning_rate": 1.974207318204299e-05,
+      "loss": 0.72,
+      "step": 4257
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6112185598731759,
+      "learning_rate": 1.9741900969783625e-05,
+      "loss": 0.6854,
+      "step": 4258
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5866407575126836,
+      "learning_rate": 1.974172870080383e-05,
+      "loss": 0.5261,
+      "step": 4259
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4474710075420347,
+      "learning_rate": 1.97415563751046e-05,
+      "loss": 0.803,
+      "step": 4260
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9811795401258956,
+      "learning_rate": 1.974138399268695e-05,
+      "loss": 0.7471,
+      "step": 4261
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5903737005623463,
+      "learning_rate": 1.9741211553551877e-05,
+      "loss": 0.5441,
+      "step": 4262
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5429852079618314,
+      "learning_rate": 1.9741039057700385e-05,
+      "loss": 0.7527,
+      "step": 4263
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8295773544583895,
+      "learning_rate": 1.974086650513348e-05,
+      "loss": 0.7595,
+      "step": 4264
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4097831403073167,
+      "learning_rate": 1.974069389585217e-05,
+      "loss": 0.8047,
+      "step": 4265
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6088011469790338,
+      "learning_rate": 1.9740521229857453e-05,
+      "loss": 0.7761,
+      "step": 4266
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.664222599328983,
+      "learning_rate": 1.9740348507150338e-05,
+      "loss": 0.715,
+      "step": 4267
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6128429397928791,
+      "learning_rate": 1.9740175727731832e-05,
+      "loss": 0.5285,
+      "step": 4268
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.07645298536239,
+      "learning_rate": 1.9740002891602936e-05,
+      "loss": 0.807,
+      "step": 4269
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4537245283017022,
+      "learning_rate": 1.9739829998764662e-05,
+      "loss": 0.8833,
+      "step": 4270
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6050914854442985,
+      "learning_rate": 1.9739657049218014e-05,
+      "loss": 0.5336,
+      "step": 4271
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.7958835689850177,
+      "learning_rate": 1.9739484042963997e-05,
+      "loss": 0.9013,
+      "step": 4272
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7493717593081004,
+      "learning_rate": 1.973931098000362e-05,
+      "loss": 0.7497,
+      "step": 4273
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8078274079102132,
+      "learning_rate": 1.9739137860337893e-05,
+      "loss": 0.6795,
+      "step": 4274
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6880442521525199,
+      "learning_rate": 1.9738964683967823e-05,
+      "loss": 0.7156,
+      "step": 4275
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6950261658352797,
+      "learning_rate": 1.9738791450894415e-05,
+      "loss": 0.8362,
+      "step": 4276
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5753527357160994,
+      "learning_rate": 1.9738618161118684e-05,
+      "loss": 0.5742,
+      "step": 4277
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6566150791735526,
+      "learning_rate": 1.973844481464163e-05,
+      "loss": 0.6949,
+      "step": 4278
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6338774216184169,
+      "learning_rate": 1.973827141146427e-05,
+      "loss": 0.6791,
+      "step": 4279
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4630078460973457,
+      "learning_rate": 1.973809795158761e-05,
+      "loss": 0.707,
+      "step": 4280
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7471094402705498,
+      "learning_rate": 1.973792443501266e-05,
+      "loss": 0.738,
+      "step": 4281
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5013800409257052,
+      "learning_rate": 1.9737750861740434e-05,
+      "loss": 0.6819,
+      "step": 4282
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6688094714472527,
+      "learning_rate": 1.9737577231771934e-05,
+      "loss": 0.5474,
+      "step": 4283
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5680681925928621,
+      "learning_rate": 1.973740354510818e-05,
+      "loss": 0.5179,
+      "step": 4284
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5264129387466614,
+      "learning_rate": 1.973722980175018e-05,
+      "loss": 0.8674,
+      "step": 4285
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4015080333446073,
+      "learning_rate": 1.9737056001698946e-05,
+      "loss": 0.904,
+      "step": 4286
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0086495594032048,
+      "learning_rate": 1.9736882144955487e-05,
+      "loss": 0.6764,
+      "step": 4287
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5444071649412126,
+      "learning_rate": 1.973670823152082e-05,
+      "loss": 0.6945,
+      "step": 4288
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6146688167733764,
+      "learning_rate": 1.9736534261395953e-05,
+      "loss": 0.5798,
+      "step": 4289
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6853295624339868,
+      "learning_rate": 1.9736360234581902e-05,
+      "loss": 0.6388,
+      "step": 4290
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5929331700299094,
+      "learning_rate": 1.9736186151079677e-05,
+      "loss": 0.7513,
+      "step": 4291
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2161436107573422,
+      "learning_rate": 1.9736012010890294e-05,
+      "loss": 0.8634,
+      "step": 4292
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4092973656583851,
+      "learning_rate": 1.9735837814014772e-05,
+      "loss": 0.7449,
+      "step": 4293
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.459216422942279,
+      "learning_rate": 1.9735663560454115e-05,
+      "loss": 0.7788,
+      "step": 4294
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.742248807301703,
+      "learning_rate": 1.9735489250209342e-05,
+      "loss": 0.787,
+      "step": 4295
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2680951463051466,
+      "learning_rate": 1.973531488328147e-05,
+      "loss": 0.7314,
+      "step": 4296
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6289953608481982,
+      "learning_rate": 1.9735140459671512e-05,
+      "loss": 0.8141,
+      "step": 4297
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.53102674316602,
+      "learning_rate": 1.9734965979380487e-05,
+      "loss": 0.7519,
+      "step": 4298
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5079404754020467,
+      "learning_rate": 1.9734791442409403e-05,
+      "loss": 0.7695,
+      "step": 4299
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.369147600521886,
+      "learning_rate": 1.9734616848759288e-05,
+      "loss": 0.6615,
+      "step": 4300
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3533742218390648,
+      "learning_rate": 1.9734442198431146e-05,
+      "loss": 0.7806,
+      "step": 4301
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6126322444952133,
+      "learning_rate": 1.9734267491426003e-05,
+      "loss": 0.78,
+      "step": 4302
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5896711726287664,
+      "learning_rate": 1.973409272774487e-05,
+      "loss": 1.0447,
+      "step": 4303
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7524913863652873,
+      "learning_rate": 1.9733917907388772e-05,
+      "loss": 0.9362,
+      "step": 4304
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5904872459902457,
+      "learning_rate": 1.9733743030358716e-05,
+      "loss": 0.5314,
+      "step": 4305
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6394487831899767,
+      "learning_rate": 1.973356809665573e-05,
+      "loss": 0.6242,
+      "step": 4306
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6050238122108997,
+      "learning_rate": 1.9733393106280826e-05,
+      "loss": 0.6059,
+      "step": 4307
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4249411791632378,
+      "learning_rate": 1.9733218059235032e-05,
+      "loss": 0.7098,
+      "step": 4308
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5621535616010351,
+      "learning_rate": 1.9733042955519356e-05,
+      "loss": 0.7161,
+      "step": 4309
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.029318443124293,
+      "learning_rate": 1.9732867795134824e-05,
+      "loss": 0.7575,
+      "step": 4310
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.215389706933073,
+      "learning_rate": 1.9732692578082453e-05,
+      "loss": 0.7737,
+      "step": 4311
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.534339452865174,
+      "learning_rate": 1.9732517304363266e-05,
+      "loss": 0.8376,
+      "step": 4312
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4925998855179412,
+      "learning_rate": 1.9732341973978283e-05,
+      "loss": 0.7954,
+      "step": 4313
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.586333115142988,
+      "learning_rate": 1.973216658692852e-05,
+      "loss": 0.8368,
+      "step": 4314
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3524146674590967,
+      "learning_rate": 1.9731991143215005e-05,
+      "loss": 0.8388,
+      "step": 4315
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7506424638449476,
+      "learning_rate": 1.9731815642838756e-05,
+      "loss": 0.7696,
+      "step": 4316
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5412702622381749,
+      "learning_rate": 1.9731640085800794e-05,
+      "loss": 0.8909,
+      "step": 4317
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8196894061857376,
+      "learning_rate": 1.9731464472102145e-05,
+      "loss": 0.8456,
+      "step": 4318
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6380963776278374,
+      "learning_rate": 1.9731288801743826e-05,
+      "loss": 0.7134,
+      "step": 4319
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.547286595312828,
+      "learning_rate": 1.9731113074726863e-05,
+      "loss": 0.8446,
+      "step": 4320
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8847763134104807,
+      "learning_rate": 1.973093729105228e-05,
+      "loss": 0.6993,
+      "step": 4321
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.724270260612745,
+      "learning_rate": 1.9730761450721103e-05,
+      "loss": 0.5409,
+      "step": 4322
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5184537863538015,
+      "learning_rate": 1.9730585553734345e-05,
+      "loss": 0.7173,
+      "step": 4323
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.373981330769667,
+      "learning_rate": 1.9730409600093042e-05,
+      "loss": 0.7799,
+      "step": 4324
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4235600321630155,
+      "learning_rate": 1.9730233589798213e-05,
+      "loss": 0.7524,
+      "step": 4325
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.489409207833165,
+      "learning_rate": 1.9730057522850882e-05,
+      "loss": 0.7233,
+      "step": 4326
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6413001737671138,
+      "learning_rate": 1.9729881399252077e-05,
+      "loss": 0.7253,
+      "step": 4327
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9848267405757618,
+      "learning_rate": 1.972970521900282e-05,
+      "loss": 0.7814,
+      "step": 4328
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6153197277763987,
+      "learning_rate": 1.972952898210414e-05,
+      "loss": 0.6475,
+      "step": 4329
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7410659965390936,
+      "learning_rate": 1.9729352688557064e-05,
+      "loss": 0.6945,
+      "step": 4330
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4349822726010735,
+      "learning_rate": 1.9729176338362613e-05,
+      "loss": 0.8359,
+      "step": 4331
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5036288164203029,
+      "learning_rate": 1.9728999931521818e-05,
+      "loss": 0.8548,
+      "step": 4332
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7168969597495387,
+      "learning_rate": 1.9728823468035708e-05,
+      "loss": 0.8559,
+      "step": 4333
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.674194456354958,
+      "learning_rate": 1.9728646947905307e-05,
+      "loss": 0.8251,
+      "step": 4334
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4253627573011682,
+      "learning_rate": 1.972847037113164e-05,
+      "loss": 0.715,
+      "step": 4335
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4568039668335142,
+      "learning_rate": 1.9728293737715742e-05,
+      "loss": 0.7442,
+      "step": 4336
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4768775443876039,
+      "learning_rate": 1.9728117047658636e-05,
+      "loss": 0.5845,
+      "step": 4337
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7311293150889675,
+      "learning_rate": 1.9727940300961354e-05,
+      "loss": 0.7222,
+      "step": 4338
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6299668483471885,
+      "learning_rate": 1.9727763497624924e-05,
+      "loss": 0.7591,
+      "step": 4339
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4036361532966861,
+      "learning_rate": 1.9727586637650373e-05,
+      "loss": 0.7364,
+      "step": 4340
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6311511200017293,
+      "learning_rate": 1.9727409721038736e-05,
+      "loss": 0.5754,
+      "step": 4341
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2307662287801766,
+      "learning_rate": 1.9727232747791037e-05,
+      "loss": 0.5848,
+      "step": 4342
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4895898385237323,
+      "learning_rate": 1.972705571790831e-05,
+      "loss": 0.7579,
+      "step": 4343
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4684304049077124,
+      "learning_rate": 1.9726878631391585e-05,
+      "loss": 0.907,
+      "step": 4344
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5110119320660484,
+      "learning_rate": 1.9726701488241895e-05,
+      "loss": 0.7611,
+      "step": 4345
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3162640830346048,
+      "learning_rate": 1.9726524288460266e-05,
+      "loss": 0.7842,
+      "step": 4346
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8186011148661088,
+      "learning_rate": 1.9726347032047735e-05,
+      "loss": 0.7496,
+      "step": 4347
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.543737306261303,
+      "learning_rate": 1.9726169719005336e-05,
+      "loss": 1.0788,
+      "step": 4348
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3547294897395656,
+      "learning_rate": 1.9725992349334092e-05,
+      "loss": 0.6449,
+      "step": 4349
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.586015867480051,
+      "learning_rate": 1.9725814923035044e-05,
+      "loss": 0.7526,
+      "step": 4350
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5698495052931474,
+      "learning_rate": 1.9725637440109223e-05,
+      "loss": 0.7488,
+      "step": 4351
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.481556934072095,
+      "learning_rate": 1.972545990055766e-05,
+      "loss": 0.8237,
+      "step": 4352
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.007320923310732,
+      "learning_rate": 1.972528230438139e-05,
+      "loss": 0.7712,
+      "step": 4353
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9686695465433524,
+      "learning_rate": 1.972510465158145e-05,
+      "loss": 0.9039,
+      "step": 4354
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8356849858030486,
+      "learning_rate": 1.972492694215887e-05,
+      "loss": 0.6973,
+      "step": 4355
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8736108007531571,
+      "learning_rate": 1.972474917611469e-05,
+      "loss": 0.7643,
+      "step": 4356
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.212636777918496,
+      "learning_rate": 1.9724571353449933e-05,
+      "loss": 0.7661,
+      "step": 4357
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6171508699124397,
+      "learning_rate": 1.972439347416565e-05,
+      "loss": 0.7479,
+      "step": 4358
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6334356348436263,
+      "learning_rate": 1.972421553826287e-05,
+      "loss": 0.5253,
+      "step": 4359
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7262348097814042,
+      "learning_rate": 1.9724037545742624e-05,
+      "loss": 0.7218,
+      "step": 4360
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.620590828964206,
+      "learning_rate": 1.9723859496605958e-05,
+      "loss": 0.8947,
+      "step": 4361
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4621113514426067,
+      "learning_rate": 1.97236813908539e-05,
+      "loss": 0.6897,
+      "step": 4362
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5769549539003591,
+      "learning_rate": 1.972350322848749e-05,
+      "loss": 0.773,
+      "step": 4363
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5267439448060307,
+      "learning_rate": 1.972332500950777e-05,
+      "loss": 0.8271,
+      "step": 4364
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4625387540508579,
+      "learning_rate": 1.9723146733915766e-05,
+      "loss": 0.76,
+      "step": 4365
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3951847774998503,
+      "learning_rate": 1.9722968401712527e-05,
+      "loss": 0.9364,
+      "step": 4366
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5811105355578255,
+      "learning_rate": 1.972279001289909e-05,
+      "loss": 0.5271,
+      "step": 4367
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.642126620485138,
+      "learning_rate": 1.972261156747649e-05,
+      "loss": 0.7123,
+      "step": 4368
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6718697577162718,
+      "learning_rate": 1.9722433065445767e-05,
+      "loss": 0.7308,
+      "step": 4369
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.475076144330942,
+      "learning_rate": 1.972225450680796e-05,
+      "loss": 0.7179,
+      "step": 4370
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5845820407417998,
+      "learning_rate": 1.972207589156411e-05,
+      "loss": 0.5356,
+      "step": 4371
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6492416852235978,
+      "learning_rate": 1.9721897219715254e-05,
+      "loss": 0.8869,
+      "step": 4372
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4338407761952987,
+      "learning_rate": 1.972171849126244e-05,
+      "loss": 0.6867,
+      "step": 4373
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.667312691255316,
+      "learning_rate": 1.9721539706206697e-05,
+      "loss": 0.86,
+      "step": 4374
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7909476242424636,
+      "learning_rate": 1.9721360864549078e-05,
+      "loss": 0.7014,
+      "step": 4375
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3293839264498235,
+      "learning_rate": 1.9721181966290614e-05,
+      "loss": 0.7751,
+      "step": 4376
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.6295061026016002,
+      "learning_rate": 1.972100301143235e-05,
+      "loss": 0.7797,
+      "step": 4377
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5473984054039442,
+      "learning_rate": 1.9720823999975333e-05,
+      "loss": 0.7819,
+      "step": 4378
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4415682834085073,
+      "learning_rate": 1.97206449319206e-05,
+      "loss": 0.708,
+      "step": 4379
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4349769282970355,
+      "learning_rate": 1.9720465807269192e-05,
+      "loss": 0.7551,
+      "step": 4380
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6601172934810493,
+      "learning_rate": 1.972028662602216e-05,
+      "loss": 0.7632,
+      "step": 4381
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6022667338478687,
+      "learning_rate": 1.9720107388180536e-05,
+      "loss": 0.5363,
+      "step": 4382
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.366705502415216,
+      "learning_rate": 1.9719928093745372e-05,
+      "loss": 0.76,
+      "step": 4383
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.309664667264486,
+      "learning_rate": 1.971974874271771e-05,
+      "loss": 0.7407,
+      "step": 4384
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4757560647429073,
+      "learning_rate": 1.9719569335098592e-05,
+      "loss": 0.8651,
+      "step": 4385
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3749115456399787,
+      "learning_rate": 1.9719389870889064e-05,
+      "loss": 0.7754,
+      "step": 4386
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.579155767985412,
+      "learning_rate": 1.9719210350090172e-05,
+      "loss": 0.6689,
+      "step": 4387
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7432777403781794,
+      "learning_rate": 1.971903077270296e-05,
+      "loss": 0.6935,
+      "step": 4388
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8286408536255798,
+      "learning_rate": 1.9718851138728474e-05,
+      "loss": 0.7657,
+      "step": 4389
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8125441390543986,
+      "learning_rate": 1.971867144816776e-05,
+      "loss": 0.6668,
+      "step": 4390
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.005748060585432,
+      "learning_rate": 1.9718491701021865e-05,
+      "loss": 0.8279,
+      "step": 4391
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3611655701040006,
+      "learning_rate": 1.971831189729183e-05,
+      "loss": 0.7346,
+      "step": 4392
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3961341976459343,
+      "learning_rate": 1.9718132036978712e-05,
+      "loss": 0.782,
+      "step": 4393
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6078026254302182,
+      "learning_rate": 1.971795212008355e-05,
+      "loss": 0.5151,
+      "step": 4394
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.865744514624002,
+      "learning_rate": 1.9717772146607396e-05,
+      "loss": 0.8478,
+      "step": 4395
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7958272243219502,
+      "learning_rate": 1.971759211655129e-05,
+      "loss": 0.7337,
+      "step": 4396
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5012609845260096,
+      "learning_rate": 1.971741202991629e-05,
+      "loss": 0.8937,
+      "step": 4397
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6087845396657349,
+      "learning_rate": 1.9717231886703444e-05,
+      "loss": 0.5464,
+      "step": 4398
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8843333607186317,
+      "learning_rate": 1.9717051686913793e-05,
+      "loss": 0.7387,
+      "step": 4399
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5726429731807,
+      "learning_rate": 1.9716871430548394e-05,
+      "loss": 0.8929,
+      "step": 4400
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.331162170177622,
+      "learning_rate": 1.971669111760829e-05,
+      "loss": 0.7585,
+      "step": 4401
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4821917992470246,
+      "learning_rate": 1.9716510748094536e-05,
+      "loss": 0.8188,
+      "step": 4402
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.489205636445236,
+      "learning_rate": 1.971633032200818e-05,
+      "loss": 0.8127,
+      "step": 4403
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6987095486586863,
+      "learning_rate": 1.971614983935027e-05,
+      "loss": 0.8186,
+      "step": 4404
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5169137709559002,
+      "learning_rate": 1.9715969300121863e-05,
+      "loss": 0.6954,
+      "step": 4405
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7359997575740602,
+      "learning_rate": 1.971578870432401e-05,
+      "loss": 0.7581,
+      "step": 4406
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8270144479002395,
+      "learning_rate": 1.9715608051957753e-05,
+      "loss": 0.7128,
+      "step": 4407
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.962763134427347,
+      "learning_rate": 1.971542734302415e-05,
+      "loss": 0.8385,
+      "step": 4408
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4010784594609644,
+      "learning_rate": 1.9715246577524258e-05,
+      "loss": 0.758,
+      "step": 4409
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.035673988145331,
+      "learning_rate": 1.9715065755459122e-05,
+      "loss": 0.7917,
+      "step": 4410
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.556983349268564,
+      "learning_rate": 1.9714884876829797e-05,
+      "loss": 0.802,
+      "step": 4411
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6293213757573048,
+      "learning_rate": 1.9714703941637333e-05,
+      "loss": 0.7044,
+      "step": 4412
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.601956780350593,
+      "learning_rate": 1.9714522949882794e-05,
+      "loss": 0.8468,
+      "step": 4413
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.646109081659802,
+      "learning_rate": 1.9714341901567223e-05,
+      "loss": 0.84,
+      "step": 4414
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7194134475081513,
+      "learning_rate": 1.971416079669168e-05,
+      "loss": 0.5864,
+      "step": 4415
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3817535867432442,
+      "learning_rate": 1.9713979635257215e-05,
+      "loss": 0.7697,
+      "step": 4416
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6931198317304461,
+      "learning_rate": 1.9713798417264885e-05,
+      "loss": 0.9045,
+      "step": 4417
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5845224421530797,
+      "learning_rate": 1.9713617142715748e-05,
+      "loss": 0.8162,
+      "step": 4418
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.58830093705431,
+      "learning_rate": 1.9713435811610855e-05,
+      "loss": 0.8221,
+      "step": 4419
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5204962461498062,
+      "learning_rate": 1.9713254423951263e-05,
+      "loss": 0.7549,
+      "step": 4420
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6120742842243783,
+      "learning_rate": 1.971307297973803e-05,
+      "loss": 0.5092,
+      "step": 4421
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7640290574213697,
+      "learning_rate": 1.971289147897221e-05,
+      "loss": 0.7218,
+      "step": 4422
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.533869728758144,
+      "learning_rate": 1.971270992165486e-05,
+      "loss": 0.8063,
+      "step": 4423
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6312910494806896,
+      "learning_rate": 1.971252830778704e-05,
+      "loss": 0.7667,
+      "step": 4424
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5258385417176994,
+      "learning_rate": 1.9712346637369803e-05,
+      "loss": 0.6991,
+      "step": 4425
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8981574232603398,
+      "learning_rate": 1.971216491040421e-05,
+      "loss": 0.8386,
+      "step": 4426
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.615658055439134,
+      "learning_rate": 1.971198312689132e-05,
+      "loss": 0.7858,
+      "step": 4427
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8121114903237947,
+      "learning_rate": 1.9711801286832186e-05,
+      "loss": 0.8574,
+      "step": 4428
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.545904984254358,
+      "learning_rate": 1.971161939022787e-05,
+      "loss": 0.8022,
+      "step": 4429
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.747407842687874,
+      "learning_rate": 1.9711437437079436e-05,
+      "loss": 0.6914,
+      "step": 4430
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5263788489059031,
+      "learning_rate": 1.9711255427387935e-05,
+      "loss": 0.7861,
+      "step": 4431
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4790513792397282,
+      "learning_rate": 1.9711073361154433e-05,
+      "loss": 0.8502,
+      "step": 4432
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3840523731981167,
+      "learning_rate": 1.9710891238379984e-05,
+      "loss": 0.726,
+      "step": 4433
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.800402662542863,
+      "learning_rate": 1.9710709059065653e-05,
+      "loss": 0.701,
+      "step": 4434
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5419423665359655,
+      "learning_rate": 1.97105268232125e-05,
+      "loss": 0.7461,
+      "step": 4435
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5809934149019397,
+      "learning_rate": 1.9710344530821586e-05,
+      "loss": 0.517,
+      "step": 4436
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7038470423764622,
+      "learning_rate": 1.9710162181893972e-05,
+      "loss": 0.9267,
+      "step": 4437
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.583574169738219,
+      "learning_rate": 1.9709979776430717e-05,
+      "loss": 0.5366,
+      "step": 4438
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5934117931452592,
+      "learning_rate": 1.9709797314432885e-05,
+      "loss": 0.7829,
+      "step": 4439
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5544188604465827,
+      "learning_rate": 1.9709614795901543e-05,
+      "loss": 0.8057,
+      "step": 4440
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6458762108199705,
+      "learning_rate": 1.9709432220837748e-05,
+      "loss": 0.7753,
+      "step": 4441
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7269661011790243,
+      "learning_rate": 1.9709249589242564e-05,
+      "loss": 0.9239,
+      "step": 4442
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.193421582710676,
+      "learning_rate": 1.970906690111706e-05,
+      "loss": 0.6404,
+      "step": 4443
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4201341699110563,
+      "learning_rate": 1.9708884156462286e-05,
+      "loss": 0.9077,
+      "step": 4444
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.429984275249054,
+      "learning_rate": 1.970870135527932e-05,
+      "loss": 0.7871,
+      "step": 4445
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4836928945968102,
+      "learning_rate": 1.970851849756922e-05,
+      "loss": 0.8081,
+      "step": 4446
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5849531980952536,
+      "learning_rate": 1.9708335583333052e-05,
+      "loss": 0.8584,
+      "step": 4447
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4376025432791955,
+      "learning_rate": 1.970815261257188e-05,
+      "loss": 0.8412,
+      "step": 4448
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6011203305553009,
+      "learning_rate": 1.970796958528677e-05,
+      "loss": 0.5967,
+      "step": 4449
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.568603417158323,
+      "learning_rate": 1.970778650147879e-05,
+      "loss": 0.5357,
+      "step": 4450
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.1213101028532115,
+      "learning_rate": 1.9707603361149e-05,
+      "loss": 0.8215,
+      "step": 4451
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.596303838858434,
+      "learning_rate": 1.970742016429847e-05,
+      "loss": 0.7018,
+      "step": 4452
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5635427928488014,
+      "learning_rate": 1.970723691092827e-05,
+      "loss": 0.56,
+      "step": 4453
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.772600609513197,
+      "learning_rate": 1.970705360103946e-05,
+      "loss": 0.705,
+      "step": 4454
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6099375467999365,
+      "learning_rate": 1.9706870234633114e-05,
+      "loss": 0.5485,
+      "step": 4455
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.381010811033315,
+      "learning_rate": 1.970668681171029e-05,
+      "loss": 0.8373,
+      "step": 4456
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5640956323002275,
+      "learning_rate": 1.970650333227207e-05,
+      "loss": 0.7966,
+      "step": 4457
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1691986537549317,
+      "learning_rate": 1.970631979631951e-05,
+      "loss": 0.8016,
+      "step": 4458
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5075638046659168,
+      "learning_rate": 1.9706136203853684e-05,
+      "loss": 0.8673,
+      "step": 4459
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.407447420934284,
+      "learning_rate": 1.970595255487566e-05,
+      "loss": 0.6803,
+      "step": 4460
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5672979524533562,
+      "learning_rate": 1.970576884938651e-05,
+      "loss": 0.5193,
+      "step": 4461
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8540298486173143,
+      "learning_rate": 1.9705585087387297e-05,
+      "loss": 0.716,
+      "step": 4462
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5123379910719152,
+      "learning_rate": 1.9705401268879095e-05,
+      "loss": 0.7745,
+      "step": 4463
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9363544978832803,
+      "learning_rate": 1.9705217393862977e-05,
+      "loss": 0.8542,
+      "step": 4464
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.614887636024806,
+      "learning_rate": 1.970503346234001e-05,
+      "loss": 0.5598,
+      "step": 4465
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.588907030983631,
+      "learning_rate": 1.9704849474311266e-05,
+      "loss": 0.8424,
+      "step": 4466
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.388732205798032,
+      "learning_rate": 1.9704665429777815e-05,
+      "loss": 0.8511,
+      "step": 4467
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6859419445409802,
+      "learning_rate": 1.9704481328740726e-05,
+      "loss": 0.8114,
+      "step": 4468
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5896058644112016,
+      "learning_rate": 1.9704297171201082e-05,
+      "loss": 0.5269,
+      "step": 4469
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8565456822213313,
+      "learning_rate": 1.970411295715994e-05,
+      "loss": 0.8359,
+      "step": 4470
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7536612974969832,
+      "learning_rate": 1.9703928686618384e-05,
+      "loss": 0.8082,
+      "step": 4471
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6313573031913084,
+      "learning_rate": 1.9703744359577483e-05,
+      "loss": 0.8927,
+      "step": 4472
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.422645623356774,
+      "learning_rate": 1.9703559976038308e-05,
+      "loss": 0.6703,
+      "step": 4473
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.605211748436238,
+      "learning_rate": 1.9703375536001937e-05,
+      "loss": 0.5414,
+      "step": 4474
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5862507751804907,
+      "learning_rate": 1.970319103946944e-05,
+      "loss": 0.5549,
+      "step": 4475
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7746071729927406,
+      "learning_rate": 1.9703006486441895e-05,
+      "loss": 0.649,
+      "step": 4476
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5081604618773123,
+      "learning_rate": 1.9702821876920374e-05,
+      "loss": 0.7296,
+      "step": 4477
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.595928633379645,
+      "learning_rate": 1.970263721090595e-05,
+      "loss": 0.7348,
+      "step": 4478
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.6392302694919048,
+      "learning_rate": 1.97024524883997e-05,
+      "loss": 0.7781,
+      "step": 4479
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.401850745172558,
+      "learning_rate": 1.9702267709402703e-05,
+      "loss": 0.821,
+      "step": 4480
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5712911517266126,
+      "learning_rate": 1.970208287391603e-05,
+      "loss": 0.7469,
+      "step": 4481
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6455976236324852,
+      "learning_rate": 1.970189798194076e-05,
+      "loss": 0.5784,
+      "step": 4482
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6469388467983865,
+      "learning_rate": 1.9701713033477965e-05,
+      "loss": 0.8772,
+      "step": 4483
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6222543740378244,
+      "learning_rate": 1.9701528028528728e-05,
+      "loss": 0.7572,
+      "step": 4484
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4126801296423706,
+      "learning_rate": 1.970134296709412e-05,
+      "loss": 0.6919,
+      "step": 4485
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4966276524164879,
+      "learning_rate": 1.970115784917523e-05,
+      "loss": 0.8217,
+      "step": 4486
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4554966853011353,
+      "learning_rate": 1.970097267477312e-05,
+      "loss": 0.7502,
+      "step": 4487
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9959923734919653,
+      "learning_rate": 1.9700787443888877e-05,
+      "loss": 0.7851,
+      "step": 4488
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6054564801339537,
+      "learning_rate": 1.9700602156523582e-05,
+      "loss": 0.7454,
+      "step": 4489
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.4079077713963226,
+      "learning_rate": 1.9700416812678303e-05,
+      "loss": 0.8815,
+      "step": 4490
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9051145987525993,
+      "learning_rate": 1.9700231412354134e-05,
+      "loss": 0.735,
+      "step": 4491
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6101204892568903,
+      "learning_rate": 1.9700045955552143e-05,
+      "loss": 0.7772,
+      "step": 4492
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.015322855646714,
+      "learning_rate": 1.9699860442273417e-05,
+      "loss": 0.7514,
+      "step": 4493
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6124692556971617,
+      "learning_rate": 1.9699674872519026e-05,
+      "loss": 0.5411,
+      "step": 4494
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8251443112118624,
+      "learning_rate": 1.969948924629006e-05,
+      "loss": 0.7609,
+      "step": 4495
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8076461514865623,
+      "learning_rate": 1.9699303563587596e-05,
+      "loss": 0.8097,
+      "step": 4496
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7562797032873756,
+      "learning_rate": 1.9699117824412718e-05,
+      "loss": 0.789,
+      "step": 4497
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5659880928901002,
+      "learning_rate": 1.9698932028766506e-05,
+      "loss": 0.7639,
+      "step": 4498
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.612695447360466,
+      "learning_rate": 1.969874617665004e-05,
+      "loss": 0.8454,
+      "step": 4499
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6809308003102754,
+      "learning_rate": 1.96985602680644e-05,
+      "loss": 0.8136,
+      "step": 4500
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.4211098799993502,
+      "learning_rate": 1.969837430301068e-05,
+      "loss": 0.7676,
+      "step": 4501
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.025112122311153,
+      "learning_rate": 1.9698188281489944e-05,
+      "loss": 0.7178,
+      "step": 4502
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6176698979478163,
+      "learning_rate": 1.969800220350329e-05,
+      "loss": 0.8913,
+      "step": 4503
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8787638031302731,
+      "learning_rate": 1.96978160690518e-05,
+      "loss": 0.76,
+      "step": 4504
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6065000272276417,
+      "learning_rate": 1.969762987813655e-05,
+      "loss": 0.5643,
+      "step": 4505
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5987503033014858,
+      "learning_rate": 1.969744363075863e-05,
+      "loss": 0.7133,
+      "step": 4506
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.295507649222083,
+      "learning_rate": 1.9697257326919123e-05,
+      "loss": 0.7021,
+      "step": 4507
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5942409477315114,
+      "learning_rate": 1.9697070966619115e-05,
+      "loss": 0.557,
+      "step": 4508
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5739731246951554,
+      "learning_rate": 1.969688454985969e-05,
+      "loss": 0.552,
+      "step": 4509
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5343318466288203,
+      "learning_rate": 1.969669807664193e-05,
+      "loss": 0.8529,
+      "step": 4510
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7571266358415925,
+      "learning_rate": 1.969651154696693e-05,
+      "loss": 0.7557,
+      "step": 4511
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.644420427214968,
+      "learning_rate": 1.9696324960835763e-05,
+      "loss": 0.6874,
+      "step": 4512
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4436749695539144,
+      "learning_rate": 1.9696138318249526e-05,
+      "loss": 0.6984,
+      "step": 4513
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5525339380337597,
+      "learning_rate": 1.96959516192093e-05,
+      "loss": 0.7741,
+      "step": 4514
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6272980755269257,
+      "learning_rate": 1.9695764863716173e-05,
+      "loss": 0.7932,
+      "step": 4515
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.247771921126095,
+      "learning_rate": 1.9695578051771238e-05,
+      "loss": 0.7225,
+      "step": 4516
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.595359626787592,
+      "learning_rate": 1.9695391183375575e-05,
+      "loss": 0.5256,
+      "step": 4517
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8827305699482924,
+      "learning_rate": 1.9695204258530273e-05,
+      "loss": 0.8239,
+      "step": 4518
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4844920668369845,
+      "learning_rate": 1.9695017277236425e-05,
+      "loss": 0.7079,
+      "step": 4519
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5146520861184207,
+      "learning_rate": 1.969483023949512e-05,
+      "loss": 0.6679,
+      "step": 4520
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6459543920281807,
+      "learning_rate": 1.969464314530744e-05,
+      "loss": 0.7362,
+      "step": 4521
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4916136383263936,
+      "learning_rate": 1.9694455994674476e-05,
+      "loss": 0.7814,
+      "step": 4522
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7475983101600938,
+      "learning_rate": 1.9694268787597328e-05,
+      "loss": 0.8175,
+      "step": 4523
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.518536860556469,
+      "learning_rate": 1.9694081524077072e-05,
+      "loss": 0.7391,
+      "step": 4524
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7115038123330504,
+      "learning_rate": 1.9693894204114803e-05,
+      "loss": 0.8866,
+      "step": 4525
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1940948484688616,
+      "learning_rate": 1.9693706827711615e-05,
+      "loss": 0.7418,
+      "step": 4526
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.5925027546993205,
+      "learning_rate": 1.9693519394868595e-05,
+      "loss": 0.7112,
+      "step": 4527
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8057376048815545,
+      "learning_rate": 1.9693331905586842e-05,
+      "loss": 0.8103,
+      "step": 4528
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6060846581174104,
+      "learning_rate": 1.9693144359867436e-05,
+      "loss": 0.842,
+      "step": 4529
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.623299088201991,
+      "learning_rate": 1.969295675771148e-05,
+      "loss": 0.7138,
+      "step": 4530
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.32502525883839,
+      "learning_rate": 1.9692769099120054e-05,
+      "loss": 0.6437,
+      "step": 4531
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8043564999218928,
+      "learning_rate": 1.9692581384094262e-05,
+      "loss": 0.6738,
+      "step": 4532
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.731409625985917,
+      "learning_rate": 1.9692393612635192e-05,
+      "loss": 0.6632,
+      "step": 4533
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4192038627336265,
+      "learning_rate": 1.9692205784743938e-05,
+      "loss": 0.7207,
+      "step": 4534
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.538032506429081,
+      "learning_rate": 1.9692017900421594e-05,
+      "loss": 0.8356,
+      "step": 4535
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6606295030651914,
+      "learning_rate": 1.969182995966925e-05,
+      "loss": 0.7341,
+      "step": 4536
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5553662889697624,
+      "learning_rate": 1.969164196248801e-05,
+      "loss": 0.7165,
+      "step": 4537
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6050212626408272,
+      "learning_rate": 1.9691453908878957e-05,
+      "loss": 0.5156,
+      "step": 4538
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4804450370791977,
+      "learning_rate": 1.969126579884319e-05,
+      "loss": 0.7079,
+      "step": 4539
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8149550843018474,
+      "learning_rate": 1.969107763238181e-05,
+      "loss": 0.83,
+      "step": 4540
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6425756925652475,
+      "learning_rate": 1.9690889409495905e-05,
+      "loss": 0.7843,
+      "step": 4541
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5357170904867352,
+      "learning_rate": 1.9690701130186576e-05,
+      "loss": 0.7812,
+      "step": 4542
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.980839931377989,
+      "learning_rate": 1.9690512794454915e-05,
+      "loss": 0.7899,
+      "step": 4543
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6449443126616317,
+      "learning_rate": 1.969032440230202e-05,
+      "loss": 0.8572,
+      "step": 4544
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3148626135872878,
+      "learning_rate": 1.969013595372899e-05,
+      "loss": 0.7111,
+      "step": 4545
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3988190817252675,
+      "learning_rate": 1.9689947448736918e-05,
+      "loss": 0.8255,
+      "step": 4546
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.625128930800244,
+      "learning_rate": 1.9689758887326905e-05,
+      "loss": 0.7721,
+      "step": 4547
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5887346309487265,
+      "learning_rate": 1.9689570269500053e-05,
+      "loss": 0.7304,
+      "step": 4548
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.467434265213514,
+      "learning_rate": 1.968938159525745e-05,
+      "loss": 0.7376,
+      "step": 4549
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0815567645439685,
+      "learning_rate": 1.9689192864600203e-05,
+      "loss": 0.8571,
+      "step": 4550
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5876493454870284,
+      "learning_rate": 1.9689004077529405e-05,
+      "loss": 0.6364,
+      "step": 4551
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3584000488384131,
+      "learning_rate": 1.968881523404616e-05,
+      "loss": 0.7354,
+      "step": 4552
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.097918865028622,
+      "learning_rate": 1.968862633415156e-05,
+      "loss": 0.8623,
+      "step": 4553
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5000293939210734,
+      "learning_rate": 1.9688437377846714e-05,
+      "loss": 0.6989,
+      "step": 4554
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4727054641755442,
+      "learning_rate": 1.968824836513272e-05,
+      "loss": 0.6903,
+      "step": 4555
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4264216498934463,
+      "learning_rate": 1.9688059296010676e-05,
+      "loss": 0.6887,
+      "step": 4556
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5079212337419998,
+      "learning_rate": 1.968787017048168e-05,
+      "loss": 0.67,
+      "step": 4557
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.488846288080075,
+      "learning_rate": 1.968768098854684e-05,
+      "loss": 0.6667,
+      "step": 4558
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5431881698791319,
+      "learning_rate": 1.9687491750207255e-05,
+      "loss": 0.6874,
+      "step": 4559
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5336813996677632,
+      "learning_rate": 1.9687302455464025e-05,
+      "loss": 0.9494,
+      "step": 4560
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5912045715789033,
+      "learning_rate": 1.9687113104318252e-05,
+      "loss": 0.503,
+      "step": 4561
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5882463667051532,
+      "learning_rate": 1.968692369677104e-05,
+      "loss": 0.7269,
+      "step": 4562
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4021785909434688,
+      "learning_rate": 1.968673423282349e-05,
+      "loss": 0.7552,
+      "step": 4563
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3671374903626479,
+      "learning_rate": 1.968654471247671e-05,
+      "loss": 0.7712,
+      "step": 4564
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4784755160613743,
+      "learning_rate": 1.9686355135731797e-05,
+      "loss": 0.8478,
+      "step": 4565
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.509702636901783,
+      "learning_rate": 1.968616550258986e-05,
+      "loss": 0.7622,
+      "step": 4566
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9647293611946932,
+      "learning_rate": 1.9685975813052e-05,
+      "loss": 0.7908,
+      "step": 4567
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2493846893678018,
+      "learning_rate": 1.968578606711932e-05,
+      "loss": 0.823,
+      "step": 4568
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.844660604971322,
+      "learning_rate": 1.968559626479293e-05,
+      "loss": 0.7991,
+      "step": 4569
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5183878558689423,
+      "learning_rate": 1.9685406406073933e-05,
+      "loss": 0.714,
+      "step": 4570
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7593003150448434,
+      "learning_rate": 1.968521649096343e-05,
+      "loss": 0.7852,
+      "step": 4571
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3230887404773286,
+      "learning_rate": 1.9685026519462535e-05,
+      "loss": 0.6779,
+      "step": 4572
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5965056641901237,
+      "learning_rate": 1.9684836491572342e-05,
+      "loss": 0.5758,
+      "step": 4573
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3189781401739658,
+      "learning_rate": 1.9684646407293974e-05,
+      "loss": 0.7211,
+      "step": 4574
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5979524502777085,
+      "learning_rate": 1.9684456266628524e-05,
+      "loss": 0.5423,
+      "step": 4575
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3545207752425892,
+      "learning_rate": 1.96842660695771e-05,
+      "loss": 0.8019,
+      "step": 4576
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4947064227578346,
+      "learning_rate": 1.9684075816140817e-05,
+      "loss": 0.7675,
+      "step": 4577
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6407791807467962,
+      "learning_rate": 1.968388550632078e-05,
+      "loss": 0.753,
+      "step": 4578
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5064080604049426,
+      "learning_rate": 1.9683695140118093e-05,
+      "loss": 0.6465,
+      "step": 4579
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.400320310235187,
+      "learning_rate": 1.9683504717533868e-05,
+      "loss": 0.7243,
+      "step": 4580
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.336348397186953,
+      "learning_rate": 1.9683314238569215e-05,
+      "loss": 0.6766,
+      "step": 4581
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4815004796087274,
+      "learning_rate": 1.9683123703225236e-05,
+      "loss": 0.7577,
+      "step": 4582
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.812727319706796,
+      "learning_rate": 1.968293311150305e-05,
+      "loss": 0.8577,
+      "step": 4583
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.528503951461526,
+      "learning_rate": 1.9682742463403755e-05,
+      "loss": 0.6799,
+      "step": 4584
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5925522807737036,
+      "learning_rate": 1.9682551758928473e-05,
+      "loss": 0.7658,
+      "step": 4585
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7289896664717144,
+      "learning_rate": 1.968236099807831e-05,
+      "loss": 0.728,
+      "step": 4586
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6166554696179336,
+      "learning_rate": 1.9682170180854373e-05,
+      "loss": 0.7741,
+      "step": 4587
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5911717750730842,
+      "learning_rate": 1.9681979307257778e-05,
+      "loss": 0.5358,
+      "step": 4588
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5992432769205083,
+      "learning_rate": 1.9681788377289633e-05,
+      "loss": 0.8477,
+      "step": 4589
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8677402107261156,
+      "learning_rate": 1.9681597390951052e-05,
+      "loss": 0.7855,
+      "step": 4590
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6810969590864004,
+      "learning_rate": 1.9681406348243147e-05,
+      "loss": 0.6586,
+      "step": 4591
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3905749279425228,
+      "learning_rate": 1.968121524916703e-05,
+      "loss": 0.766,
+      "step": 4592
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4801967118379284,
+      "learning_rate": 1.9681024093723805e-05,
+      "loss": 0.74,
+      "step": 4593
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8805141674535335,
+      "learning_rate": 1.9680832881914598e-05,
+      "loss": 0.7427,
+      "step": 4594
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.465881111352119,
+      "learning_rate": 1.9680641613740515e-05,
+      "loss": 0.6769,
+      "step": 4595
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7271284938810103,
+      "learning_rate": 1.968045028920268e-05,
+      "loss": 0.8094,
+      "step": 4596
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2237616310937565,
+      "learning_rate": 1.968025890830219e-05,
+      "loss": 0.8066,
+      "step": 4597
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4546223306597534,
+      "learning_rate": 1.968006747104017e-05,
+      "loss": 0.7778,
+      "step": 4598
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3768772181892284,
+      "learning_rate": 1.9679875977417734e-05,
+      "loss": 0.7545,
+      "step": 4599
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7171142235095533,
+      "learning_rate": 1.9679684427435993e-05,
+      "loss": 0.6706,
+      "step": 4600
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.854526755196112,
+      "learning_rate": 1.9679492821096065e-05,
+      "loss": 0.8004,
+      "step": 4601
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5542097382694318,
+      "learning_rate": 1.9679301158399067e-05,
+      "loss": 0.8491,
+      "step": 4602
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5279721426872046,
+      "learning_rate": 1.9679109439346114e-05,
+      "loss": 0.8269,
+      "step": 4603
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.63530255644273,
+      "learning_rate": 1.967891766393832e-05,
+      "loss": 0.7509,
+      "step": 4604
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7496829803134597,
+      "learning_rate": 1.9678725832176803e-05,
+      "loss": 0.8015,
+      "step": 4605
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.191108032956583,
+      "learning_rate": 1.967853394406268e-05,
+      "loss": 0.7503,
+      "step": 4606
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.500477282125818,
+      "learning_rate": 1.9678341999597066e-05,
+      "loss": 0.5591,
+      "step": 4607
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5813331328527052,
+      "learning_rate": 1.9678149998781083e-05,
+      "loss": 0.5578,
+      "step": 4608
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9268420209508899,
+      "learning_rate": 1.9677957941615847e-05,
+      "loss": 0.7405,
+      "step": 4609
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.254277632731089,
+      "learning_rate": 1.9677765828102477e-05,
+      "loss": 0.8095,
+      "step": 4610
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7303304281598482,
+      "learning_rate": 1.967757365824209e-05,
+      "loss": 0.7332,
+      "step": 4611
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.477347553641397,
+      "learning_rate": 1.9677381432035803e-05,
+      "loss": 0.8212,
+      "step": 4612
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4659912849351258,
+      "learning_rate": 1.967718914948474e-05,
+      "loss": 0.8126,
+      "step": 4613
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.725895685040191,
+      "learning_rate": 1.9676996810590016e-05,
+      "loss": 0.7253,
+      "step": 4614
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5843495609338327,
+      "learning_rate": 1.9676804415352752e-05,
+      "loss": 0.73,
+      "step": 4615
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3123781185639665,
+      "learning_rate": 1.967661196377407e-05,
+      "loss": 0.7631,
+      "step": 4616
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.730771381288979,
+      "learning_rate": 1.9676419455855093e-05,
+      "loss": 0.7384,
+      "step": 4617
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5309876285903758,
+      "learning_rate": 1.9676226891596934e-05,
+      "loss": 0.6235,
+      "step": 4618
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.087361812567534,
+      "learning_rate": 1.967603427100072e-05,
+      "loss": 0.901,
+      "step": 4619
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5490310053594818,
+      "learning_rate": 1.967584159406757e-05,
+      "loss": 0.7177,
+      "step": 4620
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7920569259074162,
+      "learning_rate": 1.967564886079861e-05,
+      "loss": 0.8684,
+      "step": 4621
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5001619927385654,
+      "learning_rate": 1.9675456071194957e-05,
+      "loss": 0.8209,
+      "step": 4622
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5912536631438503,
+      "learning_rate": 1.9675263225257736e-05,
+      "loss": 0.7512,
+      "step": 4623
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4979548243770218,
+      "learning_rate": 1.967507032298807e-05,
+      "loss": 0.8496,
+      "step": 4624
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6829453742911495,
+      "learning_rate": 1.9674877364387083e-05,
+      "loss": 0.836,
+      "step": 4625
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.5201725051442043,
+      "learning_rate": 1.9674684349455893e-05,
+      "loss": 0.7323,
+      "step": 4626
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4102682373250557,
+      "learning_rate": 1.967449127819563e-05,
+      "loss": 0.8162,
+      "step": 4627
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.295218226589827,
+      "learning_rate": 1.9674298150607415e-05,
+      "loss": 0.7654,
+      "step": 4628
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5508558470783735,
+      "learning_rate": 1.9674104966692376e-05,
+      "loss": 0.7795,
+      "step": 4629
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6723415918057511,
+      "learning_rate": 1.9673911726451632e-05,
+      "loss": 0.6764,
+      "step": 4630
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.960227676180097,
+      "learning_rate": 1.967371842988631e-05,
+      "loss": 0.7502,
+      "step": 4631
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.921544540535672,
+      "learning_rate": 1.967352507699754e-05,
+      "loss": 0.8194,
+      "step": 4632
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.54087441737435,
+      "learning_rate": 1.9673331667786445e-05,
+      "loss": 0.8733,
+      "step": 4633
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5725219540477968,
+      "learning_rate": 1.9673138202254148e-05,
+      "loss": 0.5248,
+      "step": 4634
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.557388368448773,
+      "learning_rate": 1.9672944680401776e-05,
+      "loss": 0.7149,
+      "step": 4635
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9537419419369648,
+      "learning_rate": 1.967275110223046e-05,
+      "loss": 0.8086,
+      "step": 4636
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5640731811813717,
+      "learning_rate": 1.9672557467741325e-05,
+      "loss": 0.705,
+      "step": 4637
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4366483084835862,
+      "learning_rate": 1.9672363776935503e-05,
+      "loss": 0.7688,
+      "step": 4638
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6964636665988317,
+      "learning_rate": 1.9672170029814112e-05,
+      "loss": 0.9955,
+      "step": 4639
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3844714647376288,
+      "learning_rate": 1.967197622637828e-05,
+      "loss": 0.6083,
+      "step": 4640
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8206682957483369,
+      "learning_rate": 1.967178236662915e-05,
+      "loss": 0.8003,
+      "step": 4641
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5768213216304718,
+      "learning_rate": 1.9671588450567833e-05,
+      "loss": 0.6799,
+      "step": 4642
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.484474513196825,
+      "learning_rate": 1.9671394478195472e-05,
+      "loss": 0.6812,
+      "step": 4643
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.424484174001809,
+      "learning_rate": 1.9671200449513188e-05,
+      "loss": 0.7131,
+      "step": 4644
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8002282561966099,
+      "learning_rate": 1.9671006364522115e-05,
+      "loss": 0.761,
+      "step": 4645
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.375141340832687,
+      "learning_rate": 1.9670812223223378e-05,
+      "loss": 0.8417,
+      "step": 4646
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8939107566694726,
+      "learning_rate": 1.9670618025618112e-05,
+      "loss": 0.7463,
+      "step": 4647
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5108233020039776,
+      "learning_rate": 1.9670423771707442e-05,
+      "loss": 0.6903,
+      "step": 4648
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5658747528068175,
+      "learning_rate": 1.967022946149251e-05,
+      "loss": 0.6696,
+      "step": 4649
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3401968782648765,
+      "learning_rate": 1.9670035094974435e-05,
+      "loss": 0.8172,
+      "step": 4650
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.214990130202009,
+      "learning_rate": 1.9669840672154358e-05,
+      "loss": 0.7812,
+      "step": 4651
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8524538012858525,
+      "learning_rate": 1.9669646193033406e-05,
+      "loss": 0.7821,
+      "step": 4652
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.272679834735901,
+      "learning_rate": 1.9669451657612712e-05,
+      "loss": 0.8063,
+      "step": 4653
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2590318517967096,
+      "learning_rate": 1.966925706589341e-05,
+      "loss": 0.6247,
+      "step": 4654
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5233895867979033,
+      "learning_rate": 1.966906241787663e-05,
+      "loss": 0.6223,
+      "step": 4655
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6192504679327886,
+      "learning_rate": 1.966886771356351e-05,
+      "loss": 0.5504,
+      "step": 4656
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5357027312778253,
+      "learning_rate": 1.966867295295518e-05,
+      "loss": 0.7415,
+      "step": 4657
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9486609332326137,
+      "learning_rate": 1.9668478136052776e-05,
+      "loss": 0.8875,
+      "step": 4658
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0197676685702186,
+      "learning_rate": 1.966828326285743e-05,
+      "loss": 0.729,
+      "step": 4659
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5017171795600794,
+      "learning_rate": 1.966808833337028e-05,
+      "loss": 0.7493,
+      "step": 4660
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7396899233364158,
+      "learning_rate": 1.9667893347592458e-05,
+      "loss": 0.7931,
+      "step": 4661
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.1002682551452674,
+      "learning_rate": 1.96676983055251e-05,
+      "loss": 0.7695,
+      "step": 4662
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5927592237296302,
+      "learning_rate": 1.9667503207169337e-05,
+      "loss": 0.7462,
+      "step": 4663
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.82225395875891,
+      "learning_rate": 1.9667308052526318e-05,
+      "loss": 0.9644,
+      "step": 4664
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9305614874398453,
+      "learning_rate": 1.9667112841597164e-05,
+      "loss": 0.6839,
+      "step": 4665
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4392160081408905,
+      "learning_rate": 1.9666917574383026e-05,
+      "loss": 0.8215,
+      "step": 4666
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7333661163197924,
+      "learning_rate": 1.9666722250885026e-05,
+      "loss": 0.7431,
+      "step": 4667
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3184038296904563,
+      "learning_rate": 1.9666526871104312e-05,
+      "loss": 0.6898,
+      "step": 4668
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5187561277383461,
+      "learning_rate": 1.9666331435042017e-05,
+      "loss": 0.709,
+      "step": 4669
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3550160781481335,
+      "learning_rate": 1.9666135942699278e-05,
+      "loss": 0.7448,
+      "step": 4670
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8415502047752121,
+      "learning_rate": 1.966594039407724e-05,
+      "loss": 0.8703,
+      "step": 4671
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8220409733763,
+      "learning_rate": 1.9665744789177033e-05,
+      "loss": 0.7173,
+      "step": 4672
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.572763289114503,
+      "learning_rate": 1.96655491279998e-05,
+      "loss": 0.7791,
+      "step": 4673
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.0422928256621775,
+      "learning_rate": 1.9665353410546683e-05,
+      "loss": 0.7628,
+      "step": 4674
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.718607211269204,
+      "learning_rate": 1.9665157636818816e-05,
+      "loss": 0.6425,
+      "step": 4675
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4764240095341967,
+      "learning_rate": 1.9664961806817343e-05,
+      "loss": 0.7808,
+      "step": 4676
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9951883122243506,
+      "learning_rate": 1.96647659205434e-05,
+      "loss": 0.7145,
+      "step": 4677
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.363879677688092,
+      "learning_rate": 1.9664569977998132e-05,
+      "loss": 0.8324,
+      "step": 4678
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5584808820721863,
+      "learning_rate": 1.9664373979182678e-05,
+      "loss": 0.9258,
+      "step": 4679
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5070189784950192,
+      "learning_rate": 1.9664177924098178e-05,
+      "loss": 0.7748,
+      "step": 4680
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.402795682609466,
+      "learning_rate": 1.9663981812745777e-05,
+      "loss": 0.7942,
+      "step": 4681
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9452463421477804,
+      "learning_rate": 1.966378564512661e-05,
+      "loss": 0.8119,
+      "step": 4682
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7107981002435524,
+      "learning_rate": 1.9663589421241827e-05,
+      "loss": 0.8198,
+      "step": 4683
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2811978714236405,
+      "learning_rate": 1.9663393141092565e-05,
+      "loss": 0.7089,
+      "step": 4684
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.619621481025237,
+      "learning_rate": 1.966319680467997e-05,
+      "loss": 0.7101,
+      "step": 4685
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.512537194623602,
+      "learning_rate": 1.9663000412005182e-05,
+      "loss": 0.8339,
+      "step": 4686
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.5267368523312275,
+      "learning_rate": 1.966280396306935e-05,
+      "loss": 0.8809,
+      "step": 4687
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7406425042144134,
+      "learning_rate": 1.966260745787361e-05,
+      "loss": 0.6995,
+      "step": 4688
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7779028767933833,
+      "learning_rate": 1.966241089641911e-05,
+      "loss": 0.7504,
+      "step": 4689
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.601853204970425,
+      "learning_rate": 1.9662214278707e-05,
+      "loss": 0.531,
+      "step": 4690
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7647810293182022,
+      "learning_rate": 1.9662017604738416e-05,
+      "loss": 0.7473,
+      "step": 4691
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5616044606436794,
+      "learning_rate": 1.9661820874514504e-05,
+      "loss": 0.7166,
+      "step": 4692
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3527274860921616,
+      "learning_rate": 1.9661624088036414e-05,
+      "loss": 0.6836,
+      "step": 4693
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.55376743513863,
+      "learning_rate": 1.966142724530529e-05,
+      "loss": 0.776,
+      "step": 4694
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8124251197248236,
+      "learning_rate": 1.966123034632228e-05,
+      "loss": 0.7686,
+      "step": 4695
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.591156697852557,
+      "learning_rate": 1.9661033391088526e-05,
+      "loss": 0.5237,
+      "step": 4696
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6072784684017836,
+      "learning_rate": 1.9660836379605176e-05,
+      "loss": 0.5105,
+      "step": 4697
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5934717650986596,
+      "learning_rate": 1.966063931187338e-05,
+      "loss": 0.6003,
+      "step": 4698
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3551589832439468,
+      "learning_rate": 1.9660442187894283e-05,
+      "loss": 0.8215,
+      "step": 4699
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5072759014020982,
+      "learning_rate": 1.9660245007669035e-05,
+      "loss": 0.7028,
+      "step": 4700
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6355518206715518,
+      "learning_rate": 1.966004777119878e-05,
+      "loss": 0.7317,
+      "step": 4701
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6824178936207368,
+      "learning_rate": 1.9659850478484668e-05,
+      "loss": 0.5893,
+      "step": 4702
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2226679978715955,
+      "learning_rate": 1.965965312952785e-05,
+      "loss": 0.7083,
+      "step": 4703
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.779294565629298,
+      "learning_rate": 1.965945572432947e-05,
+      "loss": 0.7014,
+      "step": 4704
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.033860001453453,
+      "learning_rate": 1.9659258262890683e-05,
+      "loss": 0.8411,
+      "step": 4705
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7547506336833212,
+      "learning_rate": 1.9659060745212638e-05,
+      "loss": 0.7369,
+      "step": 4706
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4631888253628065,
+      "learning_rate": 1.9658863171296484e-05,
+      "loss": 0.7051,
+      "step": 4707
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5819334884471177,
+      "learning_rate": 1.9658665541143366e-05,
+      "loss": 0.5121,
+      "step": 4708
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4301804361738932,
+      "learning_rate": 1.965846785475444e-05,
+      "loss": 0.754,
+      "step": 4709
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.600952678934344,
+      "learning_rate": 1.965827011213086e-05,
+      "loss": 0.5598,
+      "step": 4710
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8525188366043979,
+      "learning_rate": 1.965807231327377e-05,
+      "loss": 0.8321,
+      "step": 4711
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.5848075616286486,
+      "learning_rate": 1.965787445818433e-05,
+      "loss": 0.6964,
+      "step": 4712
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3365988394592474,
+      "learning_rate": 1.9657676546863682e-05,
+      "loss": 0.8178,
+      "step": 4713
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5534600078508456,
+      "learning_rate": 1.9657478579312984e-05,
+      "loss": 0.8069,
+      "step": 4714
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.806398300821407,
+      "learning_rate": 1.965728055553339e-05,
+      "loss": 0.6996,
+      "step": 4715
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6009142056479857,
+      "learning_rate": 1.965708247552605e-05,
+      "loss": 0.8103,
+      "step": 4716
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8603267567353108,
+      "learning_rate": 1.9656884339292123e-05,
+      "loss": 0.7373,
+      "step": 4717
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4856423541300823,
+      "learning_rate": 1.9656686146832754e-05,
+      "loss": 0.6506,
+      "step": 4718
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0443236941095284,
+      "learning_rate": 1.96564878981491e-05,
+      "loss": 0.8059,
+      "step": 4719
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4687143745177393,
+      "learning_rate": 1.965628959324232e-05,
+      "loss": 0.8827,
+      "step": 4720
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7838418378202328,
+      "learning_rate": 1.9656091232113562e-05,
+      "loss": 0.8231,
+      "step": 4721
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.969878491229573,
+      "learning_rate": 1.9655892814763984e-05,
+      "loss": 0.7501,
+      "step": 4722
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5463983072271863,
+      "learning_rate": 1.9655694341194743e-05,
+      "loss": 0.8423,
+      "step": 4723
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7051569500305201,
+      "learning_rate": 1.965549581140699e-05,
+      "loss": 0.9404,
+      "step": 4724
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.172623698228995,
+      "learning_rate": 1.9655297225401884e-05,
+      "loss": 0.7505,
+      "step": 4725
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4547357918963648,
+      "learning_rate": 1.965509858318058e-05,
+      "loss": 0.8474,
+      "step": 4726
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7420794430335433,
+      "learning_rate": 1.9654899884744238e-05,
+      "loss": 0.6901,
+      "step": 4727
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.971386213696605,
+      "learning_rate": 1.965470113009401e-05,
+      "loss": 0.6774,
+      "step": 4728
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4283586783722477,
+      "learning_rate": 1.9654502319231056e-05,
+      "loss": 0.7673,
+      "step": 4729
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5323417776694872,
+      "learning_rate": 1.9654303452156535e-05,
+      "loss": 0.7889,
+      "step": 4730
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4782580422803013,
+      "learning_rate": 1.96541045288716e-05,
+      "loss": 0.7735,
+      "step": 4731
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4951379462891028,
+      "learning_rate": 1.9653905549377413e-05,
+      "loss": 0.925,
+      "step": 4732
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.45041959356616,
+      "learning_rate": 1.965370651367513e-05,
+      "loss": 0.8232,
+      "step": 4733
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4906267287515904,
+      "learning_rate": 1.9653507421765914e-05,
+      "loss": 0.6612,
+      "step": 4734
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.329840544787214,
+      "learning_rate": 1.965330827365092e-05,
+      "loss": 0.858,
+      "step": 4735
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.500002434310313,
+      "learning_rate": 1.965310906933131e-05,
+      "loss": 0.7569,
+      "step": 4736
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3900478374273826,
+      "learning_rate": 1.9652909808808242e-05,
+      "loss": 0.6453,
+      "step": 4737
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3448661593637479,
+      "learning_rate": 1.965271049208288e-05,
+      "loss": 0.7551,
+      "step": 4738
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6766293860246613,
+      "learning_rate": 1.9652511119156375e-05,
+      "loss": 0.6503,
+      "step": 4739
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5988156822897601,
+      "learning_rate": 1.96523116900299e-05,
+      "loss": 0.7655,
+      "step": 4740
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5350714135820769,
+      "learning_rate": 1.9652112204704608e-05,
+      "loss": 0.8047,
+      "step": 4741
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4990983606852826,
+      "learning_rate": 1.9651912663181665e-05,
+      "loss": 0.9031,
+      "step": 4742
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5694758778851536,
+      "learning_rate": 1.965171306546223e-05,
+      "loss": 0.7449,
+      "step": 4743
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9217512859873445,
+      "learning_rate": 1.9651513411547466e-05,
+      "loss": 0.8563,
+      "step": 4744
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.483675554300428,
+      "learning_rate": 1.9651313701438533e-05,
+      "loss": 0.6998,
+      "step": 4745
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6167856080507876,
+      "learning_rate": 1.9651113935136596e-05,
+      "loss": 0.5719,
+      "step": 4746
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.603760613068446,
+      "learning_rate": 1.965091411264282e-05,
+      "loss": 0.5645,
+      "step": 4747
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.498813195766189,
+      "learning_rate": 1.9650714233958367e-05,
+      "loss": 0.8363,
+      "step": 4748
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6050565873985236,
+      "learning_rate": 1.96505142990844e-05,
+      "loss": 0.6561,
+      "step": 4749
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7576145282416298,
+      "learning_rate": 1.9650314308022083e-05,
+      "loss": 0.8235,
+      "step": 4750
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.629478083441846,
+      "learning_rate": 1.9650114260772583e-05,
+      "loss": 0.7684,
+      "step": 4751
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5881160994297001,
+      "learning_rate": 1.964991415733706e-05,
+      "loss": 0.8281,
+      "step": 4752
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2148138336012626,
+      "learning_rate": 1.964971399771668e-05,
+      "loss": 0.7755,
+      "step": 4753
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.63498957803821,
+      "learning_rate": 1.964951378191261e-05,
+      "loss": 0.8431,
+      "step": 4754
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5586256629935062,
+      "learning_rate": 1.9649313509926022e-05,
+      "loss": 0.7941,
+      "step": 4755
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5330089945996062,
+      "learning_rate": 1.964911318175807e-05,
+      "loss": 0.7757,
+      "step": 4756
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.499581874425071,
+      "learning_rate": 1.964891279740993e-05,
+      "loss": 0.8048,
+      "step": 4757
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4139300190030135,
+      "learning_rate": 1.9648712356882763e-05,
+      "loss": 0.748,
+      "step": 4758
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3378254737966462,
+      "learning_rate": 1.9648511860177737e-05,
+      "loss": 0.849,
+      "step": 4759
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.213558024266813,
+      "learning_rate": 1.964831130729602e-05,
+      "loss": 0.6945,
+      "step": 4760
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3404324954672713,
+      "learning_rate": 1.9648110698238782e-05,
+      "loss": 0.804,
+      "step": 4761
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2951632181024135,
+      "learning_rate": 1.964791003300719e-05,
+      "loss": 0.6634,
+      "step": 4762
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4276653036238176,
+      "learning_rate": 1.9647709311602408e-05,
+      "loss": 0.7176,
+      "step": 4763
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.235156515282581,
+      "learning_rate": 1.964750853402561e-05,
+      "loss": 0.66,
+      "step": 4764
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5819014439984993,
+      "learning_rate": 1.9647307700277963e-05,
+      "loss": 0.81,
+      "step": 4765
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7790635161739277,
+      "learning_rate": 1.9647106810360636e-05,
+      "loss": 0.7634,
+      "step": 4766
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8899192261489548,
+      "learning_rate": 1.9646905864274798e-05,
+      "loss": 0.8413,
+      "step": 4767
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.443293358140092,
+      "learning_rate": 1.964670486202162e-05,
+      "loss": 0.8015,
+      "step": 4768
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7406074872880983,
+      "learning_rate": 1.9646503803602272e-05,
+      "loss": 0.9908,
+      "step": 4769
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4628985533512184,
+      "learning_rate": 1.9646302689017925e-05,
+      "loss": 0.7604,
+      "step": 4770
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.550194224316085,
+      "learning_rate": 1.964610151826975e-05,
+      "loss": 0.8046,
+      "step": 4771
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7559052505509711,
+      "learning_rate": 1.9645900291358918e-05,
+      "loss": 0.8664,
+      "step": 4772
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5833885125062759,
+      "learning_rate": 1.9645699008286602e-05,
+      "loss": 0.7624,
+      "step": 4773
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5653094055675312,
+      "learning_rate": 1.9645497669053973e-05,
+      "loss": 0.7146,
+      "step": 4774
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4739091394594561,
+      "learning_rate": 1.9645296273662202e-05,
+      "loss": 0.8361,
+      "step": 4775
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4643119245345306,
+      "learning_rate": 1.964509482211246e-05,
+      "loss": 0.6246,
+      "step": 4776
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5825873664565833,
+      "learning_rate": 1.9644893314405924e-05,
+      "loss": 0.7766,
+      "step": 4777
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6278156777179824,
+      "learning_rate": 1.964469175054377e-05,
+      "loss": 0.7344,
+      "step": 4778
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5603032962123282,
+      "learning_rate": 1.964449013052716e-05,
+      "loss": 0.6342,
+      "step": 4779
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.499797642570719,
+      "learning_rate": 1.964428845435728e-05,
+      "loss": 0.9177,
+      "step": 4780
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4931412457826505,
+      "learning_rate": 1.9644086722035295e-05,
+      "loss": 0.7279,
+      "step": 4781
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4707050573607334,
+      "learning_rate": 1.9643884933562387e-05,
+      "loss": 0.7077,
+      "step": 4782
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6032075698666671,
+      "learning_rate": 1.9643683088939726e-05,
+      "loss": 0.5234,
+      "step": 4783
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3129564252398407,
+      "learning_rate": 1.9643481188168492e-05,
+      "loss": 0.6702,
+      "step": 4784
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6142336889889303,
+      "learning_rate": 1.9643279231249855e-05,
+      "loss": 0.5462,
+      "step": 4785
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4959086768106142,
+      "learning_rate": 1.9643077218184993e-05,
+      "loss": 0.7552,
+      "step": 4786
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6659963635393875,
+      "learning_rate": 1.9642875148975084e-05,
+      "loss": 0.7397,
+      "step": 4787
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7482810943825016,
+      "learning_rate": 1.9642673023621303e-05,
+      "loss": 0.9381,
+      "step": 4788
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.484549734672055,
+      "learning_rate": 1.9642470842124823e-05,
+      "loss": 0.8065,
+      "step": 4789
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5846041103748152,
+      "learning_rate": 1.964226860448683e-05,
+      "loss": 0.6654,
+      "step": 4790
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3164715412233274,
+      "learning_rate": 1.9642066310708493e-05,
+      "loss": 0.827,
+      "step": 4791
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.504994576167709,
+      "learning_rate": 1.9641863960790998e-05,
+      "loss": 0.754,
+      "step": 4792
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4313876031638804,
+      "learning_rate": 1.9641661554735513e-05,
+      "loss": 0.7522,
+      "step": 4793
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7157230675057151,
+      "learning_rate": 1.9641459092543224e-05,
+      "loss": 0.8837,
+      "step": 4794
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1737316063693775,
+      "learning_rate": 1.9641256574215308e-05,
+      "loss": 0.7188,
+      "step": 4795
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3373931383260584,
+      "learning_rate": 1.9641053999752942e-05,
+      "loss": 0.6959,
+      "step": 4796
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4381721184672045,
+      "learning_rate": 1.964085136915731e-05,
+      "loss": 0.8372,
+      "step": 4797
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6975827008037623,
+      "learning_rate": 1.964064868242959e-05,
+      "loss": 0.6056,
+      "step": 4798
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5038855092597005,
+      "learning_rate": 1.964044593957096e-05,
+      "loss": 0.7145,
+      "step": 4799
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4861964869533932,
+      "learning_rate": 1.96402431405826e-05,
+      "loss": 0.7744,
+      "step": 4800
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.5139249639350925,
+      "learning_rate": 1.9640040285465693e-05,
+      "loss": 0.7983,
+      "step": 4801
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4304154024644422,
+      "learning_rate": 1.963983737422142e-05,
+      "loss": 0.6395,
+      "step": 4802
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5654747544140204,
+      "learning_rate": 1.9639634406850962e-05,
+      "loss": 0.7541,
+      "step": 4803
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3809495446203546,
+      "learning_rate": 1.96394313833555e-05,
+      "loss": 0.7294,
+      "step": 4804
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4173858273353726,
+      "learning_rate": 1.9639228303736216e-05,
+      "loss": 0.7685,
+      "step": 4805
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.432902779473568,
+      "learning_rate": 1.9639025167994295e-05,
+      "loss": 0.6441,
+      "step": 4806
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4039966811559503,
+      "learning_rate": 1.9638821976130916e-05,
+      "loss": 0.7863,
+      "step": 4807
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4391053657301767,
+      "learning_rate": 1.9638618728147266e-05,
+      "loss": 0.5672,
+      "step": 4808
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6650252227333875,
+      "learning_rate": 1.9638415424044527e-05,
+      "loss": 0.7193,
+      "step": 4809
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8280910999578905,
+      "learning_rate": 1.963821206382388e-05,
+      "loss": 0.8199,
+      "step": 4810
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5085120046361564,
+      "learning_rate": 1.963800864748651e-05,
+      "loss": 0.5988,
+      "step": 4811
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.59646337858648,
+      "learning_rate": 1.9637805175033603e-05,
+      "loss": 0.7451,
+      "step": 4812
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4142190752853803,
+      "learning_rate": 1.9637601646466343e-05,
+      "loss": 0.6647,
+      "step": 4813
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.774166783897241,
+      "learning_rate": 1.9637398061785914e-05,
+      "loss": 0.9757,
+      "step": 4814
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6341501140188899,
+      "learning_rate": 1.9637194420993504e-05,
+      "loss": 0.8511,
+      "step": 4815
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9592006768717671,
+      "learning_rate": 1.9636990724090297e-05,
+      "loss": 0.727,
+      "step": 4816
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.454098611516677,
+      "learning_rate": 1.963678697107748e-05,
+      "loss": 0.8776,
+      "step": 4817
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6089571462999257,
+      "learning_rate": 1.9636583161956235e-05,
+      "loss": 0.5277,
+      "step": 4818
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5100160850918458,
+      "learning_rate": 1.9636379296727754e-05,
+      "loss": 0.6626,
+      "step": 4819
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6388515234623018,
+      "learning_rate": 1.9636175375393223e-05,
+      "loss": 0.814,
+      "step": 4820
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9309790158599243,
+      "learning_rate": 1.9635971397953826e-05,
+      "loss": 0.6941,
+      "step": 4821
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5233108920663998,
+      "learning_rate": 1.963576736441075e-05,
+      "loss": 0.6832,
+      "step": 4822
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6071001026667018,
+      "learning_rate": 1.9635563274765194e-05,
+      "loss": 0.481,
+      "step": 4823
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8827796841032418,
+      "learning_rate": 1.9635359129018334e-05,
+      "loss": 0.7802,
+      "step": 4824
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8854583405715848,
+      "learning_rate": 1.963515492717136e-05,
+      "loss": 0.6861,
+      "step": 4825
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5211015580919924,
+      "learning_rate": 1.9634950669225465e-05,
+      "loss": 0.7782,
+      "step": 4826
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.649148119009354,
+      "learning_rate": 1.9634746355181836e-05,
+      "loss": 0.7242,
+      "step": 4827
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3954310628503128,
+      "learning_rate": 1.9634541985041667e-05,
+      "loss": 0.7987,
+      "step": 4828
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5498779645705305,
+      "learning_rate": 1.9634337558806142e-05,
+      "loss": 0.6184,
+      "step": 4829
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3725815346566372,
+      "learning_rate": 1.9634133076476453e-05,
+      "loss": 0.8951,
+      "step": 4830
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.530123735632193,
+      "learning_rate": 1.9633928538053792e-05,
+      "loss": 0.5579,
+      "step": 4831
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3976245926292759,
+      "learning_rate": 1.9633723943539352e-05,
+      "loss": 0.7813,
+      "step": 4832
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9103872332256955,
+      "learning_rate": 1.9633519292934317e-05,
+      "loss": 0.7288,
+      "step": 4833
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.493390462457636,
+      "learning_rate": 1.9633314586239884e-05,
+      "loss": 0.7614,
+      "step": 4834
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7504458703771775,
+      "learning_rate": 1.9633109823457243e-05,
+      "loss": 0.7789,
+      "step": 4835
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3655035823824486,
+      "learning_rate": 1.9632905004587588e-05,
+      "loss": 0.6589,
+      "step": 4836
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9057095033121254,
+      "learning_rate": 1.963270012963211e-05,
+      "loss": 0.7845,
+      "step": 4837
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.799233857228774,
+      "learning_rate": 1.9632495198592003e-05,
+      "loss": 0.9142,
+      "step": 4838
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5468895961077054,
+      "learning_rate": 1.9632290211468458e-05,
+      "loss": 0.8545,
+      "step": 4839
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6391881568685114,
+      "learning_rate": 1.963208516826267e-05,
+      "loss": 0.8079,
+      "step": 4840
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.589452704298672,
+      "learning_rate": 1.9631880068975834e-05,
+      "loss": 0.7282,
+      "step": 4841
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6096479467324467,
+      "learning_rate": 1.9631674913609137e-05,
+      "loss": 0.7766,
+      "step": 4842
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6254364216781765,
+      "learning_rate": 1.9631469702163786e-05,
+      "loss": 0.7651,
+      "step": 4843
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.092949325737862,
+      "learning_rate": 1.9631264434640965e-05,
+      "loss": 0.7093,
+      "step": 4844
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5337160722886696,
+      "learning_rate": 1.9631059111041876e-05,
+      "loss": 0.8752,
+      "step": 4845
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6372853225992733,
+      "learning_rate": 1.9630853731367715e-05,
+      "loss": 0.5695,
+      "step": 4846
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8916723628000243,
+      "learning_rate": 1.9630648295619668e-05,
+      "loss": 0.5746,
+      "step": 4847
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4294816241591195,
+      "learning_rate": 1.963044280379894e-05,
+      "loss": 0.8065,
+      "step": 4848
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5976013359616501,
+      "learning_rate": 1.9630237255906724e-05,
+      "loss": 0.5803,
+      "step": 4849
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.855342088873177,
+      "learning_rate": 1.963003165194422e-05,
+      "loss": 0.7958,
+      "step": 4850
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5556882861928054,
+      "learning_rate": 1.9629825991912622e-05,
+      "loss": 0.7485,
+      "step": 4851
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0549110587838624,
+      "learning_rate": 1.962962027581313e-05,
+      "loss": 0.7692,
+      "step": 4852
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.037298429574677,
+      "learning_rate": 1.9629414503646936e-05,
+      "loss": 0.8826,
+      "step": 4853
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5146429126850567,
+      "learning_rate": 1.9629208675415246e-05,
+      "loss": 0.9369,
+      "step": 4854
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5380132535159162,
+      "learning_rate": 1.9629002791119252e-05,
+      "loss": 0.5102,
+      "step": 4855
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.558953234302432,
+      "learning_rate": 1.9628796850760156e-05,
+      "loss": 0.8182,
+      "step": 4856
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5137881689620507,
+      "learning_rate": 1.962859085433916e-05,
+      "loss": 0.6783,
+      "step": 4857
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4567392426525843,
+      "learning_rate": 1.9628384801857454e-05,
+      "loss": 0.7972,
+      "step": 4858
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6376922115449828,
+      "learning_rate": 1.9628178693316243e-05,
+      "loss": 0.8686,
+      "step": 4859
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.2786441917408973,
+      "learning_rate": 1.962797252871673e-05,
+      "loss": 0.7969,
+      "step": 4860
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.461418577205374,
+      "learning_rate": 1.9627766308060115e-05,
+      "loss": 0.7106,
+      "step": 4861
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5457344276175073,
+      "learning_rate": 1.9627560031347594e-05,
+      "loss": 0.8408,
+      "step": 4862
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.634609122629761,
+      "learning_rate": 1.962735369858037e-05,
+      "loss": 0.7597,
+      "step": 4863
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.502658976923086,
+      "learning_rate": 1.9627147309759646e-05,
+      "loss": 0.8303,
+      "step": 4864
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.428588092973733,
+      "learning_rate": 1.9626940864886622e-05,
+      "loss": 0.7001,
+      "step": 4865
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7106750427959716,
+      "learning_rate": 1.96267343639625e-05,
+      "loss": 0.7658,
+      "step": 4866
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.673849333251939,
+      "learning_rate": 1.9626527806988485e-05,
+      "loss": 0.9171,
+      "step": 4867
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6745371353903244,
+      "learning_rate": 1.9626321193965774e-05,
+      "loss": 0.7887,
+      "step": 4868
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4454555477333688,
+      "learning_rate": 1.9626114524895574e-05,
+      "loss": 0.6654,
+      "step": 4869
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.372823212090702,
+      "learning_rate": 1.962590779977909e-05,
+      "loss": 0.8653,
+      "step": 4870
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6748083412172956,
+      "learning_rate": 1.9625701018617524e-05,
+      "loss": 0.7817,
+      "step": 4871
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8134737886511358,
+      "learning_rate": 1.9625494181412076e-05,
+      "loss": 0.8584,
+      "step": 4872
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8892592678245133,
+      "learning_rate": 1.9625287288163954e-05,
+      "loss": 0.7239,
+      "step": 4873
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.645265490439214,
+      "learning_rate": 1.9625080338874363e-05,
+      "loss": 0.7457,
+      "step": 4874
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4164964279944225,
+      "learning_rate": 1.9624873333544507e-05,
+      "loss": 0.798,
+      "step": 4875
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.881724539617315,
+      "learning_rate": 1.9624666272175588e-05,
+      "loss": 0.7736,
+      "step": 4876
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6808939075069844,
+      "learning_rate": 1.962445915476882e-05,
+      "loss": 0.6992,
+      "step": 4877
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.49051418611284,
+      "learning_rate": 1.9624251981325403e-05,
+      "loss": 0.8886,
+      "step": 4878
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.2742085854122225,
+      "learning_rate": 1.9624044751846543e-05,
+      "loss": 0.6532,
+      "step": 4879
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.949296243702753,
+      "learning_rate": 1.9623837466333445e-05,
+      "loss": 0.5979,
+      "step": 4880
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.867993731899855,
+      "learning_rate": 1.962363012478732e-05,
+      "loss": 0.7018,
+      "step": 4881
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5148445127821033,
+      "learning_rate": 1.9623422727209376e-05,
+      "loss": 0.7517,
+      "step": 4882
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4992870774330411,
+      "learning_rate": 1.9623215273600817e-05,
+      "loss": 0.8945,
+      "step": 4883
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3550934448268632,
+      "learning_rate": 1.962300776396285e-05,
+      "loss": 0.7204,
+      "step": 4884
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.50549053508602,
+      "learning_rate": 1.9622800198296686e-05,
+      "loss": 0.7681,
+      "step": 4885
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.5011834380002123,
+      "learning_rate": 1.9622592576603536e-05,
+      "loss": 0.7284,
+      "step": 4886
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7565065423614041,
+      "learning_rate": 1.96223848988846e-05,
+      "loss": 0.7902,
+      "step": 4887
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5867808675762951,
+      "learning_rate": 1.9622177165141098e-05,
+      "loss": 0.52,
+      "step": 4888
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.384795184027605,
+      "learning_rate": 1.962196937537423e-05,
+      "loss": 0.8168,
+      "step": 4889
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4604360230271236,
+      "learning_rate": 1.962176152958521e-05,
+      "loss": 0.762,
+      "step": 4890
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.552029792069929,
+      "learning_rate": 1.962155362777525e-05,
+      "loss": 0.7009,
+      "step": 4891
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5626330992607682,
+      "learning_rate": 1.9621345669945562e-05,
+      "loss": 0.8203,
+      "step": 4892
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2376051868358533,
+      "learning_rate": 1.9621137656097347e-05,
+      "loss": 0.8174,
+      "step": 4893
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6624157434164335,
+      "learning_rate": 1.9620929586231828e-05,
+      "loss": 0.808,
+      "step": 4894
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5009565336448347,
+      "learning_rate": 1.9620721460350207e-05,
+      "loss": 0.8125,
+      "step": 4895
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.4649616022670857,
+      "learning_rate": 1.9620513278453704e-05,
+      "loss": 0.8623,
+      "step": 4896
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4027441088532895,
+      "learning_rate": 1.9620305040543523e-05,
+      "loss": 0.8175,
+      "step": 4897
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5159182684749772,
+      "learning_rate": 1.9620096746620882e-05,
+      "loss": 0.7171,
+      "step": 4898
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4901129079604067,
+      "learning_rate": 1.9619888396686992e-05,
+      "loss": 0.7679,
+      "step": 4899
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5467854369505298,
+      "learning_rate": 1.9619679990743065e-05,
+      "loss": 0.7432,
+      "step": 4900
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4527761557919068,
+      "learning_rate": 1.9619471528790317e-05,
+      "loss": 0.734,
+      "step": 4901
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.623117692048477,
+      "learning_rate": 1.961926301082996e-05,
+      "loss": 0.6855,
+      "step": 4902
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.554446037603072,
+      "learning_rate": 1.9619054436863206e-05,
+      "loss": 0.6678,
+      "step": 4903
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6072516182491886,
+      "learning_rate": 1.9618845806891276e-05,
+      "loss": 0.5524,
+      "step": 4904
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7312315341357847,
+      "learning_rate": 1.9618637120915376e-05,
+      "loss": 0.8751,
+      "step": 4905
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5935692089403264,
+      "learning_rate": 1.9618428378936728e-05,
+      "loss": 0.8597,
+      "step": 4906
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.425960997735624,
+      "learning_rate": 1.9618219580956544e-05,
+      "loss": 0.7266,
+      "step": 4907
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.709255403280767,
+      "learning_rate": 1.9618010726976044e-05,
+      "loss": 0.8226,
+      "step": 4908
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5818210561512502,
+      "learning_rate": 1.9617801816996437e-05,
+      "loss": 0.7569,
+      "step": 4909
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5285506743178072,
+      "learning_rate": 1.961759285101894e-05,
+      "loss": 0.7094,
+      "step": 4910
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4998702455245243,
+      "learning_rate": 1.961738382904478e-05,
+      "loss": 0.7842,
+      "step": 4911
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.546561299594898,
+      "learning_rate": 1.961717475107516e-05,
+      "loss": 0.8341,
+      "step": 4912
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4042918892367027,
+      "learning_rate": 1.9616965617111305e-05,
+      "loss": 0.7973,
+      "step": 4913
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7112249549226826,
+      "learning_rate": 1.9616756427154434e-05,
+      "loss": 0.8839,
+      "step": 4914
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6460799465350489,
+      "learning_rate": 1.961654718120576e-05,
+      "loss": 0.755,
+      "step": 4915
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4817359915711372,
+      "learning_rate": 1.9616337879266507e-05,
+      "loss": 0.7064,
+      "step": 4916
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5968916167595951,
+      "learning_rate": 1.9616128521337886e-05,
+      "loss": 0.5305,
+      "step": 4917
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4126012971660866,
+      "learning_rate": 1.9615919107421123e-05,
+      "loss": 0.6911,
+      "step": 4918
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6065041590815526,
+      "learning_rate": 1.9615709637517437e-05,
+      "loss": 0.632,
+      "step": 4919
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2993667526527606,
+      "learning_rate": 1.961550011162804e-05,
+      "loss": 0.7206,
+      "step": 4920
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3583989773268625,
+      "learning_rate": 1.9615290529754162e-05,
+      "loss": 0.8261,
+      "step": 4921
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7183250414290734,
+      "learning_rate": 1.961508089189702e-05,
+      "loss": 0.8933,
+      "step": 4922
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4678287647959105,
+      "learning_rate": 1.9614871198057826e-05,
+      "loss": 0.8717,
+      "step": 4923
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5715299504507036,
+      "learning_rate": 1.961466144823781e-05,
+      "loss": 0.7426,
+      "step": 4924
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3434439325669802,
+      "learning_rate": 1.9614451642438193e-05,
+      "loss": 0.583,
+      "step": 4925
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.61363108469581,
+      "learning_rate": 1.9614241780660196e-05,
+      "loss": 0.9144,
+      "step": 4926
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4226219625926297,
+      "learning_rate": 1.9614031862905034e-05,
+      "loss": 0.8018,
+      "step": 4927
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8879543533872414,
+      "learning_rate": 1.961382188917394e-05,
+      "loss": 0.8622,
+      "step": 4928
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4738791119430312,
+      "learning_rate": 1.9613611859468128e-05,
+      "loss": 0.8657,
+      "step": 4929
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6321030991307296,
+      "learning_rate": 1.9613401773788826e-05,
+      "loss": 0.5526,
+      "step": 4930
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4570740067765628,
+      "learning_rate": 1.9613191632137256e-05,
+      "loss": 0.9139,
+      "step": 4931
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8275778965783258,
+      "learning_rate": 1.961298143451464e-05,
+      "loss": 0.6439,
+      "step": 4932
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3907107762447821,
+      "learning_rate": 1.9612771180922203e-05,
+      "loss": 0.8569,
+      "step": 4933
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4393733001068627,
+      "learning_rate": 1.9612560871361163e-05,
+      "loss": 0.8297,
+      "step": 4934
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6033496913820251,
+      "learning_rate": 1.9612350505832757e-05,
+      "loss": 0.5508,
+      "step": 4935
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8408802222405316,
+      "learning_rate": 1.9612140084338204e-05,
+      "loss": 0.7848,
+      "step": 4936
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5919325612217734,
+      "learning_rate": 1.9611929606878724e-05,
+      "loss": 0.4729,
+      "step": 4937
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4754812357120821,
+      "learning_rate": 1.9611719073455547e-05,
+      "loss": 0.7581,
+      "step": 4938
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4720499343470357,
+      "learning_rate": 1.9611508484069902e-05,
+      "loss": 0.7983,
+      "step": 4939
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.613343767900312,
+      "learning_rate": 1.961129783872301e-05,
+      "loss": 0.8334,
+      "step": 4940
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7598020731687856,
+      "learning_rate": 1.9611087137416097e-05,
+      "loss": 0.7294,
+      "step": 4941
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3564444343289073,
+      "learning_rate": 1.9610876380150396e-05,
+      "loss": 0.6682,
+      "step": 4942
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.435066490785988,
+      "learning_rate": 1.9610665566927127e-05,
+      "loss": 0.9329,
+      "step": 4943
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.341652963894986,
+      "learning_rate": 1.961045469774752e-05,
+      "loss": 0.7229,
+      "step": 4944
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6060393246620395,
+      "learning_rate": 1.9610243772612804e-05,
+      "loss": 0.7094,
+      "step": 4945
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5549579641304954,
+      "learning_rate": 1.9610032791524206e-05,
+      "loss": 0.6781,
+      "step": 4946
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6085454278523683,
+      "learning_rate": 1.9609821754482957e-05,
+      "loss": 0.7724,
+      "step": 4947
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0405018117930562,
+      "learning_rate": 1.960961066149028e-05,
+      "loss": 0.8561,
+      "step": 4948
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.393063757622227,
+      "learning_rate": 1.960939951254741e-05,
+      "loss": 0.7292,
+      "step": 4949
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4296561686013742,
+      "learning_rate": 1.9609188307655572e-05,
+      "loss": 0.7198,
+      "step": 4950
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6605179101165576,
+      "learning_rate": 1.9608977046815997e-05,
+      "loss": 0.5606,
+      "step": 4951
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4218112185292637,
+      "learning_rate": 1.9608765730029917e-05,
+      "loss": 0.8718,
+      "step": 4952
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4986323435602091,
+      "learning_rate": 1.960855435729856e-05,
+      "loss": 0.8381,
+      "step": 4953
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8908971060017183,
+      "learning_rate": 1.9608342928623157e-05,
+      "loss": 0.6123,
+      "step": 4954
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7299888056246793,
+      "learning_rate": 1.960813144400494e-05,
+      "loss": 0.7425,
+      "step": 4955
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4396179771652986,
+      "learning_rate": 1.960791990344514e-05,
+      "loss": 0.8285,
+      "step": 4956
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.412237046674804,
+      "learning_rate": 1.960770830694499e-05,
+      "loss": 0.7755,
+      "step": 4957
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5040939964437539,
+      "learning_rate": 1.960749665450572e-05,
+      "loss": 0.7901,
+      "step": 4958
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.381229261318834,
+      "learning_rate": 1.9607284946128562e-05,
+      "loss": 0.8624,
+      "step": 4959
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4423629689199486,
+      "learning_rate": 1.9607073181814748e-05,
+      "loss": 0.8276,
+      "step": 4960
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4893320062693947,
+      "learning_rate": 1.9606861361565514e-05,
+      "loss": 0.7367,
+      "step": 4961
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7000262570289344,
+      "learning_rate": 1.960664948538209e-05,
+      "loss": 0.7916,
+      "step": 4962
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.345704660218831,
+      "learning_rate": 1.9606437553265713e-05,
+      "loss": 0.7745,
+      "step": 4963
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6414609898830586,
+      "learning_rate": 1.9606225565217615e-05,
+      "loss": 0.6007,
+      "step": 4964
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4704263554230073,
+      "learning_rate": 1.960601352123903e-05,
+      "loss": 0.7289,
+      "step": 4965
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.648554980240907,
+      "learning_rate": 1.9605801421331194e-05,
+      "loss": 0.7421,
+      "step": 4966
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4987326393824323,
+      "learning_rate": 1.960558926549534e-05,
+      "loss": 0.6745,
+      "step": 4967
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5825762225827824,
+      "learning_rate": 1.9605377053732705e-05,
+      "loss": 0.8607,
+      "step": 4968
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.700383815716077,
+      "learning_rate": 1.9605164786044523e-05,
+      "loss": 0.7451,
+      "step": 4969
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5374151045608084,
+      "learning_rate": 1.9604952462432032e-05,
+      "loss": 0.6644,
+      "step": 4970
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.748965949023486,
+      "learning_rate": 1.9604740082896467e-05,
+      "loss": 0.7244,
+      "step": 4971
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6362849904145501,
+      "learning_rate": 1.960452764743906e-05,
+      "loss": 0.5441,
+      "step": 4972
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.656688507374042,
+      "learning_rate": 1.960431515606106e-05,
+      "loss": 0.7867,
+      "step": 4973
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.638697664222536,
+      "learning_rate": 1.960410260876369e-05,
+      "loss": 0.8651,
+      "step": 4974
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3317354070402951,
+      "learning_rate": 1.9603890005548198e-05,
+      "loss": 0.7901,
+      "step": 4975
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5193753355375497,
+      "learning_rate": 1.9603677346415815e-05,
+      "loss": 0.7329,
+      "step": 4976
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7259291973813213,
+      "learning_rate": 1.9603464631367782e-05,
+      "loss": 0.7392,
+      "step": 4977
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.956534104186429,
+      "learning_rate": 1.960325186040534e-05,
+      "loss": 0.8747,
+      "step": 4978
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4270712996749406,
+      "learning_rate": 1.9603039033529723e-05,
+      "loss": 0.8219,
+      "step": 4979
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3769553283515676,
+      "learning_rate": 1.9602826150742174e-05,
+      "loss": 0.7099,
+      "step": 4980
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7394768652277988,
+      "learning_rate": 1.960261321204393e-05,
+      "loss": 0.8506,
+      "step": 4981
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4911874068292286,
+      "learning_rate": 1.960240021743623e-05,
+      "loss": 0.9057,
+      "step": 4982
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4659084645233147,
+      "learning_rate": 1.960218716692032e-05,
+      "loss": 0.6956,
+      "step": 4983
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4705705082146217,
+      "learning_rate": 1.9601974060497433e-05,
+      "loss": 0.7414,
+      "step": 4984
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6528050118562896,
+      "learning_rate": 1.9601760898168814e-05,
+      "loss": 0.6951,
+      "step": 4985
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.630886068953693,
+      "learning_rate": 1.9601547679935706e-05,
+      "loss": 0.7585,
+      "step": 4986
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5463079350143467,
+      "learning_rate": 1.9601334405799344e-05,
+      "loss": 0.8283,
+      "step": 4987
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5601679425298818,
+      "learning_rate": 1.9601121075760977e-05,
+      "loss": 0.7709,
+      "step": 4988
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5047180692793078,
+      "learning_rate": 1.960090768982184e-05,
+      "loss": 0.7095,
+      "step": 4989
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2210597853233796,
+      "learning_rate": 1.960069424798318e-05,
+      "loss": 0.6462,
+      "step": 4990
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5417294067711802,
+      "learning_rate": 1.9600480750246235e-05,
+      "loss": 0.7088,
+      "step": 4991
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8886439728626694,
+      "learning_rate": 1.9600267196612253e-05,
+      "loss": 0.7207,
+      "step": 4992
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6214541853833226,
+      "learning_rate": 1.960005358708248e-05,
+      "loss": 0.5611,
+      "step": 4993
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5674466891991041,
+      "learning_rate": 1.9599839921658154e-05,
+      "loss": 0.7522,
+      "step": 4994
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7433052017704105,
+      "learning_rate": 1.9599626200340523e-05,
+      "loss": 0.7773,
+      "step": 4995
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4551695086218266,
+      "learning_rate": 1.9599412423130823e-05,
+      "loss": 0.8302,
+      "step": 4996
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6368536012693777,
+      "learning_rate": 1.9599198590030308e-05,
+      "loss": 0.776,
+      "step": 4997
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6070060665851995,
+      "learning_rate": 1.959898470104022e-05,
+      "loss": 0.5237,
+      "step": 4998
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7912673273319486,
+      "learning_rate": 1.9598770756161807e-05,
+      "loss": 0.9156,
+      "step": 4999
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4815019681756036,
+      "learning_rate": 1.9598556755396307e-05,
+      "loss": 0.7569,
+      "step": 5000
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0664670885878436,
+      "learning_rate": 1.9598342698744973e-05,
+      "loss": 0.7633,
+      "step": 5001
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4744128085958677,
+      "learning_rate": 1.9598128586209047e-05,
+      "loss": 0.7836,
+      "step": 5002
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8562862441218397,
+      "learning_rate": 1.959791441778978e-05,
+      "loss": 0.9086,
+      "step": 5003
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8069255648215778,
+      "learning_rate": 1.9597700193488418e-05,
+      "loss": 0.6665,
+      "step": 5004
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.394835743085814,
+      "learning_rate": 1.9597485913306203e-05,
+      "loss": 0.7725,
+      "step": 5005
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.597232746041313,
+      "learning_rate": 1.959727157724439e-05,
+      "loss": 0.8392,
+      "step": 5006
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3878621793656147,
+      "learning_rate": 1.959705718530422e-05,
+      "loss": 0.7806,
+      "step": 5007
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2694250115545336,
+      "learning_rate": 1.9596842737486945e-05,
+      "loss": 0.826,
+      "step": 5008
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4942999340250176,
+      "learning_rate": 1.9596628233793814e-05,
+      "loss": 0.8424,
+      "step": 5009
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5579760015222859,
+      "learning_rate": 1.9596413674226076e-05,
+      "loss": 0.7432,
+      "step": 5010
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3853475837468536,
+      "learning_rate": 1.959619905878498e-05,
+      "loss": 0.804,
+      "step": 5011
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8458960956013706,
+      "learning_rate": 1.9595984387471772e-05,
+      "loss": 0.6601,
+      "step": 5012
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6987403597170962,
+      "learning_rate": 1.959576966028771e-05,
+      "loss": 0.7245,
+      "step": 5013
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6237432789178591,
+      "learning_rate": 1.9595554877234036e-05,
+      "loss": 0.5819,
+      "step": 5014
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.523118875297416,
+      "learning_rate": 1.9595340038312e-05,
+      "loss": 0.7604,
+      "step": 5015
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.594864220329519,
+      "learning_rate": 1.959512514352286e-05,
+      "loss": 0.8045,
+      "step": 5016
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6708828026686897,
+      "learning_rate": 1.9594910192867866e-05,
+      "loss": 0.7257,
+      "step": 5017
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8693974209318858,
+      "learning_rate": 1.9594695186348265e-05,
+      "loss": 0.7307,
+      "step": 5018
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5136849381571493,
+      "learning_rate": 1.959448012396531e-05,
+      "loss": 0.7567,
+      "step": 5019
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0631224968733126,
+      "learning_rate": 1.9594265005720255e-05,
+      "loss": 0.7923,
+      "step": 5020
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5546664097156055,
+      "learning_rate": 1.9594049831614352e-05,
+      "loss": 0.8698,
+      "step": 5021
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6900751181592554,
+      "learning_rate": 1.9593834601648853e-05,
+      "loss": 0.5211,
+      "step": 5022
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5155360436307332,
+      "learning_rate": 1.9593619315825013e-05,
+      "loss": 0.7535,
+      "step": 5023
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4196240530685027,
+      "learning_rate": 1.9593403974144084e-05,
+      "loss": 0.809,
+      "step": 5024
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6102870991877906,
+      "learning_rate": 1.9593188576607316e-05,
+      "loss": 0.5516,
+      "step": 5025
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7740832490351748,
+      "learning_rate": 1.959297312321597e-05,
+      "loss": 0.7682,
+      "step": 5026
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4782930024696959,
+      "learning_rate": 1.95927576139713e-05,
+      "loss": 0.6632,
+      "step": 5027
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.655134840656706,
+      "learning_rate": 1.959254204887455e-05,
+      "loss": 0.7278,
+      "step": 5028
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7800377432933832,
+      "learning_rate": 1.959232642792699e-05,
+      "loss": 0.7861,
+      "step": 5029
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.564374550394426,
+      "learning_rate": 1.959211075112987e-05,
+      "loss": 0.7383,
+      "step": 5030
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5369319984782521,
+      "learning_rate": 1.959189501848444e-05,
+      "loss": 0.8049,
+      "step": 5031
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4995091031499515,
+      "learning_rate": 1.9591679229991963e-05,
+      "loss": 0.7525,
+      "step": 5032
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5149116874168904,
+      "learning_rate": 1.959146338565369e-05,
+      "loss": 0.8415,
+      "step": 5033
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4524941732286076,
+      "learning_rate": 1.959124748547088e-05,
+      "loss": 0.7799,
+      "step": 5034
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9174875202609674,
+      "learning_rate": 1.9591031529444795e-05,
+      "loss": 0.7933,
+      "step": 5035
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4246242180244997,
+      "learning_rate": 1.9590815517576684e-05,
+      "loss": 0.7984,
+      "step": 5036
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.514283921908883,
+      "learning_rate": 1.9590599449867813e-05,
+      "loss": 0.8041,
+      "step": 5037
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6043260580817925,
+      "learning_rate": 1.959038332631943e-05,
+      "loss": 0.5761,
+      "step": 5038
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4115798072292827,
+      "learning_rate": 1.9590167146932804e-05,
+      "loss": 0.8294,
+      "step": 5039
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6000598710062717,
+      "learning_rate": 1.958995091170919e-05,
+      "loss": 0.6645,
+      "step": 5040
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3892212395592392,
+      "learning_rate": 1.958973462064984e-05,
+      "loss": 0.6913,
+      "step": 5041
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5207996641579618,
+      "learning_rate": 1.9589518273756023e-05,
+      "loss": 0.8401,
+      "step": 5042
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4948776544831355,
+      "learning_rate": 1.9589301871028993e-05,
+      "loss": 0.6551,
+      "step": 5043
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3979436015890798,
+      "learning_rate": 1.958908541247001e-05,
+      "loss": 0.7708,
+      "step": 5044
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8740854745653162,
+      "learning_rate": 1.9588868898080337e-05,
+      "loss": 0.7842,
+      "step": 5045
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6100971946073364,
+      "learning_rate": 1.9588652327861236e-05,
+      "loss": 0.8316,
+      "step": 5046
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5247293224442557,
+      "learning_rate": 1.9588435701813962e-05,
+      "loss": 0.708,
+      "step": 5047
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8115336190934297,
+      "learning_rate": 1.958821901993978e-05,
+      "loss": 0.7379,
+      "step": 5048
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5431026092278444,
+      "learning_rate": 1.9588002282239955e-05,
+      "loss": 0.7052,
+      "step": 5049
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.643026945666942,
+      "learning_rate": 1.9587785488715742e-05,
+      "loss": 0.5447,
+      "step": 5050
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6268674479185266,
+      "learning_rate": 1.9587568639368405e-05,
+      "loss": 0.7522,
+      "step": 5051
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.425796843449192,
+      "learning_rate": 1.958735173419921e-05,
+      "loss": 0.8649,
+      "step": 5052
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0245669079391693,
+      "learning_rate": 1.958713477320942e-05,
+      "loss": 0.794,
+      "step": 5053
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6550153760061501,
+      "learning_rate": 1.9586917756400293e-05,
+      "loss": 0.728,
+      "step": 5054
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6771605053001943,
+      "learning_rate": 1.95867006837731e-05,
+      "loss": 0.7233,
+      "step": 5055
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8461715218852368,
+      "learning_rate": 1.9586483555329098e-05,
+      "loss": 0.7409,
+      "step": 5056
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5609409976290047,
+      "learning_rate": 1.958626637106955e-05,
+      "loss": 0.747,
+      "step": 5057
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6211398255983338,
+      "learning_rate": 1.9586049130995728e-05,
+      "loss": 0.5596,
+      "step": 5058
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8318513819703692,
+      "learning_rate": 1.958583183510889e-05,
+      "loss": 0.9322,
+      "step": 5059
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5880629774419608,
+      "learning_rate": 1.9585614483410308e-05,
+      "loss": 0.7292,
+      "step": 5060
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8637098422310308,
+      "learning_rate": 1.9585397075901242e-05,
+      "loss": 0.835,
+      "step": 5061
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7814606674627256,
+      "learning_rate": 1.958517961258296e-05,
+      "loss": 0.9119,
+      "step": 5062
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.473523423765484,
+      "learning_rate": 1.958496209345673e-05,
+      "loss": 0.7889,
+      "step": 5063
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6007461220495178,
+      "learning_rate": 1.9584744518523812e-05,
+      "loss": 0.6823,
+      "step": 5064
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4527465572223672,
+      "learning_rate": 1.958452688778548e-05,
+      "loss": 0.7188,
+      "step": 5065
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6151229432498162,
+      "learning_rate": 1.9584309201242993e-05,
+      "loss": 0.5328,
+      "step": 5066
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7322107910811988,
+      "learning_rate": 1.958409145889763e-05,
+      "loss": 0.8673,
+      "step": 5067
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.4784591625108803,
+      "learning_rate": 1.9583873660750646e-05,
+      "loss": 0.7699,
+      "step": 5068
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9421247396607773,
+      "learning_rate": 1.958365580680332e-05,
+      "loss": 0.8574,
+      "step": 5069
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.706597414029922,
+      "learning_rate": 1.9583437897056915e-05,
+      "loss": 0.7707,
+      "step": 5070
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0763303587529127,
+      "learning_rate": 1.95832199315127e-05,
+      "loss": 0.7551,
+      "step": 5071
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.948738185661906,
+      "learning_rate": 1.9583001910171947e-05,
+      "loss": 0.7156,
+      "step": 5072
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6480202674208306,
+      "learning_rate": 1.958278383303592e-05,
+      "loss": 0.7915,
+      "step": 5073
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4501312427644741,
+      "learning_rate": 1.9582565700105893e-05,
+      "loss": 0.7651,
+      "step": 5074
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6673742589402876,
+      "learning_rate": 1.958234751138313e-05,
+      "loss": 0.8693,
+      "step": 5075
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5517257310748882,
+      "learning_rate": 1.9582129266868914e-05,
+      "loss": 0.7947,
+      "step": 5076
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4893316886412844,
+      "learning_rate": 1.9581910966564502e-05,
+      "loss": 0.7757,
+      "step": 5077
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.504760026196628,
+      "learning_rate": 1.9581692610471174e-05,
+      "loss": 0.8605,
+      "step": 5078
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.362989652458071,
+      "learning_rate": 1.9581474198590196e-05,
+      "loss": 0.8074,
+      "step": 5079
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6050989518824315,
+      "learning_rate": 1.9581255730922842e-05,
+      "loss": 0.8337,
+      "step": 5080
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.620026611587605,
+      "learning_rate": 1.9581037207470382e-05,
+      "loss": 0.7906,
+      "step": 5081
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7480010072671526,
+      "learning_rate": 1.9580818628234095e-05,
+      "loss": 0.7755,
+      "step": 5082
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.264996095573398,
+      "learning_rate": 1.9580599993215243e-05,
+      "loss": 0.6499,
+      "step": 5083
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4519520803516544,
+      "learning_rate": 1.9580381302415107e-05,
+      "loss": 0.6916,
+      "step": 5084
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6381793633909447,
+      "learning_rate": 1.9580162555834962e-05,
+      "loss": 0.5612,
+      "step": 5085
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.583462577008334,
+      "learning_rate": 1.957994375347607e-05,
+      "loss": 0.4957,
+      "step": 5086
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8079692248714845,
+      "learning_rate": 1.957972489533972e-05,
+      "loss": 0.725,
+      "step": 5087
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7407571679167766,
+      "learning_rate": 1.9579505981427177e-05,
+      "loss": 0.8465,
+      "step": 5088
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.592651699314213,
+      "learning_rate": 1.9579287011739712e-05,
+      "loss": 0.81,
+      "step": 5089
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6202995134570385,
+      "learning_rate": 1.957906798627861e-05,
+      "loss": 0.5779,
+      "step": 5090
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4649767569715466,
+      "learning_rate": 1.957884890504514e-05,
+      "loss": 0.6383,
+      "step": 5091
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.904651307560281,
+      "learning_rate": 1.957862976804058e-05,
+      "loss": 0.7223,
+      "step": 5092
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.542099150374393,
+      "learning_rate": 1.9578410575266203e-05,
+      "loss": 0.7592,
+      "step": 5093
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3169038404844478,
+      "learning_rate": 1.9578191326723292e-05,
+      "loss": 0.7807,
+      "step": 5094
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5910494016874448,
+      "learning_rate": 1.9577972022413113e-05,
+      "loss": 0.6746,
+      "step": 5095
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3115354704708744,
+      "learning_rate": 1.9577752662336952e-05,
+      "loss": 0.6932,
+      "step": 5096
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6125421836778944,
+      "learning_rate": 1.957753324649608e-05,
+      "loss": 0.7213,
+      "step": 5097
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.2429314149296653,
+      "learning_rate": 1.957731377489178e-05,
+      "loss": 0.7651,
+      "step": 5098
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.476139740565439,
+      "learning_rate": 1.9577094247525324e-05,
+      "loss": 0.8096,
+      "step": 5099
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6081710237588267,
+      "learning_rate": 1.9576874664397998e-05,
+      "loss": 0.6847,
+      "step": 5100
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3391428875051408,
+      "learning_rate": 1.9576655025511072e-05,
+      "loss": 0.7537,
+      "step": 5101
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5818960546767256,
+      "learning_rate": 1.957643533086583e-05,
+      "loss": 0.5361,
+      "step": 5102
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.579373880501854,
+      "learning_rate": 1.957621558046355e-05,
+      "loss": 0.8827,
+      "step": 5103
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5741278162943624,
+      "learning_rate": 1.9575995774305506e-05,
+      "loss": 0.5245,
+      "step": 5104
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5885668845065476,
+      "learning_rate": 1.9575775912392992e-05,
+      "loss": 0.5214,
+      "step": 5105
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.303247459710526,
+      "learning_rate": 1.9575555994727273e-05,
+      "loss": 0.7342,
+      "step": 5106
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.479314299392172,
+      "learning_rate": 1.9575336021309635e-05,
+      "loss": 0.752,
+      "step": 5107
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.500250755280733,
+      "learning_rate": 1.957511599214136e-05,
+      "loss": 0.8098,
+      "step": 5108
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5757000755127815,
+      "learning_rate": 1.957489590722373e-05,
+      "loss": 0.5408,
+      "step": 5109
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4827321712513202,
+      "learning_rate": 1.9574675766558023e-05,
+      "loss": 0.7702,
+      "step": 5110
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6685532304780206,
+      "learning_rate": 1.9574455570145524e-05,
+      "loss": 0.7724,
+      "step": 5111
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7580231344250032,
+      "learning_rate": 1.9574235317987514e-05,
+      "loss": 0.7174,
+      "step": 5112
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4167917078627716,
+      "learning_rate": 1.9574015010085273e-05,
+      "loss": 0.6624,
+      "step": 5113
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9011291652394589,
+      "learning_rate": 1.9573794646440084e-05,
+      "loss": 0.7876,
+      "step": 5114
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.016790384097132,
+      "learning_rate": 1.9573574227053233e-05,
+      "loss": 0.6915,
+      "step": 5115
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9799793200189666,
+      "learning_rate": 1.9573353751926004e-05,
+      "loss": 0.7799,
+      "step": 5116
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2869544630591543,
+      "learning_rate": 1.9573133221059676e-05,
+      "loss": 0.6761,
+      "step": 5117
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.269163423275525,
+      "learning_rate": 1.9572912634455538e-05,
+      "loss": 0.8193,
+      "step": 5118
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8134244199794574,
+      "learning_rate": 1.9572691992114867e-05,
+      "loss": 0.7805,
+      "step": 5119
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4984625683672783,
+      "learning_rate": 1.9572471294038956e-05,
+      "loss": 0.8203,
+      "step": 5120
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.938827960956828,
+      "learning_rate": 1.9572250540229088e-05,
+      "loss": 0.7165,
+      "step": 5121
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5701938644552664,
+      "learning_rate": 1.9572029730686546e-05,
+      "loss": 0.7801,
+      "step": 5122
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7043727634359827,
+      "learning_rate": 1.9571808865412617e-05,
+      "loss": 0.6432,
+      "step": 5123
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.445671475570802,
+      "learning_rate": 1.9571587944408582e-05,
+      "loss": 0.7136,
+      "step": 5124
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5474175895326834,
+      "learning_rate": 1.957136696767574e-05,
+      "loss": 0.796,
+      "step": 5125
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9690246501845214,
+      "learning_rate": 1.9571145935215358e-05,
+      "loss": 0.8503,
+      "step": 5126
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.502408095330066,
+      "learning_rate": 1.9570924847028744e-05,
+      "loss": 0.8493,
+      "step": 5127
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3711419195454146,
+      "learning_rate": 1.957070370311717e-05,
+      "loss": 0.7613,
+      "step": 5128
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.197048760268092,
+      "learning_rate": 1.9570482503481933e-05,
+      "loss": 0.6266,
+      "step": 5129
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.494935839083547,
+      "learning_rate": 1.957026124812431e-05,
+      "loss": 0.8464,
+      "step": 5130
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5180183038589719,
+      "learning_rate": 1.9570039937045604e-05,
+      "loss": 0.7997,
+      "step": 5131
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6770716794432923,
+      "learning_rate": 1.9569818570247093e-05,
+      "loss": 0.69,
+      "step": 5132
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4939347270788985,
+      "learning_rate": 1.9569597147730066e-05,
+      "loss": 0.8665,
+      "step": 5133
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5549620024517223,
+      "learning_rate": 1.9569375669495815e-05,
+      "loss": 0.8843,
+      "step": 5134
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8290916315054557,
+      "learning_rate": 1.9569154135545634e-05,
+      "loss": 0.679,
+      "step": 5135
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8135043145729584,
+      "learning_rate": 1.9568932545880804e-05,
+      "loss": 0.8561,
+      "step": 5136
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5337946655058794,
+      "learning_rate": 1.9568710900502622e-05,
+      "loss": 0.7899,
+      "step": 5137
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6600234841629764,
+      "learning_rate": 1.9568489199412373e-05,
+      "loss": 0.7922,
+      "step": 5138
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.805188274828821,
+      "learning_rate": 1.956826744261135e-05,
+      "loss": 0.8047,
+      "step": 5139
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6916772594688073,
+      "learning_rate": 1.9568045630100846e-05,
+      "loss": 0.6333,
+      "step": 5140
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6700775007827615,
+      "learning_rate": 1.9567823761882154e-05,
+      "loss": 0.6199,
+      "step": 5141
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6098114782947378,
+      "learning_rate": 1.9567601837956557e-05,
+      "loss": 0.54,
+      "step": 5142
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.463132793747267,
+      "learning_rate": 1.9567379858325358e-05,
+      "loss": 0.7418,
+      "step": 5143
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4750135934103763,
+      "learning_rate": 1.9567157822989843e-05,
+      "loss": 0.7593,
+      "step": 5144
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6187406444540267,
+      "learning_rate": 1.9566935731951307e-05,
+      "loss": 0.8065,
+      "step": 5145
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 14.424047258363178,
+      "learning_rate": 1.956671358521104e-05,
+      "loss": 0.7632,
+      "step": 5146
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5686323095436012,
+      "learning_rate": 1.9566491382770342e-05,
+      "loss": 0.9567,
+      "step": 5147
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5893153087937508,
+      "learning_rate": 1.9566269124630498e-05,
+      "loss": 0.7345,
+      "step": 5148
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1906428725100273,
+      "learning_rate": 1.9566046810792806e-05,
+      "loss": 0.68,
+      "step": 5149
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8534694588690641,
+      "learning_rate": 1.9565824441258562e-05,
+      "loss": 0.8363,
+      "step": 5150
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4568357434360857,
+      "learning_rate": 1.9565602016029062e-05,
+      "loss": 0.7691,
+      "step": 5151
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6169578846931886,
+      "learning_rate": 1.95653795351056e-05,
+      "loss": 0.5511,
+      "step": 5152
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8036076344967542,
+      "learning_rate": 1.9565156998489465e-05,
+      "loss": 0.8961,
+      "step": 5153
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4961806456016626,
+      "learning_rate": 1.956493440618196e-05,
+      "loss": 0.8471,
+      "step": 5154
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.534173937959387,
+      "learning_rate": 1.9564711758184378e-05,
+      "loss": 0.6539,
+      "step": 5155
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5314198429110657,
+      "learning_rate": 1.956448905449802e-05,
+      "loss": 0.857,
+      "step": 5156
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.598417438948261,
+      "learning_rate": 1.9564266295124174e-05,
+      "loss": 0.5738,
+      "step": 5157
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6442197957947946,
+      "learning_rate": 1.956404348006414e-05,
+      "loss": 0.713,
+      "step": 5158
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4897070148030158,
+      "learning_rate": 1.956382060931922e-05,
+      "loss": 0.7784,
+      "step": 5159
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5957871086077686,
+      "learning_rate": 1.956359768289071e-05,
+      "loss": 0.5373,
+      "step": 5160
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8261587859404387,
+      "learning_rate": 1.9563374700779904e-05,
+      "loss": 0.7988,
+      "step": 5161
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7182340358280992,
+      "learning_rate": 1.9563151662988106e-05,
+      "loss": 0.7512,
+      "step": 5162
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.946940546206993,
+      "learning_rate": 1.9562928569516607e-05,
+      "loss": 0.6992,
+      "step": 5163
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.414886768910733,
+      "learning_rate": 1.9562705420366712e-05,
+      "loss": 0.8414,
+      "step": 5164
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7262476170746446,
+      "learning_rate": 1.9562482215539722e-05,
+      "loss": 0.9227,
+      "step": 5165
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.836800080876893,
+      "learning_rate": 1.9562258955036928e-05,
+      "loss": 0.6674,
+      "step": 5166
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.643090424585294,
+      "learning_rate": 1.9562035638859638e-05,
+      "loss": 0.8859,
+      "step": 5167
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8543949466816778,
+      "learning_rate": 1.9561812267009148e-05,
+      "loss": 0.6978,
+      "step": 5168
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8665399469568065,
+      "learning_rate": 1.956158883948676e-05,
+      "loss": 0.8721,
+      "step": 5169
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1917643449575164,
+      "learning_rate": 1.9561365356293774e-05,
+      "loss": 0.6071,
+      "step": 5170
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3527421881179582,
+      "learning_rate": 1.9561141817431494e-05,
+      "loss": 0.8683,
+      "step": 5171
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9325739507512483,
+      "learning_rate": 1.9560918222901216e-05,
+      "loss": 0.7037,
+      "step": 5172
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2492654072804785,
+      "learning_rate": 1.9560694572704245e-05,
+      "loss": 0.8736,
+      "step": 5173
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9856630416369405,
+      "learning_rate": 1.9560470866841888e-05,
+      "loss": 0.7916,
+      "step": 5174
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.609026927647152,
+      "learning_rate": 1.9560247105315443e-05,
+      "loss": 0.8458,
+      "step": 5175
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6583115733158975,
+      "learning_rate": 1.9560023288126208e-05,
+      "loss": 0.6835,
+      "step": 5176
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.566090175484253,
+      "learning_rate": 1.955979941527549e-05,
+      "loss": 0.7846,
+      "step": 5177
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0619606799157837,
+      "learning_rate": 1.9559575486764596e-05,
+      "loss": 0.8034,
+      "step": 5178
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7286437483659267,
+      "learning_rate": 1.9559351502594826e-05,
+      "loss": 0.7681,
+      "step": 5179
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3726929375607375,
+      "learning_rate": 1.9559127462767487e-05,
+      "loss": 0.637,
+      "step": 5180
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6194901055040607,
+      "learning_rate": 1.9558903367283882e-05,
+      "loss": 0.5893,
+      "step": 5181
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6991075319119266,
+      "learning_rate": 1.9558679216145314e-05,
+      "loss": 0.83,
+      "step": 5182
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.374409817518744,
+      "learning_rate": 1.9558455009353086e-05,
+      "loss": 0.8265,
+      "step": 5183
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.285672968899873,
+      "learning_rate": 1.955823074690851e-05,
+      "loss": 0.7626,
+      "step": 5184
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8261301111813948,
+      "learning_rate": 1.955800642881289e-05,
+      "loss": 0.7295,
+      "step": 5185
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6074052992922969,
+      "learning_rate": 1.9557782055067528e-05,
+      "loss": 0.5158,
+      "step": 5186
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7045665876807397,
+      "learning_rate": 1.9557557625673734e-05,
+      "loss": 0.7775,
+      "step": 5187
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8550275262005995,
+      "learning_rate": 1.9557333140632816e-05,
+      "loss": 0.7765,
+      "step": 5188
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6339676392390637,
+      "learning_rate": 1.9557108599946076e-05,
+      "loss": 0.5996,
+      "step": 5189
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.516807764337316,
+      "learning_rate": 1.9556884003614824e-05,
+      "loss": 0.769,
+      "step": 5190
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4214690010821545,
+      "learning_rate": 1.9556659351640368e-05,
+      "loss": 0.9039,
+      "step": 5191
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6773362534162237,
+      "learning_rate": 1.9556434644024013e-05,
+      "loss": 0.7104,
+      "step": 5192
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.038706165132011,
+      "learning_rate": 1.9556209880767073e-05,
+      "loss": 0.8321,
+      "step": 5193
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4326488774195587,
+      "learning_rate": 1.9555985061870854e-05,
+      "loss": 0.799,
+      "step": 5194
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8065013059819448,
+      "learning_rate": 1.9555760187336665e-05,
+      "loss": 0.8823,
+      "step": 5195
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4546961086674504,
+      "learning_rate": 1.955553525716581e-05,
+      "loss": 0.69,
+      "step": 5196
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6383924938044798,
+      "learning_rate": 1.9555310271359608e-05,
+      "loss": 0.7673,
+      "step": 5197
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7287976240168454,
+      "learning_rate": 1.9555085229919366e-05,
+      "loss": 0.798,
+      "step": 5198
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5530062908599735,
+      "learning_rate": 1.955486013284639e-05,
+      "loss": 0.7993,
+      "step": 5199
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7343614168252068,
+      "learning_rate": 1.9554634980141993e-05,
+      "loss": 0.6922,
+      "step": 5200
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.777131879850625,
+      "learning_rate": 1.955440977180749e-05,
+      "loss": 0.8641,
+      "step": 5201
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2148143180531763,
+      "learning_rate": 1.9554184507844183e-05,
+      "loss": 0.6466,
+      "step": 5202
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4675783253236299,
+      "learning_rate": 1.9553959188253388e-05,
+      "loss": 0.8704,
+      "step": 5203
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.602933624995174,
+      "learning_rate": 1.9553733813036425e-05,
+      "loss": 0.5693,
+      "step": 5204
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5691796374208878,
+      "learning_rate": 1.955350838219459e-05,
+      "loss": 0.5251,
+      "step": 5205
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.561516467394073,
+      "learning_rate": 1.9553282895729212e-05,
+      "loss": 0.6745,
+      "step": 5206
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5628962099598239,
+      "learning_rate": 1.9553057353641595e-05,
+      "loss": 0.5393,
+      "step": 5207
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7720786616234487,
+      "learning_rate": 1.955283175593305e-05,
+      "loss": 0.7482,
+      "step": 5208
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7493929592766968,
+      "learning_rate": 1.95526061026049e-05,
+      "loss": 0.7804,
+      "step": 5209
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5725372097477868,
+      "learning_rate": 1.955238039365845e-05,
+      "loss": 0.8195,
+      "step": 5210
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.703939584115375,
+      "learning_rate": 1.9552154629095015e-05,
+      "loss": 0.8323,
+      "step": 5211
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6358215810062218,
+      "learning_rate": 1.9551928808915914e-05,
+      "loss": 0.7782,
+      "step": 5212
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5933737855606882,
+      "learning_rate": 1.9551702933122462e-05,
+      "loss": 0.7375,
+      "step": 5213
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7014086233073007,
+      "learning_rate": 1.9551477001715965e-05,
+      "loss": 0.6573,
+      "step": 5214
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.448347946895799,
+      "learning_rate": 1.955125101469775e-05,
+      "loss": 0.8528,
+      "step": 5215
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5947032474512913,
+      "learning_rate": 1.9551024972069127e-05,
+      "loss": 0.52,
+      "step": 5216
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5105218786124068,
+      "learning_rate": 1.955079887383141e-05,
+      "loss": 0.7033,
+      "step": 5217
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5263688240336897,
+      "learning_rate": 1.9550572719985922e-05,
+      "loss": 0.8601,
+      "step": 5218
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5886357760743995,
+      "learning_rate": 1.955034651053397e-05,
+      "loss": 0.5439,
+      "step": 5219
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6073665538039514,
+      "learning_rate": 1.9550120245476886e-05,
+      "loss": 0.8133,
+      "step": 5220
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2088576173265337,
+      "learning_rate": 1.9549893924815974e-05,
+      "loss": 0.768,
+      "step": 5221
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.95215218087086,
+      "learning_rate": 1.9549667548552557e-05,
+      "loss": 0.6924,
+      "step": 5222
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4605981712123108,
+      "learning_rate": 1.9549441116687945e-05,
+      "loss": 0.7947,
+      "step": 5223
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5407296826011745,
+      "learning_rate": 1.954921462922347e-05,
+      "loss": 0.7434,
+      "step": 5224
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4115991217523924,
+      "learning_rate": 1.9548988086160445e-05,
+      "loss": 0.7472,
+      "step": 5225
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7765736832349549,
+      "learning_rate": 1.954876148750019e-05,
+      "loss": 0.5586,
+      "step": 5226
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4677132101727366,
+      "learning_rate": 1.9548534833244017e-05,
+      "loss": 0.6697,
+      "step": 5227
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7392967658912353,
+      "learning_rate": 1.9548308123393255e-05,
+      "loss": 0.7668,
+      "step": 5228
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.8425355257096947,
+      "learning_rate": 1.9548081357949217e-05,
+      "loss": 0.8301,
+      "step": 5229
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6089304970695337,
+      "learning_rate": 1.954785453691323e-05,
+      "loss": 0.7278,
+      "step": 5230
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4920521456095026,
+      "learning_rate": 1.9547627660286607e-05,
+      "loss": 0.8371,
+      "step": 5231
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.385842713933913,
+      "learning_rate": 1.9547400728070675e-05,
+      "loss": 0.6574,
+      "step": 5232
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6304196686786763,
+      "learning_rate": 1.9547173740266753e-05,
+      "loss": 0.7361,
+      "step": 5233
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8580432708110124,
+      "learning_rate": 1.9546946696876164e-05,
+      "loss": 0.8198,
+      "step": 5234
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3730729549744827,
+      "learning_rate": 1.954671959790023e-05,
+      "loss": 0.7533,
+      "step": 5235
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.37184029107025,
+      "learning_rate": 1.9546492443340268e-05,
+      "loss": 0.8121,
+      "step": 5236
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4079243575005467,
+      "learning_rate": 1.9546265233197607e-05,
+      "loss": 0.8786,
+      "step": 5237
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5547654107721582,
+      "learning_rate": 1.954603796747357e-05,
+      "loss": 0.8019,
+      "step": 5238
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.688027523874707,
+      "learning_rate": 1.9545810646169474e-05,
+      "loss": 0.6683,
+      "step": 5239
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7153368341679063,
+      "learning_rate": 1.9545583269286646e-05,
+      "loss": 0.8048,
+      "step": 5240
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4213023166608898,
+      "learning_rate": 1.9545355836826415e-05,
+      "loss": 0.8267,
+      "step": 5241
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.8173206846489642,
+      "learning_rate": 1.9545128348790096e-05,
+      "loss": 0.8691,
+      "step": 5242
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8163752395616268,
+      "learning_rate": 1.9544900805179016e-05,
+      "loss": 0.7807,
+      "step": 5243
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.52864285856344,
+      "learning_rate": 1.9544673205994505e-05,
+      "loss": 0.829,
+      "step": 5244
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.51762576809964,
+      "learning_rate": 1.9544445551237883e-05,
+      "loss": 0.7852,
+      "step": 5245
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5153485525065156,
+      "learning_rate": 1.9544217840910483e-05,
+      "loss": 0.8308,
+      "step": 5246
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.725754270245829,
+      "learning_rate": 1.9543990075013618e-05,
+      "loss": 0.7481,
+      "step": 5247
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5752523274772976,
+      "learning_rate": 1.9543762253548623e-05,
+      "loss": 0.7055,
+      "step": 5248
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2888384596229523,
+      "learning_rate": 1.9543534376516824e-05,
+      "loss": 0.7425,
+      "step": 5249
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.281482823744628,
+      "learning_rate": 1.954330644391955e-05,
+      "loss": 0.8113,
+      "step": 5250
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5261873211260868,
+      "learning_rate": 1.9543078455758118e-05,
+      "loss": 0.7905,
+      "step": 5251
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.527012356297822,
+      "learning_rate": 1.9542850412033863e-05,
+      "loss": 0.783,
+      "step": 5252
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5229804915852534,
+      "learning_rate": 1.9542622312748113e-05,
+      "loss": 0.847,
+      "step": 5253
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.569562867697064,
+      "learning_rate": 1.9542394157902196e-05,
+      "loss": 0.7459,
+      "step": 5254
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5291915017191866,
+      "learning_rate": 1.9542165947497436e-05,
+      "loss": 0.7136,
+      "step": 5255
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3570167681635938,
+      "learning_rate": 1.954193768153517e-05,
+      "loss": 0.7759,
+      "step": 5256
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.468924181828605,
+      "learning_rate": 1.9541709360016715e-05,
+      "loss": 0.7758,
+      "step": 5257
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9441348509821812,
+      "learning_rate": 1.9541480982943413e-05,
+      "loss": 0.8466,
+      "step": 5258
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.48826007784154,
+      "learning_rate": 1.9541252550316585e-05,
+      "loss": 0.7363,
+      "step": 5259
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7609557767816175,
+      "learning_rate": 1.9541024062137565e-05,
+      "loss": 0.8151,
+      "step": 5260
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.550310140114777,
+      "learning_rate": 1.954079551840768e-05,
+      "loss": 0.7982,
+      "step": 5261
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4806634241349816,
+      "learning_rate": 1.954056691912827e-05,
+      "loss": 0.7594,
+      "step": 5262
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6339803947450415,
+      "learning_rate": 1.954033826430065e-05,
+      "loss": 0.5668,
+      "step": 5263
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6008458781737028,
+      "learning_rate": 1.9540109553926165e-05,
+      "loss": 0.5615,
+      "step": 5264
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6479322610088574,
+      "learning_rate": 1.953988078800614e-05,
+      "loss": 0.8066,
+      "step": 5265
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3653095490114349,
+      "learning_rate": 1.953965196654191e-05,
+      "loss": 0.7538,
+      "step": 5266
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9838310874205425,
+      "learning_rate": 1.9539423089534805e-05,
+      "loss": 0.7559,
+      "step": 5267
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9439082893831818,
+      "learning_rate": 1.953919415698616e-05,
+      "loss": 0.8264,
+      "step": 5268
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.573727911268515,
+      "learning_rate": 1.9538965168897307e-05,
+      "loss": 0.7173,
+      "step": 5269
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.681298509191911,
+      "learning_rate": 1.9538736125269578e-05,
+      "loss": 0.7812,
+      "step": 5270
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8663001591381954,
+      "learning_rate": 1.9538507026104304e-05,
+      "loss": 0.7491,
+      "step": 5271
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5967660116694748,
+      "learning_rate": 1.953827787140283e-05,
+      "loss": 0.5084,
+      "step": 5272
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4890533025042607,
+      "learning_rate": 1.9538048661166474e-05,
+      "loss": 0.6998,
+      "step": 5273
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7242140432109843,
+      "learning_rate": 1.9537819395396586e-05,
+      "loss": 0.7235,
+      "step": 5274
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5644076649535698,
+      "learning_rate": 1.9537590074094492e-05,
+      "loss": 0.5463,
+      "step": 5275
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3756518084356857,
+      "learning_rate": 1.9537360697261524e-05,
+      "loss": 0.8108,
+      "step": 5276
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4609553914537217,
+      "learning_rate": 1.953713126489903e-05,
+      "loss": 0.7635,
+      "step": 5277
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5068815523929422,
+      "learning_rate": 1.9536901777008332e-05,
+      "loss": 0.8175,
+      "step": 5278
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0766567119521726,
+      "learning_rate": 1.9536672233590774e-05,
+      "loss": 0.904,
+      "step": 5279
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3235821630086675,
+      "learning_rate": 1.9536442634647696e-05,
+      "loss": 0.7825,
+      "step": 5280
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6720829453758073,
+      "learning_rate": 1.9536212980180422e-05,
+      "loss": 0.6816,
+      "step": 5281
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.717754029860496,
+      "learning_rate": 1.95359832701903e-05,
+      "loss": 0.962,
+      "step": 5282
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8530274550905306,
+      "learning_rate": 1.9535753504678667e-05,
+      "loss": 0.7806,
+      "step": 5283
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6623846054723932,
+      "learning_rate": 1.9535523683646857e-05,
+      "loss": 0.7509,
+      "step": 5284
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.400623468623949,
+      "learning_rate": 1.9535293807096208e-05,
+      "loss": 0.6875,
+      "step": 5285
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5003157213507052,
+      "learning_rate": 1.953506387502806e-05,
+      "loss": 0.7158,
+      "step": 5286
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.372947707338771,
+      "learning_rate": 1.9534833887443752e-05,
+      "loss": 0.6628,
+      "step": 5287
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1587705516646034,
+      "learning_rate": 1.9534603844344623e-05,
+      "loss": 0.848,
+      "step": 5288
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8568795221931247,
+      "learning_rate": 1.9534373745732008e-05,
+      "loss": 0.7275,
+      "step": 5289
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.71785952783923,
+      "learning_rate": 1.9534143591607256e-05,
+      "loss": 0.8575,
+      "step": 5290
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6434373559032763,
+      "learning_rate": 1.9533913381971698e-05,
+      "loss": 0.5313,
+      "step": 5291
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3878197588280772,
+      "learning_rate": 1.953368311682668e-05,
+      "loss": 0.7164,
+      "step": 5292
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6027499474751087,
+      "learning_rate": 1.953345279617354e-05,
+      "loss": 0.5536,
+      "step": 5293
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3887511927685017,
+      "learning_rate": 1.9533222420013618e-05,
+      "loss": 0.7402,
+      "step": 5294
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8895226284219055,
+      "learning_rate": 1.9532991988348254e-05,
+      "loss": 0.8353,
+      "step": 5295
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5434526869967817,
+      "learning_rate": 1.9532761501178797e-05,
+      "loss": 0.7767,
+      "step": 5296
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7826119648461256,
+      "learning_rate": 1.9532530958506585e-05,
+      "loss": 0.8337,
+      "step": 5297
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5061505919304816,
+      "learning_rate": 1.953230036033296e-05,
+      "loss": 0.7656,
+      "step": 5298
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6277128010606299,
+      "learning_rate": 1.9532069706659263e-05,
+      "loss": 0.7549,
+      "step": 5299
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3864162506339666,
+      "learning_rate": 1.9531838997486838e-05,
+      "loss": 0.7815,
+      "step": 5300
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.51790008524521,
+      "learning_rate": 1.953160823281703e-05,
+      "loss": 0.9085,
+      "step": 5301
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4114077543268155,
+      "learning_rate": 1.953137741265118e-05,
+      "loss": 0.8046,
+      "step": 5302
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4487578925031275,
+      "learning_rate": 1.9531146536990634e-05,
+      "loss": 0.7118,
+      "step": 5303
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3545239095735153,
+      "learning_rate": 1.9530915605836735e-05,
+      "loss": 0.7285,
+      "step": 5304
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5035591929901218,
+      "learning_rate": 1.9530684619190827e-05,
+      "loss": 0.6661,
+      "step": 5305
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.693899736962172,
+      "learning_rate": 1.9530453577054257e-05,
+      "loss": 0.8844,
+      "step": 5306
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4921260930113658,
+      "learning_rate": 1.9530222479428368e-05,
+      "loss": 0.693,
+      "step": 5307
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5991273459647164,
+      "learning_rate": 1.9529991326314506e-05,
+      "loss": 0.8212,
+      "step": 5308
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5481622442790124,
+      "learning_rate": 1.952976011771402e-05,
+      "loss": 0.648,
+      "step": 5309
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5551761883487347,
+      "learning_rate": 1.9529528853628254e-05,
+      "loss": 0.8158,
+      "step": 5310
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.581224408700279,
+      "learning_rate": 1.952929753405855e-05,
+      "loss": 0.8394,
+      "step": 5311
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.131019279636934,
+      "learning_rate": 1.952906615900626e-05,
+      "loss": 0.5552,
+      "step": 5312
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.57712382347156,
+      "learning_rate": 1.952883472847273e-05,
+      "loss": 0.4905,
+      "step": 5313
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.810013438720967,
+      "learning_rate": 1.9528603242459308e-05,
+      "loss": 0.7517,
+      "step": 5314
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4958443773874537,
+      "learning_rate": 1.952837170096734e-05,
+      "loss": 0.789,
+      "step": 5315
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3003684218555671,
+      "learning_rate": 1.952814010399818e-05,
+      "loss": 0.766,
+      "step": 5316
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5947150946875377,
+      "learning_rate": 1.9527908451553168e-05,
+      "loss": 0.7552,
+      "step": 5317
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.39578040185991,
+      "learning_rate": 1.9527676743633654e-05,
+      "loss": 0.8356,
+      "step": 5318
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5506793650118376,
+      "learning_rate": 1.9527444980240993e-05,
+      "loss": 0.7739,
+      "step": 5319
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.683623400620894,
+      "learning_rate": 1.952721316137653e-05,
+      "loss": 0.7625,
+      "step": 5320
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9264169664974988,
+      "learning_rate": 1.952698128704162e-05,
+      "loss": 0.9316,
+      "step": 5321
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4783437741919283,
+      "learning_rate": 1.9526749357237604e-05,
+      "loss": 0.6591,
+      "step": 5322
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7144779174087232,
+      "learning_rate": 1.952651737196584e-05,
+      "loss": 0.7364,
+      "step": 5323
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9131808441024887,
+      "learning_rate": 1.9526285331227672e-05,
+      "loss": 0.7696,
+      "step": 5324
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.416379802494534,
+      "learning_rate": 1.9526053235024457e-05,
+      "loss": 0.8068,
+      "step": 5325
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.710655006329159,
+      "learning_rate": 1.9525821083357545e-05,
+      "loss": 0.8203,
+      "step": 5326
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4565513851592864,
+      "learning_rate": 1.9525588876228284e-05,
+      "loss": 0.823,
+      "step": 5327
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4193009881556833,
+      "learning_rate": 1.952535661363803e-05,
+      "loss": 0.7522,
+      "step": 5328
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4368877685226866,
+      "learning_rate": 1.9525124295588135e-05,
+      "loss": 0.8511,
+      "step": 5329
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5191123927704484,
+      "learning_rate": 1.952489192207995e-05,
+      "loss": 0.7938,
+      "step": 5330
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8362009448599657,
+      "learning_rate": 1.952465949311483e-05,
+      "loss": 0.758,
+      "step": 5331
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.416027874739459,
+      "learning_rate": 1.9524427008694125e-05,
+      "loss": 0.5489,
+      "step": 5332
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4870814323583055,
+      "learning_rate": 1.952419446881919e-05,
+      "loss": 0.807,
+      "step": 5333
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0836120340742195,
+      "learning_rate": 1.9523961873491385e-05,
+      "loss": 0.9105,
+      "step": 5334
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4586168166148055,
+      "learning_rate": 1.9523729222712053e-05,
+      "loss": 0.7016,
+      "step": 5335
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3742174091794768,
+      "learning_rate": 1.9523496516482557e-05,
+      "loss": 0.8263,
+      "step": 5336
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.958984352958084,
+      "learning_rate": 1.9523263754804248e-05,
+      "loss": 0.6978,
+      "step": 5337
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6529708564528345,
+      "learning_rate": 1.9523030937678484e-05,
+      "loss": 0.7185,
+      "step": 5338
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7123635129996524,
+      "learning_rate": 1.9522798065106615e-05,
+      "loss": 0.7492,
+      "step": 5339
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.583517423654124,
+      "learning_rate": 1.9522565137090005e-05,
+      "loss": 0.9533,
+      "step": 5340
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.989214239330313,
+      "learning_rate": 1.9522332153630006e-05,
+      "loss": 0.7689,
+      "step": 5341
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.730517307233212,
+      "learning_rate": 1.952209911472797e-05,
+      "loss": 0.7456,
+      "step": 5342
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5523674275369586,
+      "learning_rate": 1.952186602038526e-05,
+      "loss": 0.7255,
+      "step": 5343
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5574799121054256,
+      "learning_rate": 1.9521632870603235e-05,
+      "loss": 0.5004,
+      "step": 5344
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9689573731953878,
+      "learning_rate": 1.952139966538325e-05,
+      "loss": 0.798,
+      "step": 5345
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7688079913795773,
+      "learning_rate": 1.9521166404726656e-05,
+      "loss": 0.6263,
+      "step": 5346
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6122439163368302,
+      "learning_rate": 1.952093308863482e-05,
+      "loss": 0.5783,
+      "step": 5347
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4596400945083035,
+      "learning_rate": 1.9520699717109096e-05,
+      "loss": 0.8007,
+      "step": 5348
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.747635367850568,
+      "learning_rate": 1.9520466290150842e-05,
+      "loss": 0.8345,
+      "step": 5349
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5136269788178918,
+      "learning_rate": 1.9520232807761424e-05,
+      "loss": 0.7956,
+      "step": 5350
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5470680834615145,
+      "learning_rate": 1.9519999269942194e-05,
+      "loss": 0.7322,
+      "step": 5351
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8086152655878653,
+      "learning_rate": 1.9519765676694513e-05,
+      "loss": 0.7981,
+      "step": 5352
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5836995276283128,
+      "learning_rate": 1.9519532028019744e-05,
+      "loss": 0.9605,
+      "step": 5353
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7132056612200353,
+      "learning_rate": 1.9519298323919245e-05,
+      "loss": 0.6203,
+      "step": 5354
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5948116329194727,
+      "learning_rate": 1.9519064564394383e-05,
+      "loss": 0.5315,
+      "step": 5355
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4999330618597626,
+      "learning_rate": 1.9518830749446507e-05,
+      "loss": 0.7287,
+      "step": 5356
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.569689972207522,
+      "learning_rate": 1.9518596879076982e-05,
+      "loss": 0.9073,
+      "step": 5357
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5412511092274774,
+      "learning_rate": 1.9518362953287178e-05,
+      "loss": 0.6757,
+      "step": 5358
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8188109965758263,
+      "learning_rate": 1.9518128972078448e-05,
+      "loss": 0.7291,
+      "step": 5359
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.773353432351658,
+      "learning_rate": 1.951789493545216e-05,
+      "loss": 0.6193,
+      "step": 5360
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5951882102129242,
+      "learning_rate": 1.9517660843409678e-05,
+      "loss": 0.8107,
+      "step": 5361
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5528875366506747,
+      "learning_rate": 1.9517426695952358e-05,
+      "loss": 0.8531,
+      "step": 5362
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5741632157172025,
+      "learning_rate": 1.9517192493081563e-05,
+      "loss": 0.5441,
+      "step": 5363
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8621092304871765,
+      "learning_rate": 1.9516958234798663e-05,
+      "loss": 0.8435,
+      "step": 5364
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5032844009455355,
+      "learning_rate": 1.951672392110502e-05,
+      "loss": 0.8345,
+      "step": 5365
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6106004745975933,
+      "learning_rate": 1.9516489552002e-05,
+      "loss": 0.798,
+      "step": 5366
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7127876809130065,
+      "learning_rate": 1.951625512749096e-05,
+      "loss": 0.7995,
+      "step": 5367
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6636267743897144,
+      "learning_rate": 1.9516020647573272e-05,
+      "loss": 0.8443,
+      "step": 5368
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3695035806242832,
+      "learning_rate": 1.95157861122503e-05,
+      "loss": 0.6351,
+      "step": 5369
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4873394857364937,
+      "learning_rate": 1.9515551521523406e-05,
+      "loss": 0.6973,
+      "step": 5370
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2724964777883963,
+      "learning_rate": 1.9515316875393958e-05,
+      "loss": 0.8312,
+      "step": 5371
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7690551608767366,
+      "learning_rate": 1.9515082173863327e-05,
+      "loss": 0.9261,
+      "step": 5372
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5828091671335377,
+      "learning_rate": 1.9514847416932873e-05,
+      "loss": 0.5199,
+      "step": 5373
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5621532182486813,
+      "learning_rate": 1.9514612604603963e-05,
+      "loss": 0.8182,
+      "step": 5374
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7290420575990657,
+      "learning_rate": 1.9514377736877967e-05,
+      "loss": 0.7423,
+      "step": 5375
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4923619554838499,
+      "learning_rate": 1.9514142813756253e-05,
+      "loss": 0.8319,
+      "step": 5376
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4191653597346119,
+      "learning_rate": 1.9513907835240187e-05,
+      "loss": 0.7208,
+      "step": 5377
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4556174776768833,
+      "learning_rate": 1.9513672801331135e-05,
+      "loss": 0.6684,
+      "step": 5378
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4640910002307628,
+      "learning_rate": 1.951343771203047e-05,
+      "loss": 0.6117,
+      "step": 5379
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.623292338392181,
+      "learning_rate": 1.951320256733956e-05,
+      "loss": 0.7255,
+      "step": 5380
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.904230949776814,
+      "learning_rate": 1.951296736725977e-05,
+      "loss": 0.7242,
+      "step": 5381
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5499815051084007,
+      "learning_rate": 1.9512732111792477e-05,
+      "loss": 0.7939,
+      "step": 5382
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5128150815125927,
+      "learning_rate": 1.9512496800939043e-05,
+      "loss": 0.7552,
+      "step": 5383
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.673750875997401,
+      "learning_rate": 1.9512261434700842e-05,
+      "loss": 0.5436,
+      "step": 5384
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3682529683415598,
+      "learning_rate": 1.951202601307924e-05,
+      "loss": 0.8516,
+      "step": 5385
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6963106071035967,
+      "learning_rate": 1.9511790536075613e-05,
+      "loss": 0.6255,
+      "step": 5386
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9797132922989373,
+      "learning_rate": 1.951155500369133e-05,
+      "loss": 0.75,
+      "step": 5387
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.628652731753082,
+      "learning_rate": 1.9511319415927766e-05,
+      "loss": 0.7169,
+      "step": 5388
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4448312633310598,
+      "learning_rate": 1.9511083772786285e-05,
+      "loss": 0.7044,
+      "step": 5389
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.579922406141784,
+      "learning_rate": 1.9510848074268263e-05,
+      "loss": 0.6309,
+      "step": 5390
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6363784824799068,
+      "learning_rate": 1.9510612320375073e-05,
+      "loss": 0.7509,
+      "step": 5391
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5810380595758635,
+      "learning_rate": 1.951037651110809e-05,
+      "loss": 0.7297,
+      "step": 5392
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4401024367375943,
+      "learning_rate": 1.9510140646468682e-05,
+      "loss": 0.7668,
+      "step": 5393
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3510032180804625,
+      "learning_rate": 1.9509904726458223e-05,
+      "loss": 0.8132,
+      "step": 5394
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4480809490253135,
+      "learning_rate": 1.9509668751078087e-05,
+      "loss": 0.613,
+      "step": 5395
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.485053987010509,
+      "learning_rate": 1.9509432720329653e-05,
+      "loss": 0.8485,
+      "step": 5396
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4067890894646902,
+      "learning_rate": 1.9509196634214288e-05,
+      "loss": 0.7773,
+      "step": 5397
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4543076884426984,
+      "learning_rate": 1.950896049273337e-05,
+      "loss": 0.7067,
+      "step": 5398
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.340736729835222,
+      "learning_rate": 1.9508724295888277e-05,
+      "loss": 0.7543,
+      "step": 5399
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4547381976377967,
+      "learning_rate": 1.9508488043680376e-05,
+      "loss": 0.7967,
+      "step": 5400
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7497921531209069,
+      "learning_rate": 1.950825173611105e-05,
+      "loss": 0.8276,
+      "step": 5401
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7551181894377543,
+      "learning_rate": 1.950801537318167e-05,
+      "loss": 0.6359,
+      "step": 5402
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8733611752158572,
+      "learning_rate": 1.9507778954893614e-05,
+      "loss": 0.6382,
+      "step": 5403
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.474007530809813,
+      "learning_rate": 1.950754248124826e-05,
+      "loss": 0.7839,
+      "step": 5404
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6712143678045015,
+      "learning_rate": 1.9507305952246988e-05,
+      "loss": 0.847,
+      "step": 5405
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6428666748697642,
+      "learning_rate": 1.9507069367891166e-05,
+      "loss": 0.7684,
+      "step": 5406
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.7955255266056356,
+      "learning_rate": 1.950683272818218e-05,
+      "loss": 0.8233,
+      "step": 5407
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3985329399293767,
+      "learning_rate": 1.9506596033121397e-05,
+      "loss": 0.754,
+      "step": 5408
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.982703177117682,
+      "learning_rate": 1.9506359282710207e-05,
+      "loss": 0.7924,
+      "step": 5409
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3598051034537864,
+      "learning_rate": 1.950612247694998e-05,
+      "loss": 0.8266,
+      "step": 5410
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5508278004163052,
+      "learning_rate": 1.9505885615842102e-05,
+      "loss": 0.8836,
+      "step": 5411
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.789605721951554,
+      "learning_rate": 1.950564869938795e-05,
+      "loss": 0.8103,
+      "step": 5412
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9313000461953291,
+      "learning_rate": 1.95054117275889e-05,
+      "loss": 0.716,
+      "step": 5413
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3337267297576558,
+      "learning_rate": 1.9505174700446333e-05,
+      "loss": 0.8575,
+      "step": 5414
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.076981339658784,
+      "learning_rate": 1.950493761796163e-05,
+      "loss": 0.8272,
+      "step": 5415
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5122157385021995,
+      "learning_rate": 1.950470048013617e-05,
+      "loss": 0.8607,
+      "step": 5416
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6705982816837593,
+      "learning_rate": 1.9504463286971332e-05,
+      "loss": 0.5794,
+      "step": 5417
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6519682324402507,
+      "learning_rate": 1.9504226038468506e-05,
+      "loss": 0.6651,
+      "step": 5418
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4641380189025806,
+      "learning_rate": 1.950398873462906e-05,
+      "loss": 0.9143,
+      "step": 5419
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6103962274897969,
+      "learning_rate": 1.9503751375454388e-05,
+      "loss": 0.5327,
+      "step": 5420
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4578596775016233,
+      "learning_rate": 1.9503513960945864e-05,
+      "loss": 0.7517,
+      "step": 5421
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5807535136792064,
+      "learning_rate": 1.9503276491104876e-05,
+      "loss": 0.5101,
+      "step": 5422
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8497978906438755,
+      "learning_rate": 1.9503038965932802e-05,
+      "loss": 0.724,
+      "step": 5423
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5409759265581975,
+      "learning_rate": 1.9502801385431028e-05,
+      "loss": 0.8079,
+      "step": 5424
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8723501206741386,
+      "learning_rate": 1.9502563749600933e-05,
+      "loss": 0.8098,
+      "step": 5425
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5182062361027302,
+      "learning_rate": 1.9502326058443904e-05,
+      "loss": 0.8195,
+      "step": 5426
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3712910428873,
+      "learning_rate": 1.9502088311961324e-05,
+      "loss": 0.8437,
+      "step": 5427
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6054910755415617,
+      "learning_rate": 1.9501850510154578e-05,
+      "loss": 0.5547,
+      "step": 5428
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5858447451932416,
+      "learning_rate": 1.950161265302505e-05,
+      "loss": 0.7756,
+      "step": 5429
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3270048841898467,
+      "learning_rate": 1.9501374740574128e-05,
+      "loss": 0.7459,
+      "step": 5430
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4142477206127542,
+      "learning_rate": 1.9501136772803192e-05,
+      "loss": 0.7219,
+      "step": 5431
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5784736144832231,
+      "learning_rate": 1.9500898749713628e-05,
+      "loss": 0.4833,
+      "step": 5432
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6915879143411636,
+      "learning_rate": 1.9500660671306826e-05,
+      "loss": 0.839,
+      "step": 5433
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.373246798439251,
+      "learning_rate": 1.950042253758417e-05,
+      "loss": 0.7899,
+      "step": 5434
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4490033631336026,
+      "learning_rate": 1.950018434854704e-05,
+      "loss": 0.7314,
+      "step": 5435
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5051085436078766,
+      "learning_rate": 1.9499946104196835e-05,
+      "loss": 0.8972,
+      "step": 5436
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6673220166029659,
+      "learning_rate": 1.9499707804534934e-05,
+      "loss": 0.7597,
+      "step": 5437
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8689948161451198,
+      "learning_rate": 1.949946944956273e-05,
+      "loss": 0.7758,
+      "step": 5438
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7178312726176523,
+      "learning_rate": 1.9499231039281603e-05,
+      "loss": 0.7674,
+      "step": 5439
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.444584002980259,
+      "learning_rate": 1.9498992573692948e-05,
+      "loss": 0.8755,
+      "step": 5440
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6160988412797501,
+      "learning_rate": 1.949875405279815e-05,
+      "loss": 0.8527,
+      "step": 5441
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7857586098692337,
+      "learning_rate": 1.94985154765986e-05,
+      "loss": 0.6654,
+      "step": 5442
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9860268050982597,
+      "learning_rate": 1.949827684509568e-05,
+      "loss": 0.8278,
+      "step": 5443
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5669183475028298,
+      "learning_rate": 1.9498038158290792e-05,
+      "loss": 0.5208,
+      "step": 5444
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.564984664927241,
+      "learning_rate": 1.9497799416185318e-05,
+      "loss": 0.8332,
+      "step": 5445
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4951638897739334,
+      "learning_rate": 1.949756061878065e-05,
+      "loss": 0.8505,
+      "step": 5446
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.364061438960527,
+      "learning_rate": 1.949732176607817e-05,
+      "loss": 0.8723,
+      "step": 5447
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6103029711676682,
+      "learning_rate": 1.9497082858079284e-05,
+      "loss": 0.8446,
+      "step": 5448
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6232869630329593,
+      "learning_rate": 1.9496843894785372e-05,
+      "loss": 0.5526,
+      "step": 5449
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.902142282302777,
+      "learning_rate": 1.9496604876197826e-05,
+      "loss": 0.8126,
+      "step": 5450
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.42870214977034,
+      "learning_rate": 1.949636580231804e-05,
+      "loss": 0.7517,
+      "step": 5451
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9533058558786847,
+      "learning_rate": 1.949612667314741e-05,
+      "loss": 0.795,
+      "step": 5452
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8147324128598272,
+      "learning_rate": 1.949588748868732e-05,
+      "loss": 0.7318,
+      "step": 5453
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8805627112944863,
+      "learning_rate": 1.9495648248939166e-05,
+      "loss": 0.7696,
+      "step": 5454
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.695422481368967,
+      "learning_rate": 1.9495408953904346e-05,
+      "loss": 0.6674,
+      "step": 5455
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6097011896574916,
+      "learning_rate": 1.9495169603584244e-05,
+      "loss": 0.5117,
+      "step": 5456
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6990827249662654,
+      "learning_rate": 1.949493019798026e-05,
+      "loss": 0.9305,
+      "step": 5457
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4802634335402687,
+      "learning_rate": 1.9494690737093787e-05,
+      "loss": 0.6901,
+      "step": 5458
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3640883461294533,
+      "learning_rate": 1.9494451220926224e-05,
+      "loss": 0.7222,
+      "step": 5459
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5054874224225814,
+      "learning_rate": 1.9494211649478952e-05,
+      "loss": 0.6913,
+      "step": 5460
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.631735185543266,
+      "learning_rate": 1.949397202275338e-05,
+      "loss": 0.6565,
+      "step": 5461
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.659170993697747,
+      "learning_rate": 1.9493732340750892e-05,
+      "loss": 0.8283,
+      "step": 5462
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.386122066085231,
+      "learning_rate": 1.949349260347289e-05,
+      "loss": 0.7223,
+      "step": 5463
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7732304059706014,
+      "learning_rate": 1.949325281092077e-05,
+      "loss": 0.697,
+      "step": 5464
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6620084582785224,
+      "learning_rate": 1.9493012963095923e-05,
+      "loss": 0.7499,
+      "step": 5465
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1813514267874172,
+      "learning_rate": 1.9492773059999752e-05,
+      "loss": 0.8436,
+      "step": 5466
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6674898387832775,
+      "learning_rate": 1.949253310163365e-05,
+      "loss": 0.8581,
+      "step": 5467
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5952904931310588,
+      "learning_rate": 1.9492293087999016e-05,
+      "loss": 0.8115,
+      "step": 5468
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6077242923675695,
+      "learning_rate": 1.9492053019097246e-05,
+      "loss": 0.5061,
+      "step": 5469
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.6750140011932055,
+      "learning_rate": 1.949181289492974e-05,
+      "loss": 0.7311,
+      "step": 5470
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.024399318670994,
+      "learning_rate": 1.9491572715497894e-05,
+      "loss": 0.7896,
+      "step": 5471
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2393873961627526,
+      "learning_rate": 1.9491332480803106e-05,
+      "loss": 0.7782,
+      "step": 5472
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6719162019950655,
+      "learning_rate": 1.949109219084677e-05,
+      "loss": 0.6977,
+      "step": 5473
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8958741973812767,
+      "learning_rate": 1.9490851845630297e-05,
+      "loss": 0.7999,
+      "step": 5474
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7445825997411408,
+      "learning_rate": 1.9490611445155078e-05,
+      "loss": 0.6357,
+      "step": 5475
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5939671464944531,
+      "learning_rate": 1.9490370989422516e-05,
+      "loss": 0.8274,
+      "step": 5476
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4540675696818373,
+      "learning_rate": 1.9490130478434008e-05,
+      "loss": 0.7978,
+      "step": 5477
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7376540288970548,
+      "learning_rate": 1.9489889912190954e-05,
+      "loss": 0.869,
+      "step": 5478
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5143344463652841,
+      "learning_rate": 1.948964929069476e-05,
+      "loss": 0.6566,
+      "step": 5479
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.02122655163935,
+      "learning_rate": 1.948940861394682e-05,
+      "loss": 0.7142,
+      "step": 5480
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.587927405725964,
+      "learning_rate": 1.948916788194854e-05,
+      "loss": 0.5353,
+      "step": 5481
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4687750345895116,
+      "learning_rate": 1.9488927094701322e-05,
+      "loss": 0.7977,
+      "step": 5482
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8395185364120077,
+      "learning_rate": 1.9488686252206564e-05,
+      "loss": 0.7143,
+      "step": 5483
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6068115406265322,
+      "learning_rate": 1.948844535446567e-05,
+      "loss": 0.7351,
+      "step": 5484
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.955025822973758,
+      "learning_rate": 1.948820440148005e-05,
+      "loss": 0.7024,
+      "step": 5485
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.441435717522414,
+      "learning_rate": 1.948796339325109e-05,
+      "loss": 0.685,
+      "step": 5486
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6096551479007681,
+      "learning_rate": 1.948772232978021e-05,
+      "loss": 0.7499,
+      "step": 5487
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4537360529041448,
+      "learning_rate": 1.9487481211068807e-05,
+      "loss": 0.9154,
+      "step": 5488
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7009099349079173,
+      "learning_rate": 1.9487240037118283e-05,
+      "loss": 0.6519,
+      "step": 5489
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.583012482991365,
+      "learning_rate": 1.9486998807930043e-05,
+      "loss": 0.8173,
+      "step": 5490
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.005432471234288,
+      "learning_rate": 1.9486757523505495e-05,
+      "loss": 0.695,
+      "step": 5491
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.437710482979007,
+      "learning_rate": 1.9486516183846036e-05,
+      "loss": 0.7289,
+      "step": 5492
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7540742714446143,
+      "learning_rate": 1.9486274788953083e-05,
+      "loss": 0.8817,
+      "step": 5493
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5834103183317606,
+      "learning_rate": 1.948603333882803e-05,
+      "loss": 0.5376,
+      "step": 5494
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5888491848799178,
+      "learning_rate": 1.9485791833472286e-05,
+      "loss": 0.5522,
+      "step": 5495
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4727668956378364,
+      "learning_rate": 1.9485550272887265e-05,
+      "loss": 0.8175,
+      "step": 5496
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5584300250595318,
+      "learning_rate": 1.9485308657074364e-05,
+      "loss": 0.7171,
+      "step": 5497
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6218278378307804,
+      "learning_rate": 1.9485066986034992e-05,
+      "loss": 0.5422,
+      "step": 5498
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5841695447102522,
+      "learning_rate": 1.948482525977056e-05,
+      "loss": 0.7854,
+      "step": 5499
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5003161616377285,
+      "learning_rate": 1.948458347828247e-05,
+      "loss": 0.7469,
+      "step": 5500
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3087306872198494,
+      "learning_rate": 1.9484341641572134e-05,
+      "loss": 0.6784,
+      "step": 5501
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5542744488847032,
+      "learning_rate": 1.9484099749640956e-05,
+      "loss": 0.5174,
+      "step": 5502
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6012856819014818,
+      "learning_rate": 1.9483857802490348e-05,
+      "loss": 0.7994,
+      "step": 5503
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.718445493789106,
+      "learning_rate": 1.9483615800121717e-05,
+      "loss": 0.6842,
+      "step": 5504
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5832096653774044,
+      "learning_rate": 1.9483373742536475e-05,
+      "loss": 0.5212,
+      "step": 5505
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6854841419296374,
+      "learning_rate": 1.9483131629736024e-05,
+      "loss": 0.7574,
+      "step": 5506
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5282062606694857,
+      "learning_rate": 1.9482889461721782e-05,
+      "loss": 0.7107,
+      "step": 5507
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7375215726342572,
+      "learning_rate": 1.948264723849515e-05,
+      "loss": 0.6926,
+      "step": 5508
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3588283897163569,
+      "learning_rate": 1.948240496005755e-05,
+      "loss": 0.8511,
+      "step": 5509
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2704548644670504,
+      "learning_rate": 1.9482162626410383e-05,
+      "loss": 0.7839,
+      "step": 5510
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5993138932873644,
+      "learning_rate": 1.9481920237555064e-05,
+      "loss": 0.5841,
+      "step": 5511
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2220708393122575,
+      "learning_rate": 1.9481677793493005e-05,
+      "loss": 0.8156,
+      "step": 5512
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6085440656735817,
+      "learning_rate": 1.9481435294225614e-05,
+      "loss": 0.7759,
+      "step": 5513
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4997000603555128,
+      "learning_rate": 1.9481192739754307e-05,
+      "loss": 0.7384,
+      "step": 5514
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5994485576065722,
+      "learning_rate": 1.948095013008049e-05,
+      "loss": 0.762,
+      "step": 5515
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3240859030447412,
+      "learning_rate": 1.9480707465205582e-05,
+      "loss": 0.7415,
+      "step": 5516
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5181262148627979,
+      "learning_rate": 1.9480464745130997e-05,
+      "loss": 0.6907,
+      "step": 5517
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3159121239320932,
+      "learning_rate": 1.948022196985814e-05,
+      "loss": 0.7856,
+      "step": 5518
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5114231092469343,
+      "learning_rate": 1.9479979139388432e-05,
+      "loss": 0.7593,
+      "step": 5519
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.706964464792078,
+      "learning_rate": 1.9479736253723282e-05,
+      "loss": 0.781,
+      "step": 5520
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6030571936434022,
+      "learning_rate": 1.9479493312864107e-05,
+      "loss": 0.9284,
+      "step": 5521
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3804457116230386,
+      "learning_rate": 1.9479250316812323e-05,
+      "loss": 0.738,
+      "step": 5522
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3984201390300786,
+      "learning_rate": 1.9479007265569342e-05,
+      "loss": 0.7445,
+      "step": 5523
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5185477331169381,
+      "learning_rate": 1.9478764159136578e-05,
+      "loss": 0.8277,
+      "step": 5524
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4175468446160795,
+      "learning_rate": 1.9478520997515446e-05,
+      "loss": 0.8272,
+      "step": 5525
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.689971919428035,
+      "learning_rate": 1.9478277780707367e-05,
+      "loss": 0.7503,
+      "step": 5526
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7385552505294788,
+      "learning_rate": 1.9478034508713753e-05,
+      "loss": 0.7374,
+      "step": 5527
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.028182481187349,
+      "learning_rate": 1.9477791181536024e-05,
+      "loss": 0.8162,
+      "step": 5528
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4646453386779463,
+      "learning_rate": 1.9477547799175592e-05,
+      "loss": 0.7611,
+      "step": 5529
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.062728321658109,
+      "learning_rate": 1.9477304361633878e-05,
+      "loss": 0.7232,
+      "step": 5530
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6973201598233812,
+      "learning_rate": 1.9477060868912294e-05,
+      "loss": 0.5628,
+      "step": 5531
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5432910105913904,
+      "learning_rate": 1.9476817321012263e-05,
+      "loss": 0.8169,
+      "step": 5532
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4797432111424413,
+      "learning_rate": 1.94765737179352e-05,
+      "loss": 0.7565,
+      "step": 5533
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6211898107659712,
+      "learning_rate": 1.9476330059682528e-05,
+      "loss": 0.6704,
+      "step": 5534
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3570094543044042,
+      "learning_rate": 1.9476086346255657e-05,
+      "loss": 0.8743,
+      "step": 5535
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.598584545984782,
+      "learning_rate": 1.9475842577656018e-05,
+      "loss": 0.5871,
+      "step": 5536
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4512565120299215,
+      "learning_rate": 1.947559875388502e-05,
+      "loss": 0.7469,
+      "step": 5537
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7534623055302156,
+      "learning_rate": 1.9475354874944083e-05,
+      "loss": 0.6981,
+      "step": 5538
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6329151754866489,
+      "learning_rate": 1.9475110940834633e-05,
+      "loss": 0.8783,
+      "step": 5539
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7174057512611172,
+      "learning_rate": 1.947486695155809e-05,
+      "loss": 0.8139,
+      "step": 5540
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.393888167602419,
+      "learning_rate": 1.947462290711587e-05,
+      "loss": 0.8265,
+      "step": 5541
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4519750277245094,
+      "learning_rate": 1.9474378807509395e-05,
+      "loss": 0.6692,
+      "step": 5542
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.480621466060815,
+      "learning_rate": 1.9474134652740083e-05,
+      "loss": 0.8281,
+      "step": 5543
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3540835786469247,
+      "learning_rate": 1.9473890442809366e-05,
+      "loss": 0.8997,
+      "step": 5544
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3409425898253569,
+      "learning_rate": 1.9473646177718658e-05,
+      "loss": 0.8473,
+      "step": 5545
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8494479721276216,
+      "learning_rate": 1.947340185746938e-05,
+      "loss": 0.7389,
+      "step": 5546
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.472295793812743,
+      "learning_rate": 1.9473157482062958e-05,
+      "loss": 0.9877,
+      "step": 5547
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5980353887779728,
+      "learning_rate": 1.9472913051500816e-05,
+      "loss": 0.7823,
+      "step": 5548
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.052361238851147,
+      "learning_rate": 1.9472668565784375e-05,
+      "loss": 0.7526,
+      "step": 5549
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.71723833107039,
+      "learning_rate": 1.9472424024915056e-05,
+      "loss": 0.7267,
+      "step": 5550
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.956072025686748,
+      "learning_rate": 1.9472179428894288e-05,
+      "loss": 0.8172,
+      "step": 5551
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8705792891986288,
+      "learning_rate": 1.9471934777723492e-05,
+      "loss": 0.7825,
+      "step": 5552
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5986548677049175,
+      "learning_rate": 1.9471690071404096e-05,
+      "loss": 0.7642,
+      "step": 5553
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6622475822167182,
+      "learning_rate": 1.9471445309937515e-05,
+      "loss": 0.8631,
+      "step": 5554
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5747996943216176,
+      "learning_rate": 1.9471200493325186e-05,
+      "loss": 0.7122,
+      "step": 5555
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7355487351361425,
+      "learning_rate": 1.9470955621568525e-05,
+      "loss": 0.8128,
+      "step": 5556
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6060468101803358,
+      "learning_rate": 1.9470710694668966e-05,
+      "loss": 0.7312,
+      "step": 5557
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9053257604180405,
+      "learning_rate": 1.9470465712627928e-05,
+      "loss": 0.7731,
+      "step": 5558
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4309093247515656,
+      "learning_rate": 1.9470220675446844e-05,
+      "loss": 0.7253,
+      "step": 5559
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9148955672850394,
+      "learning_rate": 1.946997558312713e-05,
+      "loss": 0.7858,
+      "step": 5560
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7097271184651923,
+      "learning_rate": 1.9469730435670227e-05,
+      "loss": 0.7766,
+      "step": 5561
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.400181568154601,
+      "learning_rate": 1.9469485233077554e-05,
+      "loss": 0.8333,
+      "step": 5562
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.553970523938761,
+      "learning_rate": 1.9469239975350538e-05,
+      "loss": 0.8216,
+      "step": 5563
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5444204072543104,
+      "learning_rate": 1.946899466249061e-05,
+      "loss": 0.6847,
+      "step": 5564
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6085183834801213,
+      "learning_rate": 1.9468749294499196e-05,
+      "loss": 0.4897,
+      "step": 5565
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5855207531146832,
+      "learning_rate": 1.9468503871377728e-05,
+      "loss": 0.5285,
+      "step": 5566
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5923706738607228,
+      "learning_rate": 1.946825839312763e-05,
+      "loss": 0.7025,
+      "step": 5567
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4981875231359876,
+      "learning_rate": 1.9468012859750337e-05,
+      "loss": 0.8488,
+      "step": 5568
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.421995691755509,
+      "learning_rate": 1.9467767271247275e-05,
+      "loss": 0.7843,
+      "step": 5569
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.550631059750723,
+      "learning_rate": 1.9467521627619874e-05,
+      "loss": 0.75,
+      "step": 5570
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8257729697553637,
+      "learning_rate": 1.9467275928869565e-05,
+      "loss": 0.8161,
+      "step": 5571
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6770880508698565,
+      "learning_rate": 1.9467030174997777e-05,
+      "loss": 0.9142,
+      "step": 5572
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6076386443779174,
+      "learning_rate": 1.9466784366005943e-05,
+      "loss": 0.9709,
+      "step": 5573
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.48808112042144,
+      "learning_rate": 1.9466538501895493e-05,
+      "loss": 0.7545,
+      "step": 5574
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4061540530755567,
+      "learning_rate": 1.946629258266786e-05,
+      "loss": 0.8963,
+      "step": 5575
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6978670485783092,
+      "learning_rate": 1.9466046608324472e-05,
+      "loss": 0.8335,
+      "step": 5576
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6102234515971936,
+      "learning_rate": 1.9465800578866764e-05,
+      "loss": 0.7977,
+      "step": 5577
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.468936311648799,
+      "learning_rate": 1.9465554494296173e-05,
+      "loss": 0.7722,
+      "step": 5578
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.72556133298598,
+      "learning_rate": 1.9465308354614125e-05,
+      "loss": 0.691,
+      "step": 5579
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3874104611899798,
+      "learning_rate": 1.946506215982205e-05,
+      "loss": 0.5685,
+      "step": 5580
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5887048107926403,
+      "learning_rate": 1.946481590992139e-05,
+      "loss": 0.8245,
+      "step": 5581
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4843626817667395,
+      "learning_rate": 1.9464569604913575e-05,
+      "loss": 0.6484,
+      "step": 5582
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3082202860408776,
+      "learning_rate": 1.9464323244800042e-05,
+      "loss": 0.7046,
+      "step": 5583
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.656430115785448,
+      "learning_rate": 1.946407682958222e-05,
+      "loss": 0.6849,
+      "step": 5584
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.874564954335129,
+      "learning_rate": 1.9463830359261547e-05,
+      "loss": 0.7603,
+      "step": 5585
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5937982975272637,
+      "learning_rate": 1.946358383383946e-05,
+      "loss": 0.5671,
+      "step": 5586
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4909226156316648,
+      "learning_rate": 1.9463337253317385e-05,
+      "loss": 0.7381,
+      "step": 5587
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6087415459583289,
+      "learning_rate": 1.946309061769677e-05,
+      "loss": 0.5035,
+      "step": 5588
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5925491702333027,
+      "learning_rate": 1.9462843926979044e-05,
+      "loss": 0.5649,
+      "step": 5589
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4023395371780565,
+      "learning_rate": 1.9462597181165647e-05,
+      "loss": 0.8406,
+      "step": 5590
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6572430587233715,
+      "learning_rate": 1.946235038025801e-05,
+      "loss": 0.7947,
+      "step": 5591
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4170147238944855,
+      "learning_rate": 1.9462103524257574e-05,
+      "loss": 0.7959,
+      "step": 5592
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3991460005512157,
+      "learning_rate": 1.9461856613165775e-05,
+      "loss": 0.8216,
+      "step": 5593
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5975212682317523,
+      "learning_rate": 1.946160964698405e-05,
+      "loss": 0.7747,
+      "step": 5594
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6078629183864177,
+      "learning_rate": 1.9461362625713845e-05,
+      "loss": 0.7733,
+      "step": 5595
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6001270465369831,
+      "learning_rate": 1.9461115549356588e-05,
+      "loss": 0.5364,
+      "step": 5596
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4151271083159256,
+      "learning_rate": 1.9460868417913718e-05,
+      "loss": 0.823,
+      "step": 5597
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.582665888036888,
+      "learning_rate": 1.946062123138668e-05,
+      "loss": 0.7561,
+      "step": 5598
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5879484982409027,
+      "learning_rate": 1.9460373989776905e-05,
+      "loss": 0.8323,
+      "step": 5599
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3797742173486836,
+      "learning_rate": 1.9460126693085842e-05,
+      "loss": 0.8078,
+      "step": 5600
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.612691236544782,
+      "learning_rate": 1.9459879341314927e-05,
+      "loss": 0.7251,
+      "step": 5601
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.3823652313586012,
+      "learning_rate": 1.9459631934465596e-05,
+      "loss": 0.7226,
+      "step": 5602
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6529745385822192,
+      "learning_rate": 1.9459384472539297e-05,
+      "loss": 0.5761,
+      "step": 5603
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4221481277515868,
+      "learning_rate": 1.9459136955537465e-05,
+      "loss": 0.8016,
+      "step": 5604
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2503561457193415,
+      "learning_rate": 1.9458889383461543e-05,
+      "loss": 0.8268,
+      "step": 5605
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3174246179177755,
+      "learning_rate": 1.9458641756312973e-05,
+      "loss": 0.7977,
+      "step": 5606
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.619903056516054,
+      "learning_rate": 1.9458394074093198e-05,
+      "loss": 0.7412,
+      "step": 5607
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3644028734586857,
+      "learning_rate": 1.9458146336803653e-05,
+      "loss": 0.6837,
+      "step": 5608
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3589366129292786,
+      "learning_rate": 1.945789854444579e-05,
+      "loss": 0.8106,
+      "step": 5609
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.8817227438398003,
+      "learning_rate": 1.9457650697021046e-05,
+      "loss": 0.6717,
+      "step": 5610
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7070824022005988,
+      "learning_rate": 1.9457402794530864e-05,
+      "loss": 0.8028,
+      "step": 5611
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0517722030916965,
+      "learning_rate": 1.945715483697669e-05,
+      "loss": 0.9164,
+      "step": 5612
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8030538951355097,
+      "learning_rate": 1.9456906824359965e-05,
+      "loss": 0.8253,
+      "step": 5613
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.7046443359162513,
+      "learning_rate": 1.9456658756682136e-05,
+      "loss": 0.8511,
+      "step": 5614
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6596320423234978,
+      "learning_rate": 1.9456410633944645e-05,
+      "loss": 0.7924,
+      "step": 5615
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5819576070777102,
+      "learning_rate": 1.9456162456148938e-05,
+      "loss": 0.525,
+      "step": 5616
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.795364428506334,
+      "learning_rate": 1.9455914223296458e-05,
+      "loss": 0.7647,
+      "step": 5617
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4937895418463962,
+      "learning_rate": 1.9455665935388653e-05,
+      "loss": 0.627,
+      "step": 5618
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.622424437278253,
+      "learning_rate": 1.945541759242697e-05,
+      "loss": 0.6997,
+      "step": 5619
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5890710204944826,
+      "learning_rate": 1.945516919441285e-05,
+      "loss": 0.5772,
+      "step": 5620
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5694877977522883,
+      "learning_rate": 1.945492074134774e-05,
+      "loss": 0.653,
+      "step": 5621
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9343486097452491,
+      "learning_rate": 1.945467223323309e-05,
+      "loss": 0.7662,
+      "step": 5622
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5138244968618477,
+      "learning_rate": 1.945442367007034e-05,
+      "loss": 0.732,
+      "step": 5623
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6924642207533067,
+      "learning_rate": 1.945417505186095e-05,
+      "loss": 0.7433,
+      "step": 5624
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5982968614975839,
+      "learning_rate": 1.9453926378606354e-05,
+      "loss": 0.7553,
+      "step": 5625
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8288424877115377,
+      "learning_rate": 1.9453677650308007e-05,
+      "loss": 0.8421,
+      "step": 5626
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3924507159571093,
+      "learning_rate": 1.945342886696736e-05,
+      "loss": 0.7681,
+      "step": 5627
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7124378346084803,
+      "learning_rate": 1.945318002858585e-05,
+      "loss": 0.7871,
+      "step": 5628
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.641842473870686,
+      "learning_rate": 1.9452931135164936e-05,
+      "loss": 0.694,
+      "step": 5629
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6619594716726513,
+      "learning_rate": 1.9452682186706067e-05,
+      "loss": 0.8804,
+      "step": 5630
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0003861858102496,
+      "learning_rate": 1.945243318321069e-05,
+      "loss": 0.8348,
+      "step": 5631
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7918072751103653,
+      "learning_rate": 1.9452184124680247e-05,
+      "loss": 0.8242,
+      "step": 5632
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6260831633804849,
+      "learning_rate": 1.94519350111162e-05,
+      "loss": 0.6505,
+      "step": 5633
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7156958020466702,
+      "learning_rate": 1.9451685842519997e-05,
+      "loss": 0.6532,
+      "step": 5634
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6356709343333309,
+      "learning_rate": 1.9451436618893086e-05,
+      "loss": 0.7129,
+      "step": 5635
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.59871709860077,
+      "learning_rate": 1.9451187340236918e-05,
+      "loss": 0.6829,
+      "step": 5636
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5971567718023871,
+      "learning_rate": 1.9450938006552943e-05,
+      "loss": 0.6703,
+      "step": 5637
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6321343528821814,
+      "learning_rate": 1.945068861784262e-05,
+      "loss": 0.7661,
+      "step": 5638
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4751284887929588,
+      "learning_rate": 1.9450439174107388e-05,
+      "loss": 0.7435,
+      "step": 5639
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6957345337268108,
+      "learning_rate": 1.9450189675348712e-05,
+      "loss": 0.8539,
+      "step": 5640
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7861720916512172,
+      "learning_rate": 1.944994012156804e-05,
+      "loss": 0.7706,
+      "step": 5641
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6010855897052353,
+      "learning_rate": 1.9449690512766825e-05,
+      "loss": 0.5743,
+      "step": 5642
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4346660432411678,
+      "learning_rate": 1.944944084894652e-05,
+      "loss": 0.8326,
+      "step": 5643
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5269688837707387,
+      "learning_rate": 1.944919113010858e-05,
+      "loss": 0.6748,
+      "step": 5644
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3621983945882168,
+      "learning_rate": 1.9448941356254453e-05,
+      "loss": 0.7513,
+      "step": 5645
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.721686730714254,
+      "learning_rate": 1.94486915273856e-05,
+      "loss": 0.852,
+      "step": 5646
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5861915179325665,
+      "learning_rate": 1.9448441643503473e-05,
+      "loss": 0.5629,
+      "step": 5647
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.942864160454644,
+      "learning_rate": 1.944819170460953e-05,
+      "loss": 0.9694,
+      "step": 5648
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8069721539074706,
+      "learning_rate": 1.9447941710705223e-05,
+      "loss": 0.7789,
+      "step": 5649
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.632405161303863,
+      "learning_rate": 1.9447691661792007e-05,
+      "loss": 0.6968,
+      "step": 5650
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3515217227879295,
+      "learning_rate": 1.944744155787134e-05,
+      "loss": 0.9418,
+      "step": 5651
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6198535498852176,
+      "learning_rate": 1.944719139894468e-05,
+      "loss": 0.7282,
+      "step": 5652
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.505793258043328,
+      "learning_rate": 1.9446941185013474e-05,
+      "loss": 0.7897,
+      "step": 5653
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5996564877813394,
+      "learning_rate": 1.944669091607919e-05,
+      "loss": 0.5297,
+      "step": 5654
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4132264746698913,
+      "learning_rate": 1.944644059214328e-05,
+      "loss": 0.8385,
+      "step": 5655
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8287455347654982,
+      "learning_rate": 1.94461902132072e-05,
+      "loss": 0.6898,
+      "step": 5656
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6464399124191742,
+      "learning_rate": 1.9445939779272415e-05,
+      "loss": 0.8496,
+      "step": 5657
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5942256733870211,
+      "learning_rate": 1.9445689290340372e-05,
+      "loss": 0.5686,
+      "step": 5658
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4561688954344696,
+      "learning_rate": 1.944543874641254e-05,
+      "loss": 0.7817,
+      "step": 5659
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6583947846905005,
+      "learning_rate": 1.9445188147490373e-05,
+      "loss": 0.7618,
+      "step": 5660
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.17975129066143,
+      "learning_rate": 1.9444937493575326e-05,
+      "loss": 0.8331,
+      "step": 5661
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.554304298336325,
+      "learning_rate": 1.9444686784668868e-05,
+      "loss": 0.747,
+      "step": 5662
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6068818388155203,
+      "learning_rate": 1.944443602077245e-05,
+      "loss": 0.5399,
+      "step": 5663
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6972028815579518,
+      "learning_rate": 1.9444185201887536e-05,
+      "loss": 0.711,
+      "step": 5664
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5995176925765839,
+      "learning_rate": 1.9443934328015587e-05,
+      "loss": 0.5441,
+      "step": 5665
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0637480603403655,
+      "learning_rate": 1.944368339915806e-05,
+      "loss": 0.6773,
+      "step": 5666
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5275312065658975,
+      "learning_rate": 1.9443432415316424e-05,
+      "loss": 0.7145,
+      "step": 5667
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4824136659393485,
+      "learning_rate": 1.944318137649213e-05,
+      "loss": 0.8777,
+      "step": 5668
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.7125662016582166,
+      "learning_rate": 1.9442930282686647e-05,
+      "loss": 0.8504,
+      "step": 5669
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4510288958773203,
+      "learning_rate": 1.9442679133901433e-05,
+      "loss": 0.7997,
+      "step": 5670
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3414460981758412,
+      "learning_rate": 1.9442427930137948e-05,
+      "loss": 0.7523,
+      "step": 5671
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5223482886018191,
+      "learning_rate": 1.9442176671397663e-05,
+      "loss": 0.8383,
+      "step": 5672
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.823791503642945,
+      "learning_rate": 1.9441925357682033e-05,
+      "loss": 0.7691,
+      "step": 5673
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.49479486261619,
+      "learning_rate": 1.9441673988992525e-05,
+      "loss": 0.7351,
+      "step": 5674
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4953629407192799,
+      "learning_rate": 1.9441422565330606e-05,
+      "loss": 0.8323,
+      "step": 5675
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5209343154765647,
+      "learning_rate": 1.9441171086697732e-05,
+      "loss": 0.7202,
+      "step": 5676
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6028445632117585,
+      "learning_rate": 1.944091955309537e-05,
+      "loss": 0.8454,
+      "step": 5677
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5438409023578608,
+      "learning_rate": 1.9440667964524987e-05,
+      "loss": 0.6681,
+      "step": 5678
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7354095324491003,
+      "learning_rate": 1.9440416320988046e-05,
+      "loss": 0.7502,
+      "step": 5679
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3837208511162584,
+      "learning_rate": 1.9440164622486013e-05,
+      "loss": 0.7954,
+      "step": 5680
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.656035153036949,
+      "learning_rate": 1.9439912869020353e-05,
+      "loss": 0.9234,
+      "step": 5681
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7880170380536107,
+      "learning_rate": 1.943966106059253e-05,
+      "loss": 0.7996,
+      "step": 5682
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5773976214372093,
+      "learning_rate": 1.9439409197204012e-05,
+      "loss": 0.5681,
+      "step": 5683
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7758564973187727,
+      "learning_rate": 1.9439157278856268e-05,
+      "loss": 0.8429,
+      "step": 5684
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4694222119587514,
+      "learning_rate": 1.9438905305550756e-05,
+      "loss": 0.6854,
+      "step": 5685
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3421621439832025,
+      "learning_rate": 1.9438653277288953e-05,
+      "loss": 0.8207,
+      "step": 5686
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.085197944572304,
+      "learning_rate": 1.9438401194072324e-05,
+      "loss": 0.7197,
+      "step": 5687
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5301483090548544,
+      "learning_rate": 1.9438149055902335e-05,
+      "loss": 0.7561,
+      "step": 5688
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0832264818013306,
+      "learning_rate": 1.943789686278045e-05,
+      "loss": 0.7176,
+      "step": 5689
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3083322897566583,
+      "learning_rate": 1.9437644614708143e-05,
+      "loss": 0.8352,
+      "step": 5690
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6828451301454277,
+      "learning_rate": 1.9437392311686884e-05,
+      "loss": 0.7058,
+      "step": 5691
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4829044782735843,
+      "learning_rate": 1.9437139953718137e-05,
+      "loss": 0.7818,
+      "step": 5692
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.673577767620741,
+      "learning_rate": 1.9436887540803372e-05,
+      "loss": 0.6866,
+      "step": 5693
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9545849788777596,
+      "learning_rate": 1.943663507294406e-05,
+      "loss": 0.7364,
+      "step": 5694
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.680188763649241,
+      "learning_rate": 1.9436382550141675e-05,
+      "loss": 0.7385,
+      "step": 5695
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5463707856707116,
+      "learning_rate": 1.943612997239768e-05,
+      "loss": 0.8075,
+      "step": 5696
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6200336982923043,
+      "learning_rate": 1.943587733971355e-05,
+      "loss": 0.7899,
+      "step": 5697
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5586246121746972,
+      "learning_rate": 1.9435624652090754e-05,
+      "loss": 0.755,
+      "step": 5698
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6768631185719853,
+      "learning_rate": 1.9435371909530766e-05,
+      "loss": 0.7782,
+      "step": 5699
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8216132515544352,
+      "learning_rate": 1.9435119112035054e-05,
+      "loss": 0.8416,
+      "step": 5700
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6105066554669558,
+      "learning_rate": 1.943486625960509e-05,
+      "loss": 0.5091,
+      "step": 5701
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6321082269670613,
+      "learning_rate": 1.943461335224235e-05,
+      "loss": 0.7484,
+      "step": 5702
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.618818164300547,
+      "learning_rate": 1.9434360389948303e-05,
+      "loss": 0.556,
+      "step": 5703
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.581546543230435,
+      "learning_rate": 1.9434107372724424e-05,
+      "loss": 0.8861,
+      "step": 5704
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4904684323595407,
+      "learning_rate": 1.9433854300572184e-05,
+      "loss": 0.6516,
+      "step": 5705
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8179366167732545,
+      "learning_rate": 1.943360117349306e-05,
+      "loss": 0.7707,
+      "step": 5706
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.592316033816058,
+      "learning_rate": 1.943334799148852e-05,
+      "loss": 0.827,
+      "step": 5707
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5521563202591322,
+      "learning_rate": 1.9433094754560042e-05,
+      "loss": 0.7871,
+      "step": 5708
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8984008912422579,
+      "learning_rate": 1.94328414627091e-05,
+      "loss": 0.8359,
+      "step": 5709
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4015231129352697,
+      "learning_rate": 1.943258811593717e-05,
+      "loss": 0.8069,
+      "step": 5710
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5224272330611668,
+      "learning_rate": 1.9432334714245723e-05,
+      "loss": 0.8219,
+      "step": 5711
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5954039537316695,
+      "learning_rate": 1.943208125763624e-05,
+      "loss": 0.5481,
+      "step": 5712
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.53621884971739,
+      "learning_rate": 1.943182774611019e-05,
+      "loss": 0.8407,
+      "step": 5713
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5477860997693007,
+      "learning_rate": 1.943157417966906e-05,
+      "loss": 0.6642,
+      "step": 5714
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3630625005356827,
+      "learning_rate": 1.9431320558314313e-05,
+      "loss": 0.7264,
+      "step": 5715
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.009422152738297,
+      "learning_rate": 1.9431066882047434e-05,
+      "loss": 0.8402,
+      "step": 5716
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.680909068894146,
+      "learning_rate": 1.9430813150869896e-05,
+      "loss": 0.8921,
+      "step": 5717
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.682833266235262,
+      "learning_rate": 1.943055936478318e-05,
+      "loss": 0.7326,
+      "step": 5718
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7823001746511884,
+      "learning_rate": 1.943030552378876e-05,
+      "loss": 0.7952,
+      "step": 5719
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9119794553070624,
+      "learning_rate": 1.9430051627888117e-05,
+      "loss": 0.6485,
+      "step": 5720
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6169334596728713,
+      "learning_rate": 1.942979767708273e-05,
+      "loss": 0.7885,
+      "step": 5721
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6160052110401513,
+      "learning_rate": 1.942954367137407e-05,
+      "loss": 0.6952,
+      "step": 5722
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.456694884811978,
+      "learning_rate": 1.9429289610763624e-05,
+      "loss": 0.7818,
+      "step": 5723
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6010003236269637,
+      "learning_rate": 1.9429035495252872e-05,
+      "loss": 0.7889,
+      "step": 5724
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5658658861264313,
+      "learning_rate": 1.9428781324843285e-05,
+      "loss": 0.5012,
+      "step": 5725
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5789198594874359,
+      "learning_rate": 1.942852709953635e-05,
+      "loss": 0.7685,
+      "step": 5726
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4261682091006675,
+      "learning_rate": 1.9428272819333545e-05,
+      "loss": 0.8309,
+      "step": 5727
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6475224680859728,
+      "learning_rate": 1.9428018484236353e-05,
+      "loss": 0.819,
+      "step": 5728
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5974459550808118,
+      "learning_rate": 1.9427764094246248e-05,
+      "loss": 0.8835,
+      "step": 5729
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5903781282691689,
+      "learning_rate": 1.942750964936472e-05,
+      "loss": 0.566,
+      "step": 5730
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6129464788812181,
+      "learning_rate": 1.9427255149593245e-05,
+      "loss": 0.5293,
+      "step": 5731
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.658871350518192,
+      "learning_rate": 1.9427000594933305e-05,
+      "loss": 0.6883,
+      "step": 5732
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5767526021712661,
+      "learning_rate": 1.9426745985386385e-05,
+      "loss": 0.7594,
+      "step": 5733
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6697068428892718,
+      "learning_rate": 1.9426491320953963e-05,
+      "loss": 0.846,
+      "step": 5734
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6484288393379061,
+      "learning_rate": 1.9426236601637523e-05,
+      "loss": 0.5531,
+      "step": 5735
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6745695192313848,
+      "learning_rate": 1.9425981827438552e-05,
+      "loss": 0.6952,
+      "step": 5736
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.93839071530222,
+      "learning_rate": 1.942572699835853e-05,
+      "loss": 0.6841,
+      "step": 5737
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4356111142526709,
+      "learning_rate": 1.942547211439894e-05,
+      "loss": 0.8408,
+      "step": 5738
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4639551982497634,
+      "learning_rate": 1.9425217175561268e-05,
+      "loss": 0.7402,
+      "step": 5739
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7745841590899818,
+      "learning_rate": 1.9424962181847e-05,
+      "loss": 0.7536,
+      "step": 5740
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1450501914199944,
+      "learning_rate": 1.9424707133257615e-05,
+      "loss": 0.7202,
+      "step": 5741
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6145594994871375,
+      "learning_rate": 1.9424452029794603e-05,
+      "loss": 0.7725,
+      "step": 5742
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8821137874159253,
+      "learning_rate": 1.9424196871459448e-05,
+      "loss": 0.6653,
+      "step": 5743
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6393222756685671,
+      "learning_rate": 1.9423941658253634e-05,
+      "loss": 0.5294,
+      "step": 5744
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6614650246808433,
+      "learning_rate": 1.942368639017865e-05,
+      "loss": 0.7973,
+      "step": 5745
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6001051168352243,
+      "learning_rate": 1.9423431067235977e-05,
+      "loss": 0.5979,
+      "step": 5746
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4778872480795373,
+      "learning_rate": 1.9423175689427105e-05,
+      "loss": 0.8384,
+      "step": 5747
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5091094571687507,
+      "learning_rate": 1.9422920256753523e-05,
+      "loss": 0.6687,
+      "step": 5748
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5383131052469388,
+      "learning_rate": 1.9422664769216716e-05,
+      "loss": 0.925,
+      "step": 5749
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4163298311605255,
+      "learning_rate": 1.9422409226818175e-05,
+      "loss": 0.9023,
+      "step": 5750
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6084889985825513,
+      "learning_rate": 1.9422153629559376e-05,
+      "loss": 0.5657,
+      "step": 5751
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.306505974221004,
+      "learning_rate": 1.942189797744182e-05,
+      "loss": 0.7655,
+      "step": 5752
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.572589761091768,
+      "learning_rate": 1.942164227046699e-05,
+      "loss": 0.5271,
+      "step": 5753
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5929327433718626,
+      "learning_rate": 1.9421386508636378e-05,
+      "loss": 0.5224,
+      "step": 5754
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5793249709750223,
+      "learning_rate": 1.9421130691951472e-05,
+      "loss": 0.9082,
+      "step": 5755
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.734206048849087,
+      "learning_rate": 1.942087482041376e-05,
+      "loss": 0.7972,
+      "step": 5756
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3771996195362555,
+      "learning_rate": 1.9420618894024727e-05,
+      "loss": 0.7571,
+      "step": 5757
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6247485845468848,
+      "learning_rate": 1.942036291278587e-05,
+      "loss": 0.8132,
+      "step": 5758
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6276980070857663,
+      "learning_rate": 1.942010687669868e-05,
+      "loss": 0.691,
+      "step": 5759
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.088367447215152,
+      "learning_rate": 1.9419850785764645e-05,
+      "loss": 0.6872,
+      "step": 5760
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7177489904081278,
+      "learning_rate": 1.9419594639985257e-05,
+      "loss": 0.8222,
+      "step": 5761
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.4611662121676954,
+      "learning_rate": 1.9419338439362007e-05,
+      "loss": 0.7822,
+      "step": 5762
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5807267662440703,
+      "learning_rate": 1.9419082183896385e-05,
+      "loss": 0.7472,
+      "step": 5763
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5963138199998497,
+      "learning_rate": 1.9418825873589883e-05,
+      "loss": 0.5282,
+      "step": 5764
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4250155970605491,
+      "learning_rate": 1.9418569508443998e-05,
+      "loss": 0.7502,
+      "step": 5765
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.321200264730104,
+      "learning_rate": 1.941831308846022e-05,
+      "loss": 0.8112,
+      "step": 5766
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5929149578737616,
+      "learning_rate": 1.9418056613640038e-05,
+      "loss": 0.7948,
+      "step": 5767
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7614657940944691,
+      "learning_rate": 1.941780008398495e-05,
+      "loss": 0.7639,
+      "step": 5768
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4553721084032378,
+      "learning_rate": 1.941754349949645e-05,
+      "loss": 0.8021,
+      "step": 5769
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4717792467699258,
+      "learning_rate": 1.941728686017603e-05,
+      "loss": 0.7844,
+      "step": 5770
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4825349537245778,
+      "learning_rate": 1.9417030166025182e-05,
+      "loss": 0.7196,
+      "step": 5771
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.656781985156417,
+      "learning_rate": 1.9416773417045406e-05,
+      "loss": 0.564,
+      "step": 5772
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7721325794968077,
+      "learning_rate": 1.9416516613238193e-05,
+      "loss": 0.8985,
+      "step": 5773
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7885691527385497,
+      "learning_rate": 1.9416259754605034e-05,
+      "loss": 0.7467,
+      "step": 5774
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.568039387333085,
+      "learning_rate": 1.9416002841147438e-05,
+      "loss": 0.7716,
+      "step": 5775
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6810275866080708,
+      "learning_rate": 1.9415745872866887e-05,
+      "loss": 0.7906,
+      "step": 5776
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6194741808619681,
+      "learning_rate": 1.9415488849764884e-05,
+      "loss": 0.6886,
+      "step": 5777
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.364982128350929,
+      "learning_rate": 1.9415231771842925e-05,
+      "loss": 0.8734,
+      "step": 5778
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5849838324317116,
+      "learning_rate": 1.9414974639102505e-05,
+      "loss": 0.7935,
+      "step": 5779
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4832847546964345,
+      "learning_rate": 1.941471745154512e-05,
+      "loss": 0.8965,
+      "step": 5780
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.680658673656553,
+      "learning_rate": 1.941446020917227e-05,
+      "loss": 0.772,
+      "step": 5781
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4434833322557048,
+      "learning_rate": 1.9414202911985453e-05,
+      "loss": 0.6855,
+      "step": 5782
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5057266556410016,
+      "learning_rate": 1.9413945559986164e-05,
+      "loss": 0.7376,
+      "step": 5783
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5481683123156111,
+      "learning_rate": 1.9413688153175907e-05,
+      "loss": 0.7777,
+      "step": 5784
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5455798071620732,
+      "learning_rate": 1.9413430691556175e-05,
+      "loss": 0.8671,
+      "step": 5785
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.473847293014269,
+      "learning_rate": 1.9413173175128472e-05,
+      "loss": 0.704,
+      "step": 5786
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5023081962351137,
+      "learning_rate": 1.9412915603894294e-05,
+      "loss": 0.7312,
+      "step": 5787
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3827629896898956,
+      "learning_rate": 1.941265797785514e-05,
+      "loss": 0.7712,
+      "step": 5788
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5063299520884221,
+      "learning_rate": 1.941240029701251e-05,
+      "loss": 0.6952,
+      "step": 5789
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5468843373743224,
+      "learning_rate": 1.9412142561367906e-05,
+      "loss": 0.7986,
+      "step": 5790
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7126943812415756,
+      "learning_rate": 1.941188477092283e-05,
+      "loss": 0.7684,
+      "step": 5791
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6924909072647014,
+      "learning_rate": 1.941162692567878e-05,
+      "loss": 0.6718,
+      "step": 5792
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5352209099807292,
+      "learning_rate": 1.941136902563726e-05,
+      "loss": 0.7729,
+      "step": 5793
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7992159512931736,
+      "learning_rate": 1.941111107079977e-05,
+      "loss": 0.7993,
+      "step": 5794
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6938594141575007,
+      "learning_rate": 1.9410853061167806e-05,
+      "loss": 0.8016,
+      "step": 5795
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8304099293486877,
+      "learning_rate": 1.9410594996742884e-05,
+      "loss": 0.9041,
+      "step": 5796
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.625592175781552,
+      "learning_rate": 1.9410336877526494e-05,
+      "loss": 0.8121,
+      "step": 5797
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.49350293031761,
+      "learning_rate": 1.9410078703520147e-05,
+      "loss": 0.887,
+      "step": 5798
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.669008177736904,
+      "learning_rate": 1.9409820474725338e-05,
+      "loss": 0.7903,
+      "step": 5799
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7453600334472699,
+      "learning_rate": 1.9409562191143577e-05,
+      "loss": 0.7396,
+      "step": 5800
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4413553216056076,
+      "learning_rate": 1.9409303852776365e-05,
+      "loss": 0.7521,
+      "step": 5801
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7820069592291452,
+      "learning_rate": 1.940904545962521e-05,
+      "loss": 0.751,
+      "step": 5802
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6265808914996331,
+      "learning_rate": 1.940878701169161e-05,
+      "loss": 0.6487,
+      "step": 5803
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6147123669166198,
+      "learning_rate": 1.9408528508977076e-05,
+      "loss": 0.5289,
+      "step": 5804
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4779724155034413,
+      "learning_rate": 1.940826995148311e-05,
+      "loss": 0.731,
+      "step": 5805
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4618707308836631,
+      "learning_rate": 1.9408011339211215e-05,
+      "loss": 0.8791,
+      "step": 5806
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.5356501505052327,
+      "learning_rate": 1.94077526721629e-05,
+      "loss": 0.711,
+      "step": 5807
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5913961381130243,
+      "learning_rate": 1.940749395033967e-05,
+      "loss": 0.5258,
+      "step": 5808
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4807769676222222,
+      "learning_rate": 1.9407235173743038e-05,
+      "loss": 0.8075,
+      "step": 5809
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5534066608256685,
+      "learning_rate": 1.94069763423745e-05,
+      "loss": 0.9255,
+      "step": 5810
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.626470192582578,
+      "learning_rate": 1.9406717456235566e-05,
+      "loss": 0.7743,
+      "step": 5811
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4243976663341222,
+      "learning_rate": 1.9406458515327747e-05,
+      "loss": 0.8881,
+      "step": 5812
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6862897887355024,
+      "learning_rate": 1.940619951965255e-05,
+      "loss": 0.7908,
+      "step": 5813
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.957618394170477,
+      "learning_rate": 1.9405940469211478e-05,
+      "loss": 0.6988,
+      "step": 5814
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4961139812436846,
+      "learning_rate": 1.9405681364006047e-05,
+      "loss": 0.7377,
+      "step": 5815
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6098055986921823,
+      "learning_rate": 1.9405422204037758e-05,
+      "loss": 0.5458,
+      "step": 5816
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6130080501934845,
+      "learning_rate": 1.940516298930812e-05,
+      "loss": 0.8228,
+      "step": 5817
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5247880364228736,
+      "learning_rate": 1.940490371981865e-05,
+      "loss": 0.7557,
+      "step": 5818
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6346184123549656,
+      "learning_rate": 1.9404644395570852e-05,
+      "loss": 0.7891,
+      "step": 5819
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6262903165629419,
+      "learning_rate": 1.940438501656624e-05,
+      "loss": 0.5386,
+      "step": 5820
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8039787942047216,
+      "learning_rate": 1.9404125582806316e-05,
+      "loss": 0.8461,
+      "step": 5821
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.587185904203018,
+      "learning_rate": 1.9403866094292598e-05,
+      "loss": 0.7002,
+      "step": 5822
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.563740515108813,
+      "learning_rate": 1.9403606551026596e-05,
+      "loss": 0.7834,
+      "step": 5823
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.704336232312531,
+      "learning_rate": 1.9403346953009817e-05,
+      "loss": 0.7578,
+      "step": 5824
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8125485959852778,
+      "learning_rate": 1.9403087300243776e-05,
+      "loss": 0.7146,
+      "step": 5825
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5788277252013005,
+      "learning_rate": 1.9402827592729984e-05,
+      "loss": 0.7171,
+      "step": 5826
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.672896490437382,
+      "learning_rate": 1.9402567830469952e-05,
+      "loss": 0.8274,
+      "step": 5827
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7028431818153948,
+      "learning_rate": 1.9402308013465192e-05,
+      "loss": 0.7777,
+      "step": 5828
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.899715332655358,
+      "learning_rate": 1.9402048141717217e-05,
+      "loss": 0.7768,
+      "step": 5829
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5792497939161558,
+      "learning_rate": 1.9401788215227547e-05,
+      "loss": 0.8562,
+      "step": 5830
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5790805814611475,
+      "learning_rate": 1.9401528233997684e-05,
+      "loss": 0.7171,
+      "step": 5831
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6721821820049028,
+      "learning_rate": 1.940126819802915e-05,
+      "loss": 0.7623,
+      "step": 5832
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4529633625209641,
+      "learning_rate": 1.9401008107323457e-05,
+      "loss": 0.8104,
+      "step": 5833
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4302078286438207,
+      "learning_rate": 1.9400747961882115e-05,
+      "loss": 0.656,
+      "step": 5834
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4643473985291573,
+      "learning_rate": 1.9400487761706646e-05,
+      "loss": 0.8469,
+      "step": 5835
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2888176494370798,
+      "learning_rate": 1.9400227506798557e-05,
+      "loss": 0.7543,
+      "step": 5836
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7650392071199104,
+      "learning_rate": 1.939996719715937e-05,
+      "loss": 0.6752,
+      "step": 5837
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6770675509317485,
+      "learning_rate": 1.93997068327906e-05,
+      "loss": 0.8616,
+      "step": 5838
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8385870240027264,
+      "learning_rate": 1.9399446413693756e-05,
+      "loss": 0.7387,
+      "step": 5839
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5679810268486218,
+      "learning_rate": 1.939918593987036e-05,
+      "loss": 0.7331,
+      "step": 5840
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4022431364793881,
+      "learning_rate": 1.939892541132193e-05,
+      "loss": 0.8268,
+      "step": 5841
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7034548600110166,
+      "learning_rate": 1.9398664828049983e-05,
+      "loss": 0.7624,
+      "step": 5842
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.410548808822763,
+      "learning_rate": 1.939840419005603e-05,
+      "loss": 0.8337,
+      "step": 5843
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8443304837878154,
+      "learning_rate": 1.9398143497341593e-05,
+      "loss": 0.8329,
+      "step": 5844
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4201988804076846,
+      "learning_rate": 1.939788274990819e-05,
+      "loss": 0.7054,
+      "step": 5845
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6712914616750865,
+      "learning_rate": 1.9397621947757338e-05,
+      "loss": 0.769,
+      "step": 5846
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7855961379955065,
+      "learning_rate": 1.9397361090890554e-05,
+      "loss": 0.8092,
+      "step": 5847
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7437169700276334,
+      "learning_rate": 1.939710017930936e-05,
+      "loss": 0.7829,
+      "step": 5848
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7291113647997534,
+      "learning_rate": 1.939683921301527e-05,
+      "loss": 0.6944,
+      "step": 5849
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6254742101062853,
+      "learning_rate": 1.9396578192009812e-05,
+      "loss": 0.8309,
+      "step": 5850
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.818511702454248,
+      "learning_rate": 1.9396317116294502e-05,
+      "loss": 0.8353,
+      "step": 5851
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4459435510079635,
+      "learning_rate": 1.9396055985870855e-05,
+      "loss": 0.7065,
+      "step": 5852
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4990313040401895,
+      "learning_rate": 1.93957948007404e-05,
+      "loss": 0.788,
+      "step": 5853
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7865856544706116,
+      "learning_rate": 1.939553356090465e-05,
+      "loss": 0.9592,
+      "step": 5854
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7083610179143045,
+      "learning_rate": 1.9395272266365127e-05,
+      "loss": 0.8554,
+      "step": 5855
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3819081440394816,
+      "learning_rate": 1.9395010917123355e-05,
+      "loss": 0.7032,
+      "step": 5856
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.974034350109884,
+      "learning_rate": 1.9394749513180857e-05,
+      "loss": 0.8183,
+      "step": 5857
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8273235485427364,
+      "learning_rate": 1.9394488054539153e-05,
+      "loss": 0.7673,
+      "step": 5858
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4144055996611067,
+      "learning_rate": 1.9394226541199764e-05,
+      "loss": 0.8099,
+      "step": 5859
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5969910961859057,
+      "learning_rate": 1.9393964973164216e-05,
+      "loss": 0.7029,
+      "step": 5860
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4900180060033699,
+      "learning_rate": 1.939370335043403e-05,
+      "loss": 0.807,
+      "step": 5861
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8755862946632498,
+      "learning_rate": 1.9393441673010725e-05,
+      "loss": 0.8784,
+      "step": 5862
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3908640868701287,
+      "learning_rate": 1.939317994089583e-05,
+      "loss": 0.7437,
+      "step": 5863
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.55560965296577,
+      "learning_rate": 1.939291815409087e-05,
+      "loss": 0.7134,
+      "step": 5864
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2965439638783935,
+      "learning_rate": 1.9392656312597365e-05,
+      "loss": 0.6248,
+      "step": 5865
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6686112887631617,
+      "learning_rate": 1.939239441641684e-05,
+      "loss": 0.7584,
+      "step": 5866
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4954958731204422,
+      "learning_rate": 1.9392132465550824e-05,
+      "loss": 0.7102,
+      "step": 5867
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5660407674943695,
+      "learning_rate": 1.939187046000084e-05,
+      "loss": 0.6489,
+      "step": 5868
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5185400712974932,
+      "learning_rate": 1.939160839976841e-05,
+      "loss": 0.7554,
+      "step": 5869
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8416157660655073,
+      "learning_rate": 1.939134628485506e-05,
+      "loss": 0.7406,
+      "step": 5870
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5962852179714235,
+      "learning_rate": 1.9391084115262323e-05,
+      "loss": 0.532,
+      "step": 5871
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6221628043848684,
+      "learning_rate": 1.9390821890991723e-05,
+      "loss": 0.8098,
+      "step": 5872
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5161834624823776,
+      "learning_rate": 1.9390559612044783e-05,
+      "loss": 0.7849,
+      "step": 5873
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.72014435928059,
+      "learning_rate": 1.9390297278423028e-05,
+      "loss": 0.9123,
+      "step": 5874
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7333952794583525,
+      "learning_rate": 1.9390034890127993e-05,
+      "loss": 0.8149,
+      "step": 5875
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0925845063136648,
+      "learning_rate": 1.93897724471612e-05,
+      "loss": 0.7543,
+      "step": 5876
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.769885316051724,
+      "learning_rate": 1.938950994952418e-05,
+      "loss": 0.7758,
+      "step": 5877
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3441266733930226,
+      "learning_rate": 1.938924739721846e-05,
+      "loss": 0.8125,
+      "step": 5878
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5895387872327318,
+      "learning_rate": 1.9388984790245575e-05,
+      "loss": 0.8615,
+      "step": 5879
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5182543719899195,
+      "learning_rate": 1.938872212860704e-05,
+      "loss": 0.8496,
+      "step": 5880
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4241963077887743,
+      "learning_rate": 1.9388459412304397e-05,
+      "loss": 0.7834,
+      "step": 5881
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6114050915436934,
+      "learning_rate": 1.938819664133917e-05,
+      "loss": 0.5406,
+      "step": 5882
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4275006357604778,
+      "learning_rate": 1.938793381571289e-05,
+      "loss": 0.5947,
+      "step": 5883
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.526122798253718,
+      "learning_rate": 1.938767093542708e-05,
+      "loss": 0.7927,
+      "step": 5884
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9353383902262833,
+      "learning_rate": 1.938740800048329e-05,
+      "loss": 0.8203,
+      "step": 5885
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3639967123864831,
+      "learning_rate": 1.938714501088303e-05,
+      "loss": 0.6601,
+      "step": 5886
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.497707871283145,
+      "learning_rate": 1.9386881966627844e-05,
+      "loss": 0.7983,
+      "step": 5887
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.665786140483787,
+      "learning_rate": 1.9386618867719258e-05,
+      "loss": 0.9628,
+      "step": 5888
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5674044804861618,
+      "learning_rate": 1.9386355714158806e-05,
+      "loss": 0.7914,
+      "step": 5889
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6134278530406045,
+      "learning_rate": 1.9386092505948018e-05,
+      "loss": 0.7027,
+      "step": 5890
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5014796529985486,
+      "learning_rate": 1.9385829243088426e-05,
+      "loss": 0.6521,
+      "step": 5891
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5514553708622247,
+      "learning_rate": 1.938556592558157e-05,
+      "loss": 0.6402,
+      "step": 5892
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5938478203638085,
+      "learning_rate": 1.9385302553428974e-05,
+      "loss": 0.7383,
+      "step": 5893
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.827584571816447,
+      "learning_rate": 1.9385039126632176e-05,
+      "loss": 0.8161,
+      "step": 5894
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4472090857529556,
+      "learning_rate": 1.9384775645192707e-05,
+      "loss": 0.686,
+      "step": 5895
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8017069504279841,
+      "learning_rate": 1.9384512109112104e-05,
+      "loss": 0.7278,
+      "step": 5896
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5837794180034431,
+      "learning_rate": 1.9384248518391903e-05,
+      "loss": 0.5011,
+      "step": 5897
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7139265133620405,
+      "learning_rate": 1.9383984873033636e-05,
+      "loss": 0.7218,
+      "step": 5898
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6195817978056926,
+      "learning_rate": 1.9383721173038836e-05,
+      "loss": 0.827,
+      "step": 5899
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.668902212292186,
+      "learning_rate": 1.938345741840904e-05,
+      "loss": 0.7563,
+      "step": 5900
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.956001838028691,
+      "learning_rate": 1.9383193609145786e-05,
+      "loss": 0.9672,
+      "step": 5901
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6150727747027659,
+      "learning_rate": 1.9382929745250605e-05,
+      "loss": 0.4907,
+      "step": 5902
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.828258184760051,
+      "learning_rate": 1.938266582672504e-05,
+      "loss": 0.8938,
+      "step": 5903
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.45345488710173,
+      "learning_rate": 1.9382401853570624e-05,
+      "loss": 0.7292,
+      "step": 5904
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4885721229635007,
+      "learning_rate": 1.938213782578889e-05,
+      "loss": 0.892,
+      "step": 5905
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6789496937463382,
+      "learning_rate": 1.9381873743381384e-05,
+      "loss": 0.849,
+      "step": 5906
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6764003638694278,
+      "learning_rate": 1.9381609606349635e-05,
+      "loss": 0.919,
+      "step": 5907
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4334717285163063,
+      "learning_rate": 1.9381345414695188e-05,
+      "loss": 0.8175,
+      "step": 5908
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0419323250948644,
+      "learning_rate": 1.9381081168419577e-05,
+      "loss": 0.5865,
+      "step": 5909
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.07038709897188,
+      "learning_rate": 1.938081686752434e-05,
+      "loss": 0.8816,
+      "step": 5910
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5279931949188996,
+      "learning_rate": 1.938055251201102e-05,
+      "loss": 0.7446,
+      "step": 5911
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6114147792274356,
+      "learning_rate": 1.9380288101881152e-05,
+      "loss": 0.5877,
+      "step": 5912
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5767685440499308,
+      "learning_rate": 1.9380023637136272e-05,
+      "loss": 0.5646,
+      "step": 5913
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5881451600858711,
+      "learning_rate": 1.937975911777793e-05,
+      "loss": 0.5024,
+      "step": 5914
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1133271326630454,
+      "learning_rate": 1.937949454380766e-05,
+      "loss": 0.7351,
+      "step": 5915
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.640547171517796,
+      "learning_rate": 1.9379229915227006e-05,
+      "loss": 0.7516,
+      "step": 5916
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2986549056230372,
+      "learning_rate": 1.9378965232037503e-05,
+      "loss": 0.7068,
+      "step": 5917
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6454609849213777,
+      "learning_rate": 1.9378700494240697e-05,
+      "loss": 0.8244,
+      "step": 5918
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.586913733959164,
+      "learning_rate": 1.9378435701838127e-05,
+      "loss": 0.7362,
+      "step": 5919
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.330245021024269,
+      "learning_rate": 1.937817085483133e-05,
+      "loss": 0.7506,
+      "step": 5920
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5448450735611727,
+      "learning_rate": 1.937790595322186e-05,
+      "loss": 0.6954,
+      "step": 5921
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6374399928356768,
+      "learning_rate": 1.937764099701125e-05,
+      "loss": 0.58,
+      "step": 5922
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.628639819435197,
+      "learning_rate": 1.9377375986201047e-05,
+      "loss": 0.7236,
+      "step": 5923
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.339148604992148,
+      "learning_rate": 1.9377110920792788e-05,
+      "loss": 0.6467,
+      "step": 5924
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8415862145756827,
+      "learning_rate": 1.937684580078802e-05,
+      "loss": 0.6841,
+      "step": 5925
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6837254817167304,
+      "learning_rate": 1.9376580626188292e-05,
+      "loss": 0.7271,
+      "step": 5926
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5629284120325109,
+      "learning_rate": 1.937631539699514e-05,
+      "loss": 0.6232,
+      "step": 5927
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.510155804571618,
+      "learning_rate": 1.9376050113210113e-05,
+      "loss": 0.6607,
+      "step": 5928
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6075299950786421,
+      "learning_rate": 1.9375784774834754e-05,
+      "loss": 0.5639,
+      "step": 5929
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.607202417914829,
+      "learning_rate": 1.9375519381870608e-05,
+      "loss": 0.8445,
+      "step": 5930
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5887495216875724,
+      "learning_rate": 1.9375253934319217e-05,
+      "loss": 0.6795,
+      "step": 5931
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4138297566717133,
+      "learning_rate": 1.9374988432182133e-05,
+      "loss": 0.7205,
+      "step": 5932
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.572560026059193,
+      "learning_rate": 1.9374722875460894e-05,
+      "loss": 0.6594,
+      "step": 5933
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7224379434648793,
+      "learning_rate": 1.9374457264157055e-05,
+      "loss": 0.8018,
+      "step": 5934
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6136979194177392,
+      "learning_rate": 1.9374191598272154e-05,
+      "loss": 0.5449,
+      "step": 5935
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0466402460647157,
+      "learning_rate": 1.9373925877807744e-05,
+      "loss": 0.7298,
+      "step": 5936
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6420604952366349,
+      "learning_rate": 1.937366010276537e-05,
+      "loss": 0.534,
+      "step": 5937
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6039154562262488,
+      "learning_rate": 1.9373394273146577e-05,
+      "loss": 0.5487,
+      "step": 5938
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5250917381434503,
+      "learning_rate": 1.9373128388952914e-05,
+      "loss": 0.7961,
+      "step": 5939
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5050128123510524,
+      "learning_rate": 1.9372862450185934e-05,
+      "loss": 0.7282,
+      "step": 5940
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6967116604156782,
+      "learning_rate": 1.937259645684718e-05,
+      "loss": 0.8069,
+      "step": 5941
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8569071523504501,
+      "learning_rate": 1.9372330408938203e-05,
+      "loss": 0.7325,
+      "step": 5942
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4587486797950189,
+      "learning_rate": 1.937206430646055e-05,
+      "loss": 0.75,
+      "step": 5943
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6185650745718553,
+      "learning_rate": 1.937179814941577e-05,
+      "loss": 0.5908,
+      "step": 5944
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8880100925029735,
+      "learning_rate": 1.9371531937805417e-05,
+      "loss": 0.7609,
+      "step": 5945
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6512915097348495,
+      "learning_rate": 1.9371265671631038e-05,
+      "loss": 0.5748,
+      "step": 5946
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5370538396410873,
+      "learning_rate": 1.9370999350894178e-05,
+      "loss": 0.8436,
+      "step": 5947
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6232153122944348,
+      "learning_rate": 1.93707329755964e-05,
+      "loss": 0.8066,
+      "step": 5948
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6409043832730872,
+      "learning_rate": 1.9370466545739246e-05,
+      "loss": 0.7379,
+      "step": 5949
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6136062275415741,
+      "learning_rate": 1.9370200061324267e-05,
+      "loss": 0.857,
+      "step": 5950
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7037247599332446,
+      "learning_rate": 1.9369933522353017e-05,
+      "loss": 0.7771,
+      "step": 5951
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8995970725415638,
+      "learning_rate": 1.936966692882705e-05,
+      "loss": 0.9504,
+      "step": 5952
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6245905496592368,
+      "learning_rate": 1.9369400280747913e-05,
+      "loss": 0.7883,
+      "step": 5953
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.536899867720473,
+      "learning_rate": 1.9369133578117162e-05,
+      "loss": 0.6923,
+      "step": 5954
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5981976559249997,
+      "learning_rate": 1.936886682093635e-05,
+      "loss": 0.7649,
+      "step": 5955
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5891022675564829,
+      "learning_rate": 1.9368600009207027e-05,
+      "loss": 0.5316,
+      "step": 5956
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6774364372365564,
+      "learning_rate": 1.9368333142930754e-05,
+      "loss": 0.7141,
+      "step": 5957
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4112379117456215,
+      "learning_rate": 1.9368066222109073e-05,
+      "loss": 0.7857,
+      "step": 5958
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6294368506501746,
+      "learning_rate": 1.936779924674355e-05,
+      "loss": 0.5755,
+      "step": 5959
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.562972612879512,
+      "learning_rate": 1.936753221683573e-05,
+      "loss": 0.6578,
+      "step": 5960
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4463480914455702,
+      "learning_rate": 1.936726513238717e-05,
+      "loss": 0.6973,
+      "step": 5961
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6008894558520922,
+      "learning_rate": 1.9366997993399432e-05,
+      "loss": 0.5573,
+      "step": 5962
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6165611620780784,
+      "learning_rate": 1.9366730799874063e-05,
+      "loss": 0.554,
+      "step": 5963
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.693449258352363,
+      "learning_rate": 1.9366463551812624e-05,
+      "loss": 0.8936,
+      "step": 5964
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4131321705466504,
+      "learning_rate": 1.936619624921667e-05,
+      "loss": 0.7435,
+      "step": 5965
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0045172428068865,
+      "learning_rate": 1.936592889208775e-05,
+      "loss": 0.6942,
+      "step": 5966
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2470901020477296,
+      "learning_rate": 1.936566148042743e-05,
+      "loss": 0.7831,
+      "step": 5967
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3345990096288411,
+      "learning_rate": 1.9365394014237268e-05,
+      "loss": 0.6976,
+      "step": 5968
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.477021935817729,
+      "learning_rate": 1.9365126493518813e-05,
+      "loss": 0.6539,
+      "step": 5969
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.861818294886225,
+      "learning_rate": 1.9364858918273626e-05,
+      "loss": 0.8602,
+      "step": 5970
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.630015028720785,
+      "learning_rate": 1.9364591288503267e-05,
+      "loss": 0.7574,
+      "step": 5971
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.529643739187247,
+      "learning_rate": 1.9364323604209292e-05,
+      "loss": 0.6974,
+      "step": 5972
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.381499085706771,
+      "learning_rate": 1.936405586539326e-05,
+      "loss": 0.8574,
+      "step": 5973
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9366639446267004,
+      "learning_rate": 1.936378807205673e-05,
+      "loss": 0.7598,
+      "step": 5974
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5776271243745967,
+      "learning_rate": 1.9363520224201263e-05,
+      "loss": 0.7822,
+      "step": 5975
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5978744489720459,
+      "learning_rate": 1.9363252321828417e-05,
+      "loss": 0.5439,
+      "step": 5976
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.56093293275497,
+      "learning_rate": 1.936298436493975e-05,
+      "loss": 0.8591,
+      "step": 5977
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4880489053270387,
+      "learning_rate": 1.9362716353536825e-05,
+      "loss": 0.8383,
+      "step": 5978
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5568294620008352,
+      "learning_rate": 1.93624482876212e-05,
+      "loss": 0.6882,
+      "step": 5979
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1013994439360877,
+      "learning_rate": 1.9362180167194435e-05,
+      "loss": 0.7956,
+      "step": 5980
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4013576705685058,
+      "learning_rate": 1.9361911992258096e-05,
+      "loss": 0.8408,
+      "step": 5981
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7458894587136382,
+      "learning_rate": 1.936164376281374e-05,
+      "loss": 0.8475,
+      "step": 5982
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3081068638628246,
+      "learning_rate": 1.9361375478862933e-05,
+      "loss": 0.7912,
+      "step": 5983
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6169930763271128,
+      "learning_rate": 1.9361107140407233e-05,
+      "loss": 0.7252,
+      "step": 5984
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6292715465349166,
+      "learning_rate": 1.9360838747448202e-05,
+      "loss": 0.7418,
+      "step": 5985
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6700916618935857,
+      "learning_rate": 1.9360570299987405e-05,
+      "loss": 0.8013,
+      "step": 5986
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8205924551347208,
+      "learning_rate": 1.9360301798026404e-05,
+      "loss": 0.7696,
+      "step": 5987
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.692962519493975,
+      "learning_rate": 1.9360033241566764e-05,
+      "loss": 0.8111,
+      "step": 5988
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6224050365079461,
+      "learning_rate": 1.9359764630610044e-05,
+      "loss": 0.5526,
+      "step": 5989
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5893729645671415,
+      "learning_rate": 1.9359495965157814e-05,
+      "loss": 0.7286,
+      "step": 5990
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5067722969374144,
+      "learning_rate": 1.9359227245211633e-05,
+      "loss": 0.8163,
+      "step": 5991
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.513044774329013,
+      "learning_rate": 1.935895847077307e-05,
+      "loss": 0.6659,
+      "step": 5992
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6605055439463152,
+      "learning_rate": 1.9358689641843686e-05,
+      "loss": 0.7086,
+      "step": 5993
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7505728746817204,
+      "learning_rate": 1.935842075842505e-05,
+      "loss": 0.7687,
+      "step": 5994
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4684115824814954,
+      "learning_rate": 1.9358151820518725e-05,
+      "loss": 0.7249,
+      "step": 5995
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4781986676575156,
+      "learning_rate": 1.9357882828126278e-05,
+      "loss": 0.8206,
+      "step": 5996
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6329164013058917,
+      "learning_rate": 1.935761378124927e-05,
+      "loss": 0.729,
+      "step": 5997
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.975122343735426,
+      "learning_rate": 1.935734467988928e-05,
+      "loss": 0.6086,
+      "step": 5998
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7491416878032588,
+      "learning_rate": 1.935707552404786e-05,
+      "loss": 0.6768,
+      "step": 5999
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5699554840658323,
+      "learning_rate": 1.9356806313726584e-05,
+      "loss": 0.7995,
+      "step": 6000
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.653597499007351,
+      "learning_rate": 1.9356537048927023e-05,
+      "loss": 0.7311,
+      "step": 6001
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4061205149826068,
+      "learning_rate": 1.9356267729650738e-05,
+      "loss": 0.7496,
+      "step": 6002
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6301055032975831,
+      "learning_rate": 1.93559983558993e-05,
+      "loss": 0.5452,
+      "step": 6003
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.6025498664400892,
+      "learning_rate": 1.9355728927674284e-05,
+      "loss": 0.6579,
+      "step": 6004
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.48773012651187,
+      "learning_rate": 1.9355459444977245e-05,
+      "loss": 0.823,
+      "step": 6005
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.581896167111591,
+      "learning_rate": 1.935518990780976e-05,
+      "loss": 0.5565,
+      "step": 6006
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.552565342715294,
+      "learning_rate": 1.93549203161734e-05,
+      "loss": 0.6506,
+      "step": 6007
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.701005710391996,
+      "learning_rate": 1.9354650670069732e-05,
+      "loss": 0.7639,
+      "step": 6008
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6614734339502975,
+      "learning_rate": 1.9354380969500326e-05,
+      "loss": 0.5757,
+      "step": 6009
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2860582446335485,
+      "learning_rate": 1.9354111214466752e-05,
+      "loss": 0.6428,
+      "step": 6010
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4749267787177833,
+      "learning_rate": 1.9353841404970582e-05,
+      "loss": 0.6989,
+      "step": 6011
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.94339541878878,
+      "learning_rate": 1.9353571541013382e-05,
+      "loss": 0.772,
+      "step": 6012
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5208270284461072,
+      "learning_rate": 1.935330162259673e-05,
+      "loss": 0.7189,
+      "step": 6013
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5865485820803802,
+      "learning_rate": 1.9353031649722196e-05,
+      "loss": 0.8667,
+      "step": 6014
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5914522285266643,
+      "learning_rate": 1.9352761622391348e-05,
+      "loss": 0.7379,
+      "step": 6015
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6110102799696588,
+      "learning_rate": 1.935249154060576e-05,
+      "loss": 0.8176,
+      "step": 6016
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4908793892497378,
+      "learning_rate": 1.9352221404367012e-05,
+      "loss": 0.6741,
+      "step": 6017
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4649148628735682,
+      "learning_rate": 1.9351951213676664e-05,
+      "loss": 0.9667,
+      "step": 6018
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5679527395274702,
+      "learning_rate": 1.9351680968536298e-05,
+      "loss": 0.6968,
+      "step": 6019
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6353140724280053,
+      "learning_rate": 1.9351410668947482e-05,
+      "loss": 0.8031,
+      "step": 6020
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7461329473895821,
+      "learning_rate": 1.9351140314911798e-05,
+      "loss": 0.8078,
+      "step": 6021
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6926467358537545,
+      "learning_rate": 1.9350869906430807e-05,
+      "loss": 0.7246,
+      "step": 6022
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.315729153101309,
+      "learning_rate": 1.9350599443506097e-05,
+      "loss": 0.8227,
+      "step": 6023
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5486993134142832,
+      "learning_rate": 1.935032892613923e-05,
+      "loss": 0.7163,
+      "step": 6024
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5024239186452222,
+      "learning_rate": 1.9350058354331796e-05,
+      "loss": 0.7575,
+      "step": 6025
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5552938025020417,
+      "learning_rate": 1.9349787728085357e-05,
+      "loss": 0.7973,
+      "step": 6026
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5849926284635188,
+      "learning_rate": 1.9349517047401493e-05,
+      "loss": 0.5515,
+      "step": 6027
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5906756573145109,
+      "learning_rate": 1.9349246312281783e-05,
+      "loss": 0.7351,
+      "step": 6028
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8248320568245027,
+      "learning_rate": 1.9348975522727802e-05,
+      "loss": 0.8194,
+      "step": 6029
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.009998223285972,
+      "learning_rate": 1.9348704678741125e-05,
+      "loss": 0.7518,
+      "step": 6030
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0472809672866874,
+      "learning_rate": 1.934843378032333e-05,
+      "loss": 0.8355,
+      "step": 6031
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3880838185234703,
+      "learning_rate": 1.934816282747599e-05,
+      "loss": 0.6772,
+      "step": 6032
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5764543998196765,
+      "learning_rate": 1.9347891820200692e-05,
+      "loss": 0.7274,
+      "step": 6033
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6916498104581141,
+      "learning_rate": 1.9347620758499003e-05,
+      "loss": 0.7548,
+      "step": 6034
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0001537417836865,
+      "learning_rate": 1.9347349642372512e-05,
+      "loss": 0.8516,
+      "step": 6035
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.603111943783951,
+      "learning_rate": 1.9347078471822786e-05,
+      "loss": 0.6657,
+      "step": 6036
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5804146961991488,
+      "learning_rate": 1.9346807246851417e-05,
+      "loss": 0.7755,
+      "step": 6037
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4780148613977326,
+      "learning_rate": 1.9346535967459973e-05,
+      "loss": 0.7627,
+      "step": 6038
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9641952884188851,
+      "learning_rate": 1.9346264633650036e-05,
+      "loss": 0.7973,
+      "step": 6039
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8867629958261476,
+      "learning_rate": 1.934599324542319e-05,
+      "loss": 0.8469,
+      "step": 6040
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.280011538114615,
+      "learning_rate": 1.934572180278101e-05,
+      "loss": 0.7684,
+      "step": 6041
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5911116965631686,
+      "learning_rate": 1.934545030572508e-05,
+      "loss": 0.7343,
+      "step": 6042
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0153721692662123,
+      "learning_rate": 1.934517875425698e-05,
+      "loss": 0.7549,
+      "step": 6043
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6482922669735167,
+      "learning_rate": 1.934490714837829e-05,
+      "loss": 0.8977,
+      "step": 6044
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6341472751308852,
+      "learning_rate": 1.9344635488090593e-05,
+      "loss": 0.8185,
+      "step": 6045
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6027622845976471,
+      "learning_rate": 1.934436377339547e-05,
+      "loss": 0.5203,
+      "step": 6046
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.407120478696364,
+      "learning_rate": 1.9344092004294503e-05,
+      "loss": 0.7447,
+      "step": 6047
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.867989646696343,
+      "learning_rate": 1.9343820180789273e-05,
+      "loss": 0.8171,
+      "step": 6048
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6370806723841356,
+      "learning_rate": 1.9343548302881364e-05,
+      "loss": 0.8297,
+      "step": 6049
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4741528467114533,
+      "learning_rate": 1.9343276370572357e-05,
+      "loss": 0.8412,
+      "step": 6050
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6167896407675084,
+      "learning_rate": 1.9343004383863838e-05,
+      "loss": 0.539,
+      "step": 6051
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5180768594576708,
+      "learning_rate": 1.9342732342757392e-05,
+      "loss": 0.7074,
+      "step": 6052
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4843058730309995,
+      "learning_rate": 1.9342460247254596e-05,
+      "loss": 0.7395,
+      "step": 6053
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8130139640882723,
+      "learning_rate": 1.9342188097357042e-05,
+      "loss": 0.7479,
+      "step": 6054
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.292785056853019,
+      "learning_rate": 1.934191589306631e-05,
+      "loss": 0.8301,
+      "step": 6055
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4696135268554307,
+      "learning_rate": 1.9341643634383986e-05,
+      "loss": 0.9027,
+      "step": 6056
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7288477271913165,
+      "learning_rate": 1.9341371321311654e-05,
+      "loss": 0.8682,
+      "step": 6057
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4422144693933794,
+      "learning_rate": 1.9341098953850905e-05,
+      "loss": 0.7451,
+      "step": 6058
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6442687861858892,
+      "learning_rate": 1.934082653200332e-05,
+      "loss": 0.628,
+      "step": 6059
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3986855306107668,
+      "learning_rate": 1.934055405577048e-05,
+      "loss": 0.7279,
+      "step": 6060
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.4119594948382312,
+      "learning_rate": 1.9340281525153983e-05,
+      "loss": 0.6063,
+      "step": 6061
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5023596302679787,
+      "learning_rate": 1.9340008940155407e-05,
+      "loss": 0.7797,
+      "step": 6062
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.517886109121152,
+      "learning_rate": 1.933973630077634e-05,
+      "loss": 0.7848,
+      "step": 6063
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6122103022355874,
+      "learning_rate": 1.9339463607018375e-05,
+      "loss": 0.5289,
+      "step": 6064
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5143156680939156,
+      "learning_rate": 1.9339190858883096e-05,
+      "loss": 0.7902,
+      "step": 6065
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.489948634264315,
+      "learning_rate": 1.933891805637209e-05,
+      "loss": 0.7744,
+      "step": 6066
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.725289485360558,
+      "learning_rate": 1.9338645199486947e-05,
+      "loss": 0.8501,
+      "step": 6067
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5807935447956842,
+      "learning_rate": 1.9338372288229253e-05,
+      "loss": 0.9715,
+      "step": 6068
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6046657194607432,
+      "learning_rate": 1.93380993226006e-05,
+      "loss": 0.7714,
+      "step": 6069
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.776268532790159,
+      "learning_rate": 1.933782630260258e-05,
+      "loss": 0.6877,
+      "step": 6070
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5932131558158127,
+      "learning_rate": 1.9337553228236772e-05,
+      "loss": 0.5329,
+      "step": 6071
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4754706037248753,
+      "learning_rate": 1.933728009950478e-05,
+      "loss": 0.7422,
+      "step": 6072
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5262477965958006,
+      "learning_rate": 1.9337006916408182e-05,
+      "loss": 0.8654,
+      "step": 6073
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.473766553121725,
+      "learning_rate": 1.9336733678948577e-05,
+      "loss": 0.736,
+      "step": 6074
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.6127070167757016,
+      "learning_rate": 1.933646038712755e-05,
+      "loss": 0.7995,
+      "step": 6075
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8473709311965008,
+      "learning_rate": 1.9336187040946695e-05,
+      "loss": 0.6986,
+      "step": 6076
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9107213841435335,
+      "learning_rate": 1.9335913640407606e-05,
+      "loss": 0.864,
+      "step": 6077
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4829745137880463,
+      "learning_rate": 1.9335640185511872e-05,
+      "loss": 0.7578,
+      "step": 6078
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5024518224112904,
+      "learning_rate": 1.933536667626108e-05,
+      "loss": 0.8331,
+      "step": 6079
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8935834161137366,
+      "learning_rate": 1.933509311265683e-05,
+      "loss": 0.7442,
+      "step": 6080
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5709490273848774,
+      "learning_rate": 1.9334819494700714e-05,
+      "loss": 0.7052,
+      "step": 6081
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5972367209918842,
+      "learning_rate": 1.9334545822394327e-05,
+      "loss": 0.5044,
+      "step": 6082
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5568791282136596,
+      "learning_rate": 1.9334272095739252e-05,
+      "loss": 0.8657,
+      "step": 6083
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4808515219838985,
+      "learning_rate": 1.9333998314737093e-05,
+      "loss": 0.616,
+      "step": 6084
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5651697746683237,
+      "learning_rate": 1.9333724479389443e-05,
+      "loss": 0.869,
+      "step": 6085
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4461571225161753,
+      "learning_rate": 1.9333450589697887e-05,
+      "loss": 0.7261,
+      "step": 6086
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8368383404628956,
+      "learning_rate": 1.9333176645664035e-05,
+      "loss": 0.872,
+      "step": 6087
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.416503472313326,
+      "learning_rate": 1.9332902647289468e-05,
+      "loss": 0.6908,
+      "step": 6088
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3897658924211267,
+      "learning_rate": 1.9332628594575792e-05,
+      "loss": 0.7175,
+      "step": 6089
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.712742505496696,
+      "learning_rate": 1.933235448752459e-05,
+      "loss": 0.7112,
+      "step": 6090
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8481310707793257,
+      "learning_rate": 1.9332080326137476e-05,
+      "loss": 0.6992,
+      "step": 6091
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4877028111370663,
+      "learning_rate": 1.9331806110416027e-05,
+      "loss": 0.8687,
+      "step": 6092
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4979288388833587,
+      "learning_rate": 1.9331531840361852e-05,
+      "loss": 0.7774,
+      "step": 6093
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.724868496548334,
+      "learning_rate": 1.9331257515976545e-05,
+      "loss": 0.747,
+      "step": 6094
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6459424407827525,
+      "learning_rate": 1.93309831372617e-05,
+      "loss": 0.7571,
+      "step": 6095
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7857795511206855,
+      "learning_rate": 1.933070870421892e-05,
+      "loss": 0.7449,
+      "step": 6096
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6354513953802078,
+      "learning_rate": 1.9330434216849796e-05,
+      "loss": 0.5456,
+      "step": 6097
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5152302650063922,
+      "learning_rate": 1.9330159675155933e-05,
+      "loss": 0.6582,
+      "step": 6098
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6125052739924116,
+      "learning_rate": 1.9329885079138927e-05,
+      "loss": 0.5821,
+      "step": 6099
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4832645284932775,
+      "learning_rate": 1.9329610428800376e-05,
+      "loss": 0.841,
+      "step": 6100
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4863592570375754,
+      "learning_rate": 1.9329335724141877e-05,
+      "loss": 0.6265,
+      "step": 6101
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3580959921878795,
+      "learning_rate": 1.9329060965165032e-05,
+      "loss": 0.7703,
+      "step": 6102
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9409050880310617,
+      "learning_rate": 1.9328786151871444e-05,
+      "loss": 0.8372,
+      "step": 6103
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4917049176455333,
+      "learning_rate": 1.9328511284262705e-05,
+      "loss": 0.7547,
+      "step": 6104
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.441873183475497,
+      "learning_rate": 1.9328236362340425e-05,
+      "loss": 0.7171,
+      "step": 6105
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4295305567531862,
+      "learning_rate": 1.9327961386106198e-05,
+      "loss": 0.813,
+      "step": 6106
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4711162969214098,
+      "learning_rate": 1.9327686355561622e-05,
+      "loss": 0.8005,
+      "step": 6107
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.466367539057994,
+      "learning_rate": 1.932741127070831e-05,
+      "loss": 0.7884,
+      "step": 6108
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.460501172072052,
+      "learning_rate": 1.932713613154785e-05,
+      "loss": 0.7616,
+      "step": 6109
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.583629681808435,
+      "learning_rate": 1.9326860938081854e-05,
+      "loss": 0.7043,
+      "step": 6110
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4961936551859059,
+      "learning_rate": 1.932658569031192e-05,
+      "loss": 0.756,
+      "step": 6111
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6642379906852605,
+      "learning_rate": 1.9326310388239654e-05,
+      "loss": 0.7854,
+      "step": 6112
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.246566924900102,
+      "learning_rate": 1.9326035031866653e-05,
+      "loss": 0.7892,
+      "step": 6113
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.109604667014194,
+      "learning_rate": 1.9325759621194528e-05,
+      "loss": 0.7441,
+      "step": 6114
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.543572026321755,
+      "learning_rate": 1.9325484156224872e-05,
+      "loss": 0.7221,
+      "step": 6115
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4320675404496277,
+      "learning_rate": 1.9325208636959297e-05,
+      "loss": 0.7961,
+      "step": 6116
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8132822482096385,
+      "learning_rate": 1.932493306339941e-05,
+      "loss": 0.7819,
+      "step": 6117
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.505996736664684,
+      "learning_rate": 1.9324657435546802e-05,
+      "loss": 0.7405,
+      "step": 6118
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4371375966667659,
+      "learning_rate": 1.932438175340309e-05,
+      "loss": 0.7741,
+      "step": 6119
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4762684410555282,
+      "learning_rate": 1.9324106016969875e-05,
+      "loss": 0.7926,
+      "step": 6120
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.009638643060364,
+      "learning_rate": 1.9323830226248763e-05,
+      "loss": 0.7355,
+      "step": 6121
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.58365689871451,
+      "learning_rate": 1.932355438124136e-05,
+      "loss": 0.7667,
+      "step": 6122
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4030551271393663,
+      "learning_rate": 1.9323278481949274e-05,
+      "loss": 0.709,
+      "step": 6123
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.610650083056216,
+      "learning_rate": 1.9323002528374104e-05,
+      "loss": 0.7814,
+      "step": 6124
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.822232390172355,
+      "learning_rate": 1.9322726520517465e-05,
+      "loss": 0.7104,
+      "step": 6125
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5563183426127227,
+      "learning_rate": 1.932245045838096e-05,
+      "loss": 0.7578,
+      "step": 6126
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6773173962732015,
+      "learning_rate": 1.9322174341966196e-05,
+      "loss": 0.8818,
+      "step": 6127
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6666596082083738,
+      "learning_rate": 1.9321898171274785e-05,
+      "loss": 0.6635,
+      "step": 6128
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5910916607618535,
+      "learning_rate": 1.9321621946308328e-05,
+      "loss": 0.8473,
+      "step": 6129
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7850379152884321,
+      "learning_rate": 1.9321345667068435e-05,
+      "loss": 0.8009,
+      "step": 6130
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.708388425426282,
+      "learning_rate": 1.9321069333556722e-05,
+      "loss": 0.8569,
+      "step": 6131
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5613162387723112,
+      "learning_rate": 1.9320792945774788e-05,
+      "loss": 0.7285,
+      "step": 6132
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4889098053557697,
+      "learning_rate": 1.932051650372425e-05,
+      "loss": 0.8055,
+      "step": 6133
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6622238298025835,
+      "learning_rate": 1.932024000740671e-05,
+      "loss": 0.87,
+      "step": 6134
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5948075041245987,
+      "learning_rate": 1.9319963456823788e-05,
+      "loss": 0.5432,
+      "step": 6135
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9927758205124864,
+      "learning_rate": 1.931968685197708e-05,
+      "loss": 0.7305,
+      "step": 6136
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4543766867341543,
+      "learning_rate": 1.931941019286821e-05,
+      "loss": 0.6666,
+      "step": 6137
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4544554999714494,
+      "learning_rate": 1.931913347949878e-05,
+      "loss": 0.7808,
+      "step": 6138
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5406175462873921,
+      "learning_rate": 1.931885671187041e-05,
+      "loss": 0.785,
+      "step": 6139
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9634005600488915,
+      "learning_rate": 1.9318579889984705e-05,
+      "loss": 0.6978,
+      "step": 6140
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7550517785748632,
+      "learning_rate": 1.9318303013843272e-05,
+      "loss": 0.8076,
+      "step": 6141
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5696583465863002,
+      "learning_rate": 1.9318026083447733e-05,
+      "loss": 0.5572,
+      "step": 6142
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7163879578566283,
+      "learning_rate": 1.9317749098799695e-05,
+      "loss": 0.7464,
+      "step": 6143
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.397183686586809,
+      "learning_rate": 1.9317472059900773e-05,
+      "loss": 0.7138,
+      "step": 6144
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5195989224135384,
+      "learning_rate": 1.9317194966752576e-05,
+      "loss": 0.8206,
+      "step": 6145
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.711106593984785,
+      "learning_rate": 1.9316917819356723e-05,
+      "loss": 0.6965,
+      "step": 6146
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.575152110741561,
+      "learning_rate": 1.9316640617714823e-05,
+      "loss": 0.519,
+      "step": 6147
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5510816453373624,
+      "learning_rate": 1.9316363361828493e-05,
+      "loss": 0.7975,
+      "step": 6148
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7566299897658448,
+      "learning_rate": 1.9316086051699345e-05,
+      "loss": 0.7577,
+      "step": 6149
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5442444892544538,
+      "learning_rate": 1.9315808687328996e-05,
+      "loss": 0.7152,
+      "step": 6150
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5723526945443183,
+      "learning_rate": 1.9315531268719058e-05,
+      "loss": 0.7868,
+      "step": 6151
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6352755406529877,
+      "learning_rate": 1.931525379587115e-05,
+      "loss": 0.9054,
+      "step": 6152
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7809228868125988,
+      "learning_rate": 1.9314976268786883e-05,
+      "loss": 0.7845,
+      "step": 6153
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5761391598472508,
+      "learning_rate": 1.9314698687467876e-05,
+      "loss": 0.536,
+      "step": 6154
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5992696443312859,
+      "learning_rate": 1.9314421051915744e-05,
+      "loss": 0.5502,
+      "step": 6155
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4136367279295565,
+      "learning_rate": 1.93141433621321e-05,
+      "loss": 0.735,
+      "step": 6156
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6049905720953583,
+      "learning_rate": 1.931386561811857e-05,
+      "loss": 0.7212,
+      "step": 6157
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5800963700127315,
+      "learning_rate": 1.9313587819876765e-05,
+      "loss": 0.873,
+      "step": 6158
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.425204536655689,
+      "learning_rate": 1.93133099674083e-05,
+      "loss": 0.7776,
+      "step": 6159
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3831273338831107,
+      "learning_rate": 1.9313032060714797e-05,
+      "loss": 0.8124,
+      "step": 6160
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4114150630841822,
+      "learning_rate": 1.9312754099797876e-05,
+      "loss": 0.8953,
+      "step": 6161
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6514123196013613,
+      "learning_rate": 1.931247608465915e-05,
+      "loss": 0.8548,
+      "step": 6162
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3823478072816597,
+      "learning_rate": 1.931219801530024e-05,
+      "loss": 0.6001,
+      "step": 6163
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5243736733268944,
+      "learning_rate": 1.9311919891722764e-05,
+      "loss": 0.7831,
+      "step": 6164
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4025850884419087,
+      "learning_rate": 1.9311641713928342e-05,
+      "loss": 0.7946,
+      "step": 6165
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3566060006018663,
+      "learning_rate": 1.9311363481918594e-05,
+      "loss": 0.8047,
+      "step": 6166
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6419086996037465,
+      "learning_rate": 1.9311085195695142e-05,
+      "loss": 0.5354,
+      "step": 6167
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4561583683068884,
+      "learning_rate": 1.9310806855259604e-05,
+      "loss": 0.7534,
+      "step": 6168
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6028953683289162,
+      "learning_rate": 1.93105284606136e-05,
+      "loss": 0.5098,
+      "step": 6169
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5607145321438634,
+      "learning_rate": 1.9310250011758752e-05,
+      "loss": 0.8391,
+      "step": 6170
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.707391257901064,
+      "learning_rate": 1.9309971508696675e-05,
+      "loss": 0.6696,
+      "step": 6171
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2048799533923242,
+      "learning_rate": 1.9309692951429003e-05,
+      "loss": 0.8022,
+      "step": 6172
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5214153993471051,
+      "learning_rate": 1.9309414339957355e-05,
+      "loss": 0.7076,
+      "step": 6173
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4344902068550658,
+      "learning_rate": 1.9309135674283343e-05,
+      "loss": 0.7073,
+      "step": 6174
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6703129328612525,
+      "learning_rate": 1.9308856954408594e-05,
+      "loss": 0.8499,
+      "step": 6175
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8238699141585535,
+      "learning_rate": 1.9308578180334737e-05,
+      "loss": 0.8598,
+      "step": 6176
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.566774520227379,
+      "learning_rate": 1.930829935206339e-05,
+      "loss": 0.5107,
+      "step": 6177
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4442497871567914,
+      "learning_rate": 1.9308020469596177e-05,
+      "loss": 0.7587,
+      "step": 6178
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5048822449213353,
+      "learning_rate": 1.930774153293472e-05,
+      "loss": 0.7677,
+      "step": 6179
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6669617512992414,
+      "learning_rate": 1.9307462542080646e-05,
+      "loss": 0.7411,
+      "step": 6180
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.126922343596367,
+      "learning_rate": 1.930718349703558e-05,
+      "loss": 0.6932,
+      "step": 6181
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4587779797973701,
+      "learning_rate": 1.9306904397801144e-05,
+      "loss": 0.8138,
+      "step": 6182
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7000955814695617,
+      "learning_rate": 1.9306625244378964e-05,
+      "loss": 0.6379,
+      "step": 6183
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6691179136212908,
+      "learning_rate": 1.9306346036770667e-05,
+      "loss": 0.7557,
+      "step": 6184
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4075343347368923,
+      "learning_rate": 1.9306066774977875e-05,
+      "loss": 0.755,
+      "step": 6185
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3950635263572504,
+      "learning_rate": 1.930578745900222e-05,
+      "loss": 0.7909,
+      "step": 6186
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5937236757674154,
+      "learning_rate": 1.930550808884532e-05,
+      "loss": 0.5392,
+      "step": 6187
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6299900197487271,
+      "learning_rate": 1.9305228664508805e-05,
+      "loss": 0.7478,
+      "step": 6188
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.575332798682867,
+      "learning_rate": 1.9304949185994303e-05,
+      "loss": 0.6696,
+      "step": 6189
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4407249570941116,
+      "learning_rate": 1.9304669653303444e-05,
+      "loss": 0.7117,
+      "step": 6190
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4509101172492458,
+      "learning_rate": 1.930439006643785e-05,
+      "loss": 0.792,
+      "step": 6191
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5420121518880168,
+      "learning_rate": 1.930411042539915e-05,
+      "loss": 0.6554,
+      "step": 6192
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6179289820678313,
+      "learning_rate": 1.930383073018898e-05,
+      "loss": 0.5831,
+      "step": 6193
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0474061951067744,
+      "learning_rate": 1.9303550980808956e-05,
+      "loss": 0.6829,
+      "step": 6194
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.551852853621857,
+      "learning_rate": 1.9303271177260715e-05,
+      "loss": 0.716,
+      "step": 6195
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5462176159636605,
+      "learning_rate": 1.9302991319545883e-05,
+      "loss": 0.9185,
+      "step": 6196
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.498506867383575,
+      "learning_rate": 1.930271140766609e-05,
+      "loss": 0.844,
+      "step": 6197
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4637160828810765,
+      "learning_rate": 1.930243144162297e-05,
+      "loss": 0.7511,
+      "step": 6198
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5795382122942208,
+      "learning_rate": 1.930215142141814e-05,
+      "loss": 0.9222,
+      "step": 6199
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.49692404663763,
+      "learning_rate": 1.930187134705325e-05,
+      "loss": 0.7803,
+      "step": 6200
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4862018890635804,
+      "learning_rate": 1.930159121852991e-05,
+      "loss": 0.7614,
+      "step": 6201
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.183605624470952,
+      "learning_rate": 1.930131103584977e-05,
+      "loss": 0.6653,
+      "step": 6202
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0449156578973344,
+      "learning_rate": 1.9301030799014445e-05,
+      "loss": 0.7282,
+      "step": 6203
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5770069676686924,
+      "learning_rate": 1.9300750508025577e-05,
+      "loss": 0.781,
+      "step": 6204
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6517361557810417,
+      "learning_rate": 1.93004701628848e-05,
+      "loss": 0.7262,
+      "step": 6205
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5877935133519109,
+      "learning_rate": 1.9300189763593736e-05,
+      "loss": 0.5153,
+      "step": 6206
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5999319110668215,
+      "learning_rate": 1.9299909310154025e-05,
+      "loss": 0.8041,
+      "step": 6207
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.284206717326583,
+      "learning_rate": 1.9299628802567293e-05,
+      "loss": 0.817,
+      "step": 6208
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5770643479226205,
+      "learning_rate": 1.9299348240835182e-05,
+      "loss": 0.725,
+      "step": 6209
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5654515013372956,
+      "learning_rate": 1.9299067624959322e-05,
+      "loss": 0.681,
+      "step": 6210
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.44304241387293,
+      "learning_rate": 1.9298786954941344e-05,
+      "loss": 0.7233,
+      "step": 6211
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7467778005073942,
+      "learning_rate": 1.9298506230782886e-05,
+      "loss": 0.7067,
+      "step": 6212
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5686541292081033,
+      "learning_rate": 1.929822545248558e-05,
+      "loss": 0.7876,
+      "step": 6213
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6338583219409488,
+      "learning_rate": 1.9297944620051063e-05,
+      "loss": 0.7897,
+      "step": 6214
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8007967936218379,
+      "learning_rate": 1.9297663733480966e-05,
+      "loss": 0.7865,
+      "step": 6215
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.246191857530271,
+      "learning_rate": 1.929738279277693e-05,
+      "loss": 0.7143,
+      "step": 6216
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6040329510401203,
+      "learning_rate": 1.9297101797940586e-05,
+      "loss": 0.7492,
+      "step": 6217
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6035923089523518,
+      "learning_rate": 1.9296820748973572e-05,
+      "loss": 0.7798,
+      "step": 6218
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7423729320590366,
+      "learning_rate": 1.9296539645877524e-05,
+      "loss": 0.7635,
+      "step": 6219
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8797698214559588,
+      "learning_rate": 1.9296258488654082e-05,
+      "loss": 0.7811,
+      "step": 6220
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5519482806943214,
+      "learning_rate": 1.9295977277304875e-05,
+      "loss": 0.7641,
+      "step": 6221
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6712309561621035,
+      "learning_rate": 1.9295696011831547e-05,
+      "loss": 0.7559,
+      "step": 6222
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5691452741932421,
+      "learning_rate": 1.9295414692235738e-05,
+      "loss": 0.8726,
+      "step": 6223
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5843602368795204,
+      "learning_rate": 1.9295133318519077e-05,
+      "loss": 0.5318,
+      "step": 6224
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4375067656174143,
+      "learning_rate": 1.9294851890683208e-05,
+      "loss": 0.8048,
+      "step": 6225
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6287296235418487,
+      "learning_rate": 1.929457040872977e-05,
+      "loss": 0.5426,
+      "step": 6226
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5939971780777711,
+      "learning_rate": 1.92942888726604e-05,
+      "loss": 0.556,
+      "step": 6227
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5810707821540027,
+      "learning_rate": 1.9294007282476737e-05,
+      "loss": 0.7087,
+      "step": 6228
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6862779540915007,
+      "learning_rate": 1.929372563818042e-05,
+      "loss": 0.8328,
+      "step": 6229
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1032998754566115,
+      "learning_rate": 1.9293443939773092e-05,
+      "loss": 0.8359,
+      "step": 6230
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5963157270157053,
+      "learning_rate": 1.9293162187256392e-05,
+      "loss": 0.7255,
+      "step": 6231
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5628696903774613,
+      "learning_rate": 1.9292880380631956e-05,
+      "loss": 0.8204,
+      "step": 6232
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4666136958892866,
+      "learning_rate": 1.929259851990143e-05,
+      "loss": 0.7012,
+      "step": 6233
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.796036915280256,
+      "learning_rate": 1.9292316605066455e-05,
+      "loss": 0.7908,
+      "step": 6234
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4681188164490089,
+      "learning_rate": 1.929203463612867e-05,
+      "loss": 0.6564,
+      "step": 6235
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1996462828721377,
+      "learning_rate": 1.9291752613089717e-05,
+      "loss": 0.928,
+      "step": 6236
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0246187994283438,
+      "learning_rate": 1.929147053595124e-05,
+      "loss": 0.7505,
+      "step": 6237
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6542623354191046,
+      "learning_rate": 1.9291188404714876e-05,
+      "loss": 0.6163,
+      "step": 6238
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.522340096196448,
+      "learning_rate": 1.9290906219382277e-05,
+      "loss": 0.7537,
+      "step": 6239
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5748458611018876,
+      "learning_rate": 1.9290623979955077e-05,
+      "loss": 0.7569,
+      "step": 6240
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.2470840864229973,
+      "learning_rate": 1.9290341686434924e-05,
+      "loss": 0.7336,
+      "step": 6241
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5823293411883501,
+      "learning_rate": 1.929005933882346e-05,
+      "loss": 0.7738,
+      "step": 6242
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.620490352073305,
+      "learning_rate": 1.928977693712233e-05,
+      "loss": 0.9488,
+      "step": 6243
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6131069341682986,
+      "learning_rate": 1.9289494481333177e-05,
+      "loss": 0.7705,
+      "step": 6244
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9057993158223225,
+      "learning_rate": 1.9289211971457644e-05,
+      "loss": 0.8196,
+      "step": 6245
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5057956011732059,
+      "learning_rate": 1.928892940749738e-05,
+      "loss": 0.8127,
+      "step": 6246
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9981230331285218,
+      "learning_rate": 1.9288646789454032e-05,
+      "loss": 0.8325,
+      "step": 6247
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.572337773434345,
+      "learning_rate": 1.9288364117329233e-05,
+      "loss": 0.7701,
+      "step": 6248
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3914272084412522,
+      "learning_rate": 1.9288081391124643e-05,
+      "loss": 0.8636,
+      "step": 6249
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2996006767649322,
+      "learning_rate": 1.9287798610841903e-05,
+      "loss": 0.8397,
+      "step": 6250
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6132668949359816,
+      "learning_rate": 1.928751577648266e-05,
+      "loss": 0.5259,
+      "step": 6251
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4304796223362197,
+      "learning_rate": 1.9287232888048554e-05,
+      "loss": 0.6577,
+      "step": 6252
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6842873661777829,
+      "learning_rate": 1.9286949945541243e-05,
+      "loss": 0.7323,
+      "step": 6253
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4110588346587616,
+      "learning_rate": 1.9286666948962368e-05,
+      "loss": 0.8634,
+      "step": 6254
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6223516139843452,
+      "learning_rate": 1.9286383898313578e-05,
+      "loss": 0.797,
+      "step": 6255
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.027675950143641,
+      "learning_rate": 1.9286100793596523e-05,
+      "loss": 0.7961,
+      "step": 6256
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4567865111907734,
+      "learning_rate": 1.9285817634812844e-05,
+      "loss": 0.7488,
+      "step": 6257
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4567040325555387,
+      "learning_rate": 1.92855344219642e-05,
+      "loss": 0.8169,
+      "step": 6258
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4787420782615708,
+      "learning_rate": 1.9285251155052233e-05,
+      "loss": 0.7479,
+      "step": 6259
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4577378189820789,
+      "learning_rate": 1.928496783407859e-05,
+      "loss": 0.8847,
+      "step": 6260
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6038684788431189,
+      "learning_rate": 1.928468445904493e-05,
+      "loss": 0.5193,
+      "step": 6261
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8144611694839117,
+      "learning_rate": 1.9284401029952898e-05,
+      "loss": 0.7707,
+      "step": 6262
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8869146322860149,
+      "learning_rate": 1.928411754680414e-05,
+      "loss": 0.8511,
+      "step": 6263
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8307171702210927,
+      "learning_rate": 1.928383400960031e-05,
+      "loss": 0.8021,
+      "step": 6264
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.0183072659805514,
+      "learning_rate": 1.9283550418343063e-05,
+      "loss": 0.7272,
+      "step": 6265
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9658631900385948,
+      "learning_rate": 1.9283266773034047e-05,
+      "loss": 0.7957,
+      "step": 6266
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6419920923293092,
+      "learning_rate": 1.928298307367491e-05,
+      "loss": 0.569,
+      "step": 6267
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3718437603196199,
+      "learning_rate": 1.928269932026731e-05,
+      "loss": 0.7334,
+      "step": 6268
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5384333246051538,
+      "learning_rate": 1.928241551281289e-05,
+      "loss": 0.6509,
+      "step": 6269
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7370498447463836,
+      "learning_rate": 1.928213165131331e-05,
+      "loss": 0.6649,
+      "step": 6270
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.623032675323719,
+      "learning_rate": 1.9281847735770224e-05,
+      "loss": 0.7536,
+      "step": 6271
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6123293356159049,
+      "learning_rate": 1.928156376618528e-05,
+      "loss": 0.5581,
+      "step": 6272
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.5839665446741664,
+      "learning_rate": 1.9281279742560134e-05,
+      "loss": 0.768,
+      "step": 6273
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4115326457578858,
+      "learning_rate": 1.9280995664896435e-05,
+      "loss": 0.8333,
+      "step": 6274
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5986113533821638,
+      "learning_rate": 1.9280711533195843e-05,
+      "loss": 0.9413,
+      "step": 6275
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5833589550763254,
+      "learning_rate": 1.9280427347460013e-05,
+      "loss": 0.781,
+      "step": 6276
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.675724847836078,
+      "learning_rate": 1.9280143107690593e-05,
+      "loss": 0.8019,
+      "step": 6277
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5531660283880213,
+      "learning_rate": 1.9279858813889246e-05,
+      "loss": 0.7201,
+      "step": 6278
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1967544132041392,
+      "learning_rate": 1.927957446605762e-05,
+      "loss": 0.8401,
+      "step": 6279
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4459965322647137,
+      "learning_rate": 1.9279290064197375e-05,
+      "loss": 0.6639,
+      "step": 6280
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.072263226684105,
+      "learning_rate": 1.9279005608310165e-05,
+      "loss": 0.7075,
+      "step": 6281
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3586512947274172,
+      "learning_rate": 1.9278721098397645e-05,
+      "loss": 0.9317,
+      "step": 6282
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.386788489322101,
+      "learning_rate": 1.9278436534461476e-05,
+      "loss": 0.6392,
+      "step": 6283
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5928812181183505,
+      "learning_rate": 1.927815191650331e-05,
+      "loss": 0.8945,
+      "step": 6284
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6159525703364637,
+      "learning_rate": 1.9277867244524805e-05,
+      "loss": 0.5856,
+      "step": 6285
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.44542506423389,
+      "learning_rate": 1.927758251852762e-05,
+      "loss": 0.7468,
+      "step": 6286
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6674163794349037,
+      "learning_rate": 1.9277297738513415e-05,
+      "loss": 0.7385,
+      "step": 6287
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9128243240606426,
+      "learning_rate": 1.9277012904483846e-05,
+      "loss": 0.7631,
+      "step": 6288
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9068056295848468,
+      "learning_rate": 1.927672801644057e-05,
+      "loss": 0.7943,
+      "step": 6289
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8350578914097897,
+      "learning_rate": 1.9276443074385246e-05,
+      "loss": 0.9642,
+      "step": 6290
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.63109197785217,
+      "learning_rate": 1.927615807831953e-05,
+      "loss": 0.7132,
+      "step": 6291
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5828803359025619,
+      "learning_rate": 1.927587302824509e-05,
+      "loss": 0.5825,
+      "step": 6292
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9233783859530398,
+      "learning_rate": 1.9275587924163578e-05,
+      "loss": 0.7292,
+      "step": 6293
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5649725671380852,
+      "learning_rate": 1.927530276607666e-05,
+      "loss": 0.5862,
+      "step": 6294
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5348398496467899,
+      "learning_rate": 1.9275017553985986e-05,
+      "loss": 0.8626,
+      "step": 6295
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.468432320365119,
+      "learning_rate": 1.927473228789323e-05,
+      "loss": 0.6737,
+      "step": 6296
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6280655286746338,
+      "learning_rate": 1.927444696780004e-05,
+      "loss": 0.5378,
+      "step": 6297
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5441779718236084,
+      "learning_rate": 1.9274161593708086e-05,
+      "loss": 0.9245,
+      "step": 6298
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5876140117984208,
+      "learning_rate": 1.927387616561903e-05,
+      "loss": 0.5375,
+      "step": 6299
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6039779989956998,
+      "learning_rate": 1.927359068353453e-05,
+      "loss": 0.529,
+      "step": 6300
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3711897383390157,
+      "learning_rate": 1.9273305147456247e-05,
+      "loss": 0.6605,
+      "step": 6301
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3571724334112503,
+      "learning_rate": 1.9273019557385845e-05,
+      "loss": 0.8368,
+      "step": 6302
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5060987631011997,
+      "learning_rate": 1.927273391332499e-05,
+      "loss": 0.82,
+      "step": 6303
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.142154905830783,
+      "learning_rate": 1.927244821527534e-05,
+      "loss": 0.7448,
+      "step": 6304
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.824593495934203,
+      "learning_rate": 1.9272162463238562e-05,
+      "loss": 0.7104,
+      "step": 6305
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.449489573020591,
+      "learning_rate": 1.927187665721632e-05,
+      "loss": 0.8455,
+      "step": 6306
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7661447235737653,
+      "learning_rate": 1.9271590797210276e-05,
+      "loss": 0.6618,
+      "step": 6307
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.679313195419398,
+      "learning_rate": 1.9271304883222095e-05,
+      "loss": 0.8849,
+      "step": 6308
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.684519434528571,
+      "learning_rate": 1.927101891525344e-05,
+      "loss": 0.6659,
+      "step": 6309
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4371781301231532,
+      "learning_rate": 1.927073289330598e-05,
+      "loss": 0.9913,
+      "step": 6310
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6859235514638518,
+      "learning_rate": 1.9270446817381377e-05,
+      "loss": 0.6437,
+      "step": 6311
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5959152144348209,
+      "learning_rate": 1.9270160687481296e-05,
+      "loss": 0.5372,
+      "step": 6312
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6518204193934733,
+      "learning_rate": 1.926987450360741e-05,
+      "loss": 0.847,
+      "step": 6313
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6112477714796865,
+      "learning_rate": 1.9269588265761374e-05,
+      "loss": 0.5449,
+      "step": 6314
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.239808858958942,
+      "learning_rate": 1.9269301973944864e-05,
+      "loss": 0.7779,
+      "step": 6315
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5819294100433423,
+      "learning_rate": 1.926901562815954e-05,
+      "loss": 0.6091,
+      "step": 6316
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6024695889656337,
+      "learning_rate": 1.9268729228407076e-05,
+      "loss": 0.7534,
+      "step": 6317
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5530888414417309,
+      "learning_rate": 1.9268442774689132e-05,
+      "loss": 0.495,
+      "step": 6318
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4784795142621177,
+      "learning_rate": 1.9268156267007383e-05,
+      "loss": 0.8463,
+      "step": 6319
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5873894860748354,
+      "learning_rate": 1.9267869705363493e-05,
+      "loss": 0.5106,
+      "step": 6320
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.529250174380803,
+      "learning_rate": 1.926758308975913e-05,
+      "loss": 0.8466,
+      "step": 6321
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.891168746344474,
+      "learning_rate": 1.9267296420195966e-05,
+      "loss": 0.6706,
+      "step": 6322
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6175288675540987,
+      "learning_rate": 1.926700969667567e-05,
+      "loss": 0.5325,
+      "step": 6323
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.554004541579648,
+      "learning_rate": 1.9266722919199907e-05,
+      "loss": 0.6576,
+      "step": 6324
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5574034925066158,
+      "learning_rate": 1.926643608777035e-05,
+      "loss": 0.7046,
+      "step": 6325
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4476005685482052,
+      "learning_rate": 1.926614920238867e-05,
+      "loss": 0.8196,
+      "step": 6326
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.089538185993736,
+      "learning_rate": 1.926586226305653e-05,
+      "loss": 0.8341,
+      "step": 6327
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.82001287029067,
+      "learning_rate": 1.9265575269775612e-05,
+      "loss": 0.794,
+      "step": 6328
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3441980336665693,
+      "learning_rate": 1.926528822254758e-05,
+      "loss": 0.7244,
+      "step": 6329
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5941929770965286,
+      "learning_rate": 1.9265001121374108e-05,
+      "loss": 0.8285,
+      "step": 6330
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2431141580274003,
+      "learning_rate": 1.9264713966256864e-05,
+      "loss": 0.7691,
+      "step": 6331
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5436522466872247,
+      "learning_rate": 1.9264426757197523e-05,
+      "loss": 0.7344,
+      "step": 6332
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2735418574348156,
+      "learning_rate": 1.9264139494197755e-05,
+      "loss": 0.6514,
+      "step": 6333
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9104937200326408,
+      "learning_rate": 1.9263852177259234e-05,
+      "loss": 0.7069,
+      "step": 6334
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4330666316649632,
+      "learning_rate": 1.9263564806383637e-05,
+      "loss": 0.7027,
+      "step": 6335
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.806007152730385,
+      "learning_rate": 1.9263277381572628e-05,
+      "loss": 0.6303,
+      "step": 6336
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.742837557375174,
+      "learning_rate": 1.926298990282789e-05,
+      "loss": 0.7886,
+      "step": 6337
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5789027083682554,
+      "learning_rate": 1.9262702370151087e-05,
+      "loss": 0.6477,
+      "step": 6338
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6465287093414094,
+      "learning_rate": 1.92624147835439e-05,
+      "loss": 0.5321,
+      "step": 6339
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8282507141636297,
+      "learning_rate": 1.9262127143008e-05,
+      "loss": 0.8274,
+      "step": 6340
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5876113818244094,
+      "learning_rate": 1.9261839448545066e-05,
+      "loss": 0.5688,
+      "step": 6341
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6563830171568446,
+      "learning_rate": 1.926155170015677e-05,
+      "loss": 0.5322,
+      "step": 6342
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.211528881260836,
+      "learning_rate": 1.926126389784479e-05,
+      "loss": 0.8291,
+      "step": 6343
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.0489956507543914,
+      "learning_rate": 1.9260976041610797e-05,
+      "loss": 0.8005,
+      "step": 6344
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5780568290708442,
+      "learning_rate": 1.926068813145647e-05,
+      "loss": 0.5212,
+      "step": 6345
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5823828160070851,
+      "learning_rate": 1.9260400167383485e-05,
+      "loss": 0.519,
+      "step": 6346
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4423000646820308,
+      "learning_rate": 1.9260112149393517e-05,
+      "loss": 0.7796,
+      "step": 6347
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6475050945066212,
+      "learning_rate": 1.9259824077488248e-05,
+      "loss": 0.5444,
+      "step": 6348
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5940739203348765,
+      "learning_rate": 1.925953595166935e-05,
+      "loss": 0.5292,
+      "step": 6349
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.656042013120959,
+      "learning_rate": 1.92592477719385e-05,
+      "loss": 0.6799,
+      "step": 6350
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5510126863725577,
+      "learning_rate": 1.925895953829738e-05,
+      "loss": 0.7781,
+      "step": 6351
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6602678505730994,
+      "learning_rate": 1.9258671250747665e-05,
+      "loss": 0.7439,
+      "step": 6352
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7645381457104905,
+      "learning_rate": 1.9258382909291037e-05,
+      "loss": 0.7551,
+      "step": 6353
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7039589983827281,
+      "learning_rate": 1.925809451392917e-05,
+      "loss": 0.7655,
+      "step": 6354
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4943122374278788,
+      "learning_rate": 1.9257806064663746e-05,
+      "loss": 0.7559,
+      "step": 6355
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9989934689800994,
+      "learning_rate": 1.9257517561496443e-05,
+      "loss": 0.7973,
+      "step": 6356
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5811569754190881,
+      "learning_rate": 1.9257229004428944e-05,
+      "loss": 0.5542,
+      "step": 6357
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8624682456054447,
+      "learning_rate": 1.925694039346293e-05,
+      "loss": 0.7838,
+      "step": 6358
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4859653748785575,
+      "learning_rate": 1.9256651728600073e-05,
+      "loss": 0.8811,
+      "step": 6359
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4721726230449708,
+      "learning_rate": 1.925636300984206e-05,
+      "loss": 0.7525,
+      "step": 6360
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8382720676773052,
+      "learning_rate": 1.9256074237190573e-05,
+      "loss": 0.7673,
+      "step": 6361
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5113659290342645,
+      "learning_rate": 1.925578541064729e-05,
+      "loss": 0.7614,
+      "step": 6362
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4537712689007611,
+      "learning_rate": 1.9255496530213895e-05,
+      "loss": 0.7124,
+      "step": 6363
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5135258761009753,
+      "learning_rate": 1.925520759589207e-05,
+      "loss": 0.6697,
+      "step": 6364
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4456906167489252,
+      "learning_rate": 1.9254918607683493e-05,
+      "loss": 0.7176,
+      "step": 6365
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4187150610349497,
+      "learning_rate": 1.925462956558985e-05,
+      "loss": 0.7994,
+      "step": 6366
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5508680696358341,
+      "learning_rate": 1.9254340469612826e-05,
+      "loss": 0.6172,
+      "step": 6367
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4587522393295267,
+      "learning_rate": 1.9254051319754103e-05,
+      "loss": 0.8305,
+      "step": 6368
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3784576990331958,
+      "learning_rate": 1.925376211601536e-05,
+      "loss": 0.9216,
+      "step": 6369
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.612375667063886,
+      "learning_rate": 1.9253472858398284e-05,
+      "loss": 0.7077,
+      "step": 6370
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.788475761370955,
+      "learning_rate": 1.925318354690456e-05,
+      "loss": 0.7843,
+      "step": 6371
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5999697107947126,
+      "learning_rate": 1.9252894181535874e-05,
+      "loss": 0.5166,
+      "step": 6372
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6609774487764652,
+      "learning_rate": 1.9252604762293908e-05,
+      "loss": 0.6033,
+      "step": 6373
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0644047448457488,
+      "learning_rate": 1.9252315289180348e-05,
+      "loss": 0.7159,
+      "step": 6374
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5610465165519256,
+      "learning_rate": 1.9252025762196878e-05,
+      "loss": 0.7282,
+      "step": 6375
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.580673008922698,
+      "learning_rate": 1.9251736181345186e-05,
+      "loss": 0.7892,
+      "step": 6376
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4828442052860586,
+      "learning_rate": 1.925144654662696e-05,
+      "loss": 0.7257,
+      "step": 6377
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5191039895373373,
+      "learning_rate": 1.9251156858043877e-05,
+      "loss": 0.8419,
+      "step": 6378
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.454729991108136,
+      "learning_rate": 1.925086711559763e-05,
+      "loss": 0.9472,
+      "step": 6379
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6703848382404218,
+      "learning_rate": 1.9250577319289912e-05,
+      "loss": 0.8048,
+      "step": 6380
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5829291714655735,
+      "learning_rate": 1.92502874691224e-05,
+      "loss": 0.7244,
+      "step": 6381
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7185022480952188,
+      "learning_rate": 1.9249997565096783e-05,
+      "loss": 0.8389,
+      "step": 6382
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7533483475907696,
+      "learning_rate": 1.9249707607214757e-05,
+      "loss": 0.7638,
+      "step": 6383
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5028372590485306,
+      "learning_rate": 1.9249417595478005e-05,
+      "loss": 0.713,
+      "step": 6384
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.831501073785257,
+      "learning_rate": 1.9249127529888212e-05,
+      "loss": 0.7386,
+      "step": 6385
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8097074974741172,
+      "learning_rate": 1.924883741044707e-05,
+      "loss": 0.8122,
+      "step": 6386
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8609629910119687,
+      "learning_rate": 1.924854723715627e-05,
+      "loss": 0.6802,
+      "step": 6387
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6283846778304825,
+      "learning_rate": 1.92482570100175e-05,
+      "loss": 0.7982,
+      "step": 6388
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6193648080194112,
+      "learning_rate": 1.924796672903245e-05,
+      "loss": 0.7464,
+      "step": 6389
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5736794019036063,
+      "learning_rate": 1.924767639420281e-05,
+      "loss": 0.5359,
+      "step": 6390
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.670684673810328,
+      "learning_rate": 1.924738600553027e-05,
+      "loss": 0.8087,
+      "step": 6391
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4993939520196382,
+      "learning_rate": 1.924709556301652e-05,
+      "loss": 0.7967,
+      "step": 6392
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6974420357551192,
+      "learning_rate": 1.9246805066663255e-05,
+      "loss": 0.6763,
+      "step": 6393
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4755335854390232,
+      "learning_rate": 1.9246514516472157e-05,
+      "loss": 0.8695,
+      "step": 6394
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.861629790464776,
+      "learning_rate": 1.9246223912444933e-05,
+      "loss": 0.7612,
+      "step": 6395
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.631660033820752,
+      "learning_rate": 1.924593325458326e-05,
+      "loss": 0.7061,
+      "step": 6396
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8653151949238949,
+      "learning_rate": 1.9245642542888836e-05,
+      "loss": 0.741,
+      "step": 6397
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.641140817850573,
+      "learning_rate": 1.9245351777363355e-05,
+      "loss": 0.8973,
+      "step": 6398
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6027963242099066,
+      "learning_rate": 1.924506095800851e-05,
+      "loss": 0.7307,
+      "step": 6399
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.870649421439275,
+      "learning_rate": 1.9244770084825993e-05,
+      "loss": 0.6919,
+      "step": 6400
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6543472896014064,
+      "learning_rate": 1.92444791578175e-05,
+      "loss": 0.7162,
+      "step": 6401
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.746542135316016,
+      "learning_rate": 1.924418817698472e-05,
+      "loss": 0.8463,
+      "step": 6402
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.470285783027509,
+      "learning_rate": 1.9243897142329347e-05,
+      "loss": 0.7628,
+      "step": 6403
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6498667602400605,
+      "learning_rate": 1.9243606053853084e-05,
+      "loss": 0.8811,
+      "step": 6404
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6462798524506561,
+      "learning_rate": 1.9243314911557614e-05,
+      "loss": 0.7063,
+      "step": 6405
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5531976061420352,
+      "learning_rate": 1.9243023715444646e-05,
+      "loss": 0.7653,
+      "step": 6406
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.412475882688344,
+      "learning_rate": 1.9242732465515862e-05,
+      "loss": 0.7586,
+      "step": 6407
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6662147857110898,
+      "learning_rate": 1.9242441161772965e-05,
+      "loss": 0.7128,
+      "step": 6408
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8283875063482382,
+      "learning_rate": 1.9242149804217648e-05,
+      "loss": 0.7817,
+      "step": 6409
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.795618218546379,
+      "learning_rate": 1.9241858392851612e-05,
+      "loss": 0.6959,
+      "step": 6410
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4055544379636402,
+      "learning_rate": 1.9241566927676548e-05,
+      "loss": 0.7622,
+      "step": 6411
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1294234986916476,
+      "learning_rate": 1.9241275408694156e-05,
+      "loss": 0.7501,
+      "step": 6412
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6232834118958168,
+      "learning_rate": 1.924098383590613e-05,
+      "loss": 0.7488,
+      "step": 6413
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.469364595098089,
+      "learning_rate": 1.9240692209314176e-05,
+      "loss": 0.82,
+      "step": 6414
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4979198285972934,
+      "learning_rate": 1.9240400528919983e-05,
+      "loss": 0.7507,
+      "step": 6415
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6490592803395965,
+      "learning_rate": 1.9240108794725252e-05,
+      "loss": 0.8506,
+      "step": 6416
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4665500261218014,
+      "learning_rate": 1.9239817006731684e-05,
+      "loss": 0.6976,
+      "step": 6417
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8972170193856992,
+      "learning_rate": 1.9239525164940977e-05,
+      "loss": 0.6545,
+      "step": 6418
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6397642921699169,
+      "learning_rate": 1.9239233269354828e-05,
+      "loss": 0.5818,
+      "step": 6419
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4681984839663762,
+      "learning_rate": 1.9238941319974938e-05,
+      "loss": 0.776,
+      "step": 6420
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5805286541400624,
+      "learning_rate": 1.9238649316803004e-05,
+      "loss": 0.5377,
+      "step": 6421
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.81758716006013,
+      "learning_rate": 1.923835725984073e-05,
+      "loss": 0.8566,
+      "step": 6422
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4559770724984271,
+      "learning_rate": 1.923806514908982e-05,
+      "loss": 0.7475,
+      "step": 6423
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.58726103116939,
+      "learning_rate": 1.9237772984551966e-05,
+      "loss": 0.7673,
+      "step": 6424
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.45705933112001,
+      "learning_rate": 1.9237480766228872e-05,
+      "loss": 0.7174,
+      "step": 6425
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6335757133211286,
+      "learning_rate": 1.923718849412224e-05,
+      "loss": 0.7505,
+      "step": 6426
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.591976892962067,
+      "learning_rate": 1.9236896168233775e-05,
+      "loss": 0.8621,
+      "step": 6427
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.572880305061113,
+      "learning_rate": 1.9236603788565176e-05,
+      "loss": 0.7484,
+      "step": 6428
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8604654057655425,
+      "learning_rate": 1.9236311355118144e-05,
+      "loss": 0.7546,
+      "step": 6429
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.635905438413987,
+      "learning_rate": 1.9236018867894382e-05,
+      "loss": 0.5384,
+      "step": 6430
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9054000943517462,
+      "learning_rate": 1.9235726326895595e-05,
+      "loss": 0.8602,
+      "step": 6431
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4433454545665116,
+      "learning_rate": 1.9235433732123488e-05,
+      "loss": 0.6966,
+      "step": 6432
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4676860460718877,
+      "learning_rate": 1.923514108357976e-05,
+      "loss": 0.7331,
+      "step": 6433
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.4907383376223655,
+      "learning_rate": 1.9234848381266117e-05,
+      "loss": 0.754,
+      "step": 6434
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.402930998460524,
+      "learning_rate": 1.9234555625184265e-05,
+      "loss": 0.7348,
+      "step": 6435
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6353148098057384,
+      "learning_rate": 1.9234262815335903e-05,
+      "loss": 0.7571,
+      "step": 6436
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.434853070938428,
+      "learning_rate": 1.923396995172274e-05,
+      "loss": 0.742,
+      "step": 6437
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5686672639973618,
+      "learning_rate": 1.9233677034346484e-05,
+      "loss": 0.8569,
+      "step": 6438
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6293669553361582,
+      "learning_rate": 1.923338406320883e-05,
+      "loss": 0.5593,
+      "step": 6439
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7593891190410487,
+      "learning_rate": 1.92330910383115e-05,
+      "loss": 0.8437,
+      "step": 6440
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.9662599693098284,
+      "learning_rate": 1.9232797959656184e-05,
+      "loss": 0.9015,
+      "step": 6441
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7351849294888504,
+      "learning_rate": 1.9232504827244594e-05,
+      "loss": 0.7914,
+      "step": 6442
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7995301952320257,
+      "learning_rate": 1.9232211641078443e-05,
+      "loss": 0.9899,
+      "step": 6443
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.846590043374141,
+      "learning_rate": 1.9231918401159433e-05,
+      "loss": 0.6509,
+      "step": 6444
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4299022276742503,
+      "learning_rate": 1.923162510748927e-05,
+      "loss": 0.8055,
+      "step": 6445
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8601824127072042,
+      "learning_rate": 1.9231331760069662e-05,
+      "loss": 0.8689,
+      "step": 6446
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.429048942744327,
+      "learning_rate": 1.9231038358902317e-05,
+      "loss": 0.7072,
+      "step": 6447
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.38300457437224,
+      "learning_rate": 1.9230744903988943e-05,
+      "loss": 0.6601,
+      "step": 6448
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.187645440829301,
+      "learning_rate": 1.9230451395331253e-05,
+      "loss": 0.6831,
+      "step": 6449
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5575215484815481,
+      "learning_rate": 1.9230157832930953e-05,
+      "loss": 0.608,
+      "step": 6450
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4800142427928524,
+      "learning_rate": 1.922986421678975e-05,
+      "loss": 0.787,
+      "step": 6451
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8778323407508664,
+      "learning_rate": 1.9229570546909354e-05,
+      "loss": 0.8111,
+      "step": 6452
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9240061217182345,
+      "learning_rate": 1.922927682329148e-05,
+      "loss": 0.7505,
+      "step": 6453
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.552445554525265,
+      "learning_rate": 1.922898304593783e-05,
+      "loss": 0.6775,
+      "step": 6454
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5468509514461322,
+      "learning_rate": 1.922868921485012e-05,
+      "loss": 0.7449,
+      "step": 6455
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3734990567829548,
+      "learning_rate": 1.9228395330030064e-05,
+      "loss": 0.7707,
+      "step": 6456
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4154972778733113,
+      "learning_rate": 1.9228101391479366e-05,
+      "loss": 0.8403,
+      "step": 6457
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5813262718988935,
+      "learning_rate": 1.922780739919974e-05,
+      "loss": 0.5514,
+      "step": 6458
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6187563367506237,
+      "learning_rate": 1.9227513353192895e-05,
+      "loss": 0.6754,
+      "step": 6459
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.258361146313404,
+      "learning_rate": 1.9227219253460552e-05,
+      "loss": 0.8557,
+      "step": 6460
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.928857297082307,
+      "learning_rate": 1.9226925100004415e-05,
+      "loss": 0.7198,
+      "step": 6461
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5851799562297476,
+      "learning_rate": 1.9226630892826196e-05,
+      "loss": 0.578,
+      "step": 6462
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8008064560589725,
+      "learning_rate": 1.9226336631927613e-05,
+      "loss": 0.9116,
+      "step": 6463
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.462545617859446,
+      "learning_rate": 1.9226042317310377e-05,
+      "loss": 0.7397,
+      "step": 6464
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6352169366646788,
+      "learning_rate": 1.9225747948976205e-05,
+      "loss": 0.5488,
+      "step": 6465
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4865113734122082,
+      "learning_rate": 1.92254535269268e-05,
+      "loss": 0.7091,
+      "step": 6466
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.252250379088576,
+      "learning_rate": 1.9225159051163888e-05,
+      "loss": 0.8372,
+      "step": 6467
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4822466512741073,
+      "learning_rate": 1.9224864521689182e-05,
+      "loss": 0.7816,
+      "step": 6468
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.529692448812986,
+      "learning_rate": 1.922456993850439e-05,
+      "loss": 0.8755,
+      "step": 6469
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.545425741769067,
+      "learning_rate": 1.9224275301611234e-05,
+      "loss": 0.7056,
+      "step": 6470
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.070090351410822,
+      "learning_rate": 1.9223980611011424e-05,
+      "loss": 0.7683,
+      "step": 6471
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5192879979561487,
+      "learning_rate": 1.922368586670668e-05,
+      "loss": 0.8117,
+      "step": 6472
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5814734286130792,
+      "learning_rate": 1.9223391068698717e-05,
+      "loss": 0.8807,
+      "step": 6473
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.753556859906345,
+      "learning_rate": 1.9223096216989253e-05,
+      "loss": 0.8833,
+      "step": 6474
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9865436879626228,
+      "learning_rate": 1.922280131158e-05,
+      "loss": 0.8921,
+      "step": 6475
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5061751643463608,
+      "learning_rate": 1.9222506352472677e-05,
+      "loss": 0.7299,
+      "step": 6476
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6423619516870573,
+      "learning_rate": 1.9222211339669005e-05,
+      "loss": 0.7171,
+      "step": 6477
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9867881869197384,
+      "learning_rate": 1.9221916273170697e-05,
+      "loss": 0.8335,
+      "step": 6478
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.933644016479773,
+      "learning_rate": 1.9221621152979476e-05,
+      "loss": 0.8039,
+      "step": 6479
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.962814068517142,
+      "learning_rate": 1.9221325979097053e-05,
+      "loss": 0.7136,
+      "step": 6480
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6115802530235374,
+      "learning_rate": 1.9221030751525154e-05,
+      "loss": 0.5578,
+      "step": 6481
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4075714535459518,
+      "learning_rate": 1.9220735470265492e-05,
+      "loss": 0.743,
+      "step": 6482
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.593169732637537,
+      "learning_rate": 1.9220440135319794e-05,
+      "loss": 0.5226,
+      "step": 6483
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9767524083701513,
+      "learning_rate": 1.9220144746689768e-05,
+      "loss": 0.8253,
+      "step": 6484
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6394589548610738,
+      "learning_rate": 1.9219849304377142e-05,
+      "loss": 0.7799,
+      "step": 6485
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4543774857413398,
+      "learning_rate": 1.9219553808383636e-05,
+      "loss": 0.7372,
+      "step": 6486
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7034452541669043,
+      "learning_rate": 1.921925825871097e-05,
+      "loss": 0.6543,
+      "step": 6487
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4588047085363935,
+      "learning_rate": 1.921896265536086e-05,
+      "loss": 0.9083,
+      "step": 6488
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.3114578915077417,
+      "learning_rate": 1.9218666998335034e-05,
+      "loss": 0.8058,
+      "step": 6489
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.591253552654712,
+      "learning_rate": 1.921837128763521e-05,
+      "loss": 0.7805,
+      "step": 6490
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.057811339692527,
+      "learning_rate": 1.9218075523263104e-05,
+      "loss": 0.7469,
+      "step": 6491
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5095200588756965,
+      "learning_rate": 1.9217779705220452e-05,
+      "loss": 0.6271,
+      "step": 6492
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0331265142678503,
+      "learning_rate": 1.9217483833508962e-05,
+      "loss": 0.7429,
+      "step": 6493
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6384003282924915,
+      "learning_rate": 1.921718790813037e-05,
+      "loss": 0.794,
+      "step": 6494
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7703069858703069,
+      "learning_rate": 1.9216891929086382e-05,
+      "loss": 0.6841,
+      "step": 6495
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4512036351237967,
+      "learning_rate": 1.9216595896378737e-05,
+      "loss": 0.7841,
+      "step": 6496
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6223778970327443,
+      "learning_rate": 1.9216299810009156e-05,
+      "loss": 0.6829,
+      "step": 6497
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.506295086180594,
+      "learning_rate": 1.9216003669979356e-05,
+      "loss": 0.6746,
+      "step": 6498
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9839704914928902,
+      "learning_rate": 1.9215707476291064e-05,
+      "loss": 0.7775,
+      "step": 6499
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5177365554265427,
+      "learning_rate": 1.9215411228946005e-05,
+      "loss": 0.7822,
+      "step": 6500
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.651651191177497,
+      "learning_rate": 1.9215114927945905e-05,
+      "loss": 0.8791,
+      "step": 6501
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6573369643269893,
+      "learning_rate": 1.9214818573292492e-05,
+      "loss": 0.81,
+      "step": 6502
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.605375077117776,
+      "learning_rate": 1.9214522164987484e-05,
+      "loss": 0.8035,
+      "step": 6503
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6055528125188345,
+      "learning_rate": 1.921422570303261e-05,
+      "loss": 0.8301,
+      "step": 6504
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7346281301566142,
+      "learning_rate": 1.9213929187429602e-05,
+      "loss": 0.6811,
+      "step": 6505
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8997099007602998,
+      "learning_rate": 1.9213632618180175e-05,
+      "loss": 0.8031,
+      "step": 6506
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0649410530617947,
+      "learning_rate": 1.9213335995286065e-05,
+      "loss": 0.7407,
+      "step": 6507
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.887583924159186,
+      "learning_rate": 1.9213039318748995e-05,
+      "loss": 0.7061,
+      "step": 6508
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5795485956917614,
+      "learning_rate": 1.9212742588570694e-05,
+      "loss": 0.8116,
+      "step": 6509
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6233106139155027,
+      "learning_rate": 1.921244580475289e-05,
+      "loss": 0.7616,
+      "step": 6510
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8635217327875147,
+      "learning_rate": 1.9212148967297307e-05,
+      "loss": 0.7985,
+      "step": 6511
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3025670082016774,
+      "learning_rate": 1.921185207620568e-05,
+      "loss": 0.6166,
+      "step": 6512
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5771693111280191,
+      "learning_rate": 1.9211555131479728e-05,
+      "loss": 0.517,
+      "step": 6513
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.70028205334454,
+      "learning_rate": 1.921125813312119e-05,
+      "loss": 0.6618,
+      "step": 6514
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5846967293121499,
+      "learning_rate": 1.921096108113179e-05,
+      "loss": 0.6547,
+      "step": 6515
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6399635036782458,
+      "learning_rate": 1.921066397551326e-05,
+      "loss": 0.6706,
+      "step": 6516
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.672062655545592,
+      "learning_rate": 1.9210366816267327e-05,
+      "loss": 0.689,
+      "step": 6517
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.667726737602377,
+      "learning_rate": 1.9210069603395724e-05,
+      "loss": 0.8318,
+      "step": 6518
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9610376301448458,
+      "learning_rate": 1.920977233690018e-05,
+      "loss": 0.8218,
+      "step": 6519
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.645298433477904,
+      "learning_rate": 1.920947501678242e-05,
+      "loss": 0.8218,
+      "step": 6520
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7175656875101502,
+      "learning_rate": 1.9209177643044187e-05,
+      "loss": 0.9484,
+      "step": 6521
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.616492228604304,
+      "learning_rate": 1.9208880215687203e-05,
+      "loss": 0.7288,
+      "step": 6522
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5877846228031318,
+      "learning_rate": 1.9208582734713202e-05,
+      "loss": 0.5124,
+      "step": 6523
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5739179829730883,
+      "learning_rate": 1.920828520012392e-05,
+      "loss": 0.7351,
+      "step": 6524
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7008763448044253,
+      "learning_rate": 1.9207987611921085e-05,
+      "loss": 0.8586,
+      "step": 6525
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.689588725663411,
+      "learning_rate": 1.9207689970106434e-05,
+      "loss": 0.7174,
+      "step": 6526
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0805553576863285,
+      "learning_rate": 1.9207392274681692e-05,
+      "loss": 0.7758,
+      "step": 6527
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7179397501022937,
+      "learning_rate": 1.92070945256486e-05,
+      "loss": 0.7077,
+      "step": 6528
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.216112933569432,
+      "learning_rate": 1.920679672300889e-05,
+      "loss": 0.8648,
+      "step": 6529
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6878218517908108,
+      "learning_rate": 1.920649886676429e-05,
+      "loss": 0.8177,
+      "step": 6530
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5342272562677433,
+      "learning_rate": 1.9206200956916542e-05,
+      "loss": 0.6831,
+      "step": 6531
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4995091380083452,
+      "learning_rate": 1.9205902993467377e-05,
+      "loss": 0.8536,
+      "step": 6532
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8854631649695581,
+      "learning_rate": 1.920560497641853e-05,
+      "loss": 0.7473,
+      "step": 6533
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5430025879968736,
+      "learning_rate": 1.9205306905771735e-05,
+      "loss": 0.713,
+      "step": 6534
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5389229649121394,
+      "learning_rate": 1.920500878152873e-05,
+      "loss": 0.8115,
+      "step": 6535
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.171433079704782,
+      "learning_rate": 1.920471060369125e-05,
+      "loss": 0.7949,
+      "step": 6536
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.938476911823013,
+      "learning_rate": 1.920441237226103e-05,
+      "loss": 0.7757,
+      "step": 6537
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.417527472628822,
+      "learning_rate": 1.9204114087239806e-05,
+      "loss": 0.831,
+      "step": 6538
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6563735254137106,
+      "learning_rate": 1.920381574862932e-05,
+      "loss": 0.7089,
+      "step": 6539
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5705874129239272,
+      "learning_rate": 1.92035173564313e-05,
+      "loss": 0.8914,
+      "step": 6540
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4613178041512778,
+      "learning_rate": 1.9203218910647492e-05,
+      "loss": 0.7183,
+      "step": 6541
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6295574025899224,
+      "learning_rate": 1.920292041127963e-05,
+      "loss": 0.6708,
+      "step": 6542
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7193182447600777,
+      "learning_rate": 1.920262185832945e-05,
+      "loss": 0.6984,
+      "step": 6543
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4616985004448106,
+      "learning_rate": 1.920232325179869e-05,
+      "loss": 0.6975,
+      "step": 6544
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4387018574349892,
+      "learning_rate": 1.9202024591689094e-05,
+      "loss": 0.742,
+      "step": 6545
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5669939280274672,
+      "learning_rate": 1.9201725878002394e-05,
+      "loss": 0.7576,
+      "step": 6546
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9053044260113743,
+      "learning_rate": 1.9201427110740335e-05,
+      "loss": 0.8649,
+      "step": 6547
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3992279738890485,
+      "learning_rate": 1.9201128289904654e-05,
+      "loss": 0.819,
+      "step": 6548
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7064457290683317,
+      "learning_rate": 1.9200829415497095e-05,
+      "loss": 0.8335,
+      "step": 6549
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.600815571405012,
+      "learning_rate": 1.920053048751939e-05,
+      "loss": 0.7787,
+      "step": 6550
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5904632185263232,
+      "learning_rate": 1.9200231505973282e-05,
+      "loss": 0.6586,
+      "step": 6551
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6235409927263256,
+      "learning_rate": 1.9199932470860516e-05,
+      "loss": 0.679,
+      "step": 6552
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5462704887405572,
+      "learning_rate": 1.919963338218283e-05,
+      "loss": 0.6658,
+      "step": 6553
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.049653509659029,
+      "learning_rate": 1.9199334239941968e-05,
+      "loss": 0.6859,
+      "step": 6554
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.518850738111207,
+      "learning_rate": 1.9199035044139666e-05,
+      "loss": 0.7487,
+      "step": 6555
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.740922884874779,
+      "learning_rate": 1.919873579477767e-05,
+      "loss": 0.7723,
+      "step": 6556
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6422872627775247,
+      "learning_rate": 1.9198436491857725e-05,
+      "loss": 0.897,
+      "step": 6557
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6252661516913995,
+      "learning_rate": 1.9198137135381567e-05,
+      "loss": 0.6475,
+      "step": 6558
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7658167508179259,
+      "learning_rate": 1.9197837725350944e-05,
+      "loss": 0.812,
+      "step": 6559
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1835365518900107,
+      "learning_rate": 1.91975382617676e-05,
+      "loss": 0.8172,
+      "step": 6560
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.518874093886399,
+      "learning_rate": 1.9197238744633275e-05,
+      "loss": 0.7403,
+      "step": 6561
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0642830586973973,
+      "learning_rate": 1.919693917394971e-05,
+      "loss": 0.8136,
+      "step": 6562
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8808868940867751,
+      "learning_rate": 1.9196639549718655e-05,
+      "loss": 0.9529,
+      "step": 6563
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5856221813612568,
+      "learning_rate": 1.919633987194186e-05,
+      "loss": 0.7029,
+      "step": 6564
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9826051448214452,
+      "learning_rate": 1.9196040140621056e-05,
+      "loss": 0.7427,
+      "step": 6565
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9031036459234643,
+      "learning_rate": 1.9195740355757998e-05,
+      "loss": 0.7685,
+      "step": 6566
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5384512905049446,
+      "learning_rate": 1.9195440517354424e-05,
+      "loss": 0.7415,
+      "step": 6567
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9703967394525388,
+      "learning_rate": 1.9195140625412087e-05,
+      "loss": 0.8251,
+      "step": 6568
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6664465924296095,
+      "learning_rate": 1.919484067993273e-05,
+      "loss": 0.7491,
+      "step": 6569
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.17820790392353,
+      "learning_rate": 1.91945406809181e-05,
+      "loss": 0.8015,
+      "step": 6570
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7572656004216234,
+      "learning_rate": 1.919424062836994e-05,
+      "loss": 0.6383,
+      "step": 6571
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6655346079956335,
+      "learning_rate": 1.9193940522290008e-05,
+      "loss": 0.8349,
+      "step": 6572
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.538319452872192,
+      "learning_rate": 1.9193640362680038e-05,
+      "loss": 0.8651,
+      "step": 6573
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6461004884256718,
+      "learning_rate": 1.919334014954178e-05,
+      "loss": 0.8649,
+      "step": 6574
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.931058308800501,
+      "learning_rate": 1.919303988287699e-05,
+      "loss": 0.8544,
+      "step": 6575
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2748605301381066,
+      "learning_rate": 1.9192739562687412e-05,
+      "loss": 0.712,
+      "step": 6576
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8868296769527506,
+      "learning_rate": 1.919243918897479e-05,
+      "loss": 0.726,
+      "step": 6577
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.805152100952862,
+      "learning_rate": 1.9192138761740883e-05,
+      "loss": 0.8221,
+      "step": 6578
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9056191862833232,
+      "learning_rate": 1.9191838280987428e-05,
+      "loss": 0.8316,
+      "step": 6579
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8124354793446342,
+      "learning_rate": 1.919153774671618e-05,
+      "loss": 0.7887,
+      "step": 6580
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8445857744590692,
+      "learning_rate": 1.919123715892889e-05,
+      "loss": 0.6938,
+      "step": 6581
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5873503714166441,
+      "learning_rate": 1.919093651762731e-05,
+      "loss": 0.8874,
+      "step": 6582
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.68991039041895,
+      "learning_rate": 1.919063582281319e-05,
+      "loss": 0.7936,
+      "step": 6583
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7327913137475048,
+      "learning_rate": 1.9190335074488275e-05,
+      "loss": 0.7122,
+      "step": 6584
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.7413978140713087,
+      "learning_rate": 1.9190034272654316e-05,
+      "loss": 0.7712,
+      "step": 6585
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.4533540788901465,
+      "learning_rate": 1.9189733417313073e-05,
+      "loss": 0.7319,
+      "step": 6586
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.731725666471387,
+      "learning_rate": 1.918943250846629e-05,
+      "loss": 0.8129,
+      "step": 6587
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4728494606925793,
+      "learning_rate": 1.9189131546115725e-05,
+      "loss": 0.8067,
+      "step": 6588
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.502478771272758,
+      "learning_rate": 1.9188830530263124e-05,
+      "loss": 0.698,
+      "step": 6589
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9429239174124033,
+      "learning_rate": 1.9188529460910244e-05,
+      "loss": 0.6902,
+      "step": 6590
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6229734144492795,
+      "learning_rate": 1.9188228338058835e-05,
+      "loss": 0.587,
+      "step": 6591
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4729010052551645,
+      "learning_rate": 1.918792716171065e-05,
+      "loss": 0.7637,
+      "step": 6592
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.801937548223643,
+      "learning_rate": 1.9187625931867448e-05,
+      "loss": 0.7641,
+      "step": 6593
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.560477949331691,
+      "learning_rate": 1.9187324648530977e-05,
+      "loss": 0.6777,
+      "step": 6594
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.600433977008859,
+      "learning_rate": 1.9187023311702996e-05,
+      "loss": 0.6301,
+      "step": 6595
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.601380461621466,
+      "learning_rate": 1.9186721921385257e-05,
+      "loss": 0.6929,
+      "step": 6596
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.520298250924491,
+      "learning_rate": 1.9186420477579508e-05,
+      "loss": 0.776,
+      "step": 6597
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.768543875049444,
+      "learning_rate": 1.918611898028752e-05,
+      "loss": 0.7252,
+      "step": 6598
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9826941665865807,
+      "learning_rate": 1.9185817429511033e-05,
+      "loss": 0.7123,
+      "step": 6599
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.7086703716945584,
+      "learning_rate": 1.9185515825251806e-05,
+      "loss": 0.7324,
+      "step": 6600
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.573618887747313,
+      "learning_rate": 1.91852141675116e-05,
+      "loss": 0.7398,
+      "step": 6601
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8828592270193696,
+      "learning_rate": 1.9184912456292172e-05,
+      "loss": 0.6618,
+      "step": 6602
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.5427853175118855,
+      "learning_rate": 1.9184610691595273e-05,
+      "loss": 0.7291,
+      "step": 6603
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5685254903252138,
+      "learning_rate": 1.9184308873422666e-05,
+      "loss": 0.4948,
+      "step": 6604
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6637949866509474,
+      "learning_rate": 1.9184007001776103e-05,
+      "loss": 0.7827,
+      "step": 6605
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5344986706430348,
+      "learning_rate": 1.9183705076657343e-05,
+      "loss": 0.7676,
+      "step": 6606
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8784804259423538,
+      "learning_rate": 1.9183403098068145e-05,
+      "loss": 0.9163,
+      "step": 6607
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0668375223736377,
+      "learning_rate": 1.918310106601027e-05,
+      "loss": 0.7418,
+      "step": 6608
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.614580050304624,
+      "learning_rate": 1.9182798980485468e-05,
+      "loss": 0.5508,
+      "step": 6609
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.391134063533499,
+      "learning_rate": 1.9182496841495504e-05,
+      "loss": 0.7489,
+      "step": 6610
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7696362738611593,
+      "learning_rate": 1.9182194649042138e-05,
+      "loss": 0.779,
+      "step": 6611
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6218177432466863,
+      "learning_rate": 1.918189240312713e-05,
+      "loss": 0.7605,
+      "step": 6612
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.855881787171793,
+      "learning_rate": 1.918159010375223e-05,
+      "loss": 0.7079,
+      "step": 6613
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.825345881247662,
+      "learning_rate": 1.9181287750919214e-05,
+      "loss": 0.7906,
+      "step": 6614
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7468750643768804,
+      "learning_rate": 1.918098534462983e-05,
+      "loss": 0.6863,
+      "step": 6615
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8608854710075766,
+      "learning_rate": 1.918068288488584e-05,
+      "loss": 0.6062,
+      "step": 6616
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5659255235497276,
+      "learning_rate": 1.918038037168901e-05,
+      "loss": 0.5357,
+      "step": 6617
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5835987273386437,
+      "learning_rate": 1.9180077805041103e-05,
+      "loss": 0.5611,
+      "step": 6618
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7109037279879515,
+      "learning_rate": 1.917977518494387e-05,
+      "loss": 0.78,
+      "step": 6619
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.012612585300797,
+      "learning_rate": 1.9179472511399084e-05,
+      "loss": 0.7294,
+      "step": 6620
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6983088310831147,
+      "learning_rate": 1.91791697844085e-05,
+      "loss": 0.7338,
+      "step": 6621
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6271381693714919,
+      "learning_rate": 1.917886700397389e-05,
+      "loss": 0.4795,
+      "step": 6622
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.824727443294312,
+      "learning_rate": 1.9178564170097004e-05,
+      "loss": 0.9527,
+      "step": 6623
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6760018191621435,
+      "learning_rate": 1.9178261282779614e-05,
+      "loss": 0.8047,
+      "step": 6624
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.710161085749103,
+      "learning_rate": 1.9177958342023478e-05,
+      "loss": 0.6315,
+      "step": 6625
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9315278884366485,
+      "learning_rate": 1.9177655347830368e-05,
+      "loss": 0.7782,
+      "step": 6626
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.18282462839473,
+      "learning_rate": 1.9177352300202038e-05,
+      "loss": 0.6698,
+      "step": 6627
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9512356397558441,
+      "learning_rate": 1.917704919914026e-05,
+      "loss": 0.8997,
+      "step": 6628
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7796806667146876,
+      "learning_rate": 1.9176746044646796e-05,
+      "loss": 0.8824,
+      "step": 6629
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0098753521861994,
+      "learning_rate": 1.917644283672341e-05,
+      "loss": 0.7352,
+      "step": 6630
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5912476751326022,
+      "learning_rate": 1.9176139575371867e-05,
+      "loss": 0.5557,
+      "step": 6631
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7597291171402278,
+      "learning_rate": 1.9175836260593937e-05,
+      "loss": 0.6878,
+      "step": 6632
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6026782594694289,
+      "learning_rate": 1.9175532892391383e-05,
+      "loss": 0.5648,
+      "step": 6633
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6237585786795966,
+      "learning_rate": 1.9175229470765972e-05,
+      "loss": 0.5587,
+      "step": 6634
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.104689747053254,
+      "learning_rate": 1.917492599571947e-05,
+      "loss": 0.8457,
+      "step": 6635
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6394497213049677,
+      "learning_rate": 1.9174622467253644e-05,
+      "loss": 0.7908,
+      "step": 6636
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0502454896726117,
+      "learning_rate": 1.917431888537026e-05,
+      "loss": 0.8633,
+      "step": 6637
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5772849154823638,
+      "learning_rate": 1.917401525007109e-05,
+      "loss": 0.7081,
+      "step": 6638
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9074313606555597,
+      "learning_rate": 1.9173711561357895e-05,
+      "loss": 0.7937,
+      "step": 6639
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4318731731606522,
+      "learning_rate": 1.917340781923245e-05,
+      "loss": 0.6873,
+      "step": 6640
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.396017539811793,
+      "learning_rate": 1.917310402369652e-05,
+      "loss": 0.7831,
+      "step": 6641
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6059889078857432,
+      "learning_rate": 1.9172800174751873e-05,
+      "loss": 0.5556,
+      "step": 6642
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5561510563863932,
+      "learning_rate": 1.9172496272400282e-05,
+      "loss": 0.7327,
+      "step": 6643
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.843833283487108,
+      "learning_rate": 1.917219231664351e-05,
+      "loss": 0.805,
+      "step": 6644
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6627133692233447,
+      "learning_rate": 1.917188830748333e-05,
+      "loss": 0.8046,
+      "step": 6645
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7285952509472367,
+      "learning_rate": 1.9171584244921516e-05,
+      "loss": 0.7433,
+      "step": 6646
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.624643992249304,
+      "learning_rate": 1.9171280128959834e-05,
+      "loss": 0.6911,
+      "step": 6647
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3913213355063039,
+      "learning_rate": 1.9170975959600053e-05,
+      "loss": 0.7289,
+      "step": 6648
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8538674445958552,
+      "learning_rate": 1.917067173684395e-05,
+      "loss": 0.7159,
+      "step": 6649
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.582106758875748,
+      "learning_rate": 1.917036746069329e-05,
+      "loss": 0.7624,
+      "step": 6650
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6731370619576138,
+      "learning_rate": 1.917006313114985e-05,
+      "loss": 0.7602,
+      "step": 6651
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4209780479648468,
+      "learning_rate": 1.9169758748215396e-05,
+      "loss": 0.681,
+      "step": 6652
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7836177809035656,
+      "learning_rate": 1.9169454311891705e-05,
+      "loss": 0.7402,
+      "step": 6653
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.485494973982504,
+      "learning_rate": 1.916914982218055e-05,
+      "loss": 0.6909,
+      "step": 6654
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7351318308914303,
+      "learning_rate": 1.9168845279083697e-05,
+      "loss": 0.6717,
+      "step": 6655
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7632890142629858,
+      "learning_rate": 1.9168540682602928e-05,
+      "loss": 0.7949,
+      "step": 6656
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4864752433550297,
+      "learning_rate": 1.916823603274001e-05,
+      "loss": 0.8345,
+      "step": 6657
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.528455673167151,
+      "learning_rate": 1.916793132949672e-05,
+      "loss": 0.7527,
+      "step": 6658
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3896241586573828,
+      "learning_rate": 1.916762657287483e-05,
+      "loss": 0.8261,
+      "step": 6659
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4173247416651065,
+      "learning_rate": 1.916732176287612e-05,
+      "loss": 0.8325,
+      "step": 6660
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3858242469936732,
+      "learning_rate": 1.9167016899502357e-05,
+      "loss": 0.7448,
+      "step": 6661
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0317011347400045,
+      "learning_rate": 1.916671198275532e-05,
+      "loss": 0.7607,
+      "step": 6662
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3974593529048303,
+      "learning_rate": 1.9166407012636785e-05,
+      "loss": 0.6921,
+      "step": 6663
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7348618750210185,
+      "learning_rate": 1.9166101989148527e-05,
+      "loss": 0.7751,
+      "step": 6664
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4803836219563413,
+      "learning_rate": 1.9165796912292317e-05,
+      "loss": 0.8373,
+      "step": 6665
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9487104556298118,
+      "learning_rate": 1.9165491782069942e-05,
+      "loss": 0.767,
+      "step": 6666
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0106598295922296,
+      "learning_rate": 1.9165186598483166e-05,
+      "loss": 0.8433,
+      "step": 6667
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5550323456019597,
+      "learning_rate": 1.9164881361533775e-05,
+      "loss": 0.6568,
+      "step": 6668
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8716244189249183,
+      "learning_rate": 1.9164576071223542e-05,
+      "loss": 0.9511,
+      "step": 6669
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4331273862713885,
+      "learning_rate": 1.9164270727554247e-05,
+      "loss": 0.6844,
+      "step": 6670
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6363817648997163,
+      "learning_rate": 1.9163965330527664e-05,
+      "loss": 0.9081,
+      "step": 6671
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.325026491319115,
+      "learning_rate": 1.9163659880145575e-05,
+      "loss": 0.8375,
+      "step": 6672
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.405966918341967,
+      "learning_rate": 1.9163354376409757e-05,
+      "loss": 0.8993,
+      "step": 6673
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5946846819217193,
+      "learning_rate": 1.9163048819321987e-05,
+      "loss": 0.7456,
+      "step": 6674
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4663608560686734,
+      "learning_rate": 1.916274320888405e-05,
+      "loss": 0.6725,
+      "step": 6675
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5507105613808458,
+      "learning_rate": 1.916243754509772e-05,
+      "loss": 0.7726,
+      "step": 6676
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7705683057755868,
+      "learning_rate": 1.9162131827964774e-05,
+      "loss": 0.8285,
+      "step": 6677
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1661515436592396,
+      "learning_rate": 1.9161826057487e-05,
+      "loss": 0.7395,
+      "step": 6678
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5815993562782513,
+      "learning_rate": 1.9161520233666172e-05,
+      "loss": 0.7287,
+      "step": 6679
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7446050397897717,
+      "learning_rate": 1.9161214356504072e-05,
+      "loss": 0.8353,
+      "step": 6680
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9138863002710784,
+      "learning_rate": 1.916090842600248e-05,
+      "loss": 0.8356,
+      "step": 6681
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.311615997340161,
+      "learning_rate": 1.9160602442163182e-05,
+      "loss": 0.7721,
+      "step": 6682
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7152037559007722,
+      "learning_rate": 1.916029640498796e-05,
+      "loss": 0.6996,
+      "step": 6683
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9961707203620764,
+      "learning_rate": 1.915999031447859e-05,
+      "loss": 0.8073,
+      "step": 6684
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6108967248253435,
+      "learning_rate": 1.915968417063685e-05,
+      "loss": 0.5536,
+      "step": 6685
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.497719630826897,
+      "learning_rate": 1.9159377973464536e-05,
+      "loss": 0.9343,
+      "step": 6686
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5360793232585601,
+      "learning_rate": 1.9159071722963422e-05,
+      "loss": 0.755,
+      "step": 6687
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7729593166443545,
+      "learning_rate": 1.915876541913529e-05,
+      "loss": 0.6751,
+      "step": 6688
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7272690141574856,
+      "learning_rate": 1.9158459061981928e-05,
+      "loss": 0.7418,
+      "step": 6689
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7045336783206342,
+      "learning_rate": 1.9158152651505116e-05,
+      "loss": 0.892,
+      "step": 6690
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.694937462113882,
+      "learning_rate": 1.9157846187706642e-05,
+      "loss": 0.718,
+      "step": 6691
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3678733164188872,
+      "learning_rate": 1.9157539670588287e-05,
+      "loss": 0.5892,
+      "step": 6692
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5083087143777372,
+      "learning_rate": 1.915723310015184e-05,
+      "loss": 0.6638,
+      "step": 6693
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8619972210170792,
+      "learning_rate": 1.915692647639908e-05,
+      "loss": 0.8058,
+      "step": 6694
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8989234033312468,
+      "learning_rate": 1.9156619799331796e-05,
+      "loss": 0.8054,
+      "step": 6695
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5096107506374663,
+      "learning_rate": 1.9156313068951772e-05,
+      "loss": 0.8676,
+      "step": 6696
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2382360772331236,
+      "learning_rate": 1.9156006285260792e-05,
+      "loss": 0.7413,
+      "step": 6697
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4415796706492383,
+      "learning_rate": 1.9155699448260648e-05,
+      "loss": 0.7968,
+      "step": 6698
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3922417120000283,
+      "learning_rate": 1.9155392557953126e-05,
+      "loss": 0.6876,
+      "step": 6699
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5643208676937446,
+      "learning_rate": 1.9155085614340004e-05,
+      "loss": 0.5237,
+      "step": 6700
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5347015528595254,
+      "learning_rate": 1.915477861742308e-05,
+      "loss": 0.6367,
+      "step": 6701
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4184744303238612,
+      "learning_rate": 1.9154471567204132e-05,
+      "loss": 0.576,
+      "step": 6702
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4769076711411837,
+      "learning_rate": 1.9154164463684953e-05,
+      "loss": 0.8064,
+      "step": 6703
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5500447984552328,
+      "learning_rate": 1.915385730686733e-05,
+      "loss": 0.8306,
+      "step": 6704
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7278590403909881,
+      "learning_rate": 1.9153550096753056e-05,
+      "loss": 0.8406,
+      "step": 6705
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.815142067696035,
+      "learning_rate": 1.9153242833343912e-05,
+      "loss": 0.8317,
+      "step": 6706
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.5932739137126846,
+      "learning_rate": 1.915293551664169e-05,
+      "loss": 0.8501,
+      "step": 6707
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7735762560146364,
+      "learning_rate": 1.9152628146648183e-05,
+      "loss": 0.6848,
+      "step": 6708
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7376724740008844,
+      "learning_rate": 1.9152320723365176e-05,
+      "loss": 0.7399,
+      "step": 6709
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4888242244956802,
+      "learning_rate": 1.9152013246794455e-05,
+      "loss": 0.7794,
+      "step": 6710
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.596322986941352,
+      "learning_rate": 1.915170571693782e-05,
+      "loss": 0.909,
+      "step": 6711
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6425135915960023,
+      "learning_rate": 1.915139813379706e-05,
+      "loss": 0.7422,
+      "step": 6712
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5135850674883078,
+      "learning_rate": 1.9151090497373955e-05,
+      "loss": 0.7217,
+      "step": 6713
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7326656389071429,
+      "learning_rate": 1.915078280767031e-05,
+      "loss": 0.6844,
+      "step": 6714
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9117610605121487,
+      "learning_rate": 1.9150475064687907e-05,
+      "loss": 0.7488,
+      "step": 6715
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5922802457135142,
+      "learning_rate": 1.9150167268428544e-05,
+      "loss": 0.5419,
+      "step": 6716
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6412065069405455,
+      "learning_rate": 1.9149859418894007e-05,
+      "loss": 0.5558,
+      "step": 6717
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4601851861426687,
+      "learning_rate": 1.9149551516086095e-05,
+      "loss": 0.7021,
+      "step": 6718
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4640180013521542,
+      "learning_rate": 1.9149243560006598e-05,
+      "loss": 0.8782,
+      "step": 6719
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6477476577235342,
+      "learning_rate": 1.9148935550657306e-05,
+      "loss": 0.6907,
+      "step": 6720
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4205686628545398,
+      "learning_rate": 1.9148627488040014e-05,
+      "loss": 0.9441,
+      "step": 6721
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6675623340341708,
+      "learning_rate": 1.914831937215652e-05,
+      "loss": 0.7508,
+      "step": 6722
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6729676757993137,
+      "learning_rate": 1.9148011203008607e-05,
+      "loss": 0.8121,
+      "step": 6723
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7569260017027062,
+      "learning_rate": 1.9147702980598085e-05,
+      "loss": 0.8001,
+      "step": 6724
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8866042842375426,
+      "learning_rate": 1.9147394704926734e-05,
+      "loss": 0.6794,
+      "step": 6725
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5306873756741926,
+      "learning_rate": 1.914708637599636e-05,
+      "loss": 0.7189,
+      "step": 6726
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6715515824470764,
+      "learning_rate": 1.914677799380875e-05,
+      "loss": 0.7397,
+      "step": 6727
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4341647946295084,
+      "learning_rate": 1.9146469558365703e-05,
+      "loss": 0.7317,
+      "step": 6728
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3867622121548973,
+      "learning_rate": 1.9146161069669015e-05,
+      "loss": 0.683,
+      "step": 6729
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5925014288133498,
+      "learning_rate": 1.914585252772048e-05,
+      "loss": 0.518,
+      "step": 6730
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.576874528846159,
+      "learning_rate": 1.9145543932521898e-05,
+      "loss": 0.5218,
+      "step": 6731
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.490893112645404,
+      "learning_rate": 1.9145235284075064e-05,
+      "loss": 0.8168,
+      "step": 6732
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.734701949080198,
+      "learning_rate": 1.9144926582381777e-05,
+      "loss": 0.8026,
+      "step": 6733
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4276149099101139,
+      "learning_rate": 1.9144617827443826e-05,
+      "loss": 0.6961,
+      "step": 6734
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5499681339718387,
+      "learning_rate": 1.914430901926302e-05,
+      "loss": 0.702,
+      "step": 6735
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9490080687181046,
+      "learning_rate": 1.9144000157841146e-05,
+      "loss": 0.7082,
+      "step": 6736
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5727673782484004,
+      "learning_rate": 1.9143691243180015e-05,
+      "loss": 0.7523,
+      "step": 6737
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6058946586076441,
+      "learning_rate": 1.9143382275281416e-05,
+      "loss": 0.9079,
+      "step": 6738
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.0332305067915724,
+      "learning_rate": 1.914307325414715e-05,
+      "loss": 0.6043,
+      "step": 6739
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8296010402024938,
+      "learning_rate": 1.914276417977902e-05,
+      "loss": 0.7309,
+      "step": 6740
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6466703240799774,
+      "learning_rate": 1.9142455052178817e-05,
+      "loss": 0.7981,
+      "step": 6741
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5289249650388328,
+      "learning_rate": 1.914214587134835e-05,
+      "loss": 0.7905,
+      "step": 6742
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0996698354108343,
+      "learning_rate": 1.9141836637289414e-05,
+      "loss": 0.8022,
+      "step": 6743
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5862217654163283,
+      "learning_rate": 1.9141527350003812e-05,
+      "loss": 0.5136,
+      "step": 6744
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6274595079890773,
+      "learning_rate": 1.9141218009493344e-05,
+      "loss": 0.5291,
+      "step": 6745
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1026081272719273,
+      "learning_rate": 1.9140908615759807e-05,
+      "loss": 0.7774,
+      "step": 6746
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6198777605040646,
+      "learning_rate": 1.914059916880501e-05,
+      "loss": 0.8057,
+      "step": 6747
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.593728008199245,
+      "learning_rate": 1.9140289668630753e-05,
+      "loss": 0.5331,
+      "step": 6748
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.629825078843533,
+      "learning_rate": 1.9139980115238827e-05,
+      "loss": 0.7184,
+      "step": 6749
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8203329304965306,
+      "learning_rate": 1.913967050863105e-05,
+      "loss": 0.8647,
+      "step": 6750
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4266361032629775,
+      "learning_rate": 1.9139360848809215e-05,
+      "loss": 0.8381,
+      "step": 6751
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.601426655725888,
+      "learning_rate": 1.913905113577513e-05,
+      "loss": 0.8222,
+      "step": 6752
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1023348402653634,
+      "learning_rate": 1.9138741369530597e-05,
+      "loss": 0.7849,
+      "step": 6753
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4672357528572506,
+      "learning_rate": 1.9138431550077414e-05,
+      "loss": 0.7799,
+      "step": 6754
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.73106005909144,
+      "learning_rate": 1.913812167741739e-05,
+      "loss": 0.7636,
+      "step": 6755
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.57557630891284,
+      "learning_rate": 1.9137811751552332e-05,
+      "loss": 0.8987,
+      "step": 6756
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.90843316688987,
+      "learning_rate": 1.913750177248404e-05,
+      "loss": 0.7035,
+      "step": 6757
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3579890196709796,
+      "learning_rate": 1.913719174021432e-05,
+      "loss": 0.78,
+      "step": 6758
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8539048618611813,
+      "learning_rate": 1.9136881654744973e-05,
+      "loss": 0.6663,
+      "step": 6759
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.31264877722988,
+      "learning_rate": 1.913657151607781e-05,
+      "loss": 0.8067,
+      "step": 6760
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4705007185730348,
+      "learning_rate": 1.913626132421464e-05,
+      "loss": 0.8335,
+      "step": 6761
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.598066763206946,
+      "learning_rate": 1.913595107915726e-05,
+      "loss": 0.8082,
+      "step": 6762
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.429830962365767,
+      "learning_rate": 1.913564078090748e-05,
+      "loss": 0.8192,
+      "step": 6763
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.387734378642148,
+      "learning_rate": 1.9135330429467105e-05,
+      "loss": 0.8091,
+      "step": 6764
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6176705120826906,
+      "learning_rate": 1.913502002483795e-05,
+      "loss": 0.5567,
+      "step": 6765
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4644988672645072,
+      "learning_rate": 1.9134709567021808e-05,
+      "loss": 0.8531,
+      "step": 6766
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9451281272592946,
+      "learning_rate": 1.9134399056020504e-05,
+      "loss": 0.7256,
+      "step": 6767
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6742974544991542,
+      "learning_rate": 1.913408849183583e-05,
+      "loss": 0.7858,
+      "step": 6768
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9363523201011468,
+      "learning_rate": 1.9133777874469605e-05,
+      "loss": 0.8345,
+      "step": 6769
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3586528208741515,
+      "learning_rate": 1.913346720392363e-05,
+      "loss": 0.8339,
+      "step": 6770
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1220308473024936,
+      "learning_rate": 1.913315648019972e-05,
+      "loss": 0.7505,
+      "step": 6771
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8574373582369248,
+      "learning_rate": 1.9132845703299683e-05,
+      "loss": 0.8358,
+      "step": 6772
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3813507108845888,
+      "learning_rate": 1.9132534873225323e-05,
+      "loss": 0.758,
+      "step": 6773
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6812775959512964,
+      "learning_rate": 1.9132223989978457e-05,
+      "loss": 0.6774,
+      "step": 6774
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4072064414040193,
+      "learning_rate": 1.913191305356089e-05,
+      "loss": 0.8572,
+      "step": 6775
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.004662791423416,
+      "learning_rate": 1.9131602063974434e-05,
+      "loss": 0.7248,
+      "step": 6776
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8732924168968155,
+      "learning_rate": 1.9131291021220902e-05,
+      "loss": 0.7284,
+      "step": 6777
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8720908304858885,
+      "learning_rate": 1.91309799253021e-05,
+      "loss": 0.809,
+      "step": 6778
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6774014680190892,
+      "learning_rate": 1.9130668776219843e-05,
+      "loss": 0.6572,
+      "step": 6779
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.707027584174015,
+      "learning_rate": 1.913035757397594e-05,
+      "loss": 0.6234,
+      "step": 6780
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8834019802482826,
+      "learning_rate": 1.913004631857221e-05,
+      "loss": 0.6911,
+      "step": 6781
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.466138095880133,
+      "learning_rate": 1.9129735010010452e-05,
+      "loss": 0.8746,
+      "step": 6782
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.8552327099838504,
+      "learning_rate": 1.9129423648292492e-05,
+      "loss": 0.8925,
+      "step": 6783
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7398029785110092,
+      "learning_rate": 1.9129112233420138e-05,
+      "loss": 0.7332,
+      "step": 6784
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5739891741720771,
+      "learning_rate": 1.9128800765395196e-05,
+      "loss": 0.5252,
+      "step": 6785
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3755191743876654,
+      "learning_rate": 1.9128489244219492e-05,
+      "loss": 0.9285,
+      "step": 6786
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9961564099813038,
+      "learning_rate": 1.9128177669894832e-05,
+      "loss": 0.7638,
+      "step": 6787
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8803946069311175,
+      "learning_rate": 1.9127866042423033e-05,
+      "loss": 0.6656,
+      "step": 6788
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9791251208479914,
+      "learning_rate": 1.9127554361805902e-05,
+      "loss": 0.846,
+      "step": 6789
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7450326330609296,
+      "learning_rate": 1.9127242628045268e-05,
+      "loss": 0.7473,
+      "step": 6790
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6924242320965721,
+      "learning_rate": 1.912693084114293e-05,
+      "loss": 0.7467,
+      "step": 6791
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0183626055795436,
+      "learning_rate": 1.9126619001100715e-05,
+      "loss": 0.7569,
+      "step": 6792
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7132651202623632,
+      "learning_rate": 1.9126307107920433e-05,
+      "loss": 0.7233,
+      "step": 6793
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3890941259646414,
+      "learning_rate": 1.9125995161603904e-05,
+      "loss": 0.8101,
+      "step": 6794
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9126719675325496,
+      "learning_rate": 1.912568316215294e-05,
+      "loss": 0.662,
+      "step": 6795
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6584959691435424,
+      "learning_rate": 1.9125371109569362e-05,
+      "loss": 0.8131,
+      "step": 6796
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5140091521449432,
+      "learning_rate": 1.912505900385498e-05,
+      "loss": 0.9074,
+      "step": 6797
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.701773125570326,
+      "learning_rate": 1.912474684501162e-05,
+      "loss": 0.8459,
+      "step": 6798
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7402689197799732,
+      "learning_rate": 1.9124434633041097e-05,
+      "loss": 0.858,
+      "step": 6799
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.42764687519734,
+      "learning_rate": 1.9124122367945223e-05,
+      "loss": 0.7421,
+      "step": 6800
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6778799422052642,
+      "learning_rate": 1.912381004972582e-05,
+      "loss": 0.756,
+      "step": 6801
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.654096095415917,
+      "learning_rate": 1.9123497678384706e-05,
+      "loss": 0.7294,
+      "step": 6802
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0584642805544378,
+      "learning_rate": 1.91231852539237e-05,
+      "loss": 0.7174,
+      "step": 6803
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7980580052967752,
+      "learning_rate": 1.9122872776344622e-05,
+      "loss": 0.7199,
+      "step": 6804
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3871037723568571,
+      "learning_rate": 1.912256024564929e-05,
+      "loss": 0.8825,
+      "step": 6805
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3845543660098691,
+      "learning_rate": 1.9122247661839524e-05,
+      "loss": 0.8018,
+      "step": 6806
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.464659111337137,
+      "learning_rate": 1.9121935024917144e-05,
+      "loss": 0.7335,
+      "step": 6807
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5976255724696863,
+      "learning_rate": 1.9121622334883973e-05,
+      "loss": 0.5428,
+      "step": 6808
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.453890330022671,
+      "learning_rate": 1.9121309591741822e-05,
+      "loss": 0.7877,
+      "step": 6809
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6230541459434542,
+      "learning_rate": 1.9120996795492525e-05,
+      "loss": 0.8192,
+      "step": 6810
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4745533241938702,
+      "learning_rate": 1.9120683946137894e-05,
+      "loss": 0.7541,
+      "step": 6811
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6307466881065356,
+      "learning_rate": 1.9120371043679756e-05,
+      "loss": 0.8367,
+      "step": 6812
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7958097294339141,
+      "learning_rate": 1.912005808811993e-05,
+      "loss": 0.7571,
+      "step": 6813
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6492771133346464,
+      "learning_rate": 1.9119745079460238e-05,
+      "loss": 0.7555,
+      "step": 6814
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5649366569986305,
+      "learning_rate": 1.91194320177025e-05,
+      "loss": 0.7663,
+      "step": 6815
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5697050371487584,
+      "learning_rate": 1.9119118902848546e-05,
+      "loss": 0.8255,
+      "step": 6816
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6162356702981433,
+      "learning_rate": 1.9118805734900192e-05,
+      "loss": 0.529,
+      "step": 6817
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6157798819820703,
+      "learning_rate": 1.9118492513859264e-05,
+      "loss": 0.8535,
+      "step": 6818
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.934328983387113,
+      "learning_rate": 1.911817923972759e-05,
+      "loss": 0.8178,
+      "step": 6819
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.429554343027893,
+      "learning_rate": 1.9117865912506984e-05,
+      "loss": 0.8414,
+      "step": 6820
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7947402542204116,
+      "learning_rate": 1.9117552532199278e-05,
+      "loss": 0.6292,
+      "step": 6821
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6184117549608956,
+      "learning_rate": 1.9117239098806296e-05,
+      "loss": 0.735,
+      "step": 6822
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7034639410838357,
+      "learning_rate": 1.911692561232986e-05,
+      "loss": 0.8401,
+      "step": 6823
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6366384294119827,
+      "learning_rate": 1.9116612072771798e-05,
+      "loss": 0.5572,
+      "step": 6824
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6993654493247183,
+      "learning_rate": 1.911629848013393e-05,
+      "loss": 0.7161,
+      "step": 6825
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8593804745095166,
+      "learning_rate": 1.911598483441809e-05,
+      "loss": 0.7655,
+      "step": 6826
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5411134279247827,
+      "learning_rate": 1.91156711356261e-05,
+      "loss": 0.7576,
+      "step": 6827
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6506379128195345,
+      "learning_rate": 1.911535738375979e-05,
+      "loss": 0.7411,
+      "step": 6828
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.474313712607556,
+      "learning_rate": 1.9115043578820978e-05,
+      "loss": 0.7301,
+      "step": 6829
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.469864389748771,
+      "learning_rate": 1.9114729720811497e-05,
+      "loss": 0.7434,
+      "step": 6830
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.347559361564211,
+      "learning_rate": 1.9114415809733176e-05,
+      "loss": 0.7171,
+      "step": 6831
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3745072601459478,
+      "learning_rate": 1.911410184558784e-05,
+      "loss": 0.7976,
+      "step": 6832
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6123105643043992,
+      "learning_rate": 1.911378782837732e-05,
+      "loss": 0.5712,
+      "step": 6833
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.886960047429633,
+      "learning_rate": 1.9113473758103435e-05,
+      "loss": 0.7217,
+      "step": 6834
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6280491689266271,
+      "learning_rate": 1.9113159634768025e-05,
+      "loss": 0.5354,
+      "step": 6835
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9359065230261512,
+      "learning_rate": 1.9112845458372914e-05,
+      "loss": 0.9028,
+      "step": 6836
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8218009684912646,
+      "learning_rate": 1.9112531228919933e-05,
+      "loss": 0.7598,
+      "step": 6837
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5538295023867597,
+      "learning_rate": 1.911221694641091e-05,
+      "loss": 0.954,
+      "step": 6838
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6760416856829345,
+      "learning_rate": 1.9111902610847674e-05,
+      "loss": 0.7477,
+      "step": 6839
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3007242229905236,
+      "learning_rate": 1.9111588222232056e-05,
+      "loss": 0.8957,
+      "step": 6840
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6269781272659039,
+      "learning_rate": 1.9111273780565886e-05,
+      "loss": 0.7252,
+      "step": 6841
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3552933191289585,
+      "learning_rate": 1.9110959285850995e-05,
+      "loss": 0.7578,
+      "step": 6842
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5682792745738927,
+      "learning_rate": 1.9110644738089216e-05,
+      "loss": 0.7612,
+      "step": 6843
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3272903111092722,
+      "learning_rate": 1.911033013728238e-05,
+      "loss": 0.7766,
+      "step": 6844
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.707167550944172,
+      "learning_rate": 1.9110015483432318e-05,
+      "loss": 0.7623,
+      "step": 6845
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0058480305267317,
+      "learning_rate": 1.9109700776540857e-05,
+      "loss": 0.8308,
+      "step": 6846
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4802043524361466,
+      "learning_rate": 1.910938601660984e-05,
+      "loss": 0.831,
+      "step": 6847
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.847855026696846,
+      "learning_rate": 1.910907120364109e-05,
+      "loss": 0.7373,
+      "step": 6848
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5574148386449507,
+      "learning_rate": 1.9108756337636446e-05,
+      "loss": 0.7558,
+      "step": 6849
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6904634650217585,
+      "learning_rate": 1.9108441418597737e-05,
+      "loss": 0.8846,
+      "step": 6850
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5995474009864266,
+      "learning_rate": 1.9108126446526797e-05,
+      "loss": 0.5276,
+      "step": 6851
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5940353157058483,
+      "learning_rate": 1.9107811421425465e-05,
+      "loss": 0.5315,
+      "step": 6852
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2494152346238114,
+      "learning_rate": 1.9107496343295567e-05,
+      "loss": 0.7823,
+      "step": 6853
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.288238414656139,
+      "learning_rate": 1.9107181212138945e-05,
+      "loss": 0.5978,
+      "step": 6854
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7136950723817577,
+      "learning_rate": 1.910686602795743e-05,
+      "loss": 0.7829,
+      "step": 6855
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.734445954631157,
+      "learning_rate": 1.9106550790752857e-05,
+      "loss": 0.7544,
+      "step": 6856
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7790249508356486,
+      "learning_rate": 1.9106235500527065e-05,
+      "loss": 0.7742,
+      "step": 6857
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4986269568460708,
+      "learning_rate": 1.9105920157281883e-05,
+      "loss": 0.7696,
+      "step": 6858
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4783152736494818,
+      "learning_rate": 1.9105604761019153e-05,
+      "loss": 0.6809,
+      "step": 6859
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3321541327006756,
+      "learning_rate": 1.910528931174071e-05,
+      "loss": 0.8684,
+      "step": 6860
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8371790769094563,
+      "learning_rate": 1.9104973809448388e-05,
+      "loss": 0.8927,
+      "step": 6861
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.79572178618506,
+      "learning_rate": 1.9104658254144025e-05,
+      "loss": 0.788,
+      "step": 6862
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7609926218783474,
+      "learning_rate": 1.9104342645829464e-05,
+      "loss": 0.7179,
+      "step": 6863
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.585313297307637,
+      "learning_rate": 1.9104026984506533e-05,
+      "loss": 0.7967,
+      "step": 6864
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5839250545162422,
+      "learning_rate": 1.9103711270177077e-05,
+      "loss": 0.7127,
+      "step": 6865
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5867005352297979,
+      "learning_rate": 1.9103395502842932e-05,
+      "loss": 0.5432,
+      "step": 6866
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.691963952404856,
+      "learning_rate": 1.9103079682505938e-05,
+      "loss": 0.779,
+      "step": 6867
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8093704802720671,
+      "learning_rate": 1.9102763809167927e-05,
+      "loss": 0.785,
+      "step": 6868
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7864583573096853,
+      "learning_rate": 1.9102447882830745e-05,
+      "loss": 0.7359,
+      "step": 6869
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5978631092924982,
+      "learning_rate": 1.910213190349623e-05,
+      "loss": 0.9268,
+      "step": 6870
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.6506863326223082,
+      "learning_rate": 1.9101815871166222e-05,
+      "loss": 0.7069,
+      "step": 6871
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2555578693005973,
+      "learning_rate": 1.9101499785842563e-05,
+      "loss": 0.8157,
+      "step": 6872
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.473507052669689,
+      "learning_rate": 1.9101183647527086e-05,
+      "loss": 0.8032,
+      "step": 6873
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.450046036002067,
+      "learning_rate": 1.910086745622164e-05,
+      "loss": 0.6068,
+      "step": 6874
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6181528560145917,
+      "learning_rate": 1.910055121192806e-05,
+      "loss": 0.8376,
+      "step": 6875
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6625382295584143,
+      "learning_rate": 1.910023491464819e-05,
+      "loss": 0.9509,
+      "step": 6876
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5183102299611249,
+      "learning_rate": 1.9099918564383873e-05,
+      "loss": 0.8576,
+      "step": 6877
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5991200149146656,
+      "learning_rate": 1.9099602161136947e-05,
+      "loss": 0.8399,
+      "step": 6878
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3909933805344163,
+      "learning_rate": 1.9099285704909256e-05,
+      "loss": 0.7947,
+      "step": 6879
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7003451688938196,
+      "learning_rate": 1.9098969195702646e-05,
+      "loss": 0.7552,
+      "step": 6880
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6052605175956727,
+      "learning_rate": 1.9098652633518956e-05,
+      "loss": 0.5568,
+      "step": 6881
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.401733601663215,
+      "learning_rate": 1.9098336018360027e-05,
+      "loss": 0.7699,
+      "step": 6882
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5662140665184805,
+      "learning_rate": 1.9098019350227705e-05,
+      "loss": 0.8387,
+      "step": 6883
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8262889167388643,
+      "learning_rate": 1.9097702629123836e-05,
+      "loss": 0.7545,
+      "step": 6884
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8875793248812311,
+      "learning_rate": 1.9097385855050263e-05,
+      "loss": 0.8154,
+      "step": 6885
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7914933522410144,
+      "learning_rate": 1.9097069028008827e-05,
+      "loss": 0.815,
+      "step": 6886
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5035187077942855,
+      "learning_rate": 1.9096752148001378e-05,
+      "loss": 0.7276,
+      "step": 6887
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5659655977869138,
+      "learning_rate": 1.9096435215029757e-05,
+      "loss": 0.6914,
+      "step": 6888
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.184786644446292,
+      "learning_rate": 1.9096118229095806e-05,
+      "loss": 0.7041,
+      "step": 6889
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5145744902532905,
+      "learning_rate": 1.909580119020138e-05,
+      "loss": 0.8681,
+      "step": 6890
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6558301953305825,
+      "learning_rate": 1.9095484098348314e-05,
+      "loss": 0.5253,
+      "step": 6891
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.297216679583051,
+      "learning_rate": 1.9095166953538467e-05,
+      "loss": 0.7344,
+      "step": 6892
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7163807188570888,
+      "learning_rate": 1.9094849755773675e-05,
+      "loss": 0.7948,
+      "step": 6893
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.478656074868626,
+      "learning_rate": 1.909453250505579e-05,
+      "loss": 0.7585,
+      "step": 6894
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.183654677785586,
+      "learning_rate": 1.9094215201386656e-05,
+      "loss": 0.8829,
+      "step": 6895
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.657361486934311,
+      "learning_rate": 1.9093897844768122e-05,
+      "loss": 0.7706,
+      "step": 6896
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5807341739439658,
+      "learning_rate": 1.9093580435202037e-05,
+      "loss": 0.6742,
+      "step": 6897
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5244109330906568,
+      "learning_rate": 1.9093262972690246e-05,
+      "loss": 0.7493,
+      "step": 6898
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.49812523991525,
+      "learning_rate": 1.90929454572346e-05,
+      "loss": 0.7246,
+      "step": 6899
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5825205615151241,
+      "learning_rate": 1.9092627888836945e-05,
+      "loss": 0.5749,
+      "step": 6900
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6519207248038323,
+      "learning_rate": 1.9092310267499135e-05,
+      "loss": 0.7458,
+      "step": 6901
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6061090672159484,
+      "learning_rate": 1.9091992593223015e-05,
+      "loss": 0.5613,
+      "step": 6902
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.686706047030425,
+      "learning_rate": 1.9091674866010434e-05,
+      "loss": 0.7748,
+      "step": 6903
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7526648676027332,
+      "learning_rate": 1.9091357085863246e-05,
+      "loss": 0.8186,
+      "step": 6904
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5907239030494601,
+      "learning_rate": 1.9091039252783296e-05,
+      "loss": 0.5551,
+      "step": 6905
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4332180172332878,
+      "learning_rate": 1.909072136677244e-05,
+      "loss": 0.7713,
+      "step": 6906
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4545614167686167,
+      "learning_rate": 1.909040342783252e-05,
+      "loss": 0.7199,
+      "step": 6907
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.532120417852243,
+      "learning_rate": 1.90900854359654e-05,
+      "loss": 0.7874,
+      "step": 6908
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5621573223472158,
+      "learning_rate": 1.9089767391172925e-05,
+      "loss": 0.7137,
+      "step": 6909
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5714876717301112,
+      "learning_rate": 1.9089449293456944e-05,
+      "loss": 0.6863,
+      "step": 6910
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5632199419176702,
+      "learning_rate": 1.908913114281931e-05,
+      "loss": 0.7542,
+      "step": 6911
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8606683386678342,
+      "learning_rate": 1.908881293926188e-05,
+      "loss": 0.9233,
+      "step": 6912
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5843704147104392,
+      "learning_rate": 1.9088494682786502e-05,
+      "loss": 0.5427,
+      "step": 6913
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7427970545124918,
+      "learning_rate": 1.908817637339503e-05,
+      "loss": 0.8053,
+      "step": 6914
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.939507682749278,
+      "learning_rate": 1.908785801108932e-05,
+      "loss": 0.769,
+      "step": 6915
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8057254193621426,
+      "learning_rate": 1.9087539595871222e-05,
+      "loss": 0.73,
+      "step": 6916
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7545544048900719,
+      "learning_rate": 1.908722112774259e-05,
+      "loss": 0.8851,
+      "step": 6917
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.631257439180293,
+      "learning_rate": 1.9086902606705286e-05,
+      "loss": 0.7203,
+      "step": 6918
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5876494080429193,
+      "learning_rate": 1.9086584032761148e-05,
+      "loss": 0.6869,
+      "step": 6919
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3707124265284163,
+      "learning_rate": 1.9086265405912047e-05,
+      "loss": 0.7941,
+      "step": 6920
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5542184068142204,
+      "learning_rate": 1.9085946726159833e-05,
+      "loss": 0.6733,
+      "step": 6921
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6025680301802817,
+      "learning_rate": 1.908562799350636e-05,
+      "loss": 0.5773,
+      "step": 6922
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.548250192093988,
+      "learning_rate": 1.9085309207953483e-05,
+      "loss": 0.7914,
+      "step": 6923
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2717316870575015,
+      "learning_rate": 1.908499036950306e-05,
+      "loss": 0.728,
+      "step": 6924
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.797692853903043,
+      "learning_rate": 1.9084671478156947e-05,
+      "loss": 0.6469,
+      "step": 6925
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6524909756323773,
+      "learning_rate": 1.9084352533917e-05,
+      "loss": 0.7468,
+      "step": 6926
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4565912676229822,
+      "learning_rate": 1.908403353678508e-05,
+      "loss": 0.76,
+      "step": 6927
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1780188383651846,
+      "learning_rate": 1.9083714486763035e-05,
+      "loss": 0.6912,
+      "step": 6928
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.814680070874519,
+      "learning_rate": 1.9083395383852733e-05,
+      "loss": 0.7999,
+      "step": 6929
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7522140624190798,
+      "learning_rate": 1.9083076228056023e-05,
+      "loss": 0.8172,
+      "step": 6930
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.474812187340147,
+      "learning_rate": 1.908275701937477e-05,
+      "loss": 0.8032,
+      "step": 6931
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3626970002057102,
+      "learning_rate": 1.9082437757810827e-05,
+      "loss": 0.8428,
+      "step": 6932
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.727366664579394,
+      "learning_rate": 1.908211844336606e-05,
+      "loss": 0.7446,
+      "step": 6933
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.520572123977426,
+      "learning_rate": 1.9081799076042323e-05,
+      "loss": 0.8239,
+      "step": 6934
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.6420186383724493,
+      "learning_rate": 1.9081479655841476e-05,
+      "loss": 0.6947,
+      "step": 6935
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9808272982932416,
+      "learning_rate": 1.908116018276538e-05,
+      "loss": 0.9089,
+      "step": 6936
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3007403404336768,
+      "learning_rate": 1.9080840656815895e-05,
+      "loss": 0.7692,
+      "step": 6937
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6070915982574954,
+      "learning_rate": 1.908052107799488e-05,
+      "loss": 0.8574,
+      "step": 6938
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3056610943235172,
+      "learning_rate": 1.9080201446304198e-05,
+      "loss": 0.7694,
+      "step": 6939
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5220623721826316,
+      "learning_rate": 1.9079881761745706e-05,
+      "loss": 0.7559,
+      "step": 6940
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.555091345544476,
+      "learning_rate": 1.907956202432127e-05,
+      "loss": 0.71,
+      "step": 6941
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5707688687064805,
+      "learning_rate": 1.9079242234032747e-05,
+      "loss": 0.876,
+      "step": 6942
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7001911636727265,
+      "learning_rate": 1.9078922390882e-05,
+      "loss": 0.7644,
+      "step": 6943
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5366241974688672,
+      "learning_rate": 1.9078602494870894e-05,
+      "loss": 0.7011,
+      "step": 6944
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7568310403039944,
+      "learning_rate": 1.907828254600129e-05,
+      "loss": 0.8528,
+      "step": 6945
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.560901636573681,
+      "learning_rate": 1.9077962544275053e-05,
+      "loss": 0.7703,
+      "step": 6946
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4903281615295105,
+      "learning_rate": 1.9077642489694043e-05,
+      "loss": 0.8427,
+      "step": 6947
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4800857671200336,
+      "learning_rate": 1.9077322382260123e-05,
+      "loss": 0.721,
+      "step": 6948
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7024827454608313,
+      "learning_rate": 1.907700222197516e-05,
+      "loss": 0.7623,
+      "step": 6949
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5248575779529772,
+      "learning_rate": 1.9076682008841015e-05,
+      "loss": 0.7076,
+      "step": 6950
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5178333033737452,
+      "learning_rate": 1.9076361742859556e-05,
+      "loss": 0.814,
+      "step": 6951
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6429221189888981,
+      "learning_rate": 1.9076041424032642e-05,
+      "loss": 0.6094,
+      "step": 6952
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5357618750078084,
+      "learning_rate": 1.9075721052362145e-05,
+      "loss": 0.7463,
+      "step": 6953
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6306057278817077,
+      "learning_rate": 1.9075400627849925e-05,
+      "loss": 0.5721,
+      "step": 6954
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4594225688437716,
+      "learning_rate": 1.9075080150497848e-05,
+      "loss": 0.7851,
+      "step": 6955
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.609149492948754,
+      "learning_rate": 1.907475962030778e-05,
+      "loss": 0.8555,
+      "step": 6956
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.368278303816525,
+      "learning_rate": 1.9074439037281594e-05,
+      "loss": 0.7031,
+      "step": 6957
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6684311630737998,
+      "learning_rate": 1.907411840142115e-05,
+      "loss": 0.5767,
+      "step": 6958
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6559694330935257,
+      "learning_rate": 1.907379771272831e-05,
+      "loss": 0.7666,
+      "step": 6959
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.546451504489334,
+      "learning_rate": 1.9073476971204952e-05,
+      "loss": 0.6798,
+      "step": 6960
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5212338289714653,
+      "learning_rate": 1.9073156176852936e-05,
+      "loss": 0.7842,
+      "step": 6961
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.166668649514952,
+      "learning_rate": 1.9072835329674134e-05,
+      "loss": 0.759,
+      "step": 6962
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8333209005113913,
+      "learning_rate": 1.907251442967041e-05,
+      "loss": 0.8648,
+      "step": 6963
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6114972147039341,
+      "learning_rate": 1.9072193476843636e-05,
+      "loss": 0.655,
+      "step": 6964
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.459048425806564,
+      "learning_rate": 1.907187247119568e-05,
+      "loss": 0.7482,
+      "step": 6965
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.108999390344365,
+      "learning_rate": 1.907155141272841e-05,
+      "loss": 0.7436,
+      "step": 6966
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.47084266558786,
+      "learning_rate": 1.9071230301443698e-05,
+      "loss": 0.7403,
+      "step": 6967
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6669408590446133,
+      "learning_rate": 1.907090913734341e-05,
+      "loss": 0.7036,
+      "step": 6968
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9081044931749447,
+      "learning_rate": 1.9070587920429414e-05,
+      "loss": 0.7802,
+      "step": 6969
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4527190236408556,
+      "learning_rate": 1.9070266650703585e-05,
+      "loss": 0.8156,
+      "step": 6970
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6899785968586565,
+      "learning_rate": 1.9069945328167792e-05,
+      "loss": 0.5413,
+      "step": 6971
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6891765636339573,
+      "learning_rate": 1.906962395282391e-05,
+      "loss": 0.6778,
+      "step": 6972
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8181934301615352,
+      "learning_rate": 1.90693025246738e-05,
+      "loss": 0.7372,
+      "step": 6973
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.334297281667503,
+      "learning_rate": 1.9068981043719343e-05,
+      "loss": 0.7536,
+      "step": 6974
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6202830174851077,
+      "learning_rate": 1.9068659509962407e-05,
+      "loss": 0.8479,
+      "step": 6975
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3358019862674024,
+      "learning_rate": 1.9068337923404863e-05,
+      "loss": 0.788,
+      "step": 6976
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.799628892729751,
+      "learning_rate": 1.9068016284048584e-05,
+      "loss": 0.7603,
+      "step": 6977
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.914190011124389,
+      "learning_rate": 1.9067694591895443e-05,
+      "loss": 0.7363,
+      "step": 6978
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5985203816750014,
+      "learning_rate": 1.9067372846947315e-05,
+      "loss": 0.5385,
+      "step": 6979
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4156641709176228,
+      "learning_rate": 1.906705104920607e-05,
+      "loss": 0.7002,
+      "step": 6980
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3813421955772682,
+      "learning_rate": 1.9066729198673588e-05,
+      "loss": 0.7191,
+      "step": 6981
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4567640857477806,
+      "learning_rate": 1.9066407295351732e-05,
+      "loss": 0.8172,
+      "step": 6982
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4369612947003825,
+      "learning_rate": 1.9066085339242387e-05,
+      "loss": 0.9069,
+      "step": 6983
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4484808501835693,
+      "learning_rate": 1.906576333034742e-05,
+      "loss": 0.9569,
+      "step": 6984
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4102614358195669,
+      "learning_rate": 1.906544126866871e-05,
+      "loss": 0.762,
+      "step": 6985
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2683022771309633,
+      "learning_rate": 1.906511915420813e-05,
+      "loss": 0.6627,
+      "step": 6986
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4106584656907886,
+      "learning_rate": 1.9064796986967557e-05,
+      "loss": 0.8177,
+      "step": 6987
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6328140452282953,
+      "learning_rate": 1.9064474766948864e-05,
+      "loss": 0.6959,
+      "step": 6988
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9194445667426878,
+      "learning_rate": 1.906415249415393e-05,
+      "loss": 0.7635,
+      "step": 6989
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.837315458218375,
+      "learning_rate": 1.9063830168584633e-05,
+      "loss": 0.7414,
+      "step": 6990
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8785819716259884,
+      "learning_rate": 1.9063507790242845e-05,
+      "loss": 0.794,
+      "step": 6991
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7163505241049326,
+      "learning_rate": 1.9063185359130446e-05,
+      "loss": 0.7297,
+      "step": 6992
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5905246673228888,
+      "learning_rate": 1.906286287524931e-05,
+      "loss": 0.5154,
+      "step": 6993
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7080481980288587,
+      "learning_rate": 1.9062540338601324e-05,
+      "loss": 0.8118,
+      "step": 6994
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3419014328460892,
+      "learning_rate": 1.906221774918835e-05,
+      "loss": 0.8013,
+      "step": 6995
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.6627766420860337,
+      "learning_rate": 1.906189510701228e-05,
+      "loss": 0.8392,
+      "step": 6996
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3581583426882589,
+      "learning_rate": 1.906157241207499e-05,
+      "loss": 0.7607,
+      "step": 6997
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6271744894159051,
+      "learning_rate": 1.9061249664378353e-05,
+      "loss": 0.5385,
+      "step": 6998
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5883595972143734,
+      "learning_rate": 1.906092686392425e-05,
+      "loss": 0.8124,
+      "step": 6999
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8831651341505313,
+      "learning_rate": 1.9060604010714566e-05,
+      "loss": 0.6834,
+      "step": 7000
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.34396142260384,
+      "learning_rate": 1.9060281104751178e-05,
+      "loss": 0.7342,
+      "step": 7001
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.541321652632008,
+      "learning_rate": 1.9059958146035963e-05,
+      "loss": 0.6809,
+      "step": 7002
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7261070724079377,
+      "learning_rate": 1.90596351345708e-05,
+      "loss": 0.8089,
+      "step": 7003
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5614249952759442,
+      "learning_rate": 1.9059312070357577e-05,
+      "loss": 0.7809,
+      "step": 7004
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2096774885956942,
+      "learning_rate": 1.9058988953398166e-05,
+      "loss": 0.7818,
+      "step": 7005
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6094689843859544,
+      "learning_rate": 1.9058665783694458e-05,
+      "loss": 0.7673,
+      "step": 7006
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9580358962904225,
+      "learning_rate": 1.9058342561248332e-05,
+      "loss": 0.7865,
+      "step": 7007
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5869136330191204,
+      "learning_rate": 1.9058019286061662e-05,
+      "loss": 0.5083,
+      "step": 7008
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6400559733188351,
+      "learning_rate": 1.905769595813634e-05,
+      "loss": 0.7254,
+      "step": 7009
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7196663026037897,
+      "learning_rate": 1.9057372577474244e-05,
+      "loss": 0.8274,
+      "step": 7010
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6150825300332846,
+      "learning_rate": 1.9057049144077256e-05,
+      "loss": 0.557,
+      "step": 7011
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5019517418363766,
+      "learning_rate": 1.905672565794726e-05,
+      "loss": 0.8247,
+      "step": 7012
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.611267315726801,
+      "learning_rate": 1.9056402119086142e-05,
+      "loss": 0.7103,
+      "step": 7013
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.562229408921637,
+      "learning_rate": 1.9056078527495784e-05,
+      "loss": 0.678,
+      "step": 7014
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3750857528101625,
+      "learning_rate": 1.9055754883178068e-05,
+      "loss": 0.8026,
+      "step": 7015
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.743980251113109,
+      "learning_rate": 1.905543118613488e-05,
+      "loss": 0.7284,
+      "step": 7016
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.933973494525889,
+      "learning_rate": 1.9055107436368104e-05,
+      "loss": 0.7107,
+      "step": 7017
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4785684367641103,
+      "learning_rate": 1.905478363387963e-05,
+      "loss": 0.8057,
+      "step": 7018
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7257844019728321,
+      "learning_rate": 1.9054459778671334e-05,
+      "loss": 0.6476,
+      "step": 7019
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4867345285303453,
+      "learning_rate": 1.9054135870745106e-05,
+      "loss": 0.7021,
+      "step": 7020
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.586644132700057,
+      "learning_rate": 1.9053811910102838e-05,
+      "loss": 0.5504,
+      "step": 7021
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5068095690513175,
+      "learning_rate": 1.9053487896746406e-05,
+      "loss": 0.7935,
+      "step": 7022
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6095278628006855,
+      "learning_rate": 1.9053163830677705e-05,
+      "loss": 0.5376,
+      "step": 7023
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.781003102500761,
+      "learning_rate": 1.9052839711898614e-05,
+      "loss": 0.782,
+      "step": 7024
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.633851378503386,
+      "learning_rate": 1.9052515540411027e-05,
+      "loss": 0.5668,
+      "step": 7025
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.5523423622960326,
+      "learning_rate": 1.905219131621683e-05,
+      "loss": 0.7068,
+      "step": 7026
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2369551257935947,
+      "learning_rate": 1.9051867039317903e-05,
+      "loss": 0.7375,
+      "step": 7027
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5455039929482166,
+      "learning_rate": 1.9051542709716147e-05,
+      "loss": 0.8205,
+      "step": 7028
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5161883619606233,
+      "learning_rate": 1.9051218327413438e-05,
+      "loss": 0.7866,
+      "step": 7029
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.617659952001404,
+      "learning_rate": 1.9050893892411675e-05,
+      "loss": 0.7764,
+      "step": 7030
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5598917339641882,
+      "learning_rate": 1.905056940471274e-05,
+      "loss": 0.8012,
+      "step": 7031
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5988530131565245,
+      "learning_rate": 1.9050244864318524e-05,
+      "loss": 0.4986,
+      "step": 7032
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5739864128177229,
+      "learning_rate": 1.904992027123092e-05,
+      "loss": 0.5198,
+      "step": 7033
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6728705052481743,
+      "learning_rate": 1.9049595625451813e-05,
+      "loss": 0.8932,
+      "step": 7034
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9805171514073652,
+      "learning_rate": 1.9049270926983095e-05,
+      "loss": 0.717,
+      "step": 7035
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4990773487861453,
+      "learning_rate": 1.9048946175826658e-05,
+      "loss": 0.8134,
+      "step": 7036
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6549136803475777,
+      "learning_rate": 1.904862137198439e-05,
+      "loss": 0.5936,
+      "step": 7037
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7240624228921206,
+      "learning_rate": 1.904829651545818e-05,
+      "loss": 0.8414,
+      "step": 7038
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.607945961318973,
+      "learning_rate": 1.904797160624993e-05,
+      "loss": 0.6446,
+      "step": 7039
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.443799752712099,
+      "learning_rate": 1.904764664436152e-05,
+      "loss": 0.8176,
+      "step": 7040
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4860608388896313,
+      "learning_rate": 1.904732162979485e-05,
+      "loss": 0.7565,
+      "step": 7041
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.700570580353879,
+      "learning_rate": 1.904699656255181e-05,
+      "loss": 0.7555,
+      "step": 7042
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3680696613549566,
+      "learning_rate": 1.9046671442634288e-05,
+      "loss": 0.8548,
+      "step": 7043
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.5557189380590293,
+      "learning_rate": 1.9046346270044183e-05,
+      "loss": 0.663,
+      "step": 7044
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4900960493423796,
+      "learning_rate": 1.904602104478339e-05,
+      "loss": 0.9466,
+      "step": 7045
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4842315609033183,
+      "learning_rate": 1.904569576685379e-05,
+      "loss": 0.7238,
+      "step": 7046
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.618600614542483,
+      "learning_rate": 1.904537043625729e-05,
+      "loss": 0.7432,
+      "step": 7047
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2961315902557502,
+      "learning_rate": 1.9045045052995776e-05,
+      "loss": 0.6913,
+      "step": 7048
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.928741413038653,
+      "learning_rate": 1.904471961707115e-05,
+      "loss": 0.6965,
+      "step": 7049
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4951405889919385,
+      "learning_rate": 1.9044394128485303e-05,
+      "loss": 0.7733,
+      "step": 7050
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7993911628967278,
+      "learning_rate": 1.9044068587240128e-05,
+      "loss": 0.8118,
+      "step": 7051
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4606421127575557,
+      "learning_rate": 1.9043742993337523e-05,
+      "loss": 0.7963,
+      "step": 7052
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4512456860348741,
+      "learning_rate": 1.9043417346779386e-05,
+      "loss": 0.6183,
+      "step": 7053
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8065494209784891,
+      "learning_rate": 1.9043091647567608e-05,
+      "loss": 0.6189,
+      "step": 7054
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5341654466853467,
+      "learning_rate": 1.9042765895704085e-05,
+      "loss": 0.7457,
+      "step": 7055
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6397172666134063,
+      "learning_rate": 1.904244009119072e-05,
+      "loss": 0.8891,
+      "step": 7056
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4601319910656383,
+      "learning_rate": 1.90421142340294e-05,
+      "loss": 0.8171,
+      "step": 7057
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6744203838558214,
+      "learning_rate": 1.9041788324222035e-05,
+      "loss": 0.571,
+      "step": 7058
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9876844187660714,
+      "learning_rate": 1.9041462361770513e-05,
+      "loss": 0.6922,
+      "step": 7059
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6068003495858179,
+      "learning_rate": 1.9041136346676733e-05,
+      "loss": 0.7285,
+      "step": 7060
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.566579366444339,
+      "learning_rate": 1.9040810278942596e-05,
+      "loss": 0.7936,
+      "step": 7061
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7221013768306046,
+      "learning_rate": 1.904048415857e-05,
+      "loss": 0.8467,
+      "step": 7062
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.044142444530351,
+      "learning_rate": 1.9040157985560842e-05,
+      "loss": 0.6332,
+      "step": 7063
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8939784619883253,
+      "learning_rate": 1.9039831759917023e-05,
+      "loss": 0.7542,
+      "step": 7064
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3907077808638804,
+      "learning_rate": 1.9039505481640447e-05,
+      "loss": 0.7987,
+      "step": 7065
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7145230712544843,
+      "learning_rate": 1.9039179150733e-05,
+      "loss": 0.7484,
+      "step": 7066
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5785616359593614,
+      "learning_rate": 1.9038852767196592e-05,
+      "loss": 0.5393,
+      "step": 7067
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1850776027606478,
+      "learning_rate": 1.9038526331033122e-05,
+      "loss": 0.7734,
+      "step": 7068
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.742132621182649,
+      "learning_rate": 1.903819984224449e-05,
+      "loss": 0.7892,
+      "step": 7069
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.617611306306335,
+      "learning_rate": 1.90378733008326e-05,
+      "loss": 0.7367,
+      "step": 7070
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.461074216348779,
+      "learning_rate": 1.903754670679935e-05,
+      "loss": 0.6516,
+      "step": 7071
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.595104492779696,
+      "learning_rate": 1.9037220060146637e-05,
+      "loss": 0.8247,
+      "step": 7072
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5108338184631367,
+      "learning_rate": 1.9036893360876373e-05,
+      "loss": 0.8679,
+      "step": 7073
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4462293321704633,
+      "learning_rate": 1.9036566608990452e-05,
+      "loss": 0.8078,
+      "step": 7074
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5230482022988574,
+      "learning_rate": 1.9036239804490778e-05,
+      "loss": 0.7507,
+      "step": 7075
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.543367336362515,
+      "learning_rate": 1.9035912947379258e-05,
+      "loss": 0.7395,
+      "step": 7076
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7615783110766459,
+      "learning_rate": 1.903558603765779e-05,
+      "loss": 0.7805,
+      "step": 7077
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5763523543911975,
+      "learning_rate": 1.9035259075328278e-05,
+      "loss": 0.7176,
+      "step": 7078
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.479713710325056,
+      "learning_rate": 1.9034932060392633e-05,
+      "loss": 0.7616,
+      "step": 7079
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0453613997809117,
+      "learning_rate": 1.903460499285275e-05,
+      "loss": 0.6845,
+      "step": 7080
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5818162112797942,
+      "learning_rate": 1.9034277872710533e-05,
+      "loss": 0.8969,
+      "step": 7081
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6022269068034433,
+      "learning_rate": 1.9033950699967896e-05,
+      "loss": 0.5413,
+      "step": 7082
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8665499921205588,
+      "learning_rate": 1.9033623474626736e-05,
+      "loss": 0.679,
+      "step": 7083
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6814191952962472,
+      "learning_rate": 1.9033296196688957e-05,
+      "loss": 0.6969,
+      "step": 7084
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.015226187415716,
+      "learning_rate": 1.903296886615647e-05,
+      "loss": 0.7486,
+      "step": 7085
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4078196505124243,
+      "learning_rate": 1.9032641483031177e-05,
+      "loss": 0.7418,
+      "step": 7086
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.815145959547508,
+      "learning_rate": 1.9032314047314985e-05,
+      "loss": 0.6884,
+      "step": 7087
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4731228550721907,
+      "learning_rate": 1.9031986559009804e-05,
+      "loss": 0.7357,
+      "step": 7088
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9940045699693771,
+      "learning_rate": 1.9031659018117537e-05,
+      "loss": 0.796,
+      "step": 7089
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4856433047278927,
+      "learning_rate": 1.903133142464009e-05,
+      "loss": 0.6787,
+      "step": 7090
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6093834938564735,
+      "learning_rate": 1.9031003778579376e-05,
+      "loss": 0.8703,
+      "step": 7091
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.455477281895785,
+      "learning_rate": 1.9030676079937296e-05,
+      "loss": 0.9463,
+      "step": 7092
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3733390203668756,
+      "learning_rate": 1.903034832871576e-05,
+      "loss": 0.8063,
+      "step": 7093
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3749225483342378,
+      "learning_rate": 1.903002052491668e-05,
+      "loss": 0.8508,
+      "step": 7094
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8418742860196349,
+      "learning_rate": 1.902969266854196e-05,
+      "loss": 0.8239,
+      "step": 7095
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2827603201811355,
+      "learning_rate": 1.902936475959351e-05,
+      "loss": 0.7026,
+      "step": 7096
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.174673094141085,
+      "learning_rate": 1.902903679807324e-05,
+      "loss": 0.7449,
+      "step": 7097
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9044227367158233,
+      "learning_rate": 1.902870878398306e-05,
+      "loss": 0.7212,
+      "step": 7098
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5111344801148576,
+      "learning_rate": 1.9028380717324875e-05,
+      "loss": 0.7688,
+      "step": 7099
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4549107549823588,
+      "learning_rate": 1.9028052598100607e-05,
+      "loss": 0.8167,
+      "step": 7100
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6042559797247307,
+      "learning_rate": 1.902772442631215e-05,
+      "loss": 0.7516,
+      "step": 7101
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8941105198602888,
+      "learning_rate": 1.902739620196143e-05,
+      "loss": 0.7462,
+      "step": 7102
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6653192326113828,
+      "learning_rate": 1.902706792505035e-05,
+      "loss": 0.7171,
+      "step": 7103
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7011988300974261,
+      "learning_rate": 1.902673959558082e-05,
+      "loss": 0.8141,
+      "step": 7104
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6315207152046438,
+      "learning_rate": 1.9026411213554758e-05,
+      "loss": 0.682,
+      "step": 7105
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6651538859538195,
+      "learning_rate": 1.9026082778974065e-05,
+      "loss": 0.6713,
+      "step": 7106
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4485391799445022,
+      "learning_rate": 1.902575429184067e-05,
+      "loss": 0.8534,
+      "step": 7107
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5543863730322263,
+      "learning_rate": 1.902542575215647e-05,
+      "loss": 0.6445,
+      "step": 7108
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5904370923124207,
+      "learning_rate": 1.902509715992338e-05,
+      "loss": 0.7018,
+      "step": 7109
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4327913069113172,
+      "learning_rate": 1.9024768515143325e-05,
+      "loss": 0.8386,
+      "step": 7110
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.021153937476672,
+      "learning_rate": 1.902443981781821e-05,
+      "loss": 0.796,
+      "step": 7111
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6082624941160562,
+      "learning_rate": 1.9024111067949945e-05,
+      "loss": 0.5186,
+      "step": 7112
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5866623362739052,
+      "learning_rate": 1.902378226554045e-05,
+      "loss": 0.7185,
+      "step": 7113
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6767865190158608,
+      "learning_rate": 1.902345341059164e-05,
+      "loss": 0.8563,
+      "step": 7114
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3746835562880133,
+      "learning_rate": 1.9023124503105424e-05,
+      "loss": 0.7273,
+      "step": 7115
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7643053659871373,
+      "learning_rate": 1.9022795543083722e-05,
+      "loss": 0.8454,
+      "step": 7116
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6343443537128237,
+      "learning_rate": 1.9022466530528448e-05,
+      "loss": 0.7502,
+      "step": 7117
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5117299971582578,
+      "learning_rate": 1.9022137465441518e-05,
+      "loss": 0.6937,
+      "step": 7118
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6636982415693944,
+      "learning_rate": 1.9021808347824847e-05,
+      "loss": 0.7985,
+      "step": 7119
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.7409168145879006,
+      "learning_rate": 1.902147917768035e-05,
+      "loss": 0.6317,
+      "step": 7120
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6475464687134564,
+      "learning_rate": 1.9021149955009943e-05,
+      "loss": 0.7523,
+      "step": 7121
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.650615975446915,
+      "learning_rate": 1.902082067981555e-05,
+      "loss": 0.7678,
+      "step": 7122
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6789401154810346,
+      "learning_rate": 1.902049135209908e-05,
+      "loss": 0.7955,
+      "step": 7123
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7167586468743912,
+      "learning_rate": 1.9020161971862454e-05,
+      "loss": 0.7435,
+      "step": 7124
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6616919574499422,
+      "learning_rate": 1.9019832539107592e-05,
+      "loss": 0.7725,
+      "step": 7125
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7743545139429602,
+      "learning_rate": 1.9019503053836406e-05,
+      "loss": 0.7459,
+      "step": 7126
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4610610534808803,
+      "learning_rate": 1.901917351605082e-05,
+      "loss": 0.777,
+      "step": 7127
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6705473229374588,
+      "learning_rate": 1.901884392575275e-05,
+      "loss": 0.7174,
+      "step": 7128
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4657390678072046,
+      "learning_rate": 1.9018514282944115e-05,
+      "loss": 0.6713,
+      "step": 7129
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6716257411675914,
+      "learning_rate": 1.901818458762683e-05,
+      "loss": 0.7944,
+      "step": 7130
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6546662309226947,
+      "learning_rate": 1.9017854839802824e-05,
+      "loss": 0.7204,
+      "step": 7131
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8942049100055194,
+      "learning_rate": 1.9017525039474013e-05,
+      "loss": 0.6732,
+      "step": 7132
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6522034006066915,
+      "learning_rate": 1.9017195186642315e-05,
+      "loss": 0.5324,
+      "step": 7133
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7937611893732726,
+      "learning_rate": 1.9016865281309652e-05,
+      "loss": 0.6975,
+      "step": 7134
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7307512113797392,
+      "learning_rate": 1.9016535323477945e-05,
+      "loss": 0.7174,
+      "step": 7135
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.760192794828573,
+      "learning_rate": 1.9016205313149114e-05,
+      "loss": 0.7451,
+      "step": 7136
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8773006407055333,
+      "learning_rate": 1.901587525032508e-05,
+      "loss": 0.794,
+      "step": 7137
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1161748605804234,
+      "learning_rate": 1.9015545135007766e-05,
+      "loss": 0.7347,
+      "step": 7138
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3959747769419266,
+      "learning_rate": 1.9015214967199095e-05,
+      "loss": 0.8454,
+      "step": 7139
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.518909675372661,
+      "learning_rate": 1.901488474690099e-05,
+      "loss": 0.7125,
+      "step": 7140
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5895266698556961,
+      "learning_rate": 1.901455447411537e-05,
+      "loss": 0.5296,
+      "step": 7141
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.427777181928996,
+      "learning_rate": 1.901422414884416e-05,
+      "loss": 0.8056,
+      "step": 7142
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5018335441374004,
+      "learning_rate": 1.9013893771089283e-05,
+      "loss": 0.7739,
+      "step": 7143
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.414627274213053,
+      "learning_rate": 1.9013563340852664e-05,
+      "loss": 0.7892,
+      "step": 7144
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.852007433512306,
+      "learning_rate": 1.9013232858136225e-05,
+      "loss": 0.8402,
+      "step": 7145
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5351960899176902,
+      "learning_rate": 1.901290232294189e-05,
+      "loss": 0.7976,
+      "step": 7146
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4307074462221132,
+      "learning_rate": 1.9012571735271584e-05,
+      "loss": 0.6238,
+      "step": 7147
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7447979600198902,
+      "learning_rate": 1.9012241095127234e-05,
+      "loss": 0.8377,
+      "step": 7148
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5330766570173742,
+      "learning_rate": 1.9011910402510762e-05,
+      "loss": 0.8876,
+      "step": 7149
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6502832554844502,
+      "learning_rate": 1.9011579657424092e-05,
+      "loss": 0.8526,
+      "step": 7150
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6549531443711069,
+      "learning_rate": 1.9011248859869155e-05,
+      "loss": 0.5261,
+      "step": 7151
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5121969234655621,
+      "learning_rate": 1.9010918009847877e-05,
+      "loss": 0.8898,
+      "step": 7152
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.800039658218479,
+      "learning_rate": 1.9010587107362177e-05,
+      "loss": 0.6811,
+      "step": 7153
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9945832255807379,
+      "learning_rate": 1.9010256152413987e-05,
+      "loss": 0.7807,
+      "step": 7154
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8315943083744377,
+      "learning_rate": 1.900992514500523e-05,
+      "loss": 0.642,
+      "step": 7155
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5542235878358168,
+      "learning_rate": 1.900959408513784e-05,
+      "loss": 0.5099,
+      "step": 7156
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.467280082408238,
+      "learning_rate": 1.9009262972813746e-05,
+      "loss": 0.8161,
+      "step": 7157
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6264964956735577,
+      "learning_rate": 1.9008931808034864e-05,
+      "loss": 0.725,
+      "step": 7158
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0134276640448427,
+      "learning_rate": 1.900860059080313e-05,
+      "loss": 0.7999,
+      "step": 7159
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.363674359595841,
+      "learning_rate": 1.900826932112047e-05,
+      "loss": 0.6195,
+      "step": 7160
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5692389977418815,
+      "learning_rate": 1.9007937998988818e-05,
+      "loss": 0.7742,
+      "step": 7161
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.44018522918431,
+      "learning_rate": 1.9007606624410094e-05,
+      "loss": 0.7423,
+      "step": 7162
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5003521721958015,
+      "learning_rate": 1.9007275197386238e-05,
+      "loss": 0.7938,
+      "step": 7163
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4251487181283289,
+      "learning_rate": 1.900694371791917e-05,
+      "loss": 0.9867,
+      "step": 7164
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.750752195327869,
+      "learning_rate": 1.9006612186010825e-05,
+      "loss": 0.9095,
+      "step": 7165
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.472065876845665,
+      "learning_rate": 1.9006280601663132e-05,
+      "loss": 0.6639,
+      "step": 7166
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4443453756772648,
+      "learning_rate": 1.9005948964878024e-05,
+      "loss": 0.813,
+      "step": 7167
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4686451450191353,
+      "learning_rate": 1.900561727565743e-05,
+      "loss": 0.7313,
+      "step": 7168
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5899188269042295,
+      "learning_rate": 1.900528553400328e-05,
+      "loss": 0.7757,
+      "step": 7169
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6121274090996915,
+      "learning_rate": 1.9004953739917503e-05,
+      "loss": 0.6962,
+      "step": 7170
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.623525972352843,
+      "learning_rate": 1.900462189340204e-05,
+      "loss": 0.5464,
+      "step": 7171
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8933992903544865,
+      "learning_rate": 1.9004289994458815e-05,
+      "loss": 0.8309,
+      "step": 7172
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0848864944086287,
+      "learning_rate": 1.9003958043089763e-05,
+      "loss": 0.8227,
+      "step": 7173
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9754258862477145,
+      "learning_rate": 1.900362603929682e-05,
+      "loss": 0.6698,
+      "step": 7174
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5991856004356908,
+      "learning_rate": 1.900329398308191e-05,
+      "loss": 0.787,
+      "step": 7175
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.594351645974833,
+      "learning_rate": 1.9002961874446973e-05,
+      "loss": 0.8438,
+      "step": 7176
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3265254291349402,
+      "learning_rate": 1.9002629713393946e-05,
+      "loss": 0.681,
+      "step": 7177
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.56020275669463,
+      "learning_rate": 1.9002297499924752e-05,
+      "loss": 0.7914,
+      "step": 7178
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.424347987881678,
+      "learning_rate": 1.900196523404134e-05,
+      "loss": 0.7403,
+      "step": 7179
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3838897408098667,
+      "learning_rate": 1.900163291574563e-05,
+      "loss": 0.7315,
+      "step": 7180
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.560562633840207,
+      "learning_rate": 1.9001300545039566e-05,
+      "loss": 0.7825,
+      "step": 7181
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.504055724354095,
+      "learning_rate": 1.900096812192508e-05,
+      "loss": 0.7932,
+      "step": 7182
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6941696178718448,
+      "learning_rate": 1.9000635646404108e-05,
+      "loss": 0.832,
+      "step": 7183
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5862020757306918,
+      "learning_rate": 1.9000303118478584e-05,
+      "loss": 0.7262,
+      "step": 7184
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.659342596805009,
+      "learning_rate": 1.899997053815045e-05,
+      "loss": 0.7755,
+      "step": 7185
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3557255437207416,
+      "learning_rate": 1.8999637905421633e-05,
+      "loss": 0.7801,
+      "step": 7186
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4569367342521842,
+      "learning_rate": 1.899930522029408e-05,
+      "loss": 0.8433,
+      "step": 7187
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6188038702143572,
+      "learning_rate": 1.899897248276972e-05,
+      "loss": 0.5164,
+      "step": 7188
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.536090182499807,
+      "learning_rate": 1.8998639692850492e-05,
+      "loss": 0.8644,
+      "step": 7189
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5762725817330783,
+      "learning_rate": 1.899830685053834e-05,
+      "loss": 0.8191,
+      "step": 7190
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9469426439356,
+      "learning_rate": 1.8997973955835193e-05,
+      "loss": 0.8363,
+      "step": 7191
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6500605721888701,
+      "learning_rate": 1.899764100874299e-05,
+      "loss": 0.806,
+      "step": 7192
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7858941207171617,
+      "learning_rate": 1.8997308009263675e-05,
+      "loss": 0.7904,
+      "step": 7193
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4957965012968886,
+      "learning_rate": 1.8996974957399186e-05,
+      "loss": 0.7184,
+      "step": 7194
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5599345493046692,
+      "learning_rate": 1.8996641853151462e-05,
+      "loss": 0.7753,
+      "step": 7195
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6759507422153246,
+      "learning_rate": 1.8996308696522435e-05,
+      "loss": 0.7926,
+      "step": 7196
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7509777804222595,
+      "learning_rate": 1.8995975487514052e-05,
+      "loss": 0.896,
+      "step": 7197
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5888783170439817,
+      "learning_rate": 1.8995642226128256e-05,
+      "loss": 0.6916,
+      "step": 7198
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6869441953507578,
+      "learning_rate": 1.899530891236698e-05,
+      "loss": 0.7734,
+      "step": 7199
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3590781591533891,
+      "learning_rate": 1.8994975546232166e-05,
+      "loss": 0.6499,
+      "step": 7200
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5449222370171696,
+      "learning_rate": 1.899464212772576e-05,
+      "loss": 0.6671,
+      "step": 7201
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.632049963478865,
+      "learning_rate": 1.8994308656849697e-05,
+      "loss": 0.5444,
+      "step": 7202
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.741385773398484,
+      "learning_rate": 1.8993975133605925e-05,
+      "loss": 0.6513,
+      "step": 7203
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.005290618136668,
+      "learning_rate": 1.8993641557996377e-05,
+      "loss": 0.9245,
+      "step": 7204
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.008881795733233,
+      "learning_rate": 1.8993307930023005e-05,
+      "loss": 0.8023,
+      "step": 7205
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.344557787531115,
+      "learning_rate": 1.8992974249687747e-05,
+      "loss": 0.8722,
+      "step": 7206
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4541594813294458,
+      "learning_rate": 1.8992640516992544e-05,
+      "loss": 0.7327,
+      "step": 7207
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5373832080778347,
+      "learning_rate": 1.899230673193934e-05,
+      "loss": 0.817,
+      "step": 7208
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7115534977679996,
+      "learning_rate": 1.899197289453008e-05,
+      "loss": 0.6486,
+      "step": 7209
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6227342675857432,
+      "learning_rate": 1.8991639004766708e-05,
+      "loss": 0.5536,
+      "step": 7210
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2622898206085449,
+      "learning_rate": 1.8991305062651167e-05,
+      "loss": 0.6648,
+      "step": 7211
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.572747791774219,
+      "learning_rate": 1.8990971068185402e-05,
+      "loss": 0.5446,
+      "step": 7212
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5221752907892043,
+      "learning_rate": 1.8990637021371355e-05,
+      "loss": 0.781,
+      "step": 7213
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3722232963674275,
+      "learning_rate": 1.8990302922210976e-05,
+      "loss": 0.8563,
+      "step": 7214
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4598904937953867,
+      "learning_rate": 1.89899687707062e-05,
+      "loss": 0.6788,
+      "step": 7215
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7505570301667221,
+      "learning_rate": 1.8989634566858987e-05,
+      "loss": 0.771,
+      "step": 7216
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8506109806073612,
+      "learning_rate": 1.8989300310671273e-05,
+      "loss": 0.8323,
+      "step": 7217
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8534354441684593,
+      "learning_rate": 1.8988966002145007e-05,
+      "loss": 0.7354,
+      "step": 7218
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.052050297134738,
+      "learning_rate": 1.8988631641282134e-05,
+      "loss": 0.8081,
+      "step": 7219
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6185071275733252,
+      "learning_rate": 1.89882972280846e-05,
+      "loss": 0.6829,
+      "step": 7220
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.4004385771306733,
+      "learning_rate": 1.8987962762554358e-05,
+      "loss": 0.952,
+      "step": 7221
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5808186457124576,
+      "learning_rate": 1.898762824469335e-05,
+      "loss": 0.843,
+      "step": 7222
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4446629830971496,
+      "learning_rate": 1.8987293674503525e-05,
+      "loss": 0.8579,
+      "step": 7223
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6063405439198081,
+      "learning_rate": 1.898695905198683e-05,
+      "loss": 0.5506,
+      "step": 7224
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3870118614654103,
+      "learning_rate": 1.8986624377145216e-05,
+      "loss": 0.6321,
+      "step": 7225
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.033169769307287,
+      "learning_rate": 1.8986289649980627e-05,
+      "loss": 0.7694,
+      "step": 7226
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6142023335769888,
+      "learning_rate": 1.8985954870495016e-05,
+      "loss": 0.7058,
+      "step": 7227
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6616603081691783,
+      "learning_rate": 1.8985620038690332e-05,
+      "loss": 0.762,
+      "step": 7228
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5319850439831912,
+      "learning_rate": 1.8985285154568522e-05,
+      "loss": 0.7392,
+      "step": 7229
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5802941519551716,
+      "learning_rate": 1.8984950218131538e-05,
+      "loss": 0.7615,
+      "step": 7230
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3583322700294675,
+      "learning_rate": 1.8984615229381325e-05,
+      "loss": 0.7195,
+      "step": 7231
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6875780462089685,
+      "learning_rate": 1.8984280188319845e-05,
+      "loss": 0.7296,
+      "step": 7232
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.197355012313939,
+      "learning_rate": 1.8983945094949035e-05,
+      "loss": 0.797,
+      "step": 7233
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6732244918115338,
+      "learning_rate": 1.8983609949270854e-05,
+      "loss": 0.8149,
+      "step": 7234
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5945783197068846,
+      "learning_rate": 1.8983274751287254e-05,
+      "loss": 0.7816,
+      "step": 7235
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7519436241716122,
+      "learning_rate": 1.8982939501000182e-05,
+      "loss": 0.6789,
+      "step": 7236
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.2367754715684907,
+      "learning_rate": 1.8982604198411594e-05,
+      "loss": 0.7709,
+      "step": 7237
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.4610126471588494,
+      "learning_rate": 1.898226884352344e-05,
+      "loss": 0.7825,
+      "step": 7238
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.345038110361363,
+      "learning_rate": 1.8981933436337675e-05,
+      "loss": 0.5834,
+      "step": 7239
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7867954575465692,
+      "learning_rate": 1.8981597976856247e-05,
+      "loss": 0.7132,
+      "step": 7240
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5675365577906871,
+      "learning_rate": 1.8981262465081112e-05,
+      "loss": 0.5529,
+      "step": 7241
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.5202023534988727,
+      "learning_rate": 1.8980926901014226e-05,
+      "loss": 0.713,
+      "step": 7242
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7429290498277952,
+      "learning_rate": 1.8980591284657536e-05,
+      "loss": 0.8371,
+      "step": 7243
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.6059193696093876,
+      "learning_rate": 1.8980255616013004e-05,
+      "loss": 0.8157,
+      "step": 7244
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8308177440561364,
+      "learning_rate": 1.8979919895082578e-05,
+      "loss": 0.7395,
+      "step": 7245
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.373071879114867,
+      "learning_rate": 1.8979584121868217e-05,
+      "loss": 0.7142,
+      "step": 7246
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.425127854861233,
+      "learning_rate": 1.8979248296371876e-05,
+      "loss": 0.7656,
+      "step": 7247
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7537289241787406,
+      "learning_rate": 1.897891241859551e-05,
+      "loss": 0.6757,
+      "step": 7248
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7338208563221058,
+      "learning_rate": 1.8978576488541068e-05,
+      "loss": 0.748,
+      "step": 7249
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7767660093607207,
+      "learning_rate": 1.897824050621051e-05,
+      "loss": 0.808,
+      "step": 7250
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0448069321679703,
+      "learning_rate": 1.8977904471605798e-05,
+      "loss": 0.7576,
+      "step": 7251
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7856214180425272,
+      "learning_rate": 1.8977568384728885e-05,
+      "loss": 0.8306,
+      "step": 7252
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8205492327605146,
+      "learning_rate": 1.8977232245581722e-05,
+      "loss": 0.7387,
+      "step": 7253
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.70457890832,
+      "learning_rate": 1.8976896054166273e-05,
+      "loss": 0.847,
+      "step": 7254
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.965690439859979,
+      "learning_rate": 1.8976559810484494e-05,
+      "loss": 0.8682,
+      "step": 7255
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.1470127763291953,
+      "learning_rate": 1.8976223514538342e-05,
+      "loss": 0.62,
+      "step": 7256
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7952298459011276,
+      "learning_rate": 1.8975887166329773e-05,
+      "loss": 0.6861,
+      "step": 7257
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5037047302727202,
+      "learning_rate": 1.8975550765860745e-05,
+      "loss": 0.695,
+      "step": 7258
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4985009699161023,
+      "learning_rate": 1.8975214313133223e-05,
+      "loss": 0.8544,
+      "step": 7259
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7046191571937206,
+      "learning_rate": 1.8974877808149162e-05,
+      "loss": 0.7248,
+      "step": 7260
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9126398773968012,
+      "learning_rate": 1.897454125091052e-05,
+      "loss": 0.838,
+      "step": 7261
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.654444874902132,
+      "learning_rate": 1.8974204641419256e-05,
+      "loss": 0.915,
+      "step": 7262
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5396841193840634,
+      "learning_rate": 1.8973867979677332e-05,
+      "loss": 0.6908,
+      "step": 7263
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5589697412024701,
+      "learning_rate": 1.8973531265686706e-05,
+      "loss": 0.6503,
+      "step": 7264
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3516019178415848,
+      "learning_rate": 1.8973194499449342e-05,
+      "loss": 0.6778,
+      "step": 7265
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6429209242281514,
+      "learning_rate": 1.89728576809672e-05,
+      "loss": 0.5647,
+      "step": 7266
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8434981567590432,
+      "learning_rate": 1.897252081024224e-05,
+      "loss": 0.7374,
+      "step": 7267
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7772778736193193,
+      "learning_rate": 1.8972183887276422e-05,
+      "loss": 0.7254,
+      "step": 7268
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.67775093639667,
+      "learning_rate": 1.8971846912071707e-05,
+      "loss": 0.8869,
+      "step": 7269
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1437015066477203,
+      "learning_rate": 1.897150988463006e-05,
+      "loss": 0.6889,
+      "step": 7270
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6171205796425292,
+      "learning_rate": 1.8971172804953444e-05,
+      "loss": 0.5762,
+      "step": 7271
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4212012928629831,
+      "learning_rate": 1.897083567304382e-05,
+      "loss": 0.7799,
+      "step": 7272
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3585019892370784,
+      "learning_rate": 1.8970498488903147e-05,
+      "loss": 0.7561,
+      "step": 7273
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.903974299443821,
+      "learning_rate": 1.8970161252533393e-05,
+      "loss": 0.8112,
+      "step": 7274
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7747189775352523,
+      "learning_rate": 1.8969823963936524e-05,
+      "loss": 0.7202,
+      "step": 7275
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5271309366355739,
+      "learning_rate": 1.8969486623114496e-05,
+      "loss": 0.7011,
+      "step": 7276
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0073871238264487,
+      "learning_rate": 1.8969149230069278e-05,
+      "loss": 0.6527,
+      "step": 7277
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5957048512630247,
+      "learning_rate": 1.8968811784802833e-05,
+      "loss": 0.8032,
+      "step": 7278
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6130614226002693,
+      "learning_rate": 1.896847428731713e-05,
+      "loss": 0.7132,
+      "step": 7279
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3955733761316664,
+      "learning_rate": 1.8968136737614123e-05,
+      "loss": 0.6553,
+      "step": 7280
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5087339954257408,
+      "learning_rate": 1.8967799135695788e-05,
+      "loss": 0.6761,
+      "step": 7281
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5862795961230565,
+      "learning_rate": 1.896746148156409e-05,
+      "loss": 0.8766,
+      "step": 7282
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9821690663915095,
+      "learning_rate": 1.896712377522099e-05,
+      "loss": 0.8798,
+      "step": 7283
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.761308875212907,
+      "learning_rate": 1.8966786016668458e-05,
+      "loss": 0.8661,
+      "step": 7284
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8972803040586985,
+      "learning_rate": 1.8966448205908458e-05,
+      "loss": 0.7296,
+      "step": 7285
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7193910326283681,
+      "learning_rate": 1.8966110342942954e-05,
+      "loss": 0.7567,
+      "step": 7286
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7068897623846777,
+      "learning_rate": 1.8965772427773923e-05,
+      "loss": 0.6905,
+      "step": 7287
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.64870186427838,
+      "learning_rate": 1.8965434460403325e-05,
+      "loss": 0.786,
+      "step": 7288
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.646398080879561,
+      "learning_rate": 1.896509644083313e-05,
+      "loss": 0.8421,
+      "step": 7289
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.5119039941847436,
+      "learning_rate": 1.8964758369065303e-05,
+      "loss": 0.8857,
+      "step": 7290
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7681017883390304,
+      "learning_rate": 1.8964420245101816e-05,
+      "loss": 0.702,
+      "step": 7291
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.895416432154507,
+      "learning_rate": 1.8964082068944634e-05,
+      "loss": 0.8542,
+      "step": 7292
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7745092428406724,
+      "learning_rate": 1.896374384059573e-05,
+      "loss": 0.7824,
+      "step": 7293
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6322154408816231,
+      "learning_rate": 1.896340556005707e-05,
+      "loss": 0.5607,
+      "step": 7294
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7006413607164987,
+      "learning_rate": 1.8963067227330624e-05,
+      "loss": 0.8209,
+      "step": 7295
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4779698771883227,
+      "learning_rate": 1.8962728842418366e-05,
+      "loss": 0.75,
+      "step": 7296
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8076409471608443,
+      "learning_rate": 1.8962390405322263e-05,
+      "loss": 0.7226,
+      "step": 7297
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6482714874435076,
+      "learning_rate": 1.8962051916044288e-05,
+      "loss": 0.8589,
+      "step": 7298
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9081363080452687,
+      "learning_rate": 1.8961713374586403e-05,
+      "loss": 0.8271,
+      "step": 7299
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.153615555626147,
+      "learning_rate": 1.8961374780950588e-05,
+      "loss": 0.7199,
+      "step": 7300
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.917759315271075,
+      "learning_rate": 1.8961036135138813e-05,
+      "loss": 0.727,
+      "step": 7301
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7210652079124493,
+      "learning_rate": 1.8960697437153048e-05,
+      "loss": 0.7952,
+      "step": 7302
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7751459224304194,
+      "learning_rate": 1.8960358686995265e-05,
+      "loss": 0.7752,
+      "step": 7303
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4928903965124125,
+      "learning_rate": 1.896001988466744e-05,
+      "loss": 0.8043,
+      "step": 7304
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7831006909372535,
+      "learning_rate": 1.895968103017154e-05,
+      "loss": 0.7261,
+      "step": 7305
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.106258477524096,
+      "learning_rate": 1.8959342123509538e-05,
+      "loss": 0.7386,
+      "step": 7306
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.605800229289948,
+      "learning_rate": 1.8959003164683416e-05,
+      "loss": 0.71,
+      "step": 7307
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.852132858532304,
+      "learning_rate": 1.8958664153695138e-05,
+      "loss": 0.8591,
+      "step": 7308
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3224266994875307,
+      "learning_rate": 1.895832509054668e-05,
+      "loss": 0.74,
+      "step": 7309
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9773142457992299,
+      "learning_rate": 1.8957985975240016e-05,
+      "loss": 0.69,
+      "step": 7310
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.773286565967987,
+      "learning_rate": 1.8957646807777123e-05,
+      "loss": 0.8361,
+      "step": 7311
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6257181600562949,
+      "learning_rate": 1.8957307588159978e-05,
+      "loss": 0.8575,
+      "step": 7312
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.701847019224927,
+      "learning_rate": 1.8956968316390547e-05,
+      "loss": 0.9175,
+      "step": 7313
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.035316362684948,
+      "learning_rate": 1.8956628992470812e-05,
+      "loss": 0.8414,
+      "step": 7314
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6162373477190802,
+      "learning_rate": 1.895628961640275e-05,
+      "loss": 0.5308,
+      "step": 7315
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0346854628597457,
+      "learning_rate": 1.895595018818833e-05,
+      "loss": 0.8075,
+      "step": 7316
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9986000908367836,
+      "learning_rate": 1.8955610707829536e-05,
+      "loss": 0.6109,
+      "step": 7317
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7438929071413733,
+      "learning_rate": 1.895527117532834e-05,
+      "loss": 0.6874,
+      "step": 7318
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.490164925221238,
+      "learning_rate": 1.8954931590686722e-05,
+      "loss": 0.6582,
+      "step": 7319
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7692989931787209,
+      "learning_rate": 1.8954591953906657e-05,
+      "loss": 0.6565,
+      "step": 7320
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3638652531893602,
+      "learning_rate": 1.8954252264990122e-05,
+      "loss": 0.7889,
+      "step": 7321
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5686613644237712,
+      "learning_rate": 1.895391252393909e-05,
+      "loss": 0.8092,
+      "step": 7322
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.015537198178956,
+      "learning_rate": 1.8953572730755552e-05,
+      "loss": 0.7079,
+      "step": 7323
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.504444551179879,
+      "learning_rate": 1.8953232885441474e-05,
+      "loss": 0.78,
+      "step": 7324
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6057516341645133,
+      "learning_rate": 1.8952892987998844e-05,
+      "loss": 0.925,
+      "step": 7325
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5960310323889233,
+      "learning_rate": 1.8952553038429634e-05,
+      "loss": 0.777,
+      "step": 7326
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7073137547282586,
+      "learning_rate": 1.8952213036735826e-05,
+      "loss": 0.8466,
+      "step": 7327
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6861155193653732,
+      "learning_rate": 1.8951872982919398e-05,
+      "loss": 0.675,
+      "step": 7328
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6308395251137557,
+      "learning_rate": 1.8951532876982335e-05,
+      "loss": 0.7333,
+      "step": 7329
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5431619307637925,
+      "learning_rate": 1.895119271892661e-05,
+      "loss": 0.8311,
+      "step": 7330
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.577021929299643,
+      "learning_rate": 1.895085250875421e-05,
+      "loss": 0.832,
+      "step": 7331
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0444922915031727,
+      "learning_rate": 1.8950512246467113e-05,
+      "loss": 0.8011,
+      "step": 7332
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9077381126237172,
+      "learning_rate": 1.8950171932067297e-05,
+      "loss": 0.7938,
+      "step": 7333
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.619201153194278,
+      "learning_rate": 1.8949831565556747e-05,
+      "loss": 0.8172,
+      "step": 7334
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5657235846003388,
+      "learning_rate": 1.8949491146937446e-05,
+      "loss": 0.7254,
+      "step": 7335
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.560541643518042,
+      "learning_rate": 1.8949150676211374e-05,
+      "loss": 0.8648,
+      "step": 7336
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6157587048775175,
+      "learning_rate": 1.8948810153380514e-05,
+      "loss": 0.7632,
+      "step": 7337
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.063219618974811,
+      "learning_rate": 1.8948469578446845e-05,
+      "loss": 0.7429,
+      "step": 7338
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6364639681536358,
+      "learning_rate": 1.8948128951412356e-05,
+      "loss": 0.5627,
+      "step": 7339
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5097318289809218,
+      "learning_rate": 1.8947788272279028e-05,
+      "loss": 0.7772,
+      "step": 7340
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5753278806464373,
+      "learning_rate": 1.8947447541048844e-05,
+      "loss": 0.5212,
+      "step": 7341
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.389012483141599,
+      "learning_rate": 1.8947106757723782e-05,
+      "loss": 0.8116,
+      "step": 7342
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4465268224261592,
+      "learning_rate": 1.8946765922305835e-05,
+      "loss": 0.6801,
+      "step": 7343
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.388955251008964,
+      "learning_rate": 1.8946425034796988e-05,
+      "loss": 0.8104,
+      "step": 7344
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6190966134807305,
+      "learning_rate": 1.8946084095199216e-05,
+      "loss": 0.8257,
+      "step": 7345
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5977168066795908,
+      "learning_rate": 1.8945743103514516e-05,
+      "loss": 0.7078,
+      "step": 7346
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8805459719916187,
+      "learning_rate": 1.8945402059744865e-05,
+      "loss": 0.7743,
+      "step": 7347
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5960493944751379,
+      "learning_rate": 1.894506096389225e-05,
+      "loss": 0.51,
+      "step": 7348
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5377402514981096,
+      "learning_rate": 1.894471981595866e-05,
+      "loss": 0.6941,
+      "step": 7349
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7677244331172588,
+      "learning_rate": 1.8944378615946077e-05,
+      "loss": 0.9002,
+      "step": 7350
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8733774281787217,
+      "learning_rate": 1.8944037363856492e-05,
+      "loss": 0.6917,
+      "step": 7351
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1206810213899416,
+      "learning_rate": 1.894369605969189e-05,
+      "loss": 0.7785,
+      "step": 7352
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5464426053003593,
+      "learning_rate": 1.894335470345426e-05,
+      "loss": 0.7438,
+      "step": 7353
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5752885587677047,
+      "learning_rate": 1.894301329514558e-05,
+      "loss": 0.5363,
+      "step": 7354
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3512613671130915,
+      "learning_rate": 1.894267183476785e-05,
+      "loss": 0.8657,
+      "step": 7355
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6212290455627928,
+      "learning_rate": 1.8942330322323055e-05,
+      "loss": 0.4803,
+      "step": 7356
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8429394830475376,
+      "learning_rate": 1.894198875781318e-05,
+      "loss": 0.8265,
+      "step": 7357
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7612786549841997,
+      "learning_rate": 1.8941647141240213e-05,
+      "loss": 0.8587,
+      "step": 7358
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6951639186802099,
+      "learning_rate": 1.894130547260615e-05,
+      "loss": 0.7576,
+      "step": 7359
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.067562248539772,
+      "learning_rate": 1.8940963751912976e-05,
+      "loss": 0.6979,
+      "step": 7360
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5539607880615547,
+      "learning_rate": 1.8940621979162678e-05,
+      "loss": 0.7477,
+      "step": 7361
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4731624738477256,
+      "learning_rate": 1.8940280154357248e-05,
+      "loss": 0.7079,
+      "step": 7362
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8637023555735035,
+      "learning_rate": 1.8939938277498675e-05,
+      "loss": 0.743,
+      "step": 7363
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6011283698628525,
+      "learning_rate": 1.8939596348588955e-05,
+      "loss": 0.7242,
+      "step": 7364
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5974836121574625,
+      "learning_rate": 1.893925436763007e-05,
+      "loss": 0.5251,
+      "step": 7365
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6114371915682986,
+      "learning_rate": 1.893891233462402e-05,
+      "loss": 0.8002,
+      "step": 7366
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6102712806950976,
+      "learning_rate": 1.8938570249572793e-05,
+      "loss": 0.5381,
+      "step": 7367
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9584845926649215,
+      "learning_rate": 1.8938228112478377e-05,
+      "loss": 0.9586,
+      "step": 7368
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6090629675198564,
+      "learning_rate": 1.8937885923342768e-05,
+      "loss": 0.6779,
+      "step": 7369
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5902154739284147,
+      "learning_rate": 1.893754368216796e-05,
+      "loss": 0.4969,
+      "step": 7370
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5441807568158228,
+      "learning_rate": 1.893720138895594e-05,
+      "loss": 0.702,
+      "step": 7371
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.4071748521284646,
+      "learning_rate": 1.8936859043708705e-05,
+      "loss": 0.7128,
+      "step": 7372
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4282677251274603,
+      "learning_rate": 1.893651664642825e-05,
+      "loss": 0.6814,
+      "step": 7373
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6968992834219427,
+      "learning_rate": 1.893617419711656e-05,
+      "loss": 0.8069,
+      "step": 7374
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.269401776323918,
+      "learning_rate": 1.893583169577564e-05,
+      "loss": 0.7222,
+      "step": 7375
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5936531558719356,
+      "learning_rate": 1.8935489142407474e-05,
+      "loss": 0.8258,
+      "step": 7376
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7266398591118066,
+      "learning_rate": 1.8935146537014064e-05,
+      "loss": 0.8251,
+      "step": 7377
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0136158496123375,
+      "learning_rate": 1.8934803879597402e-05,
+      "loss": 0.7292,
+      "step": 7378
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9796267383396313,
+      "learning_rate": 1.8934461170159483e-05,
+      "loss": 0.8789,
+      "step": 7379
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8787932692905776,
+      "learning_rate": 1.8934118408702306e-05,
+      "loss": 0.7932,
+      "step": 7380
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4855082340201198,
+      "learning_rate": 1.8933775595227856e-05,
+      "loss": 0.746,
+      "step": 7381
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.824430303182885,
+      "learning_rate": 1.893343272973814e-05,
+      "loss": 0.833,
+      "step": 7382
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.705913636378365,
+      "learning_rate": 1.8933089812235153e-05,
+      "loss": 0.7477,
+      "step": 7383
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.686434885158587,
+      "learning_rate": 1.8932746842720884e-05,
+      "loss": 0.8365,
+      "step": 7384
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9298283265803549,
+      "learning_rate": 1.8932403821197338e-05,
+      "loss": 0.6879,
+      "step": 7385
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6492806448375643,
+      "learning_rate": 1.8932060747666504e-05,
+      "loss": 0.8609,
+      "step": 7386
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6427363091204312,
+      "learning_rate": 1.8931717622130392e-05,
+      "loss": 0.7912,
+      "step": 7387
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.555181037266651,
+      "learning_rate": 1.8931374444590987e-05,
+      "loss": 0.7203,
+      "step": 7388
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4996542455475803,
+      "learning_rate": 1.8931031215050295e-05,
+      "loss": 0.795,
+      "step": 7389
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5584018955915908,
+      "learning_rate": 1.893068793351031e-05,
+      "loss": 0.7033,
+      "step": 7390
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.4353407655103054,
+      "learning_rate": 1.8930344599973035e-05,
+      "loss": 0.8292,
+      "step": 7391
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5658629962909345,
+      "learning_rate": 1.8930001214440464e-05,
+      "loss": 0.7464,
+      "step": 7392
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8796331877751682,
+      "learning_rate": 1.89296577769146e-05,
+      "loss": 0.7159,
+      "step": 7393
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5255319933019047,
+      "learning_rate": 1.892931428739744e-05,
+      "loss": 0.6387,
+      "step": 7394
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.660908687694598,
+      "learning_rate": 1.8928970745890985e-05,
+      "loss": 0.6716,
+      "step": 7395
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.754607538410095,
+      "learning_rate": 1.8928627152397237e-05,
+      "loss": 0.7941,
+      "step": 7396
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3593065088124217,
+      "learning_rate": 1.8928283506918194e-05,
+      "loss": 1.0082,
+      "step": 7397
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7626220517012081,
+      "learning_rate": 1.892793980945586e-05,
+      "loss": 0.8598,
+      "step": 7398
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5463687819577998,
+      "learning_rate": 1.892759606001223e-05,
+      "loss": 0.7078,
+      "step": 7399
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5837881197447964,
+      "learning_rate": 1.8927252258589312e-05,
+      "loss": 0.5609,
+      "step": 7400
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7863108484935484,
+      "learning_rate": 1.8926908405189105e-05,
+      "loss": 0.7492,
+      "step": 7401
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5822474946831212,
+      "learning_rate": 1.8926564499813608e-05,
+      "loss": 0.5254,
+      "step": 7402
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5023327625417342,
+      "learning_rate": 1.8926220542464833e-05,
+      "loss": 0.7227,
+      "step": 7403
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6162747723857819,
+      "learning_rate": 1.8925876533144772e-05,
+      "loss": 0.5131,
+      "step": 7404
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.441301301510954,
+      "learning_rate": 1.892553247185543e-05,
+      "loss": 0.7627,
+      "step": 7405
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6952788513848798,
+      "learning_rate": 1.8925188358598815e-05,
+      "loss": 0.7726,
+      "step": 7406
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.311583053966749,
+      "learning_rate": 1.8924844193376926e-05,
+      "loss": 0.6832,
+      "step": 7407
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9896600610031072,
+      "learning_rate": 1.892449997619177e-05,
+      "loss": 0.8717,
+      "step": 7408
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6525498315423832,
+      "learning_rate": 1.8924155707045348e-05,
+      "loss": 0.7763,
+      "step": 7409
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.736705372418615,
+      "learning_rate": 1.8923811385939666e-05,
+      "loss": 0.8317,
+      "step": 7410
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9109085815350673,
+      "learning_rate": 1.892346701287673e-05,
+      "loss": 0.8467,
+      "step": 7411
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9305862773238471,
+      "learning_rate": 1.8923122587858543e-05,
+      "loss": 0.867,
+      "step": 7412
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5824637527057028,
+      "learning_rate": 1.8922778110887114e-05,
+      "loss": 0.7612,
+      "step": 7413
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4040038764547254,
+      "learning_rate": 1.8922433581964443e-05,
+      "loss": 0.8305,
+      "step": 7414
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5851397650427677,
+      "learning_rate": 1.892208900109254e-05,
+      "loss": 0.8479,
+      "step": 7415
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3683173765841676,
+      "learning_rate": 1.892174436827341e-05,
+      "loss": 0.7795,
+      "step": 7416
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6076377229475953,
+      "learning_rate": 1.892139968350906e-05,
+      "loss": 0.5123,
+      "step": 7417
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5140509403024172,
+      "learning_rate": 1.8921054946801497e-05,
+      "loss": 0.9302,
+      "step": 7418
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7631921510150705,
+      "learning_rate": 1.8920710158152727e-05,
+      "loss": 0.6463,
+      "step": 7419
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.535716043986291,
+      "learning_rate": 1.8920365317564758e-05,
+      "loss": 0.6916,
+      "step": 7420
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5826069042697692,
+      "learning_rate": 1.8920020425039598e-05,
+      "loss": 0.531,
+      "step": 7421
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0588685045247233,
+      "learning_rate": 1.8919675480579256e-05,
+      "loss": 0.8458,
+      "step": 7422
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.849499997880088,
+      "learning_rate": 1.891933048418574e-05,
+      "loss": 0.9218,
+      "step": 7423
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6117441908538719,
+      "learning_rate": 1.8918985435861056e-05,
+      "loss": 0.5622,
+      "step": 7424
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5893840669893924,
+      "learning_rate": 1.8918640335607215e-05,
+      "loss": 0.5066,
+      "step": 7425
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5498525932135063,
+      "learning_rate": 1.8918295183426226e-05,
+      "loss": 0.7454,
+      "step": 7426
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.921271374786598,
+      "learning_rate": 1.89179499793201e-05,
+      "loss": 0.7761,
+      "step": 7427
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7570222655133696,
+      "learning_rate": 1.891760472329085e-05,
+      "loss": 0.6893,
+      "step": 7428
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7030643880210816,
+      "learning_rate": 1.891725941534048e-05,
+      "loss": 0.7248,
+      "step": 7429
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4220208814549484,
+      "learning_rate": 1.8916914055471e-05,
+      "loss": 0.6081,
+      "step": 7430
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.697396352293893,
+      "learning_rate": 1.891656864368442e-05,
+      "loss": 0.7696,
+      "step": 7431
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.698341804144984,
+      "learning_rate": 1.891622317998276e-05,
+      "loss": 0.7964,
+      "step": 7432
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5662402703170355,
+      "learning_rate": 1.8915877664368026e-05,
+      "loss": 0.7933,
+      "step": 7433
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9820173646264672,
+      "learning_rate": 1.8915532096842225e-05,
+      "loss": 0.7026,
+      "step": 7434
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4177980956967322,
+      "learning_rate": 1.8915186477407376e-05,
+      "loss": 0.9121,
+      "step": 7435
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6536684015599993,
+      "learning_rate": 1.891484080606549e-05,
+      "loss": 0.5462,
+      "step": 7436
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8303209624318266,
+      "learning_rate": 1.8914495082818578e-05,
+      "loss": 0.7386,
+      "step": 7437
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4699342470625072,
+      "learning_rate": 1.891414930766865e-05,
+      "loss": 0.791,
+      "step": 7438
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.815399193271962,
+      "learning_rate": 1.891380348061773e-05,
+      "loss": 0.768,
+      "step": 7439
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.799226443106915,
+      "learning_rate": 1.891345760166782e-05,
+      "loss": 0.7848,
+      "step": 7440
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5457197313161382,
+      "learning_rate": 1.8913111670820936e-05,
+      "loss": 0.8475,
+      "step": 7441
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.501154704646998,
+      "learning_rate": 1.8912765688079097e-05,
+      "loss": 0.6833,
+      "step": 7442
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6626310814635847,
+      "learning_rate": 1.8912419653444313e-05,
+      "loss": 0.6824,
+      "step": 7443
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6717942996910697,
+      "learning_rate": 1.8912073566918597e-05,
+      "loss": 0.6538,
+      "step": 7444
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4444136796219205,
+      "learning_rate": 1.891172742850397e-05,
+      "loss": 0.6381,
+      "step": 7445
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8158896417301509,
+      "learning_rate": 1.8911381238202447e-05,
+      "loss": 0.7255,
+      "step": 7446
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6290199772841152,
+      "learning_rate": 1.891103499601604e-05,
+      "loss": 0.5826,
+      "step": 7447
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4385380693747514,
+      "learning_rate": 1.8910688701946763e-05,
+      "loss": 0.7969,
+      "step": 7448
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5911505895705806,
+      "learning_rate": 1.8910342355996637e-05,
+      "loss": 0.7244,
+      "step": 7449
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7571827331244174,
+      "learning_rate": 1.8909995958167678e-05,
+      "loss": 0.7631,
+      "step": 7450
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1007161633357976,
+      "learning_rate": 1.89096495084619e-05,
+      "loss": 0.8095,
+      "step": 7451
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7348771897695079,
+      "learning_rate": 1.8909303006881323e-05,
+      "loss": 0.8149,
+      "step": 7452
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6959034840052392,
+      "learning_rate": 1.8908956453427963e-05,
+      "loss": 0.8375,
+      "step": 7453
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.7640928942854672,
+      "learning_rate": 1.890860984810384e-05,
+      "loss": 0.7308,
+      "step": 7454
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9619888405089365,
+      "learning_rate": 1.8908263190910966e-05,
+      "loss": 0.6814,
+      "step": 7455
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.7133106249198686,
+      "learning_rate": 1.8907916481851364e-05,
+      "loss": 0.7592,
+      "step": 7456
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5814360797113645,
+      "learning_rate": 1.8907569720927055e-05,
+      "loss": 0.8253,
+      "step": 7457
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7738782592559041,
+      "learning_rate": 1.8907222908140053e-05,
+      "loss": 0.7718,
+      "step": 7458
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1543661643122842,
+      "learning_rate": 1.8906876043492382e-05,
+      "loss": 0.8063,
+      "step": 7459
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5852467120239173,
+      "learning_rate": 1.8906529126986057e-05,
+      "loss": 0.8256,
+      "step": 7460
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4469991363323735,
+      "learning_rate": 1.8906182158623097e-05,
+      "loss": 0.8154,
+      "step": 7461
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4471421502596034,
+      "learning_rate": 1.8905835138405526e-05,
+      "loss": 0.7812,
+      "step": 7462
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6958283265156826,
+      "learning_rate": 1.8905488066335366e-05,
+      "loss": 0.7448,
+      "step": 7463
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.80402628345819,
+      "learning_rate": 1.8905140942414633e-05,
+      "loss": 0.8398,
+      "step": 7464
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5532818433121864,
+      "learning_rate": 1.890479376664535e-05,
+      "loss": 0.8238,
+      "step": 7465
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6006596390517914,
+      "learning_rate": 1.890444653902954e-05,
+      "loss": 0.5135,
+      "step": 7466
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5172086719432951,
+      "learning_rate": 1.890409925956922e-05,
+      "loss": 0.7324,
+      "step": 7467
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6152767588996778,
+      "learning_rate": 1.8903751928266415e-05,
+      "loss": 0.5276,
+      "step": 7468
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.588067122927689,
+      "learning_rate": 1.890340454512315e-05,
+      "loss": 0.8701,
+      "step": 7469
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5990924232766504,
+      "learning_rate": 1.8903057110141443e-05,
+      "loss": 0.8449,
+      "step": 7470
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8088736374998124,
+      "learning_rate": 1.8902709623323322e-05,
+      "loss": 0.7461,
+      "step": 7471
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8613160988249924,
+      "learning_rate": 1.8902362084670806e-05,
+      "loss": 0.7743,
+      "step": 7472
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1105266639387037,
+      "learning_rate": 1.8902014494185918e-05,
+      "loss": 0.8143,
+      "step": 7473
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6294621657322694,
+      "learning_rate": 1.8901666851870684e-05,
+      "loss": 0.5516,
+      "step": 7474
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2491343487859994,
+      "learning_rate": 1.8901319157727127e-05,
+      "loss": 0.7994,
+      "step": 7475
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6020791992245536,
+      "learning_rate": 1.8900971411757273e-05,
+      "loss": 0.8609,
+      "step": 7476
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3314372299593424,
+      "learning_rate": 1.8900623613963144e-05,
+      "loss": 0.8064,
+      "step": 7477
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4262813871481688,
+      "learning_rate": 1.890027576434677e-05,
+      "loss": 0.7127,
+      "step": 7478
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5746513667288686,
+      "learning_rate": 1.889992786291017e-05,
+      "loss": 0.5767,
+      "step": 7479
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.59859450015418,
+      "learning_rate": 1.889957990965537e-05,
+      "loss": 0.8573,
+      "step": 7480
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7143869430748742,
+      "learning_rate": 1.88992319045844e-05,
+      "loss": 0.6467,
+      "step": 7481
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.032576018832485,
+      "learning_rate": 1.8898883847699284e-05,
+      "loss": 0.8749,
+      "step": 7482
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6160261064694466,
+      "learning_rate": 1.889853573900205e-05,
+      "loss": 0.6871,
+      "step": 7483
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5843749485230267,
+      "learning_rate": 1.8898187578494723e-05,
+      "loss": 0.768,
+      "step": 7484
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1975392449601436,
+      "learning_rate": 1.889783936617933e-05,
+      "loss": 0.6329,
+      "step": 7485
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6196110085582247,
+      "learning_rate": 1.88974911020579e-05,
+      "loss": 0.8623,
+      "step": 7486
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7713992047332494,
+      "learning_rate": 1.889714278613246e-05,
+      "loss": 0.72,
+      "step": 7487
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5097961695943671,
+      "learning_rate": 1.889679441840504e-05,
+      "loss": 0.836,
+      "step": 7488
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6100735472023421,
+      "learning_rate": 1.8896445998877664e-05,
+      "loss": 0.5855,
+      "step": 7489
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0688872978615622,
+      "learning_rate": 1.8896097527552362e-05,
+      "loss": 0.8955,
+      "step": 7490
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7864089201125237,
+      "learning_rate": 1.8895749004431168e-05,
+      "loss": 0.8313,
+      "step": 7491
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5921637748158737,
+      "learning_rate": 1.88954004295161e-05,
+      "loss": 0.8338,
+      "step": 7492
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6620206586742434,
+      "learning_rate": 1.8895051802809202e-05,
+      "loss": 0.7618,
+      "step": 7493
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8807722098447854,
+      "learning_rate": 1.8894703124312492e-05,
+      "loss": 0.8293,
+      "step": 7494
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.022866416438747,
+      "learning_rate": 1.8894354394028005e-05,
+      "loss": 0.632,
+      "step": 7495
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6838700921736596,
+      "learning_rate": 1.889400561195777e-05,
+      "loss": 0.8631,
+      "step": 7496
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8034391486366737,
+      "learning_rate": 1.889365677810382e-05,
+      "loss": 0.8115,
+      "step": 7497
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5233666096582548,
+      "learning_rate": 1.8893307892468184e-05,
+      "loss": 0.8167,
+      "step": 7498
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5835034868039541,
+      "learning_rate": 1.8892958955052893e-05,
+      "loss": 0.8379,
+      "step": 7499
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8587780480629297,
+      "learning_rate": 1.889260996585998e-05,
+      "loss": 0.7536,
+      "step": 7500
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.60034275122068,
+      "learning_rate": 1.889226092489148e-05,
+      "loss": 0.779,
+      "step": 7501
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7196344454714179,
+      "learning_rate": 1.8891911832149418e-05,
+      "loss": 0.8124,
+      "step": 7502
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.095869135941746,
+      "learning_rate": 1.889156268763583e-05,
+      "loss": 0.7796,
+      "step": 7503
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.576419190185308,
+      "learning_rate": 1.889121349135275e-05,
+      "loss": 0.7811,
+      "step": 7504
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.738178324532014,
+      "learning_rate": 1.889086424330221e-05,
+      "loss": 0.8513,
+      "step": 7505
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6902772575018963,
+      "learning_rate": 1.8890514943486242e-05,
+      "loss": 0.75,
+      "step": 7506
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.492241279500603,
+      "learning_rate": 1.8890165591906883e-05,
+      "loss": 0.7572,
+      "step": 7507
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6642035018519417,
+      "learning_rate": 1.8889816188566163e-05,
+      "loss": 0.7648,
+      "step": 7508
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6111317532173068,
+      "learning_rate": 1.8889466733466124e-05,
+      "loss": 0.8206,
+      "step": 7509
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2671973136149735,
+      "learning_rate": 1.888911722660879e-05,
+      "loss": 0.7012,
+      "step": 7510
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7901659575266289,
+      "learning_rate": 1.88887676679962e-05,
+      "loss": 0.6989,
+      "step": 7511
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6279734135523742,
+      "learning_rate": 1.8888418057630397e-05,
+      "loss": 0.742,
+      "step": 7512
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.368729162097255,
+      "learning_rate": 1.8888068395513407e-05,
+      "loss": 0.8,
+      "step": 7513
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8451027917148295,
+      "learning_rate": 1.8887718681647265e-05,
+      "loss": 0.7169,
+      "step": 7514
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.74672426900192,
+      "learning_rate": 1.8887368916034017e-05,
+      "loss": 0.7739,
+      "step": 7515
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6098878754964485,
+      "learning_rate": 1.8887019098675692e-05,
+      "loss": 0.545,
+      "step": 7516
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5902151096778512,
+      "learning_rate": 1.8886669229574326e-05,
+      "loss": 0.7416,
+      "step": 7517
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3686341928804175,
+      "learning_rate": 1.888631930873196e-05,
+      "loss": 0.7694,
+      "step": 7518
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5522363278197864,
+      "learning_rate": 1.888596933615063e-05,
+      "loss": 0.7721,
+      "step": 7519
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9601560753009521,
+      "learning_rate": 1.888561931183237e-05,
+      "loss": 0.8206,
+      "step": 7520
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.794439649831193,
+      "learning_rate": 1.8885269235779226e-05,
+      "loss": 0.8259,
+      "step": 7521
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5873917658245718,
+      "learning_rate": 1.888491910799323e-05,
+      "loss": 0.8633,
+      "step": 7522
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7380773501406948,
+      "learning_rate": 1.888456892847642e-05,
+      "loss": 0.8382,
+      "step": 7523
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4345679881604207,
+      "learning_rate": 1.8884218697230838e-05,
+      "loss": 0.5945,
+      "step": 7524
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.442419188338344,
+      "learning_rate": 1.8883868414258525e-05,
+      "loss": 0.8049,
+      "step": 7525
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6732290113648072,
+      "learning_rate": 1.8883518079561512e-05,
+      "loss": 0.9107,
+      "step": 7526
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8001013885733537,
+      "learning_rate": 1.8883167693141845e-05,
+      "loss": 0.8861,
+      "step": 7527
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6249763833786369,
+      "learning_rate": 1.8882817255001567e-05,
+      "loss": 0.5335,
+      "step": 7528
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.583201850701863,
+      "learning_rate": 1.8882466765142715e-05,
+      "loss": 0.7391,
+      "step": 7529
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5009796416371133,
+      "learning_rate": 1.8882116223567327e-05,
+      "loss": 0.7307,
+      "step": 7530
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.365956697510336,
+      "learning_rate": 1.8881765630277447e-05,
+      "loss": 0.7671,
+      "step": 7531
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1349620600153147,
+      "learning_rate": 1.8881414985275114e-05,
+      "loss": 0.7309,
+      "step": 7532
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.881855628224024,
+      "learning_rate": 1.8881064288562373e-05,
+      "loss": 0.8661,
+      "step": 7533
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6105895225623301,
+      "learning_rate": 1.888071354014126e-05,
+      "loss": 0.6415,
+      "step": 7534
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9313989999440337,
+      "learning_rate": 1.8880362740013827e-05,
+      "loss": 0.7389,
+      "step": 7535
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5951173650477737,
+      "learning_rate": 1.8880011888182108e-05,
+      "loss": 0.5718,
+      "step": 7536
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6022748480710143,
+      "learning_rate": 1.887966098464815e-05,
+      "loss": 0.7488,
+      "step": 7537
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.568414188291219,
+      "learning_rate": 1.8879310029413993e-05,
+      "loss": 0.6521,
+      "step": 7538
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6802890644162423,
+      "learning_rate": 1.8878959022481682e-05,
+      "loss": 0.8165,
+      "step": 7539
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9068048486082614,
+      "learning_rate": 1.887860796385326e-05,
+      "loss": 0.6778,
+      "step": 7540
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.157317676812071,
+      "learning_rate": 1.8878256853530772e-05,
+      "loss": 0.8322,
+      "step": 7541
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.635844275790383,
+      "learning_rate": 1.8877905691516263e-05,
+      "loss": 0.8201,
+      "step": 7542
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6069310683928526,
+      "learning_rate": 1.8877554477811775e-05,
+      "loss": 0.8686,
+      "step": 7543
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6524007522728383,
+      "learning_rate": 1.8877203212419354e-05,
+      "loss": 0.5875,
+      "step": 7544
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.361671356659305,
+      "learning_rate": 1.8876851895341044e-05,
+      "loss": 0.7633,
+      "step": 7545
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.535089258064715,
+      "learning_rate": 1.8876500526578897e-05,
+      "loss": 0.7145,
+      "step": 7546
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.751893447258385,
+      "learning_rate": 1.8876149106134947e-05,
+      "loss": 0.7006,
+      "step": 7547
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6161823194630165,
+      "learning_rate": 1.8875797634011254e-05,
+      "loss": 0.7183,
+      "step": 7548
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9517970185011888,
+      "learning_rate": 1.887544611020985e-05,
+      "loss": 0.786,
+      "step": 7549
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.573317748471186,
+      "learning_rate": 1.8875094534732793e-05,
+      "loss": 0.8205,
+      "step": 7550
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7080853470112252,
+      "learning_rate": 1.8874742907582124e-05,
+      "loss": 0.8034,
+      "step": 7551
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8994616481224806,
+      "learning_rate": 1.8874391228759894e-05,
+      "loss": 0.7621,
+      "step": 7552
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.575656368103326,
+      "learning_rate": 1.8874039498268147e-05,
+      "loss": 0.5344,
+      "step": 7553
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7853223862874603,
+      "learning_rate": 1.8873687716108935e-05,
+      "loss": 0.6764,
+      "step": 7554
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.331986293437267,
+      "learning_rate": 1.88733358822843e-05,
+      "loss": 0.8282,
+      "step": 7555
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.128430894673289,
+      "learning_rate": 1.8872983996796298e-05,
+      "loss": 0.9532,
+      "step": 7556
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6089355626539088,
+      "learning_rate": 1.8872632059646973e-05,
+      "loss": 0.5921,
+      "step": 7557
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7596143308932588,
+      "learning_rate": 1.8872280070838373e-05,
+      "loss": 0.7249,
+      "step": 7558
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.916116022831737,
+      "learning_rate": 1.8871928030372553e-05,
+      "loss": 0.6616,
+      "step": 7559
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4754404421345166,
+      "learning_rate": 1.8871575938251557e-05,
+      "loss": 0.6252,
+      "step": 7560
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7275408664408767,
+      "learning_rate": 1.887122379447744e-05,
+      "loss": 0.7888,
+      "step": 7561
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.534614337552353,
+      "learning_rate": 1.8870871599052244e-05,
+      "loss": 0.7122,
+      "step": 7562
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5631493301652133,
+      "learning_rate": 1.8870519351978026e-05,
+      "loss": 0.5327,
+      "step": 7563
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.927454628307499,
+      "learning_rate": 1.8870167053256838e-05,
+      "loss": 0.7973,
+      "step": 7564
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.589225735208528,
+      "learning_rate": 1.8869814702890733e-05,
+      "loss": 0.5416,
+      "step": 7565
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.815955201333621,
+      "learning_rate": 1.8869462300881755e-05,
+      "loss": 0.7239,
+      "step": 7566
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5198117493872452,
+      "learning_rate": 1.8869109847231956e-05,
+      "loss": 0.717,
+      "step": 7567
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5430870172004707,
+      "learning_rate": 1.8868757341943396e-05,
+      "loss": 0.8364,
+      "step": 7568
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5544806851209982,
+      "learning_rate": 1.886840478501812e-05,
+      "loss": 0.727,
+      "step": 7569
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3316415214376447,
+      "learning_rate": 1.8868052176458188e-05,
+      "loss": 0.8733,
+      "step": 7570
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.554589547051462,
+      "learning_rate": 1.8867699516265646e-05,
+      "loss": 0.7655,
+      "step": 7571
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8063708875714486,
+      "learning_rate": 1.886734680444255e-05,
+      "loss": 0.7466,
+      "step": 7572
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6179313182843171,
+      "learning_rate": 1.8866994040990954e-05,
+      "loss": 0.7598,
+      "step": 7573
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7338315003359583,
+      "learning_rate": 1.8866641225912908e-05,
+      "loss": 0.8061,
+      "step": 7574
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.283611349742605,
+      "learning_rate": 1.8866288359210474e-05,
+      "loss": 0.9321,
+      "step": 7575
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9240727703925518,
+      "learning_rate": 1.88659354408857e-05,
+      "loss": 0.7355,
+      "step": 7576
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.961810570078443,
+      "learning_rate": 1.8865582470940644e-05,
+      "loss": 0.8321,
+      "step": 7577
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6978708422417474,
+      "learning_rate": 1.886522944937736e-05,
+      "loss": 0.9123,
+      "step": 7578
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8151136184366663,
+      "learning_rate": 1.8864876376197904e-05,
+      "loss": 0.6531,
+      "step": 7579
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6632036280093139,
+      "learning_rate": 1.886452325140433e-05,
+      "loss": 0.6726,
+      "step": 7580
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9166102321250922,
+      "learning_rate": 1.8864170074998695e-05,
+      "loss": 0.6534,
+      "step": 7581
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7832497950056674,
+      "learning_rate": 1.8863816846983055e-05,
+      "loss": 0.6644,
+      "step": 7582
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5124594375686216,
+      "learning_rate": 1.886346356735947e-05,
+      "loss": 0.7767,
+      "step": 7583
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.4925814318975923,
+      "learning_rate": 1.8863110236129993e-05,
+      "loss": 0.7623,
+      "step": 7584
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7300984318589654,
+      "learning_rate": 1.886275685329668e-05,
+      "loss": 0.6348,
+      "step": 7585
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9685439570267476,
+      "learning_rate": 1.8862403418861593e-05,
+      "loss": 0.696,
+      "step": 7586
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6641510131931159,
+      "learning_rate": 1.8862049932826787e-05,
+      "loss": 0.7991,
+      "step": 7587
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4615656064999818,
+      "learning_rate": 1.886169639519432e-05,
+      "loss": 0.6904,
+      "step": 7588
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6153548521700498,
+      "learning_rate": 1.886134280596625e-05,
+      "loss": 0.6471,
+      "step": 7589
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6126591664808574,
+      "learning_rate": 1.886098916514464e-05,
+      "loss": 0.7271,
+      "step": 7590
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3710888570188096,
+      "learning_rate": 1.886063547273154e-05,
+      "loss": 0.7065,
+      "step": 7591
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5753962192683921,
+      "learning_rate": 1.886028172872902e-05,
+      "loss": 0.6385,
+      "step": 7592
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.00323750921651,
+      "learning_rate": 1.885992793313913e-05,
+      "loss": 0.6996,
+      "step": 7593
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7258380986368769,
+      "learning_rate": 1.885957408596394e-05,
+      "loss": 0.8098,
+      "step": 7594
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.42712799156934,
+      "learning_rate": 1.8859220187205505e-05,
+      "loss": 0.8729,
+      "step": 7595
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3915499465973622,
+      "learning_rate": 1.8858866236865883e-05,
+      "loss": 0.8318,
+      "step": 7596
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1243676062905577,
+      "learning_rate": 1.8858512234947138e-05,
+      "loss": 0.7721,
+      "step": 7597
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5186060987023235,
+      "learning_rate": 1.8858158181451327e-05,
+      "loss": 0.7237,
+      "step": 7598
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8228188763423838,
+      "learning_rate": 1.885780407638052e-05,
+      "loss": 0.9172,
+      "step": 7599
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9970267911770452,
+      "learning_rate": 1.885744991973677e-05,
+      "loss": 0.7564,
+      "step": 7600
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.770115403727674,
+      "learning_rate": 1.885709571152214e-05,
+      "loss": 0.7194,
+      "step": 7601
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4897998262119596,
+      "learning_rate": 1.8856741451738697e-05,
+      "loss": 0.7162,
+      "step": 7602
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1599405268461074,
+      "learning_rate": 1.88563871403885e-05,
+      "loss": 0.8163,
+      "step": 7603
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7281429305789957,
+      "learning_rate": 1.8856032777473617e-05,
+      "loss": 0.7915,
+      "step": 7604
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7596396834866443,
+      "learning_rate": 1.8855678362996105e-05,
+      "loss": 0.6582,
+      "step": 7605
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7739048056805475,
+      "learning_rate": 1.885532389695803e-05,
+      "loss": 0.8172,
+      "step": 7606
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8033397290452275,
+      "learning_rate": 1.8854969379361456e-05,
+      "loss": 0.7496,
+      "step": 7607
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7384458378394638,
+      "learning_rate": 1.8854614810208446e-05,
+      "loss": 0.6736,
+      "step": 7608
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6153324475736701,
+      "learning_rate": 1.8854260189501068e-05,
+      "loss": 0.7339,
+      "step": 7609
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5814319763141578,
+      "learning_rate": 1.8853905517241384e-05,
+      "loss": 0.8315,
+      "step": 7610
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5279601380244279,
+      "learning_rate": 1.885355079343145e-05,
+      "loss": 0.8385,
+      "step": 7611
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4487709153609554,
+      "learning_rate": 1.8853196018073352e-05,
+      "loss": 0.7863,
+      "step": 7612
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7499523760557203,
+      "learning_rate": 1.8852841191169138e-05,
+      "loss": 0.7425,
+      "step": 7613
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4854559373228728,
+      "learning_rate": 1.885248631272088e-05,
+      "loss": 0.7296,
+      "step": 7614
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0950645610491416,
+      "learning_rate": 1.8852131382730645e-05,
+      "loss": 0.804,
+      "step": 7615
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7478581617361097,
+      "learning_rate": 1.88517764012005e-05,
+      "loss": 0.784,
+      "step": 7616
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5840784254996836,
+      "learning_rate": 1.8851421368132507e-05,
+      "loss": 0.5339,
+      "step": 7617
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6600494502054075,
+      "learning_rate": 1.8851066283528738e-05,
+      "loss": 0.6427,
+      "step": 7618
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3646851022808963,
+      "learning_rate": 1.8850711147391257e-05,
+      "loss": 0.7698,
+      "step": 7619
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6153430748794022,
+      "learning_rate": 1.8850355959722138e-05,
+      "loss": 0.5758,
+      "step": 7620
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6917009056373356,
+      "learning_rate": 1.885000072052344e-05,
+      "loss": 0.604,
+      "step": 7621
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.529963399311194,
+      "learning_rate": 1.884964542979724e-05,
+      "loss": 0.6814,
+      "step": 7622
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6278204524871559,
+      "learning_rate": 1.88492900875456e-05,
+      "loss": 0.5409,
+      "step": 7623
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5860304460587105,
+      "learning_rate": 1.8848934693770594e-05,
+      "loss": 0.7283,
+      "step": 7624
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6925250638920362,
+      "learning_rate": 1.8848579248474286e-05,
+      "loss": 0.8022,
+      "step": 7625
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.480743126145777,
+      "learning_rate": 1.884822375165875e-05,
+      "loss": 0.8717,
+      "step": 7626
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7167538104600295,
+      "learning_rate": 1.8847868203326053e-05,
+      "loss": 0.7124,
+      "step": 7627
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6959502205606587,
+      "learning_rate": 1.8847512603478267e-05,
+      "loss": 0.8722,
+      "step": 7628
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2876016881068713,
+      "learning_rate": 1.8847156952117462e-05,
+      "loss": 0.7811,
+      "step": 7629
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5548521352238058,
+      "learning_rate": 1.8846801249245706e-05,
+      "loss": 0.7859,
+      "step": 7630
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9305338407348005,
+      "learning_rate": 1.8846445494865074e-05,
+      "loss": 0.7992,
+      "step": 7631
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8955456443896568,
+      "learning_rate": 1.8846089688977635e-05,
+      "loss": 0.8316,
+      "step": 7632
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7540986175070912,
+      "learning_rate": 1.884573383158546e-05,
+      "loss": 0.771,
+      "step": 7633
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3978334809338822,
+      "learning_rate": 1.8845377922690623e-05,
+      "loss": 0.7449,
+      "step": 7634
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9311386113184927,
+      "learning_rate": 1.88450219622952e-05,
+      "loss": 0.8182,
+      "step": 7635
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5703221297941214,
+      "learning_rate": 1.884466595040125e-05,
+      "loss": 0.5417,
+      "step": 7636
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8868515097623249,
+      "learning_rate": 1.8844309887010862e-05,
+      "loss": 0.7407,
+      "step": 7637
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5636435596810885,
+      "learning_rate": 1.88439537721261e-05,
+      "loss": 0.7424,
+      "step": 7638
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9714973579820447,
+      "learning_rate": 1.8843597605749035e-05,
+      "loss": 0.8188,
+      "step": 7639
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.166069761760411,
+      "learning_rate": 1.884324138788175e-05,
+      "loss": 0.8042,
+      "step": 7640
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5156531660738977,
+      "learning_rate": 1.8842885118526314e-05,
+      "loss": 0.7649,
+      "step": 7641
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6831326175302754,
+      "learning_rate": 1.8842528797684795e-05,
+      "loss": 0.6569,
+      "step": 7642
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3957634903971583,
+      "learning_rate": 1.884217242535928e-05,
+      "loss": 0.7524,
+      "step": 7643
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5712845707242236,
+      "learning_rate": 1.8841816001551836e-05,
+      "loss": 0.653,
+      "step": 7644
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4968175584686503,
+      "learning_rate": 1.8841459526264542e-05,
+      "loss": 0.7631,
+      "step": 7645
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4744422905667338,
+      "learning_rate": 1.8841102999499468e-05,
+      "loss": 0.708,
+      "step": 7646
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6239702838924803,
+      "learning_rate": 1.8840746421258694e-05,
+      "loss": 0.7123,
+      "step": 7647
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4781495794039736,
+      "learning_rate": 1.8840389791544297e-05,
+      "loss": 0.8461,
+      "step": 7648
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4683437868688238,
+      "learning_rate": 1.884003311035835e-05,
+      "loss": 0.7809,
+      "step": 7649
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9759987945032742,
+      "learning_rate": 1.8839676377702934e-05,
+      "loss": 0.7202,
+      "step": 7650
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.544682573290478,
+      "learning_rate": 1.883931959358012e-05,
+      "loss": 0.8091,
+      "step": 7651
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2341760730691242,
+      "learning_rate": 1.8838962757991993e-05,
+      "loss": 0.8065,
+      "step": 7652
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5052981325520152,
+      "learning_rate": 1.883860587094062e-05,
+      "loss": 0.781,
+      "step": 7653
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4379546961096585,
+      "learning_rate": 1.883824893242809e-05,
+      "loss": 0.6712,
+      "step": 7654
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4843597486223807,
+      "learning_rate": 1.8837891942456477e-05,
+      "loss": 0.7563,
+      "step": 7655
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6343467354877839,
+      "learning_rate": 1.883753490102786e-05,
+      "loss": 0.534,
+      "step": 7656
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9091051617787862,
+      "learning_rate": 1.8837177808144313e-05,
+      "loss": 0.7586,
+      "step": 7657
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6215440240735485,
+      "learning_rate": 1.8836820663807924e-05,
+      "loss": 0.5872,
+      "step": 7658
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9896820150629597,
+      "learning_rate": 1.8836463468020764e-05,
+      "loss": 0.6389,
+      "step": 7659
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.480230066616,
+      "learning_rate": 1.8836106220784915e-05,
+      "loss": 0.6787,
+      "step": 7660
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4987486150142264,
+      "learning_rate": 1.883574892210246e-05,
+      "loss": 0.6471,
+      "step": 7661
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9029149543249022,
+      "learning_rate": 1.8835391571975475e-05,
+      "loss": 0.7865,
+      "step": 7662
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9224531286180437,
+      "learning_rate": 1.8835034170406046e-05,
+      "loss": 0.7765,
+      "step": 7663
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.28036742790362,
+      "learning_rate": 1.883467671739625e-05,
+      "loss": 0.7479,
+      "step": 7664
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.797821138029405,
+      "learning_rate": 1.8834319212948168e-05,
+      "loss": 0.9695,
+      "step": 7665
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7526124576108912,
+      "learning_rate": 1.8833961657063887e-05,
+      "loss": 0.6381,
+      "step": 7666
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6161113510026586,
+      "learning_rate": 1.883360404974548e-05,
+      "loss": 0.5378,
+      "step": 7667
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7202525241400655,
+      "learning_rate": 1.8833246390995034e-05,
+      "loss": 0.7779,
+      "step": 7668
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3923845623191209,
+      "learning_rate": 1.883288868081463e-05,
+      "loss": 0.8283,
+      "step": 7669
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5513712306357702,
+      "learning_rate": 1.8832530919206356e-05,
+      "loss": 0.5189,
+      "step": 7670
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8946921924877829,
+      "learning_rate": 1.883217310617229e-05,
+      "loss": 0.9022,
+      "step": 7671
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.576541340850796,
+      "learning_rate": 1.883181524171451e-05,
+      "loss": 0.7771,
+      "step": 7672
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.16199886967401,
+      "learning_rate": 1.8831457325835113e-05,
+      "loss": 0.7618,
+      "step": 7673
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5274121268397969,
+      "learning_rate": 1.8831099358536172e-05,
+      "loss": 0.7418,
+      "step": 7674
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5156377823676377,
+      "learning_rate": 1.8830741339819774e-05,
+      "loss": 0.7217,
+      "step": 7675
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.592320728246458,
+      "learning_rate": 1.8830383269688005e-05,
+      "loss": 0.697,
+      "step": 7676
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5996176503658632,
+      "learning_rate": 1.8830025148142948e-05,
+      "loss": 0.5385,
+      "step": 7677
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0701099350565735,
+      "learning_rate": 1.882966697518669e-05,
+      "loss": 0.8822,
+      "step": 7678
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4722165325483134,
+      "learning_rate": 1.882930875082132e-05,
+      "loss": 0.8332,
+      "step": 7679
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6185274524268667,
+      "learning_rate": 1.8828950475048913e-05,
+      "loss": 0.5503,
+      "step": 7680
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.921473789150788,
+      "learning_rate": 1.8828592147871562e-05,
+      "loss": 0.815,
+      "step": 7681
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3381809088075176,
+      "learning_rate": 1.8828233769291353e-05,
+      "loss": 0.7755,
+      "step": 7682
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7769017960568618,
+      "learning_rate": 1.8827875339310372e-05,
+      "loss": 0.7521,
+      "step": 7683
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7039411741556243,
+      "learning_rate": 1.8827516857930706e-05,
+      "loss": 0.7394,
+      "step": 7684
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6222208546998882,
+      "learning_rate": 1.882715832515444e-05,
+      "loss": 0.7919,
+      "step": 7685
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.470003056676991,
+      "learning_rate": 1.8826799740983665e-05,
+      "loss": 0.7748,
+      "step": 7686
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5462248381300565,
+      "learning_rate": 1.882644110542047e-05,
+      "loss": 0.7159,
+      "step": 7687
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6572707934032165,
+      "learning_rate": 1.8826082418466938e-05,
+      "loss": 0.7666,
+      "step": 7688
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5205524472348346,
+      "learning_rate": 1.882572368012516e-05,
+      "loss": 0.8202,
+      "step": 7689
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4822269047454089,
+      "learning_rate": 1.8825364890397225e-05,
+      "loss": 0.8179,
+      "step": 7690
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6418350517047361,
+      "learning_rate": 1.8825006049285225e-05,
+      "loss": 0.5827,
+      "step": 7691
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3573586297395666,
+      "learning_rate": 1.882464715679124e-05,
+      "loss": 0.7342,
+      "step": 7692
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.280750974598451,
+      "learning_rate": 1.8824288212917367e-05,
+      "loss": 0.7281,
+      "step": 7693
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.4945266321963406,
+      "learning_rate": 1.8823929217665696e-05,
+      "loss": 0.8638,
+      "step": 7694
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7908363316485851,
+      "learning_rate": 1.8823570171038316e-05,
+      "loss": 0.6925,
+      "step": 7695
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4099919733634219,
+      "learning_rate": 1.8823211073037316e-05,
+      "loss": 0.8007,
+      "step": 7696
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6950972510273457,
+      "learning_rate": 1.8822851923664785e-05,
+      "loss": 0.8742,
+      "step": 7697
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6018123965200898,
+      "learning_rate": 1.882249272292282e-05,
+      "loss": 0.5072,
+      "step": 7698
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.489427225177923,
+      "learning_rate": 1.8822133470813507e-05,
+      "loss": 0.6968,
+      "step": 7699
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0974689011965872,
+      "learning_rate": 1.8821774167338944e-05,
+      "loss": 0.7149,
+      "step": 7700
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3357839488157564,
+      "learning_rate": 1.8821414812501214e-05,
+      "loss": 0.8239,
+      "step": 7701
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.718896891275068,
+      "learning_rate": 1.8821055406302418e-05,
+      "loss": 0.813,
+      "step": 7702
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5816922081323417,
+      "learning_rate": 1.882069594874464e-05,
+      "loss": 0.8224,
+      "step": 7703
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.931684913259524,
+      "learning_rate": 1.882033643982998e-05,
+      "loss": 0.8278,
+      "step": 7704
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.549475350469368,
+      "learning_rate": 1.8819976879560534e-05,
+      "loss": 0.9529,
+      "step": 7705
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5960410116683859,
+      "learning_rate": 1.8819617267938383e-05,
+      "loss": 0.8181,
+      "step": 7706
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.711809410120716,
+      "learning_rate": 1.881925760496563e-05,
+      "loss": 0.7653,
+      "step": 7707
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5467972787929363,
+      "learning_rate": 1.881889789064437e-05,
+      "loss": 0.6805,
+      "step": 7708
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3940252976353924,
+      "learning_rate": 1.881853812497669e-05,
+      "loss": 0.8642,
+      "step": 7709
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1684874998204013,
+      "learning_rate": 1.881817830796469e-05,
+      "loss": 0.6873,
+      "step": 7710
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.278040981675039,
+      "learning_rate": 1.8817818439610466e-05,
+      "loss": 0.7037,
+      "step": 7711
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1659437847461853,
+      "learning_rate": 1.8817458519916113e-05,
+      "loss": 0.7708,
+      "step": 7712
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4033250245980475,
+      "learning_rate": 1.881709854888372e-05,
+      "loss": 0.7103,
+      "step": 7713
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8248274675675564,
+      "learning_rate": 1.881673852651539e-05,
+      "loss": 0.7914,
+      "step": 7714
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4671617323624555,
+      "learning_rate": 1.881637845281322e-05,
+      "loss": 0.7964,
+      "step": 7715
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5978262414732523,
+      "learning_rate": 1.88160183277793e-05,
+      "loss": 0.7299,
+      "step": 7716
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3092864896286696,
+      "learning_rate": 1.8815658151415734e-05,
+      "loss": 0.7843,
+      "step": 7717
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6079978864093365,
+      "learning_rate": 1.8815297923724612e-05,
+      "loss": 0.7786,
+      "step": 7718
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4382140069167761,
+      "learning_rate": 1.8814937644708035e-05,
+      "loss": 0.7504,
+      "step": 7719
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.232252298162136,
+      "learning_rate": 1.8814577314368098e-05,
+      "loss": 0.6921,
+      "step": 7720
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.874638057692578,
+      "learning_rate": 1.8814216932706905e-05,
+      "loss": 0.7718,
+      "step": 7721
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5804324711112432,
+      "learning_rate": 1.8813856499726548e-05,
+      "loss": 0.6653,
+      "step": 7722
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.270122831499259,
+      "learning_rate": 1.881349601542913e-05,
+      "loss": 0.78,
+      "step": 7723
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6125624171596966,
+      "learning_rate": 1.8813135479816748e-05,
+      "loss": 0.7921,
+      "step": 7724
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.637379480439171,
+      "learning_rate": 1.88127748928915e-05,
+      "loss": 0.7347,
+      "step": 7725
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7037218286041564,
+      "learning_rate": 1.8812414254655483e-05,
+      "loss": 0.6861,
+      "step": 7726
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1984393272189178,
+      "learning_rate": 1.8812053565110804e-05,
+      "loss": 0.6462,
+      "step": 7727
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7790866712615967,
+      "learning_rate": 1.881169282425956e-05,
+      "loss": 0.5213,
+      "step": 7728
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.462558317205404,
+      "learning_rate": 1.881133203210385e-05,
+      "loss": 0.6954,
+      "step": 7729
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7683137514426233,
+      "learning_rate": 1.8810971188645775e-05,
+      "loss": 0.8279,
+      "step": 7730
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6082853836823836,
+      "learning_rate": 1.8810610293887435e-05,
+      "loss": 0.8456,
+      "step": 7731
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8165328319791665,
+      "learning_rate": 1.8810249347830935e-05,
+      "loss": 0.7151,
+      "step": 7732
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8448178381133298,
+      "learning_rate": 1.8809888350478372e-05,
+      "loss": 0.7089,
+      "step": 7733
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.59340399097738,
+      "learning_rate": 1.880952730183185e-05,
+      "loss": 0.6556,
+      "step": 7734
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6024402398100337,
+      "learning_rate": 1.880916620189347e-05,
+      "loss": 0.8365,
+      "step": 7735
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.576918054506426,
+      "learning_rate": 1.880880505066534e-05,
+      "loss": 0.7748,
+      "step": 7736
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6183044610313204,
+      "learning_rate": 1.8808443848149556e-05,
+      "loss": 0.8392,
+      "step": 7737
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4531100081027255,
+      "learning_rate": 1.880808259434822e-05,
+      "loss": 0.9167,
+      "step": 7738
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4990805740060351,
+      "learning_rate": 1.8807721289263442e-05,
+      "loss": 0.6893,
+      "step": 7739
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7268681364473768,
+      "learning_rate": 1.880735993289732e-05,
+      "loss": 0.7548,
+      "step": 7740
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4683158991554115,
+      "learning_rate": 1.8806998525251964e-05,
+      "loss": 0.8578,
+      "step": 7741
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8926154356866447,
+      "learning_rate": 1.880663706632947e-05,
+      "loss": 0.6906,
+      "step": 7742
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.042320257129368,
+      "learning_rate": 1.8806275556131948e-05,
+      "loss": 0.9403,
+      "step": 7743
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4824458565660064,
+      "learning_rate": 1.88059139946615e-05,
+      "loss": 0.5968,
+      "step": 7744
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6996141824509636,
+      "learning_rate": 1.8805552381920238e-05,
+      "loss": 0.7232,
+      "step": 7745
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4679085215033312,
+      "learning_rate": 1.880519071791026e-05,
+      "loss": 0.7511,
+      "step": 7746
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2211081398831514,
+      "learning_rate": 1.8804829002633672e-05,
+      "loss": 0.8583,
+      "step": 7747
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6433174794534986,
+      "learning_rate": 1.8804467236092584e-05,
+      "loss": 0.7044,
+      "step": 7748
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8488633807205623,
+      "learning_rate": 1.8804105418289097e-05,
+      "loss": 0.7625,
+      "step": 7749
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5642458354451776,
+      "learning_rate": 1.8803743549225323e-05,
+      "loss": 0.7578,
+      "step": 7750
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4471478461049492,
+      "learning_rate": 1.880338162890337e-05,
+      "loss": 0.7968,
+      "step": 7751
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5975674794302147,
+      "learning_rate": 1.8803019657325338e-05,
+      "loss": 0.572,
+      "step": 7752
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3891699818261987,
+      "learning_rate": 1.880265763449334e-05,
+      "loss": 0.9031,
+      "step": 7753
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7702667563695238,
+      "learning_rate": 1.880229556040948e-05,
+      "loss": 0.8053,
+      "step": 7754
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6396618661831337,
+      "learning_rate": 1.8801933435075872e-05,
+      "loss": 0.5687,
+      "step": 7755
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7720723413521278,
+      "learning_rate": 1.880157125849462e-05,
+      "loss": 0.8783,
+      "step": 7756
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5715353808036644,
+      "learning_rate": 1.8801209030667834e-05,
+      "loss": 0.5318,
+      "step": 7757
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5518488369916223,
+      "learning_rate": 1.880084675159762e-05,
+      "loss": 0.8461,
+      "step": 7758
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8183985800766453,
+      "learning_rate": 1.8800484421286094e-05,
+      "loss": 0.904,
+      "step": 7759
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4718299489130553,
+      "learning_rate": 1.880012203973536e-05,
+      "loss": 0.8447,
+      "step": 7760
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8464309428340593,
+      "learning_rate": 1.8799759606947527e-05,
+      "loss": 0.8158,
+      "step": 7761
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4745529255479308,
+      "learning_rate": 1.8799397122924708e-05,
+      "loss": 0.794,
+      "step": 7762
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3797393043664639,
+      "learning_rate": 1.8799034587669016e-05,
+      "loss": 0.857,
+      "step": 7763
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6370132427483901,
+      "learning_rate": 1.8798672001182553e-05,
+      "loss": 0.5686,
+      "step": 7764
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5237899217800674,
+      "learning_rate": 1.8798309363467446e-05,
+      "loss": 0.7595,
+      "step": 7765
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3554940598690963,
+      "learning_rate": 1.8797946674525788e-05,
+      "loss": 0.7843,
+      "step": 7766
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4374318868285378,
+      "learning_rate": 1.87975839343597e-05,
+      "loss": 0.7654,
+      "step": 7767
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5917310637490134,
+      "learning_rate": 1.8797221142971295e-05,
+      "loss": 0.7982,
+      "step": 7768
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7428039901996952,
+      "learning_rate": 1.8796858300362682e-05,
+      "loss": 0.7341,
+      "step": 7769
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6074771619742427,
+      "learning_rate": 1.879649540653598e-05,
+      "loss": 0.7087,
+      "step": 7770
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5490107884750808,
+      "learning_rate": 1.879613246149329e-05,
+      "loss": 0.8356,
+      "step": 7771
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1463759198123333,
+      "learning_rate": 1.8795769465236735e-05,
+      "loss": 0.726,
+      "step": 7772
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5516854694646756,
+      "learning_rate": 1.8795406417768426e-05,
+      "loss": 0.8981,
+      "step": 7773
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6133925273405971,
+      "learning_rate": 1.8795043319090477e-05,
+      "loss": 0.5849,
+      "step": 7774
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4532720882869004,
+      "learning_rate": 1.8794680169205e-05,
+      "loss": 0.7113,
+      "step": 7775
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.456652552460397,
+      "learning_rate": 1.879431696811411e-05,
+      "loss": 0.6832,
+      "step": 7776
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3342223149497567,
+      "learning_rate": 1.8793953715819922e-05,
+      "loss": 0.6811,
+      "step": 7777
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9005928631820403,
+      "learning_rate": 1.8793590412324555e-05,
+      "loss": 0.7645,
+      "step": 7778
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.742420252350809,
+      "learning_rate": 1.8793227057630115e-05,
+      "loss": 0.6822,
+      "step": 7779
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6004518175206148,
+      "learning_rate": 1.8792863651738727e-05,
+      "loss": 0.8546,
+      "step": 7780
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.729989201893514,
+      "learning_rate": 1.8792500194652505e-05,
+      "loss": 0.7087,
+      "step": 7781
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6816062616539287,
+      "learning_rate": 1.879213668637356e-05,
+      "loss": 0.8109,
+      "step": 7782
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8020371303592746,
+      "learning_rate": 1.879177312690401e-05,
+      "loss": 0.6955,
+      "step": 7783
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4131189536696722,
+      "learning_rate": 1.8791409516245974e-05,
+      "loss": 0.8995,
+      "step": 7784
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7345568078114513,
+      "learning_rate": 1.879104585440157e-05,
+      "loss": 0.6565,
+      "step": 7785
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.609938416385106,
+      "learning_rate": 1.8790682141372913e-05,
+      "loss": 0.8077,
+      "step": 7786
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5121801166553472,
+      "learning_rate": 1.8790318377162126e-05,
+      "loss": 0.6896,
+      "step": 7787
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7917241995793323,
+      "learning_rate": 1.8789954561771314e-05,
+      "loss": 0.6564,
+      "step": 7788
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4519053747093904,
+      "learning_rate": 1.878959069520261e-05,
+      "loss": 0.7087,
+      "step": 7789
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.422669692186041,
+      "learning_rate": 1.8789226777458126e-05,
+      "loss": 0.7245,
+      "step": 7790
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.28486971459059,
+      "learning_rate": 1.878886280853998e-05,
+      "loss": 0.7059,
+      "step": 7791
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.594734104796609,
+      "learning_rate": 1.878849878845029e-05,
+      "loss": 0.7506,
+      "step": 7792
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4400410386872866,
+      "learning_rate": 1.8788134717191182e-05,
+      "loss": 0.7744,
+      "step": 7793
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5015567947353674,
+      "learning_rate": 1.8787770594764768e-05,
+      "loss": 0.7298,
+      "step": 7794
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5424243947885565,
+      "learning_rate": 1.8787406421173173e-05,
+      "loss": 0.7838,
+      "step": 7795
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7637598494348135,
+      "learning_rate": 1.8787042196418514e-05,
+      "loss": 0.8038,
+      "step": 7796
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.598690631000168,
+      "learning_rate": 1.8786677920502913e-05,
+      "loss": 0.8406,
+      "step": 7797
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.322306149022338,
+      "learning_rate": 1.8786313593428493e-05,
+      "loss": 0.7725,
+      "step": 7798
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5471979308047321,
+      "learning_rate": 1.8785949215197377e-05,
+      "loss": 0.8668,
+      "step": 7799
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6329011862371394,
+      "learning_rate": 1.8785584785811677e-05,
+      "loss": 0.6987,
+      "step": 7800
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.558577326284949,
+      "learning_rate": 1.8785220305273526e-05,
+      "loss": 0.8178,
+      "step": 7801
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4856190099549225,
+      "learning_rate": 1.8784855773585037e-05,
+      "loss": 0.8066,
+      "step": 7802
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6368525424491736,
+      "learning_rate": 1.8784491190748338e-05,
+      "loss": 0.6762,
+      "step": 7803
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8032926663323723,
+      "learning_rate": 1.878412655676555e-05,
+      "loss": 0.6878,
+      "step": 7804
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.368607593948879,
+      "learning_rate": 1.8783761871638798e-05,
+      "loss": 0.8535,
+      "step": 7805
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6238071985633584,
+      "learning_rate": 1.87833971353702e-05,
+      "loss": 0.7594,
+      "step": 7806
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5869885211108602,
+      "learning_rate": 1.8783032347961884e-05,
+      "loss": 0.5067,
+      "step": 7807
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6669738918289594,
+      "learning_rate": 1.8782667509415974e-05,
+      "loss": 0.824,
+      "step": 7808
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.623097090006738,
+      "learning_rate": 1.8782302619734594e-05,
+      "loss": 0.7436,
+      "step": 7809
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3854450350337129,
+      "learning_rate": 1.8781937678919867e-05,
+      "loss": 0.7945,
+      "step": 7810
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.28226563219793,
+      "learning_rate": 1.878157268697392e-05,
+      "loss": 0.7789,
+      "step": 7811
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5865977930923764,
+      "learning_rate": 1.8781207643898873e-05,
+      "loss": 0.5623,
+      "step": 7812
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9148172259962053,
+      "learning_rate": 1.8780842549696857e-05,
+      "loss": 0.6615,
+      "step": 7813
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.251206148782049,
+      "learning_rate": 1.8780477404369996e-05,
+      "loss": 0.6602,
+      "step": 7814
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7360840563599078,
+      "learning_rate": 1.8780112207920415e-05,
+      "loss": 0.7869,
+      "step": 7815
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5993186534811037,
+      "learning_rate": 1.8779746960350243e-05,
+      "loss": 0.5681,
+      "step": 7816
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4825415407674047,
+      "learning_rate": 1.8779381661661603e-05,
+      "loss": 0.8464,
+      "step": 7817
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7741050379282184,
+      "learning_rate": 1.877901631185662e-05,
+      "loss": 0.7509,
+      "step": 7818
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9802941216916243,
+      "learning_rate": 1.877865091093743e-05,
+      "loss": 0.9335,
+      "step": 7819
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.454502009792444,
+      "learning_rate": 1.8778285458906152e-05,
+      "loss": 0.7797,
+      "step": 7820
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4672088779356034,
+      "learning_rate": 1.8777919955764917e-05,
+      "loss": 0.7975,
+      "step": 7821
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5801788298888981,
+      "learning_rate": 1.8777554401515855e-05,
+      "loss": 0.7527,
+      "step": 7822
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.457461671942781,
+      "learning_rate": 1.877718879616109e-05,
+      "loss": 0.7195,
+      "step": 7823
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.903434889806231,
+      "learning_rate": 1.877682313970276e-05,
+      "loss": 0.8221,
+      "step": 7824
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5084716787120163,
+      "learning_rate": 1.877645743214298e-05,
+      "loss": 0.7927,
+      "step": 7825
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.631389711093266,
+      "learning_rate": 1.8776091673483886e-05,
+      "loss": 0.8546,
+      "step": 7826
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9881425992874278,
+      "learning_rate": 1.877572586372761e-05,
+      "loss": 0.7437,
+      "step": 7827
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7991368424304732,
+      "learning_rate": 1.877536000287628e-05,
+      "loss": 0.7212,
+      "step": 7828
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.759510848846331,
+      "learning_rate": 1.8774994090932025e-05,
+      "loss": 0.8168,
+      "step": 7829
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5168615360813615,
+      "learning_rate": 1.8774628127896976e-05,
+      "loss": 0.6987,
+      "step": 7830
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5156962736025361,
+      "learning_rate": 1.8774262113773262e-05,
+      "loss": 0.6386,
+      "step": 7831
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4546221460175197,
+      "learning_rate": 1.877389604856302e-05,
+      "loss": 0.7624,
+      "step": 7832
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.491042883650415,
+      "learning_rate": 1.8773529932268374e-05,
+      "loss": 0.742,
+      "step": 7833
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5871708929040604,
+      "learning_rate": 1.8773163764891464e-05,
+      "loss": 0.8475,
+      "step": 7834
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6854612253955275,
+      "learning_rate": 1.8772797546434415e-05,
+      "loss": 0.7675,
+      "step": 7835
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8246924604001795,
+      "learning_rate": 1.8772431276899358e-05,
+      "loss": 0.6953,
+      "step": 7836
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.738034882073838,
+      "learning_rate": 1.877206495628843e-05,
+      "loss": 0.8445,
+      "step": 7837
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8061465993774664,
+      "learning_rate": 1.8771698584603766e-05,
+      "loss": 0.8124,
+      "step": 7838
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4614266661785644,
+      "learning_rate": 1.8771332161847493e-05,
+      "loss": 0.7377,
+      "step": 7839
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5977968281384662,
+      "learning_rate": 1.8770965688021747e-05,
+      "loss": 0.5201,
+      "step": 7840
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.690625111311808,
+      "learning_rate": 1.8770599163128663e-05,
+      "loss": 0.7634,
+      "step": 7841
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.946451214428192,
+      "learning_rate": 1.8770232587170374e-05,
+      "loss": 0.8352,
+      "step": 7842
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4925703880552674,
+      "learning_rate": 1.8769865960149012e-05,
+      "loss": 0.7346,
+      "step": 7843
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4036015076066508,
+      "learning_rate": 1.8769499282066716e-05,
+      "loss": 0.7872,
+      "step": 7844
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1690707285327817,
+      "learning_rate": 1.876913255292562e-05,
+      "loss": 0.7187,
+      "step": 7845
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6630682947903137,
+      "learning_rate": 1.8768765772727856e-05,
+      "loss": 0.8318,
+      "step": 7846
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6120719630512952,
+      "learning_rate": 1.876839894147556e-05,
+      "loss": 0.7205,
+      "step": 7847
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3796128559874152,
+      "learning_rate": 1.8768032059170874e-05,
+      "loss": 0.8351,
+      "step": 7848
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4891166222002554,
+      "learning_rate": 1.8767665125815927e-05,
+      "loss": 0.8531,
+      "step": 7849
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.538825957931869,
+      "learning_rate": 1.876729814141286e-05,
+      "loss": 0.7228,
+      "step": 7850
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6363597984608256,
+      "learning_rate": 1.87669311059638e-05,
+      "loss": 0.7347,
+      "step": 7851
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.649447571774451,
+      "learning_rate": 1.8766564019470898e-05,
+      "loss": 0.9319,
+      "step": 7852
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.936432759352666,
+      "learning_rate": 1.8766196881936284e-05,
+      "loss": 0.6312,
+      "step": 7853
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6559935964494175,
+      "learning_rate": 1.8765829693362097e-05,
+      "loss": 0.6325,
+      "step": 7854
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.322051226460985,
+      "learning_rate": 1.8765462453750472e-05,
+      "loss": 0.8173,
+      "step": 7855
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6141600634850626,
+      "learning_rate": 1.876509516310355e-05,
+      "loss": 0.7271,
+      "step": 7856
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.592799819210162,
+      "learning_rate": 1.8764727821423473e-05,
+      "loss": 0.5416,
+      "step": 7857
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5461275063587567,
+      "learning_rate": 1.876436042871237e-05,
+      "loss": 0.7576,
+      "step": 7858
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7170638329585726,
+      "learning_rate": 1.8763992984972388e-05,
+      "loss": 0.6925,
+      "step": 7859
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5263314168672435,
+      "learning_rate": 1.8763625490205664e-05,
+      "loss": 0.631,
+      "step": 7860
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5878544344508079,
+      "learning_rate": 1.876325794441434e-05,
+      "loss": 0.7474,
+      "step": 7861
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4818468965232878,
+      "learning_rate": 1.876289034760055e-05,
+      "loss": 0.7566,
+      "step": 7862
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6211621012356898,
+      "learning_rate": 1.8762522699766443e-05,
+      "loss": 0.7034,
+      "step": 7863
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5405769841165806,
+      "learning_rate": 1.876215500091415e-05,
+      "loss": 0.7249,
+      "step": 7864
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5649734562188685,
+      "learning_rate": 1.876178725104582e-05,
+      "loss": 0.5362,
+      "step": 7865
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6184948484898328,
+      "learning_rate": 1.8761419450163586e-05,
+      "loss": 0.5636,
+      "step": 7866
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3732039383090198,
+      "learning_rate": 1.87610515982696e-05,
+      "loss": 0.7167,
+      "step": 7867
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5617350318794883,
+      "learning_rate": 1.8760683695365994e-05,
+      "loss": 0.6534,
+      "step": 7868
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4735826939373318,
+      "learning_rate": 1.8760315741454917e-05,
+      "loss": 0.7927,
+      "step": 7869
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5185486198212843,
+      "learning_rate": 1.8759947736538507e-05,
+      "loss": 0.7123,
+      "step": 7870
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.5530701067186627,
+      "learning_rate": 1.8759579680618907e-05,
+      "loss": 0.8915,
+      "step": 7871
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5271498037321443,
+      "learning_rate": 1.8759211573698263e-05,
+      "loss": 0.8432,
+      "step": 7872
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7297511252744209,
+      "learning_rate": 1.875884341577871e-05,
+      "loss": 0.8185,
+      "step": 7873
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8614577085548627,
+      "learning_rate": 1.87584752068624e-05,
+      "loss": 0.7931,
+      "step": 7874
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3554534855316844,
+      "learning_rate": 1.8758106946951484e-05,
+      "loss": 0.8349,
+      "step": 7875
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.435304386981007,
+      "learning_rate": 1.8757738636048088e-05,
+      "loss": 0.8053,
+      "step": 7876
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4851913903640594,
+      "learning_rate": 1.8757370274154362e-05,
+      "loss": 0.7256,
+      "step": 7877
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6495265890636654,
+      "learning_rate": 1.875700186127246e-05,
+      "loss": 0.8028,
+      "step": 7878
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9223563200674039,
+      "learning_rate": 1.8756633397404518e-05,
+      "loss": 0.7469,
+      "step": 7879
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.539991645141211,
+      "learning_rate": 1.8756264882552685e-05,
+      "loss": 0.7192,
+      "step": 7880
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7379351570145665,
+      "learning_rate": 1.8755896316719104e-05,
+      "loss": 0.7827,
+      "step": 7881
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6955043389010203,
+      "learning_rate": 1.8755527699905927e-05,
+      "loss": 0.7946,
+      "step": 7882
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6951042521162106,
+      "learning_rate": 1.8755159032115292e-05,
+      "loss": 0.5425,
+      "step": 7883
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5718813015706798,
+      "learning_rate": 1.8754790313349347e-05,
+      "loss": 0.6803,
+      "step": 7884
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5160392944067282,
+      "learning_rate": 1.8754421543610247e-05,
+      "loss": 0.7143,
+      "step": 7885
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6922366691092128,
+      "learning_rate": 1.875405272290013e-05,
+      "loss": 0.7225,
+      "step": 7886
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5769860676211527,
+      "learning_rate": 1.8753683851221146e-05,
+      "loss": 0.72,
+      "step": 7887
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4124707388471205,
+      "learning_rate": 1.8753314928575444e-05,
+      "loss": 0.6684,
+      "step": 7888
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.610804021095675,
+      "learning_rate": 1.875294595496517e-05,
+      "loss": 0.7908,
+      "step": 7889
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6257591945203127,
+      "learning_rate": 1.8752576930392473e-05,
+      "loss": 0.4991,
+      "step": 7890
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6287471773398343,
+      "learning_rate": 1.8752207854859507e-05,
+      "loss": 0.8001,
+      "step": 7891
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.614687318210899,
+      "learning_rate": 1.875183872836841e-05,
+      "loss": 0.7363,
+      "step": 7892
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5626691482675155,
+      "learning_rate": 1.8751469550921342e-05,
+      "loss": 0.7452,
+      "step": 7893
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.624059750247173,
+      "learning_rate": 1.8751100322520447e-05,
+      "loss": 0.7476,
+      "step": 7894
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.618769899861393,
+      "learning_rate": 1.8750731043167873e-05,
+      "loss": 0.5628,
+      "step": 7895
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7125136460028194,
+      "learning_rate": 1.8750361712865774e-05,
+      "loss": 0.7266,
+      "step": 7896
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6330947743564492,
+      "learning_rate": 1.8749992331616296e-05,
+      "loss": 0.7429,
+      "step": 7897
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6333459703273268,
+      "learning_rate": 1.8749622899421597e-05,
+      "loss": 0.6834,
+      "step": 7898
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8124605713153934,
+      "learning_rate": 1.874925341628382e-05,
+      "loss": 0.7966,
+      "step": 7899
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6445425807897796,
+      "learning_rate": 1.874888388220512e-05,
+      "loss": 0.7578,
+      "step": 7900
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7586508316958511,
+      "learning_rate": 1.874851429718765e-05,
+      "loss": 0.8436,
+      "step": 7901
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6206249535780117,
+      "learning_rate": 1.8748144661233558e-05,
+      "loss": 0.6588,
+      "step": 7902
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.533823863921037,
+      "learning_rate": 1.8747774974345e-05,
+      "loss": 0.7072,
+      "step": 7903
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4608551666328093,
+      "learning_rate": 1.8747405236524125e-05,
+      "loss": 0.9057,
+      "step": 7904
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6440961443765636,
+      "learning_rate": 1.8747035447773086e-05,
+      "loss": 0.636,
+      "step": 7905
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4834207415453913,
+      "learning_rate": 1.874666560809404e-05,
+      "loss": 0.8821,
+      "step": 7906
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7687489932669958,
+      "learning_rate": 1.8746295717489138e-05,
+      "loss": 0.6814,
+      "step": 7907
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.359704593575843,
+      "learning_rate": 1.874592577596053e-05,
+      "loss": 0.7554,
+      "step": 7908
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9159608649260182,
+      "learning_rate": 1.8745555783510378e-05,
+      "loss": 0.7676,
+      "step": 7909
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9292843069618921,
+      "learning_rate": 1.874518574014083e-05,
+      "loss": 0.7784,
+      "step": 7910
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4046838521520515,
+      "learning_rate": 1.874481564585404e-05,
+      "loss": 0.6936,
+      "step": 7911
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5176111453071233,
+      "learning_rate": 1.8744445500652165e-05,
+      "loss": 0.7779,
+      "step": 7912
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.637871025714501,
+      "learning_rate": 1.8744075304537358e-05,
+      "loss": 0.7602,
+      "step": 7913
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3368869260780516,
+      "learning_rate": 1.8743705057511777e-05,
+      "loss": 0.7891,
+      "step": 7914
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7422627089223404,
+      "learning_rate": 1.8743334759577582e-05,
+      "loss": 0.7581,
+      "step": 7915
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4221584543434302,
+      "learning_rate": 1.874296441073692e-05,
+      "loss": 0.7202,
+      "step": 7916
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4154365521289904,
+      "learning_rate": 1.8742594010991953e-05,
+      "loss": 0.7867,
+      "step": 7917
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6021224982114726,
+      "learning_rate": 1.8742223560344835e-05,
+      "loss": 0.8389,
+      "step": 7918
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8524962873156863,
+      "learning_rate": 1.874185305879772e-05,
+      "loss": 0.7967,
+      "step": 7919
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5927231041748682,
+      "learning_rate": 1.8741482506352776e-05,
+      "loss": 0.5666,
+      "step": 7920
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5540769955844216,
+      "learning_rate": 1.8741111903012148e-05,
+      "loss": 0.7664,
+      "step": 7921
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.601351010682805,
+      "learning_rate": 1.8740741248778002e-05,
+      "loss": 0.6868,
+      "step": 7922
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8178921817365983,
+      "learning_rate": 1.8740370543652487e-05,
+      "loss": 0.6253,
+      "step": 7923
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5754624252625757,
+      "learning_rate": 1.8739999787637775e-05,
+      "loss": 0.5233,
+      "step": 7924
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2798084562349037,
+      "learning_rate": 1.8739628980736016e-05,
+      "loss": 0.8039,
+      "step": 7925
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8087361985981354,
+      "learning_rate": 1.8739258122949367e-05,
+      "loss": 0.7576,
+      "step": 7926
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4212039601426114,
+      "learning_rate": 1.873888721427999e-05,
+      "loss": 0.7931,
+      "step": 7927
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3184440347578164,
+      "learning_rate": 1.8738516254730052e-05,
+      "loss": 0.6325,
+      "step": 7928
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6207988421395543,
+      "learning_rate": 1.87381452443017e-05,
+      "loss": 0.7429,
+      "step": 7929
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4957400414797295,
+      "learning_rate": 1.87377741829971e-05,
+      "loss": 0.9009,
+      "step": 7930
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4348131020147548,
+      "learning_rate": 1.873740307081841e-05,
+      "loss": 0.7736,
+      "step": 7931
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.346563586213906,
+      "learning_rate": 1.8737031907767797e-05,
+      "loss": 0.8395,
+      "step": 7932
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4267830881438563,
+      "learning_rate": 1.873666069384742e-05,
+      "loss": 0.7671,
+      "step": 7933
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4521273448784058,
+      "learning_rate": 1.8736289429059436e-05,
+      "loss": 0.8214,
+      "step": 7934
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6257245313958636,
+      "learning_rate": 1.873591811340601e-05,
+      "loss": 0.5502,
+      "step": 7935
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7774412904311077,
+      "learning_rate": 1.87355467468893e-05,
+      "loss": 0.7849,
+      "step": 7936
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1910511857436292,
+      "learning_rate": 1.873517532951147e-05,
+      "loss": 0.6966,
+      "step": 7937
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.417157560384858,
+      "learning_rate": 1.8734803861274692e-05,
+      "loss": 0.7637,
+      "step": 7938
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3550368800642167,
+      "learning_rate": 1.8734432342181113e-05,
+      "loss": 0.6979,
+      "step": 7939
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4837176249873985,
+      "learning_rate": 1.8734060772232906e-05,
+      "loss": 0.7658,
+      "step": 7940
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5296035444998894,
+      "learning_rate": 1.873368915143223e-05,
+      "loss": 0.8056,
+      "step": 7941
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5347633843051804,
+      "learning_rate": 1.8733317479781256e-05,
+      "loss": 0.8296,
+      "step": 7942
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2918536466120603,
+      "learning_rate": 1.8732945757282137e-05,
+      "loss": 0.771,
+      "step": 7943
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.508629738588247,
+      "learning_rate": 1.8732573983937046e-05,
+      "loss": 0.8107,
+      "step": 7944
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2673868922274507,
+      "learning_rate": 1.8732202159748146e-05,
+      "loss": 0.7584,
+      "step": 7945
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5608402452550845,
+      "learning_rate": 1.8731830284717597e-05,
+      "loss": 0.7235,
+      "step": 7946
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5444746206763025,
+      "learning_rate": 1.873145835884757e-05,
+      "loss": 0.8958,
+      "step": 7947
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6011707994702096,
+      "learning_rate": 1.8731086382140228e-05,
+      "loss": 0.5436,
+      "step": 7948
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4747946036486568,
+      "learning_rate": 1.8730714354597738e-05,
+      "loss": 0.8525,
+      "step": 7949
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7098585092200425,
+      "learning_rate": 1.873034227622226e-05,
+      "loss": 0.7486,
+      "step": 7950
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7365531663080327,
+      "learning_rate": 1.8729970147015968e-05,
+      "loss": 0.7217,
+      "step": 7951
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5854068623012445,
+      "learning_rate": 1.8729597966981025e-05,
+      "loss": 0.6391,
+      "step": 7952
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4984182567565485,
+      "learning_rate": 1.8729225736119602e-05,
+      "loss": 0.7862,
+      "step": 7953
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7286577443534799,
+      "learning_rate": 1.872885345443386e-05,
+      "loss": 0.7478,
+      "step": 7954
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6163649142537186,
+      "learning_rate": 1.872848112192597e-05,
+      "loss": 0.5341,
+      "step": 7955
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4436321840051642,
+      "learning_rate": 1.8728108738598103e-05,
+      "loss": 0.7092,
+      "step": 7956
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7431067418660249,
+      "learning_rate": 1.872773630445242e-05,
+      "loss": 0.7993,
+      "step": 7957
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3874410342657166,
+      "learning_rate": 1.8727363819491097e-05,
+      "loss": 0.7458,
+      "step": 7958
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.512572297440838,
+      "learning_rate": 1.8726991283716295e-05,
+      "loss": 0.6969,
+      "step": 7959
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5737946182389027,
+      "learning_rate": 1.8726618697130186e-05,
+      "loss": 0.4724,
+      "step": 7960
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6901690758953376,
+      "learning_rate": 1.872624605973494e-05,
+      "loss": 0.6979,
+      "step": 7961
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.309216124172628,
+      "learning_rate": 1.8725873371532733e-05,
+      "loss": 0.7387,
+      "step": 7962
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8342268611808341,
+      "learning_rate": 1.872550063252572e-05,
+      "loss": 0.6883,
+      "step": 7963
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.6073250670222525,
+      "learning_rate": 1.8725127842716086e-05,
+      "loss": 0.8404,
+      "step": 7964
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8048099967567732,
+      "learning_rate": 1.8724755002105996e-05,
+      "loss": 0.7507,
+      "step": 7965
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4941544627580443,
+      "learning_rate": 1.8724382110697615e-05,
+      "loss": 0.8546,
+      "step": 7966
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5357502456785133,
+      "learning_rate": 1.8724009168493126e-05,
+      "loss": 0.7265,
+      "step": 7967
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5658878444363278,
+      "learning_rate": 1.872363617549469e-05,
+      "loss": 0.7895,
+      "step": 7968
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5753416794569236,
+      "learning_rate": 1.872326313170448e-05,
+      "loss": 0.537,
+      "step": 7969
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7750876200947026,
+      "learning_rate": 1.8722890037124674e-05,
+      "loss": 0.8722,
+      "step": 7970
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2547707500855747,
+      "learning_rate": 1.8722516891757435e-05,
+      "loss": 0.8233,
+      "step": 7971
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.723806796768845,
+      "learning_rate": 1.872214369560495e-05,
+      "loss": 0.7497,
+      "step": 7972
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6847820834584073,
+      "learning_rate": 1.8721770448669377e-05,
+      "loss": 0.7804,
+      "step": 7973
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4897584120900054,
+      "learning_rate": 1.8721397150952894e-05,
+      "loss": 0.7367,
+      "step": 7974
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3539394243862248,
+      "learning_rate": 1.8721023802457678e-05,
+      "loss": 0.8076,
+      "step": 7975
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3241592990982696,
+      "learning_rate": 1.87206504031859e-05,
+      "loss": 0.7968,
+      "step": 7976
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.420831315154834,
+      "learning_rate": 1.8720276953139737e-05,
+      "loss": 0.7558,
+      "step": 7977
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6241499214262918,
+      "learning_rate": 1.871990345232136e-05,
+      "loss": 0.7691,
+      "step": 7978
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.853740856969615,
+      "learning_rate": 1.871952990073294e-05,
+      "loss": 0.7882,
+      "step": 7979
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3164519027736432,
+      "learning_rate": 1.871915629837666e-05,
+      "loss": 0.6792,
+      "step": 7980
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.8318622690505784,
+      "learning_rate": 1.8718782645254688e-05,
+      "loss": 0.6983,
+      "step": 7981
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5616634665379245,
+      "learning_rate": 1.8718408941369208e-05,
+      "loss": 0.7965,
+      "step": 7982
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5281576997691912,
+      "learning_rate": 1.871803518672239e-05,
+      "loss": 0.7671,
+      "step": 7983
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8785816906936912,
+      "learning_rate": 1.871766138131641e-05,
+      "loss": 0.6509,
+      "step": 7984
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5529527209549296,
+      "learning_rate": 1.8717287525153443e-05,
+      "loss": 0.8613,
+      "step": 7985
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3892800529840224,
+      "learning_rate": 1.8716913618235676e-05,
+      "loss": 0.6831,
+      "step": 7986
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5456668375905465,
+      "learning_rate": 1.871653966056527e-05,
+      "loss": 0.8759,
+      "step": 7987
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.33347547903726,
+      "learning_rate": 1.8716165652144415e-05,
+      "loss": 0.6636,
+      "step": 7988
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4858167312767085,
+      "learning_rate": 1.871579159297528e-05,
+      "loss": 0.7787,
+      "step": 7989
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4225762388421557,
+      "learning_rate": 1.871541748306005e-05,
+      "loss": 0.8167,
+      "step": 7990
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5412826659408336,
+      "learning_rate": 1.87150433224009e-05,
+      "loss": 0.7472,
+      "step": 7991
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.5209107452363795,
+      "learning_rate": 1.871466911100001e-05,
+      "loss": 0.7634,
+      "step": 7992
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5023761211833395,
+      "learning_rate": 1.8714294848859555e-05,
+      "loss": 0.7277,
+      "step": 7993
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5925095094111474,
+      "learning_rate": 1.8713920535981716e-05,
+      "loss": 0.5402,
+      "step": 7994
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6220818544300312,
+      "learning_rate": 1.8713546172368676e-05,
+      "loss": 0.5365,
+      "step": 7995
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9985205140233933,
+      "learning_rate": 1.8713171758022608e-05,
+      "loss": 0.7329,
+      "step": 7996
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7301497603524485,
+      "learning_rate": 1.8712797292945698e-05,
+      "loss": 0.8546,
+      "step": 7997
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.585493065170768,
+      "learning_rate": 1.8712422777140123e-05,
+      "loss": 0.7459,
+      "step": 7998
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.725784905712372,
+      "learning_rate": 1.8712048210608062e-05,
+      "loss": 0.7161,
+      "step": 7999
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3968046709379538,
+      "learning_rate": 1.87116735933517e-05,
+      "loss": 0.706,
+      "step": 8000
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6068598076002243,
+      "learning_rate": 1.8711298925373218e-05,
+      "loss": 0.5787,
+      "step": 8001
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7378166911027249,
+      "learning_rate": 1.8710924206674795e-05,
+      "loss": 0.6701,
+      "step": 8002
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8481879616444357,
+      "learning_rate": 1.871054943725861e-05,
+      "loss": 0.8214,
+      "step": 8003
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6240198698283483,
+      "learning_rate": 1.871017461712685e-05,
+      "loss": 0.8852,
+      "step": 8004
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6227894046699993,
+      "learning_rate": 1.87097997462817e-05,
+      "loss": 0.7277,
+      "step": 8005
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5356656883272706,
+      "learning_rate": 1.8709424824725335e-05,
+      "loss": 0.7785,
+      "step": 8006
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.200574712275475,
+      "learning_rate": 1.8709049852459942e-05,
+      "loss": 0.7482,
+      "step": 8007
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.471148621723678,
+      "learning_rate": 1.87086748294877e-05,
+      "loss": 0.6982,
+      "step": 8008
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4588211844440644,
+      "learning_rate": 1.87082997558108e-05,
+      "loss": 0.806,
+      "step": 8009
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4167629524258207,
+      "learning_rate": 1.8707924631431424e-05,
+      "loss": 0.8214,
+      "step": 8010
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.429502825809229,
+      "learning_rate": 1.870754945635175e-05,
+      "loss": 0.8125,
+      "step": 8011
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.6707024623498348,
+      "learning_rate": 1.8707174230573967e-05,
+      "loss": 0.8311,
+      "step": 8012
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.749253652003758,
+      "learning_rate": 1.870679895410026e-05,
+      "loss": 0.7889,
+      "step": 8013
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5220193862338396,
+      "learning_rate": 1.870642362693281e-05,
+      "loss": 0.7503,
+      "step": 8014
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3307012880492632,
+      "learning_rate": 1.870604824907381e-05,
+      "loss": 0.7958,
+      "step": 8015
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7445408511520875,
+      "learning_rate": 1.8705672820525437e-05,
+      "loss": 0.7912,
+      "step": 8016
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6573716166642416,
+      "learning_rate": 1.870529734128988e-05,
+      "loss": 0.8042,
+      "step": 8017
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5349287184934144,
+      "learning_rate": 1.8704921811369328e-05,
+      "loss": 0.8178,
+      "step": 8018
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6154028484688132,
+      "learning_rate": 1.8704546230765964e-05,
+      "loss": 0.5809,
+      "step": 8019
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4255872498517463,
+      "learning_rate": 1.870417059948198e-05,
+      "loss": 0.7165,
+      "step": 8020
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.669938197878764,
+      "learning_rate": 1.8703794917519555e-05,
+      "loss": 0.7526,
+      "step": 8021
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4432383732048453,
+      "learning_rate": 1.8703419184880882e-05,
+      "loss": 0.7598,
+      "step": 8022
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6260853554332713,
+      "learning_rate": 1.8703043401568144e-05,
+      "loss": 0.9656,
+      "step": 8023
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8750816950239277,
+      "learning_rate": 1.8702667567583533e-05,
+      "loss": 0.8879,
+      "step": 8024
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4734142424506569,
+      "learning_rate": 1.870229168292924e-05,
+      "loss": 0.7349,
+      "step": 8025
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6556102378567141,
+      "learning_rate": 1.8701915747607446e-05,
+      "loss": 0.6949,
+      "step": 8026
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.625367700744058,
+      "learning_rate": 1.8701539761620348e-05,
+      "loss": 0.7519,
+      "step": 8027
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8722849358157436,
+      "learning_rate": 1.8701163724970125e-05,
+      "loss": 0.6983,
+      "step": 8028
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3433934922195032,
+      "learning_rate": 1.8700787637658976e-05,
+      "loss": 0.717,
+      "step": 8029
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.361784989420578,
+      "learning_rate": 1.8700411499689088e-05,
+      "loss": 0.8664,
+      "step": 8030
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5241939341564479,
+      "learning_rate": 1.8700035311062644e-05,
+      "loss": 0.7606,
+      "step": 8031
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5390401468005392,
+      "learning_rate": 1.8699659071781845e-05,
+      "loss": 0.8189,
+      "step": 8032
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6780734429152158,
+      "learning_rate": 1.8699282781848873e-05,
+      "loss": 0.6998,
+      "step": 8033
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6063858246627556,
+      "learning_rate": 1.8698906441265925e-05,
+      "loss": 0.5469,
+      "step": 8034
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5005168997894103,
+      "learning_rate": 1.8698530050035188e-05,
+      "loss": 0.8106,
+      "step": 8035
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.752724920552033,
+      "learning_rate": 1.8698153608158856e-05,
+      "loss": 0.6988,
+      "step": 8036
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.933179146192046,
+      "learning_rate": 1.869777711563912e-05,
+      "loss": 0.779,
+      "step": 8037
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3913492389249396,
+      "learning_rate": 1.8697400572478173e-05,
+      "loss": 0.8734,
+      "step": 8038
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4430027103443106,
+      "learning_rate": 1.8697023978678205e-05,
+      "loss": 0.8146,
+      "step": 8039
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.751336824886342,
+      "learning_rate": 1.869664733424141e-05,
+      "loss": 0.8735,
+      "step": 8040
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.472905250406065,
+      "learning_rate": 1.869627063916998e-05,
+      "loss": 0.8219,
+      "step": 8041
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.586701940638751,
+      "learning_rate": 1.869589389346611e-05,
+      "loss": 0.5589,
+      "step": 8042
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4267670982492422,
+      "learning_rate": 1.8695517097131993e-05,
+      "loss": 0.8725,
+      "step": 8043
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9256881801307681,
+      "learning_rate": 1.8695140250169825e-05,
+      "loss": 0.8323,
+      "step": 8044
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4135142992366017,
+      "learning_rate": 1.8694763352581792e-05,
+      "loss": 0.8768,
+      "step": 8045
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.025234504079438,
+      "learning_rate": 1.86943864043701e-05,
+      "loss": 0.8767,
+      "step": 8046
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.803371526972751,
+      "learning_rate": 1.8694009405536934e-05,
+      "loss": 0.7371,
+      "step": 8047
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8166219420609266,
+      "learning_rate": 1.8693632356084493e-05,
+      "loss": 0.8079,
+      "step": 8048
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6466099064700304,
+      "learning_rate": 1.8693255256014976e-05,
+      "loss": 0.8524,
+      "step": 8049
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7547965743843075,
+      "learning_rate": 1.8692878105330568e-05,
+      "loss": 0.9051,
+      "step": 8050
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4853464338367217,
+      "learning_rate": 1.8692500904033478e-05,
+      "loss": 0.7856,
+      "step": 8051
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5889573484137782,
+      "learning_rate": 1.869212365212589e-05,
+      "loss": 0.5616,
+      "step": 8052
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6326790133014728,
+      "learning_rate": 1.8691746349610008e-05,
+      "loss": 0.8328,
+      "step": 8053
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5687169074846363,
+      "learning_rate": 1.869136899648803e-05,
+      "loss": 0.6709,
+      "step": 8054
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.260417499344717,
+      "learning_rate": 1.8690991592762147e-05,
+      "loss": 0.7684,
+      "step": 8055
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3639578144900248,
+      "learning_rate": 1.869061413843456e-05,
+      "loss": 0.8384,
+      "step": 8056
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4038842709087356,
+      "learning_rate": 1.869023663350747e-05,
+      "loss": 0.69,
+      "step": 8057
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5769680212487904,
+      "learning_rate": 1.8689859077983064e-05,
+      "loss": 0.7458,
+      "step": 8058
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4544927881837524,
+      "learning_rate": 1.868948147186355e-05,
+      "loss": 0.7746,
+      "step": 8059
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5983451738308438,
+      "learning_rate": 1.8689103815151127e-05,
+      "loss": 0.7694,
+      "step": 8060
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6111080408528652,
+      "learning_rate": 1.8688726107847986e-05,
+      "loss": 0.5436,
+      "step": 8061
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.920376266338194,
+      "learning_rate": 1.8688348349956334e-05,
+      "loss": 0.7031,
+      "step": 8062
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5550206395014934,
+      "learning_rate": 1.8687970541478367e-05,
+      "loss": 0.9317,
+      "step": 8063
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6708326927798014,
+      "learning_rate": 1.8687592682416284e-05,
+      "loss": 0.8544,
+      "step": 8064
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5187144103390944,
+      "learning_rate": 1.8687214772772283e-05,
+      "loss": 0.6723,
+      "step": 8065
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8886105835724614,
+      "learning_rate": 1.8686836812548573e-05,
+      "loss": 0.6807,
+      "step": 8066
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6281023218900879,
+      "learning_rate": 1.8686458801747347e-05,
+      "loss": 0.5369,
+      "step": 8067
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.349180541612459,
+      "learning_rate": 1.8686080740370807e-05,
+      "loss": 0.8303,
+      "step": 8068
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5836571176330655,
+      "learning_rate": 1.8685702628421155e-05,
+      "loss": 0.5457,
+      "step": 8069
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5222363237858452,
+      "learning_rate": 1.8685324465900594e-05,
+      "loss": 0.8893,
+      "step": 8070
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6081455454017068,
+      "learning_rate": 1.868494625281132e-05,
+      "loss": 0.7583,
+      "step": 8071
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5623961070120301,
+      "learning_rate": 1.8684567989155545e-05,
+      "loss": 0.5056,
+      "step": 8072
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.539394262456835,
+      "learning_rate": 1.868418967493546e-05,
+      "loss": 0.6583,
+      "step": 8073
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.321119111216568,
+      "learning_rate": 1.8683811310153276e-05,
+      "loss": 0.7975,
+      "step": 8074
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5389203237327025,
+      "learning_rate": 1.8683432894811195e-05,
+      "loss": 0.7604,
+      "step": 8075
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8200302230762362,
+      "learning_rate": 1.8683054428911414e-05,
+      "loss": 0.8154,
+      "step": 8076
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4920947802317515,
+      "learning_rate": 1.8682675912456147e-05,
+      "loss": 0.7133,
+      "step": 8077
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.456862232284758,
+      "learning_rate": 1.868229734544759e-05,
+      "loss": 0.7449,
+      "step": 8078
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7840683429896742,
+      "learning_rate": 1.8681918727887945e-05,
+      "loss": 0.9037,
+      "step": 8079
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3509370702147843,
+      "learning_rate": 1.8681540059779425e-05,
+      "loss": 0.751,
+      "step": 8080
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.829800958888895,
+      "learning_rate": 1.8681161341124227e-05,
+      "loss": 0.8452,
+      "step": 8081
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4627057338928937,
+      "learning_rate": 1.8680782571924564e-05,
+      "loss": 0.6781,
+      "step": 8082
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4108275379681743,
+      "learning_rate": 1.868040375218263e-05,
+      "loss": 0.7464,
+      "step": 8083
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6380072546748559,
+      "learning_rate": 1.868002488190064e-05,
+      "loss": 0.772,
+      "step": 8084
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9212378449745322,
+      "learning_rate": 1.86796459610808e-05,
+      "loss": 0.7767,
+      "step": 8085
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3742200621834295,
+      "learning_rate": 1.8679266989725313e-05,
+      "loss": 0.7334,
+      "step": 8086
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9383046698643078,
+      "learning_rate": 1.8678887967836382e-05,
+      "loss": 0.8834,
+      "step": 8087
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.333457850301702,
+      "learning_rate": 1.867850889541622e-05,
+      "loss": 0.7778,
+      "step": 8088
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8267320032493481,
+      "learning_rate": 1.867812977246703e-05,
+      "loss": 0.8202,
+      "step": 8089
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8482171803016272,
+      "learning_rate": 1.8677750598991023e-05,
+      "loss": 0.8048,
+      "step": 8090
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5942412234078146,
+      "learning_rate": 1.8677371374990406e-05,
+      "loss": 0.5354,
+      "step": 8091
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5762525812142076,
+      "learning_rate": 1.8676992100467383e-05,
+      "loss": 0.8886,
+      "step": 8092
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.190497945679652,
+      "learning_rate": 1.8676612775424165e-05,
+      "loss": 0.758,
+      "step": 8093
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6156418645562394,
+      "learning_rate": 1.8676233399862963e-05,
+      "loss": 0.8061,
+      "step": 8094
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5777704593023636,
+      "learning_rate": 1.8675853973785976e-05,
+      "loss": 0.5147,
+      "step": 8095
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2561888473966656,
+      "learning_rate": 1.8675474497195428e-05,
+      "loss": 0.7145,
+      "step": 8096
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6554012114146903,
+      "learning_rate": 1.867509497009352e-05,
+      "loss": 0.7013,
+      "step": 8097
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5814278762562052,
+      "learning_rate": 1.867471539248246e-05,
+      "loss": 0.5342,
+      "step": 8098
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7417895691859557,
+      "learning_rate": 1.8674335764364464e-05,
+      "loss": 0.8385,
+      "step": 8099
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.589673338748033,
+      "learning_rate": 1.8673956085741734e-05,
+      "loss": 0.8625,
+      "step": 8100
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4416230752684442,
+      "learning_rate": 1.8673576356616488e-05,
+      "loss": 0.7534,
+      "step": 8101
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5933933807374554,
+      "learning_rate": 1.8673196576990938e-05,
+      "loss": 0.7227,
+      "step": 8102
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.476815895062054,
+      "learning_rate": 1.8672816746867285e-05,
+      "loss": 0.7514,
+      "step": 8103
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6652057604636294,
+      "learning_rate": 1.867243686624775e-05,
+      "loss": 0.7414,
+      "step": 8104
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6718746753774756,
+      "learning_rate": 1.8672056935134545e-05,
+      "loss": 0.756,
+      "step": 8105
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.580483478095097,
+      "learning_rate": 1.8671676953529875e-05,
+      "loss": 0.8005,
+      "step": 8106
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.481350901297954,
+      "learning_rate": 1.8671296921435956e-05,
+      "loss": 0.8171,
+      "step": 8107
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.482293940978749,
+      "learning_rate": 1.8670916838855005e-05,
+      "loss": 0.5707,
+      "step": 8108
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6167079286590745,
+      "learning_rate": 1.867053670578923e-05,
+      "loss": 0.5781,
+      "step": 8109
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9311723986421052,
+      "learning_rate": 1.8670156522240843e-05,
+      "loss": 0.8807,
+      "step": 8110
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6229455395651162,
+      "learning_rate": 1.866977628821206e-05,
+      "loss": 0.7237,
+      "step": 8111
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6548580960067953,
+      "learning_rate": 1.8669396003705095e-05,
+      "loss": 0.7598,
+      "step": 8112
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3596095811580078,
+      "learning_rate": 1.8669015668722162e-05,
+      "loss": 0.7266,
+      "step": 8113
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.060903066973596,
+      "learning_rate": 1.8668635283265475e-05,
+      "loss": 0.7078,
+      "step": 8114
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4384451917639915,
+      "learning_rate": 1.8668254847337252e-05,
+      "loss": 0.8576,
+      "step": 8115
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6240394950552098,
+      "learning_rate": 1.86678743609397e-05,
+      "loss": 0.5325,
+      "step": 8116
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6199266326664457,
+      "learning_rate": 1.8667493824075042e-05,
+      "loss": 0.6577,
+      "step": 8117
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6135635688092892,
+      "learning_rate": 1.866711323674549e-05,
+      "loss": 0.5727,
+      "step": 8118
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.0067619102817127,
+      "learning_rate": 1.8666732598953263e-05,
+      "loss": 0.7404,
+      "step": 8119
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5650099180112962,
+      "learning_rate": 1.8666351910700572e-05,
+      "loss": 0.8336,
+      "step": 8120
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4578360799973074,
+      "learning_rate": 1.8665971171989638e-05,
+      "loss": 0.7187,
+      "step": 8121
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.338561320553645,
+      "learning_rate": 1.8665590382822675e-05,
+      "loss": 0.6418,
+      "step": 8122
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6119162725532026,
+      "learning_rate": 1.8665209543201904e-05,
+      "loss": 0.782,
+      "step": 8123
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.166039046952531,
+      "learning_rate": 1.866482865312954e-05,
+      "loss": 0.7096,
+      "step": 8124
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8542519768691463,
+      "learning_rate": 1.8664447712607798e-05,
+      "loss": 0.7891,
+      "step": 8125
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8222690728267765,
+      "learning_rate": 1.8664066721638895e-05,
+      "loss": 0.752,
+      "step": 8126
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5821141699191257,
+      "learning_rate": 1.8663685680225056e-05,
+      "loss": 0.7865,
+      "step": 8127
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9397768085118692,
+      "learning_rate": 1.8663304588368495e-05,
+      "loss": 0.9159,
+      "step": 8128
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6065078400882884,
+      "learning_rate": 1.8662923446071436e-05,
+      "loss": 0.8108,
+      "step": 8129
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.834910344154888,
+      "learning_rate": 1.866254225333609e-05,
+      "loss": 0.6027,
+      "step": 8130
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9690865634061216,
+      "learning_rate": 1.8662161010164682e-05,
+      "loss": 0.8593,
+      "step": 8131
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5675618775006017,
+      "learning_rate": 1.8661779716559428e-05,
+      "loss": 0.7779,
+      "step": 8132
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4814676281616224,
+      "learning_rate": 1.866139837252255e-05,
+      "loss": 0.8006,
+      "step": 8133
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8419971335150338,
+      "learning_rate": 1.8661016978056273e-05,
+      "loss": 0.7496,
+      "step": 8134
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2984539025392081,
+      "learning_rate": 1.8660635533162813e-05,
+      "loss": 0.8408,
+      "step": 8135
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.069587806014136,
+      "learning_rate": 1.866025403784439e-05,
+      "loss": 0.7539,
+      "step": 8136
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4813644029182276,
+      "learning_rate": 1.8659872492103222e-05,
+      "loss": 0.6706,
+      "step": 8137
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.721513698126267,
+      "learning_rate": 1.865949089594154e-05,
+      "loss": 0.6715,
+      "step": 8138
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.309812002711926,
+      "learning_rate": 1.8659109249361557e-05,
+      "loss": 0.8485,
+      "step": 8139
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6858975179834699,
+      "learning_rate": 1.8658727552365498e-05,
+      "loss": 0.9345,
+      "step": 8140
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7762810093975587,
+      "learning_rate": 1.865834580495559e-05,
+      "loss": 0.7314,
+      "step": 8141
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4655321101049001,
+      "learning_rate": 1.865796400713405e-05,
+      "loss": 0.7,
+      "step": 8142
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.815238144073417,
+      "learning_rate": 1.86575821589031e-05,
+      "loss": 0.8697,
+      "step": 8143
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4974983469345104,
+      "learning_rate": 1.8657200260264966e-05,
+      "loss": 0.7771,
+      "step": 8144
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8279049852048244,
+      "learning_rate": 1.865681831122187e-05,
+      "loss": 0.7062,
+      "step": 8145
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4883017689824023,
+      "learning_rate": 1.865643631177604e-05,
+      "loss": 0.6555,
+      "step": 8146
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8779075759838417,
+      "learning_rate": 1.86560542619297e-05,
+      "loss": 0.807,
+      "step": 8147
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5509713016613569,
+      "learning_rate": 1.865567216168507e-05,
+      "loss": 0.8411,
+      "step": 8148
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6518685003917324,
+      "learning_rate": 1.865529001104437e-05,
+      "loss": 0.7384,
+      "step": 8149
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6657669013367211,
+      "learning_rate": 1.8654907810009838e-05,
+      "loss": 0.8499,
+      "step": 8150
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4277138042268573,
+      "learning_rate": 1.865452555858369e-05,
+      "loss": 0.8419,
+      "step": 8151
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.297411035300696,
+      "learning_rate": 1.8654143256768157e-05,
+      "loss": 0.6563,
+      "step": 8152
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4639066015646491,
+      "learning_rate": 1.865376090456546e-05,
+      "loss": 0.7095,
+      "step": 8153
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5775712109794366,
+      "learning_rate": 1.8653378501977827e-05,
+      "loss": 0.5483,
+      "step": 8154
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8305797610202128,
+      "learning_rate": 1.8652996049007485e-05,
+      "loss": 0.8117,
+      "step": 8155
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.529706678339382,
+      "learning_rate": 1.865261354565666e-05,
+      "loss": 0.6372,
+      "step": 8156
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8974638248489493,
+      "learning_rate": 1.8652230991927582e-05,
+      "loss": 0.8821,
+      "step": 8157
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.679693916319935,
+      "learning_rate": 1.8651848387822473e-05,
+      "loss": 0.7573,
+      "step": 8158
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6125178916459322,
+      "learning_rate": 1.8651465733343562e-05,
+      "loss": 0.7395,
+      "step": 8159
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.454299929999799,
+      "learning_rate": 1.8651083028493082e-05,
+      "loss": 0.7073,
+      "step": 8160
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.482682265530725,
+      "learning_rate": 1.8650700273273257e-05,
+      "loss": 0.7715,
+      "step": 8161
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.832017779554936,
+      "learning_rate": 1.8650317467686318e-05,
+      "loss": 0.8119,
+      "step": 8162
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9387930487535963,
+      "learning_rate": 1.864993461173449e-05,
+      "loss": 0.7052,
+      "step": 8163
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6005351864860726,
+      "learning_rate": 1.8649551705420004e-05,
+      "loss": 0.8037,
+      "step": 8164
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.595646694617957,
+      "learning_rate": 1.864916874874509e-05,
+      "loss": 0.8541,
+      "step": 8165
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.161613232849768,
+      "learning_rate": 1.8648785741711975e-05,
+      "loss": 0.7475,
+      "step": 8166
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9574191448050227,
+      "learning_rate": 1.8648402684322893e-05,
+      "loss": 0.7376,
+      "step": 8167
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.193856038074532,
+      "learning_rate": 1.8648019576580074e-05,
+      "loss": 0.7947,
+      "step": 8168
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.696979897040671,
+      "learning_rate": 1.8647636418485748e-05,
+      "loss": 0.7837,
+      "step": 8169
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5488489517789048,
+      "learning_rate": 1.8647253210042143e-05,
+      "loss": 0.6592,
+      "step": 8170
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6217305239763569,
+      "learning_rate": 1.8646869951251493e-05,
+      "loss": 0.7011,
+      "step": 8171
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5116527833148494,
+      "learning_rate": 1.8646486642116026e-05,
+      "loss": 0.7806,
+      "step": 8172
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5712915579748312,
+      "learning_rate": 1.864610328263798e-05,
+      "loss": 0.7124,
+      "step": 8173
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.763364228992251,
+      "learning_rate": 1.864571987281958e-05,
+      "loss": 0.6981,
+      "step": 8174
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3416134226251766,
+      "learning_rate": 1.8645336412663062e-05,
+      "loss": 0.7498,
+      "step": 8175
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6871866185131887,
+      "learning_rate": 1.864495290217066e-05,
+      "loss": 0.7702,
+      "step": 8176
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6408835260423018,
+      "learning_rate": 1.8644569341344604e-05,
+      "loss": 0.7995,
+      "step": 8177
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.850993830361467,
+      "learning_rate": 1.864418573018713e-05,
+      "loss": 0.7307,
+      "step": 8178
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5209422681386464,
+      "learning_rate": 1.864380206870047e-05,
+      "loss": 0.6833,
+      "step": 8179
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.507962220930179,
+      "learning_rate": 1.8643418356886854e-05,
+      "loss": 0.7581,
+      "step": 8180
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3818419387901053,
+      "learning_rate": 1.8643034594748525e-05,
+      "loss": 0.8449,
+      "step": 8181
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.891577698612932,
+      "learning_rate": 1.864265078228771e-05,
+      "loss": 0.7417,
+      "step": 8182
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3488113302994074,
+      "learning_rate": 1.8642266919506644e-05,
+      "loss": 0.7104,
+      "step": 8183
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4578843538572666,
+      "learning_rate": 1.864188300640757e-05,
+      "loss": 0.6718,
+      "step": 8184
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5111354928222775,
+      "learning_rate": 1.864149904299271e-05,
+      "loss": 0.7418,
+      "step": 8185
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1532642207441315,
+      "learning_rate": 1.8641115029264312e-05,
+      "loss": 0.8203,
+      "step": 8186
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.396089815738929,
+      "learning_rate": 1.8640730965224602e-05,
+      "loss": 0.8029,
+      "step": 8187
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1395901399218666,
+      "learning_rate": 1.8640346850875825e-05,
+      "loss": 0.7466,
+      "step": 8188
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6004380386239767,
+      "learning_rate": 1.863996268622021e-05,
+      "loss": 0.7591,
+      "step": 8189
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8407435027475267,
+      "learning_rate": 1.8639578471259997e-05,
+      "loss": 0.7819,
+      "step": 8190
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4550923687002502,
+      "learning_rate": 1.8639194205997423e-05,
+      "loss": 0.7939,
+      "step": 8191
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1696210963995313,
+      "learning_rate": 1.863880989043473e-05,
+      "loss": 0.7411,
+      "step": 8192
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0622110454107525,
+      "learning_rate": 1.8638425524574148e-05,
+      "loss": 0.7814,
+      "step": 8193
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5296913017198552,
+      "learning_rate": 1.8638041108417916e-05,
+      "loss": 0.7843,
+      "step": 8194
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3815039238788787,
+      "learning_rate": 1.8637656641968272e-05,
+      "loss": 0.7511,
+      "step": 8195
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6138887736282232,
+      "learning_rate": 1.863727212522746e-05,
+      "loss": 0.7499,
+      "step": 8196
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.306277370763683,
+      "learning_rate": 1.8636887558197713e-05,
+      "loss": 0.6625,
+      "step": 8197
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4772322730014718,
+      "learning_rate": 1.8636502940881276e-05,
+      "loss": 0.8453,
+      "step": 8198
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4264637612162905,
+      "learning_rate": 1.863611827328038e-05,
+      "loss": 0.8725,
+      "step": 8199
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.578190335472103,
+      "learning_rate": 1.8635733555397273e-05,
+      "loss": 0.8564,
+      "step": 8200
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8701730485434158,
+      "learning_rate": 1.863534878723419e-05,
+      "loss": 0.8472,
+      "step": 8201
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8768591598794455,
+      "learning_rate": 1.863496396879337e-05,
+      "loss": 0.8146,
+      "step": 8202
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8043392238057232,
+      "learning_rate": 1.863457910007706e-05,
+      "loss": 0.7601,
+      "step": 8203
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7556282350439818,
+      "learning_rate": 1.8634194181087494e-05,
+      "loss": 0.8051,
+      "step": 8204
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.0374670481321235,
+      "learning_rate": 1.8633809211826917e-05,
+      "loss": 0.7344,
+      "step": 8205
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3760396108614708,
+      "learning_rate": 1.8633424192297566e-05,
+      "loss": 0.8282,
+      "step": 8206
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.44207252363168,
+      "learning_rate": 1.8633039122501692e-05,
+      "loss": 0.7553,
+      "step": 8207
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6341527472281416,
+      "learning_rate": 1.8632654002441528e-05,
+      "loss": 0.5246,
+      "step": 8208
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.408121614095252,
+      "learning_rate": 1.863226883211932e-05,
+      "loss": 0.8306,
+      "step": 8209
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6977046756138554,
+      "learning_rate": 1.863188361153731e-05,
+      "loss": 0.6735,
+      "step": 8210
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6439553907498713,
+      "learning_rate": 1.8631498340697736e-05,
+      "loss": 0.554,
+      "step": 8211
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8858084742778563,
+      "learning_rate": 1.8631113019602854e-05,
+      "loss": 0.8056,
+      "step": 8212
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4036494407205848,
+      "learning_rate": 1.8630727648254896e-05,
+      "loss": 0.7519,
+      "step": 8213
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7628098823686165,
+      "learning_rate": 1.8630342226656108e-05,
+      "loss": 0.6956,
+      "step": 8214
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5786929115998883,
+      "learning_rate": 1.8629956754808734e-05,
+      "loss": 0.4987,
+      "step": 8215
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8114404004706568,
+      "learning_rate": 1.8629571232715024e-05,
+      "loss": 0.8616,
+      "step": 8216
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4894492277460625,
+      "learning_rate": 1.8629185660377216e-05,
+      "loss": 0.803,
+      "step": 8217
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3946970026909022,
+      "learning_rate": 1.8628800037797558e-05,
+      "loss": 0.7814,
+      "step": 8218
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6530257154606596,
+      "learning_rate": 1.8628414364978292e-05,
+      "loss": 0.8316,
+      "step": 8219
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3744292433930831,
+      "learning_rate": 1.8628028641921668e-05,
+      "loss": 0.6861,
+      "step": 8220
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5447519704635322,
+      "learning_rate": 1.862764286862993e-05,
+      "loss": 0.7691,
+      "step": 8221
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5389440232661076,
+      "learning_rate": 1.862725704510532e-05,
+      "loss": 0.7185,
+      "step": 8222
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5806957651502451,
+      "learning_rate": 1.8626871171350093e-05,
+      "loss": 0.6327,
+      "step": 8223
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5040196189573714,
+      "learning_rate": 1.8626485247366486e-05,
+      "loss": 0.972,
+      "step": 8224
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.47978791579869,
+      "learning_rate": 1.8626099273156753e-05,
+      "loss": 0.7967,
+      "step": 8225
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4855624360925974,
+      "learning_rate": 1.862571324872314e-05,
+      "loss": 0.691,
+      "step": 8226
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5648982229405375,
+      "learning_rate": 1.8625327174067896e-05,
+      "loss": 0.5236,
+      "step": 8227
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3267145019477904,
+      "learning_rate": 1.8624941049193262e-05,
+      "loss": 0.8406,
+      "step": 8228
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5519914902946028,
+      "learning_rate": 1.862455487410149e-05,
+      "loss": 0.908,
+      "step": 8229
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8530771192422995,
+      "learning_rate": 1.8624168648794833e-05,
+      "loss": 0.7113,
+      "step": 8230
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6148055579129579,
+      "learning_rate": 1.8623782373275535e-05,
+      "loss": 0.6899,
+      "step": 8231
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4984563093205532,
+      "learning_rate": 1.8623396047545844e-05,
+      "loss": 0.7162,
+      "step": 8232
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6198458103148212,
+      "learning_rate": 1.862300967160801e-05,
+      "loss": 0.6543,
+      "step": 8233
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8286372275815406,
+      "learning_rate": 1.8622623245464286e-05,
+      "loss": 0.7805,
+      "step": 8234
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.536621597728307,
+      "learning_rate": 1.862223676911692e-05,
+      "loss": 0.7574,
+      "step": 8235
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4644290476727508,
+      "learning_rate": 1.862185024256816e-05,
+      "loss": 0.8027,
+      "step": 8236
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5522559086671417,
+      "learning_rate": 1.862146366582026e-05,
+      "loss": 0.6773,
+      "step": 8237
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.728913065947378,
+      "learning_rate": 1.862107703887547e-05,
+      "loss": 0.7574,
+      "step": 8238
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4495011683087926,
+      "learning_rate": 1.8620690361736037e-05,
+      "loss": 0.8532,
+      "step": 8239
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.619504476378969,
+      "learning_rate": 1.8620303634404218e-05,
+      "loss": 0.7127,
+      "step": 8240
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6614552602209653,
+      "learning_rate": 1.8619916856882258e-05,
+      "loss": 0.671,
+      "step": 8241
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3822855475113973,
+      "learning_rate": 1.8619530029172417e-05,
+      "loss": 0.7813,
+      "step": 8242
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.557729162384955,
+      "learning_rate": 1.861914315127694e-05,
+      "loss": 0.7654,
+      "step": 8243
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7639275443960898,
+      "learning_rate": 1.8618756223198086e-05,
+      "loss": 0.7729,
+      "step": 8244
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5834769398478934,
+      "learning_rate": 1.8618369244938103e-05,
+      "loss": 0.6338,
+      "step": 8245
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.172033628980841,
+      "learning_rate": 1.8617982216499247e-05,
+      "loss": 0.7893,
+      "step": 8246
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1109413493499853,
+      "learning_rate": 1.861759513788377e-05,
+      "loss": 0.8064,
+      "step": 8247
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0640864872030997,
+      "learning_rate": 1.8617208009093922e-05,
+      "loss": 0.8416,
+      "step": 8248
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2101443707671122,
+      "learning_rate": 1.8616820830131966e-05,
+      "loss": 0.8052,
+      "step": 8249
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4952644963539827,
+      "learning_rate": 1.861643360100015e-05,
+      "loss": 0.6513,
+      "step": 8250
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7162942814535445,
+      "learning_rate": 1.861604632170073e-05,
+      "loss": 0.9035,
+      "step": 8251
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5592651271715543,
+      "learning_rate": 1.861565899223596e-05,
+      "loss": 0.8687,
+      "step": 8252
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3655689106165503,
+      "learning_rate": 1.8615271612608094e-05,
+      "loss": 0.6885,
+      "step": 8253
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.59898082952056,
+      "learning_rate": 1.861488418281939e-05,
+      "loss": 0.525,
+      "step": 8254
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5475525408960773,
+      "learning_rate": 1.8614496702872102e-05,
+      "loss": 0.7425,
+      "step": 8255
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6320296707415588,
+      "learning_rate": 1.8614109172768487e-05,
+      "loss": 0.5272,
+      "step": 8256
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6395909223924705,
+      "learning_rate": 1.8613721592510804e-05,
+      "loss": 0.5346,
+      "step": 8257
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7795908801770977,
+      "learning_rate": 1.8613333962101303e-05,
+      "loss": 0.7149,
+      "step": 8258
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3287475089816752,
+      "learning_rate": 1.8612946281542247e-05,
+      "loss": 0.7225,
+      "step": 8259
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.628078281399691,
+      "learning_rate": 1.861255855083589e-05,
+      "loss": 0.7994,
+      "step": 8260
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.562418050107056,
+      "learning_rate": 1.861217076998449e-05,
+      "loss": 0.8491,
+      "step": 8261
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4229862556264168,
+      "learning_rate": 1.8611782938990308e-05,
+      "loss": 0.7992,
+      "step": 8262
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4146838288140926,
+      "learning_rate": 1.86113950578556e-05,
+      "loss": 0.8744,
+      "step": 8263
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3783412867934466,
+      "learning_rate": 1.8611007126582617e-05,
+      "loss": 0.7876,
+      "step": 8264
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9216448303394387,
+      "learning_rate": 1.861061914517363e-05,
+      "loss": 0.7227,
+      "step": 8265
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.71976508354185,
+      "learning_rate": 1.861023111363089e-05,
+      "loss": 0.7628,
+      "step": 8266
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4713835246987024,
+      "learning_rate": 1.860984303195666e-05,
+      "loss": 0.8884,
+      "step": 8267
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5205487856526987,
+      "learning_rate": 1.8609454900153195e-05,
+      "loss": 0.8583,
+      "step": 8268
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6856825996889844,
+      "learning_rate": 1.860906671822276e-05,
+      "loss": 0.8088,
+      "step": 8269
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4677831032555921,
+      "learning_rate": 1.8608678486167615e-05,
+      "loss": 0.8699,
+      "step": 8270
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6702800404890088,
+      "learning_rate": 1.8608290203990015e-05,
+      "loss": 0.7098,
+      "step": 8271
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6326671349655126,
+      "learning_rate": 1.8607901871692226e-05,
+      "loss": 0.7799,
+      "step": 8272
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5105532031796658,
+      "learning_rate": 1.8607513489276507e-05,
+      "loss": 0.8261,
+      "step": 8273
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5272129617008023,
+      "learning_rate": 1.860712505674512e-05,
+      "loss": 0.7349,
+      "step": 8274
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.546700292137492,
+      "learning_rate": 1.8606736574100325e-05,
+      "loss": 0.7526,
+      "step": 8275
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4562681056185,
+      "learning_rate": 1.8606348041344383e-05,
+      "loss": 0.7143,
+      "step": 8276
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5829377202064832,
+      "learning_rate": 1.860595945847956e-05,
+      "loss": 0.7778,
+      "step": 8277
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.368979592221679,
+      "learning_rate": 1.8605570825508117e-05,
+      "loss": 0.8363,
+      "step": 8278
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.318409982902364,
+      "learning_rate": 1.8605182142432317e-05,
+      "loss": 0.7777,
+      "step": 8279
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6586377910405934,
+      "learning_rate": 1.860479340925442e-05,
+      "loss": 0.7561,
+      "step": 8280
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6264321004560827,
+      "learning_rate": 1.860440462597669e-05,
+      "loss": 0.795,
+      "step": 8281
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3470343954751498,
+      "learning_rate": 1.8604015792601395e-05,
+      "loss": 0.7978,
+      "step": 8282
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4758477544297874,
+      "learning_rate": 1.8603626909130796e-05,
+      "loss": 0.6935,
+      "step": 8283
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4902382675278019,
+      "learning_rate": 1.8603237975567157e-05,
+      "loss": 0.8244,
+      "step": 8284
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6298049342749555,
+      "learning_rate": 1.860284899191274e-05,
+      "loss": 0.5908,
+      "step": 8285
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5356857913998427,
+      "learning_rate": 1.8602459958169817e-05,
+      "loss": 0.8372,
+      "step": 8286
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7103463488614554,
+      "learning_rate": 1.8602070874340646e-05,
+      "loss": 0.7081,
+      "step": 8287
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6307020109047259,
+      "learning_rate": 1.8601681740427495e-05,
+      "loss": 0.817,
+      "step": 8288
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4501052563172252,
+      "learning_rate": 1.8601292556432632e-05,
+      "loss": 0.7816,
+      "step": 8289
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.503907285581966,
+      "learning_rate": 1.860090332235832e-05,
+      "loss": 0.7072,
+      "step": 8290
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5410322272026769,
+      "learning_rate": 1.860051403820682e-05,
+      "loss": 0.689,
+      "step": 8291
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6023794631611803,
+      "learning_rate": 1.860012470398041e-05,
+      "loss": 0.5742,
+      "step": 8292
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4309994546621032,
+      "learning_rate": 1.859973531968135e-05,
+      "loss": 0.8004,
+      "step": 8293
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6009386521804585,
+      "learning_rate": 1.859934588531191e-05,
+      "loss": 0.5281,
+      "step": 8294
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5024729155847762,
+      "learning_rate": 1.8598956400874354e-05,
+      "loss": 0.7434,
+      "step": 8295
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6091664810075765,
+      "learning_rate": 1.8598566866370952e-05,
+      "loss": 0.5398,
+      "step": 8296
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6034915703091056,
+      "learning_rate": 1.859817728180397e-05,
+      "loss": 0.5501,
+      "step": 8297
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.591157221840879,
+      "learning_rate": 1.859778764717568e-05,
+      "loss": 0.5449,
+      "step": 8298
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6100865760119687,
+      "learning_rate": 1.8597397962488347e-05,
+      "loss": 0.7335,
+      "step": 8299
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9966135989045355,
+      "learning_rate": 1.8597008227744242e-05,
+      "loss": 0.6668,
+      "step": 8300
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4469337831819693,
+      "learning_rate": 1.8596618442945634e-05,
+      "loss": 0.7836,
+      "step": 8301
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5002834015889426,
+      "learning_rate": 1.8596228608094792e-05,
+      "loss": 0.753,
+      "step": 8302
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6312607185074115,
+      "learning_rate": 1.8595838723193986e-05,
+      "loss": 0.7092,
+      "step": 8303
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6020995657917259,
+      "learning_rate": 1.859544878824548e-05,
+      "loss": 0.7419,
+      "step": 8304
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7103951246801867,
+      "learning_rate": 1.859505880325155e-05,
+      "loss": 0.7549,
+      "step": 8305
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5723507013500913,
+      "learning_rate": 1.8594668768214474e-05,
+      "loss": 0.8677,
+      "step": 8306
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6925568207421369,
+      "learning_rate": 1.859427868313651e-05,
+      "loss": 0.7443,
+      "step": 8307
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.609385466817854,
+      "learning_rate": 1.8593888548019938e-05,
+      "loss": 0.7035,
+      "step": 8308
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4334247924442713,
+      "learning_rate": 1.8593498362867023e-05,
+      "loss": 0.9522,
+      "step": 8309
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6943400586962621,
+      "learning_rate": 1.859310812768004e-05,
+      "loss": 0.5813,
+      "step": 8310
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5869648071483742,
+      "learning_rate": 1.8592717842461262e-05,
+      "loss": 0.7432,
+      "step": 8311
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.443058218576121,
+      "learning_rate": 1.859232750721296e-05,
+      "loss": 0.8172,
+      "step": 8312
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7886623504423589,
+      "learning_rate": 1.8591937121937408e-05,
+      "loss": 0.8118,
+      "step": 8313
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5470902287536679,
+      "learning_rate": 1.859154668663688e-05,
+      "loss": 0.7643,
+      "step": 8314
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5936723372274373,
+      "learning_rate": 1.8591156201313642e-05,
+      "loss": 0.8965,
+      "step": 8315
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5481715312820996,
+      "learning_rate": 1.8590765665969974e-05,
+      "loss": 0.8641,
+      "step": 8316
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6159942464820598,
+      "learning_rate": 1.859037508060815e-05,
+      "loss": 0.5304,
+      "step": 8317
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5757551252687636,
+      "learning_rate": 1.8589984445230444e-05,
+      "loss": 0.5626,
+      "step": 8318
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3858321140455585,
+      "learning_rate": 1.8589593759839125e-05,
+      "loss": 0.8433,
+      "step": 8319
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6866939604082667,
+      "learning_rate": 1.8589203024436477e-05,
+      "loss": 0.7663,
+      "step": 8320
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4177109494170352,
+      "learning_rate": 1.8588812239024766e-05,
+      "loss": 0.7814,
+      "step": 8321
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.519507111751232,
+      "learning_rate": 1.858842140360627e-05,
+      "loss": 0.8079,
+      "step": 8322
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.720159851698484,
+      "learning_rate": 1.8588030518183265e-05,
+      "loss": 0.719,
+      "step": 8323
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5440145203015618,
+      "learning_rate": 1.8587639582758032e-05,
+      "loss": 0.8183,
+      "step": 8324
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4700655312739328,
+      "learning_rate": 1.8587248597332837e-05,
+      "loss": 0.8115,
+      "step": 8325
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4954340933398116,
+      "learning_rate": 1.8586857561909967e-05,
+      "loss": 0.7207,
+      "step": 8326
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7323889199286,
+      "learning_rate": 1.8586466476491692e-05,
+      "loss": 0.6922,
+      "step": 8327
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.064763989074651,
+      "learning_rate": 1.8586075341080293e-05,
+      "loss": 0.8727,
+      "step": 8328
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4881373258856807,
+      "learning_rate": 1.858568415567804e-05,
+      "loss": 0.7876,
+      "step": 8329
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5994851562705795,
+      "learning_rate": 1.8585292920287217e-05,
+      "loss": 0.5454,
+      "step": 8330
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7325862777165615,
+      "learning_rate": 1.8584901634910105e-05,
+      "loss": 0.6928,
+      "step": 8331
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3926777346342338,
+      "learning_rate": 1.8584510299548975e-05,
+      "loss": 0.7792,
+      "step": 8332
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4797210496871487,
+      "learning_rate": 1.8584118914206105e-05,
+      "loss": 0.7859,
+      "step": 8333
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6036044216965805,
+      "learning_rate": 1.8583727478883782e-05,
+      "loss": 0.5163,
+      "step": 8334
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5352106145684954,
+      "learning_rate": 1.8583335993584275e-05,
+      "loss": 0.6816,
+      "step": 8335
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7747241719572846,
+      "learning_rate": 1.858294445830987e-05,
+      "loss": 0.8955,
+      "step": 8336
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7211345394442803,
+      "learning_rate": 1.858255287306285e-05,
+      "loss": 0.6372,
+      "step": 8337
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5886070415136984,
+      "learning_rate": 1.8582161237845486e-05,
+      "loss": 0.7895,
+      "step": 8338
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.994880777453333,
+      "learning_rate": 1.8581769552660064e-05,
+      "loss": 0.6928,
+      "step": 8339
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6970712763923739,
+      "learning_rate": 1.858137781750886e-05,
+      "loss": 0.8217,
+      "step": 8340
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7536290540849928,
+      "learning_rate": 1.8580986032394157e-05,
+      "loss": 0.8123,
+      "step": 8341
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6928256302752932,
+      "learning_rate": 1.858059419731824e-05,
+      "loss": 0.5961,
+      "step": 8342
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5616106600221569,
+      "learning_rate": 1.8580202312283383e-05,
+      "loss": 0.7442,
+      "step": 8343
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.3376694962922637,
+      "learning_rate": 1.8579810377291875e-05,
+      "loss": 0.7808,
+      "step": 8344
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5976432288368339,
+      "learning_rate": 1.8579418392345993e-05,
+      "loss": 0.8042,
+      "step": 8345
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8755128370235248,
+      "learning_rate": 1.8579026357448023e-05,
+      "loss": 0.8083,
+      "step": 8346
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.15532597842147,
+      "learning_rate": 1.8578634272600246e-05,
+      "loss": 0.8247,
+      "step": 8347
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.870590099245687,
+      "learning_rate": 1.8578242137804938e-05,
+      "loss": 0.8905,
+      "step": 8348
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.749528880797053,
+      "learning_rate": 1.8577849953064394e-05,
+      "loss": 0.7402,
+      "step": 8349
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.3626147079240716,
+      "learning_rate": 1.8577457718380893e-05,
+      "loss": 0.6958,
+      "step": 8350
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0544997872557937,
+      "learning_rate": 1.8577065433756712e-05,
+      "loss": 0.9076,
+      "step": 8351
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4862372097846572,
+      "learning_rate": 1.8576673099194145e-05,
+      "loss": 0.9283,
+      "step": 8352
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.646025185174464,
+      "learning_rate": 1.857628071469547e-05,
+      "loss": 0.8048,
+      "step": 8353
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6381536011465925,
+      "learning_rate": 1.8575888280262975e-05,
+      "loss": 0.8177,
+      "step": 8354
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9899951745635807,
+      "learning_rate": 1.8575495795898942e-05,
+      "loss": 0.7696,
+      "step": 8355
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5677711786379003,
+      "learning_rate": 1.8575103261605657e-05,
+      "loss": 0.932,
+      "step": 8356
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7222261172705504,
+      "learning_rate": 1.8574710677385412e-05,
+      "loss": 0.7355,
+      "step": 8357
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.78726562340693,
+      "learning_rate": 1.857431804324048e-05,
+      "loss": 0.8664,
+      "step": 8358
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.407566708696518,
+      "learning_rate": 1.8573925359173154e-05,
+      "loss": 0.7025,
+      "step": 8359
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0986602708960396,
+      "learning_rate": 1.8573532625185723e-05,
+      "loss": 0.8035,
+      "step": 8360
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4306868870113898,
+      "learning_rate": 1.857313984128047e-05,
+      "loss": 0.8774,
+      "step": 8361
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.5081725439624845,
+      "learning_rate": 1.8572747007459678e-05,
+      "loss": 0.7758,
+      "step": 8362
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5796431537882343,
+      "learning_rate": 1.8572354123725645e-05,
+      "loss": 0.9004,
+      "step": 8363
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6438127878400293,
+      "learning_rate": 1.857196119008065e-05,
+      "loss": 0.6369,
+      "step": 8364
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8299533037984754,
+      "learning_rate": 1.857156820652698e-05,
+      "loss": 0.8868,
+      "step": 8365
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5082967806519456,
+      "learning_rate": 1.8571175173066935e-05,
+      "loss": 0.7299,
+      "step": 8366
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3617579257318422,
+      "learning_rate": 1.8570782089702787e-05,
+      "loss": 0.7265,
+      "step": 8367
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5774761329524011,
+      "learning_rate": 1.8570388956436836e-05,
+      "loss": 0.53,
+      "step": 8368
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.393976078207383,
+      "learning_rate": 1.8569995773271366e-05,
+      "loss": 0.9011,
+      "step": 8369
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4768575593675426,
+      "learning_rate": 1.8569602540208664e-05,
+      "loss": 0.8506,
+      "step": 8370
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7369222090135823,
+      "learning_rate": 1.8569209257251028e-05,
+      "loss": 0.8362,
+      "step": 8371
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5425872868108161,
+      "learning_rate": 1.8568815924400742e-05,
+      "loss": 0.7338,
+      "step": 8372
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.431886620837422,
+      "learning_rate": 1.856842254166009e-05,
+      "loss": 0.6674,
+      "step": 8373
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4777985106469778,
+      "learning_rate": 1.8568029109031377e-05,
+      "loss": 0.765,
+      "step": 8374
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.413954314203057,
+      "learning_rate": 1.8567635626516885e-05,
+      "loss": 0.7256,
+      "step": 8375
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5828154267725365,
+      "learning_rate": 1.8567242094118902e-05,
+      "loss": 0.9066,
+      "step": 8376
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.742823139796624,
+      "learning_rate": 1.8566848511839724e-05,
+      "loss": 0.8031,
+      "step": 8377
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7325789297066894,
+      "learning_rate": 1.8566454879681646e-05,
+      "loss": 0.7165,
+      "step": 8378
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4068246373196631,
+      "learning_rate": 1.856606119764695e-05,
+      "loss": 0.7652,
+      "step": 8379
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4535160618422318,
+      "learning_rate": 1.856566746573794e-05,
+      "loss": 0.7832,
+      "step": 8380
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6709965530439408,
+      "learning_rate": 1.85652736839569e-05,
+      "loss": 0.8139,
+      "step": 8381
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.567911825375657,
+      "learning_rate": 1.8564879852306122e-05,
+      "loss": 0.9643,
+      "step": 8382
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.479925796869552,
+      "learning_rate": 1.8564485970787904e-05,
+      "loss": 0.6899,
+      "step": 8383
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.736749511868734,
+      "learning_rate": 1.8564092039404537e-05,
+      "loss": 0.752,
+      "step": 8384
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6211923269124435,
+      "learning_rate": 1.8563698058158313e-05,
+      "loss": 0.5622,
+      "step": 8385
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8407659972349093,
+      "learning_rate": 1.8563304027051532e-05,
+      "loss": 0.7067,
+      "step": 8386
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5905407858917358,
+      "learning_rate": 1.856290994608648e-05,
+      "loss": 0.7457,
+      "step": 8387
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6719345744773504,
+      "learning_rate": 1.8562515815265456e-05,
+      "loss": 0.8097,
+      "step": 8388
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6245607482050777,
+      "learning_rate": 1.8562121634590756e-05,
+      "loss": 0.7324,
+      "step": 8389
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.312163189388186,
+      "learning_rate": 1.856172740406467e-05,
+      "loss": 0.7608,
+      "step": 8390
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7725369174424848,
+      "learning_rate": 1.85613331236895e-05,
+      "loss": 0.7801,
+      "step": 8391
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7733696655808744,
+      "learning_rate": 1.8560938793467537e-05,
+      "loss": 0.6703,
+      "step": 8392
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.503531215197321,
+      "learning_rate": 1.8560544413401077e-05,
+      "loss": 0.8155,
+      "step": 8393
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4724430926720142,
+      "learning_rate": 1.8560149983492418e-05,
+      "loss": 0.7897,
+      "step": 8394
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2621946356048532,
+      "learning_rate": 1.855975550374385e-05,
+      "loss": 0.7944,
+      "step": 8395
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6120218242664954,
+      "learning_rate": 1.8559360974157683e-05,
+      "loss": 0.5748,
+      "step": 8396
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4003200942198348,
+      "learning_rate": 1.8558966394736205e-05,
+      "loss": 0.7825,
+      "step": 8397
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.945464463392347,
+      "learning_rate": 1.8558571765481712e-05,
+      "loss": 0.6738,
+      "step": 8398
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3419380373799445,
+      "learning_rate": 1.855817708639651e-05,
+      "loss": 0.7822,
+      "step": 8399
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5720223969765545,
+      "learning_rate": 1.8557782357482884e-05,
+      "loss": 0.748,
+      "step": 8400
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4855159103456796,
+      "learning_rate": 1.8557387578743146e-05,
+      "loss": 0.8377,
+      "step": 8401
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.206867803608103,
+      "learning_rate": 1.855699275017959e-05,
+      "loss": 0.708,
+      "step": 8402
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.569465594259629,
+      "learning_rate": 1.855659787179451e-05,
+      "loss": 0.6065,
+      "step": 8403
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.568641964064559,
+      "learning_rate": 1.8556202943590206e-05,
+      "loss": 0.7807,
+      "step": 8404
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7444605739450185,
+      "learning_rate": 1.855580796556898e-05,
+      "loss": 0.7468,
+      "step": 8405
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8129199996650975,
+      "learning_rate": 1.855541293773313e-05,
+      "loss": 0.7398,
+      "step": 8406
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.833158798350752,
+      "learning_rate": 1.855501786008496e-05,
+      "loss": 0.864,
+      "step": 8407
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.5130079356992483,
+      "learning_rate": 1.855462273262677e-05,
+      "loss": 0.6644,
+      "step": 8408
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.7092221817678928,
+      "learning_rate": 1.855422755536085e-05,
+      "loss": 0.7071,
+      "step": 8409
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.756591762346555,
+      "learning_rate": 1.855383232828952e-05,
+      "loss": 0.7577,
+      "step": 8410
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6549155273227214,
+      "learning_rate": 1.855343705141506e-05,
+      "loss": 0.7425,
+      "step": 8411
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5674755065776713,
+      "learning_rate": 1.855304172473979e-05,
+      "loss": 0.8249,
+      "step": 8412
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0373707343020144,
+      "learning_rate": 1.8552646348266e-05,
+      "loss": 0.7731,
+      "step": 8413
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8274473806644869,
+      "learning_rate": 1.8552250921995995e-05,
+      "loss": 0.722,
+      "step": 8414
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4413151677824358,
+      "learning_rate": 1.8551855445932077e-05,
+      "loss": 0.8411,
+      "step": 8415
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.9808833107133292,
+      "learning_rate": 1.8551459920076547e-05,
+      "loss": 0.7234,
+      "step": 8416
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5406995997326272,
+      "learning_rate": 1.8551064344431712e-05,
+      "loss": 0.6868,
+      "step": 8417
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.545316340060629,
+      "learning_rate": 1.8550668718999873e-05,
+      "loss": 0.7663,
+      "step": 8418
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9511781496690142,
+      "learning_rate": 1.8550273043783337e-05,
+      "loss": 0.7973,
+      "step": 8419
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4405848292273686,
+      "learning_rate": 1.85498773187844e-05,
+      "loss": 0.8072,
+      "step": 8420
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7375429023227837,
+      "learning_rate": 1.8549481544005373e-05,
+      "loss": 0.7565,
+      "step": 8421
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4142903294261455,
+      "learning_rate": 1.8549085719448553e-05,
+      "loss": 0.8073,
+      "step": 8422
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6444186582809637,
+      "learning_rate": 1.8548689845116255e-05,
+      "loss": 0.6796,
+      "step": 8423
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5646963651251427,
+      "learning_rate": 1.8548293921010776e-05,
+      "loss": 0.5478,
+      "step": 8424
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.362872574749432,
+      "learning_rate": 1.854789794713442e-05,
+      "loss": 0.7974,
+      "step": 8425
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6081578404744563,
+      "learning_rate": 1.8547501923489503e-05,
+      "loss": 0.8108,
+      "step": 8426
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7076143087827549,
+      "learning_rate": 1.854710585007832e-05,
+      "loss": 0.7727,
+      "step": 8427
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5067308640010058,
+      "learning_rate": 1.8546709726903178e-05,
+      "loss": 0.7595,
+      "step": 8428
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.488173057474593,
+      "learning_rate": 1.854631355396639e-05,
+      "loss": 0.907,
+      "step": 8429
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5696430217634721,
+      "learning_rate": 1.8545917331270257e-05,
+      "loss": 0.5187,
+      "step": 8430
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.840143429508426,
+      "learning_rate": 1.8545521058817088e-05,
+      "loss": 0.7547,
+      "step": 8431
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6383814772048235,
+      "learning_rate": 1.854512473660919e-05,
+      "loss": 0.6226,
+      "step": 8432
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5632087272384751,
+      "learning_rate": 1.8544728364648875e-05,
+      "loss": 0.5249,
+      "step": 8433
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7110368387691264,
+      "learning_rate": 1.8544331942938442e-05,
+      "loss": 0.809,
+      "step": 8434
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6918377023327604,
+      "learning_rate": 1.85439354714802e-05,
+      "loss": 0.7884,
+      "step": 8435
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4203375403847767,
+      "learning_rate": 1.8543538950276465e-05,
+      "loss": 0.7407,
+      "step": 8436
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7300186742829695,
+      "learning_rate": 1.8543142379329543e-05,
+      "loss": 0.8739,
+      "step": 8437
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7875305923832154,
+      "learning_rate": 1.8542745758641736e-05,
+      "loss": 0.8385,
+      "step": 8438
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.420296615948736,
+      "learning_rate": 1.8542349088215362e-05,
+      "loss": 0.8925,
+      "step": 8439
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.805944559263609,
+      "learning_rate": 1.8541952368052727e-05,
+      "loss": 0.7132,
+      "step": 8440
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9680562317087664,
+      "learning_rate": 1.8541555598156143e-05,
+      "loss": 0.8205,
+      "step": 8441
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4706967200593175,
+      "learning_rate": 1.8541158778527914e-05,
+      "loss": 0.8062,
+      "step": 8442
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7470800432976505,
+      "learning_rate": 1.8540761909170356e-05,
+      "loss": 0.7517,
+      "step": 8443
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.41850391735582,
+      "learning_rate": 1.8540364990085778e-05,
+      "loss": 0.7,
+      "step": 8444
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6271175978861667,
+      "learning_rate": 1.8539968021276493e-05,
+      "loss": 0.6844,
+      "step": 8445
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.60791630129975,
+      "learning_rate": 1.853957100274481e-05,
+      "loss": 0.7332,
+      "step": 8446
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1838981782751423,
+      "learning_rate": 1.853917393449304e-05,
+      "loss": 0.7771,
+      "step": 8447
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8428490656032763,
+      "learning_rate": 1.8538776816523495e-05,
+      "loss": 0.8328,
+      "step": 8448
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.631171899846778,
+      "learning_rate": 1.853837964883849e-05,
+      "loss": 0.6762,
+      "step": 8449
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.75775603625059,
+      "learning_rate": 1.8537982431440333e-05,
+      "loss": 0.7474,
+      "step": 8450
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3880094197808976,
+      "learning_rate": 1.8537585164331343e-05,
+      "loss": 0.6985,
+      "step": 8451
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.463086153495011,
+      "learning_rate": 1.8537187847513825e-05,
+      "loss": 0.8181,
+      "step": 8452
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6078958300568578,
+      "learning_rate": 1.85367904809901e-05,
+      "loss": 0.7327,
+      "step": 8453
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4251701654805697,
+      "learning_rate": 1.8536393064762474e-05,
+      "loss": 0.8479,
+      "step": 8454
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.623237822507077,
+      "learning_rate": 1.853599559883327e-05,
+      "loss": 0.8227,
+      "step": 8455
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7026536535042942,
+      "learning_rate": 1.8535598083204792e-05,
+      "loss": 0.7519,
+      "step": 8456
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5924573358672902,
+      "learning_rate": 1.8535200517879364e-05,
+      "loss": 0.7813,
+      "step": 8457
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8269488453993346,
+      "learning_rate": 1.8534802902859296e-05,
+      "loss": 0.7738,
+      "step": 8458
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9448242638612727,
+      "learning_rate": 1.85344052381469e-05,
+      "loss": 0.7783,
+      "step": 8459
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.584799697109192,
+      "learning_rate": 1.8534007523744498e-05,
+      "loss": 0.6668,
+      "step": 8460
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0417688857084983,
+      "learning_rate": 1.85336097596544e-05,
+      "loss": 0.7842,
+      "step": 8461
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8290397768585243,
+      "learning_rate": 1.8533211945878928e-05,
+      "loss": 0.7195,
+      "step": 8462
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5342866482048738,
+      "learning_rate": 1.8532814082420393e-05,
+      "loss": 0.6208,
+      "step": 8463
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6188964437456834,
+      "learning_rate": 1.8532416169281113e-05,
+      "loss": 0.7707,
+      "step": 8464
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0664375765455536,
+      "learning_rate": 1.8532018206463404e-05,
+      "loss": 0.8191,
+      "step": 8465
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5303370374153642,
+      "learning_rate": 1.853162019396958e-05,
+      "loss": 0.7345,
+      "step": 8466
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6018693024098151,
+      "learning_rate": 1.8531222131801965e-05,
+      "loss": 0.505,
+      "step": 8467
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6445014118087204,
+      "learning_rate": 1.8530824019962877e-05,
+      "loss": 0.7358,
+      "step": 8468
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6040891652985019,
+      "learning_rate": 1.8530425858454627e-05,
+      "loss": 0.7739,
+      "step": 8469
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8482630775196478,
+      "learning_rate": 1.853002764727954e-05,
+      "loss": 0.6532,
+      "step": 8470
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5525169390695566,
+      "learning_rate": 1.8529629386439933e-05,
+      "loss": 0.7917,
+      "step": 8471
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7025628285316734,
+      "learning_rate": 1.8529231075938117e-05,
+      "loss": 0.7223,
+      "step": 8472
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6902191539091145,
+      "learning_rate": 1.8528832715776422e-05,
+      "loss": 0.7908,
+      "step": 8473
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6232849704052935,
+      "learning_rate": 1.852843430595716e-05,
+      "loss": 0.5546,
+      "step": 8474
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4091146588501269,
+      "learning_rate": 1.8528035846482656e-05,
+      "loss": 0.733,
+      "step": 8475
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6111400437464407,
+      "learning_rate": 1.8527637337355225e-05,
+      "loss": 0.5367,
+      "step": 8476
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5494309066649004,
+      "learning_rate": 1.852723877857719e-05,
+      "loss": 0.8809,
+      "step": 8477
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7769411666478585,
+      "learning_rate": 1.8526840170150874e-05,
+      "loss": 0.7733,
+      "step": 8478
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.666577360724136,
+      "learning_rate": 1.852644151207859e-05,
+      "loss": 0.5932,
+      "step": 8479
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5944687486970417,
+      "learning_rate": 1.8526042804362666e-05,
+      "loss": 0.7949,
+      "step": 8480
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5140864562827663,
+      "learning_rate": 1.852564404700542e-05,
+      "loss": 0.7208,
+      "step": 8481
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.5703689444409923,
+      "learning_rate": 1.852524524000918e-05,
+      "loss": 0.6377,
+      "step": 8482
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6894805536892892,
+      "learning_rate": 1.852484638337626e-05,
+      "loss": 0.7461,
+      "step": 8483
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5626776503642892,
+      "learning_rate": 1.852444747710899e-05,
+      "loss": 0.7864,
+      "step": 8484
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9038928986329071,
+      "learning_rate": 1.852404852120968e-05,
+      "loss": 0.7895,
+      "step": 8485
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3337388636093037,
+      "learning_rate": 1.852364951568067e-05,
+      "loss": 0.7387,
+      "step": 8486
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4934454930772008,
+      "learning_rate": 1.8523250460524264e-05,
+      "loss": 0.8127,
+      "step": 8487
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3164656527025969,
+      "learning_rate": 1.85228513557428e-05,
+      "loss": 0.8187,
+      "step": 8488
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5520759613423962,
+      "learning_rate": 1.85224522013386e-05,
+      "loss": 0.8114,
+      "step": 8489
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3558780629807115,
+      "learning_rate": 1.8522052997313982e-05,
+      "loss": 0.7349,
+      "step": 8490
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3567234328172777,
+      "learning_rate": 1.8521653743671276e-05,
+      "loss": 0.7541,
+      "step": 8491
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.80755693704381,
+      "learning_rate": 1.8521254440412798e-05,
+      "loss": 0.8077,
+      "step": 8492
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4916593499287492,
+      "learning_rate": 1.8520855087540887e-05,
+      "loss": 0.7349,
+      "step": 8493
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5638547902944178,
+      "learning_rate": 1.8520455685057856e-05,
+      "loss": 0.6484,
+      "step": 8494
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5627093624595736,
+      "learning_rate": 1.8520056232966033e-05,
+      "loss": 0.7246,
+      "step": 8495
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0608087125994947,
+      "learning_rate": 1.8519656731267746e-05,
+      "loss": 0.776,
+      "step": 8496
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7170995620159257,
+      "learning_rate": 1.8519257179965322e-05,
+      "loss": 0.7164,
+      "step": 8497
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.579815385595484,
+      "learning_rate": 1.8518857579061083e-05,
+      "loss": 0.5267,
+      "step": 8498
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.868160674597584,
+      "learning_rate": 1.851845792855736e-05,
+      "loss": 0.7799,
+      "step": 8499
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.355172645648616,
+      "learning_rate": 1.8518058228456478e-05,
+      "loss": 0.8408,
+      "step": 8500
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.082350898228586,
+      "learning_rate": 1.851765847876076e-05,
+      "loss": 0.6527,
+      "step": 8501
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6146845549631494,
+      "learning_rate": 1.8517258679472546e-05,
+      "loss": 0.8048,
+      "step": 8502
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5263510978601817,
+      "learning_rate": 1.851685883059415e-05,
+      "loss": 0.7852,
+      "step": 8503
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9689237468059178,
+      "learning_rate": 1.8516458932127906e-05,
+      "loss": 0.8401,
+      "step": 8504
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9745582954433099,
+      "learning_rate": 1.8516058984076145e-05,
+      "loss": 0.7386,
+      "step": 8505
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.647897853757245,
+      "learning_rate": 1.851565898644119e-05,
+      "loss": 0.7004,
+      "step": 8506
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6972405119678449,
+      "learning_rate": 1.851525893922537e-05,
+      "loss": 0.8581,
+      "step": 8507
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.38143179940152,
+      "learning_rate": 1.8514858842431017e-05,
+      "loss": 0.6128,
+      "step": 8508
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.813218342893465,
+      "learning_rate": 1.8514458696060463e-05,
+      "loss": 0.7937,
+      "step": 8509
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3776324477568616,
+      "learning_rate": 1.8514058500116033e-05,
+      "loss": 0.6656,
+      "step": 8510
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3769784245086005,
+      "learning_rate": 1.851365825460006e-05,
+      "loss": 0.7162,
+      "step": 8511
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.7613320696960657,
+      "learning_rate": 1.851325795951487e-05,
+      "loss": 0.6716,
+      "step": 8512
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.591120083655021,
+      "learning_rate": 1.85128576148628e-05,
+      "loss": 0.6843,
+      "step": 8513
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.502054508381438,
+      "learning_rate": 1.851245722064618e-05,
+      "loss": 0.6388,
+      "step": 8514
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7200981509346813,
+      "learning_rate": 1.8512056776867335e-05,
+      "loss": 0.74,
+      "step": 8515
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.841299659359234,
+      "learning_rate": 1.8511656283528603e-05,
+      "loss": 0.6192,
+      "step": 8516
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7176597260334192,
+      "learning_rate": 1.851125574063231e-05,
+      "loss": 0.6736,
+      "step": 8517
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.717505802139234,
+      "learning_rate": 1.8510855148180794e-05,
+      "loss": 0.8263,
+      "step": 8518
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7875690321071755,
+      "learning_rate": 1.851045450617639e-05,
+      "loss": 0.7558,
+      "step": 8519
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4436516741953025,
+      "learning_rate": 1.8510053814621417e-05,
+      "loss": 0.7212,
+      "step": 8520
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6635189958707735,
+      "learning_rate": 1.850965307351822e-05,
+      "loss": 0.8168,
+      "step": 8521
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7102686990756009,
+      "learning_rate": 1.8509252282869127e-05,
+      "loss": 0.8535,
+      "step": 8522
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4743955231822072,
+      "learning_rate": 1.8508851442676475e-05,
+      "loss": 0.7605,
+      "step": 8523
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.324860386321777,
+      "learning_rate": 1.8508450552942597e-05,
+      "loss": 0.8001,
+      "step": 8524
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.46684637254844,
+      "learning_rate": 1.8508049613669824e-05,
+      "loss": 0.8197,
+      "step": 8525
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8124316091078294,
+      "learning_rate": 1.850764862486049e-05,
+      "loss": 0.7524,
+      "step": 8526
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6877398618286306,
+      "learning_rate": 1.850724758651694e-05,
+      "loss": 0.8771,
+      "step": 8527
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6113952271494731,
+      "learning_rate": 1.8506846498641496e-05,
+      "loss": 0.5155,
+      "step": 8528
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.277331446444745,
+      "learning_rate": 1.8506445361236502e-05,
+      "loss": 0.81,
+      "step": 8529
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7475901766402164,
+      "learning_rate": 1.8506044174304282e-05,
+      "loss": 0.6896,
+      "step": 8530
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3539480634133287,
+      "learning_rate": 1.8505642937847188e-05,
+      "loss": 0.8948,
+      "step": 8531
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4554738642832246,
+      "learning_rate": 1.8505241651867543e-05,
+      "loss": 0.6065,
+      "step": 8532
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8691409159541028,
+      "learning_rate": 1.8504840316367692e-05,
+      "loss": 0.8185,
+      "step": 8533
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.500900507877349,
+      "learning_rate": 1.8504438931349967e-05,
+      "loss": 0.7722,
+      "step": 8534
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7554574234257097,
+      "learning_rate": 1.8504037496816708e-05,
+      "loss": 0.7555,
+      "step": 8535
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4441373153632606,
+      "learning_rate": 1.8503636012770247e-05,
+      "loss": 0.7035,
+      "step": 8536
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4893166237205553,
+      "learning_rate": 1.8503234479212924e-05,
+      "loss": 0.7447,
+      "step": 8537
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.941882565869204,
+      "learning_rate": 1.850283289614708e-05,
+      "loss": 0.8074,
+      "step": 8538
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.274172929370456,
+      "learning_rate": 1.850243126357505e-05,
+      "loss": 0.7156,
+      "step": 8539
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0136549632022382,
+      "learning_rate": 1.8502029581499173e-05,
+      "loss": 0.7846,
+      "step": 8540
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5929490480261634,
+      "learning_rate": 1.850162784992179e-05,
+      "loss": 0.8082,
+      "step": 8541
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4672131027994315,
+      "learning_rate": 1.8501226068845238e-05,
+      "loss": 0.8493,
+      "step": 8542
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.709747562161606,
+      "learning_rate": 1.8500824238271855e-05,
+      "loss": 0.8158,
+      "step": 8543
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.390016234348324,
+      "learning_rate": 1.8500422358203982e-05,
+      "loss": 0.6509,
+      "step": 8544
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.548845953084473,
+      "learning_rate": 1.850002042864396e-05,
+      "loss": 0.7457,
+      "step": 8545
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7958813447415,
+      "learning_rate": 1.849961844959413e-05,
+      "loss": 0.7669,
+      "step": 8546
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3488287758746234,
+      "learning_rate": 1.8499216421056823e-05,
+      "loss": 0.706,
+      "step": 8547
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8732360351691237,
+      "learning_rate": 1.8498814343034396e-05,
+      "loss": 0.8393,
+      "step": 8548
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0367382203683135,
+      "learning_rate": 1.8498412215529175e-05,
+      "loss": 0.7106,
+      "step": 8549
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6176229914480411,
+      "learning_rate": 1.849801003854351e-05,
+      "loss": 0.5387,
+      "step": 8550
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7755871244922399,
+      "learning_rate": 1.849760781207974e-05,
+      "loss": 0.753,
+      "step": 8551
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.488983393873542,
+      "learning_rate": 1.8497205536140207e-05,
+      "loss": 0.8119,
+      "step": 8552
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5669360992546453,
+      "learning_rate": 1.8496803210727252e-05,
+      "loss": 0.7167,
+      "step": 8553
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3545183534821945,
+      "learning_rate": 1.8496400835843223e-05,
+      "loss": 0.8036,
+      "step": 8554
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3928205661341224,
+      "learning_rate": 1.849599841149045e-05,
+      "loss": 0.7466,
+      "step": 8555
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3368864682797568,
+      "learning_rate": 1.8495595937671293e-05,
+      "loss": 0.8425,
+      "step": 8556
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7762010795699776,
+      "learning_rate": 1.8495193414388084e-05,
+      "loss": 0.764,
+      "step": 8557
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6710024088374664,
+      "learning_rate": 1.849479084164317e-05,
+      "loss": 0.7589,
+      "step": 8558
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9141707520578826,
+      "learning_rate": 1.8494388219438892e-05,
+      "loss": 0.7068,
+      "step": 8559
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5700112271305804,
+      "learning_rate": 1.8493985547777598e-05,
+      "loss": 0.8566,
+      "step": 8560
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4251589551584507,
+      "learning_rate": 1.8493582826661628e-05,
+      "loss": 0.8507,
+      "step": 8561
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5958465534683353,
+      "learning_rate": 1.8493180056093334e-05,
+      "loss": 0.5457,
+      "step": 8562
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7671172302827074,
+      "learning_rate": 1.8492777236075056e-05,
+      "loss": 0.6302,
+      "step": 8563
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.506830452616911,
+      "learning_rate": 1.849237436660914e-05,
+      "loss": 0.6963,
+      "step": 8564
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4062012218067987,
+      "learning_rate": 1.849197144769793e-05,
+      "loss": 0.7464,
+      "step": 8565
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6015113908183944,
+      "learning_rate": 1.8491568479343775e-05,
+      "loss": 0.722,
+      "step": 8566
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5096721723957465,
+      "learning_rate": 1.849116546154902e-05,
+      "loss": 0.8812,
+      "step": 8567
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.358109125521107,
+      "learning_rate": 1.849076239431601e-05,
+      "loss": 0.6988,
+      "step": 8568
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4158268369517146,
+      "learning_rate": 1.8490359277647096e-05,
+      "loss": 0.7468,
+      "step": 8569
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7859046153088913,
+      "learning_rate": 1.8489956111544624e-05,
+      "loss": 0.7603,
+      "step": 8570
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4592751357346903,
+      "learning_rate": 1.8489552896010932e-05,
+      "loss": 0.6444,
+      "step": 8571
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8925205885968328,
+      "learning_rate": 1.8489149631048378e-05,
+      "loss": 0.7849,
+      "step": 8572
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.588687805529893,
+      "learning_rate": 1.848874631665931e-05,
+      "loss": 0.8313,
+      "step": 8573
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7077373391664612,
+      "learning_rate": 1.8488342952846074e-05,
+      "loss": 0.7289,
+      "step": 8574
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.03731373870377,
+      "learning_rate": 1.8487939539611016e-05,
+      "loss": 0.7835,
+      "step": 8575
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.555390824620319,
+      "learning_rate": 1.8487536076956488e-05,
+      "loss": 0.7479,
+      "step": 8576
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3498614790580294,
+      "learning_rate": 1.8487132564884835e-05,
+      "loss": 0.6985,
+      "step": 8577
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6244047428954003,
+      "learning_rate": 1.848672900339841e-05,
+      "loss": 0.8956,
+      "step": 8578
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3322972785115716,
+      "learning_rate": 1.848632539249957e-05,
+      "loss": 0.7667,
+      "step": 8579
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6263442765739959,
+      "learning_rate": 1.8485921732190644e-05,
+      "loss": 0.5266,
+      "step": 8580
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9681172212220086,
+      "learning_rate": 1.8485518022474004e-05,
+      "loss": 0.7645,
+      "step": 8581
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4240255115224745,
+      "learning_rate": 1.8485114263351988e-05,
+      "loss": 0.794,
+      "step": 8582
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4649269993391778,
+      "learning_rate": 1.8484710454826952e-05,
+      "loss": 0.7984,
+      "step": 8583
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6249707153813258,
+      "learning_rate": 1.8484306596901243e-05,
+      "loss": 0.84,
+      "step": 8584
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7588681767357204,
+      "learning_rate": 1.8483902689577217e-05,
+      "loss": 0.7258,
+      "step": 8585
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.84228400154985,
+      "learning_rate": 1.8483498732857224e-05,
+      "loss": 0.613,
+      "step": 8586
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5837152583727317,
+      "learning_rate": 1.8483094726743613e-05,
+      "loss": 0.5458,
+      "step": 8587
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.759706617033123,
+      "learning_rate": 1.8482690671238738e-05,
+      "loss": 0.5882,
+      "step": 8588
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5617311604364605,
+      "learning_rate": 1.8482286566344957e-05,
+      "loss": 0.7233,
+      "step": 8589
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3813472440950918,
+      "learning_rate": 1.8481882412064616e-05,
+      "loss": 0.7799,
+      "step": 8590
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0772447777532275,
+      "learning_rate": 1.848147820840007e-05,
+      "loss": 0.7981,
+      "step": 8591
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8193667290773081,
+      "learning_rate": 1.8481073955353673e-05,
+      "loss": 0.6556,
+      "step": 8592
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6495287192801318,
+      "learning_rate": 1.8480669652927778e-05,
+      "loss": 0.7765,
+      "step": 8593
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5446209554183625,
+      "learning_rate": 1.848026530112474e-05,
+      "loss": 0.6365,
+      "step": 8594
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.559756020493595,
+      "learning_rate": 1.8479860899946912e-05,
+      "loss": 0.6856,
+      "step": 8595
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.756246962401961,
+      "learning_rate": 1.847945644939665e-05,
+      "loss": 0.5979,
+      "step": 8596
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.825686248968407,
+      "learning_rate": 1.8479051949476307e-05,
+      "loss": 0.7712,
+      "step": 8597
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1816276762509998,
+      "learning_rate": 1.847864740018824e-05,
+      "loss": 0.8339,
+      "step": 8598
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5846772867444342,
+      "learning_rate": 1.8478242801534803e-05,
+      "loss": 0.5735,
+      "step": 8599
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5976433368356039,
+      "learning_rate": 1.8477838153518354e-05,
+      "loss": 0.5414,
+      "step": 8600
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6422766956389825,
+      "learning_rate": 1.8477433456141242e-05,
+      "loss": 0.8156,
+      "step": 8601
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8738850547801047,
+      "learning_rate": 1.8477028709405834e-05,
+      "loss": 0.7494,
+      "step": 8602
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2686521696632238,
+      "learning_rate": 1.8476623913314476e-05,
+      "loss": 0.7219,
+      "step": 8603
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5899995236110354,
+      "learning_rate": 1.8476219067869534e-05,
+      "loss": 0.7789,
+      "step": 8604
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6801989505900983,
+      "learning_rate": 1.847581417307336e-05,
+      "loss": 0.6918,
+      "step": 8605
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.497718553462571,
+      "learning_rate": 1.8475409228928314e-05,
+      "loss": 0.7612,
+      "step": 8606
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.618099557028153,
+      "learning_rate": 1.847500423543675e-05,
+      "loss": 0.7957,
+      "step": 8607
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5973273747055569,
+      "learning_rate": 1.847459919260103e-05,
+      "loss": 0.5414,
+      "step": 8608
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6088988376943913,
+      "learning_rate": 1.8474194100423508e-05,
+      "loss": 0.7272,
+      "step": 8609
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3487660001091502,
+      "learning_rate": 1.8473788958906548e-05,
+      "loss": 0.7432,
+      "step": 8610
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4729411587486407,
+      "learning_rate": 1.8473383768052504e-05,
+      "loss": 0.6212,
+      "step": 8611
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8395987186749307,
+      "learning_rate": 1.8472978527863737e-05,
+      "loss": 0.7372,
+      "step": 8612
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5907946116495475,
+      "learning_rate": 1.8472573238342606e-05,
+      "loss": 0.7998,
+      "step": 8613
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5429177198278428,
+      "learning_rate": 1.8472167899491472e-05,
+      "loss": 0.9015,
+      "step": 8614
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4451087107835439,
+      "learning_rate": 1.8471762511312693e-05,
+      "loss": 0.8357,
+      "step": 8615
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.450592340993747,
+      "learning_rate": 1.8471357073808635e-05,
+      "loss": 0.8428,
+      "step": 8616
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.467433411803132,
+      "learning_rate": 1.8470951586981648e-05,
+      "loss": 0.9301,
+      "step": 8617
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3252695148671074,
+      "learning_rate": 1.8470546050834104e-05,
+      "loss": 0.7846,
+      "step": 8618
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.8726225600880633,
+      "learning_rate": 1.8470140465368355e-05,
+      "loss": 0.8203,
+      "step": 8619
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3893455161513892,
+      "learning_rate": 1.8469734830586773e-05,
+      "loss": 0.813,
+      "step": 8620
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1585742207623277,
+      "learning_rate": 1.8469329146491708e-05,
+      "loss": 0.6734,
+      "step": 8621
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7410909736544067,
+      "learning_rate": 1.846892341308553e-05,
+      "loss": 0.7077,
+      "step": 8622
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.5949875217527603,
+      "learning_rate": 1.8468517630370594e-05,
+      "loss": 0.6423,
+      "step": 8623
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.495330937449576,
+      "learning_rate": 1.846811179834927e-05,
+      "loss": 0.9316,
+      "step": 8624
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.918383657864159,
+      "learning_rate": 1.846770591702392e-05,
+      "loss": 0.7928,
+      "step": 8625
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0047873595005568,
+      "learning_rate": 1.8467299986396903e-05,
+      "loss": 0.8614,
+      "step": 8626
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.458587066856993,
+      "learning_rate": 1.8466894006470588e-05,
+      "loss": 0.6444,
+      "step": 8627
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3572139552557982,
+      "learning_rate": 1.846648797724733e-05,
+      "loss": 0.6564,
+      "step": 8628
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6307979623879243,
+      "learning_rate": 1.8466081898729507e-05,
+      "loss": 0.5557,
+      "step": 8629
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4813449127257017,
+      "learning_rate": 1.8465675770919468e-05,
+      "loss": 0.8682,
+      "step": 8630
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4183449039977458,
+      "learning_rate": 1.8465269593819586e-05,
+      "loss": 0.7486,
+      "step": 8631
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7519637596540265,
+      "learning_rate": 1.8464863367432225e-05,
+      "loss": 0.6963,
+      "step": 8632
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7418742334735398,
+      "learning_rate": 1.8464457091759748e-05,
+      "loss": 0.7098,
+      "step": 8633
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7386878281675677,
+      "learning_rate": 1.8464050766804526e-05,
+      "loss": 0.7919,
+      "step": 8634
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6741782822327747,
+      "learning_rate": 1.8463644392568916e-05,
+      "loss": 0.6351,
+      "step": 8635
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.845750632247062,
+      "learning_rate": 1.846323796905529e-05,
+      "loss": 0.8146,
+      "step": 8636
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8552697070078787,
+      "learning_rate": 1.8462831496266013e-05,
+      "loss": 0.8003,
+      "step": 8637
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4224419174914085,
+      "learning_rate": 1.8462424974203452e-05,
+      "loss": 0.7218,
+      "step": 8638
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6610668109761384,
+      "learning_rate": 1.8462018402869977e-05,
+      "loss": 0.774,
+      "step": 8639
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6060881782954286,
+      "learning_rate": 1.8461611782267947e-05,
+      "loss": 0.7353,
+      "step": 8640
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.353110030633092,
+      "learning_rate": 1.8461205112399736e-05,
+      "loss": 0.7719,
+      "step": 8641
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5323970552539916,
+      "learning_rate": 1.8460798393267713e-05,
+      "loss": 0.6595,
+      "step": 8642
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.992255459764633,
+      "learning_rate": 1.846039162487424e-05,
+      "loss": 0.6845,
+      "step": 8643
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4630583061346356,
+      "learning_rate": 1.8459984807221686e-05,
+      "loss": 0.7931,
+      "step": 8644
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4922834146777926,
+      "learning_rate": 1.845957794031243e-05,
+      "loss": 0.6859,
+      "step": 8645
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6703685957802432,
+      "learning_rate": 1.8459171024148826e-05,
+      "loss": 0.6748,
+      "step": 8646
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5605691757431923,
+      "learning_rate": 1.845876405873325e-05,
+      "loss": 0.8252,
+      "step": 8647
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.140424578374548,
+      "learning_rate": 1.8458357044068073e-05,
+      "loss": 0.7995,
+      "step": 8648
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8740852773469292,
+      "learning_rate": 1.8457949980155666e-05,
+      "loss": 0.8478,
+      "step": 8649
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4068379450025328,
+      "learning_rate": 1.8457542866998394e-05,
+      "loss": 0.7803,
+      "step": 8650
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5105177070216986,
+      "learning_rate": 1.8457135704598634e-05,
+      "loss": 0.7794,
+      "step": 8651
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7848723073936341,
+      "learning_rate": 1.8456728492958748e-05,
+      "loss": 0.7607,
+      "step": 8652
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0983768681387813,
+      "learning_rate": 1.845632123208111e-05,
+      "loss": 0.7608,
+      "step": 8653
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6277056283936486,
+      "learning_rate": 1.8455913921968095e-05,
+      "loss": 0.7577,
+      "step": 8654
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5255269914603415,
+      "learning_rate": 1.845550656262207e-05,
+      "loss": 0.7736,
+      "step": 8655
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6226799534250367,
+      "learning_rate": 1.8455099154045416e-05,
+      "loss": 0.8634,
+      "step": 8656
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5844820192082248,
+      "learning_rate": 1.8454691696240493e-05,
+      "loss": 0.5252,
+      "step": 8657
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5901260537041597,
+      "learning_rate": 1.8454284189209674e-05,
+      "loss": 0.7452,
+      "step": 8658
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8290590351735492,
+      "learning_rate": 1.8453876632955342e-05,
+      "loss": 0.731,
+      "step": 8659
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8348092404697385,
+      "learning_rate": 1.845346902747986e-05,
+      "loss": 0.7282,
+      "step": 8660
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.592869069120789,
+      "learning_rate": 1.8453061372785605e-05,
+      "loss": 0.5595,
+      "step": 8661
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6151964667938925,
+      "learning_rate": 1.8452653668874954e-05,
+      "loss": 0.7723,
+      "step": 8662
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2741154728850583,
+      "learning_rate": 1.8452245915750276e-05,
+      "loss": 0.7555,
+      "step": 8663
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5702916872675257,
+      "learning_rate": 1.8451838113413946e-05,
+      "loss": 0.8048,
+      "step": 8664
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.715835887840482,
+      "learning_rate": 1.8451430261868336e-05,
+      "loss": 0.7502,
+      "step": 8665
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6218319802935933,
+      "learning_rate": 1.8451022361115828e-05,
+      "loss": 0.5562,
+      "step": 8666
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7188704985964334,
+      "learning_rate": 1.845061441115879e-05,
+      "loss": 0.7783,
+      "step": 8667
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9074758705576702,
+      "learning_rate": 1.84502064119996e-05,
+      "loss": 0.8536,
+      "step": 8668
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6348524587567073,
+      "learning_rate": 1.844979836364063e-05,
+      "loss": 0.8254,
+      "step": 8669
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8745899926709748,
+      "learning_rate": 1.8449390266084263e-05,
+      "loss": 0.7099,
+      "step": 8670
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6146347916710062,
+      "learning_rate": 1.8448982119332868e-05,
+      "loss": 0.532,
+      "step": 8671
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4296726408797615,
+      "learning_rate": 1.8448573923388825e-05,
+      "loss": 0.8836,
+      "step": 8672
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4981537242107794,
+      "learning_rate": 1.844816567825451e-05,
+      "loss": 0.8271,
+      "step": 8673
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.9206540148602658,
+      "learning_rate": 1.8447757383932302e-05,
+      "loss": 0.7102,
+      "step": 8674
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5594280854227007,
+      "learning_rate": 1.8447349040424572e-05,
+      "loss": 0.7092,
+      "step": 8675
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6806026956444644,
+      "learning_rate": 1.8446940647733705e-05,
+      "loss": 0.8295,
+      "step": 8676
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8641237878953207,
+      "learning_rate": 1.8446532205862072e-05,
+      "loss": 0.6753,
+      "step": 8677
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6042029729562036,
+      "learning_rate": 1.8446123714812055e-05,
+      "loss": 0.5415,
+      "step": 8678
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6633317745249923,
+      "learning_rate": 1.8445715174586033e-05,
+      "loss": 0.7218,
+      "step": 8679
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.563051008800868,
+      "learning_rate": 1.8445306585186384e-05,
+      "loss": 0.5196,
+      "step": 8680
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5821868932697514,
+      "learning_rate": 1.8444897946615486e-05,
+      "loss": 0.5071,
+      "step": 8681
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7533707406699517,
+      "learning_rate": 1.8444489258875718e-05,
+      "loss": 0.7126,
+      "step": 8682
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5698720262902236,
+      "learning_rate": 1.844408052196946e-05,
+      "loss": 0.5513,
+      "step": 8683
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 10.777234572846261,
+      "learning_rate": 1.8443671735899092e-05,
+      "loss": 0.7608,
+      "step": 8684
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7635626466506253,
+      "learning_rate": 1.8443262900666995e-05,
+      "loss": 0.7728,
+      "step": 8685
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5940478208473488,
+      "learning_rate": 1.8442854016275545e-05,
+      "loss": 0.8014,
+      "step": 8686
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4722101041036868,
+      "learning_rate": 1.8442445082727127e-05,
+      "loss": 0.7843,
+      "step": 8687
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6223988476145444,
+      "learning_rate": 1.8442036100024124e-05,
+      "loss": 0.5629,
+      "step": 8688
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.822369434049088,
+      "learning_rate": 1.8441627068168914e-05,
+      "loss": 0.7066,
+      "step": 8689
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3651054079281868,
+      "learning_rate": 1.8441217987163874e-05,
+      "loss": 0.6291,
+      "step": 8690
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.792653497569738,
+      "learning_rate": 1.8440808857011392e-05,
+      "loss": 0.8629,
+      "step": 8691
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6808661433278842,
+      "learning_rate": 1.844039967771385e-05,
+      "loss": 0.7762,
+      "step": 8692
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8455838947348724,
+      "learning_rate": 1.843999044927363e-05,
+      "loss": 0.7772,
+      "step": 8693
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9885551261537353,
+      "learning_rate": 1.843958117169311e-05,
+      "loss": 0.7409,
+      "step": 8694
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5257364628204004,
+      "learning_rate": 1.843917184497468e-05,
+      "loss": 0.807,
+      "step": 8695
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.403498743795569,
+      "learning_rate": 1.843876246912072e-05,
+      "loss": 0.8108,
+      "step": 8696
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5645795645682188,
+      "learning_rate": 1.843835304413361e-05,
+      "loss": 0.8477,
+      "step": 8697
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6279913400555599,
+      "learning_rate": 1.8437943570015737e-05,
+      "loss": 0.5624,
+      "step": 8698
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4290435265622976,
+      "learning_rate": 1.843753404676949e-05,
+      "loss": 0.7883,
+      "step": 8699
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5355126168190756,
+      "learning_rate": 1.8437124474397247e-05,
+      "loss": 0.7698,
+      "step": 8700
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4854773370634073,
+      "learning_rate": 1.843671485290139e-05,
+      "loss": 0.8719,
+      "step": 8701
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.751034461247019,
+      "learning_rate": 1.843630518228431e-05,
+      "loss": 0.7318,
+      "step": 8702
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5287780268557087,
+      "learning_rate": 1.843589546254839e-05,
+      "loss": 0.7582,
+      "step": 8703
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8191370240388411,
+      "learning_rate": 1.8435485693696015e-05,
+      "loss": 0.7921,
+      "step": 8704
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4655356298096305,
+      "learning_rate": 1.8435075875729577e-05,
+      "loss": 0.84,
+      "step": 8705
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6001449215751621,
+      "learning_rate": 1.843466600865145e-05,
+      "loss": 0.5014,
+      "step": 8706
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4770176337104879,
+      "learning_rate": 1.843425609246403e-05,
+      "loss": 0.7722,
+      "step": 8707
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.341438606378244,
+      "learning_rate": 1.8433846127169704e-05,
+      "loss": 0.6414,
+      "step": 8708
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5917614241248776,
+      "learning_rate": 1.843343611277085e-05,
+      "loss": 0.529,
+      "step": 8709
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.459399355289646,
+      "learning_rate": 1.8433026049269866e-05,
+      "loss": 0.7548,
+      "step": 8710
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.768229964825935,
+      "learning_rate": 1.8432615936669133e-05,
+      "loss": 0.7442,
+      "step": 8711
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4472667968019062,
+      "learning_rate": 1.8432205774971036e-05,
+      "loss": 0.6524,
+      "step": 8712
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6366266836728716,
+      "learning_rate": 1.8431795564177975e-05,
+      "loss": 0.5425,
+      "step": 8713
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5310220405771022,
+      "learning_rate": 1.8431385304292324e-05,
+      "loss": 0.73,
+      "step": 8714
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8989334894299361,
+      "learning_rate": 1.8430974995316482e-05,
+      "loss": 0.8994,
+      "step": 8715
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4163213161988688,
+      "learning_rate": 1.8430564637252835e-05,
+      "loss": 0.7441,
+      "step": 8716
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4987539018965719,
+      "learning_rate": 1.843015423010377e-05,
+      "loss": 0.7765,
+      "step": 8717
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6182419344184025,
+      "learning_rate": 1.8429743773871678e-05,
+      "loss": 0.7459,
+      "step": 8718
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3133496136609748,
+      "learning_rate": 1.842933326855895e-05,
+      "loss": 0.6855,
+      "step": 8719
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9067136382257897,
+      "learning_rate": 1.842892271416797e-05,
+      "loss": 0.8039,
+      "step": 8720
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6861734031760014,
+      "learning_rate": 1.842851211070114e-05,
+      "loss": 0.8902,
+      "step": 8721
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8314015342667214,
+      "learning_rate": 1.8428101458160843e-05,
+      "loss": 0.8276,
+      "step": 8722
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3849969052931934,
+      "learning_rate": 1.842769075654947e-05,
+      "loss": 0.7967,
+      "step": 8723
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.82144562996526,
+      "learning_rate": 1.842728000586941e-05,
+      "loss": 0.7382,
+      "step": 8724
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.607299893322642,
+      "learning_rate": 1.8426869206123062e-05,
+      "loss": 0.6476,
+      "step": 8725
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.325923618916826,
+      "learning_rate": 1.8426458357312814e-05,
+      "loss": 0.7462,
+      "step": 8726
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6026414317696092,
+      "learning_rate": 1.8426047459441053e-05,
+      "loss": 0.6437,
+      "step": 8727
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.459929453010345,
+      "learning_rate": 1.842563651251018e-05,
+      "loss": 0.8014,
+      "step": 8728
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7926377280086707,
+      "learning_rate": 1.842522551652258e-05,
+      "loss": 0.7399,
+      "step": 8729
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.471882542120367,
+      "learning_rate": 1.8424814471480658e-05,
+      "loss": 0.7505,
+      "step": 8730
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.27663098829182,
+      "learning_rate": 1.842440337738679e-05,
+      "loss": 0.7736,
+      "step": 8731
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4506088762851477,
+      "learning_rate": 1.842399223424338e-05,
+      "loss": 0.7331,
+      "step": 8732
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5354817535386776,
+      "learning_rate": 1.8423581042052823e-05,
+      "loss": 0.7111,
+      "step": 8733
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5464135486249484,
+      "learning_rate": 1.842316980081751e-05,
+      "loss": 0.7661,
+      "step": 8734
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.425745708674981,
+      "learning_rate": 1.8422758510539837e-05,
+      "loss": 0.6615,
+      "step": 8735
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1182230262405475,
+      "learning_rate": 1.8422347171222194e-05,
+      "loss": 0.6045,
+      "step": 8736
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6188496352373378,
+      "learning_rate": 1.8421935782866982e-05,
+      "loss": 0.7849,
+      "step": 8737
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7068049628976376,
+      "learning_rate": 1.842152434547659e-05,
+      "loss": 0.83,
+      "step": 8738
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7181032062497412,
+      "learning_rate": 1.8421112859053417e-05,
+      "loss": 0.6401,
+      "step": 8739
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.984972067200197,
+      "learning_rate": 1.8420701323599865e-05,
+      "loss": 0.815,
+      "step": 8740
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6021104882453311,
+      "learning_rate": 1.8420289739118318e-05,
+      "loss": 0.5247,
+      "step": 8741
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3740954983354925,
+      "learning_rate": 1.841987810561118e-05,
+      "loss": 0.7166,
+      "step": 8742
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5010870907263687,
+      "learning_rate": 1.8419466423080847e-05,
+      "loss": 0.7379,
+      "step": 8743
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6728114115308812,
+      "learning_rate": 1.8419054691529715e-05,
+      "loss": 0.8243,
+      "step": 8744
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5968332224486304,
+      "learning_rate": 1.841864291096018e-05,
+      "loss": 0.8089,
+      "step": 8745
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6290572125647995,
+      "learning_rate": 1.841823108137464e-05,
+      "loss": 0.8089,
+      "step": 8746
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6179819970651987,
+      "learning_rate": 1.8417819202775497e-05,
+      "loss": 0.5483,
+      "step": 8747
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4157070259267472,
+      "learning_rate": 1.841740727516514e-05,
+      "loss": 0.8356,
+      "step": 8748
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4740924346835824,
+      "learning_rate": 1.8416995298545978e-05,
+      "loss": 0.8566,
+      "step": 8749
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6624564955004824,
+      "learning_rate": 1.84165832729204e-05,
+      "loss": 0.7704,
+      "step": 8750
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7055197977330199,
+      "learning_rate": 1.8416171198290812e-05,
+      "loss": 0.7232,
+      "step": 8751
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6989843274652747,
+      "learning_rate": 1.8415759074659614e-05,
+      "loss": 0.8689,
+      "step": 8752
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5754559575397482,
+      "learning_rate": 1.8415346902029196e-05,
+      "loss": 0.6411,
+      "step": 8753
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8498392779816963,
+      "learning_rate": 1.8414934680401968e-05,
+      "loss": 0.7054,
+      "step": 8754
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7751355136516724,
+      "learning_rate": 1.8414522409780328e-05,
+      "loss": 0.7023,
+      "step": 8755
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.564590982865808,
+      "learning_rate": 1.8414110090166672e-05,
+      "loss": 0.8573,
+      "step": 8756
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6027641728968145,
+      "learning_rate": 1.84136977215634e-05,
+      "loss": 0.803,
+      "step": 8757
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6003484186327195,
+      "learning_rate": 1.8413285303972922e-05,
+      "loss": 0.7391,
+      "step": 8758
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.0143802723033803,
+      "learning_rate": 1.8412872837397627e-05,
+      "loss": 0.7836,
+      "step": 8759
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4391908870271821,
+      "learning_rate": 1.841246032183993e-05,
+      "loss": 0.7189,
+      "step": 8760
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7273571127893774,
+      "learning_rate": 1.8412047757302222e-05,
+      "loss": 0.7698,
+      "step": 8761
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.726794575740867,
+      "learning_rate": 1.8411635143786906e-05,
+      "loss": 0.7202,
+      "step": 8762
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5996610465709578,
+      "learning_rate": 1.841122248129639e-05,
+      "loss": 0.7442,
+      "step": 8763
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5234184642016635,
+      "learning_rate": 1.8410809769833075e-05,
+      "loss": 0.7185,
+      "step": 8764
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7629289070664353,
+      "learning_rate": 1.8410397009399363e-05,
+      "loss": 0.8046,
+      "step": 8765
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8923454579459638,
+      "learning_rate": 1.8409984199997653e-05,
+      "loss": 0.6988,
+      "step": 8766
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.706518575933321,
+      "learning_rate": 1.8409571341630358e-05,
+      "loss": 0.7813,
+      "step": 8767
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.164679497599597,
+      "learning_rate": 1.8409158434299872e-05,
+      "loss": 0.7129,
+      "step": 8768
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8582262157820613,
+      "learning_rate": 1.8408745478008602e-05,
+      "loss": 0.7564,
+      "step": 8769
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6134012561379047,
+      "learning_rate": 1.840833247275896e-05,
+      "loss": 0.5381,
+      "step": 8770
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5009227853399967,
+      "learning_rate": 1.8407919418553337e-05,
+      "loss": 0.823,
+      "step": 8771
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.430443661519295,
+      "learning_rate": 1.840750631539415e-05,
+      "loss": 0.7392,
+      "step": 8772
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4353053206708257,
+      "learning_rate": 1.84070931632838e-05,
+      "loss": 0.7492,
+      "step": 8773
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.419604897157884,
+      "learning_rate": 1.840667996222469e-05,
+      "loss": 0.6609,
+      "step": 8774
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.351593973922193,
+      "learning_rate": 1.8406266712219225e-05,
+      "loss": 0.6862,
+      "step": 8775
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4930146672111957,
+      "learning_rate": 1.840585341326982e-05,
+      "loss": 0.8406,
+      "step": 8776
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8832406872886218,
+      "learning_rate": 1.840544006537887e-05,
+      "loss": 0.7734,
+      "step": 8777
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0623135908905836,
+      "learning_rate": 1.840502666854879e-05,
+      "loss": 0.7643,
+      "step": 8778
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3749573731241465,
+      "learning_rate": 1.840461322278198e-05,
+      "loss": 0.6428,
+      "step": 8779
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6130335358529153,
+      "learning_rate": 1.8404199728080854e-05,
+      "loss": 0.983,
+      "step": 8780
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.401558072984686,
+      "learning_rate": 1.8403786184447818e-05,
+      "loss": 0.8125,
+      "step": 8781
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5732592314588002,
+      "learning_rate": 1.8403372591885272e-05,
+      "loss": 0.5229,
+      "step": 8782
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5900355765004117,
+      "learning_rate": 1.8402958950395637e-05,
+      "loss": 0.574,
+      "step": 8783
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5591565467089696,
+      "learning_rate": 1.8402545259981312e-05,
+      "loss": 0.7411,
+      "step": 8784
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5746087475072084,
+      "learning_rate": 1.8402131520644708e-05,
+      "loss": 0.8248,
+      "step": 8785
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.711821917138534,
+      "learning_rate": 1.8401717732388234e-05,
+      "loss": 0.7236,
+      "step": 8786
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6656457077319659,
+      "learning_rate": 1.8401303895214303e-05,
+      "loss": 0.5796,
+      "step": 8787
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3981475306764048,
+      "learning_rate": 1.8400890009125314e-05,
+      "loss": 0.7152,
+      "step": 8788
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6675564434105137,
+      "learning_rate": 1.8400476074123686e-05,
+      "loss": 0.7657,
+      "step": 8789
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3720805150379005,
+      "learning_rate": 1.840006209021183e-05,
+      "loss": 0.6324,
+      "step": 8790
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.674496525587475,
+      "learning_rate": 1.839964805739215e-05,
+      "loss": 0.6824,
+      "step": 8791
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6849938954917252,
+      "learning_rate": 1.8399233975667062e-05,
+      "loss": 0.7745,
+      "step": 8792
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.52238568513939,
+      "learning_rate": 1.8398819845038973e-05,
+      "loss": 0.8553,
+      "step": 8793
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4155687942876298,
+      "learning_rate": 1.83984056655103e-05,
+      "loss": 0.7923,
+      "step": 8794
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4110167471815407,
+      "learning_rate": 1.8397991437083445e-05,
+      "loss": 0.7746,
+      "step": 8795
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7144622159872707,
+      "learning_rate": 1.8397577159760828e-05,
+      "loss": 0.7414,
+      "step": 8796
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9592534514426836,
+      "learning_rate": 1.8397162833544856e-05,
+      "loss": 0.6053,
+      "step": 8797
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.458819458867121,
+      "learning_rate": 1.8396748458437945e-05,
+      "loss": 0.7139,
+      "step": 8798
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8195741392918032,
+      "learning_rate": 1.8396334034442506e-05,
+      "loss": 0.7079,
+      "step": 8799
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.747487924511343,
+      "learning_rate": 1.839591956156095e-05,
+      "loss": 0.7821,
+      "step": 8800
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8331493958809375,
+      "learning_rate": 1.8395505039795696e-05,
+      "loss": 0.7214,
+      "step": 8801
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6495591576795117,
+      "learning_rate": 1.8395090469149153e-05,
+      "loss": 0.7154,
+      "step": 8802
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5715229667323822,
+      "learning_rate": 1.8394675849623733e-05,
+      "loss": 0.75,
+      "step": 8803
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.685879770414059,
+      "learning_rate": 1.8394261181221853e-05,
+      "loss": 0.8025,
+      "step": 8804
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4038180295170994,
+      "learning_rate": 1.8393846463945925e-05,
+      "loss": 0.7677,
+      "step": 8805
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3535112977216748,
+      "learning_rate": 1.839343169779837e-05,
+      "loss": 0.7609,
+      "step": 8806
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2239508891904682,
+      "learning_rate": 1.83930168827816e-05,
+      "loss": 0.7928,
+      "step": 8807
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9345907381607574,
+      "learning_rate": 1.8392602018898023e-05,
+      "loss": 0.7489,
+      "step": 8808
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5741871116142705,
+      "learning_rate": 1.8392187106150058e-05,
+      "loss": 0.7618,
+      "step": 8809
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5365399524221808,
+      "learning_rate": 1.8391772144540127e-05,
+      "loss": 0.665,
+      "step": 8810
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4067846619193904,
+      "learning_rate": 1.8391357134070642e-05,
+      "loss": 0.8307,
+      "step": 8811
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6715381222068502,
+      "learning_rate": 1.839094207474402e-05,
+      "loss": 0.5864,
+      "step": 8812
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.596680268340152,
+      "learning_rate": 1.8390526966562673e-05,
+      "loss": 0.5573,
+      "step": 8813
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.697458596859443,
+      "learning_rate": 1.8390111809529024e-05,
+      "loss": 0.6798,
+      "step": 8814
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6228304723748854,
+      "learning_rate": 1.8389696603645487e-05,
+      "loss": 0.8125,
+      "step": 8815
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.267756083785115,
+      "learning_rate": 1.8389281348914482e-05,
+      "loss": 0.7655,
+      "step": 8816
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3332193581213931,
+      "learning_rate": 1.8388866045338424e-05,
+      "loss": 0.7078,
+      "step": 8817
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5032807443968557,
+      "learning_rate": 1.8388450692919733e-05,
+      "loss": 0.7365,
+      "step": 8818
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.760598312233199,
+      "learning_rate": 1.8388035291660823e-05,
+      "loss": 0.6474,
+      "step": 8819
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.53079395168042,
+      "learning_rate": 1.8387619841564118e-05,
+      "loss": 0.8391,
+      "step": 8820
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.048803165218298,
+      "learning_rate": 1.8387204342632034e-05,
+      "loss": 0.8499,
+      "step": 8821
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6229998975756603,
+      "learning_rate": 1.8386788794866995e-05,
+      "loss": 0.856,
+      "step": 8822
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8300997372335068,
+      "learning_rate": 1.8386373198271412e-05,
+      "loss": 0.8294,
+      "step": 8823
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7763010290608288,
+      "learning_rate": 1.8385957552847707e-05,
+      "loss": 0.7818,
+      "step": 8824
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5346555486695967,
+      "learning_rate": 1.8385541858598306e-05,
+      "loss": 0.7,
+      "step": 8825
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6689355250764806,
+      "learning_rate": 1.8385126115525627e-05,
+      "loss": 0.8172,
+      "step": 8826
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8168770920422208,
+      "learning_rate": 1.8384710323632083e-05,
+      "loss": 0.7775,
+      "step": 8827
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5629588697214114,
+      "learning_rate": 1.838429448292011e-05,
+      "loss": 0.6355,
+      "step": 8828
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6035221920789358,
+      "learning_rate": 1.838387859339211e-05,
+      "loss": 0.8898,
+      "step": 8829
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5844912407275912,
+      "learning_rate": 1.8383462655050525e-05,
+      "loss": 0.5137,
+      "step": 8830
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2545270996034041,
+      "learning_rate": 1.838304666789776e-05,
+      "loss": 0.7386,
+      "step": 8831
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7242161721505254,
+      "learning_rate": 1.8382630631936243e-05,
+      "loss": 0.836,
+      "step": 8832
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8495618395354707,
+      "learning_rate": 1.8382214547168396e-05,
+      "loss": 0.846,
+      "step": 8833
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3408477964279455,
+      "learning_rate": 1.8381798413596646e-05,
+      "loss": 0.705,
+      "step": 8834
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.722981425241597,
+      "learning_rate": 1.838138223122341e-05,
+      "loss": 0.7798,
+      "step": 8835
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4529216523373907,
+      "learning_rate": 1.8380966000051113e-05,
+      "loss": 0.7651,
+      "step": 8836
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.594285125997128,
+      "learning_rate": 1.8380549720082175e-05,
+      "loss": 0.8243,
+      "step": 8837
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.493306853266308,
+      "learning_rate": 1.838013339131903e-05,
+      "loss": 0.7116,
+      "step": 8838
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8937323150966798,
+      "learning_rate": 1.837971701376409e-05,
+      "loss": 0.6715,
+      "step": 8839
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6620985037706437,
+      "learning_rate": 1.837930058741979e-05,
+      "loss": 0.7593,
+      "step": 8840
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.932637651392002,
+      "learning_rate": 1.8378884112288543e-05,
+      "loss": 0.789,
+      "step": 8841
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6197368435612886,
+      "learning_rate": 1.8378467588372785e-05,
+      "loss": 0.5414,
+      "step": 8842
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8464036578545986,
+      "learning_rate": 1.8378051015674933e-05,
+      "loss": 0.7617,
+      "step": 8843
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5718852418683689,
+      "learning_rate": 1.837763439419742e-05,
+      "loss": 0.7417,
+      "step": 8844
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4649534995966937,
+      "learning_rate": 1.8377217723942662e-05,
+      "loss": 0.7615,
+      "step": 8845
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5896456726985424,
+      "learning_rate": 1.837680100491309e-05,
+      "loss": 0.817,
+      "step": 8846
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6183896487823706,
+      "learning_rate": 1.8376384237111134e-05,
+      "loss": 0.7684,
+      "step": 8847
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5959021923518568,
+      "learning_rate": 1.8375967420539215e-05,
+      "loss": 0.7578,
+      "step": 8848
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.2043245867396717,
+      "learning_rate": 1.8375550555199764e-05,
+      "loss": 0.7414,
+      "step": 8849
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7863108131057992,
+      "learning_rate": 1.8375133641095206e-05,
+      "loss": 0.7076,
+      "step": 8850
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5184216463089952,
+      "learning_rate": 1.8374716678227964e-05,
+      "loss": 0.7907,
+      "step": 8851
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8020558996489124,
+      "learning_rate": 1.8374299666600474e-05,
+      "loss": 0.7359,
+      "step": 8852
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5215441325902024,
+      "learning_rate": 1.837388260621516e-05,
+      "loss": 0.6665,
+      "step": 8853
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7170571824761172,
+      "learning_rate": 1.8373465497074448e-05,
+      "loss": 0.8275,
+      "step": 8854
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5288586243106208,
+      "learning_rate": 1.837304833918077e-05,
+      "loss": 0.6588,
+      "step": 8855
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5970255321177261,
+      "learning_rate": 1.8372631132536552e-05,
+      "loss": 0.5382,
+      "step": 8856
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5551745074879266,
+      "learning_rate": 1.837221387714423e-05,
+      "loss": 0.8263,
+      "step": 8857
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.629172883678975,
+      "learning_rate": 1.8371796573006223e-05,
+      "loss": 0.5625,
+      "step": 8858
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.015194721907379,
+      "learning_rate": 1.8371379220124968e-05,
+      "loss": 0.7333,
+      "step": 8859
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9041064952783904,
+      "learning_rate": 1.8370961818502893e-05,
+      "loss": 0.8402,
+      "step": 8860
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4861631017505543,
+      "learning_rate": 1.837054436814243e-05,
+      "loss": 0.736,
+      "step": 8861
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5678484230866103,
+      "learning_rate": 1.8370126869046002e-05,
+      "loss": 0.8504,
+      "step": 8862
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3281692079635363,
+      "learning_rate": 1.836970932121605e-05,
+      "loss": 0.724,
+      "step": 8863
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.511213771827203,
+      "learning_rate": 1.8369291724655e-05,
+      "loss": 0.8309,
+      "step": 8864
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6056525200263025,
+      "learning_rate": 1.836887407936528e-05,
+      "loss": 0.7888,
+      "step": 8865
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6022813370806062,
+      "learning_rate": 1.8368456385349333e-05,
+      "loss": 0.8358,
+      "step": 8866
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.527679407543483,
+      "learning_rate": 1.8368038642609577e-05,
+      "loss": 0.697,
+      "step": 8867
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.479023796426288,
+      "learning_rate": 1.8367620851148455e-05,
+      "loss": 0.7144,
+      "step": 8868
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8343945822529708,
+      "learning_rate": 1.8367203010968393e-05,
+      "loss": 0.8503,
+      "step": 8869
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5967475564903821,
+      "learning_rate": 1.8366785122071828e-05,
+      "loss": 0.6591,
+      "step": 8870
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7809686470915815,
+      "learning_rate": 1.8366367184461188e-05,
+      "loss": 0.7693,
+      "step": 8871
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8524750710248505,
+      "learning_rate": 1.8365949198138913e-05,
+      "loss": 0.9051,
+      "step": 8872
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4102040200124484,
+      "learning_rate": 1.836553116310743e-05,
+      "loss": 0.873,
+      "step": 8873
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4963359708465607,
+      "learning_rate": 1.8365113079369176e-05,
+      "loss": 0.8407,
+      "step": 8874
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4358671914314833,
+      "learning_rate": 1.8364694946926586e-05,
+      "loss": 0.6498,
+      "step": 8875
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8766590328157857,
+      "learning_rate": 1.8364276765782095e-05,
+      "loss": 0.7152,
+      "step": 8876
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5992663406063097,
+      "learning_rate": 1.8363858535938136e-05,
+      "loss": 0.7826,
+      "step": 8877
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4846731911246969,
+      "learning_rate": 1.8363440257397145e-05,
+      "loss": 0.7221,
+      "step": 8878
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.418490217804683,
+      "learning_rate": 1.8363021930161555e-05,
+      "loss": 0.7005,
+      "step": 8879
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6937166647633166,
+      "learning_rate": 1.8362603554233807e-05,
+      "loss": 0.8226,
+      "step": 8880
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5398631044689381,
+      "learning_rate": 1.8362185129616333e-05,
+      "loss": 0.7562,
+      "step": 8881
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.936147652434251,
+      "learning_rate": 1.8361766656311568e-05,
+      "loss": 0.8473,
+      "step": 8882
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.404823214242814,
+      "learning_rate": 1.836134813432195e-05,
+      "loss": 0.7921,
+      "step": 8883
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4440426954973233,
+      "learning_rate": 1.8360929563649918e-05,
+      "loss": 0.8838,
+      "step": 8884
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.611682641575209,
+      "learning_rate": 1.8360510944297905e-05,
+      "loss": 0.5602,
+      "step": 8885
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6176456034409163,
+      "learning_rate": 1.836009227626835e-05,
+      "loss": 0.8132,
+      "step": 8886
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8238976869017889,
+      "learning_rate": 1.8359673559563696e-05,
+      "loss": 0.7471,
+      "step": 8887
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7363881631541471,
+      "learning_rate": 1.8359254794186368e-05,
+      "loss": 0.7938,
+      "step": 8888
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4675116426980341,
+      "learning_rate": 1.8358835980138822e-05,
+      "loss": 0.7417,
+      "step": 8889
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3671331794518868,
+      "learning_rate": 1.8358417117423476e-05,
+      "loss": 0.7447,
+      "step": 8890
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3757321245392864,
+      "learning_rate": 1.8357998206042788e-05,
+      "loss": 0.8283,
+      "step": 8891
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6829695571332455,
+      "learning_rate": 1.8357579245999182e-05,
+      "loss": 0.8299,
+      "step": 8892
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5291404792401277,
+      "learning_rate": 1.8357160237295106e-05,
+      "loss": 0.7952,
+      "step": 8893
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.560305284409028,
+      "learning_rate": 1.8356741179933e-05,
+      "loss": 0.766,
+      "step": 8894
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8831161606136855,
+      "learning_rate": 1.83563220739153e-05,
+      "loss": 0.7853,
+      "step": 8895
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9995839943416602,
+      "learning_rate": 1.835590291924444e-05,
+      "loss": 0.8476,
+      "step": 8896
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6236367991994163,
+      "learning_rate": 1.8355483715922876e-05,
+      "loss": 0.806,
+      "step": 8897
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.316774276637283,
+      "learning_rate": 1.835506446395304e-05,
+      "loss": 0.7009,
+      "step": 8898
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4155525212558802,
+      "learning_rate": 1.835464516333737e-05,
+      "loss": 0.7167,
+      "step": 8899
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.824292173403862,
+      "learning_rate": 1.8354225814078313e-05,
+      "loss": 0.694,
+      "step": 8900
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3551142484547152,
+      "learning_rate": 1.835380641617831e-05,
+      "loss": 0.8326,
+      "step": 8901
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9492930851528685,
+      "learning_rate": 1.83533869696398e-05,
+      "loss": 0.8538,
+      "step": 8902
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6250774769018423,
+      "learning_rate": 1.8352967474465223e-05,
+      "loss": 0.8258,
+      "step": 8903
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.56702909167834,
+      "learning_rate": 1.8352547930657027e-05,
+      "loss": 0.8906,
+      "step": 8904
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6277749965942742,
+      "learning_rate": 1.8352128338217654e-05,
+      "loss": 0.5071,
+      "step": 8905
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4528839340239126,
+      "learning_rate": 1.8351708697149543e-05,
+      "loss": 0.6954,
+      "step": 8906
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5895941948728898,
+      "learning_rate": 1.8351289007455145e-05,
+      "loss": 0.7737,
+      "step": 8907
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7634117297909593,
+      "learning_rate": 1.835086926913689e-05,
+      "loss": 0.7946,
+      "step": 8908
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6405674270230886,
+      "learning_rate": 1.8350449482197236e-05,
+      "loss": 0.7349,
+      "step": 8909
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3445138723465984,
+      "learning_rate": 1.8350029646638625e-05,
+      "loss": 0.7764,
+      "step": 8910
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3551691563139128,
+      "learning_rate": 1.834960976246349e-05,
+      "loss": 0.7147,
+      "step": 8911
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8000544864748504,
+      "learning_rate": 1.8349189829674288e-05,
+      "loss": 0.6851,
+      "step": 8912
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6603447321701763,
+      "learning_rate": 1.834876984827346e-05,
+      "loss": 0.8005,
+      "step": 8913
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.513924034269204,
+      "learning_rate": 1.8348349818263445e-05,
+      "loss": 0.7491,
+      "step": 8914
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4043626129390465,
+      "learning_rate": 1.83479297396467e-05,
+      "loss": 0.7043,
+      "step": 8915
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5963473570929232,
+      "learning_rate": 1.8347509612425666e-05,
+      "loss": 0.7788,
+      "step": 8916
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.409062589950399,
+      "learning_rate": 1.8347089436602786e-05,
+      "loss": 0.7961,
+      "step": 8917
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6146132704524344,
+      "learning_rate": 1.8346669212180507e-05,
+      "loss": 0.5517,
+      "step": 8918
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4210283661724106,
+      "learning_rate": 1.834624893916128e-05,
+      "loss": 0.8005,
+      "step": 8919
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6092393092708278,
+      "learning_rate": 1.834582861754755e-05,
+      "loss": 0.5613,
+      "step": 8920
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.544105117600115,
+      "learning_rate": 1.834540824734176e-05,
+      "loss": 0.8385,
+      "step": 8921
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.379157454923239,
+      "learning_rate": 1.8344987828546364e-05,
+      "loss": 0.7944,
+      "step": 8922
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.223763185671339,
+      "learning_rate": 1.8344567361163804e-05,
+      "loss": 0.7567,
+      "step": 8923
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6745643118699776,
+      "learning_rate": 1.8344146845196534e-05,
+      "loss": 0.8077,
+      "step": 8924
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7419384835495104,
+      "learning_rate": 1.8343726280646996e-05,
+      "loss": 0.8531,
+      "step": 8925
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7198437506483026,
+      "learning_rate": 1.8343305667517644e-05,
+      "loss": 0.6566,
+      "step": 8926
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5935425255070735,
+      "learning_rate": 1.8342885005810926e-05,
+      "loss": 0.4835,
+      "step": 8927
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.61285099351362,
+      "learning_rate": 1.834246429552929e-05,
+      "loss": 0.7007,
+      "step": 8928
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8787577565082276,
+      "learning_rate": 1.8342043536675185e-05,
+      "loss": 0.7643,
+      "step": 8929
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4437140130045016,
+      "learning_rate": 1.8341622729251062e-05,
+      "loss": 0.7733,
+      "step": 8930
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.476875233503763,
+      "learning_rate": 1.834120187325937e-05,
+      "loss": 0.8241,
+      "step": 8931
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4295314395196952,
+      "learning_rate": 1.834078096870256e-05,
+      "loss": 0.7009,
+      "step": 8932
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5521899109716053,
+      "learning_rate": 1.834036001558308e-05,
+      "loss": 0.555,
+      "step": 8933
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6482655010827407,
+      "learning_rate": 1.833993901390339e-05,
+      "loss": 0.705,
+      "step": 8934
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5391617755955196,
+      "learning_rate": 1.8339517963665928e-05,
+      "loss": 0.7077,
+      "step": 8935
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.487623621064261,
+      "learning_rate": 1.8339096864873153e-05,
+      "loss": 0.7832,
+      "step": 8936
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.625455122251674,
+      "learning_rate": 1.833867571752752e-05,
+      "loss": 0.7651,
+      "step": 8937
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8492255632925336,
+      "learning_rate": 1.8338254521631476e-05,
+      "loss": 0.6526,
+      "step": 8938
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6013517355010034,
+      "learning_rate": 1.8337833277187472e-05,
+      "loss": 0.5368,
+      "step": 8939
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6139735136209509,
+      "learning_rate": 1.8337411984197965e-05,
+      "loss": 0.546,
+      "step": 8940
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7538684291176114,
+      "learning_rate": 1.8336990642665405e-05,
+      "loss": 0.8615,
+      "step": 8941
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5023638310007228,
+      "learning_rate": 1.8336569252592243e-05,
+      "loss": 0.7751,
+      "step": 8942
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.727658355561472,
+      "learning_rate": 1.8336147813980936e-05,
+      "loss": 0.8799,
+      "step": 8943
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.621288075482038,
+      "learning_rate": 1.8335726326833943e-05,
+      "loss": 0.5542,
+      "step": 8944
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.8901456718441896,
+      "learning_rate": 1.8335304791153707e-05,
+      "loss": 0.7729,
+      "step": 8945
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4830988911593668,
+      "learning_rate": 1.833488320694269e-05,
+      "loss": 0.9199,
+      "step": 8946
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.393008342991801,
+      "learning_rate": 1.8334461574203346e-05,
+      "loss": 0.795,
+      "step": 8947
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8015371137551182,
+      "learning_rate": 1.8334039892938126e-05,
+      "loss": 0.738,
+      "step": 8948
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6575993846280663,
+      "learning_rate": 1.8333618163149485e-05,
+      "loss": 0.7642,
+      "step": 8949
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7814786112642897,
+      "learning_rate": 1.8333196384839884e-05,
+      "loss": 0.7279,
+      "step": 8950
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1712575321265333,
+      "learning_rate": 1.833277455801177e-05,
+      "loss": 0.8327,
+      "step": 8951
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4677112652670354,
+      "learning_rate": 1.8332352682667613e-05,
+      "loss": 0.8213,
+      "step": 8952
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4274207441880673,
+      "learning_rate": 1.8331930758809856e-05,
+      "loss": 0.7269,
+      "step": 8953
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5189745428967993,
+      "learning_rate": 1.833150878644096e-05,
+      "loss": 0.8253,
+      "step": 8954
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4113178430817235,
+      "learning_rate": 1.8331086765563384e-05,
+      "loss": 0.6576,
+      "step": 8955
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4669351585989243,
+      "learning_rate": 1.833066469617958e-05,
+      "loss": 0.7444,
+      "step": 8956
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4948916914576338,
+      "learning_rate": 1.8330242578292013e-05,
+      "loss": 0.5599,
+      "step": 8957
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9688201123589588,
+      "learning_rate": 1.832982041190313e-05,
+      "loss": 0.7978,
+      "step": 8958
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.831270504932569,
+      "learning_rate": 1.83293981970154e-05,
+      "loss": 0.7006,
+      "step": 8959
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5688278309019554,
+      "learning_rate": 1.8328975933631277e-05,
+      "loss": 0.807,
+      "step": 8960
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3744777706588494,
+      "learning_rate": 1.832855362175322e-05,
+      "loss": 0.7794,
+      "step": 8961
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.395971953613642,
+      "learning_rate": 1.832813126138368e-05,
+      "loss": 0.7483,
+      "step": 8962
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4564515874671735,
+      "learning_rate": 1.8327708852525132e-05,
+      "loss": 0.7278,
+      "step": 8963
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5508012184596351,
+      "learning_rate": 1.832728639518002e-05,
+      "loss": 0.6334,
+      "step": 8964
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.79367333788815,
+      "learning_rate": 1.8326863889350816e-05,
+      "loss": 0.7556,
+      "step": 8965
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7649156768308747,
+      "learning_rate": 1.832644133503997e-05,
+      "loss": 0.7637,
+      "step": 8966
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5685784751010357,
+      "learning_rate": 1.832601873224995e-05,
+      "loss": 0.732,
+      "step": 8967
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7388904650026025,
+      "learning_rate": 1.832559608098321e-05,
+      "loss": 0.7336,
+      "step": 8968
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4058939404674669,
+      "learning_rate": 1.8325173381242213e-05,
+      "loss": 0.713,
+      "step": 8969
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.671596355407363,
+      "learning_rate": 1.8324750633029424e-05,
+      "loss": 0.8006,
+      "step": 8970
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.775196913131223,
+      "learning_rate": 1.83243278363473e-05,
+      "loss": 0.8571,
+      "step": 8971
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5204215887209118,
+      "learning_rate": 1.8323904991198304e-05,
+      "loss": 0.812,
+      "step": 8972
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9596926049178882,
+      "learning_rate": 1.8323482097584896e-05,
+      "loss": 0.8501,
+      "step": 8973
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.648691735853873,
+      "learning_rate": 1.8323059155509542e-05,
+      "loss": 0.7808,
+      "step": 8974
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4671918665605888,
+      "learning_rate": 1.8322636164974705e-05,
+      "loss": 0.6376,
+      "step": 8975
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7578058922497746,
+      "learning_rate": 1.832221312598284e-05,
+      "loss": 0.7418,
+      "step": 8976
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.622270681583303,
+      "learning_rate": 1.8321790038536415e-05,
+      "loss": 0.8753,
+      "step": 8977
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4902303922736373,
+      "learning_rate": 1.83213669026379e-05,
+      "loss": 0.8357,
+      "step": 8978
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.595512063339051,
+      "learning_rate": 1.8320943718289747e-05,
+      "loss": 0.8013,
+      "step": 8979
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5252163810171824,
+      "learning_rate": 1.832052048549443e-05,
+      "loss": 0.7599,
+      "step": 8980
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0665433093140924,
+      "learning_rate": 1.8320097204254402e-05,
+      "loss": 0.7805,
+      "step": 8981
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9077423215679583,
+      "learning_rate": 1.831967387457214e-05,
+      "loss": 0.8337,
+      "step": 8982
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3522815558412957,
+      "learning_rate": 1.83192504964501e-05,
+      "loss": 0.7483,
+      "step": 8983
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5944219103262594,
+      "learning_rate": 1.831882706989075e-05,
+      "loss": 0.51,
+      "step": 8984
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7739651361049869,
+      "learning_rate": 1.8318403594896557e-05,
+      "loss": 0.7541,
+      "step": 8985
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7058233906934739,
+      "learning_rate": 1.8317980071469982e-05,
+      "loss": 0.7508,
+      "step": 8986
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6941849376743519,
+      "learning_rate": 1.8317556499613492e-05,
+      "loss": 0.8466,
+      "step": 8987
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.629477497660111,
+      "learning_rate": 1.831713287932956e-05,
+      "loss": 0.8248,
+      "step": 8988
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4255279663361604,
+      "learning_rate": 1.8316709210620646e-05,
+      "loss": 0.7596,
+      "step": 8989
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8802875749767027,
+      "learning_rate": 1.8316285493489213e-05,
+      "loss": 0.791,
+      "step": 8990
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8372056154615597,
+      "learning_rate": 1.831586172793774e-05,
+      "loss": 0.7545,
+      "step": 8991
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6621768324310526,
+      "learning_rate": 1.831543791396868e-05,
+      "loss": 0.6884,
+      "step": 8992
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6158124885022742,
+      "learning_rate": 1.831501405158451e-05,
+      "loss": 0.5388,
+      "step": 8993
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4880472035707166,
+      "learning_rate": 1.83145901407877e-05,
+      "loss": 0.7697,
+      "step": 8994
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6168915281343055,
+      "learning_rate": 1.831416618158071e-05,
+      "loss": 0.813,
+      "step": 8995
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.720197317002048,
+      "learning_rate": 1.8313742173966017e-05,
+      "loss": 0.6829,
+      "step": 8996
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3435815737593964,
+      "learning_rate": 1.831331811794608e-05,
+      "loss": 0.7387,
+      "step": 8997
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5111092194203581,
+      "learning_rate": 1.831289401352337e-05,
+      "loss": 0.718,
+      "step": 8998
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.383026140313138,
+      "learning_rate": 1.8312469860700366e-05,
+      "loss": 0.8709,
+      "step": 8999
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9855411845737574,
+      "learning_rate": 1.831204565947953e-05,
+      "loss": 0.6965,
+      "step": 9000
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5959420079614712,
+      "learning_rate": 1.831162140986333e-05,
+      "loss": 0.7326,
+      "step": 9001
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3432821827146608,
+      "learning_rate": 1.831119711185424e-05,
+      "loss": 0.7077,
+      "step": 9002
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4283273129448621,
+      "learning_rate": 1.831077276545473e-05,
+      "loss": 0.7337,
+      "step": 9003
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4305622788423167,
+      "learning_rate": 1.831034837066727e-05,
+      "loss": 0.7267,
+      "step": 9004
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3741086454378644,
+      "learning_rate": 1.830992392749433e-05,
+      "loss": 0.8317,
+      "step": 9005
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6024535219635407,
+      "learning_rate": 1.8309499435938377e-05,
+      "loss": 0.5581,
+      "step": 9006
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.903765062271132,
+      "learning_rate": 1.8309074896001893e-05,
+      "loss": 0.7029,
+      "step": 9007
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3472836996451072,
+      "learning_rate": 1.8308650307687347e-05,
+      "loss": 0.6132,
+      "step": 9008
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1306624363869773,
+      "learning_rate": 1.83082256709972e-05,
+      "loss": 0.6334,
+      "step": 9009
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3750862679474878,
+      "learning_rate": 1.8307800985933938e-05,
+      "loss": 0.8295,
+      "step": 9010
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.607077625753503,
+      "learning_rate": 1.830737625250003e-05,
+      "loss": 0.5587,
+      "step": 9011
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5582430801315141,
+      "learning_rate": 1.8306951470697946e-05,
+      "loss": 0.8226,
+      "step": 9012
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4981736086001443,
+      "learning_rate": 1.830652664053016e-05,
+      "loss": 0.7281,
+      "step": 9013
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.551206129730089,
+      "learning_rate": 1.8306101761999147e-05,
+      "loss": 0.8438,
+      "step": 9014
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1890529652051025,
+      "learning_rate": 1.8305676835107378e-05,
+      "loss": 0.7413,
+      "step": 9015
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5404184036249244,
+      "learning_rate": 1.830525185985733e-05,
+      "loss": 0.6923,
+      "step": 9016
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3698837629430065,
+      "learning_rate": 1.8304826836251477e-05,
+      "loss": 0.6774,
+      "step": 9017
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.49429696043743,
+      "learning_rate": 1.8304401764292292e-05,
+      "loss": 0.8374,
+      "step": 9018
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5528483479715796,
+      "learning_rate": 1.830397664398225e-05,
+      "loss": 0.6929,
+      "step": 9019
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4371866454064548,
+      "learning_rate": 1.830355147532383e-05,
+      "loss": 0.8144,
+      "step": 9020
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7824174977848926,
+      "learning_rate": 1.83031262583195e-05,
+      "loss": 0.6855,
+      "step": 9021
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3620461744408987,
+      "learning_rate": 1.8302700992971744e-05,
+      "loss": 0.7297,
+      "step": 9022
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6448157600626443,
+      "learning_rate": 1.8302275679283033e-05,
+      "loss": 0.7094,
+      "step": 9023
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7100798450405053,
+      "learning_rate": 1.8301850317255843e-05,
+      "loss": 0.7407,
+      "step": 9024
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.625024228982102,
+      "learning_rate": 1.8301424906892653e-05,
+      "loss": 0.7449,
+      "step": 9025
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3874441498993715,
+      "learning_rate": 1.830099944819594e-05,
+      "loss": 0.7133,
+      "step": 9026
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4292049973037557,
+      "learning_rate": 1.830057394116818e-05,
+      "loss": 0.7994,
+      "step": 9027
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6037502962695267,
+      "learning_rate": 1.8300148385811848e-05,
+      "loss": 0.5259,
+      "step": 9028
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5661812741161796,
+      "learning_rate": 1.8299722782129428e-05,
+      "loss": 0.8923,
+      "step": 9029
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4995551404640965,
+      "learning_rate": 1.8299297130123395e-05,
+      "loss": 0.8142,
+      "step": 9030
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3752607376514017,
+      "learning_rate": 1.8298871429796223e-05,
+      "loss": 0.8045,
+      "step": 9031
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3944764332689628,
+      "learning_rate": 1.8298445681150395e-05,
+      "loss": 0.7572,
+      "step": 9032
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7556362580945795,
+      "learning_rate": 1.829801988418839e-05,
+      "loss": 0.7297,
+      "step": 9033
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3708063094038399,
+      "learning_rate": 1.8297594038912686e-05,
+      "loss": 0.826,
+      "step": 9034
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4972973096891788,
+      "learning_rate": 1.8297168145325762e-05,
+      "loss": 0.7893,
+      "step": 9035
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5326484838870762,
+      "learning_rate": 1.82967422034301e-05,
+      "loss": 0.7882,
+      "step": 9036
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3695314129981229,
+      "learning_rate": 1.8296316213228174e-05,
+      "loss": 0.7207,
+      "step": 9037
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6880409114774055,
+      "learning_rate": 1.829589017472247e-05,
+      "loss": 0.7194,
+      "step": 9038
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.582202031489627,
+      "learning_rate": 1.829546408791547e-05,
+      "loss": 0.79,
+      "step": 9039
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5512666613972053,
+      "learning_rate": 1.829503795280965e-05,
+      "loss": 0.853,
+      "step": 9040
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4995032532167916,
+      "learning_rate": 1.8294611769407493e-05,
+      "loss": 0.7379,
+      "step": 9041
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.642310130904947,
+      "learning_rate": 1.829418553771148e-05,
+      "loss": 0.6724,
+      "step": 9042
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5110404794045966,
+      "learning_rate": 1.8293759257724095e-05,
+      "loss": 0.8739,
+      "step": 9043
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4558581938863864,
+      "learning_rate": 1.8293332929447817e-05,
+      "loss": 0.7774,
+      "step": 9044
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.4302593091493443,
+      "learning_rate": 1.8292906552885122e-05,
+      "loss": 0.7842,
+      "step": 9045
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5976187644960733,
+      "learning_rate": 1.8292480128038507e-05,
+      "loss": 0.8251,
+      "step": 9046
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4690097253094627,
+      "learning_rate": 1.8292053654910447e-05,
+      "loss": 0.7625,
+      "step": 9047
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5502917981143347,
+      "learning_rate": 1.8291627133503424e-05,
+      "loss": 0.6819,
+      "step": 9048
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6097282020700892,
+      "learning_rate": 1.829120056381992e-05,
+      "loss": 0.5473,
+      "step": 9049
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4331481346122692,
+      "learning_rate": 1.8290773945862428e-05,
+      "loss": 0.8167,
+      "step": 9050
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5097974432318035,
+      "learning_rate": 1.829034727963342e-05,
+      "loss": 0.7649,
+      "step": 9051
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6377911612605428,
+      "learning_rate": 1.828992056513539e-05,
+      "loss": 0.5709,
+      "step": 9052
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3742076115156132,
+      "learning_rate": 1.8289493802370813e-05,
+      "loss": 0.6777,
+      "step": 9053
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.123740801573065,
+      "learning_rate": 1.8289066991342182e-05,
+      "loss": 0.7782,
+      "step": 9054
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.387784156597445,
+      "learning_rate": 1.8288640132051975e-05,
+      "loss": 0.8039,
+      "step": 9055
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4340816130128529,
+      "learning_rate": 1.8288213224502685e-05,
+      "loss": 0.8045,
+      "step": 9056
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6249102097934374,
+      "learning_rate": 1.828778626869679e-05,
+      "loss": 0.7078,
+      "step": 9057
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4313580562373855,
+      "learning_rate": 1.828735926463678e-05,
+      "loss": 0.7645,
+      "step": 9058
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5417286270686203,
+      "learning_rate": 1.8286932212325143e-05,
+      "loss": 0.7423,
+      "step": 9059
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6453863627672953,
+      "learning_rate": 1.828650511176436e-05,
+      "loss": 0.8245,
+      "step": 9060
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8424795808601433,
+      "learning_rate": 1.8286077962956926e-05,
+      "loss": 0.7443,
+      "step": 9061
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4166893465980122,
+      "learning_rate": 1.828565076590532e-05,
+      "loss": 0.682,
+      "step": 9062
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.738326405808655,
+      "learning_rate": 1.828522352061203e-05,
+      "loss": 0.7896,
+      "step": 9063
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5644544537700231,
+      "learning_rate": 1.828479622707955e-05,
+      "loss": 0.8423,
+      "step": 9064
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5195494048496128,
+      "learning_rate": 1.828436888531036e-05,
+      "loss": 0.6773,
+      "step": 9065
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.688896063095537,
+      "learning_rate": 1.828394149530695e-05,
+      "loss": 0.7526,
+      "step": 9066
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5695964802010196,
+      "learning_rate": 1.8283514057071813e-05,
+      "loss": 0.7219,
+      "step": 9067
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5102706802104495,
+      "learning_rate": 1.8283086570607434e-05,
+      "loss": 0.8141,
+      "step": 9068
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4986811024384417,
+      "learning_rate": 1.8282659035916302e-05,
+      "loss": 0.7115,
+      "step": 9069
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.524133052609901,
+      "learning_rate": 1.828223145300091e-05,
+      "loss": 0.7219,
+      "step": 9070
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.08836926524265,
+      "learning_rate": 1.828180382186374e-05,
+      "loss": 0.8601,
+      "step": 9071
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5836404579080952,
+      "learning_rate": 1.8281376142507288e-05,
+      "loss": 0.7693,
+      "step": 9072
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.531780316863185,
+      "learning_rate": 1.8280948414934042e-05,
+      "loss": 0.7181,
+      "step": 9073
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7673963989283783,
+      "learning_rate": 1.828052063914649e-05,
+      "loss": 0.717,
+      "step": 9074
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4155037256082608,
+      "learning_rate": 1.828009281514713e-05,
+      "loss": 0.6591,
+      "step": 9075
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4327323694653435,
+      "learning_rate": 1.8279664942938448e-05,
+      "loss": 0.7442,
+      "step": 9076
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3720575492513856,
+      "learning_rate": 1.8279237022522932e-05,
+      "loss": 0.7157,
+      "step": 9077
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.651138875494261,
+      "learning_rate": 1.8278809053903076e-05,
+      "loss": 0.7872,
+      "step": 9078
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7530898501248886,
+      "learning_rate": 1.8278381037081377e-05,
+      "loss": 0.7287,
+      "step": 9079
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5562890885455885,
+      "learning_rate": 1.827795297206032e-05,
+      "loss": 0.6774,
+      "step": 9080
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3354956415165087,
+      "learning_rate": 1.8277524858842404e-05,
+      "loss": 0.8244,
+      "step": 9081
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7483357176298269,
+      "learning_rate": 1.827709669743011e-05,
+      "loss": 0.8347,
+      "step": 9082
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2880609380345496,
+      "learning_rate": 1.8276668487825946e-05,
+      "loss": 0.7553,
+      "step": 9083
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0223921820641575,
+      "learning_rate": 1.8276240230032396e-05,
+      "loss": 0.6881,
+      "step": 9084
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.625417956865393,
+      "learning_rate": 1.8275811924051955e-05,
+      "loss": 0.6544,
+      "step": 9085
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5408401977803339,
+      "learning_rate": 1.8275383569887112e-05,
+      "loss": 0.7547,
+      "step": 9086
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6434193990687114,
+      "learning_rate": 1.827495516754037e-05,
+      "loss": 0.5549,
+      "step": 9087
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6168110067096355,
+      "learning_rate": 1.827452671701422e-05,
+      "loss": 0.5303,
+      "step": 9088
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.53143673528778,
+      "learning_rate": 1.8274098218311154e-05,
+      "loss": 0.9046,
+      "step": 9089
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.713862183665785,
+      "learning_rate": 1.827366967143367e-05,
+      "loss": 0.7302,
+      "step": 9090
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5619574118631725,
+      "learning_rate": 1.8273241076384264e-05,
+      "loss": 0.6902,
+      "step": 9091
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5005349190164612,
+      "learning_rate": 1.8272812433165424e-05,
+      "loss": 0.6783,
+      "step": 9092
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5209177819940192,
+      "learning_rate": 1.8272383741779652e-05,
+      "loss": 0.7465,
+      "step": 9093
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8307267318045335,
+      "learning_rate": 1.8271955002229446e-05,
+      "loss": 0.6691,
+      "step": 9094
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4492756327410758,
+      "learning_rate": 1.8271526214517298e-05,
+      "loss": 0.8089,
+      "step": 9095
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4649521249874446,
+      "learning_rate": 1.8271097378645703e-05,
+      "loss": 0.8046,
+      "step": 9096
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5287041411800721,
+      "learning_rate": 1.827066849461716e-05,
+      "loss": 0.6282,
+      "step": 9097
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.886245693365644,
+      "learning_rate": 1.827023956243417e-05,
+      "loss": 0.6885,
+      "step": 9098
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4059730017107002,
+      "learning_rate": 1.826981058209923e-05,
+      "loss": 0.7513,
+      "step": 9099
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3524959758000725,
+      "learning_rate": 1.826938155361483e-05,
+      "loss": 0.7181,
+      "step": 9100
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.405912088495796,
+      "learning_rate": 1.8268952476983473e-05,
+      "loss": 0.6715,
+      "step": 9101
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6219336128709156,
+      "learning_rate": 1.8268523352207656e-05,
+      "loss": 0.7388,
+      "step": 9102
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5832312797003278,
+      "learning_rate": 1.826809417928988e-05,
+      "loss": 0.6917,
+      "step": 9103
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6775392704281777,
+      "learning_rate": 1.826766495823264e-05,
+      "loss": 0.8044,
+      "step": 9104
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5584212881901407,
+      "learning_rate": 1.826723568903844e-05,
+      "loss": 0.6938,
+      "step": 9105
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7317819538783064,
+      "learning_rate": 1.8266806371709773e-05,
+      "loss": 0.9076,
+      "step": 9106
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.45730129846375,
+      "learning_rate": 1.8266377006249144e-05,
+      "loss": 0.7761,
+      "step": 9107
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3458537301187135,
+      "learning_rate": 1.826594759265905e-05,
+      "loss": 0.8818,
+      "step": 9108
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4223618793840764,
+      "learning_rate": 1.8265518130941992e-05,
+      "loss": 0.8037,
+      "step": 9109
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.597262370779503,
+      "learning_rate": 1.8265088621100474e-05,
+      "loss": 0.7653,
+      "step": 9110
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9240434227885201,
+      "learning_rate": 1.8264659063136988e-05,
+      "loss": 0.8304,
+      "step": 9111
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5980934700531793,
+      "learning_rate": 1.8264229457054045e-05,
+      "loss": 0.5261,
+      "step": 9112
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4564984776290586,
+      "learning_rate": 1.826379980285414e-05,
+      "loss": 0.811,
+      "step": 9113
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.787593141750081,
+      "learning_rate": 1.826337010053978e-05,
+      "loss": 0.7738,
+      "step": 9114
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.560105026420486,
+      "learning_rate": 1.8262940350113456e-05,
+      "loss": 0.7649,
+      "step": 9115
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.712063345896835,
+      "learning_rate": 1.8262510551577678e-05,
+      "loss": 0.6787,
+      "step": 9116
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8978121547832976,
+      "learning_rate": 1.8262080704934955e-05,
+      "loss": 0.7698,
+      "step": 9117
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6025047250439267,
+      "learning_rate": 1.8261650810187778e-05,
+      "loss": 0.545,
+      "step": 9118
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4701236249589829,
+      "learning_rate": 1.8261220867338653e-05,
+      "loss": 0.6691,
+      "step": 9119
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.439590437635025,
+      "learning_rate": 1.8260790876390085e-05,
+      "loss": 0.8052,
+      "step": 9120
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.650856041011251,
+      "learning_rate": 1.826036083734458e-05,
+      "loss": 0.7703,
+      "step": 9121
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4009262521886243,
+      "learning_rate": 1.8259930750204637e-05,
+      "loss": 0.8564,
+      "step": 9122
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7071462006091416,
+      "learning_rate": 1.825950061497276e-05,
+      "loss": 0.7273,
+      "step": 9123
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.7325895007800716,
+      "learning_rate": 1.8259070431651457e-05,
+      "loss": 0.9437,
+      "step": 9124
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3555202404139461,
+      "learning_rate": 1.8258640200243235e-05,
+      "loss": 0.6085,
+      "step": 9125
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3178120207890107,
+      "learning_rate": 1.825820992075059e-05,
+      "loss": 0.8434,
+      "step": 9126
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.587712292203379,
+      "learning_rate": 1.8257779593176034e-05,
+      "loss": 0.8228,
+      "step": 9127
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2857679302027085,
+      "learning_rate": 1.8257349217522072e-05,
+      "loss": 0.7322,
+      "step": 9128
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6724715322070167,
+      "learning_rate": 1.8256918793791208e-05,
+      "loss": 0.7761,
+      "step": 9129
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.057039467006,
+      "learning_rate": 1.825648832198595e-05,
+      "loss": 0.7736,
+      "step": 9130
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4807526906813522,
+      "learning_rate": 1.82560578021088e-05,
+      "loss": 0.8363,
+      "step": 9131
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.518844845911968,
+      "learning_rate": 1.8255627234162267e-05,
+      "loss": 0.7079,
+      "step": 9132
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3911749134410554,
+      "learning_rate": 1.8255196618148863e-05,
+      "loss": 0.825,
+      "step": 9133
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4341077553448967,
+      "learning_rate": 1.825476595407109e-05,
+      "loss": 0.7202,
+      "step": 9134
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4671333086246998,
+      "learning_rate": 1.8254335241931453e-05,
+      "loss": 0.8187,
+      "step": 9135
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5809696004452027,
+      "learning_rate": 1.8253904481732464e-05,
+      "loss": 0.6488,
+      "step": 9136
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4681085084047176,
+      "learning_rate": 1.825347367347663e-05,
+      "loss": 0.7319,
+      "step": 9137
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.865624696026271,
+      "learning_rate": 1.8253042817166458e-05,
+      "loss": 0.6188,
+      "step": 9138
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6454734525397081,
+      "learning_rate": 1.825261191280446e-05,
+      "loss": 0.7386,
+      "step": 9139
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4610066765548297,
+      "learning_rate": 1.8252180960393145e-05,
+      "loss": 0.803,
+      "step": 9140
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3809964877716885,
+      "learning_rate": 1.8251749959935017e-05,
+      "loss": 0.8085,
+      "step": 9141
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6228323002406627,
+      "learning_rate": 1.8251318911432586e-05,
+      "loss": 0.5969,
+      "step": 9142
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3818258164889439,
+      "learning_rate": 1.8250887814888368e-05,
+      "loss": 0.7345,
+      "step": 9143
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6399609048858679,
+      "learning_rate": 1.8250456670304866e-05,
+      "loss": 0.7832,
+      "step": 9144
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4269764985432176,
+      "learning_rate": 1.8250025477684598e-05,
+      "loss": 0.7688,
+      "step": 9145
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.354913573675777,
+      "learning_rate": 1.8249594237030064e-05,
+      "loss": 0.7103,
+      "step": 9146
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.964550414157708,
+      "learning_rate": 1.8249162948343784e-05,
+      "loss": 0.7868,
+      "step": 9147
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5386070647012953,
+      "learning_rate": 1.8248731611628266e-05,
+      "loss": 0.7706,
+      "step": 9148
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4186023806858845,
+      "learning_rate": 1.8248300226886018e-05,
+      "loss": 0.79,
+      "step": 9149
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5671164255532526,
+      "learning_rate": 1.8247868794119556e-05,
+      "loss": 0.7704,
+      "step": 9150
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.005688049141873,
+      "learning_rate": 1.8247437313331394e-05,
+      "loss": 0.834,
+      "step": 9151
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5961319707076865,
+      "learning_rate": 1.8247005784524034e-05,
+      "loss": 0.5139,
+      "step": 9152
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5746323444357453,
+      "learning_rate": 1.82465742077e-05,
+      "loss": 0.5063,
+      "step": 9153
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1986074383889447,
+      "learning_rate": 1.8246142582861802e-05,
+      "loss": 0.7754,
+      "step": 9154
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5236576579950483,
+      "learning_rate": 1.8245710910011945e-05,
+      "loss": 0.818,
+      "step": 9155
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8469180817949593,
+      "learning_rate": 1.8245279189152952e-05,
+      "loss": 0.7286,
+      "step": 9156
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4255200233903862,
+      "learning_rate": 1.8244847420287334e-05,
+      "loss": 0.8065,
+      "step": 9157
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0879539627422234,
+      "learning_rate": 1.8244415603417603e-05,
+      "loss": 0.8082,
+      "step": 9158
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2453448005198955,
+      "learning_rate": 1.8243983738546276e-05,
+      "loss": 0.7286,
+      "step": 9159
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3798735900635342,
+      "learning_rate": 1.8243551825675864e-05,
+      "loss": 0.6981,
+      "step": 9160
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5415946005639771,
+      "learning_rate": 1.8243119864808886e-05,
+      "loss": 0.6012,
+      "step": 9161
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.220863381022759,
+      "learning_rate": 1.824268785594785e-05,
+      "loss": 0.7247,
+      "step": 9162
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5536806500561278,
+      "learning_rate": 1.824225579909528e-05,
+      "loss": 0.7551,
+      "step": 9163
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6925915124408677,
+      "learning_rate": 1.824182369425368e-05,
+      "loss": 0.7929,
+      "step": 9164
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0145666160109803,
+      "learning_rate": 1.824139154142558e-05,
+      "loss": 0.6815,
+      "step": 9165
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4996057549869461,
+      "learning_rate": 1.8240959340613488e-05,
+      "loss": 0.7633,
+      "step": 9166
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4442084523389587,
+      "learning_rate": 1.8240527091819923e-05,
+      "loss": 0.7219,
+      "step": 9167
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3561866995968919,
+      "learning_rate": 1.82400947950474e-05,
+      "loss": 0.9094,
+      "step": 9168
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6678560889671845,
+      "learning_rate": 1.8239662450298433e-05,
+      "loss": 0.8423,
+      "step": 9169
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.570249526615439,
+      "learning_rate": 1.8239230057575542e-05,
+      "loss": 0.7063,
+      "step": 9170
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5123305071807205,
+      "learning_rate": 1.823879761688125e-05,
+      "loss": 0.9607,
+      "step": 9171
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4505109906593887,
+      "learning_rate": 1.8238365128218066e-05,
+      "loss": 0.5839,
+      "step": 9172
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7733557260561388,
+      "learning_rate": 1.8237932591588513e-05,
+      "loss": 0.6898,
+      "step": 9173
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4839305642027143,
+      "learning_rate": 1.823750000699511e-05,
+      "loss": 0.8373,
+      "step": 9174
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4293305859111483,
+      "learning_rate": 1.8237067374440372e-05,
+      "loss": 0.757,
+      "step": 9175
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.316066674036021,
+      "learning_rate": 1.823663469392682e-05,
+      "loss": 0.7966,
+      "step": 9176
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.616220190707453,
+      "learning_rate": 1.8236201965456975e-05,
+      "loss": 0.7153,
+      "step": 9177
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0716622895290615,
+      "learning_rate": 1.8235769189033354e-05,
+      "loss": 0.7956,
+      "step": 9178
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5701101732421363,
+      "learning_rate": 1.8235336364658475e-05,
+      "loss": 0.7773,
+      "step": 9179
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.772211310463824,
+      "learning_rate": 1.8234903492334864e-05,
+      "loss": 0.7623,
+      "step": 9180
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3999127671851952,
+      "learning_rate": 1.8234470572065034e-05,
+      "loss": 0.7515,
+      "step": 9181
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.397507501468022,
+      "learning_rate": 1.8234037603851512e-05,
+      "loss": 0.7509,
+      "step": 9182
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6665175872062203,
+      "learning_rate": 1.8233604587696818e-05,
+      "loss": 0.774,
+      "step": 9183
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5223169613311989,
+      "learning_rate": 1.823317152360347e-05,
+      "loss": 0.8355,
+      "step": 9184
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8256459769652038,
+      "learning_rate": 1.823273841157399e-05,
+      "loss": 0.7064,
+      "step": 9185
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5477459306842076,
+      "learning_rate": 1.82323052516109e-05,
+      "loss": 0.7497,
+      "step": 9186
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4342620249465736,
+      "learning_rate": 1.823187204371672e-05,
+      "loss": 0.8441,
+      "step": 9187
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2657821706350045,
+      "learning_rate": 1.8231438787893978e-05,
+      "loss": 0.7119,
+      "step": 9188
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3917170212557421,
+      "learning_rate": 1.8231005484145194e-05,
+      "loss": 0.7141,
+      "step": 9189
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4235426299266383,
+      "learning_rate": 1.8230572132472888e-05,
+      "loss": 0.8619,
+      "step": 9190
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6253522533471811,
+      "learning_rate": 1.8230138732879588e-05,
+      "loss": 0.528,
+      "step": 9191
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8957333590021799,
+      "learning_rate": 1.8229705285367812e-05,
+      "loss": 0.7659,
+      "step": 9192
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.522057890580144,
+      "learning_rate": 1.822927178994009e-05,
+      "loss": 0.6629,
+      "step": 9193
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0486035971713648,
+      "learning_rate": 1.822883824659894e-05,
+      "loss": 0.6905,
+      "step": 9194
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.996052646783132,
+      "learning_rate": 1.8228404655346885e-05,
+      "loss": 0.7444,
+      "step": 9195
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8969719825026028,
+      "learning_rate": 1.8227971016186457e-05,
+      "loss": 0.7123,
+      "step": 9196
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3797993377302447,
+      "learning_rate": 1.8227537329120177e-05,
+      "loss": 0.7144,
+      "step": 9197
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6135259240183912,
+      "learning_rate": 1.8227103594150568e-05,
+      "loss": 0.8366,
+      "step": 9198
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7825032042348785,
+      "learning_rate": 1.822666981128016e-05,
+      "loss": 0.6564,
+      "step": 9199
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6472010832064334,
+      "learning_rate": 1.8226235980511476e-05,
+      "loss": 0.741,
+      "step": 9200
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4971643132666592,
+      "learning_rate": 1.8225802101847038e-05,
+      "loss": 0.7753,
+      "step": 9201
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8502737267961487,
+      "learning_rate": 1.8225368175289378e-05,
+      "loss": 0.8291,
+      "step": 9202
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.662203370302843,
+      "learning_rate": 1.822493420084102e-05,
+      "loss": 0.8117,
+      "step": 9203
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4552707953391322,
+      "learning_rate": 1.822450017850449e-05,
+      "loss": 0.7122,
+      "step": 9204
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6415686141329393,
+      "learning_rate": 1.8224066108282318e-05,
+      "loss": 0.7041,
+      "step": 9205
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.707210532661092,
+      "learning_rate": 1.8223631990177027e-05,
+      "loss": 0.6976,
+      "step": 9206
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4750904435475007,
+      "learning_rate": 1.822319782419115e-05,
+      "loss": 0.6228,
+      "step": 9207
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0552157395378474,
+      "learning_rate": 1.822276361032721e-05,
+      "loss": 0.7489,
+      "step": 9208
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3752144586623936,
+      "learning_rate": 1.822232934858774e-05,
+      "loss": 0.7351,
+      "step": 9209
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6058283464071443,
+      "learning_rate": 1.822189503897526e-05,
+      "loss": 0.5566,
+      "step": 9210
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.545051570449914,
+      "learning_rate": 1.822146068149231e-05,
+      "loss": 0.6976,
+      "step": 9211
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4161859762863405,
+      "learning_rate": 1.822102627614141e-05,
+      "loss": 0.7906,
+      "step": 9212
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6003583587145629,
+      "learning_rate": 1.822059182292509e-05,
+      "loss": 0.5097,
+      "step": 9213
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6496528100646373,
+      "learning_rate": 1.8220157321845887e-05,
+      "loss": 0.9005,
+      "step": 9214
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5785998987429442,
+      "learning_rate": 1.821972277290632e-05,
+      "loss": 0.8119,
+      "step": 9215
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.174502552508816,
+      "learning_rate": 1.8219288176108928e-05,
+      "loss": 0.688,
+      "step": 9216
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6710745207827356,
+      "learning_rate": 1.821885353145624e-05,
+      "loss": 0.6942,
+      "step": 9217
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5700110245081826,
+      "learning_rate": 1.821841883895078e-05,
+      "loss": 0.8829,
+      "step": 9218
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.347917271534969,
+      "learning_rate": 1.8217984098595086e-05,
+      "loss": 0.7441,
+      "step": 9219
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.376320643194207,
+      "learning_rate": 1.8217549310391687e-05,
+      "loss": 0.7316,
+      "step": 9220
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3415437375687689,
+      "learning_rate": 1.8217114474343113e-05,
+      "loss": 0.892,
+      "step": 9221
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2870752844000553,
+      "learning_rate": 1.82166795904519e-05,
+      "loss": 0.741,
+      "step": 9222
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6459512478491316,
+      "learning_rate": 1.8216244658720573e-05,
+      "loss": 0.8117,
+      "step": 9223
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4729637446725299,
+      "learning_rate": 1.821580967915167e-05,
+      "loss": 0.7519,
+      "step": 9224
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3681961034929198,
+      "learning_rate": 1.8215374651747723e-05,
+      "loss": 0.7153,
+      "step": 9225
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4423382632520976,
+      "learning_rate": 1.8214939576511265e-05,
+      "loss": 0.7369,
+      "step": 9226
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3660532250943234,
+      "learning_rate": 1.8214504453444825e-05,
+      "loss": 0.6698,
+      "step": 9227
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.335353151196786,
+      "learning_rate": 1.821406928255094e-05,
+      "loss": 0.9394,
+      "step": 9228
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.408400277432437,
+      "learning_rate": 1.8213634063832146e-05,
+      "loss": 0.8755,
+      "step": 9229
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6051987449507893,
+      "learning_rate": 1.8213198797290973e-05,
+      "loss": 0.5113,
+      "step": 9230
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5605003183084392,
+      "learning_rate": 1.821276348292995e-05,
+      "loss": 0.4954,
+      "step": 9231
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3397291656080792,
+      "learning_rate": 1.8212328120751626e-05,
+      "loss": 0.8486,
+      "step": 9232
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4072539153785355,
+      "learning_rate": 1.8211892710758525e-05,
+      "loss": 0.7549,
+      "step": 9233
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5942379815107184,
+      "learning_rate": 1.8211457252953185e-05,
+      "loss": 0.7285,
+      "step": 9234
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6276829994829574,
+      "learning_rate": 1.821102174733814e-05,
+      "loss": 0.8679,
+      "step": 9235
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4442098839332245,
+      "learning_rate": 1.8210586193915928e-05,
+      "loss": 0.7553,
+      "step": 9236
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5908677742895143,
+      "learning_rate": 1.8210150592689083e-05,
+      "loss": 0.5624,
+      "step": 9237
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4900367350778267,
+      "learning_rate": 1.8209714943660143e-05,
+      "loss": 0.7957,
+      "step": 9238
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2537742067479307,
+      "learning_rate": 1.8209279246831643e-05,
+      "loss": 0.7701,
+      "step": 9239
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.560898865081742,
+      "learning_rate": 1.8208843502206118e-05,
+      "loss": 0.7413,
+      "step": 9240
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5387789061926747,
+      "learning_rate": 1.8208407709786107e-05,
+      "loss": 0.8173,
+      "step": 9241
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4165472834465027,
+      "learning_rate": 1.8207971869574154e-05,
+      "loss": 0.813,
+      "step": 9242
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4332570377936327,
+      "learning_rate": 1.8207535981572785e-05,
+      "loss": 0.7579,
+      "step": 9243
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5712456488998356,
+      "learning_rate": 1.8207100045784544e-05,
+      "loss": 0.7332,
+      "step": 9244
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.420386759943386,
+      "learning_rate": 1.8206664062211966e-05,
+      "loss": 0.8126,
+      "step": 9245
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.699061367480361,
+      "learning_rate": 1.8206228030857595e-05,
+      "loss": 0.6515,
+      "step": 9246
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5708747273811383,
+      "learning_rate": 1.8205791951723966e-05,
+      "loss": 0.884,
+      "step": 9247
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.921830056398712,
+      "learning_rate": 1.8205355824813617e-05,
+      "loss": 0.8357,
+      "step": 9248
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5651087215753792,
+      "learning_rate": 1.8204919650129088e-05,
+      "loss": 0.7439,
+      "step": 9249
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3724633240069994,
+      "learning_rate": 1.8204483427672924e-05,
+      "loss": 0.686,
+      "step": 9250
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5221470570878464,
+      "learning_rate": 1.8204047157447652e-05,
+      "loss": 0.6664,
+      "step": 9251
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3870908023692259,
+      "learning_rate": 1.8203610839455828e-05,
+      "loss": 0.8273,
+      "step": 9252
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0667104467305886,
+      "learning_rate": 1.820317447369998e-05,
+      "loss": 0.7321,
+      "step": 9253
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.588867946198435,
+      "learning_rate": 1.8202738060182655e-05,
+      "loss": 0.5556,
+      "step": 9254
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4532561497763143,
+      "learning_rate": 1.8202301598906388e-05,
+      "loss": 0.7881,
+      "step": 9255
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6092037643691878,
+      "learning_rate": 1.820186508987373e-05,
+      "loss": 0.535,
+      "step": 9256
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6866924828225802,
+      "learning_rate": 1.8201428533087214e-05,
+      "loss": 0.7466,
+      "step": 9257
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3010113789504885,
+      "learning_rate": 1.820099192854938e-05,
+      "loss": 0.8572,
+      "step": 9258
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5831356331588975,
+      "learning_rate": 1.8200555276262784e-05,
+      "loss": 0.8568,
+      "step": 9259
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.435303734554582,
+      "learning_rate": 1.8200118576229953e-05,
+      "loss": 0.7806,
+      "step": 9260
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4878577462233125,
+      "learning_rate": 1.8199681828453438e-05,
+      "loss": 0.9221,
+      "step": 9261
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9030849610508653,
+      "learning_rate": 1.8199245032935775e-05,
+      "loss": 0.6276,
+      "step": 9262
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.378891083475162,
+      "learning_rate": 1.8198808189679518e-05,
+      "loss": 0.896,
+      "step": 9263
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2701226624163393,
+      "learning_rate": 1.81983712986872e-05,
+      "loss": 0.8084,
+      "step": 9264
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4226925353139739,
+      "learning_rate": 1.819793435996137e-05,
+      "loss": 0.7767,
+      "step": 9265
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9705556183713289,
+      "learning_rate": 1.819749737350457e-05,
+      "loss": 0.7892,
+      "step": 9266
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.587178143563683,
+      "learning_rate": 1.8197060339319345e-05,
+      "loss": 0.5784,
+      "step": 9267
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6277491171834024,
+      "learning_rate": 1.8196623257408238e-05,
+      "loss": 0.7689,
+      "step": 9268
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3135587493494876,
+      "learning_rate": 1.8196186127773797e-05,
+      "loss": 0.8042,
+      "step": 9269
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2353033350635798,
+      "learning_rate": 1.8195748950418567e-05,
+      "loss": 0.836,
+      "step": 9270
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3008262383559437,
+      "learning_rate": 1.819531172534509e-05,
+      "loss": 0.7932,
+      "step": 9271
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5674940546242688,
+      "learning_rate": 1.8194874452555914e-05,
+      "loss": 0.9287,
+      "step": 9272
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5771155687185618,
+      "learning_rate": 1.8194437132053586e-05,
+      "loss": 0.5388,
+      "step": 9273
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5840451894028788,
+      "learning_rate": 1.8193999763840647e-05,
+      "loss": 0.5657,
+      "step": 9274
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4149409306111076,
+      "learning_rate": 1.819356234791965e-05,
+      "loss": 0.8699,
+      "step": 9275
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6853231677387863,
+      "learning_rate": 1.819312488429314e-05,
+      "loss": 0.7405,
+      "step": 9276
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.357892958610797,
+      "learning_rate": 1.819268737296366e-05,
+      "loss": 0.7259,
+      "step": 9277
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7853184855254844,
+      "learning_rate": 1.8192249813933762e-05,
+      "loss": 0.7631,
+      "step": 9278
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3067823789397837,
+      "learning_rate": 1.819181220720599e-05,
+      "loss": 0.6794,
+      "step": 9279
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.469794744146926,
+      "learning_rate": 1.81913745527829e-05,
+      "loss": 0.7528,
+      "step": 9280
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7553395559610736,
+      "learning_rate": 1.8190936850667028e-05,
+      "loss": 0.8782,
+      "step": 9281
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.555705033072683,
+      "learning_rate": 1.819049910086093e-05,
+      "loss": 0.7252,
+      "step": 9282
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6263963383559343,
+      "learning_rate": 1.819006130336715e-05,
+      "loss": 0.5543,
+      "step": 9283
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6870255599785935,
+      "learning_rate": 1.818962345818825e-05,
+      "loss": 0.8569,
+      "step": 9284
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2016736747970596,
+      "learning_rate": 1.818918556532676e-05,
+      "loss": 0.8011,
+      "step": 9285
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8287381780721716,
+      "learning_rate": 1.818874762478524e-05,
+      "loss": 0.7225,
+      "step": 9286
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6224227177931845,
+      "learning_rate": 1.818830963656624e-05,
+      "loss": 0.7841,
+      "step": 9287
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4124103605486737,
+      "learning_rate": 1.818787160067231e-05,
+      "loss": 0.6954,
+      "step": 9288
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.770097600579949,
+      "learning_rate": 1.8187433517106e-05,
+      "loss": 0.7316,
+      "step": 9289
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.711853351908192,
+      "learning_rate": 1.8186995385869857e-05,
+      "loss": 0.6374,
+      "step": 9290
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7467138196127134,
+      "learning_rate": 1.818655720696644e-05,
+      "loss": 0.7621,
+      "step": 9291
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6393086237275917,
+      "learning_rate": 1.818611898039829e-05,
+      "loss": 0.5463,
+      "step": 9292
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6450786125014543,
+      "learning_rate": 1.8185680706167966e-05,
+      "loss": 0.6611,
+      "step": 9293
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5643082512214135,
+      "learning_rate": 1.8185242384278016e-05,
+      "loss": 0.7968,
+      "step": 9294
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8432534425120204,
+      "learning_rate": 1.8184804014730994e-05,
+      "loss": 0.6817,
+      "step": 9295
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.370184316818695,
+      "learning_rate": 1.8184365597529454e-05,
+      "loss": 0.7548,
+      "step": 9296
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6947684418284887,
+      "learning_rate": 1.8183927132675945e-05,
+      "loss": 0.7739,
+      "step": 9297
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4139757315398085,
+      "learning_rate": 1.818348862017302e-05,
+      "loss": 0.8904,
+      "step": 9298
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5628920695769668,
+      "learning_rate": 1.8183050060023233e-05,
+      "loss": 0.7219,
+      "step": 9299
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.459004743049598,
+      "learning_rate": 1.818261145222914e-05,
+      "loss": 0.8426,
+      "step": 9300
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.521982425108232,
+      "learning_rate": 1.8182172796793292e-05,
+      "loss": 0.7922,
+      "step": 9301
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5876805802487803,
+      "learning_rate": 1.8181734093718245e-05,
+      "loss": 0.7475,
+      "step": 9302
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4422075456425372,
+      "learning_rate": 1.818129534300655e-05,
+      "loss": 0.8013,
+      "step": 9303
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6732667322121333,
+      "learning_rate": 1.8180856544660763e-05,
+      "loss": 0.7889,
+      "step": 9304
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8998817863858914,
+      "learning_rate": 1.8180417698683437e-05,
+      "loss": 0.8633,
+      "step": 9305
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8312943354660733,
+      "learning_rate": 1.817997880507713e-05,
+      "loss": 0.7451,
+      "step": 9306
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5163838137490468,
+      "learning_rate": 1.81795398638444e-05,
+      "loss": 0.8146,
+      "step": 9307
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3309200697793475,
+      "learning_rate": 1.81791008749878e-05,
+      "loss": 0.8745,
+      "step": 9308
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4961010251532862,
+      "learning_rate": 1.817866183850988e-05,
+      "loss": 0.8365,
+      "step": 9309
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.3208326642991337,
+      "learning_rate": 1.8178222754413206e-05,
+      "loss": 0.8255,
+      "step": 9310
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.3277670883310524,
+      "learning_rate": 1.8177783622700328e-05,
+      "loss": 0.6922,
+      "step": 9311
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7764562907194428,
+      "learning_rate": 1.8177344443373806e-05,
+      "loss": 0.816,
+      "step": 9312
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5736523693526611,
+      "learning_rate": 1.8176905216436192e-05,
+      "loss": 0.5282,
+      "step": 9313
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6140956895456999,
+      "learning_rate": 1.817646594189005e-05,
+      "loss": 0.563,
+      "step": 9314
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.9984886187204616,
+      "learning_rate": 1.8176026619737935e-05,
+      "loss": 0.783,
+      "step": 9315
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.033737174753558,
+      "learning_rate": 1.8175587249982405e-05,
+      "loss": 0.8698,
+      "step": 9316
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5645815921143587,
+      "learning_rate": 1.8175147832626018e-05,
+      "loss": 0.8076,
+      "step": 9317
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4993154811012268,
+      "learning_rate": 1.8174708367671326e-05,
+      "loss": 0.7251,
+      "step": 9318
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6479531498879771,
+      "learning_rate": 1.81742688551209e-05,
+      "loss": 0.7221,
+      "step": 9319
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4404082941026437,
+      "learning_rate": 1.8173829294977294e-05,
+      "loss": 0.6625,
+      "step": 9320
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8711493647945103,
+      "learning_rate": 1.8173389687243063e-05,
+      "loss": 0.9097,
+      "step": 9321
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.614992434996809,
+      "learning_rate": 1.8172950031920772e-05,
+      "loss": 0.8034,
+      "step": 9322
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.3251745587697545,
+      "learning_rate": 1.8172510329012974e-05,
+      "loss": 0.7198,
+      "step": 9323
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.566064851793105,
+      "learning_rate": 1.8172070578522235e-05,
+      "loss": 0.6865,
+      "step": 9324
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5940742506505847,
+      "learning_rate": 1.8171630780451115e-05,
+      "loss": 0.7809,
+      "step": 9325
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5575620887766095,
+      "learning_rate": 1.8171190934802175e-05,
+      "loss": 0.7949,
+      "step": 9326
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.669107813261067,
+      "learning_rate": 1.8170751041577972e-05,
+      "loss": 0.7565,
+      "step": 9327
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6188280737305032,
+      "learning_rate": 1.817031110078107e-05,
+      "loss": 0.5266,
+      "step": 9328
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.558138924752506,
+      "learning_rate": 1.816987111241403e-05,
+      "loss": 0.7632,
+      "step": 9329
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8624315588301599,
+      "learning_rate": 1.8169431076479415e-05,
+      "loss": 0.8119,
+      "step": 9330
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.556447435760504,
+      "learning_rate": 1.8168990992979785e-05,
+      "loss": 0.7608,
+      "step": 9331
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4881150503670173,
+      "learning_rate": 1.8168550861917702e-05,
+      "loss": 0.7028,
+      "step": 9332
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8238968413432115,
+      "learning_rate": 1.8168110683295733e-05,
+      "loss": 0.8438,
+      "step": 9333
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.0070554098598024,
+      "learning_rate": 1.8167670457116434e-05,
+      "loss": 0.7233,
+      "step": 9334
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5719286205214026,
+      "learning_rate": 1.8167230183382375e-05,
+      "loss": 0.5234,
+      "step": 9335
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.610607319266717,
+      "learning_rate": 1.8166789862096114e-05,
+      "loss": 0.8025,
+      "step": 9336
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6218825096858651,
+      "learning_rate": 1.8166349493260218e-05,
+      "loss": 0.7787,
+      "step": 9337
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5899979146255223,
+      "learning_rate": 1.816590907687725e-05,
+      "loss": 0.7412,
+      "step": 9338
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3980791857578927,
+      "learning_rate": 1.816546861294977e-05,
+      "loss": 0.7618,
+      "step": 9339
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6301015395845433,
+      "learning_rate": 1.8165028101480354e-05,
+      "loss": 0.9011,
+      "step": 9340
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.745062280084906,
+      "learning_rate": 1.8164587542471555e-05,
+      "loss": 0.7931,
+      "step": 9341
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7057290372521623,
+      "learning_rate": 1.816414693592594e-05,
+      "loss": 0.8037,
+      "step": 9342
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8410230006353983,
+      "learning_rate": 1.816370628184608e-05,
+      "loss": 0.6952,
+      "step": 9343
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6057388868032088,
+      "learning_rate": 1.816326558023454e-05,
+      "loss": 0.6945,
+      "step": 9344
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5618866814603423,
+      "learning_rate": 1.816282483109388e-05,
+      "loss": 0.8445,
+      "step": 9345
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7767538392854494,
+      "learning_rate": 1.816238403442667e-05,
+      "loss": 0.7591,
+      "step": 9346
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4454169307133682,
+      "learning_rate": 1.8161943190235476e-05,
+      "loss": 0.7769,
+      "step": 9347
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.578560155756511,
+      "learning_rate": 1.8161502298522865e-05,
+      "loss": 0.4873,
+      "step": 9348
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4912337926614876,
+      "learning_rate": 1.81610613592914e-05,
+      "loss": 0.8917,
+      "step": 9349
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5872603708528978,
+      "learning_rate": 1.816062037254366e-05,
+      "loss": 0.724,
+      "step": 9350
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5564084481199683,
+      "learning_rate": 1.8160179338282196e-05,
+      "loss": 0.5425,
+      "step": 9351
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4111846214117563,
+      "learning_rate": 1.815973825650959e-05,
+      "loss": 0.7238,
+      "step": 9352
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5795482594372063,
+      "learning_rate": 1.8159297127228403e-05,
+      "loss": 0.8598,
+      "step": 9353
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6184678275856763,
+      "learning_rate": 1.8158855950441206e-05,
+      "loss": 0.9524,
+      "step": 9354
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4360951273835294,
+      "learning_rate": 1.8158414726150565e-05,
+      "loss": 0.7542,
+      "step": 9355
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7908644484275227,
+      "learning_rate": 1.815797345435905e-05,
+      "loss": 0.7672,
+      "step": 9356
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4409604827717493,
+      "learning_rate": 1.815753213506923e-05,
+      "loss": 0.8286,
+      "step": 9357
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6213670236853036,
+      "learning_rate": 1.815709076828368e-05,
+      "loss": 0.5065,
+      "step": 9358
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.393633598233794,
+      "learning_rate": 1.815664935400496e-05,
+      "loss": 0.795,
+      "step": 9359
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.509746596015308,
+      "learning_rate": 1.815620789223565e-05,
+      "loss": 0.7898,
+      "step": 9360
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5382984024092834,
+      "learning_rate": 1.8155766382978314e-05,
+      "loss": 0.6763,
+      "step": 9361
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6563532521200597,
+      "learning_rate": 1.815532482623552e-05,
+      "loss": 0.9374,
+      "step": 9362
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.116017457479476,
+      "learning_rate": 1.8154883222009844e-05,
+      "loss": 0.7729,
+      "step": 9363
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8958671997123249,
+      "learning_rate": 1.815444157030386e-05,
+      "loss": 0.7612,
+      "step": 9364
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.838531445313196,
+      "learning_rate": 1.8153999871120135e-05,
+      "loss": 0.6259,
+      "step": 9365
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.615497337080408,
+      "learning_rate": 1.8153558124461236e-05,
+      "loss": 0.7608,
+      "step": 9366
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5980347219580899,
+      "learning_rate": 1.8153116330329745e-05,
+      "loss": 0.5691,
+      "step": 9367
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3386369431752054,
+      "learning_rate": 1.815267448872823e-05,
+      "loss": 0.7592,
+      "step": 9368
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8735768140199096,
+      "learning_rate": 1.815223259965926e-05,
+      "loss": 0.88,
+      "step": 9369
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7249141188842654,
+      "learning_rate": 1.8151790663125412e-05,
+      "loss": 0.9106,
+      "step": 9370
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.582551961067466,
+      "learning_rate": 1.8151348679129257e-05,
+      "loss": 0.7035,
+      "step": 9371
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6955997796008664,
+      "learning_rate": 1.815090664767337e-05,
+      "loss": 0.7706,
+      "step": 9372
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7972732861360394,
+      "learning_rate": 1.8150464568760327e-05,
+      "loss": 0.794,
+      "step": 9373
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9969804202311783,
+      "learning_rate": 1.8150022442392695e-05,
+      "loss": 0.8658,
+      "step": 9374
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6475750393804456,
+      "learning_rate": 1.814958026857305e-05,
+      "loss": 0.7228,
+      "step": 9375
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5828224008701732,
+      "learning_rate": 1.8149138047303977e-05,
+      "loss": 0.7996,
+      "step": 9376
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5957265345342835,
+      "learning_rate": 1.8148695778588034e-05,
+      "loss": 0.6963,
+      "step": 9377
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1720403614986816,
+      "learning_rate": 1.814825346242781e-05,
+      "loss": 0.7347,
+      "step": 9378
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4756619095731693,
+      "learning_rate": 1.814781109882587e-05,
+      "loss": 0.8452,
+      "step": 9379
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2947857629456077,
+      "learning_rate": 1.8147368687784798e-05,
+      "loss": 0.9406,
+      "step": 9380
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3674410998845263,
+      "learning_rate": 1.8146926229307166e-05,
+      "loss": 0.764,
+      "step": 9381
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5068102154136351,
+      "learning_rate": 1.814648372339555e-05,
+      "loss": 0.7319,
+      "step": 9382
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5992455772366845,
+      "learning_rate": 1.8146041170052524e-05,
+      "loss": 0.791,
+      "step": 9383
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.610719177935142,
+      "learning_rate": 1.814559856928067e-05,
+      "loss": 0.6564,
+      "step": 9384
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6761915248901655,
+      "learning_rate": 1.814515592108256e-05,
+      "loss": 0.7379,
+      "step": 9385
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7401975878084837,
+      "learning_rate": 1.814471322546077e-05,
+      "loss": 0.7895,
+      "step": 9386
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6794354978947335,
+      "learning_rate": 1.8144270482417887e-05,
+      "loss": 0.759,
+      "step": 9387
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4919573306734315,
+      "learning_rate": 1.8143827691956483e-05,
+      "loss": 0.8321,
+      "step": 9388
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5228054373040107,
+      "learning_rate": 1.8143384854079132e-05,
+      "loss": 0.7314,
+      "step": 9389
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7913197650829744,
+      "learning_rate": 1.8142941968788422e-05,
+      "loss": 0.6977,
+      "step": 9390
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.4612296038396275,
+      "learning_rate": 1.8142499036086922e-05,
+      "loss": 0.7947,
+      "step": 9391
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.543860641770979,
+      "learning_rate": 1.8142056055977212e-05,
+      "loss": 0.8712,
+      "step": 9392
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5968924523454662,
+      "learning_rate": 1.8141613028461877e-05,
+      "loss": 0.8161,
+      "step": 9393
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0986376650031007,
+      "learning_rate": 1.8141169953543495e-05,
+      "loss": 0.761,
+      "step": 9394
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3353726846289025,
+      "learning_rate": 1.8140726831224644e-05,
+      "loss": 0.7245,
+      "step": 9395
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6524462388752337,
+      "learning_rate": 1.8140283661507904e-05,
+      "loss": 0.7357,
+      "step": 9396
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.284556076773407,
+      "learning_rate": 1.813984044439585e-05,
+      "loss": 0.727,
+      "step": 9397
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3923850480640911,
+      "learning_rate": 1.8139397179891075e-05,
+      "loss": 0.8336,
+      "step": 9398
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7042029911904297,
+      "learning_rate": 1.8138953867996147e-05,
+      "loss": 0.8501,
+      "step": 9399
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8586239772645292,
+      "learning_rate": 1.8138510508713656e-05,
+      "loss": 0.7315,
+      "step": 9400
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5249681329758036,
+      "learning_rate": 1.813806710204618e-05,
+      "loss": 0.8024,
+      "step": 9401
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9076745180050352,
+      "learning_rate": 1.81376236479963e-05,
+      "loss": 0.8537,
+      "step": 9402
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7149654663587084,
+      "learning_rate": 1.81371801465666e-05,
+      "loss": 0.7347,
+      "step": 9403
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8857925861651432,
+      "learning_rate": 1.813673659775966e-05,
+      "loss": 0.7607,
+      "step": 9404
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.220004267968745,
+      "learning_rate": 1.813629300157806e-05,
+      "loss": 0.7327,
+      "step": 9405
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5417664370210191,
+      "learning_rate": 1.8135849358024392e-05,
+      "loss": 0.8047,
+      "step": 9406
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6919193303676987,
+      "learning_rate": 1.813540566710123e-05,
+      "loss": 0.7708,
+      "step": 9407
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6406075531489863,
+      "learning_rate": 1.8134961928811164e-05,
+      "loss": 0.5639,
+      "step": 9408
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.16883886322476,
+      "learning_rate": 1.813451814315677e-05,
+      "loss": 0.858,
+      "step": 9409
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6211503326263959,
+      "learning_rate": 1.8134074310140638e-05,
+      "loss": 0.8601,
+      "step": 9410
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.646243879518899,
+      "learning_rate": 1.813363042976535e-05,
+      "loss": 0.7106,
+      "step": 9411
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0161647598498114,
+      "learning_rate": 1.8133186502033493e-05,
+      "loss": 0.8962,
+      "step": 9412
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5994363252319663,
+      "learning_rate": 1.8132742526947648e-05,
+      "loss": 0.8889,
+      "step": 9413
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5513617376080566,
+      "learning_rate": 1.8132298504510398e-05,
+      "loss": 0.8804,
+      "step": 9414
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6134573805084467,
+      "learning_rate": 1.8131854434724335e-05,
+      "loss": 0.8021,
+      "step": 9415
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.528570372870546,
+      "learning_rate": 1.8131410317592037e-05,
+      "loss": 0.6891,
+      "step": 9416
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.532163038881645,
+      "learning_rate": 1.8130966153116096e-05,
+      "loss": 0.684,
+      "step": 9417
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.29946963244,
+      "learning_rate": 1.81305219412991e-05,
+      "loss": 0.6612,
+      "step": 9418
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.068338025324986,
+      "learning_rate": 1.8130077682143627e-05,
+      "loss": 0.7989,
+      "step": 9419
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4569962896986393,
+      "learning_rate": 1.812963337565227e-05,
+      "loss": 0.8122,
+      "step": 9420
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5200742380814243,
+      "learning_rate": 1.8129189021827608e-05,
+      "loss": 0.7353,
+      "step": 9421
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.654324621908171,
+      "learning_rate": 1.8128744620672237e-05,
+      "loss": 0.7302,
+      "step": 9422
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2788309188401028,
+      "learning_rate": 1.8128300172188744e-05,
+      "loss": 0.7575,
+      "step": 9423
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6204063367837704,
+      "learning_rate": 1.812785567637971e-05,
+      "loss": 0.5479,
+      "step": 9424
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6189146271865313,
+      "learning_rate": 1.812741113324773e-05,
+      "loss": 0.7563,
+      "step": 9425
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3724050868148494,
+      "learning_rate": 1.8126966542795387e-05,
+      "loss": 0.7203,
+      "step": 9426
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6946293905774277,
+      "learning_rate": 1.8126521905025272e-05,
+      "loss": 0.7603,
+      "step": 9427
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.666163619991442,
+      "learning_rate": 1.812607721993998e-05,
+      "loss": 0.6831,
+      "step": 9428
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5257356211907738,
+      "learning_rate": 1.8125632487542084e-05,
+      "loss": 0.6746,
+      "step": 9429
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6246121146344489,
+      "learning_rate": 1.812518770783419e-05,
+      "loss": 0.5734,
+      "step": 9430
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.442033133060432,
+      "learning_rate": 1.8124742880818877e-05,
+      "loss": 0.7036,
+      "step": 9431
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4737573095879661,
+      "learning_rate": 1.812429800649874e-05,
+      "loss": 0.8853,
+      "step": 9432
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2337303363411602,
+      "learning_rate": 1.812385308487637e-05,
+      "loss": 0.8385,
+      "step": 9433
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6753902982316435,
+      "learning_rate": 1.8123408115954354e-05,
+      "loss": 0.8062,
+      "step": 9434
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.149820694368214,
+      "learning_rate": 1.812296309973528e-05,
+      "loss": 0.6688,
+      "step": 9435
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2220397438360053,
+      "learning_rate": 1.812251803622175e-05,
+      "loss": 0.7308,
+      "step": 9436
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5340307292192943,
+      "learning_rate": 1.8122072925416347e-05,
+      "loss": 0.7402,
+      "step": 9437
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5365658033201077,
+      "learning_rate": 1.8121627767321658e-05,
+      "loss": 0.6717,
+      "step": 9438
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5233100988573116,
+      "learning_rate": 1.8121182561940288e-05,
+      "loss": 0.6612,
+      "step": 9439
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4199742780042235,
+      "learning_rate": 1.8120737309274818e-05,
+      "loss": 0.768,
+      "step": 9440
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5611686160089493,
+      "learning_rate": 1.8120292009327843e-05,
+      "loss": 0.7598,
+      "step": 9441
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4710556182090373,
+      "learning_rate": 1.811984666210196e-05,
+      "loss": 0.7487,
+      "step": 9442
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.60357654104224,
+      "learning_rate": 1.811940126759976e-05,
+      "loss": 0.8749,
+      "step": 9443
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.436337939464927,
+      "learning_rate": 1.8118955825823833e-05,
+      "loss": 0.8257,
+      "step": 9444
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5919224113941884,
+      "learning_rate": 1.8118510336776776e-05,
+      "loss": 0.5612,
+      "step": 9445
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5390641908081142,
+      "learning_rate": 1.8118064800461182e-05,
+      "loss": 0.7692,
+      "step": 9446
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3754307230331388,
+      "learning_rate": 1.8117619216879645e-05,
+      "loss": 0.7547,
+      "step": 9447
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.360357801707786,
+      "learning_rate": 1.8117173586034757e-05,
+      "loss": 0.7294,
+      "step": 9448
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.502389780524956,
+      "learning_rate": 1.8116727907929113e-05,
+      "loss": 0.7168,
+      "step": 9449
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3134112896929109,
+      "learning_rate": 1.8116282182565313e-05,
+      "loss": 0.6475,
+      "step": 9450
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5253715255603324,
+      "learning_rate": 1.8115836409945947e-05,
+      "loss": 0.7562,
+      "step": 9451
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4181734062717946,
+      "learning_rate": 1.8115390590073612e-05,
+      "loss": 0.753,
+      "step": 9452
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4997260201954101,
+      "learning_rate": 1.8114944722950903e-05,
+      "loss": 0.6727,
+      "step": 9453
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4898656789594067,
+      "learning_rate": 1.8114498808580418e-05,
+      "loss": 0.6994,
+      "step": 9454
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0015454442492926,
+      "learning_rate": 1.8114052846964753e-05,
+      "loss": 0.7366,
+      "step": 9455
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.434290310238893,
+      "learning_rate": 1.8113606838106503e-05,
+      "loss": 0.7109,
+      "step": 9456
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7276653868682088,
+      "learning_rate": 1.8113160782008263e-05,
+      "loss": 0.6508,
+      "step": 9457
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7917190128022757,
+      "learning_rate": 1.8112714678672632e-05,
+      "loss": 0.814,
+      "step": 9458
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6541314489912493,
+      "learning_rate": 1.811226852810221e-05,
+      "loss": 0.8158,
+      "step": 9459
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.2020045111024324,
+      "learning_rate": 1.811182233029959e-05,
+      "loss": 0.8899,
+      "step": 9460
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4167528423856444,
+      "learning_rate": 1.8111376085267376e-05,
+      "loss": 0.7541,
+      "step": 9461
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4935475576029662,
+      "learning_rate": 1.811092979300816e-05,
+      "loss": 0.8558,
+      "step": 9462
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4260678805013374,
+      "learning_rate": 1.8110483453524543e-05,
+      "loss": 0.6634,
+      "step": 9463
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6033865742446152,
+      "learning_rate": 1.811003706681912e-05,
+      "loss": 0.6815,
+      "step": 9464
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3884785067310519,
+      "learning_rate": 1.81095906328945e-05,
+      "loss": 0.7284,
+      "step": 9465
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6022448948294541,
+      "learning_rate": 1.8109144151753274e-05,
+      "loss": 0.543,
+      "step": 9466
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.593563879995798,
+      "learning_rate": 1.810869762339804e-05,
+      "loss": 0.7204,
+      "step": 9467
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5761023801208645,
+      "learning_rate": 1.8108251047831405e-05,
+      "loss": 0.742,
+      "step": 9468
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.43759747996989,
+      "learning_rate": 1.8107804425055967e-05,
+      "loss": 0.7225,
+      "step": 9469
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.473752640435971,
+      "learning_rate": 1.810735775507432e-05,
+      "loss": 0.8131,
+      "step": 9470
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 12.048650961818627,
+      "learning_rate": 1.8106911037889072e-05,
+      "loss": 0.7443,
+      "step": 9471
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3759882355061868,
+      "learning_rate": 1.810646427350282e-05,
+      "loss": 0.8506,
+      "step": 9472
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5610894058721543,
+      "learning_rate": 1.8106017461918165e-05,
+      "loss": 0.8324,
+      "step": 9473
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3616302767979978,
+      "learning_rate": 1.8105570603137714e-05,
+      "loss": 0.7615,
+      "step": 9474
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.597620270914232,
+      "learning_rate": 1.810512369716406e-05,
+      "loss": 0.7659,
+      "step": 9475
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5812510696169892,
+      "learning_rate": 1.8104676743999816e-05,
+      "loss": 0.6852,
+      "step": 9476
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1191020454120215,
+      "learning_rate": 1.8104229743647573e-05,
+      "loss": 0.6924,
+      "step": 9477
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4616999747018338,
+      "learning_rate": 1.8103782696109943e-05,
+      "loss": 0.81,
+      "step": 9478
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6124596116085352,
+      "learning_rate": 1.8103335601389522e-05,
+      "loss": 0.914,
+      "step": 9479
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6336937164819553,
+      "learning_rate": 1.8102888459488916e-05,
+      "loss": 0.507,
+      "step": 9480
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.607614394032951,
+      "learning_rate": 1.810244127041073e-05,
+      "loss": 0.8645,
+      "step": 9481
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.356974147128741,
+      "learning_rate": 1.810199403415756e-05,
+      "loss": 0.7508,
+      "step": 9482
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.511220616368607,
+      "learning_rate": 1.8101546750732022e-05,
+      "loss": 0.7854,
+      "step": 9483
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6946695664900853,
+      "learning_rate": 1.810109942013671e-05,
+      "loss": 0.7188,
+      "step": 9484
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8716623072348173,
+      "learning_rate": 1.8100652042374232e-05,
+      "loss": 0.7462,
+      "step": 9485
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.603463339938732,
+      "learning_rate": 1.8100204617447195e-05,
+      "loss": 0.7019,
+      "step": 9486
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0043952983710693,
+      "learning_rate": 1.80997571453582e-05,
+      "loss": 0.8197,
+      "step": 9487
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6748613885604005,
+      "learning_rate": 1.8099309626109857e-05,
+      "loss": 0.7664,
+      "step": 9488
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.16842176786358,
+      "learning_rate": 1.809886205970477e-05,
+      "loss": 0.733,
+      "step": 9489
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4807682182883162,
+      "learning_rate": 1.8098414446145543e-05,
+      "loss": 0.8076,
+      "step": 9490
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.602522831303,
+      "learning_rate": 1.8097966785434786e-05,
+      "loss": 0.8099,
+      "step": 9491
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5359910927367353,
+      "learning_rate": 1.8097519077575095e-05,
+      "loss": 0.8125,
+      "step": 9492
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7346554738265678,
+      "learning_rate": 1.809707132256909e-05,
+      "loss": 0.7399,
+      "step": 9493
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.926132022014732,
+      "learning_rate": 1.8096623520419372e-05,
+      "loss": 0.7748,
+      "step": 9494
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4078486878971037,
+      "learning_rate": 1.8096175671128547e-05,
+      "loss": 0.8829,
+      "step": 9495
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.3788262362515282,
+      "learning_rate": 1.8095727774699224e-05,
+      "loss": 0.7922,
+      "step": 9496
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5635154912198663,
+      "learning_rate": 1.8095279831134012e-05,
+      "loss": 0.8517,
+      "step": 9497
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8183606880383125,
+      "learning_rate": 1.809483184043552e-05,
+      "loss": 0.6769,
+      "step": 9498
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4274610978348707,
+      "learning_rate": 1.809438380260635e-05,
+      "loss": 0.6766,
+      "step": 9499
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6280390442158261,
+      "learning_rate": 1.8093935717649117e-05,
+      "loss": 0.5509,
+      "step": 9500
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8788275702525072,
+      "learning_rate": 1.8093487585566426e-05,
+      "loss": 0.7661,
+      "step": 9501
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5221086575894485,
+      "learning_rate": 1.809303940636089e-05,
+      "loss": 0.818,
+      "step": 9502
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.382362289424077,
+      "learning_rate": 1.8092591180035114e-05,
+      "loss": 0.8493,
+      "step": 9503
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5888174433953739,
+      "learning_rate": 1.809214290659171e-05,
+      "loss": 0.5604,
+      "step": 9504
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5288253045566107,
+      "learning_rate": 1.809169458603329e-05,
+      "loss": 0.6993,
+      "step": 9505
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.545119427418098,
+      "learning_rate": 1.8091246218362464e-05,
+      "loss": 0.7414,
+      "step": 9506
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5915636313449826,
+      "learning_rate": 1.809079780358184e-05,
+      "loss": 0.7448,
+      "step": 9507
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6005424280635987,
+      "learning_rate": 1.809034934169403e-05,
+      "loss": 0.6271,
+      "step": 9508
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8686437522977506,
+      "learning_rate": 1.808990083270164e-05,
+      "loss": 0.75,
+      "step": 9509
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4254147206673866,
+      "learning_rate": 1.808945227660729e-05,
+      "loss": 0.5534,
+      "step": 9510
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.608819455094175,
+      "learning_rate": 1.808900367341359e-05,
+      "loss": 0.8706,
+      "step": 9511
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6428855136340288,
+      "learning_rate": 1.8088555023123144e-05,
+      "loss": 0.7818,
+      "step": 9512
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3507219950244354,
+      "learning_rate": 1.8088106325738572e-05,
+      "loss": 0.6937,
+      "step": 9513
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9582966706926381,
+      "learning_rate": 1.8087657581262488e-05,
+      "loss": 0.5288,
+      "step": 9514
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3420958889875265,
+      "learning_rate": 1.8087208789697498e-05,
+      "loss": 0.8225,
+      "step": 9515
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8477173738199537,
+      "learning_rate": 1.8086759951046217e-05,
+      "loss": 0.7858,
+      "step": 9516
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.086517117294093,
+      "learning_rate": 1.8086311065311258e-05,
+      "loss": 0.5974,
+      "step": 9517
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.000949488732327,
+      "learning_rate": 1.808586213249524e-05,
+      "loss": 0.823,
+      "step": 9518
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3746280655914513,
+      "learning_rate": 1.808541315260077e-05,
+      "loss": 0.8195,
+      "step": 9519
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3769765565701324,
+      "learning_rate": 1.808496412563047e-05,
+      "loss": 0.5839,
+      "step": 9520
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5188943446451146,
+      "learning_rate": 1.8084515051586943e-05,
+      "loss": 0.7943,
+      "step": 9521
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5825344491500895,
+      "learning_rate": 1.808406593047281e-05,
+      "loss": 0.798,
+      "step": 9522
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5425109854413372,
+      "learning_rate": 1.808361676229069e-05,
+      "loss": 0.7787,
+      "step": 9523
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3856823877028237,
+      "learning_rate": 1.808316754704319e-05,
+      "loss": 0.7406,
+      "step": 9524
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9434999657549874,
+      "learning_rate": 1.8082718284732932e-05,
+      "loss": 0.8176,
+      "step": 9525
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7035984410673962,
+      "learning_rate": 1.808226897536253e-05,
+      "loss": 0.5439,
+      "step": 9526
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5926584956805678,
+      "learning_rate": 1.8081819618934597e-05,
+      "loss": 0.7518,
+      "step": 9527
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5776337278152375,
+      "learning_rate": 1.8081370215451752e-05,
+      "loss": 0.8162,
+      "step": 9528
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.590098482910735,
+      "learning_rate": 1.8080920764916613e-05,
+      "loss": 0.75,
+      "step": 9529
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.50976996977139,
+      "learning_rate": 1.8080471267331792e-05,
+      "loss": 0.7274,
+      "step": 9530
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3835880536247918,
+      "learning_rate": 1.808002172269991e-05,
+      "loss": 0.8037,
+      "step": 9531
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5457083164180148,
+      "learning_rate": 1.8079572131023585e-05,
+      "loss": 0.8341,
+      "step": 9532
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.515945435693423,
+      "learning_rate": 1.807912249230543e-05,
+      "loss": 0.7528,
+      "step": 9533
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5276099869748807,
+      "learning_rate": 1.807867280654807e-05,
+      "loss": 0.8418,
+      "step": 9534
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5383499204693583,
+      "learning_rate": 1.807822307375412e-05,
+      "loss": 0.8386,
+      "step": 9535
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5422233456371217,
+      "learning_rate": 1.807777329392619e-05,
+      "loss": 0.6965,
+      "step": 9536
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6322591705199472,
+      "learning_rate": 1.8077323467066913e-05,
+      "loss": 0.8591,
+      "step": 9537
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6536111216624496,
+      "learning_rate": 1.80768735931789e-05,
+      "loss": 0.6865,
+      "step": 9538
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6977070223924617,
+      "learning_rate": 1.807642367226477e-05,
+      "loss": 0.7431,
+      "step": 9539
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6535927573271332,
+      "learning_rate": 1.8075973704327148e-05,
+      "loss": 0.8193,
+      "step": 9540
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.075546150388123,
+      "learning_rate": 1.8075523689368646e-05,
+      "loss": 0.7584,
+      "step": 9541
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4646889816519906,
+      "learning_rate": 1.8075073627391892e-05,
+      "loss": 0.7275,
+      "step": 9542
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8912061273713494,
+      "learning_rate": 1.8074623518399503e-05,
+      "loss": 0.7871,
+      "step": 9543
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5207790037285611,
+      "learning_rate": 1.8074173362394095e-05,
+      "loss": 0.7519,
+      "step": 9544
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5561441135087861,
+      "learning_rate": 1.8073723159378298e-05,
+      "loss": 0.8337,
+      "step": 9545
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.741528623857063,
+      "learning_rate": 1.8073272909354727e-05,
+      "loss": 0.6953,
+      "step": 9546
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4095699796851993,
+      "learning_rate": 1.8072822612326007e-05,
+      "loss": 0.789,
+      "step": 9547
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4600409917588146,
+      "learning_rate": 1.8072372268294755e-05,
+      "loss": 0.8471,
+      "step": 9548
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5925197827970691,
+      "learning_rate": 1.8071921877263597e-05,
+      "loss": 0.5284,
+      "step": 9549
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5207275464203325,
+      "learning_rate": 1.8071471439235154e-05,
+      "loss": 0.826,
+      "step": 9550
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.290285086777015,
+      "learning_rate": 1.807102095421205e-05,
+      "loss": 0.7248,
+      "step": 9551
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6226439273894586,
+      "learning_rate": 1.807057042219691e-05,
+      "loss": 0.5197,
+      "step": 9552
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.525355833263319,
+      "learning_rate": 1.8070119843192348e-05,
+      "loss": 0.8216,
+      "step": 9553
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4674901004720018,
+      "learning_rate": 1.8069669217200995e-05,
+      "loss": 0.7676,
+      "step": 9554
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4999221270228313,
+      "learning_rate": 1.8069218544225474e-05,
+      "loss": 0.8819,
+      "step": 9555
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7917469463450266,
+      "learning_rate": 1.8068767824268407e-05,
+      "loss": 0.7182,
+      "step": 9556
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.493005875929779,
+      "learning_rate": 1.8068317057332422e-05,
+      "loss": 0.7967,
+      "step": 9557
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5324276781564066,
+      "learning_rate": 1.8067866243420136e-05,
+      "loss": 0.7019,
+      "step": 9558
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5594065622929705,
+      "learning_rate": 1.806741538253418e-05,
+      "loss": 0.8698,
+      "step": 9559
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3747747749604218,
+      "learning_rate": 1.806696447467718e-05,
+      "loss": 0.8393,
+      "step": 9560
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5745597056652744,
+      "learning_rate": 1.806651351985176e-05,
+      "loss": 0.5839,
+      "step": 9561
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.690682074349396,
+      "learning_rate": 1.806606251806054e-05,
+      "loss": 0.8187,
+      "step": 9562
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4464441434187347,
+      "learning_rate": 1.8065611469306152e-05,
+      "loss": 0.751,
+      "step": 9563
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7579654748649824,
+      "learning_rate": 1.806516037359122e-05,
+      "loss": 0.7716,
+      "step": 9564
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4878497502552963,
+      "learning_rate": 1.806470923091837e-05,
+      "loss": 0.7994,
+      "step": 9565
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5535713233250923,
+      "learning_rate": 1.806425804129023e-05,
+      "loss": 0.7752,
+      "step": 9566
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.561581668609613,
+      "learning_rate": 1.8063806804709433e-05,
+      "loss": 0.8636,
+      "step": 9567
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8467413575853753,
+      "learning_rate": 1.8063355521178594e-05,
+      "loss": 0.7203,
+      "step": 9568
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0331572270016682,
+      "learning_rate": 1.8062904190700344e-05,
+      "loss": 0.7384,
+      "step": 9569
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3685498261307534,
+      "learning_rate": 1.8062452813277316e-05,
+      "loss": 0.6835,
+      "step": 9570
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6079599641116166,
+      "learning_rate": 1.8062001388912135e-05,
+      "loss": 0.5656,
+      "step": 9571
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.1345398520177676,
+      "learning_rate": 1.806154991760743e-05,
+      "loss": 0.8055,
+      "step": 9572
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8413606279328871,
+      "learning_rate": 1.806109839936583e-05,
+      "loss": 0.6698,
+      "step": 9573
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7086879491198754,
+      "learning_rate": 1.806064683418996e-05,
+      "loss": 0.6617,
+      "step": 9574
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6906898937770312,
+      "learning_rate": 1.8060195222082457e-05,
+      "loss": 0.8394,
+      "step": 9575
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.736739470022483,
+      "learning_rate": 1.805974356304594e-05,
+      "loss": 0.738,
+      "step": 9576
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4137803151050155,
+      "learning_rate": 1.8059291857083048e-05,
+      "loss": 0.8459,
+      "step": 9577
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4782120194984556,
+      "learning_rate": 1.8058840104196406e-05,
+      "loss": 0.7925,
+      "step": 9578
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2090325549221204,
+      "learning_rate": 1.8058388304388646e-05,
+      "loss": 0.8392,
+      "step": 9579
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.684941904161243,
+      "learning_rate": 1.80579364576624e-05,
+      "loss": 0.7362,
+      "step": 9580
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6284982987186531,
+      "learning_rate": 1.8057484564020294e-05,
+      "loss": 0.7289,
+      "step": 9581
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4454633582487484,
+      "learning_rate": 1.805703262346496e-05,
+      "loss": 0.6736,
+      "step": 9582
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5788776860707048,
+      "learning_rate": 1.8056580635999033e-05,
+      "loss": 0.6961,
+      "step": 9583
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3825959815523476,
+      "learning_rate": 1.8056128601625146e-05,
+      "loss": 0.7437,
+      "step": 9584
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4387040776923903,
+      "learning_rate": 1.8055676520345928e-05,
+      "loss": 0.6871,
+      "step": 9585
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.661629108392251,
+      "learning_rate": 1.8055224392164005e-05,
+      "loss": 0.7854,
+      "step": 9586
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6499355767443886,
+      "learning_rate": 1.805477221708202e-05,
+      "loss": 0.7561,
+      "step": 9587
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9742138320720095,
+      "learning_rate": 1.80543199951026e-05,
+      "loss": 0.9158,
+      "step": 9588
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.525555103039096,
+      "learning_rate": 1.8053867726228376e-05,
+      "loss": 0.8065,
+      "step": 9589
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6280827143213643,
+      "learning_rate": 1.8053415410461988e-05,
+      "loss": 0.8671,
+      "step": 9590
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0809392586035544,
+      "learning_rate": 1.8052963047806065e-05,
+      "loss": 0.7953,
+      "step": 9591
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4781674227364532,
+      "learning_rate": 1.8052510638263244e-05,
+      "loss": 0.7866,
+      "step": 9592
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1150289446126678,
+      "learning_rate": 1.8052058181836152e-05,
+      "loss": 0.6596,
+      "step": 9593
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4691798656351998,
+      "learning_rate": 1.805160567852743e-05,
+      "loss": 0.8015,
+      "step": 9594
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4588657059437131,
+      "learning_rate": 1.805115312833971e-05,
+      "loss": 0.7411,
+      "step": 9595
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3118107609294403,
+      "learning_rate": 1.8050700531275632e-05,
+      "loss": 0.743,
+      "step": 9596
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.764672945616798,
+      "learning_rate": 1.8050247887337825e-05,
+      "loss": 0.6433,
+      "step": 9597
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5230752590696681,
+      "learning_rate": 1.8049795196528924e-05,
+      "loss": 0.7071,
+      "step": 9598
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6525536090308999,
+      "learning_rate": 1.8049342458851563e-05,
+      "loss": 0.8383,
+      "step": 9599
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7890137763379161,
+      "learning_rate": 1.8048889674308387e-05,
+      "loss": 0.743,
+      "step": 9600
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6150666409285717,
+      "learning_rate": 1.8048436842902025e-05,
+      "loss": 0.7,
+      "step": 9601
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8828845117304607,
+      "learning_rate": 1.8047983964635118e-05,
+      "loss": 0.7227,
+      "step": 9602
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3249071845259897,
+      "learning_rate": 1.80475310395103e-05,
+      "loss": 0.7244,
+      "step": 9603
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3832731292575084,
+      "learning_rate": 1.8047078067530205e-05,
+      "loss": 0.8261,
+      "step": 9604
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8201507589283616,
+      "learning_rate": 1.804662504869748e-05,
+      "loss": 0.7155,
+      "step": 9605
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5035693288393652,
+      "learning_rate": 1.804617198301475e-05,
+      "loss": 0.7347,
+      "step": 9606
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5094022944511978,
+      "learning_rate": 1.8045718870484662e-05,
+      "loss": 0.7648,
+      "step": 9607
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.660862343620965,
+      "learning_rate": 1.804526571110985e-05,
+      "loss": 0.7688,
+      "step": 9608
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3408990454279621,
+      "learning_rate": 1.8044812504892958e-05,
+      "loss": 0.6232,
+      "step": 9609
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7391498730287493,
+      "learning_rate": 1.8044359251836615e-05,
+      "loss": 0.6906,
+      "step": 9610
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4090756033203293,
+      "learning_rate": 1.8043905951943472e-05,
+      "loss": 0.7449,
+      "step": 9611
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6175470283933213,
+      "learning_rate": 1.8043452605216157e-05,
+      "loss": 0.765,
+      "step": 9612
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5264444271495674,
+      "learning_rate": 1.804299921165732e-05,
+      "loss": 0.8381,
+      "step": 9613
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4808705845508465,
+      "learning_rate": 1.804254577126959e-05,
+      "loss": 0.726,
+      "step": 9614
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.564578912253175,
+      "learning_rate": 1.8042092284055615e-05,
+      "loss": 0.7399,
+      "step": 9615
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.459007866990193,
+      "learning_rate": 1.804163875001803e-05,
+      "loss": 0.8068,
+      "step": 9616
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6137286461401225,
+      "learning_rate": 1.8041185169159478e-05,
+      "loss": 0.5163,
+      "step": 9617
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4313446813183728,
+      "learning_rate": 1.8040731541482603e-05,
+      "loss": 0.7486,
+      "step": 9618
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7522294888686665,
+      "learning_rate": 1.804027786699004e-05,
+      "loss": 0.836,
+      "step": 9619
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7144311707834279,
+      "learning_rate": 1.8039824145684438e-05,
+      "loss": 0.799,
+      "step": 9620
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.463225608080982,
+      "learning_rate": 1.803937037756843e-05,
+      "loss": 0.6059,
+      "step": 9621
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.376293605679011,
+      "learning_rate": 1.8038916562644666e-05,
+      "loss": 0.804,
+      "step": 9622
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9800145452920745,
+      "learning_rate": 1.8038462700915782e-05,
+      "loss": 0.7655,
+      "step": 9623
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6003196485899907,
+      "learning_rate": 1.8038008792384426e-05,
+      "loss": 0.5466,
+      "step": 9624
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.090431551651559,
+      "learning_rate": 1.8037554837053238e-05,
+      "loss": 0.7216,
+      "step": 9625
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4094393439349924,
+      "learning_rate": 1.803710083492486e-05,
+      "loss": 0.8162,
+      "step": 9626
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.455853206832574,
+      "learning_rate": 1.8036646786001935e-05,
+      "loss": 0.7579,
+      "step": 9627
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.073738987951009,
+      "learning_rate": 1.803619269028711e-05,
+      "loss": 0.7609,
+      "step": 9628
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3133402406002763,
+      "learning_rate": 1.8035738547783025e-05,
+      "loss": 0.7438,
+      "step": 9629
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3882192451720747,
+      "learning_rate": 1.803528435849233e-05,
+      "loss": 0.7721,
+      "step": 9630
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4552717687857775,
+      "learning_rate": 1.8034830122417658e-05,
+      "loss": 0.6568,
+      "step": 9631
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.795637609263585,
+      "learning_rate": 1.803437583956167e-05,
+      "loss": 0.6763,
+      "step": 9632
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5988422276679848,
+      "learning_rate": 1.8033921509926998e-05,
+      "loss": 0.7695,
+      "step": 9633
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5972354793538655,
+      "learning_rate": 1.8033467133516288e-05,
+      "loss": 0.566,
+      "step": 9634
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6065585615907043,
+      "learning_rate": 1.8033012710332195e-05,
+      "loss": 0.6948,
+      "step": 9635
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4122039006808336,
+      "learning_rate": 1.8032558240377355e-05,
+      "loss": 0.6793,
+      "step": 9636
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5073397702881353,
+      "learning_rate": 1.803210372365442e-05,
+      "loss": 0.7391,
+      "step": 9637
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8657213921194478,
+      "learning_rate": 1.8031649160166035e-05,
+      "loss": 0.8772,
+      "step": 9638
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7730882065157982,
+      "learning_rate": 1.8031194549914845e-05,
+      "loss": 0.6609,
+      "step": 9639
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8596290000694475,
+      "learning_rate": 1.8030739892903498e-05,
+      "loss": 0.8178,
+      "step": 9640
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7771371312614488,
+      "learning_rate": 1.803028518913464e-05,
+      "loss": 0.7006,
+      "step": 9641
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5744713699336086,
+      "learning_rate": 1.8029830438610916e-05,
+      "loss": 0.6967,
+      "step": 9642
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5158650900572683,
+      "learning_rate": 1.802937564133498e-05,
+      "loss": 0.807,
+      "step": 9643
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.478262825325476,
+      "learning_rate": 1.8028920797309478e-05,
+      "loss": 0.7227,
+      "step": 9644
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.700255571274389,
+      "learning_rate": 1.8028465906537055e-05,
+      "loss": 0.756,
+      "step": 9645
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.392203596351711,
+      "learning_rate": 1.8028010969020364e-05,
+      "loss": 0.7387,
+      "step": 9646
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.662011312827225,
+      "learning_rate": 1.802755598476205e-05,
+      "loss": 0.6066,
+      "step": 9647
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5243447932936447,
+      "learning_rate": 1.8027100953764766e-05,
+      "loss": 0.8045,
+      "step": 9648
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1853673196912125,
+      "learning_rate": 1.802664587603116e-05,
+      "loss": 0.8322,
+      "step": 9649
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1239792097422523,
+      "learning_rate": 1.8026190751563874e-05,
+      "loss": 0.6702,
+      "step": 9650
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4789246656849513,
+      "learning_rate": 1.8025735580365568e-05,
+      "loss": 0.7323,
+      "step": 9651
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7941114436049181,
+      "learning_rate": 1.802528036243889e-05,
+      "loss": 0.719,
+      "step": 9652
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.559952748335208,
+      "learning_rate": 1.8024825097786487e-05,
+      "loss": 0.4974,
+      "step": 9653
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4820839484327661,
+      "learning_rate": 1.802436978641101e-05,
+      "loss": 0.7148,
+      "step": 9654
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4716118320199298,
+      "learning_rate": 1.8023914428315115e-05,
+      "loss": 0.798,
+      "step": 9655
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7014120707320428,
+      "learning_rate": 1.802345902350145e-05,
+      "loss": 0.6665,
+      "step": 9656
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.522308514809931,
+      "learning_rate": 1.8023003571972664e-05,
+      "loss": 0.7497,
+      "step": 9657
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.768035042941075,
+      "learning_rate": 1.8022548073731413e-05,
+      "loss": 0.6557,
+      "step": 9658
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.255758376558576,
+      "learning_rate": 1.8022092528780348e-05,
+      "loss": 0.7422,
+      "step": 9659
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5397398646936447,
+      "learning_rate": 1.8021636937122117e-05,
+      "loss": 0.8068,
+      "step": 9660
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4301199922900518,
+      "learning_rate": 1.8021181298759377e-05,
+      "loss": 0.8349,
+      "step": 9661
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.661636116765199,
+      "learning_rate": 1.802072561369478e-05,
+      "loss": 0.6128,
+      "step": 9662
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7625808157217517,
+      "learning_rate": 1.8020269881930982e-05,
+      "loss": 0.7641,
+      "step": 9663
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6460984989737735,
+      "learning_rate": 1.8019814103470633e-05,
+      "loss": 0.8154,
+      "step": 9664
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6376489122356612,
+      "learning_rate": 1.8019358278316386e-05,
+      "loss": 0.5878,
+      "step": 9665
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5602037140963674,
+      "learning_rate": 1.8018902406470894e-05,
+      "loss": 0.8634,
+      "step": 9666
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5047433281886788,
+      "learning_rate": 1.8018446487936815e-05,
+      "loss": 0.7057,
+      "step": 9667
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6431670953902462,
+      "learning_rate": 1.8017990522716804e-05,
+      "loss": 0.7314,
+      "step": 9668
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3432498874250727,
+      "learning_rate": 1.8017534510813513e-05,
+      "loss": 0.6885,
+      "step": 9669
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.80130378142717,
+      "learning_rate": 1.8017078452229597e-05,
+      "loss": 0.7401,
+      "step": 9670
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6059967448680297,
+      "learning_rate": 1.8016622346967714e-05,
+      "loss": 0.8274,
+      "step": 9671
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5803617080156294,
+      "learning_rate": 1.8016166195030514e-05,
+      "loss": 0.5208,
+      "step": 9672
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.569065202136503,
+      "learning_rate": 1.801570999642066e-05,
+      "loss": 0.5107,
+      "step": 9673
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3410443842547903,
+      "learning_rate": 1.8015253751140803e-05,
+      "loss": 0.7964,
+      "step": 9674
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6506489804913598,
+      "learning_rate": 1.80147974591936e-05,
+      "loss": 0.7188,
+      "step": 9675
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4992331425800827,
+      "learning_rate": 1.801434112058171e-05,
+      "loss": 0.9631,
+      "step": 9676
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4561012847042452,
+      "learning_rate": 1.801388473530779e-05,
+      "loss": 0.7378,
+      "step": 9677
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7910715359978329,
+      "learning_rate": 1.8013428303374495e-05,
+      "loss": 0.8704,
+      "step": 9678
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6066130465763003,
+      "learning_rate": 1.801297182478448e-05,
+      "loss": 0.5287,
+      "step": 9679
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5729608074170647,
+      "learning_rate": 1.8012515299540413e-05,
+      "loss": 0.8231,
+      "step": 9680
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4900199117623971,
+      "learning_rate": 1.8012058727644943e-05,
+      "loss": 0.7654,
+      "step": 9681
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5205410900464158,
+      "learning_rate": 1.8011602109100724e-05,
+      "loss": 0.7236,
+      "step": 9682
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6447131408403513,
+      "learning_rate": 1.8011145443910426e-05,
+      "loss": 0.7997,
+      "step": 9683
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5473603686885802,
+      "learning_rate": 1.8010688732076706e-05,
+      "loss": 0.6451,
+      "step": 9684
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.9804496639291407,
+      "learning_rate": 1.8010231973602218e-05,
+      "loss": 0.7153,
+      "step": 9685
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.581543934008191,
+      "learning_rate": 1.8009775168489622e-05,
+      "loss": 0.7747,
+      "step": 9686
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6295142075606566,
+      "learning_rate": 1.8009318316741577e-05,
+      "loss": 0.7659,
+      "step": 9687
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1282733421290008,
+      "learning_rate": 1.8008861418360746e-05,
+      "loss": 0.6582,
+      "step": 9688
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.610138503783987,
+      "learning_rate": 1.8008404473349793e-05,
+      "loss": 0.7959,
+      "step": 9689
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6080045657044593,
+      "learning_rate": 1.800794748171137e-05,
+      "loss": 0.5797,
+      "step": 9690
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.466847210199326,
+      "learning_rate": 1.8007490443448144e-05,
+      "loss": 0.7996,
+      "step": 9691
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.423742549127264,
+      "learning_rate": 1.800703335856277e-05,
+      "loss": 0.7727,
+      "step": 9692
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8134725275387014,
+      "learning_rate": 1.8006576227057914e-05,
+      "loss": 0.8143,
+      "step": 9693
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4019618742100461,
+      "learning_rate": 1.8006119048936236e-05,
+      "loss": 0.813,
+      "step": 9694
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4907201058211001,
+      "learning_rate": 1.80056618242004e-05,
+      "loss": 0.7624,
+      "step": 9695
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.58381769433528,
+      "learning_rate": 1.8005204552853068e-05,
+      "loss": 0.8565,
+      "step": 9696
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5939402613971283,
+      "learning_rate": 1.8004747234896896e-05,
+      "loss": 0.8141,
+      "step": 9697
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4275402312863406,
+      "learning_rate": 1.8004289870334552e-05,
+      "loss": 0.7302,
+      "step": 9698
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7454314221882965,
+      "learning_rate": 1.8003832459168703e-05,
+      "loss": 0.7572,
+      "step": 9699
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9215864075402926,
+      "learning_rate": 1.8003375001402005e-05,
+      "loss": 0.8385,
+      "step": 9700
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7829513413504285,
+      "learning_rate": 1.800291749703712e-05,
+      "loss": 0.7658,
+      "step": 9701
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4227440031826206,
+      "learning_rate": 1.8002459946076717e-05,
+      "loss": 0.6127,
+      "step": 9702
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6911610802631074,
+      "learning_rate": 1.8002002348523463e-05,
+      "loss": 0.6952,
+      "step": 9703
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.463373769024642,
+      "learning_rate": 1.8001544704380013e-05,
+      "loss": 0.8041,
+      "step": 9704
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4391578803865812,
+      "learning_rate": 1.800108701364904e-05,
+      "loss": 0.7302,
+      "step": 9705
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8758683838466146,
+      "learning_rate": 1.8000629276333204e-05,
+      "loss": 0.8376,
+      "step": 9706
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.4304759533792484,
+      "learning_rate": 1.8000171492435172e-05,
+      "loss": 0.7683,
+      "step": 9707
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5903505971009038,
+      "learning_rate": 1.799971366195761e-05,
+      "loss": 0.8134,
+      "step": 9708
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4034578288898847,
+      "learning_rate": 1.7999255784903178e-05,
+      "loss": 0.8257,
+      "step": 9709
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.062875569088874,
+      "learning_rate": 1.799879786127455e-05,
+      "loss": 0.7408,
+      "step": 9710
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6193610121782185,
+      "learning_rate": 1.7998339891074385e-05,
+      "loss": 0.5778,
+      "step": 9711
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.616017131848877,
+      "learning_rate": 1.7997881874305356e-05,
+      "loss": 0.7536,
+      "step": 9712
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5507876799127938,
+      "learning_rate": 1.7997423810970128e-05,
+      "loss": 0.8398,
+      "step": 9713
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.549489012475381,
+      "learning_rate": 1.7996965701071364e-05,
+      "loss": 0.6533,
+      "step": 9714
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4225835081309504,
+      "learning_rate": 1.7996507544611735e-05,
+      "loss": 0.7338,
+      "step": 9715
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4387044852095376,
+      "learning_rate": 1.7996049341593903e-05,
+      "loss": 0.7308,
+      "step": 9716
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1530515391937715,
+      "learning_rate": 1.7995591092020544e-05,
+      "loss": 0.8351,
+      "step": 9717
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4334120469921967,
+      "learning_rate": 1.7995132795894323e-05,
+      "loss": 0.7605,
+      "step": 9718
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7700695423756174,
+      "learning_rate": 1.799467445321791e-05,
+      "loss": 0.8564,
+      "step": 9719
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.365743636710706,
+      "learning_rate": 1.7994216063993968e-05,
+      "loss": 0.6771,
+      "step": 9720
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0879715913497434,
+      "learning_rate": 1.7993757628225166e-05,
+      "loss": 0.7412,
+      "step": 9721
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0126678232632162,
+      "learning_rate": 1.799329914591418e-05,
+      "loss": 0.731,
+      "step": 9722
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4060666763663832,
+      "learning_rate": 1.799284061706368e-05,
+      "loss": 0.7447,
+      "step": 9723
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.477894300303262,
+      "learning_rate": 1.7992382041676326e-05,
+      "loss": 0.7138,
+      "step": 9724
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4811703371782712,
+      "learning_rate": 1.7991923419754793e-05,
+      "loss": 0.699,
+      "step": 9725
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.023155900562226,
+      "learning_rate": 1.799146475130175e-05,
+      "loss": 0.7141,
+      "step": 9726
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4269120470009717,
+      "learning_rate": 1.7991006036319876e-05,
+      "loss": 0.6968,
+      "step": 9727
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6044761090937205,
+      "learning_rate": 1.799054727481183e-05,
+      "loss": 0.6296,
+      "step": 9728
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.146347427448616,
+      "learning_rate": 1.799008846678029e-05,
+      "loss": 0.6614,
+      "step": 9729
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5525736104763959,
+      "learning_rate": 1.7989629612227924e-05,
+      "loss": 0.8322,
+      "step": 9730
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.322999282766118,
+      "learning_rate": 1.7989170711157406e-05,
+      "loss": 0.735,
+      "step": 9731
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7208022127885105,
+      "learning_rate": 1.7988711763571406e-05,
+      "loss": 0.7914,
+      "step": 9732
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2449495736302696,
+      "learning_rate": 1.7988252769472597e-05,
+      "loss": 0.687,
+      "step": 9733
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6087650985259065,
+      "learning_rate": 1.798779372886365e-05,
+      "loss": 0.5471,
+      "step": 9734
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5208648701477385,
+      "learning_rate": 1.798733464174724e-05,
+      "loss": 0.7851,
+      "step": 9735
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.885181031752809,
+      "learning_rate": 1.798687550812604e-05,
+      "loss": 0.7057,
+      "step": 9736
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.397414862547881,
+      "learning_rate": 1.7986416328002723e-05,
+      "loss": 0.6685,
+      "step": 9737
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6350252010991904,
+      "learning_rate": 1.798595710137996e-05,
+      "loss": 0.7642,
+      "step": 9738
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.463307782944879,
+      "learning_rate": 1.7985497828260425e-05,
+      "loss": 0.7693,
+      "step": 9739
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5481672855016415,
+      "learning_rate": 1.7985038508646794e-05,
+      "loss": 0.769,
+      "step": 9740
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.374771188222201,
+      "learning_rate": 1.7984579142541743e-05,
+      "loss": 0.714,
+      "step": 9741
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4528577776510312,
+      "learning_rate": 1.7984119729947944e-05,
+      "loss": 0.8046,
+      "step": 9742
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6580325028054703,
+      "learning_rate": 1.7983660270868074e-05,
+      "loss": 0.8312,
+      "step": 9743
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.678537450413087,
+      "learning_rate": 1.7983200765304802e-05,
+      "loss": 0.7889,
+      "step": 9744
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.574016736739929,
+      "learning_rate": 1.798274121326081e-05,
+      "loss": 0.7659,
+      "step": 9745
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6081876836444533,
+      "learning_rate": 1.798228161473877e-05,
+      "loss": 0.758,
+      "step": 9746
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7326891268380336,
+      "learning_rate": 1.798182196974136e-05,
+      "loss": 0.6848,
+      "step": 9747
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.484402120287168,
+      "learning_rate": 1.7981362278271258e-05,
+      "loss": 0.7308,
+      "step": 9748
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.570239810314292,
+      "learning_rate": 1.7980902540331134e-05,
+      "loss": 0.8027,
+      "step": 9749
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.717240518188628,
+      "learning_rate": 1.798044275592367e-05,
+      "loss": 0.8506,
+      "step": 9750
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.889237083875291,
+      "learning_rate": 1.797998292505154e-05,
+      "loss": 0.8009,
+      "step": 9751
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.576725660549237,
+      "learning_rate": 1.7979523047717427e-05,
+      "loss": 0.7786,
+      "step": 9752
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5989479144066723,
+      "learning_rate": 1.7979063123924004e-05,
+      "loss": 0.8012,
+      "step": 9753
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6886562276524513,
+      "learning_rate": 1.7978603153673944e-05,
+      "loss": 0.7554,
+      "step": 9754
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5918884925050594,
+      "learning_rate": 1.7978143136969938e-05,
+      "loss": 0.5505,
+      "step": 9755
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5798663711124148,
+      "learning_rate": 1.7977683073814655e-05,
+      "loss": 0.7206,
+      "step": 9756
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4190387562048594,
+      "learning_rate": 1.797722296421077e-05,
+      "loss": 0.8726,
+      "step": 9757
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6906709313023536,
+      "learning_rate": 1.7976762808160972e-05,
+      "loss": 0.7557,
+      "step": 9758
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4470022609800512,
+      "learning_rate": 1.7976302605667936e-05,
+      "loss": 0.7139,
+      "step": 9759
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4125862684719646,
+      "learning_rate": 1.797584235673434e-05,
+      "loss": 0.6486,
+      "step": 9760
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.143423271767127,
+      "learning_rate": 1.7975382061362863e-05,
+      "loss": 0.7924,
+      "step": 9761
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9008853853541785,
+      "learning_rate": 1.797492171955619e-05,
+      "loss": 0.7693,
+      "step": 9762
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7670448813472157,
+      "learning_rate": 1.7974461331316994e-05,
+      "loss": 0.8195,
+      "step": 9763
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7964282526666437,
+      "learning_rate": 1.797400089664796e-05,
+      "loss": 0.6076,
+      "step": 9764
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.217792767271505,
+      "learning_rate": 1.797354041555177e-05,
+      "loss": 0.7419,
+      "step": 9765
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4136799294717806,
+      "learning_rate": 1.7973079888031102e-05,
+      "loss": 0.7237,
+      "step": 9766
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4107053070784419,
+      "learning_rate": 1.7972619314088643e-05,
+      "loss": 0.8055,
+      "step": 9767
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.6470427642596763,
+      "learning_rate": 1.7972158693727064e-05,
+      "loss": 0.8869,
+      "step": 9768
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.366577219675564,
+      "learning_rate": 1.7971698026949055e-05,
+      "loss": 0.7929,
+      "step": 9769
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.535235962820281,
+      "learning_rate": 1.79712373137573e-05,
+      "loss": 0.6781,
+      "step": 9770
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6182556825241026,
+      "learning_rate": 1.7970776554154472e-05,
+      "loss": 0.6819,
+      "step": 9771
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8037027523792721,
+      "learning_rate": 1.7970315748143264e-05,
+      "loss": 0.8056,
+      "step": 9772
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4156752765781853,
+      "learning_rate": 1.7969854895726354e-05,
+      "loss": 0.7916,
+      "step": 9773
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5462879619077854,
+      "learning_rate": 1.796939399690642e-05,
+      "loss": 0.7228,
+      "step": 9774
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8608515882600034,
+      "learning_rate": 1.7968933051686157e-05,
+      "loss": 0.7438,
+      "step": 9775
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.782237700793329,
+      "learning_rate": 1.7968472060068242e-05,
+      "loss": 0.8132,
+      "step": 9776
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5323497927498817,
+      "learning_rate": 1.796801102205536e-05,
+      "loss": 0.721,
+      "step": 9777
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3719797073726867,
+      "learning_rate": 1.7967549937650193e-05,
+      "loss": 0.6879,
+      "step": 9778
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5347873963363832,
+      "learning_rate": 1.796708880685543e-05,
+      "loss": 0.7983,
+      "step": 9779
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6274130294140211,
+      "learning_rate": 1.796662762967375e-05,
+      "loss": 0.5858,
+      "step": 9780
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.305003159276482,
+      "learning_rate": 1.7966166406107847e-05,
+      "loss": 0.7433,
+      "step": 9781
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8064406375368123,
+      "learning_rate": 1.7965705136160397e-05,
+      "loss": 0.6789,
+      "step": 9782
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4993880340111256,
+      "learning_rate": 1.7965243819834092e-05,
+      "loss": 0.7482,
+      "step": 9783
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8001104879405674,
+      "learning_rate": 1.7964782457131615e-05,
+      "loss": 0.7432,
+      "step": 9784
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9062642046011253,
+      "learning_rate": 1.7964321048055656e-05,
+      "loss": 0.8032,
+      "step": 9785
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5753148559892203,
+      "learning_rate": 1.7963859592608896e-05,
+      "loss": 0.7951,
+      "step": 9786
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4344138211395887,
+      "learning_rate": 1.796339809079402e-05,
+      "loss": 0.7078,
+      "step": 9787
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9877370759523405,
+      "learning_rate": 1.7962936542613723e-05,
+      "loss": 0.8087,
+      "step": 9788
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4602761389548637,
+      "learning_rate": 1.7962474948070688e-05,
+      "loss": 0.7782,
+      "step": 9789
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5391644994083,
+      "learning_rate": 1.7962013307167602e-05,
+      "loss": 0.8998,
+      "step": 9790
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9488657778494265,
+      "learning_rate": 1.7961551619907158e-05,
+      "loss": 0.7716,
+      "step": 9791
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4889120509219615,
+      "learning_rate": 1.7961089886292036e-05,
+      "loss": 0.6727,
+      "step": 9792
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.460608554472402,
+      "learning_rate": 1.7960628106324927e-05,
+      "loss": 0.7777,
+      "step": 9793
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.038004993384776,
+      "learning_rate": 1.7960166280008528e-05,
+      "loss": 0.7217,
+      "step": 9794
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3833925242101053,
+      "learning_rate": 1.795970440734551e-05,
+      "loss": 0.766,
+      "step": 9795
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6206949574646125,
+      "learning_rate": 1.795924248833858e-05,
+      "loss": 0.5725,
+      "step": 9796
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1604094048168774,
+      "learning_rate": 1.795878052299042e-05,
+      "loss": 0.6473,
+      "step": 9797
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3496255962505241,
+      "learning_rate": 1.7958318511303717e-05,
+      "loss": 0.662,
+      "step": 9798
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.788636583866124,
+      "learning_rate": 1.795785645328117e-05,
+      "loss": 0.744,
+      "step": 9799
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5158385524011315,
+      "learning_rate": 1.795739434892546e-05,
+      "loss": 0.8467,
+      "step": 9800
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6935959359885844,
+      "learning_rate": 1.7956932198239278e-05,
+      "loss": 0.696,
+      "step": 9801
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5602089086763198,
+      "learning_rate": 1.7956470001225322e-05,
+      "loss": 0.7292,
+      "step": 9802
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3708628535480132,
+      "learning_rate": 1.7956007757886275e-05,
+      "loss": 0.7117,
+      "step": 9803
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5597242648184724,
+      "learning_rate": 1.7955545468224832e-05,
+      "loss": 0.8157,
+      "step": 9804
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.700747157279284,
+      "learning_rate": 1.7955083132243687e-05,
+      "loss": 0.762,
+      "step": 9805
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5871569562589448,
+      "learning_rate": 1.795462074994553e-05,
+      "loss": 0.5446,
+      "step": 9806
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.593624001638407,
+      "learning_rate": 1.795415832133305e-05,
+      "loss": 0.7818,
+      "step": 9807
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6477122720516686,
+      "learning_rate": 1.7953695846408945e-05,
+      "loss": 0.7822,
+      "step": 9808
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3588316718460762,
+      "learning_rate": 1.7953233325175904e-05,
+      "loss": 0.7645,
+      "step": 9809
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.443264487335706,
+      "learning_rate": 1.795277075763662e-05,
+      "loss": 0.7865,
+      "step": 9810
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6299773433502613,
+      "learning_rate": 1.7952308143793783e-05,
+      "loss": 0.7383,
+      "step": 9811
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.593753464093228,
+      "learning_rate": 1.7951845483650093e-05,
+      "loss": 0.5438,
+      "step": 9812
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.399422271528744,
+      "learning_rate": 1.7951382777208243e-05,
+      "loss": 0.7795,
+      "step": 9813
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6437506327121063,
+      "learning_rate": 1.7950920024470923e-05,
+      "loss": 0.7995,
+      "step": 9814
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.366180714117173,
+      "learning_rate": 1.795045722544083e-05,
+      "loss": 0.8141,
+      "step": 9815
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4122455856644258,
+      "learning_rate": 1.794999438012066e-05,
+      "loss": 0.6878,
+      "step": 9816
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.529729410689447,
+      "learning_rate": 1.7949531488513106e-05,
+      "loss": 0.8208,
+      "step": 9817
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6449127981695264,
+      "learning_rate": 1.794906855062086e-05,
+      "loss": 0.787,
+      "step": 9818
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.486654676037116,
+      "learning_rate": 1.794860556644662e-05,
+      "loss": 0.7824,
+      "step": 9819
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3525889631475534,
+      "learning_rate": 1.7948142535993084e-05,
+      "loss": 0.7935,
+      "step": 9820
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.560901261947748,
+      "learning_rate": 1.7947679459262945e-05,
+      "loss": 0.8836,
+      "step": 9821
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3040558207705844,
+      "learning_rate": 1.79472163362589e-05,
+      "loss": 0.5854,
+      "step": 9822
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6000342922477502,
+      "learning_rate": 1.7946753166983642e-05,
+      "loss": 0.727,
+      "step": 9823
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7980543001857034,
+      "learning_rate": 1.7946289951439876e-05,
+      "loss": 0.7067,
+      "step": 9824
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.547120084253619,
+      "learning_rate": 1.7945826689630293e-05,
+      "loss": 0.7664,
+      "step": 9825
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9533138524111504,
+      "learning_rate": 1.7945363381557594e-05,
+      "loss": 0.8479,
+      "step": 9826
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.730450223939676,
+      "learning_rate": 1.794490002722447e-05,
+      "loss": 0.735,
+      "step": 9827
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.734929470702125,
+      "learning_rate": 1.7944436626633625e-05,
+      "loss": 0.7454,
+      "step": 9828
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4193118929278086,
+      "learning_rate": 1.7943973179787753e-05,
+      "loss": 0.7618,
+      "step": 9829
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3427512793555036,
+      "learning_rate": 1.7943509686689556e-05,
+      "loss": 0.6894,
+      "step": 9830
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5982860616298877,
+      "learning_rate": 1.7943046147341727e-05,
+      "loss": 0.7333,
+      "step": 9831
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5162340008240018,
+      "learning_rate": 1.7942582561746973e-05,
+      "loss": 0.9049,
+      "step": 9832
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6655884175753335,
+      "learning_rate": 1.7942118929907987e-05,
+      "loss": 0.7884,
+      "step": 9833
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2041384863504523,
+      "learning_rate": 1.7941655251827467e-05,
+      "loss": 0.8817,
+      "step": 9834
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4046108876257306,
+      "learning_rate": 1.794119152750812e-05,
+      "loss": 0.8179,
+      "step": 9835
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7212493698441298,
+      "learning_rate": 1.794072775695264e-05,
+      "loss": 0.7992,
+      "step": 9836
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.476785241091948,
+      "learning_rate": 1.7940263940163732e-05,
+      "loss": 0.8304,
+      "step": 9837
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.678734041240648,
+      "learning_rate": 1.793980007714409e-05,
+      "loss": 0.872,
+      "step": 9838
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4078335916306641,
+      "learning_rate": 1.7939336167896416e-05,
+      "loss": 0.8343,
+      "step": 9839
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4785974997714126,
+      "learning_rate": 1.7938872212423417e-05,
+      "loss": 0.6791,
+      "step": 9840
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9359679982914073,
+      "learning_rate": 1.793840821072779e-05,
+      "loss": 0.8959,
+      "step": 9841
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7931520113306592,
+      "learning_rate": 1.7937944162812238e-05,
+      "loss": 0.7086,
+      "step": 9842
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5173828682578463,
+      "learning_rate": 1.7937480068679462e-05,
+      "loss": 0.7869,
+      "step": 9843
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4378090333152072,
+      "learning_rate": 1.7937015928332165e-05,
+      "loss": 0.7032,
+      "step": 9844
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7741607376998452,
+      "learning_rate": 1.7936551741773043e-05,
+      "loss": 0.7546,
+      "step": 9845
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4661280090386906,
+      "learning_rate": 1.7936087509004812e-05,
+      "loss": 0.7748,
+      "step": 9846
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3053684540164912,
+      "learning_rate": 1.793562323003016e-05,
+      "loss": 0.6105,
+      "step": 9847
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7024978722222173,
+      "learning_rate": 1.79351589048518e-05,
+      "loss": 0.7379,
+      "step": 9848
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5165670249379823,
+      "learning_rate": 1.7934694533472433e-05,
+      "loss": 0.7704,
+      "step": 9849
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8664192910986144,
+      "learning_rate": 1.7934230115894765e-05,
+      "loss": 0.7925,
+      "step": 9850
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7174726612175433,
+      "learning_rate": 1.7933765652121493e-05,
+      "loss": 0.793,
+      "step": 9851
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8698423027114643,
+      "learning_rate": 1.793330114215533e-05,
+      "loss": 0.8455,
+      "step": 9852
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3373580166579646,
+      "learning_rate": 1.7932836585998973e-05,
+      "loss": 0.8902,
+      "step": 9853
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7572322035793337,
+      "learning_rate": 1.7932371983655125e-05,
+      "loss": 0.7354,
+      "step": 9854
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.542803104848005,
+      "learning_rate": 1.7931907335126505e-05,
+      "loss": 0.6413,
+      "step": 9855
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.7796786317099116,
+      "learning_rate": 1.7931442640415805e-05,
+      "loss": 0.735,
+      "step": 9856
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6902801737241782,
+      "learning_rate": 1.7930977899525736e-05,
+      "loss": 0.6447,
+      "step": 9857
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5563775015776633,
+      "learning_rate": 1.7930513112459002e-05,
+      "loss": 0.8054,
+      "step": 9858
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7438492556284404,
+      "learning_rate": 1.7930048279218306e-05,
+      "loss": 0.7161,
+      "step": 9859
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9339496090167743,
+      "learning_rate": 1.7929583399806363e-05,
+      "loss": 0.6363,
+      "step": 9860
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8447714889775948,
+      "learning_rate": 1.7929118474225876e-05,
+      "loss": 0.7667,
+      "step": 9861
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5042475103607407,
+      "learning_rate": 1.7928653502479546e-05,
+      "loss": 0.766,
+      "step": 9862
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3932659547851283,
+      "learning_rate": 1.792818848457009e-05,
+      "loss": 0.7184,
+      "step": 9863
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7197857723828176,
+      "learning_rate": 1.792772342050021e-05,
+      "loss": 0.8794,
+      "step": 9864
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6323163471137,
+      "learning_rate": 1.7927258310272613e-05,
+      "loss": 0.5386,
+      "step": 9865
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5947986825715208,
+      "learning_rate": 1.7926793153890006e-05,
+      "loss": 0.8008,
+      "step": 9866
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0171995854542044,
+      "learning_rate": 1.7926327951355106e-05,
+      "loss": 0.7718,
+      "step": 9867
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8550905582274293,
+      "learning_rate": 1.7925862702670608e-05,
+      "loss": 0.783,
+      "step": 9868
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7214744303302105,
+      "learning_rate": 1.7925397407839233e-05,
+      "loss": 0.6785,
+      "step": 9869
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4382940619759004,
+      "learning_rate": 1.7924932066863682e-05,
+      "loss": 0.9116,
+      "step": 9870
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3769136402986546,
+      "learning_rate": 1.792446667974667e-05,
+      "loss": 0.8376,
+      "step": 9871
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3548378924908466,
+      "learning_rate": 1.7924001246490907e-05,
+      "loss": 0.7522,
+      "step": 9872
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4715294313124543,
+      "learning_rate": 1.7923535767099093e-05,
+      "loss": 0.7439,
+      "step": 9873
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.474253418473689,
+      "learning_rate": 1.792307024157395e-05,
+      "loss": 0.8242,
+      "step": 9874
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4878636552256885,
+      "learning_rate": 1.792260466991818e-05,
+      "loss": 0.8543,
+      "step": 9875
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5938882379481705,
+      "learning_rate": 1.79221390521345e-05,
+      "loss": 0.81,
+      "step": 9876
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5185642125127659,
+      "learning_rate": 1.792167338822562e-05,
+      "loss": 0.7042,
+      "step": 9877
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9232063435139497,
+      "learning_rate": 1.792120767819425e-05,
+      "loss": 0.8876,
+      "step": 9878
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6177356680477566,
+      "learning_rate": 1.7920741922043093e-05,
+      "loss": 0.7258,
+      "step": 9879
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9204540973507531,
+      "learning_rate": 1.7920276119774876e-05,
+      "loss": 0.6735,
+      "step": 9880
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5163455403608224,
+      "learning_rate": 1.7919810271392305e-05,
+      "loss": 0.798,
+      "step": 9881
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6734034062075966,
+      "learning_rate": 1.7919344376898088e-05,
+      "loss": 0.6095,
+      "step": 9882
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.298047175124957,
+      "learning_rate": 1.791887843629494e-05,
+      "loss": 0.7535,
+      "step": 9883
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9653493676694513,
+      "learning_rate": 1.7918412449585577e-05,
+      "loss": 0.7728,
+      "step": 9884
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5921393646202227,
+      "learning_rate": 1.7917946416772708e-05,
+      "loss": 0.844,
+      "step": 9885
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5018844703611725,
+      "learning_rate": 1.791748033785905e-05,
+      "loss": 0.82,
+      "step": 9886
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5623511041910008,
+      "learning_rate": 1.7917014212847314e-05,
+      "loss": 0.6442,
+      "step": 9887
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6282029809536215,
+      "learning_rate": 1.7916548041740213e-05,
+      "loss": 0.565,
+      "step": 9888
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4009590975480026,
+      "learning_rate": 1.7916081824540466e-05,
+      "loss": 0.6856,
+      "step": 9889
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5927173229593963,
+      "learning_rate": 1.7915615561250783e-05,
+      "loss": 0.8865,
+      "step": 9890
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5976320487369704,
+      "learning_rate": 1.791514925187388e-05,
+      "loss": 0.7491,
+      "step": 9891
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7759929745399994,
+      "learning_rate": 1.791468289641247e-05,
+      "loss": 0.7341,
+      "step": 9892
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7988412668134974,
+      "learning_rate": 1.7914216494869278e-05,
+      "loss": 0.6451,
+      "step": 9893
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4968180318960895,
+      "learning_rate": 1.7913750047247004e-05,
+      "loss": 0.7662,
+      "step": 9894
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4526969784771868,
+      "learning_rate": 1.7913283553548374e-05,
+      "loss": 0.8417,
+      "step": 9895
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6034222242505214,
+      "learning_rate": 1.7912817013776102e-05,
+      "loss": 0.567,
+      "step": 9896
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8108464683995815,
+      "learning_rate": 1.7912350427932905e-05,
+      "loss": 0.8729,
+      "step": 9897
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.31824940349177,
+      "learning_rate": 1.79118837960215e-05,
+      "loss": 0.6393,
+      "step": 9898
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8011678571315362,
+      "learning_rate": 1.79114171180446e-05,
+      "loss": 0.6795,
+      "step": 9899
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.397160993159976,
+      "learning_rate": 1.7910950394004926e-05,
+      "loss": 0.8146,
+      "step": 9900
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6137533735197509,
+      "learning_rate": 1.7910483623905195e-05,
+      "loss": 0.5086,
+      "step": 9901
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4694776197965176,
+      "learning_rate": 1.791001680774812e-05,
+      "loss": 0.847,
+      "step": 9902
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4556191093183644,
+      "learning_rate": 1.7909549945536427e-05,
+      "loss": 0.8,
+      "step": 9903
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5882492366867274,
+      "learning_rate": 1.790908303727283e-05,
+      "loss": 0.7959,
+      "step": 9904
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6961701174498238,
+      "learning_rate": 1.7908616082960047e-05,
+      "loss": 0.8528,
+      "step": 9905
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7037203618401693,
+      "learning_rate": 1.7908149082600796e-05,
+      "loss": 0.648,
+      "step": 9906
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5231278046823906,
+      "learning_rate": 1.79076820361978e-05,
+      "loss": 0.88,
+      "step": 9907
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1774345500869607,
+      "learning_rate": 1.790721494375377e-05,
+      "loss": 0.8064,
+      "step": 9908
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.495633360933925,
+      "learning_rate": 1.790674780527144e-05,
+      "loss": 0.7808,
+      "step": 9909
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3646264270610646,
+      "learning_rate": 1.7906280620753517e-05,
+      "loss": 0.7308,
+      "step": 9910
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.010466388510752,
+      "learning_rate": 1.790581339020272e-05,
+      "loss": 0.6208,
+      "step": 9911
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3201248712174256,
+      "learning_rate": 1.790534611362178e-05,
+      "loss": 0.8694,
+      "step": 9912
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5365077835952963,
+      "learning_rate": 1.7904878791013412e-05,
+      "loss": 0.6989,
+      "step": 9913
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6187917896873187,
+      "learning_rate": 1.7904411422380336e-05,
+      "loss": 0.7716,
+      "step": 9914
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.285713029480323,
+      "learning_rate": 1.7903944007725273e-05,
+      "loss": 0.7755,
+      "step": 9915
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1865702924115524,
+      "learning_rate": 1.790347654705095e-05,
+      "loss": 0.7332,
+      "step": 9916
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5978569745976948,
+      "learning_rate": 1.790300904036008e-05,
+      "loss": 0.5585,
+      "step": 9917
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6877789430587173,
+      "learning_rate": 1.790254148765539e-05,
+      "loss": 0.8597,
+      "step": 9918
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5256232513834738,
+      "learning_rate": 1.7902073888939603e-05,
+      "loss": 0.833,
+      "step": 9919
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5958784465349756,
+      "learning_rate": 1.7901606244215436e-05,
+      "loss": 0.6571,
+      "step": 9920
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4524147041243816,
+      "learning_rate": 1.7901138553485623e-05,
+      "loss": 0.8207,
+      "step": 9921
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3842890882960976,
+      "learning_rate": 1.7900670816752875e-05,
+      "loss": 0.7774,
+      "step": 9922
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.374847972711247,
+      "learning_rate": 1.790020303401992e-05,
+      "loss": 0.6858,
+      "step": 9923
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7746121051353205,
+      "learning_rate": 1.7899735205289483e-05,
+      "loss": 0.8647,
+      "step": 9924
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6196956613710244,
+      "learning_rate": 1.7899267330564287e-05,
+      "loss": 0.5281,
+      "step": 9925
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5362275208968015,
+      "learning_rate": 1.7898799409847055e-05,
+      "loss": 0.7865,
+      "step": 9926
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.026588915444083,
+      "learning_rate": 1.7898331443140512e-05,
+      "loss": 0.8014,
+      "step": 9927
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.459757578602787,
+      "learning_rate": 1.7897863430447384e-05,
+      "loss": 0.8345,
+      "step": 9928
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6329420226004692,
+      "learning_rate": 1.7897395371770393e-05,
+      "loss": 0.705,
+      "step": 9929
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5406099707825085,
+      "learning_rate": 1.789692726711227e-05,
+      "loss": 0.7973,
+      "step": 9930
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4147210632815095,
+      "learning_rate": 1.789645911647573e-05,
+      "loss": 0.7422,
+      "step": 9931
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6275914972904952,
+      "learning_rate": 1.789599091986351e-05,
+      "loss": 0.9353,
+      "step": 9932
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4901817147329948,
+      "learning_rate": 1.7895522677278327e-05,
+      "loss": 0.6512,
+      "step": 9933
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.401213106484642,
+      "learning_rate": 1.7895054388722913e-05,
+      "loss": 0.6719,
+      "step": 9934
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.298244073520242,
+      "learning_rate": 1.7894586054199993e-05,
+      "loss": 0.7941,
+      "step": 9935
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.417866243762378,
+      "learning_rate": 1.7894117673712295e-05,
+      "loss": 0.8405,
+      "step": 9936
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6507344226405842,
+      "learning_rate": 1.7893649247262543e-05,
+      "loss": 0.7968,
+      "step": 9937
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6813210722841434,
+      "learning_rate": 1.7893180774853467e-05,
+      "loss": 0.8737,
+      "step": 9938
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4029678773299101,
+      "learning_rate": 1.7892712256487792e-05,
+      "loss": 0.8166,
+      "step": 9939
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6399147371489093,
+      "learning_rate": 1.789224369216825e-05,
+      "loss": 0.7818,
+      "step": 9940
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.563473371731403,
+      "learning_rate": 1.7891775081897564e-05,
+      "loss": 0.5253,
+      "step": 9941
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3212737013069447,
+      "learning_rate": 1.7891306425678467e-05,
+      "loss": 0.7764,
+      "step": 9942
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4378186557346935,
+      "learning_rate": 1.7890837723513686e-05,
+      "loss": 0.7728,
+      "step": 9943
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5330257579026898,
+      "learning_rate": 1.7890368975405947e-05,
+      "loss": 0.7628,
+      "step": 9944
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3970820017539705,
+      "learning_rate": 1.7889900181357983e-05,
+      "loss": 0.7118,
+      "step": 9945
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4937278419094957,
+      "learning_rate": 1.7889431341372523e-05,
+      "loss": 0.7117,
+      "step": 9946
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.334108719994962,
+      "learning_rate": 1.7888962455452296e-05,
+      "loss": 0.7518,
+      "step": 9947
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4433082107142363,
+      "learning_rate": 1.788849352360003e-05,
+      "loss": 0.8615,
+      "step": 9948
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5774990575601928,
+      "learning_rate": 1.7888024545818458e-05,
+      "loss": 0.8323,
+      "step": 9949
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8023969051809428,
+      "learning_rate": 1.788755552211031e-05,
+      "loss": 0.7639,
+      "step": 9950
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6123198032343335,
+      "learning_rate": 1.7887086452478318e-05,
+      "loss": 0.566,
+      "step": 9951
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.578514292373584,
+      "learning_rate": 1.788661733692521e-05,
+      "loss": 0.7884,
+      "step": 9952
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5480738041239817,
+      "learning_rate": 1.788614817545372e-05,
+      "loss": 0.868,
+      "step": 9953
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.964862341732405,
+      "learning_rate": 1.7885678968066576e-05,
+      "loss": 0.8198,
+      "step": 9954
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3247048301587205,
+      "learning_rate": 1.7885209714766517e-05,
+      "loss": 0.7104,
+      "step": 9955
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5789374427316512,
+      "learning_rate": 1.788474041555627e-05,
+      "loss": 0.5218,
+      "step": 9956
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4035998977940587,
+      "learning_rate": 1.7884271070438563e-05,
+      "loss": 0.671,
+      "step": 9957
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6368141993699146,
+      "learning_rate": 1.7883801679416138e-05,
+      "loss": 0.7782,
+      "step": 9958
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5390573121463138,
+      "learning_rate": 1.7883332242491726e-05,
+      "loss": 0.7793,
+      "step": 9959
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.705293718426901,
+      "learning_rate": 1.7882862759668052e-05,
+      "loss": 0.7525,
+      "step": 9960
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8183174365826682,
+      "learning_rate": 1.7882393230947857e-05,
+      "loss": 0.7842,
+      "step": 9961
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4982318138731867,
+      "learning_rate": 1.7881923656333876e-05,
+      "loss": 0.6674,
+      "step": 9962
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6023744135907336,
+      "learning_rate": 1.788145403582884e-05,
+      "loss": 0.6109,
+      "step": 9963
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5590743160601992,
+      "learning_rate": 1.788098436943548e-05,
+      "loss": 0.8195,
+      "step": 9964
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5272354752300394,
+      "learning_rate": 1.7880514657156534e-05,
+      "loss": 0.7558,
+      "step": 9965
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7058335230013477,
+      "learning_rate": 1.788004489899474e-05,
+      "loss": 0.6594,
+      "step": 9966
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1482478099902758,
+      "learning_rate": 1.7879575094952826e-05,
+      "loss": 0.919,
+      "step": 9967
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6010510070983331,
+      "learning_rate": 1.7879105245033533e-05,
+      "loss": 0.5648,
+      "step": 9968
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5844120949275293,
+      "learning_rate": 1.7878635349239593e-05,
+      "loss": 0.8159,
+      "step": 9969
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6491940152111675,
+      "learning_rate": 1.7878165407573747e-05,
+      "loss": 0.6886,
+      "step": 9970
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.487821045148349,
+      "learning_rate": 1.7877695420038725e-05,
+      "loss": 0.835,
+      "step": 9971
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3123956879704182,
+      "learning_rate": 1.7877225386637266e-05,
+      "loss": 0.7102,
+      "step": 9972
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5901092498284206,
+      "learning_rate": 1.787675530737211e-05,
+      "loss": 0.7645,
+      "step": 9973
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2940090145206675,
+      "learning_rate": 1.7876285182245992e-05,
+      "loss": 0.7293,
+      "step": 9974
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4322526525935328,
+      "learning_rate": 1.7875815011261645e-05,
+      "loss": 0.762,
+      "step": 9975
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0111204920334207,
+      "learning_rate": 1.7875344794421808e-05,
+      "loss": 0.6623,
+      "step": 9976
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.3607513652161978,
+      "learning_rate": 1.7874874531729225e-05,
+      "loss": 0.6997,
+      "step": 9977
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4853558277234888,
+      "learning_rate": 1.7874404223186626e-05,
+      "loss": 0.7307,
+      "step": 9978
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4724143950645898,
+      "learning_rate": 1.7873933868796753e-05,
+      "loss": 0.6922,
+      "step": 9979
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7546534214995857,
+      "learning_rate": 1.7873463468562343e-05,
+      "loss": 0.8152,
+      "step": 9980
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4657371356307574,
+      "learning_rate": 1.7872993022486137e-05,
+      "loss": 0.8905,
+      "step": 9981
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7133672701404363,
+      "learning_rate": 1.7872522530570876e-05,
+      "loss": 0.8029,
+      "step": 9982
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7535072755292163,
+      "learning_rate": 1.7872051992819295e-05,
+      "loss": 0.7718,
+      "step": 9983
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6198564442327783,
+      "learning_rate": 1.7871581409234134e-05,
+      "loss": 0.724,
+      "step": 9984
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.515534924234578,
+      "learning_rate": 1.787111077981813e-05,
+      "loss": 0.6847,
+      "step": 9985
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9527937654120837,
+      "learning_rate": 1.787064010457403e-05,
+      "loss": 0.7952,
+      "step": 9986
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.3124935196470786,
+      "learning_rate": 1.7870169383504574e-05,
+      "loss": 0.6842,
+      "step": 9987
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3704122430956691,
+      "learning_rate": 1.78696986166125e-05,
+      "loss": 0.7448,
+      "step": 9988
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3791129369063893,
+      "learning_rate": 1.7869227803900548e-05,
+      "loss": 0.8579,
+      "step": 9989
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4486661793246542,
+      "learning_rate": 1.786875694537146e-05,
+      "loss": 0.7379,
+      "step": 9990
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.566986076100083,
+      "learning_rate": 1.7868286041027977e-05,
+      "loss": 0.6956,
+      "step": 9991
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3568425741795436,
+      "learning_rate": 1.7867815090872845e-05,
+      "loss": 0.6948,
+      "step": 9992
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2850627424005152,
+      "learning_rate": 1.78673440949088e-05,
+      "loss": 0.729,
+      "step": 9993
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9706100737131058,
+      "learning_rate": 1.7866873053138588e-05,
+      "loss": 0.7779,
+      "step": 9994
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4381398933468046,
+      "learning_rate": 1.786640196556495e-05,
+      "loss": 0.7977,
+      "step": 9995
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5952756990594591,
+      "learning_rate": 1.786593083219063e-05,
+      "loss": 0.5585,
+      "step": 9996
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6173493336302556,
+      "learning_rate": 1.7865459653018368e-05,
+      "loss": 0.7898,
+      "step": 9997
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5735135362402066,
+      "learning_rate": 1.7864988428050912e-05,
+      "loss": 0.7087,
+      "step": 9998
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.61819992689538,
+      "learning_rate": 1.7864517157291e-05,
+      "loss": 0.7074,
+      "step": 9999
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.448308499505195,
+      "learning_rate": 1.7864045840741383e-05,
+      "loss": 0.7763,
+      "step": 10000
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6051495386319684,
+      "learning_rate": 1.78635744784048e-05,
+      "loss": 0.571,
+      "step": 10001
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4679838511437031,
+      "learning_rate": 1.7863103070283995e-05,
+      "loss": 0.7578,
+      "step": 10002
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6707340254744885,
+      "learning_rate": 1.7862631616381717e-05,
+      "loss": 0.724,
+      "step": 10003
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3611054008770453,
+      "learning_rate": 1.786216011670071e-05,
+      "loss": 0.6496,
+      "step": 10004
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.348106356067243,
+      "learning_rate": 1.7861688571243715e-05,
+      "loss": 0.7325,
+      "step": 10005
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6330423459892214,
+      "learning_rate": 1.786121698001348e-05,
+      "loss": 0.6739,
+      "step": 10006
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6227095472005433,
+      "learning_rate": 1.786074534301275e-05,
+      "loss": 0.558,
+      "step": 10007
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.743272419933936,
+      "learning_rate": 1.7860273660244275e-05,
+      "loss": 0.7621,
+      "step": 10008
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5485305792992796,
+      "learning_rate": 1.7859801931710798e-05,
+      "loss": 0.7686,
+      "step": 10009
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3217465648257603,
+      "learning_rate": 1.7859330157415065e-05,
+      "loss": 0.7052,
+      "step": 10010
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4008256625841826,
+      "learning_rate": 1.785885833735982e-05,
+      "loss": 0.7592,
+      "step": 10011
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3199796360922527,
+      "learning_rate": 1.785838647154782e-05,
+      "loss": 0.8245,
+      "step": 10012
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6064024273005628,
+      "learning_rate": 1.7857914559981805e-05,
+      "loss": 0.8825,
+      "step": 10013
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7776374517242397,
+      "learning_rate": 1.785744260266452e-05,
+      "loss": 0.7629,
+      "step": 10014
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4172183084225374,
+      "learning_rate": 1.7856970599598716e-05,
+      "loss": 0.7321,
+      "step": 10015
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4882718735321139,
+      "learning_rate": 1.7856498550787144e-05,
+      "loss": 0.7765,
+      "step": 10016
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4817343374435035,
+      "learning_rate": 1.7856026456232553e-05,
+      "loss": 0.7847,
+      "step": 10017
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7338834726857284,
+      "learning_rate": 1.7855554315937685e-05,
+      "loss": 0.7553,
+      "step": 10018
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9612583180863505,
+      "learning_rate": 1.785508212990529e-05,
+      "loss": 0.858,
+      "step": 10019
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.80618184804183,
+      "learning_rate": 1.7854609898138126e-05,
+      "loss": 0.9306,
+      "step": 10020
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.057335076272263,
+      "learning_rate": 1.7854137620638933e-05,
+      "loss": 0.9345,
+      "step": 10021
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6805723875019027,
+      "learning_rate": 1.7853665297410465e-05,
+      "loss": 0.7909,
+      "step": 10022
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.3473284513178405,
+      "learning_rate": 1.7853192928455475e-05,
+      "loss": 0.7536,
+      "step": 10023
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4362301939258324,
+      "learning_rate": 1.78527205137767e-05,
+      "loss": 0.764,
+      "step": 10024
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5289456641628256,
+      "learning_rate": 1.785224805337691e-05,
+      "loss": 0.7123,
+      "step": 10025
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6572704454962552,
+      "learning_rate": 1.785177554725884e-05,
+      "loss": 0.699,
+      "step": 10026
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4756058547546658,
+      "learning_rate": 1.7851302995425247e-05,
+      "loss": 0.7849,
+      "step": 10027
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6423690007248444,
+      "learning_rate": 1.7850830397878888e-05,
+      "loss": 0.8407,
+      "step": 10028
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.566549059703599,
+      "learning_rate": 1.7850357754622503e-05,
+      "loss": 0.549,
+      "step": 10029
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4811601983661478,
+      "learning_rate": 1.7849885065658852e-05,
+      "loss": 0.7927,
+      "step": 10030
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6121627044790838,
+      "learning_rate": 1.7849412330990686e-05,
+      "loss": 0.5527,
+      "step": 10031
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4303192081689933,
+      "learning_rate": 1.7848939550620755e-05,
+      "loss": 0.6408,
+      "step": 10032
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.707128637443635,
+      "learning_rate": 1.784846672455181e-05,
+      "loss": 0.7445,
+      "step": 10033
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4479490503547081,
+      "learning_rate": 1.7847993852786612e-05,
+      "loss": 0.8489,
+      "step": 10034
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.095608377324377,
+      "learning_rate": 1.7847520935327905e-05,
+      "loss": 0.8233,
+      "step": 10035
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.520718007149774,
+      "learning_rate": 1.784704797217845e-05,
+      "loss": 0.695,
+      "step": 10036
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2590372007746478,
+      "learning_rate": 1.7846574963340995e-05,
+      "loss": 0.7657,
+      "step": 10037
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5593142537561466,
+      "learning_rate": 1.78461019088183e-05,
+      "loss": 0.8074,
+      "step": 10038
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5271925507803612,
+      "learning_rate": 1.7845628808613112e-05,
+      "loss": 0.7368,
+      "step": 10039
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3193951594738242,
+      "learning_rate": 1.784515566272819e-05,
+      "loss": 0.767,
+      "step": 10040
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4549884346267299,
+      "learning_rate": 1.7844682471166288e-05,
+      "loss": 0.831,
+      "step": 10041
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4758203013532483,
+      "learning_rate": 1.784420923393016e-05,
+      "loss": 0.7377,
+      "step": 10042
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.792855869582508,
+      "learning_rate": 1.7843735951022565e-05,
+      "loss": 0.7878,
+      "step": 10043
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6594532327114908,
+      "learning_rate": 1.7843262622446253e-05,
+      "loss": 0.7773,
+      "step": 10044
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6636449131239301,
+      "learning_rate": 1.7842789248203985e-05,
+      "loss": 0.8662,
+      "step": 10045
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2762687127920855,
+      "learning_rate": 1.7842315828298514e-05,
+      "loss": 0.6158,
+      "step": 10046
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8931599615840813,
+      "learning_rate": 1.78418423627326e-05,
+      "loss": 0.7581,
+      "step": 10047
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5029938964422962,
+      "learning_rate": 1.7841368851508995e-05,
+      "loss": 0.7811,
+      "step": 10048
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2841589669282436,
+      "learning_rate": 1.7840895294630457e-05,
+      "loss": 0.7808,
+      "step": 10049
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.459621178871132,
+      "learning_rate": 1.7840421692099746e-05,
+      "loss": 0.7186,
+      "step": 10050
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.514710928750644,
+      "learning_rate": 1.7839948043919617e-05,
+      "loss": 0.7035,
+      "step": 10051
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.460136796941447,
+      "learning_rate": 1.783947435009283e-05,
+      "loss": 0.7269,
+      "step": 10052
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3533991483086936,
+      "learning_rate": 1.7839000610622138e-05,
+      "loss": 0.75,
+      "step": 10053
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6381219886685128,
+      "learning_rate": 1.7838526825510304e-05,
+      "loss": 0.8111,
+      "step": 10054
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2762881963317416,
+      "learning_rate": 1.7838052994760087e-05,
+      "loss": 0.667,
+      "step": 10055
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5157477795869658,
+      "learning_rate": 1.783757911837424e-05,
+      "loss": 0.7644,
+      "step": 10056
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2494067176119672,
+      "learning_rate": 1.783710519635553e-05,
+      "loss": 0.7141,
+      "step": 10057
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4279836582665821,
+      "learning_rate": 1.783663122870671e-05,
+      "loss": 0.7295,
+      "step": 10058
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6473284716806385,
+      "learning_rate": 1.7836157215430544e-05,
+      "loss": 0.7605,
+      "step": 10059
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5162417579181857,
+      "learning_rate": 1.7835683156529787e-05,
+      "loss": 0.6837,
+      "step": 10060
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6781148655141582,
+      "learning_rate": 1.7835209052007206e-05,
+      "loss": 0.9657,
+      "step": 10061
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5520874262041717,
+      "learning_rate": 1.7834734901865554e-05,
+      "loss": 0.8675,
+      "step": 10062
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4177050778853546,
+      "learning_rate": 1.7834260706107597e-05,
+      "loss": 0.7095,
+      "step": 10063
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7501908195711395,
+      "learning_rate": 1.7833786464736092e-05,
+      "loss": 0.7709,
+      "step": 10064
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5437847511234453,
+      "learning_rate": 1.7833312177753805e-05,
+      "loss": 0.7977,
+      "step": 10065
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6243224191972192,
+      "learning_rate": 1.7832837845163492e-05,
+      "loss": 0.823,
+      "step": 10066
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.443661829956304,
+      "learning_rate": 1.7832363466967913e-05,
+      "loss": 0.7695,
+      "step": 10067
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3466460867885015,
+      "learning_rate": 1.783188904316984e-05,
+      "loss": 0.7818,
+      "step": 10068
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.330114894091676,
+      "learning_rate": 1.783141457377203e-05,
+      "loss": 0.7556,
+      "step": 10069
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.38858225749941,
+      "learning_rate": 1.7830940058777244e-05,
+      "loss": 0.8171,
+      "step": 10070
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5892626830710312,
+      "learning_rate": 1.7830465498188244e-05,
+      "loss": 0.7116,
+      "step": 10071
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5827422791057952,
+      "learning_rate": 1.7829990892007794e-05,
+      "loss": 0.6928,
+      "step": 10072
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4202268111228618,
+      "learning_rate": 1.7829516240238662e-05,
+      "loss": 0.8188,
+      "step": 10073
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.4092061317939013,
+      "learning_rate": 1.7829041542883606e-05,
+      "loss": 0.8228,
+      "step": 10074
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3037909450497094,
+      "learning_rate": 1.7828566799945394e-05,
+      "loss": 0.7988,
+      "step": 10075
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5235571857427925,
+      "learning_rate": 1.782809201142678e-05,
+      "loss": 0.83,
+      "step": 10076
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7612725322717004,
+      "learning_rate": 1.7827617177330544e-05,
+      "loss": 0.7058,
+      "step": 10077
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3397204377596579,
+      "learning_rate": 1.7827142297659435e-05,
+      "loss": 0.7192,
+      "step": 10078
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7262780574823624,
+      "learning_rate": 1.7826667372416233e-05,
+      "loss": 0.7478,
+      "step": 10079
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3097842984936847,
+      "learning_rate": 1.782619240160369e-05,
+      "loss": 0.7827,
+      "step": 10080
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.534848755905161,
+      "learning_rate": 1.782571738522458e-05,
+      "loss": 0.6575,
+      "step": 10081
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7996943609096692,
+      "learning_rate": 1.7825242323281666e-05,
+      "loss": 0.8225,
+      "step": 10082
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0056123157115464,
+      "learning_rate": 1.7824767215777715e-05,
+      "loss": 0.7148,
+      "step": 10083
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6755431682634703,
+      "learning_rate": 1.7824292062715488e-05,
+      "loss": 0.7604,
+      "step": 10084
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3121983753022597,
+      "learning_rate": 1.7823816864097758e-05,
+      "loss": 0.5843,
+      "step": 10085
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6622336696756388,
+      "learning_rate": 1.7823341619927288e-05,
+      "loss": 0.6194,
+      "step": 10086
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4792248966688912,
+      "learning_rate": 1.782286633020685e-05,
+      "loss": 0.7464,
+      "step": 10087
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.512660037625113,
+      "learning_rate": 1.7822390994939205e-05,
+      "loss": 0.7359,
+      "step": 10088
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7292246471011665,
+      "learning_rate": 1.782191561412712e-05,
+      "loss": 0.7544,
+      "step": 10089
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3939938801516967,
+      "learning_rate": 1.782144018777337e-05,
+      "loss": 0.8284,
+      "step": 10090
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.422184995663378,
+      "learning_rate": 1.7820964715880716e-05,
+      "loss": 0.8563,
+      "step": 10091
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4751974995986061,
+      "learning_rate": 1.7820489198451932e-05,
+      "loss": 0.7115,
+      "step": 10092
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9399679295880197,
+      "learning_rate": 1.7820013635489783e-05,
+      "loss": 0.7455,
+      "step": 10093
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4557264599053168,
+      "learning_rate": 1.781953802699704e-05,
+      "loss": 0.6683,
+      "step": 10094
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5261328250852928,
+      "learning_rate": 1.7819062372976467e-05,
+      "loss": 0.7218,
+      "step": 10095
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6552676832066022,
+      "learning_rate": 1.7818586673430842e-05,
+      "loss": 0.7683,
+      "step": 10096
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7491903602629937,
+      "learning_rate": 1.781811092836293e-05,
+      "loss": 0.8488,
+      "step": 10097
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5186788225340055,
+      "learning_rate": 1.78176351377755e-05,
+      "loss": 0.8583,
+      "step": 10098
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6661963794258736,
+      "learning_rate": 1.781715930167132e-05,
+      "loss": 0.5183,
+      "step": 10099
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5580889436341747,
+      "learning_rate": 1.7816683420053164e-05,
+      "loss": 0.7416,
+      "step": 10100
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4453238500692316,
+      "learning_rate": 1.7816207492923807e-05,
+      "loss": 0.8145,
+      "step": 10101
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5837742293130197,
+      "learning_rate": 1.781573152028601e-05,
+      "loss": 0.7701,
+      "step": 10102
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0271267189338933,
+      "learning_rate": 1.7815255502142553e-05,
+      "loss": 0.9218,
+      "step": 10103
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4836663104116619,
+      "learning_rate": 1.7814779438496207e-05,
+      "loss": 0.8193,
+      "step": 10104
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6729295555168817,
+      "learning_rate": 1.781430332934974e-05,
+      "loss": 0.6233,
+      "step": 10105
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7785651504705708,
+      "learning_rate": 1.781382717470592e-05,
+      "loss": 0.5925,
+      "step": 10106
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4620489896425732,
+      "learning_rate": 1.7813350974567528e-05,
+      "loss": 0.7343,
+      "step": 10107
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.555143388810556,
+      "learning_rate": 1.781287472893733e-05,
+      "loss": 0.7386,
+      "step": 10108
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6017857613207147,
+      "learning_rate": 1.7812398437818107e-05,
+      "loss": 0.8461,
+      "step": 10109
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5013380118722652,
+      "learning_rate": 1.7811922101212622e-05,
+      "loss": 0.7902,
+      "step": 10110
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1749165609817833,
+      "learning_rate": 1.7811445719123654e-05,
+      "loss": 0.7925,
+      "step": 10111
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9861259604896757,
+      "learning_rate": 1.781096929155398e-05,
+      "loss": 0.5913,
+      "step": 10112
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4187234963108248,
+      "learning_rate": 1.7810492818506365e-05,
+      "loss": 0.7376,
+      "step": 10113
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8141638715602482,
+      "learning_rate": 1.7810016299983594e-05,
+      "loss": 0.6627,
+      "step": 10114
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5171908875523121,
+      "learning_rate": 1.780953973598843e-05,
+      "loss": 0.7981,
+      "step": 10115
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5317443959390287,
+      "learning_rate": 1.7809063126523653e-05,
+      "loss": 0.7193,
+      "step": 10116
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4384043974313685,
+      "learning_rate": 1.780858647159204e-05,
+      "loss": 0.7305,
+      "step": 10117
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5836877558350357,
+      "learning_rate": 1.7808109771196364e-05,
+      "loss": 0.5288,
+      "step": 10118
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6685146715479415,
+      "learning_rate": 1.7807633025339403e-05,
+      "loss": 0.8125,
+      "step": 10119
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.618602731418713,
+      "learning_rate": 1.780715623402393e-05,
+      "loss": 0.8203,
+      "step": 10120
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6068727168572394,
+      "learning_rate": 1.780667939725272e-05,
+      "loss": 0.6909,
+      "step": 10121
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.60435545610123,
+      "learning_rate": 1.780620251502855e-05,
+      "loss": 0.5932,
+      "step": 10122
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6463691496570647,
+      "learning_rate": 1.78057255873542e-05,
+      "loss": 0.6955,
+      "step": 10123
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5673096286821,
+      "learning_rate": 1.780524861423244e-05,
+      "loss": 0.734,
+      "step": 10124
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.494584460561133,
+      "learning_rate": 1.7804771595666057e-05,
+      "loss": 0.7974,
+      "step": 10125
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5417262617755785,
+      "learning_rate": 1.7804294531657822e-05,
+      "loss": 0.7954,
+      "step": 10126
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6188120233519043,
+      "learning_rate": 1.7803817422210512e-05,
+      "loss": 0.7795,
+      "step": 10127
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4645449623621196,
+      "learning_rate": 1.7803340267326907e-05,
+      "loss": 0.7642,
+      "step": 10128
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3675254272594561,
+      "learning_rate": 1.7802863067009783e-05,
+      "loss": 0.7056,
+      "step": 10129
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9879674787605794,
+      "learning_rate": 1.7802385821261922e-05,
+      "loss": 0.6755,
+      "step": 10130
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3906616074017852,
+      "learning_rate": 1.78019085300861e-05,
+      "loss": 0.8447,
+      "step": 10131
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3229914287740208,
+      "learning_rate": 1.7801431193485097e-05,
+      "loss": 0.815,
+      "step": 10132
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6862451202332858,
+      "learning_rate": 1.780095381146169e-05,
+      "loss": 0.8235,
+      "step": 10133
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5646420559161767,
+      "learning_rate": 1.7800476384018658e-05,
+      "loss": 0.8254,
+      "step": 10134
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4283567028433586,
+      "learning_rate": 1.7799998911158786e-05,
+      "loss": 0.6672,
+      "step": 10135
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4077680711903895,
+      "learning_rate": 1.7799521392884854e-05,
+      "loss": 0.6879,
+      "step": 10136
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.848163053005286,
+      "learning_rate": 1.7799043829199635e-05,
+      "loss": 0.8095,
+      "step": 10137
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7626891897301917,
+      "learning_rate": 1.7798566220105912e-05,
+      "loss": 0.7729,
+      "step": 10138
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5537830449902703,
+      "learning_rate": 1.7798088565606473e-05,
+      "loss": 0.8095,
+      "step": 10139
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6262749380269097,
+      "learning_rate": 1.779761086570409e-05,
+      "loss": 0.6701,
+      "step": 10140
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.583335097904739,
+      "learning_rate": 1.779713312040155e-05,
+      "loss": 0.7363,
+      "step": 10141
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6008694485245617,
+      "learning_rate": 1.7796655329701627e-05,
+      "loss": 0.5915,
+      "step": 10142
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.448762827242016,
+      "learning_rate": 1.7796177493607113e-05,
+      "loss": 0.9034,
+      "step": 10143
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6211689283100496,
+      "learning_rate": 1.7795699612120784e-05,
+      "loss": 0.5497,
+      "step": 10144
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6114483167062892,
+      "learning_rate": 1.7795221685245425e-05,
+      "loss": 0.7609,
+      "step": 10145
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.502315170193274,
+      "learning_rate": 1.779474371298382e-05,
+      "loss": 0.8022,
+      "step": 10146
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9705826765032601,
+      "learning_rate": 1.7794265695338743e-05,
+      "loss": 0.6743,
+      "step": 10147
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0309138335508132,
+      "learning_rate": 1.779378763231299e-05,
+      "loss": 0.7152,
+      "step": 10148
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5278694177155236,
+      "learning_rate": 1.779330952390933e-05,
+      "loss": 0.8253,
+      "step": 10149
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.728195119252923,
+      "learning_rate": 1.779283137013056e-05,
+      "loss": 0.8039,
+      "step": 10150
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5403351550989435,
+      "learning_rate": 1.779235317097946e-05,
+      "loss": 0.6337,
+      "step": 10151
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7188274214725334,
+      "learning_rate": 1.779187492645881e-05,
+      "loss": 0.6263,
+      "step": 10152
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.623062093030914,
+      "learning_rate": 1.77913966365714e-05,
+      "loss": 0.7836,
+      "step": 10153
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4052746264754714,
+      "learning_rate": 1.779091830132001e-05,
+      "loss": 0.7434,
+      "step": 10154
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.854226558220504,
+      "learning_rate": 1.7790439920707424e-05,
+      "loss": 0.7706,
+      "step": 10155
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.401905007888,
+      "learning_rate": 1.7789961494736438e-05,
+      "loss": 0.7617,
+      "step": 10156
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7713077103455743,
+      "learning_rate": 1.7789483023409822e-05,
+      "loss": 0.717,
+      "step": 10157
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4239252680588623,
+      "learning_rate": 1.7789004506730376e-05,
+      "loss": 0.7215,
+      "step": 10158
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8761891841490195,
+      "learning_rate": 1.7788525944700873e-05,
+      "loss": 0.7711,
+      "step": 10159
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.3146298201479523,
+      "learning_rate": 1.778804733732411e-05,
+      "loss": 0.7013,
+      "step": 10160
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6122879588167744,
+      "learning_rate": 1.778756868460287e-05,
+      "loss": 0.5656,
+      "step": 10161
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3959465138657239,
+      "learning_rate": 1.778708998653994e-05,
+      "loss": 0.6196,
+      "step": 10162
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4389166653403596,
+      "learning_rate": 1.7786611243138104e-05,
+      "loss": 0.8017,
+      "step": 10163
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.615067705633958,
+      "learning_rate": 1.7786132454400155e-05,
+      "loss": 0.7233,
+      "step": 10164
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5701043383783555,
+      "learning_rate": 1.778565362032888e-05,
+      "loss": 0.7014,
+      "step": 10165
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5614850355425627,
+      "learning_rate": 1.778517474092706e-05,
+      "loss": 0.7231,
+      "step": 10166
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3331451505304228,
+      "learning_rate": 1.778469581619749e-05,
+      "loss": 0.7492,
+      "step": 10167
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4335805408819504,
+      "learning_rate": 1.7784216846142955e-05,
+      "loss": 0.6826,
+      "step": 10168
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7313753909065424,
+      "learning_rate": 1.7783737830766247e-05,
+      "loss": 0.7413,
+      "step": 10169
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.028411221550382,
+      "learning_rate": 1.7783258770070153e-05,
+      "loss": 0.7688,
+      "step": 10170
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6485546748771136,
+      "learning_rate": 1.778277966405746e-05,
+      "loss": 0.7983,
+      "step": 10171
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7290459505623381,
+      "learning_rate": 1.7782300512730965e-05,
+      "loss": 0.8858,
+      "step": 10172
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6231980230840545,
+      "learning_rate": 1.7781821316093445e-05,
+      "loss": 0.5766,
+      "step": 10173
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4834624184691234,
+      "learning_rate": 1.7781342074147706e-05,
+      "loss": 0.8462,
+      "step": 10174
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.46302281899584,
+      "learning_rate": 1.7780862786896526e-05,
+      "loss": 0.8279,
+      "step": 10175
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5597769271602977,
+      "learning_rate": 1.7780383454342696e-05,
+      "loss": 0.7686,
+      "step": 10176
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5182292290966948,
+      "learning_rate": 1.7779904076489013e-05,
+      "loss": 0.8044,
+      "step": 10177
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6806312396835181,
+      "learning_rate": 1.7779424653338268e-05,
+      "loss": 0.8165,
+      "step": 10178
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3066510570185237,
+      "learning_rate": 1.7778945184893245e-05,
+      "loss": 0.7099,
+      "step": 10179
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.510061157234008,
+      "learning_rate": 1.7778465671156744e-05,
+      "loss": 0.8092,
+      "step": 10180
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5854431720040385,
+      "learning_rate": 1.777798611213155e-05,
+      "loss": 0.5039,
+      "step": 10181
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5999218003158937,
+      "learning_rate": 1.777750650782046e-05,
+      "loss": 0.759,
+      "step": 10182
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.400786153241364,
+      "learning_rate": 1.7777026858226265e-05,
+      "loss": 0.8613,
+      "step": 10183
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4941464533684377,
+      "learning_rate": 1.7776547163351758e-05,
+      "loss": 0.8272,
+      "step": 10184
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4368294269486057,
+      "learning_rate": 1.777606742319973e-05,
+      "loss": 0.8433,
+      "step": 10185
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5036456854814102,
+      "learning_rate": 1.777558763777298e-05,
+      "loss": 0.7578,
+      "step": 10186
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.649335799674089,
+      "learning_rate": 1.777510780707429e-05,
+      "loss": 0.827,
+      "step": 10187
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.484429822085064,
+      "learning_rate": 1.7774627931106464e-05,
+      "loss": 0.8382,
+      "step": 10188
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6146712137914607,
+      "learning_rate": 1.7774148009872295e-05,
+      "loss": 0.7437,
+      "step": 10189
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3654666564338778,
+      "learning_rate": 1.7773668043374573e-05,
+      "loss": 0.7166,
+      "step": 10190
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3654164185566111,
+      "learning_rate": 1.7773188031616096e-05,
+      "loss": 0.7289,
+      "step": 10191
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4002004886753479,
+      "learning_rate": 1.7772707974599656e-05,
+      "loss": 0.7739,
+      "step": 10192
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.04471744104519,
+      "learning_rate": 1.777222787232805e-05,
+      "loss": 0.7613,
+      "step": 10193
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3867102353873908,
+      "learning_rate": 1.7771747724804073e-05,
+      "loss": 0.8092,
+      "step": 10194
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.412022182549176,
+      "learning_rate": 1.777126753203052e-05,
+      "loss": 0.6733,
+      "step": 10195
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8943340279816783,
+      "learning_rate": 1.7770787294010185e-05,
+      "loss": 0.7529,
+      "step": 10196
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4015875354847798,
+      "learning_rate": 1.7770307010745867e-05,
+      "loss": 0.727,
+      "step": 10197
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5819660961365074,
+      "learning_rate": 1.7769826682240365e-05,
+      "loss": 0.8246,
+      "step": 10198
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.650315892212324,
+      "learning_rate": 1.7769346308496466e-05,
+      "loss": 0.8125,
+      "step": 10199
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6772009579288638,
+      "learning_rate": 1.776886588951698e-05,
+      "loss": 0.7712,
+      "step": 10200
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4379475802140704,
+      "learning_rate": 1.7768385425304692e-05,
+      "loss": 0.7552,
+      "step": 10201
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.649725996780058,
+      "learning_rate": 1.7767904915862405e-05,
+      "loss": 0.6626,
+      "step": 10202
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6852071744110915,
+      "learning_rate": 1.776742436119292e-05,
+      "loss": 0.7638,
+      "step": 10203
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6074377218548,
+      "learning_rate": 1.776694376129903e-05,
+      "loss": 0.5602,
+      "step": 10204
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7659884615737835,
+      "learning_rate": 1.776646311618353e-05,
+      "loss": 0.6911,
+      "step": 10205
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.922061309955328,
+      "learning_rate": 1.776598242584923e-05,
+      "loss": 0.7188,
+      "step": 10206
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4311389071304108,
+      "learning_rate": 1.7765501690298913e-05,
+      "loss": 0.7965,
+      "step": 10207
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6923669565755923,
+      "learning_rate": 1.7765020909535395e-05,
+      "loss": 0.8487,
+      "step": 10208
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5475038156989918,
+      "learning_rate": 1.7764540083561462e-05,
+      "loss": 0.7209,
+      "step": 10209
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.67593478526091,
+      "learning_rate": 1.776405921237992e-05,
+      "loss": 0.9531,
+      "step": 10210
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8220618646727103,
+      "learning_rate": 1.776357829599357e-05,
+      "loss": 0.7065,
+      "step": 10211
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7237719766948967,
+      "learning_rate": 1.7763097334405204e-05,
+      "loss": 0.8461,
+      "step": 10212
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2733740240650622,
+      "learning_rate": 1.776261632761763e-05,
+      "loss": 0.6903,
+      "step": 10213
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6021651344816888,
+      "learning_rate": 1.7762135275633645e-05,
+      "loss": 0.5862,
+      "step": 10214
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4689187679691873,
+      "learning_rate": 1.7761654178456053e-05,
+      "loss": 0.6628,
+      "step": 10215
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5919985190575277,
+      "learning_rate": 1.776117303608765e-05,
+      "loss": 0.7878,
+      "step": 10216
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.552763400616252,
+      "learning_rate": 1.7760691848531246e-05,
+      "loss": 0.8429,
+      "step": 10217
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5219375266223847,
+      "learning_rate": 1.7760210615789633e-05,
+      "loss": 0.7866,
+      "step": 10218
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7081247771918857,
+      "learning_rate": 1.7759729337865618e-05,
+      "loss": 0.6911,
+      "step": 10219
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6180323713719941,
+      "learning_rate": 1.7759248014762e-05,
+      "loss": 0.5286,
+      "step": 10220
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.792668811586469,
+      "learning_rate": 1.7758766646481587e-05,
+      "loss": 0.765,
+      "step": 10221
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.435444907213018,
+      "learning_rate": 1.7758285233027174e-05,
+      "loss": 0.7983,
+      "step": 10222
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7254598775638055,
+      "learning_rate": 1.7757803774401572e-05,
+      "loss": 0.6805,
+      "step": 10223
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8050941492390538,
+      "learning_rate": 1.775732227060758e-05,
+      "loss": 0.6664,
+      "step": 10224
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.604011657430159,
+      "learning_rate": 1.7756840721648e-05,
+      "loss": 0.5415,
+      "step": 10225
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4470537326627093,
+      "learning_rate": 1.7756359127525638e-05,
+      "loss": 0.8446,
+      "step": 10226
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3547948430365242,
+      "learning_rate": 1.7755877488243297e-05,
+      "loss": 0.813,
+      "step": 10227
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1647721194391085,
+      "learning_rate": 1.7755395803803783e-05,
+      "loss": 0.7298,
+      "step": 10228
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1340989429466206,
+      "learning_rate": 1.7754914074209898e-05,
+      "loss": 0.6975,
+      "step": 10229
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3591583333802664,
+      "learning_rate": 1.7754432299464448e-05,
+      "loss": 0.7299,
+      "step": 10230
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6476671111097143,
+      "learning_rate": 1.775395047957024e-05,
+      "loss": 0.7884,
+      "step": 10231
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5062681279430048,
+      "learning_rate": 1.7753468614530075e-05,
+      "loss": 0.6306,
+      "step": 10232
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6275200491748198,
+      "learning_rate": 1.7752986704346764e-05,
+      "loss": 0.5363,
+      "step": 10233
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5839399889088349,
+      "learning_rate": 1.7752504749023108e-05,
+      "loss": 0.5011,
+      "step": 10234
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.583795522134508,
+      "learning_rate": 1.7752022748561917e-05,
+      "loss": 0.7486,
+      "step": 10235
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7620115178793956,
+      "learning_rate": 1.7751540702965993e-05,
+      "loss": 0.7714,
+      "step": 10236
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5823256884999274,
+      "learning_rate": 1.7751058612238147e-05,
+      "loss": 0.5449,
+      "step": 10237
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.630932264257085,
+      "learning_rate": 1.775057647638118e-05,
+      "loss": 0.8875,
+      "step": 10238
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5904834363476135,
+      "learning_rate": 1.7750094295397907e-05,
+      "loss": 0.7868,
+      "step": 10239
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4882879587829343,
+      "learning_rate": 1.774961206929113e-05,
+      "loss": 0.8737,
+      "step": 10240
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5783820673766513,
+      "learning_rate": 1.7749129798063657e-05,
+      "loss": 0.6619,
+      "step": 10241
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7186667762782932,
+      "learning_rate": 1.77486474817183e-05,
+      "loss": 0.7685,
+      "step": 10242
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8003190794204373,
+      "learning_rate": 1.774816512025786e-05,
+      "loss": 0.7871,
+      "step": 10243
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6439494151200416,
+      "learning_rate": 1.7747682713685154e-05,
+      "loss": 0.6603,
+      "step": 10244
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4569552149509284,
+      "learning_rate": 1.7747200262002983e-05,
+      "loss": 0.6887,
+      "step": 10245
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4861604043580272,
+      "learning_rate": 1.774671776521416e-05,
+      "loss": 0.7258,
+      "step": 10246
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3970088610608447,
+      "learning_rate": 1.7746235223321492e-05,
+      "loss": 0.808,
+      "step": 10247
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8394182906879617,
+      "learning_rate": 1.7745752636327793e-05,
+      "loss": 0.6486,
+      "step": 10248
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7104490652633537,
+      "learning_rate": 1.774527000423587e-05,
+      "loss": 0.6795,
+      "step": 10249
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5380693785120434,
+      "learning_rate": 1.7744787327048533e-05,
+      "loss": 0.6371,
+      "step": 10250
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5406387370347638,
+      "learning_rate": 1.7744304604768588e-05,
+      "loss": 0.8048,
+      "step": 10251
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.6053716409385665,
+      "learning_rate": 1.7743821837398854e-05,
+      "loss": 0.7831,
+      "step": 10252
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6912313763409426,
+      "learning_rate": 1.7743339024942135e-05,
+      "loss": 0.8212,
+      "step": 10253
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4857109899717051,
+      "learning_rate": 1.7742856167401248e-05,
+      "loss": 0.8995,
+      "step": 10254
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4305824434707077,
+      "learning_rate": 1.7742373264778997e-05,
+      "loss": 0.7113,
+      "step": 10255
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6549638235544162,
+      "learning_rate": 1.7741890317078205e-05,
+      "loss": 0.6042,
+      "step": 10256
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4271401796971428,
+      "learning_rate": 1.774140732430167e-05,
+      "loss": 0.6607,
+      "step": 10257
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6977091670775653,
+      "learning_rate": 1.774092428645221e-05,
+      "loss": 0.7943,
+      "step": 10258
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.629215492200516,
+      "learning_rate": 1.7740441203532638e-05,
+      "loss": 0.6699,
+      "step": 10259
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6822422051891366,
+      "learning_rate": 1.7739958075545772e-05,
+      "loss": 0.7751,
+      "step": 10260
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4762039304340873,
+      "learning_rate": 1.7739474902494416e-05,
+      "loss": 0.6144,
+      "step": 10261
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7991903677495507,
+      "learning_rate": 1.7738991684381387e-05,
+      "loss": 0.7064,
+      "step": 10262
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8556015736242777,
+      "learning_rate": 1.77385084212095e-05,
+      "loss": 0.6746,
+      "step": 10263
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2168306301859118,
+      "learning_rate": 1.7738025112981565e-05,
+      "loss": 0.8384,
+      "step": 10264
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6428314941483971,
+      "learning_rate": 1.7737541759700398e-05,
+      "loss": 0.5586,
+      "step": 10265
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4921729271978212,
+      "learning_rate": 1.7737058361368814e-05,
+      "loss": 0.6132,
+      "step": 10266
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3612826086101382,
+      "learning_rate": 1.7736574917989627e-05,
+      "loss": 0.7142,
+      "step": 10267
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6354670991815337,
+      "learning_rate": 1.773609142956565e-05,
+      "loss": 0.7067,
+      "step": 10268
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.503306622811543,
+      "learning_rate": 1.77356078960997e-05,
+      "loss": 0.8038,
+      "step": 10269
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6862188483601048,
+      "learning_rate": 1.773512431759459e-05,
+      "loss": 0.5538,
+      "step": 10270
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.338066507887761,
+      "learning_rate": 1.7734640694053143e-05,
+      "loss": 0.8473,
+      "step": 10271
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5123957161881478,
+      "learning_rate": 1.7734157025478164e-05,
+      "loss": 0.7661,
+      "step": 10272
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5929307843555843,
+      "learning_rate": 1.7733673311872478e-05,
+      "loss": 0.5658,
+      "step": 10273
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6868231757467294,
+      "learning_rate": 1.7733189553238893e-05,
+      "loss": 0.7695,
+      "step": 10274
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5737248828065633,
+      "learning_rate": 1.7732705749580235e-05,
+      "loss": 0.7507,
+      "step": 10275
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6120509184380337,
+      "learning_rate": 1.7732221900899313e-05,
+      "loss": 0.5447,
+      "step": 10276
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.568979266504299,
+      "learning_rate": 1.773173800719895e-05,
+      "loss": 0.6587,
+      "step": 10277
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.520567643073966,
+      "learning_rate": 1.7731254068481955e-05,
+      "loss": 0.7466,
+      "step": 10278
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.614941715416006,
+      "learning_rate": 1.7730770084751156e-05,
+      "loss": 0.5373,
+      "step": 10279
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7373342945985888,
+      "learning_rate": 1.7730286056009363e-05,
+      "loss": 0.7169,
+      "step": 10280
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6520764028345671,
+      "learning_rate": 1.77298019822594e-05,
+      "loss": 0.9226,
+      "step": 10281
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.384114947992362,
+      "learning_rate": 1.7729317863504082e-05,
+      "loss": 0.6859,
+      "step": 10282
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7391262001132743,
+      "learning_rate": 1.7728833699746226e-05,
+      "loss": 0.7289,
+      "step": 10283
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.316818258842207,
+      "learning_rate": 1.7728349490988653e-05,
+      "loss": 0.847,
+      "step": 10284
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0275477166533626,
+      "learning_rate": 1.7727865237234187e-05,
+      "loss": 0.7683,
+      "step": 10285
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2872456981426967,
+      "learning_rate": 1.7727380938485642e-05,
+      "loss": 0.6565,
+      "step": 10286
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6607565377795144,
+      "learning_rate": 1.7726896594745834e-05,
+      "loss": 0.701,
+      "step": 10287
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3631391260010775,
+      "learning_rate": 1.7726412206017593e-05,
+      "loss": 0.6754,
+      "step": 10288
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6641463383270578,
+      "learning_rate": 1.772592777230373e-05,
+      "loss": 0.7546,
+      "step": 10289
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3831986130434897,
+      "learning_rate": 1.7725443293607074e-05,
+      "loss": 0.6973,
+      "step": 10290
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6119728692001069,
+      "learning_rate": 1.7724958769930437e-05,
+      "loss": 0.5767,
+      "step": 10291
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4988151481885679,
+      "learning_rate": 1.7724474201276646e-05,
+      "loss": 0.66,
+      "step": 10292
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4262858290277398,
+      "learning_rate": 1.772398958764852e-05,
+      "loss": 0.8063,
+      "step": 10293
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.576735654469523,
+      "learning_rate": 1.7723504929048886e-05,
+      "loss": 0.5806,
+      "step": 10294
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6512578352283773,
+      "learning_rate": 1.7723020225480554e-05,
+      "loss": 0.7807,
+      "step": 10295
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6161213094732725,
+      "learning_rate": 1.7722535476946357e-05,
+      "loss": 0.6932,
+      "step": 10296
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3878464395026477,
+      "learning_rate": 1.7722050683449114e-05,
+      "loss": 0.7862,
+      "step": 10297
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.228755839811138,
+      "learning_rate": 1.7721565844991643e-05,
+      "loss": 0.6819,
+      "step": 10298
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.463389883447295,
+      "learning_rate": 1.7721080961576776e-05,
+      "loss": 0.8394,
+      "step": 10299
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5959305887030822,
+      "learning_rate": 1.7720596033207327e-05,
+      "loss": 0.5609,
+      "step": 10300
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.468405725033194,
+      "learning_rate": 1.7720111059886127e-05,
+      "loss": 0.7955,
+      "step": 10301
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6099416085796312,
+      "learning_rate": 1.7719626041615992e-05,
+      "loss": 0.6623,
+      "step": 10302
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3541935608475462,
+      "learning_rate": 1.7719140978399755e-05,
+      "loss": 0.6943,
+      "step": 10303
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.432725116933046,
+      "learning_rate": 1.7718655870240234e-05,
+      "loss": 0.6962,
+      "step": 10304
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.362822053370688,
+      "learning_rate": 1.7718170717140256e-05,
+      "loss": 0.7947,
+      "step": 10305
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4367213315294591,
+      "learning_rate": 1.771768551910264e-05,
+      "loss": 0.7768,
+      "step": 10306
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5942004745339542,
+      "learning_rate": 1.771720027613022e-05,
+      "loss": 0.7247,
+      "step": 10307
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.550891929586136,
+      "learning_rate": 1.7716714988225815e-05,
+      "loss": 0.9455,
+      "step": 10308
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6698554389078066,
+      "learning_rate": 1.7716229655392255e-05,
+      "loss": 0.808,
+      "step": 10309
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6695328475867552,
+      "learning_rate": 1.771574427763236e-05,
+      "loss": 0.9033,
+      "step": 10310
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.374503183391302,
+      "learning_rate": 1.771525885494896e-05,
+      "loss": 0.8484,
+      "step": 10311
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6329883121341155,
+      "learning_rate": 1.7714773387344878e-05,
+      "loss": 0.6581,
+      "step": 10312
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9294970783099603,
+      "learning_rate": 1.7714287874822945e-05,
+      "loss": 0.7295,
+      "step": 10313
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.570486610629307,
+      "learning_rate": 1.7713802317385985e-05,
+      "loss": 0.7431,
+      "step": 10314
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6281394221265189,
+      "learning_rate": 1.7713316715036827e-05,
+      "loss": 0.7283,
+      "step": 10315
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.511370747635008,
+      "learning_rate": 1.7712831067778298e-05,
+      "loss": 0.7685,
+      "step": 10316
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6302755582133468,
+      "learning_rate": 1.771234537561322e-05,
+      "loss": 0.5467,
+      "step": 10317
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7354844446010487,
+      "learning_rate": 1.771185963854443e-05,
+      "loss": 0.7707,
+      "step": 10318
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5016673864818324,
+      "learning_rate": 1.771137385657475e-05,
+      "loss": 0.6629,
+      "step": 10319
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3518942421249798,
+      "learning_rate": 1.771088802970701e-05,
+      "loss": 0.6933,
+      "step": 10320
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.574659637562272,
+      "learning_rate": 1.771040215794404e-05,
+      "loss": 0.7827,
+      "step": 10321
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4763393152218547,
+      "learning_rate": 1.7709916241288667e-05,
+      "loss": 0.7495,
+      "step": 10322
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.665424098804787,
+      "learning_rate": 1.770943027974372e-05,
+      "loss": 0.8231,
+      "step": 10323
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3540218344797872,
+      "learning_rate": 1.770894427331203e-05,
+      "loss": 0.7904,
+      "step": 10324
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.906074678432672,
+      "learning_rate": 1.7708458221996426e-05,
+      "loss": 0.6963,
+      "step": 10325
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4607247543969901,
+      "learning_rate": 1.7707972125799738e-05,
+      "loss": 0.6138,
+      "step": 10326
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6013893126116514,
+      "learning_rate": 1.7707485984724795e-05,
+      "loss": 0.8402,
+      "step": 10327
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.6921595378521843,
+      "learning_rate": 1.7706999798774426e-05,
+      "loss": 0.7953,
+      "step": 10328
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5046347677705378,
+      "learning_rate": 1.7706513567951467e-05,
+      "loss": 0.7242,
+      "step": 10329
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8271989458766311,
+      "learning_rate": 1.7706027292258747e-05,
+      "loss": 0.8017,
+      "step": 10330
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3938341124169982,
+      "learning_rate": 1.7705540971699095e-05,
+      "loss": 0.7492,
+      "step": 10331
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5673331940044726,
+      "learning_rate": 1.7705054606275343e-05,
+      "loss": 0.7462,
+      "step": 10332
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5084799439851206,
+      "learning_rate": 1.7704568195990324e-05,
+      "loss": 0.6385,
+      "step": 10333
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5672682759847956,
+      "learning_rate": 1.770408174084687e-05,
+      "loss": 0.7246,
+      "step": 10334
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4805409553149858,
+      "learning_rate": 1.7703595240847815e-05,
+      "loss": 0.7847,
+      "step": 10335
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5638685442245774,
+      "learning_rate": 1.770310869599599e-05,
+      "loss": 0.4877,
+      "step": 10336
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.59112188528397,
+      "learning_rate": 1.770262210629422e-05,
+      "loss": 0.7778,
+      "step": 10337
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4664392680543135,
+      "learning_rate": 1.7702135471745358e-05,
+      "loss": 0.8276,
+      "step": 10338
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4728445489159607,
+      "learning_rate": 1.7701648792352215e-05,
+      "loss": 0.8043,
+      "step": 10339
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6368091950644813,
+      "learning_rate": 1.7701162068117636e-05,
+      "loss": 0.5635,
+      "step": 10340
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5642158727831993,
+      "learning_rate": 1.7700675299044457e-05,
+      "loss": 0.5264,
+      "step": 10341
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3589138101724194,
+      "learning_rate": 1.7700188485135506e-05,
+      "loss": 0.7078,
+      "step": 10342
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9709492080872169,
+      "learning_rate": 1.769970162639362e-05,
+      "loss": 0.7234,
+      "step": 10343
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5667595678727018,
+      "learning_rate": 1.7699214722821634e-05,
+      "loss": 0.6037,
+      "step": 10344
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5787559402450329,
+      "learning_rate": 1.769872777442238e-05,
+      "loss": 0.6891,
+      "step": 10345
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5065998188796457,
+      "learning_rate": 1.7698240781198696e-05,
+      "loss": 0.761,
+      "step": 10346
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5576093935846238,
+      "learning_rate": 1.769775374315342e-05,
+      "loss": 0.7987,
+      "step": 10347
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.42861160621671,
+      "learning_rate": 1.7697266660289383e-05,
+      "loss": 0.7763,
+      "step": 10348
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.529684099752912,
+      "learning_rate": 1.769677953260942e-05,
+      "loss": 0.8165,
+      "step": 10349
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4782202796234292,
+      "learning_rate": 1.7696292360116373e-05,
+      "loss": 0.6531,
+      "step": 10350
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5951926160710452,
+      "learning_rate": 1.7695805142813073e-05,
+      "loss": 0.5066,
+      "step": 10351
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.5354239869638246,
+      "learning_rate": 1.769531788070236e-05,
+      "loss": 0.6558,
+      "step": 10352
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.583738305185593,
+      "learning_rate": 1.7694830573787075e-05,
+      "loss": 0.5403,
+      "step": 10353
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2759673564085157,
+      "learning_rate": 1.769434322207004e-05,
+      "loss": 0.7261,
+      "step": 10354
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5707013858130285,
+      "learning_rate": 1.7693855825554113e-05,
+      "loss": 0.6709,
+      "step": 10355
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3579823833276818,
+      "learning_rate": 1.7693368384242115e-05,
+      "loss": 0.7107,
+      "step": 10356
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6176111483636412,
+      "learning_rate": 1.7692880898136894e-05,
+      "loss": 0.5535,
+      "step": 10357
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4355957452733754,
+      "learning_rate": 1.7692393367241283e-05,
+      "loss": 0.7309,
+      "step": 10358
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4713985938355052,
+      "learning_rate": 1.7691905791558123e-05,
+      "loss": 0.9629,
+      "step": 10359
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3493212420545562,
+      "learning_rate": 1.7691418171090254e-05,
+      "loss": 0.7003,
+      "step": 10360
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4442584481841787,
+      "learning_rate": 1.7690930505840513e-05,
+      "loss": 0.8026,
+      "step": 10361
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4952312054169294,
+      "learning_rate": 1.7690442795811734e-05,
+      "loss": 0.7428,
+      "step": 10362
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.215686470641155,
+      "learning_rate": 1.7689955041006768e-05,
+      "loss": 0.8003,
+      "step": 10363
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5691452317970435,
+      "learning_rate": 1.7689467241428446e-05,
+      "loss": 0.5537,
+      "step": 10364
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4944448127922483,
+      "learning_rate": 1.7688979397079612e-05,
+      "loss": 0.6632,
+      "step": 10365
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0985427554659233,
+      "learning_rate": 1.7688491507963108e-05,
+      "loss": 0.8398,
+      "step": 10366
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5389669487933069,
+      "learning_rate": 1.7688003574081773e-05,
+      "loss": 0.7252,
+      "step": 10367
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6858214675485659,
+      "learning_rate": 1.7687515595438443e-05,
+      "loss": 0.8247,
+      "step": 10368
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5324062750464993,
+      "learning_rate": 1.7687027572035965e-05,
+      "loss": 0.6801,
+      "step": 10369
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6020296168179681,
+      "learning_rate": 1.768653950387718e-05,
+      "loss": 0.5488,
+      "step": 10370
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2975284269094713,
+      "learning_rate": 1.768605139096493e-05,
+      "loss": 0.7427,
+      "step": 10371
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4970847365601436,
+      "learning_rate": 1.7685563233302053e-05,
+      "loss": 0.7946,
+      "step": 10372
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3618934862741205,
+      "learning_rate": 1.7685075030891393e-05,
+      "loss": 0.76,
+      "step": 10373
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4031365147781871,
+      "learning_rate": 1.7684586783735793e-05,
+      "loss": 0.7624,
+      "step": 10374
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9339539400727042,
+      "learning_rate": 1.7684098491838097e-05,
+      "loss": 0.7683,
+      "step": 10375
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5709580328688325,
+      "learning_rate": 1.7683610155201148e-05,
+      "loss": 0.5429,
+      "step": 10376
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7191776807016386,
+      "learning_rate": 1.7683121773827787e-05,
+      "loss": 0.8448,
+      "step": 10377
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6134301801038908,
+      "learning_rate": 1.7682633347720858e-05,
+      "loss": 0.5886,
+      "step": 10378
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4159664692650276,
+      "learning_rate": 1.7682144876883207e-05,
+      "loss": 0.8296,
+      "step": 10379
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3133442672836184,
+      "learning_rate": 1.7681656361317676e-05,
+      "loss": 0.7959,
+      "step": 10380
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6991409367438972,
+      "learning_rate": 1.7681167801027107e-05,
+      "loss": 0.8072,
+      "step": 10381
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6180735725111122,
+      "learning_rate": 1.7680679196014348e-05,
+      "loss": 0.7641,
+      "step": 10382
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7460366043993003,
+      "learning_rate": 1.7680190546282244e-05,
+      "loss": 0.817,
+      "step": 10383
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7482162766781353,
+      "learning_rate": 1.767970185183364e-05,
+      "loss": 0.9495,
+      "step": 10384
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.64047304959232,
+      "learning_rate": 1.7679213112671378e-05,
+      "loss": 0.6258,
+      "step": 10385
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.344223280346486,
+      "learning_rate": 1.7678724328798308e-05,
+      "loss": 0.6684,
+      "step": 10386
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8913914749784222,
+      "learning_rate": 1.7678235500217274e-05,
+      "loss": 0.7302,
+      "step": 10387
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7662445838113374,
+      "learning_rate": 1.7677746626931122e-05,
+      "loss": 0.7986,
+      "step": 10388
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6359181767773487,
+      "learning_rate": 1.7677257708942697e-05,
+      "loss": 0.724,
+      "step": 10389
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0882542376792808,
+      "learning_rate": 1.767676874625485e-05,
+      "loss": 0.8202,
+      "step": 10390
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7863193446857155,
+      "learning_rate": 1.7676279738870423e-05,
+      "loss": 0.8031,
+      "step": 10391
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.485742577127214,
+      "learning_rate": 1.7675790686792264e-05,
+      "loss": 0.7597,
+      "step": 10392
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5632727808198066,
+      "learning_rate": 1.7675301590023225e-05,
+      "loss": 0.7011,
+      "step": 10393
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9420404843178956,
+      "learning_rate": 1.7674812448566147e-05,
+      "loss": 0.8628,
+      "step": 10394
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6324321466950233,
+      "learning_rate": 1.7674323262423883e-05,
+      "loss": 0.9222,
+      "step": 10395
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.441884310127338,
+      "learning_rate": 1.767383403159928e-05,
+      "loss": 0.7935,
+      "step": 10396
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5024624424744226,
+      "learning_rate": 1.767334475609518e-05,
+      "loss": 0.8389,
+      "step": 10397
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6212742939591929,
+      "learning_rate": 1.7672855435914442e-05,
+      "loss": 0.753,
+      "step": 10398
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7836894256793958,
+      "learning_rate": 1.767236607105991e-05,
+      "loss": 0.8219,
+      "step": 10399
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6841074200814141,
+      "learning_rate": 1.7671876661534436e-05,
+      "loss": 0.5414,
+      "step": 10400
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.55888379042665,
+      "learning_rate": 1.7671387207340864e-05,
+      "loss": 0.7154,
+      "step": 10401
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7839572390978156,
+      "learning_rate": 1.767089770848205e-05,
+      "loss": 0.7683,
+      "step": 10402
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.58037956210869,
+      "learning_rate": 1.7670408164960833e-05,
+      "loss": 0.7751,
+      "step": 10403
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2775979481544861,
+      "learning_rate": 1.766991857678008e-05,
+      "loss": 0.7463,
+      "step": 10404
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.575689784401476,
+      "learning_rate": 1.766942894394263e-05,
+      "loss": 0.6995,
+      "step": 10405
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8168754595056964,
+      "learning_rate": 1.7668939266451337e-05,
+      "loss": 0.7503,
+      "step": 10406
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6202169280737968,
+      "learning_rate": 1.7668449544309053e-05,
+      "loss": 0.8257,
+      "step": 10407
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.59268170526112,
+      "learning_rate": 1.7667959777518622e-05,
+      "loss": 0.5315,
+      "step": 10408
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6278166406933476,
+      "learning_rate": 1.7667469966082908e-05,
+      "loss": 0.7056,
+      "step": 10409
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.521578412566651,
+      "learning_rate": 1.7666980110004756e-05,
+      "loss": 0.8239,
+      "step": 10410
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6853619746691582,
+      "learning_rate": 1.7666490209287015e-05,
+      "loss": 0.8122,
+      "step": 10411
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.441689106698333,
+      "learning_rate": 1.7666000263932542e-05,
+      "loss": 0.7188,
+      "step": 10412
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7745498895866345,
+      "learning_rate": 1.766551027394419e-05,
+      "loss": 0.7941,
+      "step": 10413
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8957288489266404,
+      "learning_rate": 1.7665020239324808e-05,
+      "loss": 0.6761,
+      "step": 10414
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3833122308277772,
+      "learning_rate": 1.7664530160077252e-05,
+      "loss": 0.7359,
+      "step": 10415
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4098046979407135,
+      "learning_rate": 1.7664040036204375e-05,
+      "loss": 0.7822,
+      "step": 10416
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.189902351283917,
+      "learning_rate": 1.766354986770903e-05,
+      "loss": 0.8042,
+      "step": 10417
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5129360554496392,
+      "learning_rate": 1.7663059654594073e-05,
+      "loss": 0.762,
+      "step": 10418
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.3422585954086115,
+      "learning_rate": 1.766256939686236e-05,
+      "loss": 0.8962,
+      "step": 10419
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.598381541115182,
+      "learning_rate": 1.7662079094516735e-05,
+      "loss": 0.7113,
+      "step": 10420
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6829512889054266,
+      "learning_rate": 1.7661588747560062e-05,
+      "loss": 0.6736,
+      "step": 10421
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7571311738841084,
+      "learning_rate": 1.7661098355995195e-05,
+      "loss": 0.7941,
+      "step": 10422
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5991578391818286,
+      "learning_rate": 1.7660607919824988e-05,
+      "loss": 0.791,
+      "step": 10423
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.503807754908158,
+      "learning_rate": 1.76601174390523e-05,
+      "loss": 0.7857,
+      "step": 10424
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7427422490230817,
+      "learning_rate": 1.7659626913679977e-05,
+      "loss": 0.6844,
+      "step": 10425
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5510997737680503,
+      "learning_rate": 1.765913634371088e-05,
+      "loss": 0.7254,
+      "step": 10426
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.625012156409591,
+      "learning_rate": 1.765864572914787e-05,
+      "loss": 0.6358,
+      "step": 10427
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6342106883745586,
+      "learning_rate": 1.7658155069993802e-05,
+      "loss": 0.5599,
+      "step": 10428
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.620672932218367,
+      "learning_rate": 1.7657664366251527e-05,
+      "loss": 0.5386,
+      "step": 10429
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3074134885630755,
+      "learning_rate": 1.7657173617923905e-05,
+      "loss": 0.7173,
+      "step": 10430
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4788549701310476,
+      "learning_rate": 1.7656682825013794e-05,
+      "loss": 0.8026,
+      "step": 10431
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4664931534850318,
+      "learning_rate": 1.7656191987524055e-05,
+      "loss": 0.7849,
+      "step": 10432
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6730824133847404,
+      "learning_rate": 1.765570110545754e-05,
+      "loss": 0.8482,
+      "step": 10433
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.595311602339293,
+      "learning_rate": 1.765521017881711e-05,
+      "loss": 0.5613,
+      "step": 10434
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.626945254591902,
+      "learning_rate": 1.7654719207605625e-05,
+      "loss": 0.6176,
+      "step": 10435
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6117255224907925,
+      "learning_rate": 1.7654228191825937e-05,
+      "loss": 0.5259,
+      "step": 10436
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3370250641520423,
+      "learning_rate": 1.7653737131480914e-05,
+      "loss": 0.7515,
+      "step": 10437
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.589268530663085,
+      "learning_rate": 1.7653246026573404e-05,
+      "loss": 0.7,
+      "step": 10438
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5740617553982262,
+      "learning_rate": 1.7652754877106275e-05,
+      "loss": 0.5336,
+      "step": 10439
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6163245160572148,
+      "learning_rate": 1.7652263683082388e-05,
+      "loss": 0.8444,
+      "step": 10440
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.460275756979482,
+      "learning_rate": 1.7651772444504594e-05,
+      "loss": 0.6609,
+      "step": 10441
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.605606120810369,
+      "learning_rate": 1.765128116137576e-05,
+      "loss": 0.6755,
+      "step": 10442
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5650458599319288,
+      "learning_rate": 1.765078983369875e-05,
+      "loss": 0.6994,
+      "step": 10443
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.641415848848492,
+      "learning_rate": 1.765029846147641e-05,
+      "loss": 0.718,
+      "step": 10444
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4561770401912886,
+      "learning_rate": 1.764980704471162e-05,
+      "loss": 0.6877,
+      "step": 10445
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8452995445329539,
+      "learning_rate": 1.7649315583407226e-05,
+      "loss": 0.8613,
+      "step": 10446
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6118231861060892,
+      "learning_rate": 1.7648824077566095e-05,
+      "loss": 0.8198,
+      "step": 10447
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5837510044391934,
+      "learning_rate": 1.7648332527191086e-05,
+      "loss": 0.7155,
+      "step": 10448
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8196650581926344,
+      "learning_rate": 1.764784093228507e-05,
+      "loss": 0.7555,
+      "step": 10449
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7312235951215509,
+      "learning_rate": 1.7647349292850898e-05,
+      "loss": 0.5755,
+      "step": 10450
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.5312683895298154,
+      "learning_rate": 1.7646857608891438e-05,
+      "loss": 0.9262,
+      "step": 10451
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.509910043780723,
+      "learning_rate": 1.7646365880409554e-05,
+      "loss": 0.814,
+      "step": 10452
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6279190995215834,
+      "learning_rate": 1.7645874107408107e-05,
+      "loss": 0.5392,
+      "step": 10453
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5968930848299333,
+      "learning_rate": 1.764538228988996e-05,
+      "loss": 0.5621,
+      "step": 10454
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.370793044047612,
+      "learning_rate": 1.7644890427857975e-05,
+      "loss": 0.7842,
+      "step": 10455
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5410976590914283,
+      "learning_rate": 1.764439852131502e-05,
+      "loss": 0.9107,
+      "step": 10456
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6088989410746548,
+      "learning_rate": 1.7643906570263958e-05,
+      "loss": 0.7702,
+      "step": 10457
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5870985053604407,
+      "learning_rate": 1.764341457470765e-05,
+      "loss": 0.6249,
+      "step": 10458
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5857401693731,
+      "learning_rate": 1.7642922534648962e-05,
+      "loss": 0.6931,
+      "step": 10459
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.8521539628913413,
+      "learning_rate": 1.7642430450090758e-05,
+      "loss": 0.754,
+      "step": 10460
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.802480339896078,
+      "learning_rate": 1.7641938321035904e-05,
+      "loss": 0.7256,
+      "step": 10461
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.668215314917643,
+      "learning_rate": 1.7641446147487267e-05,
+      "loss": 0.8065,
+      "step": 10462
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.656763430556504,
+      "learning_rate": 1.7640953929447712e-05,
+      "loss": 0.7584,
+      "step": 10463
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.612510744512541,
+      "learning_rate": 1.7640461666920104e-05,
+      "loss": 0.8127,
+      "step": 10464
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.378286382105406,
+      "learning_rate": 1.763996935990731e-05,
+      "loss": 0.7135,
+      "step": 10465
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6560758148751398,
+      "learning_rate": 1.7639477008412192e-05,
+      "loss": 0.7044,
+      "step": 10466
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3497675057756464,
+      "learning_rate": 1.7638984612437623e-05,
+      "loss": 0.8121,
+      "step": 10467
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.675935706833883,
+      "learning_rate": 1.7638492171986468e-05,
+      "loss": 0.7611,
+      "step": 10468
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2331526886698163,
+      "learning_rate": 1.763799968706159e-05,
+      "loss": 0.8204,
+      "step": 10469
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4474424807015278,
+      "learning_rate": 1.7637507157665867e-05,
+      "loss": 0.7071,
+      "step": 10470
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6451985917796128,
+      "learning_rate": 1.763701458380215e-05,
+      "loss": 0.7069,
+      "step": 10471
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.346187591450042,
+      "learning_rate": 1.7636521965473324e-05,
+      "loss": 0.7187,
+      "step": 10472
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5335167090247122,
+      "learning_rate": 1.7636029302682244e-05,
+      "loss": 0.7963,
+      "step": 10473
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.58069572382561,
+      "learning_rate": 1.7635536595431784e-05,
+      "loss": 0.5849,
+      "step": 10474
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9583959184421447,
+      "learning_rate": 1.7635043843724817e-05,
+      "loss": 0.7259,
+      "step": 10475
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2681729465457754,
+      "learning_rate": 1.7634551047564205e-05,
+      "loss": 0.7536,
+      "step": 10476
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.438782068656764,
+      "learning_rate": 1.763405820695282e-05,
+      "loss": 0.7772,
+      "step": 10477
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0041321885443186,
+      "learning_rate": 1.7633565321893527e-05,
+      "loss": 0.6867,
+      "step": 10478
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5807229448775029,
+      "learning_rate": 1.7633072392389204e-05,
+      "loss": 0.8505,
+      "step": 10479
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6525469563706723,
+      "learning_rate": 1.763257941844272e-05,
+      "loss": 0.7817,
+      "step": 10480
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6037422286521648,
+      "learning_rate": 1.7632086400056936e-05,
+      "loss": 0.8707,
+      "step": 10481
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6386740853522201,
+      "learning_rate": 1.7631593337234733e-05,
+      "loss": 0.8052,
+      "step": 10482
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5055003881397135,
+      "learning_rate": 1.7631100229978977e-05,
+      "loss": 0.705,
+      "step": 10483
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.104072932808653,
+      "learning_rate": 1.763060707829254e-05,
+      "loss": 0.6484,
+      "step": 10484
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8494825251921447,
+      "learning_rate": 1.7630113882178287e-05,
+      "loss": 0.8077,
+      "step": 10485
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5709740059285213,
+      "learning_rate": 1.7629620641639102e-05,
+      "loss": 0.6956,
+      "step": 10486
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.604090366208842,
+      "learning_rate": 1.7629127356677847e-05,
+      "loss": 0.7783,
+      "step": 10487
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5560549773957328,
+      "learning_rate": 1.7628634027297397e-05,
+      "loss": 0.7649,
+      "step": 10488
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.635580353711189,
+      "learning_rate": 1.7628140653500628e-05,
+      "loss": 0.6868,
+      "step": 10489
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4479292245367712,
+      "learning_rate": 1.7627647235290407e-05,
+      "loss": 0.7244,
+      "step": 10490
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.697976209459185,
+      "learning_rate": 1.762715377266961e-05,
+      "loss": 0.6634,
+      "step": 10491
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4911020575906901,
+      "learning_rate": 1.7626660265641105e-05,
+      "loss": 0.7191,
+      "step": 10492
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5008787078152692,
+      "learning_rate": 1.7626166714207774e-05,
+      "loss": 0.7344,
+      "step": 10493
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.837300117285209,
+      "learning_rate": 1.762567311837248e-05,
+      "loss": 0.772,
+      "step": 10494
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5542525334131823,
+      "learning_rate": 1.762517947813811e-05,
+      "loss": 0.7364,
+      "step": 10495
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1106931252579475,
+      "learning_rate": 1.7624685793507526e-05,
+      "loss": 0.8239,
+      "step": 10496
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5273852187302972,
+      "learning_rate": 1.762419206448361e-05,
+      "loss": 0.7494,
+      "step": 10497
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.752412582957191,
+      "learning_rate": 1.7623698291069235e-05,
+      "loss": 0.881,
+      "step": 10498
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3906188433476883,
+      "learning_rate": 1.762320447326727e-05,
+      "loss": 0.8521,
+      "step": 10499
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7054948088441022,
+      "learning_rate": 1.7622710611080597e-05,
+      "loss": 0.6976,
+      "step": 10500
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3959767804397625,
+      "learning_rate": 1.762221670451209e-05,
+      "loss": 0.7704,
+      "step": 10501
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5571762350683146,
+      "learning_rate": 1.7621722753564628e-05,
+      "loss": 0.5289,
+      "step": 10502
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.877906642329242,
+      "learning_rate": 1.762122875824108e-05,
+      "loss": 0.7464,
+      "step": 10503
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4192153350857726,
+      "learning_rate": 1.7620734718544326e-05,
+      "loss": 0.6924,
+      "step": 10504
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4866408893089884,
+      "learning_rate": 1.762024063447724e-05,
+      "loss": 0.7746,
+      "step": 10505
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3667891272165889,
+      "learning_rate": 1.76197465060427e-05,
+      "loss": 0.8701,
+      "step": 10506
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.393435428665005,
+      "learning_rate": 1.7619252333243586e-05,
+      "loss": 0.8008,
+      "step": 10507
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.422669695510335,
+      "learning_rate": 1.7618758116082772e-05,
+      "loss": 0.7656,
+      "step": 10508
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4667520530905922,
+      "learning_rate": 1.7618263854563137e-05,
+      "loss": 0.8088,
+      "step": 10509
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6450037635255321,
+      "learning_rate": 1.7617769548687555e-05,
+      "loss": 0.6888,
+      "step": 10510
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3858602571501084,
+      "learning_rate": 1.761727519845891e-05,
+      "loss": 0.8336,
+      "step": 10511
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3731285415329133,
+      "learning_rate": 1.7616780803880077e-05,
+      "loss": 0.7614,
+      "step": 10512
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3108013791987077,
+      "learning_rate": 1.7616286364953935e-05,
+      "loss": 0.6921,
+      "step": 10513
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.634743074683673,
+      "learning_rate": 1.7615791881683358e-05,
+      "loss": 0.5539,
+      "step": 10514
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6742966909372812,
+      "learning_rate": 1.7615297354071233e-05,
+      "loss": 0.8078,
+      "step": 10515
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5330002882805827,
+      "learning_rate": 1.7614802782120433e-05,
+      "loss": 0.781,
+      "step": 10516
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8182548410697488,
+      "learning_rate": 1.7614308165833845e-05,
+      "loss": 0.6831,
+      "step": 10517
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.531147118166635,
+      "learning_rate": 1.7613813505214343e-05,
+      "loss": 0.7698,
+      "step": 10518
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7039324936847304,
+      "learning_rate": 1.7613318800264808e-05,
+      "loss": 0.7668,
+      "step": 10519
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.5697864941577433,
+      "learning_rate": 1.761282405098812e-05,
+      "loss": 0.9287,
+      "step": 10520
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5065475387475185,
+      "learning_rate": 1.7612329257387163e-05,
+      "loss": 0.7695,
+      "step": 10521
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5839772296003102,
+      "learning_rate": 1.761183441946481e-05,
+      "loss": 0.5474,
+      "step": 10522
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4895946578547545,
+      "learning_rate": 1.761133953722395e-05,
+      "loss": 0.6593,
+      "step": 10523
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2731830325626776,
+      "learning_rate": 1.7610844610667463e-05,
+      "loss": 0.6735,
+      "step": 10524
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.602705720089424,
+      "learning_rate": 1.7610349639798226e-05,
+      "loss": 0.6983,
+      "step": 10525
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5991025554228238,
+      "learning_rate": 1.7609854624619124e-05,
+      "loss": 0.5556,
+      "step": 10526
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4925005828954663,
+      "learning_rate": 1.7609359565133043e-05,
+      "loss": 0.6926,
+      "step": 10527
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7469806921274607,
+      "learning_rate": 1.7608864461342856e-05,
+      "loss": 0.8584,
+      "step": 10528
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4440199405442555,
+      "learning_rate": 1.7608369313251453e-05,
+      "loss": 0.7947,
+      "step": 10529
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8731670077429798,
+      "learning_rate": 1.7607874120861715e-05,
+      "loss": 0.6973,
+      "step": 10530
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.222738784729322,
+      "learning_rate": 1.7607378884176528e-05,
+      "loss": 0.8109,
+      "step": 10531
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7020674464882461,
+      "learning_rate": 1.760688360319877e-05,
+      "loss": 0.724,
+      "step": 10532
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5713828582467003,
+      "learning_rate": 1.760638827793133e-05,
+      "loss": 0.5402,
+      "step": 10533
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5386201960948036,
+      "learning_rate": 1.7605892908377088e-05,
+      "loss": 0.7219,
+      "step": 10534
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6550047769743994,
+      "learning_rate": 1.760539749453893e-05,
+      "loss": 0.8572,
+      "step": 10535
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7698009650509492,
+      "learning_rate": 1.7604902036419737e-05,
+      "loss": 0.7314,
+      "step": 10536
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5184439594888017,
+      "learning_rate": 1.76044065340224e-05,
+      "loss": 0.6919,
+      "step": 10537
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.727426905551681,
+      "learning_rate": 1.7603910987349803e-05,
+      "loss": 0.7138,
+      "step": 10538
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2834184728798383,
+      "learning_rate": 1.7603415396404822e-05,
+      "loss": 0.6955,
+      "step": 10539
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6392899025965937,
+      "learning_rate": 1.7602919761190355e-05,
+      "loss": 0.6767,
+      "step": 10540
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6660126841493015,
+      "learning_rate": 1.760242408170928e-05,
+      "loss": 0.8932,
+      "step": 10541
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7713770846061203,
+      "learning_rate": 1.760192835796448e-05,
+      "loss": 0.7099,
+      "step": 10542
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8633988978964129,
+      "learning_rate": 1.7601432589958856e-05,
+      "loss": 0.8756,
+      "step": 10543
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4493908311452963,
+      "learning_rate": 1.760093677769528e-05,
+      "loss": 0.7365,
+      "step": 10544
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5776919753817908,
+      "learning_rate": 1.7600440921176643e-05,
+      "loss": 0.5343,
+      "step": 10545
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8991556662815103,
+      "learning_rate": 1.7599945020405834e-05,
+      "loss": 0.8169,
+      "step": 10546
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5301374398937848,
+      "learning_rate": 1.7599449075385736e-05,
+      "loss": 0.8048,
+      "step": 10547
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5930980028572346,
+      "learning_rate": 1.7598953086119243e-05,
+      "loss": 0.6233,
+      "step": 10548
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6791408594590678,
+      "learning_rate": 1.759845705260924e-05,
+      "loss": 0.8871,
+      "step": 10549
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.4454550599072107,
+      "learning_rate": 1.759796097485861e-05,
+      "loss": 0.7822,
+      "step": 10550
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6412962330601806,
+      "learning_rate": 1.759746485287025e-05,
+      "loss": 0.7519,
+      "step": 10551
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.617772805319687,
+      "learning_rate": 1.7596968686647042e-05,
+      "loss": 0.8517,
+      "step": 10552
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8948307927969958,
+      "learning_rate": 1.759647247619188e-05,
+      "loss": 0.7248,
+      "step": 10553
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.521916677955005,
+      "learning_rate": 1.7595976221507647e-05,
+      "loss": 0.7618,
+      "step": 10554
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.673934311583809,
+      "learning_rate": 1.7595479922597238e-05,
+      "loss": 0.7302,
+      "step": 10555
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1434564483830814,
+      "learning_rate": 1.7594983579463537e-05,
+      "loss": 0.818,
+      "step": 10556
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6824084976246578,
+      "learning_rate": 1.759448719210944e-05,
+      "loss": 0.7294,
+      "step": 10557
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8084593905479722,
+      "learning_rate": 1.7593990760537836e-05,
+      "loss": 0.7247,
+      "step": 10558
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4445951846049943,
+      "learning_rate": 1.759349428475161e-05,
+      "loss": 0.7362,
+      "step": 10559
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7225212288433114,
+      "learning_rate": 1.7592997764753655e-05,
+      "loss": 0.7414,
+      "step": 10560
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.6414047301300423,
+      "learning_rate": 1.7592501200546865e-05,
+      "loss": 0.7225,
+      "step": 10561
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8147522789159818,
+      "learning_rate": 1.759200459213413e-05,
+      "loss": 0.8551,
+      "step": 10562
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.72293701666067,
+      "learning_rate": 1.7591507939518342e-05,
+      "loss": 0.7364,
+      "step": 10563
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7850429793869231,
+      "learning_rate": 1.7591011242702388e-05,
+      "loss": 0.9196,
+      "step": 10564
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4221129856569914,
+      "learning_rate": 1.7590514501689163e-05,
+      "loss": 0.6248,
+      "step": 10565
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.042695584805636,
+      "learning_rate": 1.759001771648156e-05,
+      "loss": 0.6855,
+      "step": 10566
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.661958108346784,
+      "learning_rate": 1.7589520887082474e-05,
+      "loss": 0.8068,
+      "step": 10567
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4266504341391277,
+      "learning_rate": 1.7589024013494796e-05,
+      "loss": 0.6891,
+      "step": 10568
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4285653167974068,
+      "learning_rate": 1.758852709572141e-05,
+      "loss": 0.8089,
+      "step": 10569
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4381869431762349,
+      "learning_rate": 1.7588030133765222e-05,
+      "loss": 0.7678,
+      "step": 10570
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.528255183906745,
+      "learning_rate": 1.7587533127629123e-05,
+      "loss": 0.7138,
+      "step": 10571
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4251143767566286,
+      "learning_rate": 1.7587036077315996e-05,
+      "loss": 0.8522,
+      "step": 10572
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6396467374177393,
+      "learning_rate": 1.758653898282875e-05,
+      "loss": 0.7662,
+      "step": 10573
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6911529197284176,
+      "learning_rate": 1.758604184417027e-05,
+      "loss": 0.9091,
+      "step": 10574
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4504922488468348,
+      "learning_rate": 1.7585544661343454e-05,
+      "loss": 0.8275,
+      "step": 10575
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6194724303590106,
+      "learning_rate": 1.758504743435119e-05,
+      "loss": 0.5165,
+      "step": 10576
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.33795767273715,
+      "learning_rate": 1.7584550163196382e-05,
+      "loss": 0.8367,
+      "step": 10577
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9436588337581036,
+      "learning_rate": 1.758405284788192e-05,
+      "loss": 0.6432,
+      "step": 10578
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6327397770854133,
+      "learning_rate": 1.7583555488410703e-05,
+      "loss": 0.5672,
+      "step": 10579
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3802435457455426,
+      "learning_rate": 1.7583058084785626e-05,
+      "loss": 0.6643,
+      "step": 10580
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.636705014557963,
+      "learning_rate": 1.7582560637009584e-05,
+      "loss": 0.5195,
+      "step": 10581
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4726893951256925,
+      "learning_rate": 1.7582063145085473e-05,
+      "loss": 0.8059,
+      "step": 10582
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3568206272965257,
+      "learning_rate": 1.7581565609016186e-05,
+      "loss": 0.6462,
+      "step": 10583
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4390928021541145,
+      "learning_rate": 1.758106802880463e-05,
+      "loss": 0.7307,
+      "step": 10584
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6032093356729289,
+      "learning_rate": 1.7580570404453693e-05,
+      "loss": 0.5253,
+      "step": 10585
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6023369630232664,
+      "learning_rate": 1.7580072735966275e-05,
+      "loss": 0.8156,
+      "step": 10586
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4055559179145227,
+      "learning_rate": 1.7579575023345273e-05,
+      "loss": 0.6959,
+      "step": 10587
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5747520900127054,
+      "learning_rate": 1.7579077266593587e-05,
+      "loss": 0.563,
+      "step": 10588
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7034433818247687,
+      "learning_rate": 1.757857946571411e-05,
+      "loss": 0.8334,
+      "step": 10589
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5345207506472802,
+      "learning_rate": 1.7578081620709747e-05,
+      "loss": 0.7652,
+      "step": 10590
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5464769506801692,
+      "learning_rate": 1.7577583731583396e-05,
+      "loss": 0.7806,
+      "step": 10591
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8730508941915796,
+      "learning_rate": 1.757708579833795e-05,
+      "loss": 0.7322,
+      "step": 10592
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.709483529125834,
+      "learning_rate": 1.7576587820976315e-05,
+      "loss": 0.7588,
+      "step": 10593
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3273555289196701,
+      "learning_rate": 1.7576089799501384e-05,
+      "loss": 0.7185,
+      "step": 10594
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1350682629008326,
+      "learning_rate": 1.757559173391606e-05,
+      "loss": 0.7564,
+      "step": 10595
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8687994604757756,
+      "learning_rate": 1.7575093624223243e-05,
+      "loss": 0.792,
+      "step": 10596
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.611164061554863,
+      "learning_rate": 1.757459547042583e-05,
+      "loss": 0.6629,
+      "step": 10597
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.632098282503504,
+      "learning_rate": 1.757409727252673e-05,
+      "loss": 0.6703,
+      "step": 10598
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6168680261597725,
+      "learning_rate": 1.7573599030528834e-05,
+      "loss": 0.8017,
+      "step": 10599
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7117496932115095,
+      "learning_rate": 1.7573100744435044e-05,
+      "loss": 0.6643,
+      "step": 10600
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.402933653654505,
+      "learning_rate": 1.757260241424827e-05,
+      "loss": 0.7853,
+      "step": 10601
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3589913969658256,
+      "learning_rate": 1.7572104039971405e-05,
+      "loss": 0.7418,
+      "step": 10602
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2885218424020723,
+      "learning_rate": 1.757160562160735e-05,
+      "loss": 0.6587,
+      "step": 10603
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.6728451178842896,
+      "learning_rate": 1.7571107159159013e-05,
+      "loss": 0.8334,
+      "step": 10604
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5079951777878127,
+      "learning_rate": 1.757060865262929e-05,
+      "loss": 0.7806,
+      "step": 10605
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0483457180315203,
+      "learning_rate": 1.757011010202109e-05,
+      "loss": 0.7497,
+      "step": 10606
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6020676066374879,
+      "learning_rate": 1.756961150733731e-05,
+      "loss": 0.5722,
+      "step": 10607
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3016228402529635,
+      "learning_rate": 1.756911286858086e-05,
+      "loss": 0.7714,
+      "step": 10608
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6229228039966257,
+      "learning_rate": 1.7568614185754635e-05,
+      "loss": 0.6913,
+      "step": 10609
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.429996133686004,
+      "learning_rate": 1.7568115458861542e-05,
+      "loss": 0.8114,
+      "step": 10610
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.314077395955266,
+      "learning_rate": 1.7567616687904484e-05,
+      "loss": 0.7413,
+      "step": 10611
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4675790331681815,
+      "learning_rate": 1.7567117872886368e-05,
+      "loss": 0.6676,
+      "step": 10612
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.596019273669374,
+      "learning_rate": 1.7566619013810094e-05,
+      "loss": 0.7923,
+      "step": 10613
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3243804314320025,
+      "learning_rate": 1.7566120110678572e-05,
+      "loss": 0.759,
+      "step": 10614
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6473529061939299,
+      "learning_rate": 1.7565621163494697e-05,
+      "loss": 0.579,
+      "step": 10615
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.559603488934035,
+      "learning_rate": 1.7565122172261384e-05,
+      "loss": 0.7378,
+      "step": 10616
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4452193643893272,
+      "learning_rate": 1.7564623136981532e-05,
+      "loss": 0.6838,
+      "step": 10617
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0326865880845537,
+      "learning_rate": 1.7564124057658057e-05,
+      "loss": 0.5853,
+      "step": 10618
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.559957931353598,
+      "learning_rate": 1.756362493429385e-05,
+      "loss": 0.6512,
+      "step": 10619
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6373314792319587,
+      "learning_rate": 1.7563125766891826e-05,
+      "loss": 0.7471,
+      "step": 10620
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.152297204539939,
+      "learning_rate": 1.7562626555454887e-05,
+      "loss": 0.7666,
+      "step": 10621
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.521011784041685,
+      "learning_rate": 1.7562127299985944e-05,
+      "loss": 0.7638,
+      "step": 10622
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9786823266597662,
+      "learning_rate": 1.7561628000487902e-05,
+      "loss": 0.833,
+      "step": 10623
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7051226230333105,
+      "learning_rate": 1.7561128656963667e-05,
+      "loss": 0.8434,
+      "step": 10624
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7450940286904042,
+      "learning_rate": 1.756062926941615e-05,
+      "loss": 0.7547,
+      "step": 10625
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.996696435808158,
+      "learning_rate": 1.7560129837848255e-05,
+      "loss": 0.7738,
+      "step": 10626
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.7556546940571782,
+      "learning_rate": 1.755963036226289e-05,
+      "loss": 0.7334,
+      "step": 10627
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.3090267611764865,
+      "learning_rate": 1.755913084266296e-05,
+      "loss": 0.855,
+      "step": 10628
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3348247195916982,
+      "learning_rate": 1.755863127905138e-05,
+      "loss": 0.7294,
+      "step": 10629
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6333107313676489,
+      "learning_rate": 1.7558131671431056e-05,
+      "loss": 0.5607,
+      "step": 10630
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0380165731064435,
+      "learning_rate": 1.7557632019804894e-05,
+      "loss": 0.708,
+      "step": 10631
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5161661363647225,
+      "learning_rate": 1.755713232417581e-05,
+      "loss": 0.7472,
+      "step": 10632
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.937306866548201,
+      "learning_rate": 1.7556632584546705e-05,
+      "loss": 0.747,
+      "step": 10633
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6307245173370679,
+      "learning_rate": 1.7556132800920494e-05,
+      "loss": 0.5568,
+      "step": 10634
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5798429335104348,
+      "learning_rate": 1.7555632973300087e-05,
+      "loss": 0.5181,
+      "step": 10635
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4214962144793437,
+      "learning_rate": 1.755513310168839e-05,
+      "loss": 0.7964,
+      "step": 10636
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4594246254426486,
+      "learning_rate": 1.755463318608832e-05,
+      "loss": 0.6294,
+      "step": 10637
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4680896665125063,
+      "learning_rate": 1.7554133226502782e-05,
+      "loss": 0.7063,
+      "step": 10638
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4780580562354113,
+      "learning_rate": 1.755363322293469e-05,
+      "loss": 0.8236,
+      "step": 10639
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5996249201105714,
+      "learning_rate": 1.755313317538695e-05,
+      "loss": 0.7077,
+      "step": 10640
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4815096587957892,
+      "learning_rate": 1.755263308386248e-05,
+      "loss": 0.7343,
+      "step": 10641
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6239153757077911,
+      "learning_rate": 1.755213294836419e-05,
+      "loss": 0.57,
+      "step": 10642
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.703886838627051,
+      "learning_rate": 1.755163276889499e-05,
+      "loss": 0.7022,
+      "step": 10643
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6158826570223997,
+      "learning_rate": 1.7551132545457793e-05,
+      "loss": 0.4824,
+      "step": 10644
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.424188889328519,
+      "learning_rate": 1.7550632278055514e-05,
+      "loss": 0.7002,
+      "step": 10645
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4263441045057323,
+      "learning_rate": 1.7550131966691057e-05,
+      "loss": 0.8285,
+      "step": 10646
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.49539092965897,
+      "learning_rate": 1.7549631611367345e-05,
+      "loss": 0.7693,
+      "step": 10647
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8663071671000717,
+      "learning_rate": 1.7549131212087288e-05,
+      "loss": 0.6497,
+      "step": 10648
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5380603249722054,
+      "learning_rate": 1.75486307688538e-05,
+      "loss": 0.9178,
+      "step": 10649
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7995796723359296,
+      "learning_rate": 1.7548130281669793e-05,
+      "loss": 0.7581,
+      "step": 10650
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9347901101792209,
+      "learning_rate": 1.7547629750538182e-05,
+      "loss": 0.8419,
+      "step": 10651
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.384280095852449,
+      "learning_rate": 1.7547129175461882e-05,
+      "loss": 0.7605,
+      "step": 10652
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4334627810615135,
+      "learning_rate": 1.7546628556443806e-05,
+      "loss": 0.7144,
+      "step": 10653
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.334028003228432,
+      "learning_rate": 1.754612789348687e-05,
+      "loss": 0.623,
+      "step": 10654
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6277180433160986,
+      "learning_rate": 1.7545627186593986e-05,
+      "loss": 0.5806,
+      "step": 10655
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2900668589687565,
+      "learning_rate": 1.754512643576807e-05,
+      "loss": 0.7183,
+      "step": 10656
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6660178515080066,
+      "learning_rate": 1.7544625641012044e-05,
+      "loss": 0.7942,
+      "step": 10657
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2898487438139565,
+      "learning_rate": 1.7544124802328814e-05,
+      "loss": 0.7743,
+      "step": 10658
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.312548640679787,
+      "learning_rate": 1.7543623919721305e-05,
+      "loss": 0.733,
+      "step": 10659
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8096454583239034,
+      "learning_rate": 1.754312299319243e-05,
+      "loss": 0.7656,
+      "step": 10660
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4156339695035278,
+      "learning_rate": 1.7542622022745102e-05,
+      "loss": 0.7452,
+      "step": 10661
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4242816524765882,
+      "learning_rate": 1.754212100838224e-05,
+      "loss": 0.8535,
+      "step": 10662
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4439987845044322,
+      "learning_rate": 1.7541619950106764e-05,
+      "loss": 0.8603,
+      "step": 10663
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7652950832233851,
+      "learning_rate": 1.7541118847921586e-05,
+      "loss": 0.8228,
+      "step": 10664
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.451820659858435,
+      "learning_rate": 1.754061770182963e-05,
+      "loss": 0.6922,
+      "step": 10665
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4301004274656886,
+      "learning_rate": 1.754011651183381e-05,
+      "loss": 0.6281,
+      "step": 10666
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4454914708338713,
+      "learning_rate": 1.7539615277937045e-05,
+      "loss": 0.8048,
+      "step": 10667
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3538821252707003,
+      "learning_rate": 1.753911400014225e-05,
+      "loss": 0.7651,
+      "step": 10668
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6282096953902323,
+      "learning_rate": 1.753861267845235e-05,
+      "loss": 0.7263,
+      "step": 10669
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.413071899413947,
+      "learning_rate": 1.7538111312870255e-05,
+      "loss": 0.7876,
+      "step": 10670
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9679110727749995,
+      "learning_rate": 1.7537609903398895e-05,
+      "loss": 0.8621,
+      "step": 10671
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.037635261777443,
+      "learning_rate": 1.7537108450041182e-05,
+      "loss": 0.6838,
+      "step": 10672
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5742517138815886,
+      "learning_rate": 1.7536606952800034e-05,
+      "loss": 0.5116,
+      "step": 10673
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3955257400667698,
+      "learning_rate": 1.753610541167838e-05,
+      "loss": 0.7816,
+      "step": 10674
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.851680364645797,
+      "learning_rate": 1.7535603826679132e-05,
+      "loss": 0.724,
+      "step": 10675
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4877544878986193,
+      "learning_rate": 1.7535102197805212e-05,
+      "loss": 0.5969,
+      "step": 10676
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3439263571957798,
+      "learning_rate": 1.753460052505954e-05,
+      "loss": 0.6936,
+      "step": 10677
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5632831140168661,
+      "learning_rate": 1.753409880844504e-05,
+      "loss": 0.8245,
+      "step": 10678
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6230955966905132,
+      "learning_rate": 1.7533597047964633e-05,
+      "loss": 0.739,
+      "step": 10679
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4306771683621553,
+      "learning_rate": 1.7533095243621237e-05,
+      "loss": 0.8065,
+      "step": 10680
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.555660914600789,
+      "learning_rate": 1.7532593395417775e-05,
+      "loss": 0.9169,
+      "step": 10681
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7029771193776342,
+      "learning_rate": 1.7532091503357172e-05,
+      "loss": 0.7617,
+      "step": 10682
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6608054142429358,
+      "learning_rate": 1.7531589567442347e-05,
+      "loss": 0.7749,
+      "step": 10683
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9222201270158839,
+      "learning_rate": 1.7531087587676223e-05,
+      "loss": 0.7278,
+      "step": 10684
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3522341017312904,
+      "learning_rate": 1.753058556406172e-05,
+      "loss": 0.6793,
+      "step": 10685
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6073666818352186,
+      "learning_rate": 1.7530083496601763e-05,
+      "loss": 0.8016,
+      "step": 10686
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6929816503003992,
+      "learning_rate": 1.7529581385299283e-05,
+      "loss": 0.8625,
+      "step": 10687
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.828445437244411,
+      "learning_rate": 1.7529079230157192e-05,
+      "loss": 0.7989,
+      "step": 10688
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6316778446088736,
+      "learning_rate": 1.7528577031178416e-05,
+      "loss": 0.6337,
+      "step": 10689
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.603842100600178,
+      "learning_rate": 1.7528074788365885e-05,
+      "loss": 0.9548,
+      "step": 10690
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4216885717341101,
+      "learning_rate": 1.7527572501722516e-05,
+      "loss": 0.8683,
+      "step": 10691
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9908994477570499,
+      "learning_rate": 1.7527070171251236e-05,
+      "loss": 0.682,
+      "step": 10692
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.252571779605476,
+      "learning_rate": 1.7526567796954973e-05,
+      "loss": 0.7135,
+      "step": 10693
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5133785955619592,
+      "learning_rate": 1.752606537883665e-05,
+      "loss": 0.7129,
+      "step": 10694
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3375002697040272,
+      "learning_rate": 1.752556291689919e-05,
+      "loss": 0.8262,
+      "step": 10695
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2750507547178604,
+      "learning_rate": 1.752506041114552e-05,
+      "loss": 0.732,
+      "step": 10696
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5392948449293196,
+      "learning_rate": 1.7524557861578567e-05,
+      "loss": 0.7988,
+      "step": 10697
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5217131658071679,
+      "learning_rate": 1.7524055268201254e-05,
+      "loss": 0.7722,
+      "step": 10698
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5396993608981877,
+      "learning_rate": 1.752355263101651e-05,
+      "loss": 0.7346,
+      "step": 10699
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4740125170182874,
+      "learning_rate": 1.7523049950027263e-05,
+      "loss": 0.8364,
+      "step": 10700
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4544045373126036,
+      "learning_rate": 1.7522547225236434e-05,
+      "loss": 0.8141,
+      "step": 10701
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5224863123602819,
+      "learning_rate": 1.7522044456646957e-05,
+      "loss": 0.8965,
+      "step": 10702
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4168127550365839,
+      "learning_rate": 1.7521541644261754e-05,
+      "loss": 0.8474,
+      "step": 10703
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4348709029884197,
+      "learning_rate": 1.7521038788083755e-05,
+      "loss": 0.7584,
+      "step": 10704
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5466000613248818,
+      "learning_rate": 1.7520535888115887e-05,
+      "loss": 0.803,
+      "step": 10705
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4360074061183754,
+      "learning_rate": 1.7520032944361074e-05,
+      "loss": 0.7484,
+      "step": 10706
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.60564169086779,
+      "learning_rate": 1.751952995682225e-05,
+      "loss": 0.701,
+      "step": 10707
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1223116385126817,
+      "learning_rate": 1.7519026925502344e-05,
+      "loss": 0.7753,
+      "step": 10708
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6723330934438205,
+      "learning_rate": 1.7518523850404285e-05,
+      "loss": 0.8136,
+      "step": 10709
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4386713348447409,
+      "learning_rate": 1.7518020731530998e-05,
+      "loss": 0.8914,
+      "step": 10710
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5113052029129583,
+      "learning_rate": 1.7517517568885413e-05,
+      "loss": 0.7011,
+      "step": 10711
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6952826608831348,
+      "learning_rate": 1.751701436247046e-05,
+      "loss": 0.7802,
+      "step": 10712
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3760365178084308,
+      "learning_rate": 1.7516511112289068e-05,
+      "loss": 0.8281,
+      "step": 10713
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3430658140809957,
+      "learning_rate": 1.751600781834417e-05,
+      "loss": 0.715,
+      "step": 10714
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4164618964850393,
+      "learning_rate": 1.75155044806387e-05,
+      "loss": 0.8325,
+      "step": 10715
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5819821776574313,
+      "learning_rate": 1.7515001099175578e-05,
+      "loss": 0.5422,
+      "step": 10716
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3960078683175123,
+      "learning_rate": 1.751449767395774e-05,
+      "loss": 0.705,
+      "step": 10717
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6854824602537655,
+      "learning_rate": 1.751399420498812e-05,
+      "loss": 0.9279,
+      "step": 10718
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3513615173376157,
+      "learning_rate": 1.751349069226965e-05,
+      "loss": 0.7325,
+      "step": 10719
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7887806796989962,
+      "learning_rate": 1.7512987135805254e-05,
+      "loss": 0.7425,
+      "step": 10720
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4769242599777146,
+      "learning_rate": 1.7512483535597868e-05,
+      "loss": 0.6747,
+      "step": 10721
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4747155579371984,
+      "learning_rate": 1.751197989165043e-05,
+      "loss": 1.0306,
+      "step": 10722
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5250493178849116,
+      "learning_rate": 1.7511476203965857e-05,
+      "loss": 0.7217,
+      "step": 10723
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6001603165931217,
+      "learning_rate": 1.75109724725471e-05,
+      "loss": 0.9336,
+      "step": 10724
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.405330904863946,
+      "learning_rate": 1.751046869739708e-05,
+      "loss": 0.8423,
+      "step": 10725
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5829947833343385,
+      "learning_rate": 1.7509964878518736e-05,
+      "loss": 0.7243,
+      "step": 10726
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4232571917095098,
+      "learning_rate": 1.7509461015914992e-05,
+      "loss": 0.7207,
+      "step": 10727
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5409244601248622,
+      "learning_rate": 1.7508957109588794e-05,
+      "loss": 0.8008,
+      "step": 10728
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5762332462225497,
+      "learning_rate": 1.7508453159543072e-05,
+      "loss": 0.7402,
+      "step": 10729
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9276370355003922,
+      "learning_rate": 1.7507949165780753e-05,
+      "loss": 0.854,
+      "step": 10730
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4148734130389775,
+      "learning_rate": 1.750744512830478e-05,
+      "loss": 0.7648,
+      "step": 10731
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6695181325172825,
+      "learning_rate": 1.7506941047118084e-05,
+      "loss": 0.6852,
+      "step": 10732
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5908423847713856,
+      "learning_rate": 1.7506436922223602e-05,
+      "loss": 0.5574,
+      "step": 10733
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6018801706465948,
+      "learning_rate": 1.750593275362427e-05,
+      "loss": 0.6716,
+      "step": 10734
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3380401556154313,
+      "learning_rate": 1.750542854132302e-05,
+      "loss": 0.7201,
+      "step": 10735
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6519404212193642,
+      "learning_rate": 1.7504924285322782e-05,
+      "loss": 0.5729,
+      "step": 10736
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7835805578523658,
+      "learning_rate": 1.7504419985626504e-05,
+      "loss": 0.6509,
+      "step": 10737
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.458855855691075,
+      "learning_rate": 1.7503915642237116e-05,
+      "loss": 0.8432,
+      "step": 10738
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6457684649779377,
+      "learning_rate": 1.7503411255157558e-05,
+      "loss": 0.5713,
+      "step": 10739
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7395039629088054,
+      "learning_rate": 1.750290682439076e-05,
+      "loss": 0.8404,
+      "step": 10740
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5828627455399915,
+      "learning_rate": 1.7502402349939667e-05,
+      "loss": 0.7228,
+      "step": 10741
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.4336518811756696,
+      "learning_rate": 1.750189783180721e-05,
+      "loss": 0.6895,
+      "step": 10742
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7772936032089848,
+      "learning_rate": 1.7501393269996332e-05,
+      "loss": 0.6249,
+      "step": 10743
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4355312201398907,
+      "learning_rate": 1.7500888664509962e-05,
+      "loss": 0.7847,
+      "step": 10744
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5874333997945864,
+      "learning_rate": 1.7500384015351048e-05,
+      "loss": 0.7462,
+      "step": 10745
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5908226116130437,
+      "learning_rate": 1.7499879322522524e-05,
+      "loss": 0.7722,
+      "step": 10746
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5932887464369362,
+      "learning_rate": 1.7499374586027323e-05,
+      "loss": 0.534,
+      "step": 10747
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.50696680940118,
+      "learning_rate": 1.7498869805868394e-05,
+      "loss": 0.7192,
+      "step": 10748
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.597724729141295,
+      "learning_rate": 1.749836498204867e-05,
+      "loss": 0.7565,
+      "step": 10749
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6258054315573782,
+      "learning_rate": 1.749786011457109e-05,
+      "loss": 0.7275,
+      "step": 10750
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4864477777418346,
+      "learning_rate": 1.7497355203438594e-05,
+      "loss": 0.7694,
+      "step": 10751
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.813413891000816,
+      "learning_rate": 1.7496850248654123e-05,
+      "loss": 0.7646,
+      "step": 10752
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.212431647086428,
+      "learning_rate": 1.7496345250220617e-05,
+      "loss": 0.7908,
+      "step": 10753
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.770103842240031,
+      "learning_rate": 1.7495840208141015e-05,
+      "loss": 0.6865,
+      "step": 10754
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.668318026119609,
+      "learning_rate": 1.749533512241826e-05,
+      "loss": 0.8151,
+      "step": 10755
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.547895057316066,
+      "learning_rate": 1.7494829993055288e-05,
+      "loss": 0.8166,
+      "step": 10756
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5947180787800777,
+      "learning_rate": 1.7494324820055045e-05,
+      "loss": 0.7232,
+      "step": 10757
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3189577921747133,
+      "learning_rate": 1.7493819603420468e-05,
+      "loss": 0.8864,
+      "step": 10758
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1702413680724404,
+      "learning_rate": 1.7493314343154503e-05,
+      "loss": 0.7906,
+      "step": 10759
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3943516626958614,
+      "learning_rate": 1.7492809039260086e-05,
+      "loss": 0.7026,
+      "step": 10760
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.576141294187655,
+      "learning_rate": 1.7492303691740166e-05,
+      "loss": 0.7066,
+      "step": 10761
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.575752688480855,
+      "learning_rate": 1.7491798300597684e-05,
+      "loss": 0.7602,
+      "step": 10762
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.009191405253442,
+      "learning_rate": 1.7491292865835576e-05,
+      "loss": 0.7675,
+      "step": 10763
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5511740007116668,
+      "learning_rate": 1.749078738745679e-05,
+      "loss": 0.7282,
+      "step": 10764
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5802323622186797,
+      "learning_rate": 1.7490281865464268e-05,
+      "loss": 0.5224,
+      "step": 10765
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3841322442483932,
+      "learning_rate": 1.748977629986096e-05,
+      "loss": 0.6944,
+      "step": 10766
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4935746419499016,
+      "learning_rate": 1.7489270690649794e-05,
+      "loss": 0.7267,
+      "step": 10767
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3637742013395473,
+      "learning_rate": 1.748876503783373e-05,
+      "loss": 0.9508,
+      "step": 10768
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.354580595868509,
+      "learning_rate": 1.7488259341415702e-05,
+      "loss": 0.7802,
+      "step": 10769
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.970626476027022,
+      "learning_rate": 1.7487753601398655e-05,
+      "loss": 0.7116,
+      "step": 10770
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4414783133134863,
+      "learning_rate": 1.748724781778554e-05,
+      "loss": 0.8235,
+      "step": 10771
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7351458355572515,
+      "learning_rate": 1.7486741990579298e-05,
+      "loss": 0.9652,
+      "step": 10772
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5934785511889087,
+      "learning_rate": 1.748623611978287e-05,
+      "loss": 0.5134,
+      "step": 10773
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5556945762149368,
+      "learning_rate": 1.7485730205399208e-05,
+      "loss": 0.7579,
+      "step": 10774
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6445913213414511,
+      "learning_rate": 1.7485224247431255e-05,
+      "loss": 0.6309,
+      "step": 10775
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0546071875881275,
+      "learning_rate": 1.7484718245881957e-05,
+      "loss": 0.7537,
+      "step": 10776
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.4250124771321446,
+      "learning_rate": 1.7484212200754256e-05,
+      "loss": 0.7247,
+      "step": 10777
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5575845569689359,
+      "learning_rate": 1.7483706112051105e-05,
+      "loss": 0.5419,
+      "step": 10778
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2638908156191333,
+      "learning_rate": 1.7483199979775447e-05,
+      "loss": 0.8138,
+      "step": 10779
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3766338033505372,
+      "learning_rate": 1.7482693803930228e-05,
+      "loss": 0.8654,
+      "step": 10780
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5916696056442912,
+      "learning_rate": 1.7482187584518397e-05,
+      "loss": 0.5414,
+      "step": 10781
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.658024785061115,
+      "learning_rate": 1.7481681321542906e-05,
+      "loss": 0.6697,
+      "step": 10782
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2556181126281925,
+      "learning_rate": 1.7481175015006693e-05,
+      "loss": 0.7196,
+      "step": 10783
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6186575820997753,
+      "learning_rate": 1.748066866491271e-05,
+      "loss": 0.6807,
+      "step": 10784
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6528704950443753,
+      "learning_rate": 1.748016227126391e-05,
+      "loss": 0.7865,
+      "step": 10785
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.056376359425968,
+      "learning_rate": 1.7479655834063235e-05,
+      "loss": 0.6674,
+      "step": 10786
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.645680597735787,
+      "learning_rate": 1.7479149353313635e-05,
+      "loss": 0.7774,
+      "step": 10787
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5288358824774426,
+      "learning_rate": 1.747864282901806e-05,
+      "loss": 0.6289,
+      "step": 10788
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.392521650562568,
+      "learning_rate": 1.7478136261179458e-05,
+      "loss": 0.7362,
+      "step": 10789
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.557600760391962,
+      "learning_rate": 1.747762964980078e-05,
+      "loss": 0.6079,
+      "step": 10790
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.215690690856242,
+      "learning_rate": 1.7477122994884973e-05,
+      "loss": 0.5372,
+      "step": 10791
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4301124913976604,
+      "learning_rate": 1.747661629643499e-05,
+      "loss": 0.664,
+      "step": 10792
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5621808821049867,
+      "learning_rate": 1.7476109554453783e-05,
+      "loss": 0.6776,
+      "step": 10793
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3326731347600351,
+      "learning_rate": 1.7475602768944297e-05,
+      "loss": 0.6888,
+      "step": 10794
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.350474647778329,
+      "learning_rate": 1.747509593990948e-05,
+      "loss": 0.7659,
+      "step": 10795
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.212453079888006,
+      "learning_rate": 1.7474589067352296e-05,
+      "loss": 0.6219,
+      "step": 10796
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6316168611089847,
+      "learning_rate": 1.7474082151275683e-05,
+      "loss": 0.7491,
+      "step": 10797
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.50164701806517,
+      "learning_rate": 1.74735751916826e-05,
+      "loss": 0.8465,
+      "step": 10798
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7361735544593864,
+      "learning_rate": 1.747306818857599e-05,
+      "loss": 0.701,
+      "step": 10799
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6548912528450599,
+      "learning_rate": 1.7472561141958818e-05,
+      "loss": 0.5903,
+      "step": 10800
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6847676391455333,
+      "learning_rate": 1.7472054051834028e-05,
+      "loss": 0.7778,
+      "step": 10801
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4853600901230215,
+      "learning_rate": 1.7471546918204575e-05,
+      "loss": 0.7925,
+      "step": 10802
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7014420875351794,
+      "learning_rate": 1.7471039741073407e-05,
+      "loss": 0.7478,
+      "step": 10803
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.565522913964532,
+      "learning_rate": 1.7470532520443483e-05,
+      "loss": 0.8588,
+      "step": 10804
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9533881897847354,
+      "learning_rate": 1.7470025256317752e-05,
+      "loss": 0.7121,
+      "step": 10805
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3520571950088405,
+      "learning_rate": 1.7469517948699167e-05,
+      "loss": 0.8379,
+      "step": 10806
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5461335250066004,
+      "learning_rate": 1.7469010597590688e-05,
+      "loss": 0.7368,
+      "step": 10807
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.447890117563504,
+      "learning_rate": 1.7468503202995263e-05,
+      "loss": 0.7631,
+      "step": 10808
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6176549536695394,
+      "learning_rate": 1.7467995764915847e-05,
+      "loss": 0.5451,
+      "step": 10809
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6472151742974257,
+      "learning_rate": 1.7467488283355396e-05,
+      "loss": 0.7193,
+      "step": 10810
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7478195849323181,
+      "learning_rate": 1.7466980758316864e-05,
+      "loss": 0.8689,
+      "step": 10811
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.371265928580865,
+      "learning_rate": 1.7466473189803203e-05,
+      "loss": 0.7363,
+      "step": 10812
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3687355011010212,
+      "learning_rate": 1.7465965577817375e-05,
+      "loss": 0.7342,
+      "step": 10813
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6059185105359396,
+      "learning_rate": 1.746545792236233e-05,
+      "loss": 0.5603,
+      "step": 10814
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.562229331003008,
+      "learning_rate": 1.7464950223441026e-05,
+      "loss": 0.7144,
+      "step": 10815
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5777119261740817,
+      "learning_rate": 1.7464442481056418e-05,
+      "loss": 0.4829,
+      "step": 10816
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2852864737650145,
+      "learning_rate": 1.7463934695211463e-05,
+      "loss": 0.7087,
+      "step": 10817
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.57637580283897,
+      "learning_rate": 1.7463426865909118e-05,
+      "loss": 0.767,
+      "step": 10818
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.709427018010233,
+      "learning_rate": 1.746291899315234e-05,
+      "loss": 0.7764,
+      "step": 10819
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.023744992532309,
+      "learning_rate": 1.7462411076944084e-05,
+      "loss": 0.678,
+      "step": 10820
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5725635762537085,
+      "learning_rate": 1.7461903117287305e-05,
+      "loss": 0.5447,
+      "step": 10821
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2852001408472746,
+      "learning_rate": 1.746139511418497e-05,
+      "loss": 0.7077,
+      "step": 10822
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4694884210766874,
+      "learning_rate": 1.7460887067640025e-05,
+      "loss": 0.7464,
+      "step": 10823
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4213046633293513,
+      "learning_rate": 1.7460378977655432e-05,
+      "loss": 0.7377,
+      "step": 10824
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4082622910393219,
+      "learning_rate": 1.7459870844234154e-05,
+      "loss": 0.6971,
+      "step": 10825
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6219668272282495,
+      "learning_rate": 1.745936266737915e-05,
+      "loss": 0.853,
+      "step": 10826
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0522709408161153,
+      "learning_rate": 1.745885444709337e-05,
+      "loss": 0.7503,
+      "step": 10827
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3292839970598596,
+      "learning_rate": 1.7458346183379778e-05,
+      "loss": 0.7134,
+      "step": 10828
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3934327973329326,
+      "learning_rate": 1.7457837876241332e-05,
+      "loss": 0.7183,
+      "step": 10829
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5404612729553915,
+      "learning_rate": 1.7457329525680996e-05,
+      "loss": 0.8137,
+      "step": 10830
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6337163067398437,
+      "learning_rate": 1.7456821131701724e-05,
+      "loss": 0.7431,
+      "step": 10831
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4639679790233215,
+      "learning_rate": 1.745631269430648e-05,
+      "loss": 0.7773,
+      "step": 10832
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7326586279255372,
+      "learning_rate": 1.7455804213498223e-05,
+      "loss": 0.8827,
+      "step": 10833
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3405743575630944,
+      "learning_rate": 1.7455295689279913e-05,
+      "loss": 0.7194,
+      "step": 10834
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3473359236292017,
+      "learning_rate": 1.745478712165451e-05,
+      "loss": 0.8532,
+      "step": 10835
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0603749841931287,
+      "learning_rate": 1.745427851062498e-05,
+      "loss": 0.7355,
+      "step": 10836
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4229721572426872,
+      "learning_rate": 1.7453769856194277e-05,
+      "loss": 0.7238,
+      "step": 10837
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4329155509139448,
+      "learning_rate": 1.7453261158365367e-05,
+      "loss": 0.6024,
+      "step": 10838
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4340180323128313,
+      "learning_rate": 1.745275241714121e-05,
+      "loss": 0.8164,
+      "step": 10839
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.447277616322959,
+      "learning_rate": 1.745224363252477e-05,
+      "loss": 0.7212,
+      "step": 10840
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.468491143127518,
+      "learning_rate": 1.745173480451901e-05,
+      "loss": 0.754,
+      "step": 10841
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4966459849672837,
+      "learning_rate": 1.745122593312689e-05,
+      "loss": 0.7391,
+      "step": 10842
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4137515440051411,
+      "learning_rate": 1.745071701835137e-05,
+      "loss": 0.7366,
+      "step": 10843
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.283670470474925,
+      "learning_rate": 1.7450208060195423e-05,
+      "loss": 0.7306,
+      "step": 10844
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4695360615806372,
+      "learning_rate": 1.7449699058662e-05,
+      "loss": 0.821,
+      "step": 10845
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.55487576949897,
+      "learning_rate": 1.7449190013754075e-05,
+      "loss": 0.8517,
+      "step": 10846
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.441145907028962,
+      "learning_rate": 1.7448680925474605e-05,
+      "loss": 0.7251,
+      "step": 10847
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3479094975167651,
+      "learning_rate": 1.7448171793826553e-05,
+      "loss": 0.7046,
+      "step": 10848
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0884027749657097,
+      "learning_rate": 1.744766261881289e-05,
+      "loss": 0.7132,
+      "step": 10849
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.600442424294212,
+      "learning_rate": 1.7447153400436577e-05,
+      "loss": 0.6755,
+      "step": 10850
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6482080274203839,
+      "learning_rate": 1.744664413870058e-05,
+      "loss": 0.6905,
+      "step": 10851
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8040508242063678,
+      "learning_rate": 1.7446134833607863e-05,
+      "loss": 0.7918,
+      "step": 10852
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8247928669614508,
+      "learning_rate": 1.7445625485161393e-05,
+      "loss": 0.6492,
+      "step": 10853
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5876458431047846,
+      "learning_rate": 1.7445116093364133e-05,
+      "loss": 0.8337,
+      "step": 10854
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5715897710150039,
+      "learning_rate": 1.744460665821905e-05,
+      "loss": 0.563,
+      "step": 10855
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9015660347437713,
+      "learning_rate": 1.7444097179729112e-05,
+      "loss": 0.7436,
+      "step": 10856
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6406518912222519,
+      "learning_rate": 1.7443587657897278e-05,
+      "loss": 0.7012,
+      "step": 10857
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5997330257009441,
+      "learning_rate": 1.7443078092726525e-05,
+      "loss": 0.574,
+      "step": 10858
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4088171383312107,
+      "learning_rate": 1.7442568484219812e-05,
+      "loss": 0.834,
+      "step": 10859
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5523329684354836,
+      "learning_rate": 1.7442058832380113e-05,
+      "loss": 0.7828,
+      "step": 10860
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1880587620419423,
+      "learning_rate": 1.7441549137210386e-05,
+      "loss": 0.7886,
+      "step": 10861
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.75120293702833,
+      "learning_rate": 1.744103939871361e-05,
+      "loss": 0.6855,
+      "step": 10862
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.425819328837387,
+      "learning_rate": 1.744052961689274e-05,
+      "loss": 0.849,
+      "step": 10863
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.58551252140098,
+      "learning_rate": 1.7440019791750757e-05,
+      "loss": 0.493,
+      "step": 10864
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4680998860156786,
+      "learning_rate": 1.743950992329062e-05,
+      "loss": 0.8638,
+      "step": 10865
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3851503244111476,
+      "learning_rate": 1.74390000115153e-05,
+      "loss": 0.7563,
+      "step": 10866
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6404865612806843,
+      "learning_rate": 1.743849005642777e-05,
+      "loss": 0.693,
+      "step": 10867
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3237441730972732,
+      "learning_rate": 1.7437980058030993e-05,
+      "loss": 0.8507,
+      "step": 10868
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.540373720099756,
+      "learning_rate": 1.7437470016327944e-05,
+      "loss": 0.6732,
+      "step": 10869
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4277048146122935,
+      "learning_rate": 1.743695993132159e-05,
+      "loss": 0.8255,
+      "step": 10870
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4653640334506346,
+      "learning_rate": 1.74364498030149e-05,
+      "loss": 0.7833,
+      "step": 10871
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2459536624324477,
+      "learning_rate": 1.7435939631410845e-05,
+      "loss": 0.7439,
+      "step": 10872
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9097122393848243,
+      "learning_rate": 1.7435429416512394e-05,
+      "loss": 0.7641,
+      "step": 10873
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6177802477788404,
+      "learning_rate": 1.7434919158322516e-05,
+      "loss": 0.5364,
+      "step": 10874
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8419342875134006,
+      "learning_rate": 1.743440885684419e-05,
+      "loss": 0.7294,
+      "step": 10875
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5497102197094699,
+      "learning_rate": 1.743389851208038e-05,
+      "loss": 0.8126,
+      "step": 10876
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.554306092654595,
+      "learning_rate": 1.7433388124034062e-05,
+      "loss": 0.7892,
+      "step": 10877
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4096365605305021,
+      "learning_rate": 1.74328776927082e-05,
+      "loss": 0.7367,
+      "step": 10878
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5815453419755152,
+      "learning_rate": 1.7432367218105772e-05,
+      "loss": 0.8457,
+      "step": 10879
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9670276767453154,
+      "learning_rate": 1.7431856700229752e-05,
+      "loss": 0.8002,
+      "step": 10880
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4425292980746385,
+      "learning_rate": 1.7431346139083108e-05,
+      "loss": 0.7521,
+      "step": 10881
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5972328948311817,
+      "learning_rate": 1.7430835534668814e-05,
+      "loss": 0.5758,
+      "step": 10882
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3204526288479936,
+      "learning_rate": 1.7430324886989843e-05,
+      "loss": 0.8161,
+      "step": 10883
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.662158859067324,
+      "learning_rate": 1.7429814196049166e-05,
+      "loss": 0.7798,
+      "step": 10884
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3693001300247925,
+      "learning_rate": 1.7429303461849764e-05,
+      "loss": 0.8099,
+      "step": 10885
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.531751668238913,
+      "learning_rate": 1.74287926843946e-05,
+      "loss": 0.6999,
+      "step": 10886
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.311448458194745,
+      "learning_rate": 1.7428281863686654e-05,
+      "loss": 0.6685,
+      "step": 10887
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.587161404492896,
+      "learning_rate": 1.74277709997289e-05,
+      "loss": 0.5404,
+      "step": 10888
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4512598028249013,
+      "learning_rate": 1.742726009252431e-05,
+      "loss": 0.6545,
+      "step": 10889
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6149259128317088,
+      "learning_rate": 1.7426749142075863e-05,
+      "loss": 0.543,
+      "step": 10890
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6150747055804653,
+      "learning_rate": 1.742623814838653e-05,
+      "loss": 0.6873,
+      "step": 10891
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.512654998492143,
+      "learning_rate": 1.742572711145929e-05,
+      "loss": 0.7363,
+      "step": 10892
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5231171098464,
+      "learning_rate": 1.7425216031297113e-05,
+      "loss": 0.7598,
+      "step": 10893
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4496041529940966,
+      "learning_rate": 1.7424704907902975e-05,
+      "loss": 0.7206,
+      "step": 10894
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.698110548667653,
+      "learning_rate": 1.742419374127986e-05,
+      "loss": 0.7098,
+      "step": 10895
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.608163717249762,
+      "learning_rate": 1.7423682531430736e-05,
+      "loss": 0.723,
+      "step": 10896
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6107437379781755,
+      "learning_rate": 1.742317127835858e-05,
+      "loss": 0.7643,
+      "step": 10897
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4568384244643227,
+      "learning_rate": 1.7422659982066373e-05,
+      "loss": 0.8276,
+      "step": 10898
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3301291245921392,
+      "learning_rate": 1.7422148642557088e-05,
+      "loss": 0.7763,
+      "step": 10899
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.603599057151866,
+      "learning_rate": 1.7421637259833707e-05,
+      "loss": 0.5376,
+      "step": 10900
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.351415005469645,
+      "learning_rate": 1.7421125833899203e-05,
+      "loss": 0.6618,
+      "step": 10901
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.726156069262924,
+      "learning_rate": 1.7420614364756553e-05,
+      "loss": 0.6842,
+      "step": 10902
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5256160912867953,
+      "learning_rate": 1.7420102852408735e-05,
+      "loss": 0.9131,
+      "step": 10903
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4108080465561015,
+      "learning_rate": 1.7419591296858733e-05,
+      "loss": 0.8269,
+      "step": 10904
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.597707949996584,
+      "learning_rate": 1.741907969810952e-05,
+      "loss": 0.5254,
+      "step": 10905
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4606437500938005,
+      "learning_rate": 1.7418568056164077e-05,
+      "loss": 0.7869,
+      "step": 10906
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.488054488931763,
+      "learning_rate": 1.7418056371025383e-05,
+      "loss": 0.5976,
+      "step": 10907
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5954094072170114,
+      "learning_rate": 1.7417544642696414e-05,
+      "loss": 0.554,
+      "step": 10908
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8302116020706263,
+      "learning_rate": 1.741703287118015e-05,
+      "loss": 0.9197,
+      "step": 10909
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.425660488957861,
+      "learning_rate": 1.7416521056479577e-05,
+      "loss": 0.8727,
+      "step": 10910
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.822294528695285,
+      "learning_rate": 1.7416009198597668e-05,
+      "loss": 0.6455,
+      "step": 10911
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4023218312894041,
+      "learning_rate": 1.7415497297537407e-05,
+      "loss": 0.7749,
+      "step": 10912
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.353156211577425,
+      "learning_rate": 1.7414985353301773e-05,
+      "loss": 0.7805,
+      "step": 10913
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6164102716733623,
+      "learning_rate": 1.7414473365893744e-05,
+      "loss": 0.576,
+      "step": 10914
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.793959124598391,
+      "learning_rate": 1.7413961335316305e-05,
+      "loss": 0.8726,
+      "step": 10915
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4910336216515094,
+      "learning_rate": 1.7413449261572437e-05,
+      "loss": 0.8707,
+      "step": 10916
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6375812259647224,
+      "learning_rate": 1.741293714466512e-05,
+      "loss": 0.7051,
+      "step": 10917
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4130971685748674,
+      "learning_rate": 1.7412424984597334e-05,
+      "loss": 0.7254,
+      "step": 10918
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5900866058068893,
+      "learning_rate": 1.7411912781372066e-05,
+      "loss": 0.5313,
+      "step": 10919
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4246074704436023,
+      "learning_rate": 1.7411400534992292e-05,
+      "loss": 0.6586,
+      "step": 10920
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1871973112953644,
+      "learning_rate": 1.7410888245461e-05,
+      "loss": 0.6366,
+      "step": 10921
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5650464289647739,
+      "learning_rate": 1.7410375912781173e-05,
+      "loss": 0.8092,
+      "step": 10922
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6193855216234312,
+      "learning_rate": 1.7409863536955788e-05,
+      "loss": 0.839,
+      "step": 10923
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5842705771339582,
+      "learning_rate": 1.7409351117987833e-05,
+      "loss": 0.6807,
+      "step": 10924
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3739223772537104,
+      "learning_rate": 1.7408838655880287e-05,
+      "loss": 0.7889,
+      "step": 10925
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4402447741227529,
+      "learning_rate": 1.7408326150636138e-05,
+      "loss": 0.9265,
+      "step": 10926
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5953320825443638,
+      "learning_rate": 1.740781360225837e-05,
+      "loss": 0.7105,
+      "step": 10927
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5023625724384422,
+      "learning_rate": 1.7407301010749967e-05,
+      "loss": 0.6368,
+      "step": 10928
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.39129847708175,
+      "learning_rate": 1.740678837611391e-05,
+      "loss": 0.7514,
+      "step": 10929
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.131734596396421,
+      "learning_rate": 1.7406275698353184e-05,
+      "loss": 0.8107,
+      "step": 10930
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.575720368176868,
+      "learning_rate": 1.740576297747078e-05,
+      "loss": 0.7591,
+      "step": 10931
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.3206590749756546,
+      "learning_rate": 1.7405250213469678e-05,
+      "loss": 0.7703,
+      "step": 10932
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5092515136025795,
+      "learning_rate": 1.7404737406352866e-05,
+      "loss": 0.6667,
+      "step": 10933
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.396352088146183,
+      "learning_rate": 1.7404224556123324e-05,
+      "loss": 0.6965,
+      "step": 10934
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8201686119942473,
+      "learning_rate": 1.740371166278405e-05,
+      "loss": 0.7541,
+      "step": 10935
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5738632029583458,
+      "learning_rate": 1.7403198726338017e-05,
+      "loss": 0.5256,
+      "step": 10936
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.387484551868026,
+      "learning_rate": 1.740268574678822e-05,
+      "loss": 0.819,
+      "step": 10937
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.463849357729568,
+      "learning_rate": 1.7402172724137643e-05,
+      "loss": 0.8327,
+      "step": 10938
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3499827724362172,
+      "learning_rate": 1.7401659658389268e-05,
+      "loss": 0.7553,
+      "step": 10939
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6633105279840759,
+      "learning_rate": 1.7401146549546093e-05,
+      "loss": 0.7478,
+      "step": 10940
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5491815646820957,
+      "learning_rate": 1.7400633397611097e-05,
+      "loss": 0.804,
+      "step": 10941
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7685640511476832,
+      "learning_rate": 1.7400120202587273e-05,
+      "loss": 0.7733,
+      "step": 10942
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3809888062162496,
+      "learning_rate": 1.7399606964477604e-05,
+      "loss": 0.7826,
+      "step": 10943
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0683033211375914,
+      "learning_rate": 1.7399093683285082e-05,
+      "loss": 0.7103,
+      "step": 10944
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3466483489019858,
+      "learning_rate": 1.739858035901269e-05,
+      "loss": 0.7285,
+      "step": 10945
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8054307564435461,
+      "learning_rate": 1.7398066991663426e-05,
+      "loss": 0.7908,
+      "step": 10946
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5573055745906954,
+      "learning_rate": 1.7397553581240272e-05,
+      "loss": 0.6417,
+      "step": 10947
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2426227979557491,
+      "learning_rate": 1.7397040127746215e-05,
+      "loss": 0.8072,
+      "step": 10948
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7622716736881288,
+      "learning_rate": 1.7396526631184256e-05,
+      "loss": 0.7211,
+      "step": 10949
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.7251211245727034,
+      "learning_rate": 1.739601309155737e-05,
+      "loss": 0.6971,
+      "step": 10950
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4308707040000572,
+      "learning_rate": 1.739549950886856e-05,
+      "loss": 0.746,
+      "step": 10951
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6914914552511834,
+      "learning_rate": 1.7394985883120807e-05,
+      "loss": 0.5417,
+      "step": 10952
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6023950282679769,
+      "learning_rate": 1.7394472214317107e-05,
+      "loss": 0.7091,
+      "step": 10953
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.883834923366844,
+      "learning_rate": 1.7393958502460447e-05,
+      "loss": 0.7151,
+      "step": 10954
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4448375506702902,
+      "learning_rate": 1.739344474755382e-05,
+      "loss": 0.734,
+      "step": 10955
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3516297337344045,
+      "learning_rate": 1.7392930949600217e-05,
+      "loss": 0.6675,
+      "step": 10956
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3933294415132762,
+      "learning_rate": 1.739241710860263e-05,
+      "loss": 0.7329,
+      "step": 10957
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4733480246996533,
+      "learning_rate": 1.739190322456405e-05,
+      "loss": 0.7861,
+      "step": 10958
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6712854165588922,
+      "learning_rate": 1.7391389297487472e-05,
+      "loss": 0.6745,
+      "step": 10959
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3657136523388442,
+      "learning_rate": 1.739087532737588e-05,
+      "loss": 0.7796,
+      "step": 10960
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.508110168476865,
+      "learning_rate": 1.7390361314232275e-05,
+      "loss": 0.8428,
+      "step": 10961
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2406736604756947,
+      "learning_rate": 1.7389847258059646e-05,
+      "loss": 0.802,
+      "step": 10962
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3332112770986748,
+      "learning_rate": 1.738933315886099e-05,
+      "loss": 0.78,
+      "step": 10963
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3922485820226111,
+      "learning_rate": 1.7388819016639292e-05,
+      "loss": 0.7464,
+      "step": 10964
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7741914533829828,
+      "learning_rate": 1.738830483139755e-05,
+      "loss": 0.6895,
+      "step": 10965
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6105978324441949,
+      "learning_rate": 1.7387790603138764e-05,
+      "loss": 0.5114,
+      "step": 10966
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.532068183852193,
+      "learning_rate": 1.7387276331865916e-05,
+      "loss": 0.6689,
+      "step": 10967
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4357177324031674,
+      "learning_rate": 1.738676201758201e-05,
+      "loss": 0.7926,
+      "step": 10968
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7318030859164116,
+      "learning_rate": 1.7386247660290038e-05,
+      "loss": 0.5818,
+      "step": 10969
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.450029333931958,
+      "learning_rate": 1.738573325999299e-05,
+      "loss": 0.734,
+      "step": 10970
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6233047082619765,
+      "learning_rate": 1.7385218816693864e-05,
+      "loss": 0.7888,
+      "step": 10971
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5767684150232987,
+      "learning_rate": 1.738470433039566e-05,
+      "loss": 0.5745,
+      "step": 10972
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.602652405023472,
+      "learning_rate": 1.738418980110137e-05,
+      "loss": 0.6375,
+      "step": 10973
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4636685736561437,
+      "learning_rate": 1.7383675228813986e-05,
+      "loss": 0.6952,
+      "step": 10974
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.9065050855792602,
+      "learning_rate": 1.7383160613536508e-05,
+      "loss": 0.7215,
+      "step": 10975
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4969647794663323,
+      "learning_rate": 1.738264595527193e-05,
+      "loss": 0.9081,
+      "step": 10976
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5870238272045567,
+      "learning_rate": 1.7382131254023252e-05,
+      "loss": 0.8781,
+      "step": 10977
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.508506212063503,
+      "learning_rate": 1.7381616509793468e-05,
+      "loss": 0.7853,
+      "step": 10978
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1420457514022027,
+      "learning_rate": 1.7381101722585577e-05,
+      "loss": 0.7502,
+      "step": 10979
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6006790241518781,
+      "learning_rate": 1.7380586892402572e-05,
+      "loss": 0.5195,
+      "step": 10980
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4208000463039328,
+      "learning_rate": 1.7380072019247455e-05,
+      "loss": 0.7581,
+      "step": 10981
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.325474306836847,
+      "learning_rate": 1.7379557103123223e-05,
+      "loss": 0.7254,
+      "step": 10982
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7041474421221994,
+      "learning_rate": 1.7379042144032873e-05,
+      "loss": 0.9105,
+      "step": 10983
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5473047688726392,
+      "learning_rate": 1.7378527141979402e-05,
+      "loss": 0.4917,
+      "step": 10984
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4269424422396404,
+      "learning_rate": 1.7378012096965813e-05,
+      "loss": 0.8452,
+      "step": 10985
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4936792676794555,
+      "learning_rate": 1.73774970089951e-05,
+      "loss": 0.8485,
+      "step": 10986
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6063233114642455,
+      "learning_rate": 1.7376981878070265e-05,
+      "loss": 0.5439,
+      "step": 10987
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7006285401830448,
+      "learning_rate": 1.73764667041943e-05,
+      "loss": 0.799,
+      "step": 10988
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.963107762437042,
+      "learning_rate": 1.737595148737022e-05,
+      "loss": 0.8386,
+      "step": 10989
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3649508162659951,
+      "learning_rate": 1.7375436227601007e-05,
+      "loss": 0.6737,
+      "step": 10990
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2214477238657135,
+      "learning_rate": 1.7374920924889675e-05,
+      "loss": 0.7446,
+      "step": 10991
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6337173651582072,
+      "learning_rate": 1.7374405579239217e-05,
+      "loss": 0.5464,
+      "step": 10992
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4234802390120354,
+      "learning_rate": 1.737389019065263e-05,
+      "loss": 0.7749,
+      "step": 10993
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6540613591490116,
+      "learning_rate": 1.7373374759132927e-05,
+      "loss": 0.6931,
+      "step": 10994
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5606242455991117,
+      "learning_rate": 1.73728592846831e-05,
+      "loss": 0.6722,
+      "step": 10995
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5893641086441265,
+      "learning_rate": 1.7372343767306148e-05,
+      "loss": 0.5399,
+      "step": 10996
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2558317199247617,
+      "learning_rate": 1.737182820700508e-05,
+      "loss": 0.7244,
+      "step": 10997
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7939746512593853,
+      "learning_rate": 1.7371312603782893e-05,
+      "loss": 0.7426,
+      "step": 10998
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6412412335661737,
+      "learning_rate": 1.737079695764259e-05,
+      "loss": 0.8846,
+      "step": 10999
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4442621298504383,
+      "learning_rate": 1.7370281268587175e-05,
+      "loss": 0.8592,
+      "step": 11000
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.620199873163993,
+      "learning_rate": 1.736976553661965e-05,
+      "loss": 0.8277,
+      "step": 11001
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3536783587007615,
+      "learning_rate": 1.7369249761743015e-05,
+      "loss": 0.8124,
+      "step": 11002
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8183230152989902,
+      "learning_rate": 1.7368733943960278e-05,
+      "loss": 0.6182,
+      "step": 11003
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5096436699157518,
+      "learning_rate": 1.736821808327443e-05,
+      "loss": 0.7986,
+      "step": 11004
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5780357123446818,
+      "learning_rate": 1.7367702179688494e-05,
+      "loss": 0.5118,
+      "step": 11005
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5656706539722962,
+      "learning_rate": 1.7367186233205457e-05,
+      "loss": 0.6641,
+      "step": 11006
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5106470767129725,
+      "learning_rate": 1.7366670243828333e-05,
+      "loss": 0.8656,
+      "step": 11007
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5061143686168184,
+      "learning_rate": 1.736615421156012e-05,
+      "loss": 0.8341,
+      "step": 11008
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5821785635910395,
+      "learning_rate": 1.7365638136403827e-05,
+      "loss": 0.7805,
+      "step": 11009
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.383025327183449,
+      "learning_rate": 1.7365122018362456e-05,
+      "loss": 0.708,
+      "step": 11010
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5906405189139846,
+      "learning_rate": 1.736460585743901e-05,
+      "loss": 0.5594,
+      "step": 11011
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4040164826750499,
+      "learning_rate": 1.7364089653636503e-05,
+      "loss": 0.7776,
+      "step": 11012
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.439376589934225,
+      "learning_rate": 1.736357340695793e-05,
+      "loss": 0.6796,
+      "step": 11013
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5121998013451976,
+      "learning_rate": 1.7363057117406304e-05,
+      "loss": 0.7993,
+      "step": 11014
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8096534569785119,
+      "learning_rate": 1.7362540784984627e-05,
+      "loss": 0.8387,
+      "step": 11015
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.642887774013293,
+      "learning_rate": 1.7362024409695908e-05,
+      "loss": 0.7509,
+      "step": 11016
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6477106847055114,
+      "learning_rate": 1.736150799154315e-05,
+      "loss": 0.7556,
+      "step": 11017
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6501234319258087,
+      "learning_rate": 1.7360991530529363e-05,
+      "loss": 0.804,
+      "step": 11018
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7516959031460626,
+      "learning_rate": 1.7360475026657552e-05,
+      "loss": 0.7264,
+      "step": 11019
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4513988798852808,
+      "learning_rate": 1.7359958479930726e-05,
+      "loss": 0.729,
+      "step": 11020
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.592761317687257,
+      "learning_rate": 1.7359441890351893e-05,
+      "loss": 0.5116,
+      "step": 11021
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4394518449856406,
+      "learning_rate": 1.735892525792406e-05,
+      "loss": 0.6866,
+      "step": 11022
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3415098337259597,
+      "learning_rate": 1.735840858265023e-05,
+      "loss": 0.8529,
+      "step": 11023
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6283842513223246,
+      "learning_rate": 1.7357891864533418e-05,
+      "loss": 0.7734,
+      "step": 11024
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5734018525918658,
+      "learning_rate": 1.7357375103576628e-05,
+      "loss": 0.7439,
+      "step": 11025
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.285109910901318,
+      "learning_rate": 1.735685829978287e-05,
+      "loss": 0.7326,
+      "step": 11026
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7738855161656153,
+      "learning_rate": 1.735634145315516e-05,
+      "loss": 0.7239,
+      "step": 11027
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4862595156702636,
+      "learning_rate": 1.7355824563696496e-05,
+      "loss": 0.6895,
+      "step": 11028
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8894902739680937,
+      "learning_rate": 1.7355307631409894e-05,
+      "loss": 0.6299,
+      "step": 11029
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6011260124035525,
+      "learning_rate": 1.7354790656298364e-05,
+      "loss": 0.6815,
+      "step": 11030
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3719787126726912,
+      "learning_rate": 1.7354273638364914e-05,
+      "loss": 0.6581,
+      "step": 11031
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5406110165089226,
+      "learning_rate": 1.7353756577612552e-05,
+      "loss": 0.7092,
+      "step": 11032
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6407862593418927,
+      "learning_rate": 1.735323947404429e-05,
+      "loss": 0.7207,
+      "step": 11033
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3707050823601865,
+      "learning_rate": 1.7352722327663146e-05,
+      "loss": 0.7528,
+      "step": 11034
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6450469108819115,
+      "learning_rate": 1.735220513847212e-05,
+      "loss": 0.6691,
+      "step": 11035
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5022222110030798,
+      "learning_rate": 1.735168790647423e-05,
+      "loss": 0.8514,
+      "step": 11036
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6063463838242222,
+      "learning_rate": 1.735117063167248e-05,
+      "loss": 0.849,
+      "step": 11037
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7683663325109553,
+      "learning_rate": 1.7350653314069898e-05,
+      "loss": 0.6632,
+      "step": 11038
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3786876048228747,
+      "learning_rate": 1.7350135953669477e-05,
+      "loss": 0.8403,
+      "step": 11039
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4633149879216338,
+      "learning_rate": 1.7349618550474237e-05,
+      "loss": 0.7542,
+      "step": 11040
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4259202808635758,
+      "learning_rate": 1.7349101104487195e-05,
+      "loss": 0.7895,
+      "step": 11041
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9882798025668371,
+      "learning_rate": 1.734858361571136e-05,
+      "loss": 0.7889,
+      "step": 11042
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4960952248835266,
+      "learning_rate": 1.7348066084149743e-05,
+      "loss": 0.766,
+      "step": 11043
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4456089389243882,
+      "learning_rate": 1.7347548509805357e-05,
+      "loss": 0.709,
+      "step": 11044
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4038555798338213,
+      "learning_rate": 1.7347030892681217e-05,
+      "loss": 0.7386,
+      "step": 11045
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4437459338330594,
+      "learning_rate": 1.734651323278034e-05,
+      "loss": 0.7269,
+      "step": 11046
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2692357304841595,
+      "learning_rate": 1.7345995530105737e-05,
+      "loss": 0.7053,
+      "step": 11047
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6366736683233184,
+      "learning_rate": 1.734547778466042e-05,
+      "loss": 0.5661,
+      "step": 11048
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5209425255862234,
+      "learning_rate": 1.734495999644741e-05,
+      "loss": 0.7582,
+      "step": 11049
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.520101035439223,
+      "learning_rate": 1.7344442165469714e-05,
+      "loss": 0.938,
+      "step": 11050
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.635049599670588,
+      "learning_rate": 1.7343924291730352e-05,
+      "loss": 0.7487,
+      "step": 11051
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5957532930941396,
+      "learning_rate": 1.7343406375232333e-05,
+      "loss": 0.576,
+      "step": 11052
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3237025892687626,
+      "learning_rate": 1.7342888415978683e-05,
+      "loss": 0.7299,
+      "step": 11053
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.376857818260447,
+      "learning_rate": 1.734237041397241e-05,
+      "loss": 0.7277,
+      "step": 11054
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6205278224656923,
+      "learning_rate": 1.734185236921653e-05,
+      "loss": 0.8863,
+      "step": 11055
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4308411670552141,
+      "learning_rate": 1.7341334281714067e-05,
+      "loss": 0.7402,
+      "step": 11056
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.540343437140956,
+      "learning_rate": 1.7340816151468026e-05,
+      "loss": 0.854,
+      "step": 11057
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4905343344614206,
+      "learning_rate": 1.734029797848143e-05,
+      "loss": 0.7574,
+      "step": 11058
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7437879773731708,
+      "learning_rate": 1.73397797627573e-05,
+      "loss": 0.8129,
+      "step": 11059
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3691252410854857,
+      "learning_rate": 1.7339261504298645e-05,
+      "loss": 0.7601,
+      "step": 11060
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.452008307905085,
+      "learning_rate": 1.7338743203108484e-05,
+      "loss": 0.6882,
+      "step": 11061
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5853509117862646,
+      "learning_rate": 1.7338224859189837e-05,
+      "loss": 0.467,
+      "step": 11062
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5886575119304569,
+      "learning_rate": 1.7337706472545727e-05,
+      "loss": 0.5702,
+      "step": 11063
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3403196649445592,
+      "learning_rate": 1.733718804317916e-05,
+      "loss": 0.7227,
+      "step": 11064
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7243640424842286,
+      "learning_rate": 1.7336669571093165e-05,
+      "loss": 0.7672,
+      "step": 11065
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6947523816100667,
+      "learning_rate": 1.7336151056290758e-05,
+      "loss": 0.6708,
+      "step": 11066
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.458617558997164,
+      "learning_rate": 1.7335632498774956e-05,
+      "loss": 0.6841,
+      "step": 11067
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.591270438532569,
+      "learning_rate": 1.7335113898548775e-05,
+      "loss": 0.4913,
+      "step": 11068
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.383671563180492,
+      "learning_rate": 1.7334595255615243e-05,
+      "loss": 0.706,
+      "step": 11069
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6266376033528522,
+      "learning_rate": 1.7334076569977372e-05,
+      "loss": 0.7118,
+      "step": 11070
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3789785197299327,
+      "learning_rate": 1.733355784163819e-05,
+      "loss": 0.7862,
+      "step": 11071
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5659664595744887,
+      "learning_rate": 1.733303907060071e-05,
+      "loss": 0.8732,
+      "step": 11072
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4860363847553928,
+      "learning_rate": 1.7332520256867954e-05,
+      "loss": 0.8284,
+      "step": 11073
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4293368274492608,
+      "learning_rate": 1.7332001400442946e-05,
+      "loss": 0.7028,
+      "step": 11074
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4323887692891097,
+      "learning_rate": 1.7331482501328704e-05,
+      "loss": 0.8386,
+      "step": 11075
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.659361825578528,
+      "learning_rate": 1.733096355952825e-05,
+      "loss": 0.797,
+      "step": 11076
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5064621938998246,
+      "learning_rate": 1.7330444575044603e-05,
+      "loss": 0.844,
+      "step": 11077
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.528799911397917,
+      "learning_rate": 1.7329925547880785e-05,
+      "loss": 0.7042,
+      "step": 11078
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7269900947772003,
+      "learning_rate": 1.7329406478039825e-05,
+      "loss": 0.6797,
+      "step": 11079
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3627736771355778,
+      "learning_rate": 1.732888736552474e-05,
+      "loss": 0.7352,
+      "step": 11080
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.497701399844847,
+      "learning_rate": 1.732836821033855e-05,
+      "loss": 0.7657,
+      "step": 11081
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.456011976187928,
+      "learning_rate": 1.732784901248428e-05,
+      "loss": 0.7012,
+      "step": 11082
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5611977692841503,
+      "learning_rate": 1.7327329771964952e-05,
+      "loss": 0.8687,
+      "step": 11083
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6427404902865843,
+      "learning_rate": 1.7326810488783593e-05,
+      "loss": 0.8101,
+      "step": 11084
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4520125193314382,
+      "learning_rate": 1.732629116294322e-05,
+      "loss": 0.7139,
+      "step": 11085
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7128760854233527,
+      "learning_rate": 1.7325771794446865e-05,
+      "loss": 0.8085,
+      "step": 11086
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3952114771607058,
+      "learning_rate": 1.7325252383297547e-05,
+      "loss": 0.7656,
+      "step": 11087
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4132018349984645,
+      "learning_rate": 1.7324732929498287e-05,
+      "loss": 0.7917,
+      "step": 11088
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3800530362722692,
+      "learning_rate": 1.7324213433052115e-05,
+      "loss": 0.6967,
+      "step": 11089
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.983582301984668,
+      "learning_rate": 1.7323693893962055e-05,
+      "loss": 0.8329,
+      "step": 11090
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.324612896122133,
+      "learning_rate": 1.732317431223113e-05,
+      "loss": 0.6807,
+      "step": 11091
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2916855099050186,
+      "learning_rate": 1.732265468786236e-05,
+      "loss": 0.7976,
+      "step": 11092
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5891661656863902,
+      "learning_rate": 1.732213502085878e-05,
+      "loss": 0.6665,
+      "step": 11093
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4154860940927774,
+      "learning_rate": 1.7321615311223413e-05,
+      "loss": 0.8193,
+      "step": 11094
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6359278546795486,
+      "learning_rate": 1.7321095558959284e-05,
+      "loss": 0.7159,
+      "step": 11095
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8550924224437098,
+      "learning_rate": 1.732057576406942e-05,
+      "loss": 0.6753,
+      "step": 11096
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7678996221759251,
+      "learning_rate": 1.7320055926556843e-05,
+      "loss": 0.7521,
+      "step": 11097
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7356336622075446,
+      "learning_rate": 1.7319536046424583e-05,
+      "loss": 0.6687,
+      "step": 11098
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.673732876571427,
+      "learning_rate": 1.731901612367567e-05,
+      "loss": 0.7908,
+      "step": 11099
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4803353747660244,
+      "learning_rate": 1.7318496158313124e-05,
+      "loss": 0.6292,
+      "step": 11100
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.571844949350462,
+      "learning_rate": 1.731797615033998e-05,
+      "loss": 0.8165,
+      "step": 11101
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5787947994718476,
+      "learning_rate": 1.731745609975926e-05,
+      "loss": 0.533,
+      "step": 11102
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7265316588637065,
+      "learning_rate": 1.7316936006573994e-05,
+      "loss": 0.7431,
+      "step": 11103
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7589653544987442,
+      "learning_rate": 1.731641587078721e-05,
+      "loss": 0.7538,
+      "step": 11104
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4542790924734987,
+      "learning_rate": 1.7315895692401937e-05,
+      "loss": 0.7805,
+      "step": 11105
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.431011941813201,
+      "learning_rate": 1.7315375471421203e-05,
+      "loss": 0.7408,
+      "step": 11106
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5954323516390805,
+      "learning_rate": 1.7314855207848037e-05,
+      "loss": 0.7193,
+      "step": 11107
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6982080131638229,
+      "learning_rate": 1.731433490168547e-05,
+      "loss": 0.7181,
+      "step": 11108
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2963359265942276,
+      "learning_rate": 1.7313814552936527e-05,
+      "loss": 0.7574,
+      "step": 11109
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.404813222516506,
+      "learning_rate": 1.7313294161604236e-05,
+      "loss": 0.8226,
+      "step": 11110
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2125162176190536,
+      "learning_rate": 1.7312773727691637e-05,
+      "loss": 0.7876,
+      "step": 11111
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3412278458705735,
+      "learning_rate": 1.7312253251201754e-05,
+      "loss": 0.8323,
+      "step": 11112
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6237681161397053,
+      "learning_rate": 1.7311732732137617e-05,
+      "loss": 0.7401,
+      "step": 11113
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4943467960783012,
+      "learning_rate": 1.7311212170502254e-05,
+      "loss": 0.6613,
+      "step": 11114
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4414976923455844,
+      "learning_rate": 1.7310691566298702e-05,
+      "loss": 0.7304,
+      "step": 11115
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3126136399891661,
+      "learning_rate": 1.7310170919529993e-05,
+      "loss": 0.7316,
+      "step": 11116
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.714657349908025,
+      "learning_rate": 1.7309650230199148e-05,
+      "loss": 0.7568,
+      "step": 11117
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4325114396652139,
+      "learning_rate": 1.7309129498309208e-05,
+      "loss": 0.7403,
+      "step": 11118
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.253990406427589,
+      "learning_rate": 1.73086087238632e-05,
+      "loss": 0.6974,
+      "step": 11119
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5722355592357754,
+      "learning_rate": 1.730808790686416e-05,
+      "loss": 0.516,
+      "step": 11120
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.947353843331249,
+      "learning_rate": 1.7307567047315118e-05,
+      "loss": 0.6563,
+      "step": 11121
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5996525238869912,
+      "learning_rate": 1.7307046145219108e-05,
+      "loss": 0.9599,
+      "step": 11122
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5171475931441947,
+      "learning_rate": 1.730652520057916e-05,
+      "loss": 0.7299,
+      "step": 11123
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9225210907339263,
+      "learning_rate": 1.730600421339831e-05,
+      "loss": 0.7232,
+      "step": 11124
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5664828112055937,
+      "learning_rate": 1.7305483183679592e-05,
+      "loss": 0.697,
+      "step": 11125
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2894874253163877,
+      "learning_rate": 1.7304962111426036e-05,
+      "loss": 0.7149,
+      "step": 11126
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.473321489567968,
+      "learning_rate": 1.730444099664068e-05,
+      "loss": 0.7501,
+      "step": 11127
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3598287638905464,
+      "learning_rate": 1.7303919839326554e-05,
+      "loss": 0.7626,
+      "step": 11128
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3148583858572174,
+      "learning_rate": 1.7303398639486696e-05,
+      "loss": 0.6796,
+      "step": 11129
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3488924636531616,
+      "learning_rate": 1.7302877397124137e-05,
+      "loss": 0.673,
+      "step": 11130
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4839107378937493,
+      "learning_rate": 1.7302356112241913e-05,
+      "loss": 0.779,
+      "step": 11131
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5805470557055046,
+      "learning_rate": 1.7301834784843064e-05,
+      "loss": 0.8665,
+      "step": 11132
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5387138776330005,
+      "learning_rate": 1.7301313414930618e-05,
+      "loss": 0.721,
+      "step": 11133
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4371255378584293,
+      "learning_rate": 1.7300792002507614e-05,
+      "loss": 0.7793,
+      "step": 11134
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8252259120211805,
+      "learning_rate": 1.730027054757709e-05,
+      "loss": 0.6286,
+      "step": 11135
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6051760789336244,
+      "learning_rate": 1.7299749050142074e-05,
+      "loss": 0.5455,
+      "step": 11136
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.519652618041402,
+      "learning_rate": 1.7299227510205613e-05,
+      "loss": 0.7609,
+      "step": 11137
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.571171205745188,
+      "learning_rate": 1.7298705927770737e-05,
+      "loss": 0.507,
+      "step": 11138
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4324345568771906,
+      "learning_rate": 1.7298184302840485e-05,
+      "loss": 0.7171,
+      "step": 11139
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.433594052286467,
+      "learning_rate": 1.729766263541789e-05,
+      "loss": 0.6558,
+      "step": 11140
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6461319161336767,
+      "learning_rate": 1.7297140925505996e-05,
+      "loss": 0.7697,
+      "step": 11141
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.330944742596503,
+      "learning_rate": 1.7296619173107837e-05,
+      "loss": 0.7057,
+      "step": 11142
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3076217153863774,
+      "learning_rate": 1.7296097378226452e-05,
+      "loss": 0.7144,
+      "step": 11143
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4701475372620967,
+      "learning_rate": 1.7295575540864878e-05,
+      "loss": 0.7223,
+      "step": 11144
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4363539596151538,
+      "learning_rate": 1.729505366102615e-05,
+      "loss": 0.6257,
+      "step": 11145
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.592918535548156,
+      "learning_rate": 1.7294531738713313e-05,
+      "loss": 0.7966,
+      "step": 11146
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.523385140068644,
+      "learning_rate": 1.7294009773929405e-05,
+      "loss": 0.9047,
+      "step": 11147
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.3205382916036643,
+      "learning_rate": 1.7293487766677457e-05,
+      "loss": 0.6819,
+      "step": 11148
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.033883412540314,
+      "learning_rate": 1.7292965716960518e-05,
+      "loss": 0.8371,
+      "step": 11149
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5275031440821318,
+      "learning_rate": 1.7292443624781624e-05,
+      "loss": 0.8292,
+      "step": 11150
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.570011959383109,
+      "learning_rate": 1.729192149014381e-05,
+      "loss": 0.6232,
+      "step": 11151
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4499757421586452,
+      "learning_rate": 1.729139931305013e-05,
+      "loss": 0.64,
+      "step": 11152
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4797407514301166,
+      "learning_rate": 1.7290877093503606e-05,
+      "loss": 0.7637,
+      "step": 11153
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5543862082544864,
+      "learning_rate": 1.729035483150729e-05,
+      "loss": 0.7209,
+      "step": 11154
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7201060045433378,
+      "learning_rate": 1.728983252706422e-05,
+      "loss": 0.6572,
+      "step": 11155
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5106593794970413,
+      "learning_rate": 1.7289310180177438e-05,
+      "loss": 0.7066,
+      "step": 11156
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4804650474204961,
+      "learning_rate": 1.7288787790849984e-05,
+      "loss": 0.8153,
+      "step": 11157
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6425443250843605,
+      "learning_rate": 1.72882653590849e-05,
+      "loss": 0.5614,
+      "step": 11158
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0286475826704016,
+      "learning_rate": 1.7287742884885225e-05,
+      "loss": 0.842,
+      "step": 11159
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8368942423672816,
+      "learning_rate": 1.728722036825401e-05,
+      "loss": 0.6867,
+      "step": 11160
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5793620922533658,
+      "learning_rate": 1.7286697809194285e-05,
+      "loss": 0.7261,
+      "step": 11161
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.392579927851244,
+      "learning_rate": 1.72861752077091e-05,
+      "loss": 0.7471,
+      "step": 11162
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9895832456167228,
+      "learning_rate": 1.7285652563801493e-05,
+      "loss": 0.718,
+      "step": 11163
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4961598231785145,
+      "learning_rate": 1.7285129877474516e-05,
+      "loss": 0.7527,
+      "step": 11164
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.328070666231994,
+      "learning_rate": 1.7284607148731206e-05,
+      "loss": 0.7837,
+      "step": 11165
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5570582205178867,
+      "learning_rate": 1.7284084377574605e-05,
+      "loss": 0.819,
+      "step": 11166
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4860356698682704,
+      "learning_rate": 1.728356156400776e-05,
+      "loss": 0.8199,
+      "step": 11167
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6243418912058556,
+      "learning_rate": 1.7283038708033707e-05,
+      "loss": 0.7581,
+      "step": 11168
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5441711323672807,
+      "learning_rate": 1.72825158096555e-05,
+      "loss": 0.8019,
+      "step": 11169
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3279463095796942,
+      "learning_rate": 1.728199286887619e-05,
+      "loss": 0.8596,
+      "step": 11170
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.684284305195595,
+      "learning_rate": 1.72814698856988e-05,
+      "loss": 0.8773,
+      "step": 11171
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.676354617740859,
+      "learning_rate": 1.7280946860126394e-05,
+      "loss": 0.743,
+      "step": 11172
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5296519625030776,
+      "learning_rate": 1.7280423792162008e-05,
+      "loss": 0.8762,
+      "step": 11173
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.507875940339873,
+      "learning_rate": 1.727990068180869e-05,
+      "loss": 0.7286,
+      "step": 11174
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.463174526942452,
+      "learning_rate": 1.7279377529069484e-05,
+      "loss": 0.6999,
+      "step": 11175
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3671143434847142,
+      "learning_rate": 1.7278854333947438e-05,
+      "loss": 0.723,
+      "step": 11176
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4168259291468286,
+      "learning_rate": 1.72783310964456e-05,
+      "loss": 0.7882,
+      "step": 11177
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7320002884585315,
+      "learning_rate": 1.727780781656701e-05,
+      "loss": 0.7723,
+      "step": 11178
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.568032075320077,
+      "learning_rate": 1.727728449431472e-05,
+      "loss": 0.7566,
+      "step": 11179
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3540399001632237,
+      "learning_rate": 1.7276761129691776e-05,
+      "loss": 0.8097,
+      "step": 11180
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3232211396348674,
+      "learning_rate": 1.7276237722701227e-05,
+      "loss": 0.673,
+      "step": 11181
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5598143274038816,
+      "learning_rate": 1.7275714273346117e-05,
+      "loss": 0.6612,
+      "step": 11182
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4172221614727596,
+      "learning_rate": 1.7275190781629494e-05,
+      "loss": 0.811,
+      "step": 11183
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.633072756913551,
+      "learning_rate": 1.7274667247554407e-05,
+      "loss": 0.5389,
+      "step": 11184
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3342620683318236,
+      "learning_rate": 1.7274143671123905e-05,
+      "loss": 0.8887,
+      "step": 11185
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9699496523421083,
+      "learning_rate": 1.7273620052341037e-05,
+      "loss": 0.8676,
+      "step": 11186
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6479781002145273,
+      "learning_rate": 1.727309639120885e-05,
+      "loss": 0.7848,
+      "step": 11187
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6549571693572032,
+      "learning_rate": 1.727257268773039e-05,
+      "loss": 0.8466,
+      "step": 11188
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5448270944059843,
+      "learning_rate": 1.7272048941908713e-05,
+      "loss": 0.789,
+      "step": 11189
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8965870277177808,
+      "learning_rate": 1.7271525153746865e-05,
+      "loss": 0.7145,
+      "step": 11190
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7838218226453244,
+      "learning_rate": 1.7271001323247892e-05,
+      "loss": 0.8332,
+      "step": 11191
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.502413211419416,
+      "learning_rate": 1.727047745041485e-05,
+      "loss": 0.8276,
+      "step": 11192
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7321261502939649,
+      "learning_rate": 1.7269953535250782e-05,
+      "loss": 0.8213,
+      "step": 11193
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4295542493067352,
+      "learning_rate": 1.726942957775875e-05,
+      "loss": 0.7398,
+      "step": 11194
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.218388257074388,
+      "learning_rate": 1.726890557794179e-05,
+      "loss": 0.8642,
+      "step": 11195
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.407948181102794,
+      "learning_rate": 1.7268381535802966e-05,
+      "loss": 0.7911,
+      "step": 11196
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3590751305127622,
+      "learning_rate": 1.726785745134532e-05,
+      "loss": 0.7282,
+      "step": 11197
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5507039959452514,
+      "learning_rate": 1.7267333324571912e-05,
+      "loss": 0.5141,
+      "step": 11198
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5995359408287089,
+      "learning_rate": 1.7266809155485783e-05,
+      "loss": 0.7165,
+      "step": 11199
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6418724520713015,
+      "learning_rate": 1.7266284944089994e-05,
+      "loss": 0.6979,
+      "step": 11200
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.640251906279664,
+      "learning_rate": 1.7265760690387594e-05,
+      "loss": 0.7641,
+      "step": 11201
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4029812175293863,
+      "learning_rate": 1.7265236394381634e-05,
+      "loss": 0.8339,
+      "step": 11202
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.361436348313516,
+      "learning_rate": 1.7264712056075167e-05,
+      "loss": 0.8051,
+      "step": 11203
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.609833021647166,
+      "learning_rate": 1.7264187675471246e-05,
+      "loss": 0.7708,
+      "step": 11204
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5770482499741848,
+      "learning_rate": 1.7263663252572923e-05,
+      "loss": 0.5046,
+      "step": 11205
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.9379627606828995,
+      "learning_rate": 1.7263138787383255e-05,
+      "loss": 0.841,
+      "step": 11206
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6416344138687256,
+      "learning_rate": 1.7262614279905295e-05,
+      "loss": 0.649,
+      "step": 11207
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4735548291643685,
+      "learning_rate": 1.7262089730142094e-05,
+      "loss": 0.7295,
+      "step": 11208
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8061965530809847,
+      "learning_rate": 1.7261565138096706e-05,
+      "loss": 0.8278,
+      "step": 11209
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7392702541888234,
+      "learning_rate": 1.7261040503772187e-05,
+      "loss": 0.758,
+      "step": 11210
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4445181292874585,
+      "learning_rate": 1.7260515827171596e-05,
+      "loss": 0.722,
+      "step": 11211
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4214137529156488,
+      "learning_rate": 1.7259991108297976e-05,
+      "loss": 0.8349,
+      "step": 11212
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6768256781658366,
+      "learning_rate": 1.7259466347154393e-05,
+      "loss": 0.7549,
+      "step": 11213
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0561690200456884,
+      "learning_rate": 1.72589415437439e-05,
+      "loss": 0.6669,
+      "step": 11214
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1454230314516796,
+      "learning_rate": 1.725841669806955e-05,
+      "loss": 0.6996,
+      "step": 11215
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.178443889967482,
+      "learning_rate": 1.72578918101344e-05,
+      "loss": 0.7561,
+      "step": 11216
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5757107777680592,
+      "learning_rate": 1.7257366879941508e-05,
+      "loss": 0.7779,
+      "step": 11217
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.354282343062933,
+      "learning_rate": 1.7256841907493928e-05,
+      "loss": 0.7593,
+      "step": 11218
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5903841758763835,
+      "learning_rate": 1.7256316892794712e-05,
+      "loss": 0.7913,
+      "step": 11219
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7990359800074507,
+      "learning_rate": 1.7255791835846926e-05,
+      "loss": 0.8181,
+      "step": 11220
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8127173578968212,
+      "learning_rate": 1.725526673665362e-05,
+      "loss": 0.737,
+      "step": 11221
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.179485826098571,
+      "learning_rate": 1.7254741595217856e-05,
+      "loss": 0.7378,
+      "step": 11222
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3785003384617038,
+      "learning_rate": 1.725421641154269e-05,
+      "loss": 0.7737,
+      "step": 11223
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6819276600897053,
+      "learning_rate": 1.725369118563118e-05,
+      "loss": 0.6767,
+      "step": 11224
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5045660683680542,
+      "learning_rate": 1.725316591748638e-05,
+      "loss": 0.7871,
+      "step": 11225
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3978724423408186,
+      "learning_rate": 1.7252640607111354e-05,
+      "loss": 0.7486,
+      "step": 11226
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.384889390729515,
+      "learning_rate": 1.725211525450916e-05,
+      "loss": 0.7172,
+      "step": 11227
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4776852848209492,
+      "learning_rate": 1.725158985968285e-05,
+      "loss": 0.6519,
+      "step": 11228
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8857987181085019,
+      "learning_rate": 1.725106442263549e-05,
+      "loss": 0.7245,
+      "step": 11229
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6209829212185977,
+      "learning_rate": 1.7250538943370142e-05,
+      "loss": 0.5342,
+      "step": 11230
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5582168944350492,
+      "learning_rate": 1.7250013421889857e-05,
+      "loss": 0.8191,
+      "step": 11231
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1781679666049776,
+      "learning_rate": 1.7249487858197698e-05,
+      "loss": 0.8666,
+      "step": 11232
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3684545515286433,
+      "learning_rate": 1.7248962252296725e-05,
+      "loss": 0.7872,
+      "step": 11233
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.862281455983775,
+      "learning_rate": 1.724843660419e-05,
+      "loss": 0.7418,
+      "step": 11234
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.551364986117984,
+      "learning_rate": 1.7247910913880583e-05,
+      "loss": 0.9177,
+      "step": 11235
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4929013110737388,
+      "learning_rate": 1.7247385181371535e-05,
+      "loss": 0.825,
+      "step": 11236
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4384115475173715,
+      "learning_rate": 1.7246859406665916e-05,
+      "loss": 0.8953,
+      "step": 11237
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6831768701329368,
+      "learning_rate": 1.7246333589766786e-05,
+      "loss": 0.9282,
+      "step": 11238
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.511716884716156,
+      "learning_rate": 1.7245807730677206e-05,
+      "loss": 0.7,
+      "step": 11239
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.355482642287831,
+      "learning_rate": 1.724528182940024e-05,
+      "loss": 0.8261,
+      "step": 11240
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.5820092492885873,
+      "learning_rate": 1.724475588593895e-05,
+      "loss": 0.5744,
+      "step": 11241
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4995480801605316,
+      "learning_rate": 1.7244229900296398e-05,
+      "loss": 0.7701,
+      "step": 11242
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5996214316170213,
+      "learning_rate": 1.724370387247565e-05,
+      "loss": 0.5718,
+      "step": 11243
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.675562230994187,
+      "learning_rate": 1.7243177802479758e-05,
+      "loss": 0.8451,
+      "step": 11244
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.470703721798311,
+      "learning_rate": 1.7242651690311794e-05,
+      "loss": 0.8693,
+      "step": 11245
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5990950282717087,
+      "learning_rate": 1.724212553597482e-05,
+      "loss": 0.7025,
+      "step": 11246
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.018858086641468,
+      "learning_rate": 1.7241599339471894e-05,
+      "loss": 0.8646,
+      "step": 11247
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.528570481017187,
+      "learning_rate": 1.724107310080609e-05,
+      "loss": 0.7818,
+      "step": 11248
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.33280066450015,
+      "learning_rate": 1.724054681998046e-05,
+      "loss": 0.8232,
+      "step": 11249
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4869905556064318,
+      "learning_rate": 1.724002049699808e-05,
+      "loss": 0.8094,
+      "step": 11250
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4070763530119954,
+      "learning_rate": 1.7239494131862003e-05,
+      "loss": 0.8291,
+      "step": 11251
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.459567287350975,
+      "learning_rate": 1.7238967724575305e-05,
+      "loss": 0.6437,
+      "step": 11252
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.836983480826327,
+      "learning_rate": 1.723844127514104e-05,
+      "loss": 0.987,
+      "step": 11253
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2152327046895668,
+      "learning_rate": 1.7237914783562283e-05,
+      "loss": 0.8372,
+      "step": 11254
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4090489629775331,
+      "learning_rate": 1.723738824984209e-05,
+      "loss": 0.6283,
+      "step": 11255
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5265275334192936,
+      "learning_rate": 1.7236861673983534e-05,
+      "loss": 0.7483,
+      "step": 11256
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.381731216553609,
+      "learning_rate": 1.7236335055989676e-05,
+      "loss": 0.8135,
+      "step": 11257
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.907498105501799,
+      "learning_rate": 1.7235808395863586e-05,
+      "loss": 0.7504,
+      "step": 11258
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5575309130769068,
+      "learning_rate": 1.723528169360833e-05,
+      "loss": 0.8188,
+      "step": 11259
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3552723066935906,
+      "learning_rate": 1.723475494922697e-05,
+      "loss": 0.765,
+      "step": 11260
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6260706615256758,
+      "learning_rate": 1.723422816272258e-05,
+      "loss": 0.5829,
+      "step": 11261
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.317145761516306,
+      "learning_rate": 1.723370133409822e-05,
+      "loss": 0.7579,
+      "step": 11262
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6524411202265978,
+      "learning_rate": 1.723317446335696e-05,
+      "loss": 0.5662,
+      "step": 11263
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6200267324148299,
+      "learning_rate": 1.7232647550501875e-05,
+      "loss": 0.5458,
+      "step": 11264
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3811717054588948,
+      "learning_rate": 1.723212059553602e-05,
+      "loss": 0.734,
+      "step": 11265
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2920116668032053,
+      "learning_rate": 1.723159359846247e-05,
+      "loss": 0.7717,
+      "step": 11266
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6132949223748296,
+      "learning_rate": 1.7231066559284293e-05,
+      "loss": 0.5379,
+      "step": 11267
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.525059911681492,
+      "learning_rate": 1.723053947800456e-05,
+      "loss": 0.7783,
+      "step": 11268
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.783098842629143,
+      "learning_rate": 1.7230012354626334e-05,
+      "loss": 0.702,
+      "step": 11269
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.725505608277093,
+      "learning_rate": 1.722948518915269e-05,
+      "loss": 0.7435,
+      "step": 11270
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4882411743811614,
+      "learning_rate": 1.722895798158669e-05,
+      "loss": 0.8399,
+      "step": 11271
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8000724759609015,
+      "learning_rate": 1.7228430731931407e-05,
+      "loss": 0.7818,
+      "step": 11272
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4747670788288927,
+      "learning_rate": 1.7227903440189914e-05,
+      "loss": 0.6753,
+      "step": 11273
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5146662732875174,
+      "learning_rate": 1.722737610636528e-05,
+      "loss": 0.6098,
+      "step": 11274
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5518692883965035,
+      "learning_rate": 1.7226848730460574e-05,
+      "loss": 0.754,
+      "step": 11275
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4706321312356812,
+      "learning_rate": 1.722632131247887e-05,
+      "loss": 0.779,
+      "step": 11276
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6307597240091511,
+      "learning_rate": 1.722579385242323e-05,
+      "loss": 0.7061,
+      "step": 11277
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.558494220432488,
+      "learning_rate": 1.7225266350296733e-05,
+      "loss": 0.5102,
+      "step": 11278
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4134012134021616,
+      "learning_rate": 1.7224738806102447e-05,
+      "loss": 0.6248,
+      "step": 11279
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9765734632044547,
+      "learning_rate": 1.7224211219843444e-05,
+      "loss": 0.7542,
+      "step": 11280
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8095992986370655,
+      "learning_rate": 1.7223683591522798e-05,
+      "loss": 0.8429,
+      "step": 11281
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.473809405099696,
+      "learning_rate": 1.7223155921143578e-05,
+      "loss": 0.6645,
+      "step": 11282
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6642965820550504,
+      "learning_rate": 1.7222628208708857e-05,
+      "loss": 0.8065,
+      "step": 11283
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3946709270117177,
+      "learning_rate": 1.722210045422171e-05,
+      "loss": 0.7769,
+      "step": 11284
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.425869748719891,
+      "learning_rate": 1.7221572657685205e-05,
+      "loss": 0.7224,
+      "step": 11285
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8576016756090699,
+      "learning_rate": 1.722104481910242e-05,
+      "loss": 0.7932,
+      "step": 11286
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4839758343146763,
+      "learning_rate": 1.722051693847642e-05,
+      "loss": 0.769,
+      "step": 11287
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5800787033604379,
+      "learning_rate": 1.721998901581029e-05,
+      "loss": 0.5012,
+      "step": 11288
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.3461692076758234,
+      "learning_rate": 1.7219461051107102e-05,
+      "loss": 0.8111,
+      "step": 11289
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.486337005576016,
+      "learning_rate": 1.721893304436992e-05,
+      "loss": 0.8401,
+      "step": 11290
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.18067999960402,
+      "learning_rate": 1.7218404995601823e-05,
+      "loss": 0.8975,
+      "step": 11291
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3225424167497808,
+      "learning_rate": 1.721787690480589e-05,
+      "loss": 0.7937,
+      "step": 11292
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8706626248726304,
+      "learning_rate": 1.7217348771985194e-05,
+      "loss": 0.6759,
+      "step": 11293
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3870854313750653,
+      "learning_rate": 1.7216820597142804e-05,
+      "loss": 0.6861,
+      "step": 11294
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6634611950831493,
+      "learning_rate": 1.7216292380281802e-05,
+      "loss": 0.7729,
+      "step": 11295
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7853487691754657,
+      "learning_rate": 1.721576412140526e-05,
+      "loss": 0.7408,
+      "step": 11296
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5041817388586622,
+      "learning_rate": 1.721523582051626e-05,
+      "loss": 0.7321,
+      "step": 11297
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4398778645887893,
+      "learning_rate": 1.7214707477617866e-05,
+      "loss": 0.7416,
+      "step": 11298
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5229131116761243,
+      "learning_rate": 1.7214179092713165e-05,
+      "loss": 0.6164,
+      "step": 11299
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.407600766123194,
+      "learning_rate": 1.7213650665805224e-05,
+      "loss": 0.7509,
+      "step": 11300
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.984095210937341,
+      "learning_rate": 1.7213122196897135e-05,
+      "loss": 0.6999,
+      "step": 11301
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8397301077356996,
+      "learning_rate": 1.7212593685991956e-05,
+      "loss": 0.8563,
+      "step": 11302
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5776975512819091,
+      "learning_rate": 1.7212065133092778e-05,
+      "loss": 0.535,
+      "step": 11303
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6025447133102051,
+      "learning_rate": 1.7211536538202673e-05,
+      "loss": 0.5617,
+      "step": 11304
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.586076697401157,
+      "learning_rate": 1.7211007901324714e-05,
+      "loss": 0.7562,
+      "step": 11305
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.50565288657487,
+      "learning_rate": 1.721047922246199e-05,
+      "loss": 0.7469,
+      "step": 11306
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3865845659389933,
+      "learning_rate": 1.720995050161757e-05,
+      "loss": 0.7977,
+      "step": 11307
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.495068431049686,
+      "learning_rate": 1.7209421738794536e-05,
+      "loss": 0.779,
+      "step": 11308
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.563036120367707,
+      "learning_rate": 1.720889293399597e-05,
+      "loss": 0.6897,
+      "step": 11309
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4659982605287594,
+      "learning_rate": 1.720836408722494e-05,
+      "loss": 0.7909,
+      "step": 11310
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6079731215261668,
+      "learning_rate": 1.7207835198484537e-05,
+      "loss": 0.5272,
+      "step": 11311
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6899587586040248,
+      "learning_rate": 1.720730626777783e-05,
+      "loss": 0.6889,
+      "step": 11312
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.467027357274612,
+      "learning_rate": 1.720677729510791e-05,
+      "loss": 0.7436,
+      "step": 11313
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.212431854264559,
+      "learning_rate": 1.720624828047785e-05,
+      "loss": 0.7497,
+      "step": 11314
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3769327281183512,
+      "learning_rate": 1.7205719223890732e-05,
+      "loss": 0.6744,
+      "step": 11315
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7831232099383938,
+      "learning_rate": 1.7205190125349633e-05,
+      "loss": 0.6379,
+      "step": 11316
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4210317106663912,
+      "learning_rate": 1.7204660984857638e-05,
+      "loss": 0.6752,
+      "step": 11317
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.281594788358051,
+      "learning_rate": 1.7204131802417826e-05,
+      "loss": 0.8128,
+      "step": 11318
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7946931862632967,
+      "learning_rate": 1.7203602578033275e-05,
+      "loss": 0.6905,
+      "step": 11319
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4697149730737125,
+      "learning_rate": 1.7203073311707072e-05,
+      "loss": 0.6929,
+      "step": 11320
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3322225738225935,
+      "learning_rate": 1.720254400344229e-05,
+      "loss": 0.6942,
+      "step": 11321
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8317095126285818,
+      "learning_rate": 1.720201465324202e-05,
+      "loss": 0.7843,
+      "step": 11322
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4994596640679372,
+      "learning_rate": 1.7201485261109345e-05,
+      "loss": 0.8167,
+      "step": 11323
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5647166866303117,
+      "learning_rate": 1.720095582704734e-05,
+      "loss": 0.7948,
+      "step": 11324
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.639744219251184,
+      "learning_rate": 1.7200426351059093e-05,
+      "loss": 0.7432,
+      "step": 11325
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2811007761871904,
+      "learning_rate": 1.719989683314768e-05,
+      "loss": 0.7854,
+      "step": 11326
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3722343400974442,
+      "learning_rate": 1.7199367273316187e-05,
+      "loss": 0.7881,
+      "step": 11327
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2809105312340032,
+      "learning_rate": 1.71988376715677e-05,
+      "loss": 0.8434,
+      "step": 11328
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0008431427705715,
+      "learning_rate": 1.71983080279053e-05,
+      "loss": 0.8572,
+      "step": 11329
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2572433861539967,
+      "learning_rate": 1.7197778342332075e-05,
+      "loss": 0.6446,
+      "step": 11330
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8343982984959906,
+      "learning_rate": 1.7197248614851103e-05,
+      "loss": 0.7771,
+      "step": 11331
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4477488202460822,
+      "learning_rate": 1.7196718845465472e-05,
+      "loss": 0.6967,
+      "step": 11332
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5357111026517365,
+      "learning_rate": 1.7196189034178267e-05,
+      "loss": 0.7808,
+      "step": 11333
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.671262855273182,
+      "learning_rate": 1.719565918099257e-05,
+      "loss": 0.7905,
+      "step": 11334
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5467562525476204,
+      "learning_rate": 1.7195129285911465e-05,
+      "loss": 0.8355,
+      "step": 11335
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6486236028176908,
+      "learning_rate": 1.7194599348938042e-05,
+      "loss": 0.7411,
+      "step": 11336
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4654290245592436,
+      "learning_rate": 1.719406937007538e-05,
+      "loss": 0.7718,
+      "step": 11337
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4348857614798591,
+      "learning_rate": 1.7193539349326573e-05,
+      "loss": 0.7023,
+      "step": 11338
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4987681011215834,
+      "learning_rate": 1.71930092866947e-05,
+      "loss": 0.7117,
+      "step": 11339
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4260355761152343,
+      "learning_rate": 1.7192479182182853e-05,
+      "loss": 0.8019,
+      "step": 11340
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9537135065857185,
+      "learning_rate": 1.7191949035794113e-05,
+      "loss": 0.7328,
+      "step": 11341
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4042179253924785,
+      "learning_rate": 1.7191418847531567e-05,
+      "loss": 0.7884,
+      "step": 11342
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5389779450092318,
+      "learning_rate": 1.7190888617398306e-05,
+      "loss": 0.8966,
+      "step": 11343
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.58303028391668,
+      "learning_rate": 1.7190358345397412e-05,
+      "loss": 0.8048,
+      "step": 11344
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1006069028303918,
+      "learning_rate": 1.718982803153198e-05,
+      "loss": 0.7812,
+      "step": 11345
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.534961917979408,
+      "learning_rate": 1.7189297675805088e-05,
+      "loss": 0.8399,
+      "step": 11346
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5284351794068773,
+      "learning_rate": 1.718876727821983e-05,
+      "loss": 0.7739,
+      "step": 11347
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.309703987194345,
+      "learning_rate": 1.7188236838779297e-05,
+      "loss": 0.7904,
+      "step": 11348
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.71953297509526,
+      "learning_rate": 1.718770635748657e-05,
+      "loss": 0.7898,
+      "step": 11349
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7623527763502176,
+      "learning_rate": 1.718717583434474e-05,
+      "loss": 0.8801,
+      "step": 11350
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3624192808189344,
+      "learning_rate": 1.71866452693569e-05,
+      "loss": 0.7483,
+      "step": 11351
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2718756416893298,
+      "learning_rate": 1.718611466252613e-05,
+      "loss": 0.6965,
+      "step": 11352
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5952419177123223,
+      "learning_rate": 1.718558401385553e-05,
+      "loss": 0.7538,
+      "step": 11353
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5245187298654768,
+      "learning_rate": 1.7185053323348187e-05,
+      "loss": 0.6887,
+      "step": 11354
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.0354081587464465,
+      "learning_rate": 1.7184522591007184e-05,
+      "loss": 0.8012,
+      "step": 11355
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4690830577616785,
+      "learning_rate": 1.718399181683562e-05,
+      "loss": 0.6855,
+      "step": 11356
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7691632141880207,
+      "learning_rate": 1.718346100083658e-05,
+      "loss": 0.7539,
+      "step": 11357
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5767844980058903,
+      "learning_rate": 1.7182930143013154e-05,
+      "loss": 0.5376,
+      "step": 11358
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6561765138029085,
+      "learning_rate": 1.7182399243368436e-05,
+      "loss": 0.6391,
+      "step": 11359
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4087703092991466,
+      "learning_rate": 1.7181868301905515e-05,
+      "loss": 0.7457,
+      "step": 11360
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5674807370457157,
+      "learning_rate": 1.7181337318627485e-05,
+      "loss": 0.5144,
+      "step": 11361
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6290186138246225,
+      "learning_rate": 1.7180806293537435e-05,
+      "loss": 0.789,
+      "step": 11362
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.47672316144327,
+      "learning_rate": 1.7180275226638455e-05,
+      "loss": 0.802,
+      "step": 11363
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7268731484788045,
+      "learning_rate": 1.7179744117933642e-05,
+      "loss": 0.764,
+      "step": 11364
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.572439102804579,
+      "learning_rate": 1.7179212967426083e-05,
+      "loss": 0.5608,
+      "step": 11365
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5907318396138261,
+      "learning_rate": 1.7178681775118874e-05,
+      "loss": 0.7667,
+      "step": 11366
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.44594588621671,
+      "learning_rate": 1.717815054101511e-05,
+      "loss": 0.7347,
+      "step": 11367
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.463418590293488,
+      "learning_rate": 1.717761926511788e-05,
+      "loss": 0.801,
+      "step": 11368
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.611047265454063,
+      "learning_rate": 1.7177087947430275e-05,
+      "loss": 0.7874,
+      "step": 11369
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5574087877047498,
+      "learning_rate": 1.717655658795539e-05,
+      "loss": 0.8342,
+      "step": 11370
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.590091892901797,
+      "learning_rate": 1.717602518669633e-05,
+      "loss": 0.5676,
+      "step": 11371
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3392222049307243,
+      "learning_rate": 1.717549374365617e-05,
+      "loss": 0.738,
+      "step": 11372
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6243971144485273,
+      "learning_rate": 1.717496225883802e-05,
+      "loss": 0.5527,
+      "step": 11373
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.383119409194021,
+      "learning_rate": 1.717443073224496e-05,
+      "loss": 0.6993,
+      "step": 11374
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.064628042525441,
+      "learning_rate": 1.7173899163880105e-05,
+      "loss": 0.6982,
+      "step": 11375
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4007818445765667,
+      "learning_rate": 1.7173367553746526e-05,
+      "loss": 0.7984,
+      "step": 11376
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6255076358318405,
+      "learning_rate": 1.7172835901847338e-05,
+      "loss": 0.7669,
+      "step": 11377
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6043661055444259,
+      "learning_rate": 1.7172304208185627e-05,
+      "loss": 0.5654,
+      "step": 11378
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4370893495966028,
+      "learning_rate": 1.7171772472764488e-05,
+      "loss": 0.8574,
+      "step": 11379
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.461729096092866,
+      "learning_rate": 1.717124069558702e-05,
+      "loss": 0.7397,
+      "step": 11380
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5411762039170172,
+      "learning_rate": 1.717070887665632e-05,
+      "loss": 0.8501,
+      "step": 11381
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.8673593815562888,
+      "learning_rate": 1.717017701597548e-05,
+      "loss": 0.746,
+      "step": 11382
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5168042644467297,
+      "learning_rate": 1.7169645113547598e-05,
+      "loss": 0.7922,
+      "step": 11383
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.3597207074981297,
+      "learning_rate": 1.7169113169375778e-05,
+      "loss": 0.8229,
+      "step": 11384
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3047479016410037,
+      "learning_rate": 1.7168581183463106e-05,
+      "loss": 0.748,
+      "step": 11385
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4704340331622192,
+      "learning_rate": 1.7168049155812686e-05,
+      "loss": 0.6381,
+      "step": 11386
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6239501751499581,
+      "learning_rate": 1.7167517086427618e-05,
+      "loss": 0.5696,
+      "step": 11387
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5667070061192834,
+      "learning_rate": 1.716698497531099e-05,
+      "loss": 0.6945,
+      "step": 11388
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5875944791137421,
+      "learning_rate": 1.7166452822465913e-05,
+      "loss": 0.5262,
+      "step": 11389
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4118217783127822,
+      "learning_rate": 1.7165920627895477e-05,
+      "loss": 0.792,
+      "step": 11390
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.7667002055028145,
+      "learning_rate": 1.7165388391602783e-05,
+      "loss": 0.7363,
+      "step": 11391
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5666707029147178,
+      "learning_rate": 1.7164856113590925e-05,
+      "loss": 0.8683,
+      "step": 11392
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6080156363952882,
+      "learning_rate": 1.716432379386301e-05,
+      "loss": 0.8267,
+      "step": 11393
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5780834132342225,
+      "learning_rate": 1.7163791432422133e-05,
+      "loss": 0.7942,
+      "step": 11394
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6725895120928327,
+      "learning_rate": 1.7163259029271392e-05,
+      "loss": 0.7805,
+      "step": 11395
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8604963187627994,
+      "learning_rate": 1.716272658441389e-05,
+      "loss": 0.6788,
+      "step": 11396
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4355629685794893,
+      "learning_rate": 1.716219409785273e-05,
+      "loss": 0.6457,
+      "step": 11397
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7585082151344662,
+      "learning_rate": 1.7161661569591004e-05,
+      "loss": 0.7226,
+      "step": 11398
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.1994567936757683,
+      "learning_rate": 1.716112899963182e-05,
+      "loss": 0.7887,
+      "step": 11399
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6993923490975156,
+      "learning_rate": 1.7160596387978273e-05,
+      "loss": 0.828,
+      "step": 11400
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3321960046596355,
+      "learning_rate": 1.7160063734633468e-05,
+      "loss": 0.7691,
+      "step": 11401
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.595184389107336,
+      "learning_rate": 1.7159531039600507e-05,
+      "loss": 0.7894,
+      "step": 11402
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4201191492780163,
+      "learning_rate": 1.715899830288249e-05,
+      "loss": 0.8146,
+      "step": 11403
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.477738736012383,
+      "learning_rate": 1.7158465524482512e-05,
+      "loss": 0.7624,
+      "step": 11404
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5377143550849244,
+      "learning_rate": 1.715793270440369e-05,
+      "loss": 0.7888,
+      "step": 11405
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3876602813049308,
+      "learning_rate": 1.715739984264911e-05,
+      "loss": 0.7874,
+      "step": 11406
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4146150316947537,
+      "learning_rate": 1.7156866939221888e-05,
+      "loss": 0.7083,
+      "step": 11407
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3419516850889606,
+      "learning_rate": 1.715633399412512e-05,
+      "loss": 0.7312,
+      "step": 11408
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4138262447768175,
+      "learning_rate": 1.7155801007361908e-05,
+      "loss": 0.6643,
+      "step": 11409
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.210489296572203,
+      "learning_rate": 1.7155267978935355e-05,
+      "loss": 0.7462,
+      "step": 11410
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4537739967505614,
+      "learning_rate": 1.7154734908848568e-05,
+      "loss": 0.8083,
+      "step": 11411
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3959292471144538,
+      "learning_rate": 1.715420179710465e-05,
+      "loss": 0.8636,
+      "step": 11412
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6169110837831624,
+      "learning_rate": 1.7153668643706704e-05,
+      "loss": 0.5331,
+      "step": 11413
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3673200752329255,
+      "learning_rate": 1.7153135448657834e-05,
+      "loss": 0.9298,
+      "step": 11414
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4763279379641092,
+      "learning_rate": 1.7152602211961146e-05,
+      "loss": 0.9287,
+      "step": 11415
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5172117290478297,
+      "learning_rate": 1.7152068933619742e-05,
+      "loss": 0.7624,
+      "step": 11416
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4956240082580097,
+      "learning_rate": 1.715153561363673e-05,
+      "loss": 0.6921,
+      "step": 11417
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6145896459026665,
+      "learning_rate": 1.7151002252015213e-05,
+      "loss": 0.5354,
+      "step": 11418
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3783912441371995,
+      "learning_rate": 1.7150468848758296e-05,
+      "loss": 0.8,
+      "step": 11419
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4724152907781256,
+      "learning_rate": 1.7149935403869086e-05,
+      "loss": 0.7736,
+      "step": 11420
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6096794833907716,
+      "learning_rate": 1.714940191735069e-05,
+      "loss": 0.5263,
+      "step": 11421
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4658779743426735,
+      "learning_rate": 1.714886838920621e-05,
+      "loss": 0.6112,
+      "step": 11422
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3955020673032432,
+      "learning_rate": 1.7148334819438758e-05,
+      "loss": 0.7978,
+      "step": 11423
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4349196169152147,
+      "learning_rate": 1.7147801208051432e-05,
+      "loss": 0.77,
+      "step": 11424
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2157743986767717,
+      "learning_rate": 1.7147267555047352e-05,
+      "loss": 0.8466,
+      "step": 11425
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5042347186752574,
+      "learning_rate": 1.7146733860429614e-05,
+      "loss": 0.6433,
+      "step": 11426
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4816768190078144,
+      "learning_rate": 1.7146200124201327e-05,
+      "loss": 0.6631,
+      "step": 11427
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6585271494459282,
+      "learning_rate": 1.71456663463656e-05,
+      "loss": 0.5134,
+      "step": 11428
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.742550954248683,
+      "learning_rate": 1.7145132526925543e-05,
+      "loss": 0.6268,
+      "step": 11429
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3989613080693861,
+      "learning_rate": 1.714459866588426e-05,
+      "loss": 0.7545,
+      "step": 11430
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3934927975307494,
+      "learning_rate": 1.7144064763244864e-05,
+      "loss": 0.7077,
+      "step": 11431
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.536207279480192,
+      "learning_rate": 1.7143530819010463e-05,
+      "loss": 0.7392,
+      "step": 11432
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5404023202203057,
+      "learning_rate": 1.7142996833184163e-05,
+      "loss": 0.8171,
+      "step": 11433
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5754776045127081,
+      "learning_rate": 1.7142462805769073e-05,
+      "loss": 0.7365,
+      "step": 11434
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5047824313718199,
+      "learning_rate": 1.71419287367683e-05,
+      "loss": 0.8336,
+      "step": 11435
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4378636236310824,
+      "learning_rate": 1.7141394626184965e-05,
+      "loss": 0.6888,
+      "step": 11436
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.485889070024807,
+      "learning_rate": 1.7140860474022163e-05,
+      "loss": 0.7869,
+      "step": 11437
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5880403453498915,
+      "learning_rate": 1.7140326280283013e-05,
+      "loss": 0.768,
+      "step": 11438
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4348150958144485,
+      "learning_rate": 1.7139792044970622e-05,
+      "loss": 0.6512,
+      "step": 11439
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2239253544882405,
+      "learning_rate": 1.71392577680881e-05,
+      "loss": 0.7466,
+      "step": 11440
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4827008250355864,
+      "learning_rate": 1.7138723449638562e-05,
+      "loss": 0.8394,
+      "step": 11441
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6701287944838379,
+      "learning_rate": 1.7138189089625113e-05,
+      "loss": 0.8011,
+      "step": 11442
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4610833904231495,
+      "learning_rate": 1.7137654688050868e-05,
+      "loss": 0.7995,
+      "step": 11443
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3925022685624693,
+      "learning_rate": 1.713712024491894e-05,
+      "loss": 0.875,
+      "step": 11444
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4344385083522981,
+      "learning_rate": 1.7136585760232436e-05,
+      "loss": 0.615,
+      "step": 11445
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.877430071179558,
+      "learning_rate": 1.713605123399447e-05,
+      "loss": 0.7107,
+      "step": 11446
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5860211904664525,
+      "learning_rate": 1.7135516666208156e-05,
+      "loss": 0.8484,
+      "step": 11447
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5179860962993044,
+      "learning_rate": 1.7134982056876604e-05,
+      "loss": 0.7646,
+      "step": 11448
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8099917004737607,
+      "learning_rate": 1.713444740600293e-05,
+      "loss": 0.833,
+      "step": 11449
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4536136863092544,
+      "learning_rate": 1.7133912713590243e-05,
+      "loss": 0.7813,
+      "step": 11450
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3034224643766934,
+      "learning_rate": 1.7133377979641655e-05,
+      "loss": 0.6398,
+      "step": 11451
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3609330802707755,
+      "learning_rate": 1.7132843204160284e-05,
+      "loss": 0.7933,
+      "step": 11452
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2633387105210463,
+      "learning_rate": 1.713230838714924e-05,
+      "loss": 0.743,
+      "step": 11453
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.280698644527162,
+      "learning_rate": 1.7131773528611643e-05,
+      "loss": 0.7025,
+      "step": 11454
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5784520438343217,
+      "learning_rate": 1.71312386285506e-05,
+      "loss": 0.7012,
+      "step": 11455
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7456711902617523,
+      "learning_rate": 1.713070368696923e-05,
+      "loss": 0.968,
+      "step": 11456
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.432473740826573,
+      "learning_rate": 1.713016870387064e-05,
+      "loss": 0.7968,
+      "step": 11457
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5645948195478636,
+      "learning_rate": 1.7129633679257956e-05,
+      "loss": 0.7312,
+      "step": 11458
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5700798988882132,
+      "learning_rate": 1.7129098613134285e-05,
+      "loss": 0.789,
+      "step": 11459
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5994856973382618,
+      "learning_rate": 1.7128563505502744e-05,
+      "loss": 0.5334,
+      "step": 11460
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4470051462242426,
+      "learning_rate": 1.712802835636645e-05,
+      "loss": 0.7425,
+      "step": 11461
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3919298958456774,
+      "learning_rate": 1.712749316572852e-05,
+      "loss": 0.7392,
+      "step": 11462
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4047451792789214,
+      "learning_rate": 1.7126957933592067e-05,
+      "loss": 0.6262,
+      "step": 11463
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4875308672187535,
+      "learning_rate": 1.7126422659960208e-05,
+      "loss": 0.7567,
+      "step": 11464
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.649999597856627,
+      "learning_rate": 1.712588734483606e-05,
+      "loss": 0.7632,
+      "step": 11465
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4851862002430514,
+      "learning_rate": 1.712535198822274e-05,
+      "loss": 0.6688,
+      "step": 11466
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.511076204479466,
+      "learning_rate": 1.7124816590123368e-05,
+      "loss": 0.7727,
+      "step": 11467
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.960987649149951,
+      "learning_rate": 1.7124281150541054e-05,
+      "loss": 0.7062,
+      "step": 11468
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.610628247192149,
+      "learning_rate": 1.712374566947892e-05,
+      "loss": 0.5488,
+      "step": 11469
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3517957127037945,
+      "learning_rate": 1.7123210146940087e-05,
+      "loss": 0.7782,
+      "step": 11470
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7044559324809843,
+      "learning_rate": 1.7122674582927668e-05,
+      "loss": 0.7953,
+      "step": 11471
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6130223962868029,
+      "learning_rate": 1.712213897744478e-05,
+      "loss": 0.553,
+      "step": 11472
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6904410824682257,
+      "learning_rate": 1.7121603330494547e-05,
+      "loss": 0.7105,
+      "step": 11473
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4646287671623095,
+      "learning_rate": 1.712106764208008e-05,
+      "loss": 0.7624,
+      "step": 11474
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5673583955270396,
+      "learning_rate": 1.7120531912204507e-05,
+      "loss": 0.723,
+      "step": 11475
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4548736750956657,
+      "learning_rate": 1.711999614087094e-05,
+      "loss": 0.8454,
+      "step": 11476
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8560443937179414,
+      "learning_rate": 1.7119460328082506e-05,
+      "loss": 0.746,
+      "step": 11477
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4188247316515608,
+      "learning_rate": 1.7118924473842318e-05,
+      "loss": 0.7967,
+      "step": 11478
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7624554808256554,
+      "learning_rate": 1.7118388578153498e-05,
+      "loss": 0.7838,
+      "step": 11479
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4297539705500901,
+      "learning_rate": 1.7117852641019162e-05,
+      "loss": 0.6353,
+      "step": 11480
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3243599083687645,
+      "learning_rate": 1.711731666244244e-05,
+      "loss": 0.7507,
+      "step": 11481
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4953992367629891,
+      "learning_rate": 1.7116780642426444e-05,
+      "loss": 0.7305,
+      "step": 11482
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2034569308803862,
+      "learning_rate": 1.71162445809743e-05,
+      "loss": 0.6849,
+      "step": 11483
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1237496131707534,
+      "learning_rate": 1.7115708478089124e-05,
+      "loss": 0.7998,
+      "step": 11484
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4033857501236362,
+      "learning_rate": 1.7115172333774043e-05,
+      "loss": 0.7844,
+      "step": 11485
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.555803494508082,
+      "learning_rate": 1.7114636148032176e-05,
+      "loss": 0.6672,
+      "step": 11486
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.7862104107022385,
+      "learning_rate": 1.7114099920866644e-05,
+      "loss": 0.7221,
+      "step": 11487
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4480440646981658,
+      "learning_rate": 1.7113563652280568e-05,
+      "loss": 0.7224,
+      "step": 11488
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3574887381224827,
+      "learning_rate": 1.7113027342277077e-05,
+      "loss": 0.7603,
+      "step": 11489
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9800479589657036,
+      "learning_rate": 1.7112490990859286e-05,
+      "loss": 0.8272,
+      "step": 11490
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5549590141798102,
+      "learning_rate": 1.711195459803032e-05,
+      "loss": 0.6372,
+      "step": 11491
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5681991998246676,
+      "learning_rate": 1.71114181637933e-05,
+      "loss": 0.7895,
+      "step": 11492
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.514971358758323,
+      "learning_rate": 1.711088168815136e-05,
+      "loss": 0.7935,
+      "step": 11493
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.663922195498726,
+      "learning_rate": 1.711034517110761e-05,
+      "loss": 0.6988,
+      "step": 11494
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.686973845706106,
+      "learning_rate": 1.710980861266518e-05,
+      "loss": 0.6534,
+      "step": 11495
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.138272122919336,
+      "learning_rate": 1.7109272012827195e-05,
+      "loss": 0.6371,
+      "step": 11496
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.562363957055666,
+      "learning_rate": 1.7108735371596776e-05,
+      "loss": 0.8152,
+      "step": 11497
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3706590735223085,
+      "learning_rate": 1.7108198688977048e-05,
+      "loss": 0.776,
+      "step": 11498
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.474362316340316,
+      "learning_rate": 1.7107661964971137e-05,
+      "loss": 0.7637,
+      "step": 11499
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5843896476738153,
+      "learning_rate": 1.710712519958217e-05,
+      "loss": 0.7845,
+      "step": 11500
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5431607018920181,
+      "learning_rate": 1.710658839281327e-05,
+      "loss": 0.711,
+      "step": 11501
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3555449701092743,
+      "learning_rate": 1.710605154466756e-05,
+      "loss": 0.8861,
+      "step": 11502
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.456365539722692,
+      "learning_rate": 1.7105514655148173e-05,
+      "loss": 0.7652,
+      "step": 11503
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4105952962217814,
+      "learning_rate": 1.710497772425823e-05,
+      "loss": 0.7323,
+      "step": 11504
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1453890951165686,
+      "learning_rate": 1.710444075200085e-05,
+      "loss": 0.8131,
+      "step": 11505
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7364798848005762,
+      "learning_rate": 1.710390373837917e-05,
+      "loss": 0.6776,
+      "step": 11506
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6080593003136,
+      "learning_rate": 1.7103366683396316e-05,
+      "loss": 0.5186,
+      "step": 11507
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3330868015645214,
+      "learning_rate": 1.7102829587055413e-05,
+      "loss": 0.7045,
+      "step": 11508
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6229946842530024,
+      "learning_rate": 1.710229244935958e-05,
+      "loss": 0.5431,
+      "step": 11509
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.903443918962678,
+      "learning_rate": 1.7101755270311957e-05,
+      "loss": 0.7962,
+      "step": 11510
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.273641352652018,
+      "learning_rate": 1.7101218049915666e-05,
+      "loss": 0.7633,
+      "step": 11511
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.640317919370743,
+      "learning_rate": 1.7100680788173836e-05,
+      "loss": 0.7337,
+      "step": 11512
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6196850325743655,
+      "learning_rate": 1.7100143485089596e-05,
+      "loss": 0.6541,
+      "step": 11513
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3612888253779796,
+      "learning_rate": 1.7099606140666068e-05,
+      "loss": 0.7744,
+      "step": 11514
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8603158943005969,
+      "learning_rate": 1.709906875490639e-05,
+      "loss": 0.7898,
+      "step": 11515
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.414997120858285,
+      "learning_rate": 1.7098531327813684e-05,
+      "loss": 0.7485,
+      "step": 11516
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4871177272875837,
+      "learning_rate": 1.7097993859391082e-05,
+      "loss": 0.7578,
+      "step": 11517
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4354917058573848,
+      "learning_rate": 1.709745634964171e-05,
+      "loss": 0.6767,
+      "step": 11518
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.410252873175777,
+      "learning_rate": 1.7096918798568705e-05,
+      "loss": 0.6536,
+      "step": 11519
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4372120224112461,
+      "learning_rate": 1.7096381206175188e-05,
+      "loss": 0.6607,
+      "step": 11520
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.1930358628063074,
+      "learning_rate": 1.7095843572464294e-05,
+      "loss": 0.8799,
+      "step": 11521
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.5415332344818493,
+      "learning_rate": 1.7095305897439152e-05,
+      "loss": 0.6979,
+      "step": 11522
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.720204166753596,
+      "learning_rate": 1.7094768181102895e-05,
+      "loss": 0.7019,
+      "step": 11523
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4306713365062307,
+      "learning_rate": 1.709423042345865e-05,
+      "loss": 0.8468,
+      "step": 11524
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3816659932913211,
+      "learning_rate": 1.7093692624509545e-05,
+      "loss": 0.6402,
+      "step": 11525
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.002332281335637,
+      "learning_rate": 1.709315478425872e-05,
+      "loss": 0.7387,
+      "step": 11526
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.599870478818685,
+      "learning_rate": 1.7092616902709304e-05,
+      "loss": 0.6967,
+      "step": 11527
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7397351095227955,
+      "learning_rate": 1.7092078979864425e-05,
+      "loss": 0.781,
+      "step": 11528
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3078599450116717,
+      "learning_rate": 1.7091541015727216e-05,
+      "loss": 0.6388,
+      "step": 11529
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.5170898099701806,
+      "learning_rate": 1.709100301030081e-05,
+      "loss": 0.7817,
+      "step": 11530
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.372774570652944,
+      "learning_rate": 1.709046496358834e-05,
+      "loss": 0.8787,
+      "step": 11531
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5923238664819439,
+      "learning_rate": 1.7089926875592938e-05,
+      "loss": 0.5297,
+      "step": 11532
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4687591446774304,
+      "learning_rate": 1.7089388746317738e-05,
+      "loss": 0.7721,
+      "step": 11533
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5915716057704266,
+      "learning_rate": 1.708885057576587e-05,
+      "loss": 0.5596,
+      "step": 11534
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.473039166344788,
+      "learning_rate": 1.7088312363940472e-05,
+      "loss": 0.7251,
+      "step": 11535
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5188207053631815,
+      "learning_rate": 1.7087774110844674e-05,
+      "loss": 0.8765,
+      "step": 11536
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.352082229825315,
+      "learning_rate": 1.7087235816481614e-05,
+      "loss": 0.7514,
+      "step": 11537
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5326327580283978,
+      "learning_rate": 1.7086697480854423e-05,
+      "loss": 0.777,
+      "step": 11538
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.752282380776017,
+      "learning_rate": 1.708615910396623e-05,
+      "loss": 0.5869,
+      "step": 11539
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.343037116669315,
+      "learning_rate": 1.708562068582018e-05,
+      "loss": 0.7301,
+      "step": 11540
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.469736472639783,
+      "learning_rate": 1.7085082226419402e-05,
+      "loss": 0.8554,
+      "step": 11541
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.629776087949399,
+      "learning_rate": 1.7084543725767035e-05,
+      "loss": 0.5931,
+      "step": 11542
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5434536671194532,
+      "learning_rate": 1.7084005183866206e-05,
+      "loss": 0.8407,
+      "step": 11543
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7486349194037543,
+      "learning_rate": 1.708346660072006e-05,
+      "loss": 0.6422,
+      "step": 11544
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3704928954732674,
+      "learning_rate": 1.7082927976331726e-05,
+      "loss": 0.8087,
+      "step": 11545
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.40454000113348,
+      "learning_rate": 1.7082389310704347e-05,
+      "loss": 0.6566,
+      "step": 11546
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5805338252917103,
+      "learning_rate": 1.7081850603841053e-05,
+      "loss": 0.764,
+      "step": 11547
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5846024197994912,
+      "learning_rate": 1.708131185574498e-05,
+      "loss": 0.7784,
+      "step": 11548
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5083724784728936,
+      "learning_rate": 1.7080773066419272e-05,
+      "loss": 0.7217,
+      "step": 11549
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.39969307383026,
+      "learning_rate": 1.708023423586706e-05,
+      "loss": 0.7181,
+      "step": 11550
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0037845448898275,
+      "learning_rate": 1.7079695364091483e-05,
+      "loss": 0.7658,
+      "step": 11551
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4386118902405587,
+      "learning_rate": 1.7079156451095677e-05,
+      "loss": 0.7809,
+      "step": 11552
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.474881079550695,
+      "learning_rate": 1.707861749688278e-05,
+      "loss": 0.7534,
+      "step": 11553
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7845538210714171,
+      "learning_rate": 1.707807850145593e-05,
+      "loss": 0.7461,
+      "step": 11554
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7267477412229317,
+      "learning_rate": 1.707753946481827e-05,
+      "loss": 0.7511,
+      "step": 11555
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4775897433657057,
+      "learning_rate": 1.7077000386972934e-05,
+      "loss": 0.8295,
+      "step": 11556
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.454778121077201,
+      "learning_rate": 1.707646126792306e-05,
+      "loss": 0.7335,
+      "step": 11557
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.767718180479062,
+      "learning_rate": 1.7075922107671786e-05,
+      "loss": 0.8844,
+      "step": 11558
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5943713868033602,
+      "learning_rate": 1.7075382906222254e-05,
+      "loss": 0.7929,
+      "step": 11559
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6063092507387162,
+      "learning_rate": 1.7074843663577607e-05,
+      "loss": 0.5569,
+      "step": 11560
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6840435086160923,
+      "learning_rate": 1.7074304379740975e-05,
+      "loss": 0.722,
+      "step": 11561
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.540760731257221,
+      "learning_rate": 1.7073765054715506e-05,
+      "loss": 0.6666,
+      "step": 11562
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4827320436659681,
+      "learning_rate": 1.7073225688504337e-05,
+      "loss": 0.7619,
+      "step": 11563
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5428486152267509,
+      "learning_rate": 1.707268628111061e-05,
+      "loss": 0.8129,
+      "step": 11564
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.78529154588414,
+      "learning_rate": 1.7072146832537464e-05,
+      "loss": 0.7201,
+      "step": 11565
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7647441507113937,
+      "learning_rate": 1.707160734278804e-05,
+      "loss": 0.8483,
+      "step": 11566
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2760370171797608,
+      "learning_rate": 1.7071067811865477e-05,
+      "loss": 0.9322,
+      "step": 11567
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4189177890602007,
+      "learning_rate": 1.707052823977292e-05,
+      "loss": 0.6595,
+      "step": 11568
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.796580942095878,
+      "learning_rate": 1.7069988626513508e-05,
+      "loss": 0.8438,
+      "step": 11569
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.667930085819511,
+      "learning_rate": 1.7069448972090387e-05,
+      "loss": 0.7101,
+      "step": 11570
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3018892352255156,
+      "learning_rate": 1.7068909276506692e-05,
+      "loss": 0.7955,
+      "step": 11571
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4631784362014804,
+      "learning_rate": 1.7068369539765572e-05,
+      "loss": 0.6675,
+      "step": 11572
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.471328492306765,
+      "learning_rate": 1.7067829761870166e-05,
+      "loss": 0.8325,
+      "step": 11573
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5769663906806244,
+      "learning_rate": 1.7067289942823617e-05,
+      "loss": 0.788,
+      "step": 11574
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6158035788433127,
+      "learning_rate": 1.7066750082629067e-05,
+      "loss": 0.5833,
+      "step": 11575
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6340201624351784,
+      "learning_rate": 1.7066210181289666e-05,
+      "loss": 0.8017,
+      "step": 11576
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4606311618442764,
+      "learning_rate": 1.7065670238808545e-05,
+      "loss": 0.9312,
+      "step": 11577
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7102571060535947,
+      "learning_rate": 1.706513025518886e-05,
+      "loss": 0.7179,
+      "step": 11578
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5991962791014265,
+      "learning_rate": 1.706459023043375e-05,
+      "loss": 0.7355,
+      "step": 11579
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5670848851375223,
+      "learning_rate": 1.7064050164546356e-05,
+      "loss": 0.6788,
+      "step": 11580
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4800207235690603,
+      "learning_rate": 1.7063510057529825e-05,
+      "loss": 0.7751,
+      "step": 11581
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4687072202535347,
+      "learning_rate": 1.706296990938731e-05,
+      "loss": 0.7752,
+      "step": 11582
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5120717479634804,
+      "learning_rate": 1.7062429720121937e-05,
+      "loss": 0.8003,
+      "step": 11583
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7130864761951559,
+      "learning_rate": 1.706188948973687e-05,
+      "loss": 0.8144,
+      "step": 11584
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.600507530055533,
+      "learning_rate": 1.7061349218235243e-05,
+      "loss": 0.5936,
+      "step": 11585
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5390338278564963,
+      "learning_rate": 1.7060808905620207e-05,
+      "loss": 0.7812,
+      "step": 11586
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5835322897685622,
+      "learning_rate": 1.7060268551894904e-05,
+      "loss": 0.8239,
+      "step": 11587
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4940925353642696,
+      "learning_rate": 1.7059728157062485e-05,
+      "loss": 0.7653,
+      "step": 11588
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2977564015974674,
+      "learning_rate": 1.7059187721126093e-05,
+      "loss": 0.7626,
+      "step": 11589
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6989006054576652,
+      "learning_rate": 1.7058647244088873e-05,
+      "loss": 0.6724,
+      "step": 11590
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5057765619064487,
+      "learning_rate": 1.7058106725953977e-05,
+      "loss": 0.8316,
+      "step": 11591
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.716944112688179,
+      "learning_rate": 1.7057566166724545e-05,
+      "loss": 0.7485,
+      "step": 11592
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3616025649248253,
+      "learning_rate": 1.7057025566403733e-05,
+      "loss": 0.7476,
+      "step": 11593
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6012519370988677,
+      "learning_rate": 1.705648492499468e-05,
+      "loss": 0.5491,
+      "step": 11594
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.58686713840727,
+      "learning_rate": 1.7055944242500543e-05,
+      "loss": 0.7912,
+      "step": 11595
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.888727650702102,
+      "learning_rate": 1.705540351892446e-05,
+      "loss": 0.86,
+      "step": 11596
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.324807046430144,
+      "learning_rate": 1.7054862754269586e-05,
+      "loss": 0.7246,
+      "step": 11597
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.6170693850819897,
+      "learning_rate": 1.7054321948539066e-05,
+      "loss": 0.8644,
+      "step": 11598
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6226794587416158,
+      "learning_rate": 1.705378110173605e-05,
+      "loss": 0.6964,
+      "step": 11599
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3495088210972113,
+      "learning_rate": 1.705324021386369e-05,
+      "loss": 0.6432,
+      "step": 11600
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4724702072410738,
+      "learning_rate": 1.705269928492513e-05,
+      "loss": 0.9119,
+      "step": 11601
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.238092117339563,
+      "learning_rate": 1.7052158314923524e-05,
+      "loss": 0.7127,
+      "step": 11602
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2984557537618777,
+      "learning_rate": 1.705161730386202e-05,
+      "loss": 0.706,
+      "step": 11603
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.618385050440361,
+      "learning_rate": 1.7051076251743763e-05,
+      "loss": 0.5822,
+      "step": 11604
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.682283432663127,
+      "learning_rate": 1.705053515857191e-05,
+      "loss": 0.8472,
+      "step": 11605
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5475473941039168,
+      "learning_rate": 1.7049994024349613e-05,
+      "loss": 0.8421,
+      "step": 11606
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3177809279074917,
+      "learning_rate": 1.7049452849080014e-05,
+      "loss": 0.8419,
+      "step": 11607
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5493462864766039,
+      "learning_rate": 1.704891163276627e-05,
+      "loss": 0.7622,
+      "step": 11608
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5981964342975612,
+      "learning_rate": 1.7048370375411533e-05,
+      "loss": 0.8263,
+      "step": 11609
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4879013182828738,
+      "learning_rate": 1.704782907701895e-05,
+      "loss": 0.6971,
+      "step": 11610
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.424013980930104,
+      "learning_rate": 1.7047287737591672e-05,
+      "loss": 0.7102,
+      "step": 11611
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5668024023376732,
+      "learning_rate": 1.7046746357132858e-05,
+      "loss": 0.758,
+      "step": 11612
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3701120710710728,
+      "learning_rate": 1.7046204935645655e-05,
+      "loss": 0.7942,
+      "step": 11613
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5944364726287663,
+      "learning_rate": 1.7045663473133215e-05,
+      "loss": 0.5704,
+      "step": 11614
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2734246155955127,
+      "learning_rate": 1.7045121969598694e-05,
+      "loss": 0.6661,
+      "step": 11615
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3148370340095765,
+      "learning_rate": 1.7044580425045244e-05,
+      "loss": 0.7888,
+      "step": 11616
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3699043496742846,
+      "learning_rate": 1.7044038839476013e-05,
+      "loss": 0.672,
+      "step": 11617
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2752985312780134,
+      "learning_rate": 1.7043497212894154e-05,
+      "loss": 0.7639,
+      "step": 11618
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4065417509872244,
+      "learning_rate": 1.704295554530283e-05,
+      "loss": 0.807,
+      "step": 11619
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.365201076766099,
+      "learning_rate": 1.7042413836705187e-05,
+      "loss": 0.8106,
+      "step": 11620
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4498299547964406,
+      "learning_rate": 1.7041872087104383e-05,
+      "loss": 0.8375,
+      "step": 11621
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.48930622477482,
+      "learning_rate": 1.7041330296503566e-05,
+      "loss": 0.6694,
+      "step": 11622
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5551363704682906,
+      "learning_rate": 1.70407884649059e-05,
+      "loss": 0.6609,
+      "step": 11623
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6796739834055177,
+      "learning_rate": 1.704024659231453e-05,
+      "loss": 0.6988,
+      "step": 11624
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5185139706092265,
+      "learning_rate": 1.703970467873262e-05,
+      "loss": 0.7676,
+      "step": 11625
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6198576209352582,
+      "learning_rate": 1.7039162724163317e-05,
+      "loss": 0.5738,
+      "step": 11626
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3628693068997935,
+      "learning_rate": 1.703862072860978e-05,
+      "loss": 0.7737,
+      "step": 11627
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4243648255166241,
+      "learning_rate": 1.7038078692075166e-05,
+      "loss": 0.6632,
+      "step": 11628
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4986912804739951,
+      "learning_rate": 1.703753661456263e-05,
+      "loss": 0.8524,
+      "step": 11629
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7241346799768364,
+      "learning_rate": 1.703699449607533e-05,
+      "loss": 0.6953,
+      "step": 11630
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.152869001129191,
+      "learning_rate": 1.7036452336616413e-05,
+      "loss": 0.7699,
+      "step": 11631
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6282997521606889,
+      "learning_rate": 1.703591013618905e-05,
+      "loss": 0.5458,
+      "step": 11632
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4275680137383617,
+      "learning_rate": 1.703536789479639e-05,
+      "loss": 0.885,
+      "step": 11633
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2809431459380716,
+      "learning_rate": 1.7034825612441585e-05,
+      "loss": 0.8462,
+      "step": 11634
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5871943033159777,
+      "learning_rate": 1.7034283289127806e-05,
+      "loss": 0.827,
+      "step": 11635
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5762140742281578,
+      "learning_rate": 1.7033740924858194e-05,
+      "loss": 0.8051,
+      "step": 11636
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9685674956301313,
+      "learning_rate": 1.7033198519635922e-05,
+      "loss": 0.7397,
+      "step": 11637
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3902309549900518,
+      "learning_rate": 1.7032656073464137e-05,
+      "loss": 0.6699,
+      "step": 11638
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3870433106867104,
+      "learning_rate": 1.7032113586346003e-05,
+      "loss": 0.7297,
+      "step": 11639
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5555786216816638,
+      "learning_rate": 1.7031571058284678e-05,
+      "loss": 0.4678,
+      "step": 11640
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9047153539670048,
+      "learning_rate": 1.7031028489283324e-05,
+      "loss": 0.578,
+      "step": 11641
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.400300986274225,
+      "learning_rate": 1.703048587934509e-05,
+      "loss": 0.7093,
+      "step": 11642
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9538569946921953,
+      "learning_rate": 1.7029943228473147e-05,
+      "loss": 0.8345,
+      "step": 11643
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5211033418012978,
+      "learning_rate": 1.7029400536670646e-05,
+      "loss": 0.7527,
+      "step": 11644
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0997632556626056,
+      "learning_rate": 1.702885780394075e-05,
+      "loss": 0.7962,
+      "step": 11645
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4166317100195216,
+      "learning_rate": 1.7028315030286617e-05,
+      "loss": 0.7491,
+      "step": 11646
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3820253279053278,
+      "learning_rate": 1.7027772215711414e-05,
+      "loss": 0.5773,
+      "step": 11647
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4796266279989105,
+      "learning_rate": 1.702722936021829e-05,
+      "loss": 0.6849,
+      "step": 11648
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.394911607627807,
+      "learning_rate": 1.7026686463810413e-05,
+      "loss": 0.8006,
+      "step": 11649
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4806678946062821,
+      "learning_rate": 1.7026143526490947e-05,
+      "loss": 0.664,
+      "step": 11650
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.611847745158114,
+      "learning_rate": 1.7025600548263043e-05,
+      "loss": 0.4973,
+      "step": 11651
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.174776748339176,
+      "learning_rate": 1.702505752912987e-05,
+      "loss": 0.7431,
+      "step": 11652
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3989478841732301,
+      "learning_rate": 1.7024514469094592e-05,
+      "loss": 0.6135,
+      "step": 11653
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4337420206915772,
+      "learning_rate": 1.7023971368160362e-05,
+      "loss": 0.7174,
+      "step": 11654
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5618667757946652,
+      "learning_rate": 1.7023428226330348e-05,
+      "loss": 0.7373,
+      "step": 11655
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7894431022115045,
+      "learning_rate": 1.7022885043607714e-05,
+      "loss": 0.7823,
+      "step": 11656
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3620408277337175,
+      "learning_rate": 1.7022341819995622e-05,
+      "loss": 0.7089,
+      "step": 11657
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.451761298169909,
+      "learning_rate": 1.7021798555497225e-05,
+      "loss": 0.7176,
+      "step": 11658
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.619058634245651,
+      "learning_rate": 1.70212552501157e-05,
+      "loss": 0.7614,
+      "step": 11659
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8329585758934743,
+      "learning_rate": 1.7020711903854202e-05,
+      "loss": 0.8098,
+      "step": 11660
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5873739783269967,
+      "learning_rate": 1.7020168516715894e-05,
+      "loss": 0.5792,
+      "step": 11661
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6758658310622825,
+      "learning_rate": 1.7019625088703944e-05,
+      "loss": 0.5907,
+      "step": 11662
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6493182898427909,
+      "learning_rate": 1.701908161982152e-05,
+      "loss": 0.5403,
+      "step": 11663
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4483470963354759,
+      "learning_rate": 1.7018538110071773e-05,
+      "loss": 0.7562,
+      "step": 11664
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4181787955374912,
+      "learning_rate": 1.7017994559457878e-05,
+      "loss": 0.7244,
+      "step": 11665
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6018135800406461,
+      "learning_rate": 1.7017450967982995e-05,
+      "loss": 0.521,
+      "step": 11666
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5471143572625614,
+      "learning_rate": 1.7016907335650292e-05,
+      "loss": 0.659,
+      "step": 11667
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5667533164829217,
+      "learning_rate": 1.701636366246293e-05,
+      "loss": 0.5802,
+      "step": 11668
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5601679934965854,
+      "learning_rate": 1.7015819948424085e-05,
+      "loss": 0.5723,
+      "step": 11669
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4223539953357571,
+      "learning_rate": 1.7015276193536912e-05,
+      "loss": 0.7065,
+      "step": 11670
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4863300955352992,
+      "learning_rate": 1.7014732397804578e-05,
+      "loss": 0.8206,
+      "step": 11671
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4053703668837807,
+      "learning_rate": 1.7014188561230252e-05,
+      "loss": 0.6831,
+      "step": 11672
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5160139158704034,
+      "learning_rate": 1.7013644683817098e-05,
+      "loss": 0.7169,
+      "step": 11673
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4539027334306016,
+      "learning_rate": 1.7013100765568284e-05,
+      "loss": 0.7632,
+      "step": 11674
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5900063113165583,
+      "learning_rate": 1.701255680648698e-05,
+      "loss": 0.7603,
+      "step": 11675
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4117094306156275,
+      "learning_rate": 1.701201280657635e-05,
+      "loss": 0.6359,
+      "step": 11676
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6702593366836567,
+      "learning_rate": 1.7011468765839558e-05,
+      "loss": 0.7023,
+      "step": 11677
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4092971993017838,
+      "learning_rate": 1.701092468427978e-05,
+      "loss": 0.7445,
+      "step": 11678
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3037604346125746,
+      "learning_rate": 1.7010380561900174e-05,
+      "loss": 0.7118,
+      "step": 11679
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6654052889994544,
+      "learning_rate": 1.7009836398703913e-05,
+      "loss": 0.8237,
+      "step": 11680
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5071924214900019,
+      "learning_rate": 1.700929219469417e-05,
+      "loss": 0.7954,
+      "step": 11681
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7436276960036268,
+      "learning_rate": 1.7008747949874105e-05,
+      "loss": 0.6883,
+      "step": 11682
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6417128910887844,
+      "learning_rate": 1.7008203664246892e-05,
+      "loss": 0.6894,
+      "step": 11683
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.452428502338928,
+      "learning_rate": 1.70076593378157e-05,
+      "loss": 0.7772,
+      "step": 11684
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.5236942428790585,
+      "learning_rate": 1.7007114970583694e-05,
+      "loss": 0.7865,
+      "step": 11685
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5770795011029916,
+      "learning_rate": 1.7006570562554047e-05,
+      "loss": 0.5324,
+      "step": 11686
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4599117147309568,
+      "learning_rate": 1.700602611372993e-05,
+      "loss": 0.6536,
+      "step": 11687
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4463830437364664,
+      "learning_rate": 1.7005481624114506e-05,
+      "loss": 0.7991,
+      "step": 11688
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6094640027534513,
+      "learning_rate": 1.7004937093710953e-05,
+      "loss": 0.7643,
+      "step": 11689
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.554736761176149,
+      "learning_rate": 1.700439252252244e-05,
+      "loss": 0.8544,
+      "step": 11690
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3752365935113626,
+      "learning_rate": 1.7003847910552134e-05,
+      "loss": 0.8601,
+      "step": 11691
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.979998771032264,
+      "learning_rate": 1.7003303257803207e-05,
+      "loss": 0.8669,
+      "step": 11692
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.774276289322531,
+      "learning_rate": 1.7002758564278832e-05,
+      "loss": 0.7967,
+      "step": 11693
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5143292522777934,
+      "learning_rate": 1.700221382998218e-05,
+      "loss": 0.8155,
+      "step": 11694
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5709883362088529,
+      "learning_rate": 1.7001669054916424e-05,
+      "loss": 0.5355,
+      "step": 11695
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6052705865770649,
+      "learning_rate": 1.700112423908473e-05,
+      "loss": 0.5293,
+      "step": 11696
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6045965442271155,
+      "learning_rate": 1.700057938249028e-05,
+      "loss": 0.5185,
+      "step": 11697
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3581837862930375,
+      "learning_rate": 1.700003448513624e-05,
+      "loss": 0.8001,
+      "step": 11698
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.744483251598425,
+      "learning_rate": 1.6999489547025782e-05,
+      "loss": 0.6843,
+      "step": 11699
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.694065134842388,
+      "learning_rate": 1.6998944568162075e-05,
+      "loss": 0.6215,
+      "step": 11700
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7739316968493708,
+      "learning_rate": 1.69983995485483e-05,
+      "loss": 0.7863,
+      "step": 11701
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5314858587210745,
+      "learning_rate": 1.6997854488187632e-05,
+      "loss": 0.8407,
+      "step": 11702
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4705096623830975,
+      "learning_rate": 1.699730938708323e-05,
+      "loss": 0.7482,
+      "step": 11703
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5891922807261086,
+      "learning_rate": 1.6996764245238282e-05,
+      "loss": 0.5519,
+      "step": 11704
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0338742704954846,
+      "learning_rate": 1.6996219062655963e-05,
+      "loss": 0.8198,
+      "step": 11705
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6166160163917933,
+      "learning_rate": 1.6995673839339435e-05,
+      "loss": 0.8358,
+      "step": 11706
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3467312226499586,
+      "learning_rate": 1.6995128575291882e-05,
+      "loss": 0.7051,
+      "step": 11707
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.313607015987113,
+      "learning_rate": 1.699458327051647e-05,
+      "loss": 0.7995,
+      "step": 11708
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6887049224695119,
+      "learning_rate": 1.6994037925016386e-05,
+      "loss": 0.7708,
+      "step": 11709
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.527233448879644,
+      "learning_rate": 1.69934925387948e-05,
+      "loss": 0.755,
+      "step": 11710
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5162263334167494,
+      "learning_rate": 1.699294711185488e-05,
+      "loss": 0.8107,
+      "step": 11711
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7236832380854723,
+      "learning_rate": 1.6992401644199815e-05,
+      "loss": 0.7267,
+      "step": 11712
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4803995501226848,
+      "learning_rate": 1.6991856135832767e-05,
+      "loss": 0.7736,
+      "step": 11713
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.298316873140626,
+      "learning_rate": 1.6991310586756923e-05,
+      "loss": 0.843,
+      "step": 11714
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9258496943660042,
+      "learning_rate": 1.6990764996975456e-05,
+      "loss": 0.7089,
+      "step": 11715
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.548973610107872,
+      "learning_rate": 1.699021936649154e-05,
+      "loss": 0.7658,
+      "step": 11716
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.276028238332976,
+      "learning_rate": 1.6989673695308354e-05,
+      "loss": 0.6634,
+      "step": 11717
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4070781540306494,
+      "learning_rate": 1.6989127983429077e-05,
+      "loss": 0.8556,
+      "step": 11718
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7510197858553718,
+      "learning_rate": 1.698858223085688e-05,
+      "loss": 0.7692,
+      "step": 11719
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6622306757144588,
+      "learning_rate": 1.698803643759495e-05,
+      "loss": 0.5466,
+      "step": 11720
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0157857769149254,
+      "learning_rate": 1.698749060364646e-05,
+      "loss": 0.671,
+      "step": 11721
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.564488847302476,
+      "learning_rate": 1.698694472901458e-05,
+      "loss": 0.5497,
+      "step": 11722
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.135471253734256,
+      "learning_rate": 1.69863988137025e-05,
+      "loss": 0.7382,
+      "step": 11723
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.690425790432956,
+      "learning_rate": 1.6985852857713397e-05,
+      "loss": 0.6856,
+      "step": 11724
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5619598146697673,
+      "learning_rate": 1.6985306861050443e-05,
+      "loss": 0.8658,
+      "step": 11725
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6671473409993323,
+      "learning_rate": 1.6984760823716818e-05,
+      "loss": 0.8087,
+      "step": 11726
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3175092773568322,
+      "learning_rate": 1.6984214745715712e-05,
+      "loss": 0.6816,
+      "step": 11727
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.641657372507591,
+      "learning_rate": 1.6983668627050292e-05,
+      "loss": 0.8264,
+      "step": 11728
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2414216103574163,
+      "learning_rate": 1.6983122467723743e-05,
+      "loss": 0.8366,
+      "step": 11729
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.488545100197124,
+      "learning_rate": 1.6982576267739247e-05,
+      "loss": 0.7948,
+      "step": 11730
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6572142143437856,
+      "learning_rate": 1.6982030027099975e-05,
+      "loss": 0.6671,
+      "step": 11731
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.509833613865315,
+      "learning_rate": 1.698148374580912e-05,
+      "loss": 0.7814,
+      "step": 11732
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4857832276803902,
+      "learning_rate": 1.6980937423869852e-05,
+      "loss": 0.72,
+      "step": 11733
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4074944838022716,
+      "learning_rate": 1.698039106128536e-05,
+      "loss": 0.6692,
+      "step": 11734
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0476548274224307,
+      "learning_rate": 1.697984465805882e-05,
+      "loss": 0.7443,
+      "step": 11735
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7301679488413164,
+      "learning_rate": 1.6979298214193412e-05,
+      "loss": 0.7608,
+      "step": 11736
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7449916647763173,
+      "learning_rate": 1.697875172969232e-05,
+      "loss": 0.7302,
+      "step": 11737
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4839232632924049,
+      "learning_rate": 1.6978205204558728e-05,
+      "loss": 0.7672,
+      "step": 11738
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.319683517850875,
+      "learning_rate": 1.6977658638795816e-05,
+      "loss": 0.8268,
+      "step": 11739
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.756272213101259,
+      "learning_rate": 1.697711203240677e-05,
+      "loss": 0.8769,
+      "step": 11740
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.278740184476881,
+      "learning_rate": 1.6976565385394763e-05,
+      "loss": 0.6714,
+      "step": 11741
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4676852955061797,
+      "learning_rate": 1.6976018697762985e-05,
+      "loss": 0.7828,
+      "step": 11742
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.663560151268322,
+      "learning_rate": 1.6975471969514617e-05,
+      "loss": 0.6849,
+      "step": 11743
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.354940197293562,
+      "learning_rate": 1.6974925200652846e-05,
+      "loss": 0.6146,
+      "step": 11744
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9079033173856226,
+      "learning_rate": 1.697437839118085e-05,
+      "loss": 0.81,
+      "step": 11745
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6076175379090148,
+      "learning_rate": 1.6973831541101816e-05,
+      "loss": 0.5496,
+      "step": 11746
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6058688238028644,
+      "learning_rate": 1.6973284650418923e-05,
+      "loss": 0.8398,
+      "step": 11747
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2497414485212697,
+      "learning_rate": 1.6972737719135365e-05,
+      "loss": 0.6993,
+      "step": 11748
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5465526276785608,
+      "learning_rate": 1.697219074725432e-05,
+      "loss": 0.7377,
+      "step": 11749
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5879591127423014,
+      "learning_rate": 1.6971643734778967e-05,
+      "loss": 0.7639,
+      "step": 11750
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4048708621831822,
+      "learning_rate": 1.6971096681712502e-05,
+      "loss": 0.7623,
+      "step": 11751
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5586341624429207,
+      "learning_rate": 1.6970549588058106e-05,
+      "loss": 0.7016,
+      "step": 11752
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6821371941473025,
+      "learning_rate": 1.6970002453818963e-05,
+      "loss": 0.6224,
+      "step": 11753
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.334807434287545,
+      "learning_rate": 1.6969455278998254e-05,
+      "loss": 0.8022,
+      "step": 11754
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4179208186607621,
+      "learning_rate": 1.6968908063599175e-05,
+      "loss": 0.728,
+      "step": 11755
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.068265135915638,
+      "learning_rate": 1.6968360807624904e-05,
+      "loss": 0.7601,
+      "step": 11756
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4444618671740925,
+      "learning_rate": 1.6967813511078633e-05,
+      "loss": 0.683,
+      "step": 11757
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.480923455993554,
+      "learning_rate": 1.6967266173963545e-05,
+      "loss": 0.7189,
+      "step": 11758
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8544105140456149,
+      "learning_rate": 1.6966718796282827e-05,
+      "loss": 0.81,
+      "step": 11759
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0705245130856733,
+      "learning_rate": 1.6966171378039665e-05,
+      "loss": 0.815,
+      "step": 11760
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4562241987793572,
+      "learning_rate": 1.696562391923725e-05,
+      "loss": 0.6375,
+      "step": 11761
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2985716515226358,
+      "learning_rate": 1.6965076419878764e-05,
+      "loss": 0.5922,
+      "step": 11762
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5109904438061519,
+      "learning_rate": 1.69645288799674e-05,
+      "loss": 0.6232,
+      "step": 11763
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3731520054397393,
+      "learning_rate": 1.6963981299506348e-05,
+      "loss": 0.9507,
+      "step": 11764
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5083739382245527,
+      "learning_rate": 1.6963433678498786e-05,
+      "loss": 0.779,
+      "step": 11765
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3175202654593803,
+      "learning_rate": 1.696288601694791e-05,
+      "loss": 0.7278,
+      "step": 11766
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5759523610778079,
+      "learning_rate": 1.696233831485691e-05,
+      "loss": 0.5443,
+      "step": 11767
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6000422118390868,
+      "learning_rate": 1.6961790572228967e-05,
+      "loss": 0.5449,
+      "step": 11768
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6771857822394023,
+      "learning_rate": 1.6961242789067282e-05,
+      "loss": 0.7272,
+      "step": 11769
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.610471965942079,
+      "learning_rate": 1.6960694965375033e-05,
+      "loss": 0.556,
+      "step": 11770
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5711824312196923,
+      "learning_rate": 1.6960147101155415e-05,
+      "loss": 0.7561,
+      "step": 11771
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.694326359959846,
+      "learning_rate": 1.695959919641162e-05,
+      "loss": 0.7312,
+      "step": 11772
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5637668313090299,
+      "learning_rate": 1.695905125114683e-05,
+      "loss": 0.5389,
+      "step": 11773
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4068350512656231,
+      "learning_rate": 1.6958503265364243e-05,
+      "loss": 0.806,
+      "step": 11774
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5273935741447562,
+      "learning_rate": 1.6957955239067046e-05,
+      "loss": 0.6952,
+      "step": 11775
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2269921749927266,
+      "learning_rate": 1.695740717225843e-05,
+      "loss": 0.8214,
+      "step": 11776
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9532429298173184,
+      "learning_rate": 1.695685906494159e-05,
+      "loss": 0.7431,
+      "step": 11777
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.450580710478634,
+      "learning_rate": 1.6956310917119712e-05,
+      "loss": 0.7052,
+      "step": 11778
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.367334372878146,
+      "learning_rate": 1.695576272879599e-05,
+      "loss": 0.7554,
+      "step": 11779
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4167602273086124,
+      "learning_rate": 1.6955214499973616e-05,
+      "loss": 0.8162,
+      "step": 11780
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.389513036041801,
+      "learning_rate": 1.695466623065578e-05,
+      "loss": 0.7297,
+      "step": 11781
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4218249122008222,
+      "learning_rate": 1.6954117920845672e-05,
+      "loss": 0.8907,
+      "step": 11782
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5984791844869901,
+      "learning_rate": 1.6953569570546494e-05,
+      "loss": 0.7834,
+      "step": 11783
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4878367538694683,
+      "learning_rate": 1.695302117976143e-05,
+      "loss": 0.5831,
+      "step": 11784
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5444548108226137,
+      "learning_rate": 1.6952472748493675e-05,
+      "loss": 0.6954,
+      "step": 11785
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5531542446319437,
+      "learning_rate": 1.6951924276746425e-05,
+      "loss": 0.7796,
+      "step": 11786
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5312472245326696,
+      "learning_rate": 1.6951375764522868e-05,
+      "loss": 0.6478,
+      "step": 11787
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5421045351427785,
+      "learning_rate": 1.69508272118262e-05,
+      "loss": 0.6786,
+      "step": 11788
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5947290987431435,
+      "learning_rate": 1.695027861865962e-05,
+      "loss": 0.789,
+      "step": 11789
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4027586996973362,
+      "learning_rate": 1.6949729985026314e-05,
+      "loss": 0.8064,
+      "step": 11790
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5796832579788439,
+      "learning_rate": 1.6949181310929476e-05,
+      "loss": 0.5376,
+      "step": 11791
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3565511754227235,
+      "learning_rate": 1.694863259637231e-05,
+      "loss": 0.5944,
+      "step": 11792
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4161256466454837,
+      "learning_rate": 1.6948083841358004e-05,
+      "loss": 0.7392,
+      "step": 11793
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.586758058171676,
+      "learning_rate": 1.6947535045889754e-05,
+      "loss": 0.5647,
+      "step": 11794
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2984040185903565,
+      "learning_rate": 1.6946986209970754e-05,
+      "loss": 0.8438,
+      "step": 11795
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8895662572721827,
+      "learning_rate": 1.6946437333604203e-05,
+      "loss": 0.7052,
+      "step": 11796
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2908408028959426,
+      "learning_rate": 1.6945888416793296e-05,
+      "loss": 0.6297,
+      "step": 11797
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3574269598562898,
+      "learning_rate": 1.694533945954122e-05,
+      "loss": 0.7311,
+      "step": 11798
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4713459622831215,
+      "learning_rate": 1.6944790461851186e-05,
+      "loss": 0.6861,
+      "step": 11799
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4585988807026098,
+      "learning_rate": 1.694424142372638e-05,
+      "loss": 0.7583,
+      "step": 11800
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5994780464041,
+      "learning_rate": 1.6943692345170005e-05,
+      "loss": 0.5952,
+      "step": 11801
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6456354858389176,
+      "learning_rate": 1.6943143226185252e-05,
+      "loss": 0.5786,
+      "step": 11802
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8781726285627396,
+      "learning_rate": 1.694259406677532e-05,
+      "loss": 0.7824,
+      "step": 11803
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.342553630110343,
+      "learning_rate": 1.6942044866943414e-05,
+      "loss": 0.6766,
+      "step": 11804
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5824235348005207,
+      "learning_rate": 1.6941495626692718e-05,
+      "loss": 0.8078,
+      "step": 11805
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.7364398020954384,
+      "learning_rate": 1.694094634602644e-05,
+      "loss": 0.7382,
+      "step": 11806
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.60475042314243,
+      "learning_rate": 1.6940397024947773e-05,
+      "loss": 0.8658,
+      "step": 11807
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.392986523477705,
+      "learning_rate": 1.693984766345992e-05,
+      "loss": 0.8475,
+      "step": 11808
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9567892230073622,
+      "learning_rate": 1.6939298261566075e-05,
+      "loss": 0.7222,
+      "step": 11809
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6366161922210805,
+      "learning_rate": 1.6938748819269436e-05,
+      "loss": 0.746,
+      "step": 11810
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3815368712458462,
+      "learning_rate": 1.6938199336573208e-05,
+      "loss": 0.7481,
+      "step": 11811
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6152997439542716,
+      "learning_rate": 1.6937649813480586e-05,
+      "loss": 0.8619,
+      "step": 11812
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4410497629733485,
+      "learning_rate": 1.6937100249994772e-05,
+      "loss": 0.7561,
+      "step": 11813
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5927625925423143,
+      "learning_rate": 1.693655064611896e-05,
+      "loss": 0.7308,
+      "step": 11814
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6295398495210485,
+      "learning_rate": 1.693600100185636e-05,
+      "loss": 0.7509,
+      "step": 11815
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5631507308079733,
+      "learning_rate": 1.693545131721016e-05,
+      "loss": 0.5069,
+      "step": 11816
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.875256303068137,
+      "learning_rate": 1.693490159218357e-05,
+      "loss": 0.7545,
+      "step": 11817
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4584972898433193,
+      "learning_rate": 1.6934351826779787e-05,
+      "loss": 0.7005,
+      "step": 11818
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5612741775592807,
+      "learning_rate": 1.6933802021002012e-05,
+      "loss": 0.8901,
+      "step": 11819
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5908410895749399,
+      "learning_rate": 1.6933252174853446e-05,
+      "loss": 0.5563,
+      "step": 11820
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.811187350484187,
+      "learning_rate": 1.6932702288337293e-05,
+      "loss": 0.7021,
+      "step": 11821
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7014634362279641,
+      "learning_rate": 1.693215236145675e-05,
+      "loss": 0.805,
+      "step": 11822
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3955179202265635,
+      "learning_rate": 1.693160239421502e-05,
+      "loss": 0.6577,
+      "step": 11823
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0419324134497225,
+      "learning_rate": 1.693105238661531e-05,
+      "loss": 0.7412,
+      "step": 11824
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5621910039951884,
+      "learning_rate": 1.6930502338660817e-05,
+      "loss": 0.5318,
+      "step": 11825
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.596732827492199,
+      "learning_rate": 1.6929952250354747e-05,
+      "loss": 0.7776,
+      "step": 11826
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6167752250546519,
+      "learning_rate": 1.6929402121700298e-05,
+      "loss": 0.5808,
+      "step": 11827
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8573043062488988,
+      "learning_rate": 1.692885195270068e-05,
+      "loss": 0.8646,
+      "step": 11828
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4740470774342254,
+      "learning_rate": 1.6928301743359087e-05,
+      "loss": 0.711,
+      "step": 11829
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.510457198114129,
+      "learning_rate": 1.692775149367873e-05,
+      "loss": 0.8372,
+      "step": 11830
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7628355353856275,
+      "learning_rate": 1.692720120366281e-05,
+      "loss": 0.7096,
+      "step": 11831
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6899008720479975,
+      "learning_rate": 1.6926650873314535e-05,
+      "loss": 0.733,
+      "step": 11832
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5832478429818901,
+      "learning_rate": 1.69261005026371e-05,
+      "loss": 0.6652,
+      "step": 11833
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5897666960222,
+      "learning_rate": 1.692555009163372e-05,
+      "loss": 0.7033,
+      "step": 11834
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8420546034826213,
+      "learning_rate": 1.6924999640307596e-05,
+      "loss": 0.7104,
+      "step": 11835
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4792544373762846,
+      "learning_rate": 1.6924449148661927e-05,
+      "loss": 0.8084,
+      "step": 11836
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5246168002603209,
+      "learning_rate": 1.6923898616699925e-05,
+      "loss": 0.8247,
+      "step": 11837
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2230943763216515,
+      "learning_rate": 1.6923348044424793e-05,
+      "loss": 0.7421,
+      "step": 11838
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.184328244799249,
+      "learning_rate": 1.6922797431839736e-05,
+      "loss": 0.7555,
+      "step": 11839
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6684715993500314,
+      "learning_rate": 1.692224677894796e-05,
+      "loss": 0.8727,
+      "step": 11840
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3121082129253983,
+      "learning_rate": 1.6921696085752674e-05,
+      "loss": 0.7136,
+      "step": 11841
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.397925380342564,
+      "learning_rate": 1.692114535225708e-05,
+      "loss": 0.609,
+      "step": 11842
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8724413582659631,
+      "learning_rate": 1.6920594578464386e-05,
+      "loss": 0.7634,
+      "step": 11843
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4243182733944133,
+      "learning_rate": 1.6920043764377804e-05,
+      "loss": 0.8948,
+      "step": 11844
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4450089107714184,
+      "learning_rate": 1.6919492910000534e-05,
+      "loss": 0.6489,
+      "step": 11845
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4104534466269256,
+      "learning_rate": 1.6918942015335785e-05,
+      "loss": 0.7731,
+      "step": 11846
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3881918022266635,
+      "learning_rate": 1.6918391080386763e-05,
+      "loss": 0.7858,
+      "step": 11847
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0852951852392865,
+      "learning_rate": 1.691784010515668e-05,
+      "loss": 0.73,
+      "step": 11848
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.595369539812452,
+      "learning_rate": 1.6917289089648742e-05,
+      "loss": 0.5106,
+      "step": 11849
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7410089157190178,
+      "learning_rate": 1.6916738033866153e-05,
+      "loss": 0.7148,
+      "step": 11850
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8845113739592354,
+      "learning_rate": 1.6916186937812132e-05,
+      "loss": 0.7676,
+      "step": 11851
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4412637633643886,
+      "learning_rate": 1.6915635801489876e-05,
+      "loss": 0.7452,
+      "step": 11852
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.447266164266186,
+      "learning_rate": 1.6915084624902602e-05,
+      "loss": 0.6982,
+      "step": 11853
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4288119692498082,
+      "learning_rate": 1.6914533408053513e-05,
+      "loss": 0.7294,
+      "step": 11854
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.588140124352993,
+      "learning_rate": 1.6913982150945823e-05,
+      "loss": 0.6302,
+      "step": 11855
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4006703245753485,
+      "learning_rate": 1.691343085358274e-05,
+      "loss": 0.8077,
+      "step": 11856
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3267286245017018,
+      "learning_rate": 1.691287951596747e-05,
+      "loss": 0.764,
+      "step": 11857
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5720428636446107,
+      "learning_rate": 1.6912328138103237e-05,
+      "loss": 0.8327,
+      "step": 11858
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.27916081629561,
+      "learning_rate": 1.6911776719993232e-05,
+      "loss": 0.8166,
+      "step": 11859
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5906731913258052,
+      "learning_rate": 1.6911225261640678e-05,
+      "loss": 0.8869,
+      "step": 11860
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5188032301530232,
+      "learning_rate": 1.691067376304878e-05,
+      "loss": 0.8175,
+      "step": 11861
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6004395369970712,
+      "learning_rate": 1.6910122224220757e-05,
+      "loss": 0.5168,
+      "step": 11862
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.748824542497566,
+      "learning_rate": 1.6909570645159808e-05,
+      "loss": 0.7236,
+      "step": 11863
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6590773765775106,
+      "learning_rate": 1.6909019025869158e-05,
+      "loss": 0.7699,
+      "step": 11864
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5151369395541707,
+      "learning_rate": 1.6908467366352004e-05,
+      "loss": 0.8426,
+      "step": 11865
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.561674695756764,
+      "learning_rate": 1.690791566661157e-05,
+      "loss": 0.7659,
+      "step": 11866
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7223564250891585,
+      "learning_rate": 1.690736392665107e-05,
+      "loss": 0.7319,
+      "step": 11867
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6866639960306131,
+      "learning_rate": 1.6906812146473703e-05,
+      "loss": 0.6902,
+      "step": 11868
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3554995428676313,
+      "learning_rate": 1.6906260326082692e-05,
+      "loss": 0.7007,
+      "step": 11869
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4860191073324718,
+      "learning_rate": 1.6905708465481242e-05,
+      "loss": 0.7853,
+      "step": 11870
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3929540603248278,
+      "learning_rate": 1.6905156564672575e-05,
+      "loss": 0.6368,
+      "step": 11871
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3423916567547052,
+      "learning_rate": 1.6904604623659898e-05,
+      "loss": 0.7782,
+      "step": 11872
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4572933587632801,
+      "learning_rate": 1.690405264244643e-05,
+      "loss": 0.8423,
+      "step": 11873
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0582483603468043,
+      "learning_rate": 1.690350062103538e-05,
+      "loss": 0.7528,
+      "step": 11874
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.239561909098904,
+      "learning_rate": 1.6902948559429962e-05,
+      "loss": 0.6564,
+      "step": 11875
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4269236794946343,
+      "learning_rate": 1.6902396457633397e-05,
+      "loss": 0.7308,
+      "step": 11876
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3721531809547773,
+      "learning_rate": 1.6901844315648893e-05,
+      "loss": 0.8849,
+      "step": 11877
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2993231662345877,
+      "learning_rate": 1.690129213347966e-05,
+      "loss": 0.7352,
+      "step": 11878
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4872935871403845,
+      "learning_rate": 1.690073991112893e-05,
+      "loss": 0.6986,
+      "step": 11879
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0202362970589256,
+      "learning_rate": 1.69001876485999e-05,
+      "loss": 0.74,
+      "step": 11880
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.535712361818195,
+      "learning_rate": 1.6899635345895796e-05,
+      "loss": 0.6815,
+      "step": 11881
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.422862894762392,
+      "learning_rate": 1.689908300301983e-05,
+      "loss": 0.756,
+      "step": 11882
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.863666099994029,
+      "learning_rate": 1.6898530619975218e-05,
+      "loss": 0.6355,
+      "step": 11883
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9697551029888956,
+      "learning_rate": 1.6897978196765174e-05,
+      "loss": 0.6461,
+      "step": 11884
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.251261056319089,
+      "learning_rate": 1.689742573339292e-05,
+      "loss": 0.6769,
+      "step": 11885
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.663308410203845,
+      "learning_rate": 1.689687322986167e-05,
+      "loss": 0.8296,
+      "step": 11886
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6042067498943928,
+      "learning_rate": 1.689632068617464e-05,
+      "loss": 0.8033,
+      "step": 11887
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5406185309151592,
+      "learning_rate": 1.6895768102335045e-05,
+      "loss": 0.8319,
+      "step": 11888
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6353946275979392,
+      "learning_rate": 1.689521547834611e-05,
+      "loss": 0.9824,
+      "step": 11889
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4539583100306912,
+      "learning_rate": 1.6894662814211046e-05,
+      "loss": 0.8009,
+      "step": 11890
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6066671448670304,
+      "learning_rate": 1.689411010993307e-05,
+      "loss": 0.5289,
+      "step": 11891
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.422126366746332,
+      "learning_rate": 1.6893557365515404e-05,
+      "loss": 0.8859,
+      "step": 11892
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.65412323944858,
+      "learning_rate": 1.6893004580961264e-05,
+      "loss": 0.6516,
+      "step": 11893
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1530970750079532,
+      "learning_rate": 1.689245175627387e-05,
+      "loss": 0.6287,
+      "step": 11894
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.293911537582068,
+      "learning_rate": 1.689189889145644e-05,
+      "loss": 0.7001,
+      "step": 11895
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5947352679319976,
+      "learning_rate": 1.689134598651219e-05,
+      "loss": 0.5296,
+      "step": 11896
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.154167117481628,
+      "learning_rate": 1.6890793041444345e-05,
+      "loss": 0.5921,
+      "step": 11897
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3729486457766495,
+      "learning_rate": 1.689024005625612e-05,
+      "loss": 0.6071,
+      "step": 11898
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9097036093660464,
+      "learning_rate": 1.6889687030950736e-05,
+      "loss": 0.7646,
+      "step": 11899
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8777649642423375,
+      "learning_rate": 1.688913396553141e-05,
+      "loss": 0.7024,
+      "step": 11900
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4419889574070108,
+      "learning_rate": 1.688858086000137e-05,
+      "loss": 0.7848,
+      "step": 11901
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6539690915826966,
+      "learning_rate": 1.688802771436383e-05,
+      "loss": 0.7763,
+      "step": 11902
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4816206806837275,
+      "learning_rate": 1.688747452862201e-05,
+      "loss": 0.6354,
+      "step": 11903
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5318299823980928,
+      "learning_rate": 1.6886921302779134e-05,
+      "loss": 0.7533,
+      "step": 11904
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4629515163928934,
+      "learning_rate": 1.6886368036838422e-05,
+      "loss": 0.7951,
+      "step": 11905
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.90946403350004,
+      "learning_rate": 1.6885814730803095e-05,
+      "loss": 0.7134,
+      "step": 11906
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5062614517879023,
+      "learning_rate": 1.6885261384676373e-05,
+      "loss": 0.712,
+      "step": 11907
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4217398134987675,
+      "learning_rate": 1.688470799846148e-05,
+      "loss": 0.7953,
+      "step": 11908
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5408843325563573,
+      "learning_rate": 1.6884154572161636e-05,
+      "loss": 0.6737,
+      "step": 11909
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2581057874249924,
+      "learning_rate": 1.6883601105780068e-05,
+      "loss": 0.6839,
+      "step": 11910
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5919640076710766,
+      "learning_rate": 1.6883047599319994e-05,
+      "loss": 0.5175,
+      "step": 11911
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.5071550458735983,
+      "learning_rate": 1.6882494052784637e-05,
+      "loss": 0.7154,
+      "step": 11912
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5127607985595153,
+      "learning_rate": 1.688194046617722e-05,
+      "loss": 0.8244,
+      "step": 11913
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.450300888160451,
+      "learning_rate": 1.688138683950097e-05,
+      "loss": 0.7415,
+      "step": 11914
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6130544119323045,
+      "learning_rate": 1.6880833172759105e-05,
+      "loss": 0.89,
+      "step": 11915
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3865577662866448,
+      "learning_rate": 1.688027946595485e-05,
+      "loss": 0.7333,
+      "step": 11916
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4445001348221935,
+      "learning_rate": 1.687972571909143e-05,
+      "loss": 0.8208,
+      "step": 11917
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7569036901086164,
+      "learning_rate": 1.6879171932172074e-05,
+      "loss": 0.752,
+      "step": 11918
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3433195546693744,
+      "learning_rate": 1.6878618105199995e-05,
+      "loss": 0.6878,
+      "step": 11919
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.12338921389602,
+      "learning_rate": 1.6878064238178423e-05,
+      "loss": 0.8027,
+      "step": 11920
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.390967507765086,
+      "learning_rate": 1.6877510331110586e-05,
+      "loss": 0.7178,
+      "step": 11921
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.353120012247806,
+      "learning_rate": 1.6876956383999707e-05,
+      "loss": 0.6908,
+      "step": 11922
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6952639081963525,
+      "learning_rate": 1.687640239684901e-05,
+      "loss": 0.7799,
+      "step": 11923
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3542478666534086,
+      "learning_rate": 1.6875848369661722e-05,
+      "loss": 0.7345,
+      "step": 11924
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4524600683720386,
+      "learning_rate": 1.6875294302441067e-05,
+      "loss": 0.6874,
+      "step": 11925
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.357069428810661,
+      "learning_rate": 1.687474019519027e-05,
+      "loss": 0.6157,
+      "step": 11926
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.2314480261989056,
+      "learning_rate": 1.6874186047912565e-05,
+      "loss": 0.8618,
+      "step": 11927
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3171014923702908,
+      "learning_rate": 1.6873631860611165e-05,
+      "loss": 0.7931,
+      "step": 11928
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4788268841473458,
+      "learning_rate": 1.687307763328931e-05,
+      "loss": 0.7607,
+      "step": 11929
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.641404188663842,
+      "learning_rate": 1.6872523365950218e-05,
+      "loss": 0.7566,
+      "step": 11930
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9983508478378753,
+      "learning_rate": 1.6871969058597118e-05,
+      "loss": 0.7235,
+      "step": 11931
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.42912978039091,
+      "learning_rate": 1.687141471123324e-05,
+      "loss": 0.8506,
+      "step": 11932
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3745986877402077,
+      "learning_rate": 1.6870860323861813e-05,
+      "loss": 0.6452,
+      "step": 11933
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5889793401831889,
+      "learning_rate": 1.6870305896486058e-05,
+      "loss": 0.4906,
+      "step": 11934
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1992128574364282,
+      "learning_rate": 1.686975142910921e-05,
+      "loss": 0.7883,
+      "step": 11935
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5153734612653063,
+      "learning_rate": 1.686919692173449e-05,
+      "loss": 0.6054,
+      "step": 11936
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3529313922709296,
+      "learning_rate": 1.6868642374365135e-05,
+      "loss": 0.7892,
+      "step": 11937
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7698456668964409,
+      "learning_rate": 1.686808778700437e-05,
+      "loss": 0.7594,
+      "step": 11938
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.829334014531702,
+      "learning_rate": 1.686753315965542e-05,
+      "loss": 0.7316,
+      "step": 11939
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.320276897714162,
+      "learning_rate": 1.686697849232152e-05,
+      "loss": 0.7065,
+      "step": 11940
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4559965711319307,
+      "learning_rate": 1.6866423785005895e-05,
+      "loss": 0.7931,
+      "step": 11941
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6317852386727254,
+      "learning_rate": 1.6865869037711776e-05,
+      "loss": 0.7417,
+      "step": 11942
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5795565851504274,
+      "learning_rate": 1.68653142504424e-05,
+      "loss": 0.7946,
+      "step": 11943
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7734023659140183,
+      "learning_rate": 1.6864759423200984e-05,
+      "loss": 0.773,
+      "step": 11944
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6136908828622947,
+      "learning_rate": 1.6864204555990768e-05,
+      "loss": 0.7621,
+      "step": 11945
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3770742305316526,
+      "learning_rate": 1.6863649648814982e-05,
+      "loss": 0.824,
+      "step": 11946
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6684536100584986,
+      "learning_rate": 1.686309470167685e-05,
+      "loss": 0.7476,
+      "step": 11947
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5402940477688416,
+      "learning_rate": 1.686253971457961e-05,
+      "loss": 0.6645,
+      "step": 11948
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6416948005486085,
+      "learning_rate": 1.6861984687526494e-05,
+      "loss": 0.7787,
+      "step": 11949
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4762725589562424,
+      "learning_rate": 1.6861429620520725e-05,
+      "loss": 0.9075,
+      "step": 11950
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5997057665892271,
+      "learning_rate": 1.6860874513565544e-05,
+      "loss": 0.5033,
+      "step": 11951
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.180565879047667,
+      "learning_rate": 1.6860319366664182e-05,
+      "loss": 0.8181,
+      "step": 11952
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.386506390219948,
+      "learning_rate": 1.6859764179819866e-05,
+      "loss": 0.75,
+      "step": 11953
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5828928620835154,
+      "learning_rate": 1.6859208953035832e-05,
+      "loss": 0.819,
+      "step": 11954
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.305601742662539,
+      "learning_rate": 1.685865368631531e-05,
+      "loss": 0.7446,
+      "step": 11955
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4676986340449532,
+      "learning_rate": 1.6858098379661538e-05,
+      "loss": 0.8364,
+      "step": 11956
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5623245538618504,
+      "learning_rate": 1.685754303307774e-05,
+      "loss": 0.7386,
+      "step": 11957
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5616585278629007,
+      "learning_rate": 1.6856987646567163e-05,
+      "loss": 0.5415,
+      "step": 11958
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6037475506104707,
+      "learning_rate": 1.685643222013303e-05,
+      "loss": 0.5268,
+      "step": 11959
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4598168383925993,
+      "learning_rate": 1.6855876753778574e-05,
+      "loss": 0.7539,
+      "step": 11960
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3442183875275684,
+      "learning_rate": 1.6855321247507034e-05,
+      "loss": 0.7542,
+      "step": 11961
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5962121212867075,
+      "learning_rate": 1.6854765701321648e-05,
+      "loss": 0.5113,
+      "step": 11962
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.470109274353551,
+      "learning_rate": 1.6854210115225642e-05,
+      "loss": 0.765,
+      "step": 11963
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3972096872507103,
+      "learning_rate": 1.6853654489222258e-05,
+      "loss": 0.7013,
+      "step": 11964
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.021887989348197,
+      "learning_rate": 1.6853098823314728e-05,
+      "loss": 0.8191,
+      "step": 11965
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.606070640462235,
+      "learning_rate": 1.685254311750628e-05,
+      "loss": 0.9217,
+      "step": 11966
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.451713974116606,
+      "learning_rate": 1.6851987371800164e-05,
+      "loss": 0.7827,
+      "step": 11967
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7291283435785134,
+      "learning_rate": 1.6851431586199604e-05,
+      "loss": 0.778,
+      "step": 11968
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4838626652369615,
+      "learning_rate": 1.6850875760707842e-05,
+      "loss": 0.6945,
+      "step": 11969
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6178283344839637,
+      "learning_rate": 1.6850319895328112e-05,
+      "loss": 0.7397,
+      "step": 11970
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9238409785701183,
+      "learning_rate": 1.6849763990063654e-05,
+      "loss": 0.6888,
+      "step": 11971
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3391856218423857,
+      "learning_rate": 1.6849208044917696e-05,
+      "loss": 0.6496,
+      "step": 11972
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.295643651241791,
+      "learning_rate": 1.6848652059893485e-05,
+      "loss": 0.7901,
+      "step": 11973
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5171246824758737,
+      "learning_rate": 1.6848096034994253e-05,
+      "loss": 0.627,
+      "step": 11974
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5026363287077213,
+      "learning_rate": 1.684753997022323e-05,
+      "loss": 0.7475,
+      "step": 11975
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.689638990189755,
+      "learning_rate": 1.6846983865583674e-05,
+      "loss": 0.6375,
+      "step": 11976
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6132516220743143,
+      "learning_rate": 1.6846427721078798e-05,
+      "loss": 0.5834,
+      "step": 11977
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.431535135945947,
+      "learning_rate": 1.684587153671186e-05,
+      "loss": 0.7889,
+      "step": 11978
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6314967406801775,
+      "learning_rate": 1.6845315312486088e-05,
+      "loss": 0.8044,
+      "step": 11979
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.615809099279836,
+      "learning_rate": 1.6844759048404724e-05,
+      "loss": 0.8406,
+      "step": 11980
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4532960898976957,
+      "learning_rate": 1.6844202744471006e-05,
+      "loss": 0.7359,
+      "step": 11981
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.447717695332161,
+      "learning_rate": 1.684364640068817e-05,
+      "loss": 0.703,
+      "step": 11982
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.289100871962224,
+      "learning_rate": 1.684309001705946e-05,
+      "loss": 0.7332,
+      "step": 11983
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3747121109443161,
+      "learning_rate": 1.6842533593588116e-05,
+      "loss": 0.7879,
+      "step": 11984
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4098097555568923,
+      "learning_rate": 1.684197713027737e-05,
+      "loss": 0.8261,
+      "step": 11985
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6513139560989525,
+      "learning_rate": 1.684142062713047e-05,
+      "loss": 0.7612,
+      "step": 11986
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9139352092045834,
+      "learning_rate": 1.6840864084150658e-05,
+      "loss": 0.836,
+      "step": 11987
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6490440124700276,
+      "learning_rate": 1.684030750134116e-05,
+      "loss": 0.7746,
+      "step": 11988
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.564572826731781,
+      "learning_rate": 1.683975087870523e-05,
+      "loss": 0.7764,
+      "step": 11989
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.594588914451415,
+      "learning_rate": 1.683919421624611e-05,
+      "loss": 0.6935,
+      "step": 11990
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4399823048847913,
+      "learning_rate": 1.683863751396703e-05,
+      "loss": 0.7564,
+      "step": 11991
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5544846688048417,
+      "learning_rate": 1.683808077187124e-05,
+      "loss": 0.7932,
+      "step": 11992
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7488784854524586,
+      "learning_rate": 1.6837523989961973e-05,
+      "loss": 0.645,
+      "step": 11993
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3781305451180876,
+      "learning_rate": 1.6836967168242483e-05,
+      "loss": 0.7791,
+      "step": 11994
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9248573526921449,
+      "learning_rate": 1.6836410306716005e-05,
+      "loss": 0.8293,
+      "step": 11995
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3038266210331573,
+      "learning_rate": 1.6835853405385778e-05,
+      "loss": 0.8395,
+      "step": 11996
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8609367685599065,
+      "learning_rate": 1.6835296464255052e-05,
+      "loss": 0.7833,
+      "step": 11997
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3190933780132932,
+      "learning_rate": 1.6834739483327066e-05,
+      "loss": 0.8855,
+      "step": 11998
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7753471409283585,
+      "learning_rate": 1.6834182462605062e-05,
+      "loss": 0.6981,
+      "step": 11999
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.808119844011668,
+      "learning_rate": 1.6833625402092286e-05,
+      "loss": 0.7009,
+      "step": 12000
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4718906878561329,
+      "learning_rate": 1.6833068301791977e-05,
+      "loss": 0.7729,
+      "step": 12001
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3617867759864295,
+      "learning_rate": 1.683251116170738e-05,
+      "loss": 0.7505,
+      "step": 12002
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5919814301891021,
+      "learning_rate": 1.683195398184174e-05,
+      "loss": 0.5657,
+      "step": 12003
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3371189306943956,
+      "learning_rate": 1.6831396762198303e-05,
+      "loss": 0.7716,
+      "step": 12004
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4969138550014736,
+      "learning_rate": 1.683083950278031e-05,
+      "loss": 0.7782,
+      "step": 12005
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5843914623481388,
+      "learning_rate": 1.6830282203591007e-05,
+      "loss": 0.7631,
+      "step": 12006
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.673089664462347,
+      "learning_rate": 1.682972486463364e-05,
+      "loss": 0.6453,
+      "step": 12007
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5353331372759347,
+      "learning_rate": 1.682916748591145e-05,
+      "loss": 0.7813,
+      "step": 12008
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9766191518127454,
+      "learning_rate": 1.6828610067427686e-05,
+      "loss": 0.7836,
+      "step": 12009
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4461287240603096,
+      "learning_rate": 1.6828052609185592e-05,
+      "loss": 0.7463,
+      "step": 12010
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.149336901148863,
+      "learning_rate": 1.6827495111188413e-05,
+      "loss": 0.7072,
+      "step": 12011
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9255615156408548,
+      "learning_rate": 1.6826937573439395e-05,
+      "loss": 0.6632,
+      "step": 12012
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4179251135665512,
+      "learning_rate": 1.682637999594179e-05,
+      "loss": 0.7518,
+      "step": 12013
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5983852107344036,
+      "learning_rate": 1.6825822378698836e-05,
+      "loss": 0.5397,
+      "step": 12014
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.608093121420355,
+      "learning_rate": 1.6825264721713787e-05,
+      "loss": 0.7991,
+      "step": 12015
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4735872049981253,
+      "learning_rate": 1.6824707024989885e-05,
+      "loss": 0.8016,
+      "step": 12016
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8695222990191716,
+      "learning_rate": 1.6824149288530374e-05,
+      "loss": 0.7428,
+      "step": 12017
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7963344278093905,
+      "learning_rate": 1.6823591512338508e-05,
+      "loss": 0.8048,
+      "step": 12018
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4761174699813404,
+      "learning_rate": 1.682303369641753e-05,
+      "loss": 0.739,
+      "step": 12019
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0619510226642794,
+      "learning_rate": 1.6822475840770692e-05,
+      "loss": 0.7383,
+      "step": 12020
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6843437829726717,
+      "learning_rate": 1.6821917945401237e-05,
+      "loss": 0.7495,
+      "step": 12021
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4728409930579682,
+      "learning_rate": 1.6821360010312418e-05,
+      "loss": 0.7846,
+      "step": 12022
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7832254024885352,
+      "learning_rate": 1.682080203550748e-05,
+      "loss": 0.7243,
+      "step": 12023
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7959675128095625,
+      "learning_rate": 1.6820244020989672e-05,
+      "loss": 0.7588,
+      "step": 12024
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5374465837569415,
+      "learning_rate": 1.681968596676225e-05,
+      "loss": 0.7843,
+      "step": 12025
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.708301496133165,
+      "learning_rate": 1.681912787282845e-05,
+      "loss": 0.7625,
+      "step": 12026
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.49687576240963,
+      "learning_rate": 1.6818569739191535e-05,
+      "loss": 0.671,
+      "step": 12027
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5250394574506365,
+      "learning_rate": 1.6818011565854744e-05,
+      "loss": 0.7515,
+      "step": 12028
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4446967621015114,
+      "learning_rate": 1.6817453352821333e-05,
+      "loss": 0.7792,
+      "step": 12029
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6325488696540125,
+      "learning_rate": 1.681689510009455e-05,
+      "loss": 0.785,
+      "step": 12030
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5081138099776177,
+      "learning_rate": 1.6816336807677644e-05,
+      "loss": 0.7238,
+      "step": 12031
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6374821205056123,
+      "learning_rate": 1.681577847557387e-05,
+      "loss": 0.6006,
+      "step": 12032
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.100716936707162,
+      "learning_rate": 1.6815220103786472e-05,
+      "loss": 0.8076,
+      "step": 12033
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4376526549792197,
+      "learning_rate": 1.6814661692318707e-05,
+      "loss": 0.7573,
+      "step": 12034
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5115493544567737,
+      "learning_rate": 1.6814103241173825e-05,
+      "loss": 0.6955,
+      "step": 12035
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3894052160912227,
+      "learning_rate": 1.6813544750355074e-05,
+      "loss": 0.798,
+      "step": 12036
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.611131109002109,
+      "learning_rate": 1.6812986219865712e-05,
+      "loss": 0.7516,
+      "step": 12037
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.540876186174823,
+      "learning_rate": 1.6812427649708984e-05,
+      "loss": 0.95,
+      "step": 12038
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8560291343926476,
+      "learning_rate": 1.681186903988815e-05,
+      "loss": 0.8479,
+      "step": 12039
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.623452402762898,
+      "learning_rate": 1.6811310390406454e-05,
+      "loss": 0.6766,
+      "step": 12040
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3135501473250377,
+      "learning_rate": 1.681075170126715e-05,
+      "loss": 0.6872,
+      "step": 12041
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8803291514082945,
+      "learning_rate": 1.6810192972473497e-05,
+      "loss": 0.7319,
+      "step": 12042
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6649151694529536,
+      "learning_rate": 1.6809634204028745e-05,
+      "loss": 0.7193,
+      "step": 12043
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.5952745220994657,
+      "learning_rate": 1.6809075395936142e-05,
+      "loss": 0.6401,
+      "step": 12044
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6059888151621413,
+      "learning_rate": 1.6808516548198955e-05,
+      "loss": 0.8436,
+      "step": 12045
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3911616216285436,
+      "learning_rate": 1.6807957660820423e-05,
+      "loss": 0.7857,
+      "step": 12046
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5158851165631553,
+      "learning_rate": 1.6807398733803804e-05,
+      "loss": 0.7866,
+      "step": 12047
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3264428639051256,
+      "learning_rate": 1.680683976715236e-05,
+      "loss": 0.7148,
+      "step": 12048
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.361751734080845,
+      "learning_rate": 1.6806280760869338e-05,
+      "loss": 0.7763,
+      "step": 12049
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7093031552036768,
+      "learning_rate": 1.6805721714957995e-05,
+      "loss": 0.7418,
+      "step": 12050
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.5156084075659217,
+      "learning_rate": 1.6805162629421583e-05,
+      "loss": 0.8853,
+      "step": 12051
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2656333908666437,
+      "learning_rate": 1.6804603504263362e-05,
+      "loss": 0.7886,
+      "step": 12052
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2003867576318343,
+      "learning_rate": 1.6804044339486584e-05,
+      "loss": 0.7675,
+      "step": 12053
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8328304489489402,
+      "learning_rate": 1.6803485135094506e-05,
+      "loss": 0.7854,
+      "step": 12054
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3076598984993504,
+      "learning_rate": 1.6802925891090383e-05,
+      "loss": 0.7067,
+      "step": 12055
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5934895193617534,
+      "learning_rate": 1.680236660747747e-05,
+      "loss": 0.557,
+      "step": 12056
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.585649520291515,
+      "learning_rate": 1.680180728425903e-05,
+      "loss": 0.6897,
+      "step": 12057
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4527473124110144,
+      "learning_rate": 1.6801247921438313e-05,
+      "loss": 0.8801,
+      "step": 12058
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.717961846869724,
+      "learning_rate": 1.6800688519018575e-05,
+      "loss": 0.7446,
+      "step": 12059
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4845211850648832,
+      "learning_rate": 1.680012907700308e-05,
+      "loss": 0.7258,
+      "step": 12060
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6266523021344754,
+      "learning_rate": 1.6799569595395074e-05,
+      "loss": 0.6739,
+      "step": 12061
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7792646952588398,
+      "learning_rate": 1.6799010074197827e-05,
+      "loss": 0.7925,
+      "step": 12062
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5090261131160796,
+      "learning_rate": 1.6798450513414586e-05,
+      "loss": 0.8002,
+      "step": 12063
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3203376551027335,
+      "learning_rate": 1.6797890913048615e-05,
+      "loss": 0.7355,
+      "step": 12064
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0337858786054763,
+      "learning_rate": 1.6797331273103172e-05,
+      "loss": 0.7658,
+      "step": 12065
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6047136127093704,
+      "learning_rate": 1.6796771593581515e-05,
+      "loss": 0.7552,
+      "step": 12066
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7191993595096304,
+      "learning_rate": 1.67962118744869e-05,
+      "loss": 0.7943,
+      "step": 12067
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4634788152694578,
+      "learning_rate": 1.6795652115822592e-05,
+      "loss": 0.7207,
+      "step": 12068
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3454917950826846,
+      "learning_rate": 1.6795092317591845e-05,
+      "loss": 0.78,
+      "step": 12069
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4650795441631157,
+      "learning_rate": 1.6794532479797917e-05,
+      "loss": 0.7492,
+      "step": 12070
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.359465596897487,
+      "learning_rate": 1.6793972602444068e-05,
+      "loss": 0.6932,
+      "step": 12071
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6973896966385384,
+      "learning_rate": 1.679341268553356e-05,
+      "loss": 0.6559,
+      "step": 12072
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3083470031067583,
+      "learning_rate": 1.6792852729069655e-05,
+      "loss": 0.6938,
+      "step": 12073
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4201039426758562,
+      "learning_rate": 1.679229273305561e-05,
+      "loss": 0.6345,
+      "step": 12074
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7787161782907208,
+      "learning_rate": 1.6791732697494686e-05,
+      "loss": 0.8157,
+      "step": 12075
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6291693849878043,
+      "learning_rate": 1.6791172622390143e-05,
+      "loss": 0.5433,
+      "step": 12076
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5398974016533276,
+      "learning_rate": 1.6790612507745248e-05,
+      "loss": 0.8854,
+      "step": 12077
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3382451579757821,
+      "learning_rate": 1.6790052353563254e-05,
+      "loss": 0.6976,
+      "step": 12078
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9460063606452631,
+      "learning_rate": 1.6789492159847423e-05,
+      "loss": 0.8032,
+      "step": 12079
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4867532492951008,
+      "learning_rate": 1.6788931926601023e-05,
+      "loss": 0.7785,
+      "step": 12080
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5699660219562199,
+      "learning_rate": 1.6788371653827308e-05,
+      "loss": 0.5042,
+      "step": 12081
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5297334219299699,
+      "learning_rate": 1.6787811341529545e-05,
+      "loss": 0.7656,
+      "step": 12082
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5342560479708027,
+      "learning_rate": 1.6787250989711e-05,
+      "loss": 0.75,
+      "step": 12083
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5257899686452676,
+      "learning_rate": 1.6786690598374925e-05,
+      "loss": 0.7846,
+      "step": 12084
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7090280409922398,
+      "learning_rate": 1.678613016752459e-05,
+      "loss": 0.8817,
+      "step": 12085
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5280330677667928,
+      "learning_rate": 1.678556969716326e-05,
+      "loss": 0.8822,
+      "step": 12086
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.659897514606412,
+      "learning_rate": 1.678500918729419e-05,
+      "loss": 0.8233,
+      "step": 12087
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.627951373001148,
+      "learning_rate": 1.6784448637920654e-05,
+      "loss": 0.7826,
+      "step": 12088
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7752052684928363,
+      "learning_rate": 1.6783888049045905e-05,
+      "loss": 0.7324,
+      "step": 12089
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5658768038943028,
+      "learning_rate": 1.6783327420673217e-05,
+      "loss": 0.5516,
+      "step": 12090
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4322406438037658,
+      "learning_rate": 1.6782766752805843e-05,
+      "loss": 0.7737,
+      "step": 12091
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.590309847974064,
+      "learning_rate": 1.678220604544706e-05,
+      "loss": 0.6875,
+      "step": 12092
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5379222242613582,
+      "learning_rate": 1.6781645298600122e-05,
+      "loss": 0.708,
+      "step": 12093
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5863416992036978,
+      "learning_rate": 1.67810845122683e-05,
+      "loss": 0.7651,
+      "step": 12094
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4723510272822227,
+      "learning_rate": 1.6780523686454856e-05,
+      "loss": 0.7951,
+      "step": 12095
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9902904005521933,
+      "learning_rate": 1.6779962821163055e-05,
+      "loss": 0.8175,
+      "step": 12096
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3840591165931455,
+      "learning_rate": 1.6779401916396167e-05,
+      "loss": 0.8037,
+      "step": 12097
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.532068651819748,
+      "learning_rate": 1.6778840972157453e-05,
+      "loss": 0.6913,
+      "step": 12098
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.310302707193638,
+      "learning_rate": 1.677827998845018e-05,
+      "loss": 0.8807,
+      "step": 12099
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5223779719109485,
+      "learning_rate": 1.6777718965277618e-05,
+      "loss": 0.7234,
+      "step": 12100
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4194985435500096,
+      "learning_rate": 1.677715790264303e-05,
+      "loss": 0.8125,
+      "step": 12101
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3726260471630722,
+      "learning_rate": 1.677659680054968e-05,
+      "loss": 0.7149,
+      "step": 12102
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7633174637950633,
+      "learning_rate": 1.677603565900084e-05,
+      "loss": 0.733,
+      "step": 12103
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5142723968045362,
+      "learning_rate": 1.6775474477999775e-05,
+      "loss": 0.8158,
+      "step": 12104
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5685245157731802,
+      "learning_rate": 1.6774913257549753e-05,
+      "loss": 0.6599,
+      "step": 12105
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6878924773754016,
+      "learning_rate": 1.677435199765404e-05,
+      "loss": 0.6894,
+      "step": 12106
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5934522193875607,
+      "learning_rate": 1.6773790698315904e-05,
+      "loss": 0.7691,
+      "step": 12107
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5884067508938353,
+      "learning_rate": 1.6773229359538615e-05,
+      "loss": 0.9089,
+      "step": 12108
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3845479999163122,
+      "learning_rate": 1.677266798132544e-05,
+      "loss": 0.7804,
+      "step": 12109
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3741330578882107,
+      "learning_rate": 1.6772106563679652e-05,
+      "loss": 0.6377,
+      "step": 12110
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8873146247219104,
+      "learning_rate": 1.677154510660451e-05,
+      "loss": 0.8094,
+      "step": 12111
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1185949743062826,
+      "learning_rate": 1.677098361010329e-05,
+      "loss": 0.7794,
+      "step": 12112
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4131269423364732,
+      "learning_rate": 1.677042207417926e-05,
+      "loss": 0.6914,
+      "step": 12113
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.466446312507922,
+      "learning_rate": 1.676986049883569e-05,
+      "loss": 0.7647,
+      "step": 12114
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8081560759102835,
+      "learning_rate": 1.676929888407585e-05,
+      "loss": 0.8266,
+      "step": 12115
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4535255699284757,
+      "learning_rate": 1.6768737229903006e-05,
+      "loss": 0.7256,
+      "step": 12116
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.31458534938046,
+      "learning_rate": 1.6768175536320433e-05,
+      "loss": 0.8438,
+      "step": 12117
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.715124279133395,
+      "learning_rate": 1.67676138033314e-05,
+      "loss": 0.7205,
+      "step": 12118
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5175152243032857,
+      "learning_rate": 1.6767052030939176e-05,
+      "loss": 0.8545,
+      "step": 12119
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6051867195378855,
+      "learning_rate": 1.676649021914703e-05,
+      "loss": 0.7114,
+      "step": 12120
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6243697591403091,
+      "learning_rate": 1.676592836795824e-05,
+      "loss": 0.5484,
+      "step": 12121
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.918554010215171,
+      "learning_rate": 1.6765366477376067e-05,
+      "loss": 0.7702,
+      "step": 12122
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.584945966160108,
+      "learning_rate": 1.6764804547403794e-05,
+      "loss": 0.8666,
+      "step": 12123
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7925753846772843,
+      "learning_rate": 1.6764242578044685e-05,
+      "loss": 0.847,
+      "step": 12124
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6321404987494095,
+      "learning_rate": 1.6763680569302013e-05,
+      "loss": 0.4957,
+      "step": 12125
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4067911724161195,
+      "learning_rate": 1.6763118521179055e-05,
+      "loss": 0.7094,
+      "step": 12126
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6348146882241028,
+      "learning_rate": 1.6762556433679077e-05,
+      "loss": 0.7105,
+      "step": 12127
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4540637586633125,
+      "learning_rate": 1.6761994306805352e-05,
+      "loss": 0.8734,
+      "step": 12128
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5353430550041052,
+      "learning_rate": 1.676143214056116e-05,
+      "loss": 0.6464,
+      "step": 12129
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7615400438322182,
+      "learning_rate": 1.676086993494977e-05,
+      "loss": 0.7767,
+      "step": 12130
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9334203538188128,
+      "learning_rate": 1.676030768997445e-05,
+      "loss": 0.7192,
+      "step": 12131
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3079028000944632,
+      "learning_rate": 1.6759745405638483e-05,
+      "loss": 0.5932,
+      "step": 12132
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5446137755336127,
+      "learning_rate": 1.6759183081945132e-05,
+      "loss": 0.8795,
+      "step": 12133
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.653202723035308,
+      "learning_rate": 1.6758620718897686e-05,
+      "loss": 0.793,
+      "step": 12134
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7270218592905644,
+      "learning_rate": 1.6758058316499404e-05,
+      "loss": 0.6645,
+      "step": 12135
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5981408123343525,
+      "learning_rate": 1.675749587475357e-05,
+      "loss": 0.6022,
+      "step": 12136
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5742186518760353,
+      "learning_rate": 1.6756933393663454e-05,
+      "loss": 0.822,
+      "step": 12137
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2922887997097365,
+      "learning_rate": 1.6756370873232333e-05,
+      "loss": 0.6606,
+      "step": 12138
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2974975851100248,
+      "learning_rate": 1.6755808313463483e-05,
+      "loss": 0.7428,
+      "step": 12139
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5864840193125664,
+      "learning_rate": 1.6755245714360176e-05,
+      "loss": 0.7268,
+      "step": 12140
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6462749280672409,
+      "learning_rate": 1.675468307592569e-05,
+      "loss": 0.6095,
+      "step": 12141
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6512492287968776,
+      "learning_rate": 1.67541203981633e-05,
+      "loss": 0.7649,
+      "step": 12142
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3075859094884141,
+      "learning_rate": 1.6753557681076283e-05,
+      "loss": 0.7215,
+      "step": 12143
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5720763252364496,
+      "learning_rate": 1.675299492466792e-05,
+      "loss": 0.6933,
+      "step": 12144
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6228963416262343,
+      "learning_rate": 1.6752432128941478e-05,
+      "loss": 0.6979,
+      "step": 12145
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4252355862911665,
+      "learning_rate": 1.675186929390024e-05,
+      "loss": 0.8144,
+      "step": 12146
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3940753326734476,
+      "learning_rate": 1.6751306419547477e-05,
+      "loss": 0.7245,
+      "step": 12147
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5024830566410499,
+      "learning_rate": 1.6750743505886477e-05,
+      "loss": 0.8005,
+      "step": 12148
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3294584900979622,
+      "learning_rate": 1.6750180552920505e-05,
+      "loss": 0.7555,
+      "step": 12149
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.458190380825793,
+      "learning_rate": 1.674961756065285e-05,
+      "loss": 0.6784,
+      "step": 12150
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.7036855815408884,
+      "learning_rate": 1.674905452908678e-05,
+      "loss": 0.7943,
+      "step": 12151
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6027867583179485,
+      "learning_rate": 1.6748491458225583e-05,
+      "loss": 0.7984,
+      "step": 12152
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5454155724571408,
+      "learning_rate": 1.674792834807253e-05,
+      "loss": 0.7238,
+      "step": 12153
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4571095173456383,
+      "learning_rate": 1.67473651986309e-05,
+      "loss": 0.848,
+      "step": 12154
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.845598836434738,
+      "learning_rate": 1.6746802009903973e-05,
+      "loss": 0.6857,
+      "step": 12155
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.37089858341291,
+      "learning_rate": 1.674623878189503e-05,
+      "loss": 0.7591,
+      "step": 12156
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4349629440901397,
+      "learning_rate": 1.6745675514607352e-05,
+      "loss": 0.7541,
+      "step": 12157
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8639323193931663,
+      "learning_rate": 1.674511220804421e-05,
+      "loss": 0.8127,
+      "step": 12158
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5912356874838156,
+      "learning_rate": 1.674454886220889e-05,
+      "loss": 0.5392,
+      "step": 12159
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.403543573983103,
+      "learning_rate": 1.6743985477104674e-05,
+      "loss": 0.7255,
+      "step": 12160
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5622653845389216,
+      "learning_rate": 1.674342205273484e-05,
+      "loss": 0.7644,
+      "step": 12161
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.576006407061204,
+      "learning_rate": 1.6742858589102667e-05,
+      "loss": 0.6475,
+      "step": 12162
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2917404558164736,
+      "learning_rate": 1.6742295086211437e-05,
+      "loss": 0.7836,
+      "step": 12163
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4409279553009315,
+      "learning_rate": 1.674173154406443e-05,
+      "loss": 0.8595,
+      "step": 12164
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.372021224418202,
+      "learning_rate": 1.6741167962664926e-05,
+      "loss": 0.7853,
+      "step": 12165
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.886778723158633,
+      "learning_rate": 1.674060434201621e-05,
+      "loss": 0.8471,
+      "step": 12166
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3571961381894984,
+      "learning_rate": 1.674004068212156e-05,
+      "loss": 0.7838,
+      "step": 12167
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3149652439410693,
+      "learning_rate": 1.6739476982984262e-05,
+      "loss": 0.6773,
+      "step": 12168
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4167753094978308,
+      "learning_rate": 1.6738913244607596e-05,
+      "loss": 0.6861,
+      "step": 12169
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4820875632158765,
+      "learning_rate": 1.6738349466994837e-05,
+      "loss": 0.8084,
+      "step": 12170
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.332864125800148,
+      "learning_rate": 1.673778565014928e-05,
+      "loss": 0.765,
+      "step": 12171
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7962844473832897,
+      "learning_rate": 1.6737221794074198e-05,
+      "loss": 0.7015,
+      "step": 12172
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.585817173036929,
+      "learning_rate": 1.673665789877288e-05,
+      "loss": 0.5314,
+      "step": 12173
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7471745832998835,
+      "learning_rate": 1.6736093964248607e-05,
+      "loss": 0.8805,
+      "step": 12174
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5405550354412347,
+      "learning_rate": 1.673552999050466e-05,
+      "loss": 0.7474,
+      "step": 12175
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6709041194398342,
+      "learning_rate": 1.673496597754433e-05,
+      "loss": 0.6866,
+      "step": 12176
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5473105547581334,
+      "learning_rate": 1.673440192537089e-05,
+      "loss": 0.7335,
+      "step": 12177
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.532400154062661,
+      "learning_rate": 1.6733837833987634e-05,
+      "loss": 0.7092,
+      "step": 12178
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4448506847022233,
+      "learning_rate": 1.6733273703397837e-05,
+      "loss": 0.7196,
+      "step": 12179
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.547291968921952,
+      "learning_rate": 1.6732709533604795e-05,
+      "loss": 0.6677,
+      "step": 12180
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5055315543288201,
+      "learning_rate": 1.6732145324611782e-05,
+      "loss": 0.8713,
+      "step": 12181
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.490854696790821,
+      "learning_rate": 1.6731581076422093e-05,
+      "loss": 0.7557,
+      "step": 12182
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6144142179227128,
+      "learning_rate": 1.6731016789039004e-05,
+      "loss": 0.7226,
+      "step": 12183
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.3620443924236056,
+      "learning_rate": 1.6730452462465804e-05,
+      "loss": 0.8351,
+      "step": 12184
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3583097852103296,
+      "learning_rate": 1.672988809670578e-05,
+      "loss": 0.75,
+      "step": 12185
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5088614360416794,
+      "learning_rate": 1.6729323691762215e-05,
+      "loss": 0.7763,
+      "step": 12186
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6387024585576027,
+      "learning_rate": 1.6728759247638403e-05,
+      "loss": 0.8389,
+      "step": 12187
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4165959848467342,
+      "learning_rate": 1.6728194764337616e-05,
+      "loss": 0.6911,
+      "step": 12188
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5789298346260602,
+      "learning_rate": 1.6727630241863154e-05,
+      "loss": 0.5076,
+      "step": 12189
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5063544799520296,
+      "learning_rate": 1.6727065680218298e-05,
+      "loss": 0.712,
+      "step": 12190
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8086899165284644,
+      "learning_rate": 1.6726501079406333e-05,
+      "loss": 0.7986,
+      "step": 12191
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5174025935050794,
+      "learning_rate": 1.6725936439430556e-05,
+      "loss": 0.8255,
+      "step": 12192
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7176575200481903,
+      "learning_rate": 1.672537176029424e-05,
+      "loss": 0.8879,
+      "step": 12193
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.60398797727686,
+      "learning_rate": 1.6724807042000683e-05,
+      "loss": 0.545,
+      "step": 12194
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5949923879610828,
+      "learning_rate": 1.6724242284553173e-05,
+      "loss": 0.5588,
+      "step": 12195
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.598179488199667,
+      "learning_rate": 1.6723677487954994e-05,
+      "loss": 0.6545,
+      "step": 12196
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.56754459332734,
+      "learning_rate": 1.6723112652209435e-05,
+      "loss": 0.7106,
+      "step": 12197
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.601501974355594,
+      "learning_rate": 1.6722547777319788e-05,
+      "loss": 0.9341,
+      "step": 12198
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.312717253644393,
+      "learning_rate": 1.6721982863289337e-05,
+      "loss": 0.6762,
+      "step": 12199
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6102826330695366,
+      "learning_rate": 1.6721417910121373e-05,
+      "loss": 0.6556,
+      "step": 12200
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.711095625754748,
+      "learning_rate": 1.672085291781919e-05,
+      "loss": 0.7902,
+      "step": 12201
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5668131639121998,
+      "learning_rate": 1.6720287886386065e-05,
+      "loss": 0.7247,
+      "step": 12202
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.513339598173228,
+      "learning_rate": 1.6719722815825304e-05,
+      "loss": 0.8281,
+      "step": 12203
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7241296961338701,
+      "learning_rate": 1.6719157706140187e-05,
+      "loss": 0.7667,
+      "step": 12204
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4188694997659463,
+      "learning_rate": 1.671859255733401e-05,
+      "loss": 0.7629,
+      "step": 12205
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5765570768552004,
+      "learning_rate": 1.6718027369410057e-05,
+      "loss": 0.6571,
+      "step": 12206
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4257997639091728,
+      "learning_rate": 1.6717462142371622e-05,
+      "loss": 0.7266,
+      "step": 12207
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.576409165248384,
+      "learning_rate": 1.6716896876221996e-05,
+      "loss": 0.772,
+      "step": 12208
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3030690967945087,
+      "learning_rate": 1.6716331570964465e-05,
+      "loss": 0.7656,
+      "step": 12209
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.32684423621965,
+      "learning_rate": 1.6715766226602334e-05,
+      "loss": 0.8391,
+      "step": 12210
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.337476155352302,
+      "learning_rate": 1.671520084313888e-05,
+      "loss": 0.7038,
+      "step": 12211
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5416396659362137,
+      "learning_rate": 1.6714635420577405e-05,
+      "loss": 0.797,
+      "step": 12212
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8596395881766807,
+      "learning_rate": 1.6714069958921193e-05,
+      "loss": 0.776,
+      "step": 12213
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4747283347766191,
+      "learning_rate": 1.6713504458173538e-05,
+      "loss": 0.6959,
+      "step": 12214
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4294247403202713,
+      "learning_rate": 1.671293891833774e-05,
+      "loss": 0.7354,
+      "step": 12215
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.355925349100905,
+      "learning_rate": 1.6712373339417082e-05,
+      "loss": 0.7908,
+      "step": 12216
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8328259019423179,
+      "learning_rate": 1.6711807721414865e-05,
+      "loss": 0.7442,
+      "step": 12217
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5326545494941883,
+      "learning_rate": 1.6711242064334377e-05,
+      "loss": 0.828,
+      "step": 12218
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.522257858682721,
+      "learning_rate": 1.6710676368178913e-05,
+      "loss": 0.7591,
+      "step": 12219
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5713359408778932,
+      "learning_rate": 1.6710110632951765e-05,
+      "loss": 0.5637,
+      "step": 12220
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.042851251664237,
+      "learning_rate": 1.6709544858656232e-05,
+      "loss": 0.7908,
+      "step": 12221
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4461488946812853,
+      "learning_rate": 1.67089790452956e-05,
+      "loss": 0.7255,
+      "step": 12222
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4297784797683757,
+      "learning_rate": 1.670841319287317e-05,
+      "loss": 0.7897,
+      "step": 12223
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4506945054917586,
+      "learning_rate": 1.6707847301392237e-05,
+      "loss": 0.8786,
+      "step": 12224
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4360823609037419,
+      "learning_rate": 1.670728137085609e-05,
+      "loss": 0.7474,
+      "step": 12225
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6757435883854674,
+      "learning_rate": 1.670671540126803e-05,
+      "loss": 0.765,
+      "step": 12226
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4494386449554255,
+      "learning_rate": 1.670614939263135e-05,
+      "loss": 0.7608,
+      "step": 12227
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.620386838272102,
+      "learning_rate": 1.670558334494934e-05,
+      "loss": 0.6944,
+      "step": 12228
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6087875368123186,
+      "learning_rate": 1.6705017258225306e-05,
+      "loss": 0.5204,
+      "step": 12229
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5899212718727946,
+      "learning_rate": 1.6704451132462534e-05,
+      "loss": 0.5708,
+      "step": 12230
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5336774467445071,
+      "learning_rate": 1.670388496766433e-05,
+      "loss": 0.6473,
+      "step": 12231
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.518641527010407,
+      "learning_rate": 1.6703318763833985e-05,
+      "loss": 0.6515,
+      "step": 12232
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4138095887162745,
+      "learning_rate": 1.670275252097479e-05,
+      "loss": 0.7081,
+      "step": 12233
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3007143567884023,
+      "learning_rate": 1.6702186239090058e-05,
+      "loss": 0.7359,
+      "step": 12234
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6857516429351365,
+      "learning_rate": 1.6701619918183066e-05,
+      "loss": 0.6649,
+      "step": 12235
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3926164787963466,
+      "learning_rate": 1.6701053558257125e-05,
+      "loss": 0.5787,
+      "step": 12236
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.568479094209259,
+      "learning_rate": 1.670048715931553e-05,
+      "loss": 0.767,
+      "step": 12237
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.614331773372803,
+      "learning_rate": 1.6699920721361576e-05,
+      "loss": 0.7392,
+      "step": 12238
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5788015183921995,
+      "learning_rate": 1.669935424439856e-05,
+      "loss": 0.7174,
+      "step": 12239
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.283508761484908,
+      "learning_rate": 1.6698787728429786e-05,
+      "loss": 0.8112,
+      "step": 12240
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5936509749821988,
+      "learning_rate": 1.6698221173458547e-05,
+      "loss": 0.8118,
+      "step": 12241
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2654389459651254,
+      "learning_rate": 1.6697654579488147e-05,
+      "loss": 0.8193,
+      "step": 12242
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.828147897377304,
+      "learning_rate": 1.669708794652188e-05,
+      "loss": 0.7766,
+      "step": 12243
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4671623126588995,
+      "learning_rate": 1.6696521274563046e-05,
+      "loss": 0.7869,
+      "step": 12244
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6387677449524154,
+      "learning_rate": 1.6695954563614945e-05,
+      "loss": 0.8035,
+      "step": 12245
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6463259225552047,
+      "learning_rate": 1.6695387813680873e-05,
+      "loss": 0.8517,
+      "step": 12246
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3747483042511666,
+      "learning_rate": 1.6694821024764142e-05,
+      "loss": 0.8208,
+      "step": 12247
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1966236612165557,
+      "learning_rate": 1.6694254196868036e-05,
+      "loss": 0.6337,
+      "step": 12248
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7186359656398238,
+      "learning_rate": 1.6693687329995864e-05,
+      "loss": 0.9043,
+      "step": 12249
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.623912691978888,
+      "learning_rate": 1.669312042415093e-05,
+      "loss": 0.5407,
+      "step": 12250
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3585926616008417,
+      "learning_rate": 1.6692553479336522e-05,
+      "loss": 0.7579,
+      "step": 12251
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5482592965506854,
+      "learning_rate": 1.6691986495555955e-05,
+      "loss": 0.7693,
+      "step": 12252
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.537644541752576,
+      "learning_rate": 1.6691419472812523e-05,
+      "loss": 0.8059,
+      "step": 12253
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.622750041847824,
+      "learning_rate": 1.669085241110953e-05,
+      "loss": 0.6501,
+      "step": 12254
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4617192433239237,
+      "learning_rate": 1.669028531045027e-05,
+      "loss": 0.6218,
+      "step": 12255
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4667995149536128,
+      "learning_rate": 1.6689718170838058e-05,
+      "loss": 0.8637,
+      "step": 12256
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6782698718985503,
+      "learning_rate": 1.6689150992276183e-05,
+      "loss": 0.8196,
+      "step": 12257
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2659960684739668,
+      "learning_rate": 1.6688583774767954e-05,
+      "loss": 0.7351,
+      "step": 12258
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2941402481899236,
+      "learning_rate": 1.6688016518316676e-05,
+      "loss": 0.7744,
+      "step": 12259
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9579221423339734,
+      "learning_rate": 1.6687449222925644e-05,
+      "loss": 0.7505,
+      "step": 12260
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.562899957407383,
+      "learning_rate": 1.668688188859817e-05,
+      "loss": 0.7301,
+      "step": 12261
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5880980780321737,
+      "learning_rate": 1.668631451533755e-05,
+      "loss": 0.5548,
+      "step": 12262
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5684127066759632,
+      "learning_rate": 1.668574710314709e-05,
+      "loss": 0.6977,
+      "step": 12263
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6087491548214804,
+      "learning_rate": 1.6685179652030096e-05,
+      "loss": 0.5575,
+      "step": 12264
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.597092044395716,
+      "learning_rate": 1.6684612161989866e-05,
+      "loss": 0.562,
+      "step": 12265
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2507806120058296,
+      "learning_rate": 1.668404463302971e-05,
+      "loss": 0.7075,
+      "step": 12266
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5954287550839826,
+      "learning_rate": 1.6683477065152928e-05,
+      "loss": 0.5879,
+      "step": 12267
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0045671972782255,
+      "learning_rate": 1.668290945836283e-05,
+      "loss": 0.8957,
+      "step": 12268
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3307342426396147,
+      "learning_rate": 1.6682341812662715e-05,
+      "loss": 0.7951,
+      "step": 12269
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5754384239400467,
+      "learning_rate": 1.6681774128055893e-05,
+      "loss": 0.7174,
+      "step": 12270
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4228965088468113,
+      "learning_rate": 1.6681206404545665e-05,
+      "loss": 0.8096,
+      "step": 12271
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5020672899962026,
+      "learning_rate": 1.6680638642135337e-05,
+      "loss": 0.8679,
+      "step": 12272
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1534481465310265,
+      "learning_rate": 1.6680070840828215e-05,
+      "loss": 0.7183,
+      "step": 12273
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2765699977844982,
+      "learning_rate": 1.6679503000627607e-05,
+      "loss": 0.7355,
+      "step": 12274
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2641317108046513,
+      "learning_rate": 1.6678935121536817e-05,
+      "loss": 0.7483,
+      "step": 12275
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5721970611984226,
+      "learning_rate": 1.6678367203559154e-05,
+      "loss": 0.7282,
+      "step": 12276
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1794256304631636,
+      "learning_rate": 1.6677799246697924e-05,
+      "loss": 0.7895,
+      "step": 12277
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5579841822689708,
+      "learning_rate": 1.667723125095643e-05,
+      "loss": 0.8621,
+      "step": 12278
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7829425086789101,
+      "learning_rate": 1.667666321633798e-05,
+      "loss": 0.7263,
+      "step": 12279
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3379379617255087,
+      "learning_rate": 1.6676095142845883e-05,
+      "loss": 0.7319,
+      "step": 12280
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4991923941186978,
+      "learning_rate": 1.6675527030483452e-05,
+      "loss": 0.848,
+      "step": 12281
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5177731134987502,
+      "learning_rate": 1.6674958879253986e-05,
+      "loss": 0.801,
+      "step": 12282
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.542215715725451,
+      "learning_rate": 1.6674390689160795e-05,
+      "loss": 0.8657,
+      "step": 12283
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.592775739375997,
+      "learning_rate": 1.667382246020719e-05,
+      "loss": 0.7653,
+      "step": 12284
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4700152046987562,
+      "learning_rate": 1.667325419239648e-05,
+      "loss": 0.7628,
+      "step": 12285
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.569566705201822,
+      "learning_rate": 1.6672685885731964e-05,
+      "loss": 0.7621,
+      "step": 12286
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.35591582785619,
+      "learning_rate": 1.6672117540216962e-05,
+      "loss": 0.7141,
+      "step": 12287
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.703381592354569,
+      "learning_rate": 1.6671549155854783e-05,
+      "loss": 0.7719,
+      "step": 12288
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.397100979468942,
+      "learning_rate": 1.6670980732648726e-05,
+      "loss": 0.6139,
+      "step": 12289
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5095887819654665,
+      "learning_rate": 1.6670412270602115e-05,
+      "loss": 0.8841,
+      "step": 12290
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.612899089632388,
+      "learning_rate": 1.6669843769718245e-05,
+      "loss": 0.8703,
+      "step": 12291
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6137414382092228,
+      "learning_rate": 1.6669275230000438e-05,
+      "loss": 0.6972,
+      "step": 12292
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6042235817438484,
+      "learning_rate": 1.6668706651451994e-05,
+      "loss": 0.664,
+      "step": 12293
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6721840414959888,
+      "learning_rate": 1.6668138034076235e-05,
+      "loss": 0.7628,
+      "step": 12294
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2873586530196781,
+      "learning_rate": 1.666756937787646e-05,
+      "loss": 0.7162,
+      "step": 12295
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3669306386690268,
+      "learning_rate": 1.6667000682855988e-05,
+      "loss": 0.8254,
+      "step": 12296
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5584425562787487,
+      "learning_rate": 1.666643194901812e-05,
+      "loss": 0.7907,
+      "step": 12297
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3733776153492263,
+      "learning_rate": 1.6665863176366186e-05,
+      "loss": 0.8153,
+      "step": 12298
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.133546436564965,
+      "learning_rate": 1.6665294364903482e-05,
+      "loss": 0.7469,
+      "step": 12299
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5404775041469387,
+      "learning_rate": 1.666472551463332e-05,
+      "loss": 0.7589,
+      "step": 12300
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6685495488216784,
+      "learning_rate": 1.666415662555902e-05,
+      "loss": 0.8624,
+      "step": 12301
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5008930164704628,
+      "learning_rate": 1.666358769768389e-05,
+      "loss": 0.7494,
+      "step": 12302
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7377987732429252,
+      "learning_rate": 1.666301873101124e-05,
+      "loss": 0.6968,
+      "step": 12303
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5587414897010041,
+      "learning_rate": 1.6662449725544386e-05,
+      "loss": 0.6134,
+      "step": 12304
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.569133387513511,
+      "learning_rate": 1.6661880681286647e-05,
+      "loss": 0.6648,
+      "step": 12305
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3177663540186413,
+      "learning_rate": 1.6661311598241325e-05,
+      "loss": 0.7274,
+      "step": 12306
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.5500057801230605,
+      "learning_rate": 1.6660742476411736e-05,
+      "loss": 0.7664,
+      "step": 12307
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6687652644822866,
+      "learning_rate": 1.6660173315801194e-05,
+      "loss": 0.6894,
+      "step": 12308
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5205766956104008,
+      "learning_rate": 1.6659604116413016e-05,
+      "loss": 0.7715,
+      "step": 12309
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5105947677527596,
+      "learning_rate": 1.665903487825052e-05,
+      "loss": 0.7865,
+      "step": 12310
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3017683814047019,
+      "learning_rate": 1.6658465601317008e-05,
+      "loss": 0.7283,
+      "step": 12311
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8056551980683782,
+      "learning_rate": 1.6657896285615804e-05,
+      "loss": 0.8284,
+      "step": 12312
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6228506506010454,
+      "learning_rate": 1.6657326931150222e-05,
+      "loss": 0.7609,
+      "step": 12313
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4419439521907143,
+      "learning_rate": 1.6656757537923572e-05,
+      "loss": 0.8068,
+      "step": 12314
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4506942586288507,
+      "learning_rate": 1.6656188105939172e-05,
+      "loss": 0.7869,
+      "step": 12315
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.386339107909207,
+      "learning_rate": 1.665561863520034e-05,
+      "loss": 0.7447,
+      "step": 12316
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.480408891769001,
+      "learning_rate": 1.665504912571039e-05,
+      "loss": 0.7822,
+      "step": 12317
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5673268388211645,
+      "learning_rate": 1.665447957747263e-05,
+      "loss": 0.79,
+      "step": 12318
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.391133218044886,
+      "learning_rate": 1.665390999049039e-05,
+      "loss": 0.8125,
+      "step": 12319
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4972383985920328,
+      "learning_rate": 1.6653340364766978e-05,
+      "loss": 0.7779,
+      "step": 12320
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4000795615801773,
+      "learning_rate": 1.6652770700305708e-05,
+      "loss": 0.8458,
+      "step": 12321
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4683477705729873,
+      "learning_rate": 1.6652200997109906e-05,
+      "loss": 0.7428,
+      "step": 12322
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3060802505441653,
+      "learning_rate": 1.665163125518288e-05,
+      "loss": 0.8462,
+      "step": 12323
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0153668876584745,
+      "learning_rate": 1.6651061474527953e-05,
+      "loss": 0.7747,
+      "step": 12324
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4722580836975232,
+      "learning_rate": 1.665049165514844e-05,
+      "loss": 0.7891,
+      "step": 12325
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5378410615111537,
+      "learning_rate": 1.6649921797047658e-05,
+      "loss": 0.8129,
+      "step": 12326
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6034570031323796,
+      "learning_rate": 1.664935190022893e-05,
+      "loss": 0.7742,
+      "step": 12327
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.355065224140374,
+      "learning_rate": 1.6648781964695565e-05,
+      "loss": 0.74,
+      "step": 12328
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.800429838290615,
+      "learning_rate": 1.664821199045089e-05,
+      "loss": 0.7355,
+      "step": 12329
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.393852316937475,
+      "learning_rate": 1.6647641977498215e-05,
+      "loss": 0.7547,
+      "step": 12330
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5737162710001213,
+      "learning_rate": 1.6647071925840868e-05,
+      "loss": 0.7081,
+      "step": 12331
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5986122072426485,
+      "learning_rate": 1.6646501835482162e-05,
+      "loss": 0.591,
+      "step": 12332
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2918305839920485,
+      "learning_rate": 1.664593170642542e-05,
+      "loss": 0.7296,
+      "step": 12333
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4167156264561842,
+      "learning_rate": 1.6645361538673962e-05,
+      "loss": 0.7543,
+      "step": 12334
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5874565763399995,
+      "learning_rate": 1.66447913322311e-05,
+      "loss": 0.9362,
+      "step": 12335
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4974994448445273,
+      "learning_rate": 1.6644221087100164e-05,
+      "loss": 0.8101,
+      "step": 12336
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5892000974545051,
+      "learning_rate": 1.6643650803284462e-05,
+      "loss": 0.559,
+      "step": 12337
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4781673291540156,
+      "learning_rate": 1.6643080480787328e-05,
+      "loss": 0.7274,
+      "step": 12338
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1791798801420896,
+      "learning_rate": 1.6642510119612073e-05,
+      "loss": 0.645,
+      "step": 12339
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3432643412001444,
+      "learning_rate": 1.6641939719762023e-05,
+      "loss": 0.6624,
+      "step": 12340
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.447169230199966,
+      "learning_rate": 1.6641369281240497e-05,
+      "loss": 0.837,
+      "step": 12341
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6058502026730737,
+      "learning_rate": 1.6640798804050816e-05,
+      "loss": 0.5528,
+      "step": 12342
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5101356856477681,
+      "learning_rate": 1.6640228288196303e-05,
+      "loss": 0.793,
+      "step": 12343
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3987581479763687,
+      "learning_rate": 1.6639657733680276e-05,
+      "loss": 0.6894,
+      "step": 12344
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.2191680000748035,
+      "learning_rate": 1.663908714050606e-05,
+      "loss": 0.7731,
+      "step": 12345
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4046186714509301,
+      "learning_rate": 1.6638516508676977e-05,
+      "loss": 0.8969,
+      "step": 12346
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5230981042954481,
+      "learning_rate": 1.6637945838196356e-05,
+      "loss": 0.7457,
+      "step": 12347
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6193816841018385,
+      "learning_rate": 1.6637375129067502e-05,
+      "loss": 0.5662,
+      "step": 12348
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4372445327772714,
+      "learning_rate": 1.6636804381293755e-05,
+      "loss": 0.7556,
+      "step": 12349
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6725932359963875,
+      "learning_rate": 1.6636233594878434e-05,
+      "loss": 0.8071,
+      "step": 12350
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6631725120343939,
+      "learning_rate": 1.6635662769824854e-05,
+      "loss": 0.802,
+      "step": 12351
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7198095543720833,
+      "learning_rate": 1.6635091906136346e-05,
+      "loss": 0.8302,
+      "step": 12352
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.539253118334753,
+      "learning_rate": 1.6634521003816235e-05,
+      "loss": 0.8169,
+      "step": 12353
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6483307083742365,
+      "learning_rate": 1.663395006286784e-05,
+      "loss": 0.5837,
+      "step": 12354
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.651385656976255,
+      "learning_rate": 1.6633379083294487e-05,
+      "loss": 0.7962,
+      "step": 12355
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6065900035829113,
+      "learning_rate": 1.6632808065099503e-05,
+      "loss": 0.5491,
+      "step": 12356
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.477636385583988,
+      "learning_rate": 1.663223700828621e-05,
+      "loss": 0.7351,
+      "step": 12357
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5069792556635961,
+      "learning_rate": 1.663166591285793e-05,
+      "loss": 0.7793,
+      "step": 12358
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7566747763772625,
+      "learning_rate": 1.6631094778817997e-05,
+      "loss": 0.6443,
+      "step": 12359
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8231988383574427,
+      "learning_rate": 1.6630523606169726e-05,
+      "loss": 0.8189,
+      "step": 12360
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3680871161536108,
+      "learning_rate": 1.6629952394916447e-05,
+      "loss": 0.7211,
+      "step": 12361
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.417331438575726,
+      "learning_rate": 1.6629381145061492e-05,
+      "loss": 0.8677,
+      "step": 12362
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6881238975677237,
+      "learning_rate": 1.6628809856608177e-05,
+      "loss": 0.7897,
+      "step": 12363
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3999481864451946,
+      "learning_rate": 1.6628238529559834e-05,
+      "loss": 0.8059,
+      "step": 12364
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6679620928057082,
+      "learning_rate": 1.6627667163919788e-05,
+      "loss": 0.7124,
+      "step": 12365
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4012894141151542,
+      "learning_rate": 1.6627095759691364e-05,
+      "loss": 0.8204,
+      "step": 12366
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4392708516554273,
+      "learning_rate": 1.662652431687789e-05,
+      "loss": 0.8275,
+      "step": 12367
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3266896032282942,
+      "learning_rate": 1.6625952835482696e-05,
+      "loss": 0.8697,
+      "step": 12368
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4173316662696793,
+      "learning_rate": 1.6625381315509103e-05,
+      "loss": 0.7584,
+      "step": 12369
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6926754424639465,
+      "learning_rate": 1.6624809756960445e-05,
+      "loss": 0.746,
+      "step": 12370
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6024811809223989,
+      "learning_rate": 1.6624238159840043e-05,
+      "loss": 0.7775,
+      "step": 12371
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4453523111055673,
+      "learning_rate": 1.6623666524151233e-05,
+      "loss": 0.8547,
+      "step": 12372
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6190633849298928,
+      "learning_rate": 1.6623094849897342e-05,
+      "loss": 0.7305,
+      "step": 12373
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4454494766319328,
+      "learning_rate": 1.6622523137081693e-05,
+      "loss": 0.682,
+      "step": 12374
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4996612040762287,
+      "learning_rate": 1.6621951385707617e-05,
+      "loss": 0.7944,
+      "step": 12375
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6105574296186121,
+      "learning_rate": 1.6621379595778444e-05,
+      "loss": 0.5697,
+      "step": 12376
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5917812452232754,
+      "learning_rate": 1.6620807767297505e-05,
+      "loss": 0.7826,
+      "step": 12377
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5725594083572119,
+      "learning_rate": 1.6620235900268127e-05,
+      "loss": 0.8455,
+      "step": 12378
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4028020685588893,
+      "learning_rate": 1.6619663994693634e-05,
+      "loss": 0.7357,
+      "step": 12379
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6477931990853998,
+      "learning_rate": 1.6619092050577367e-05,
+      "loss": 0.6944,
+      "step": 12380
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3503374558480572,
+      "learning_rate": 1.661852006792265e-05,
+      "loss": 0.7968,
+      "step": 12381
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7165543280164672,
+      "learning_rate": 1.6617948046732814e-05,
+      "loss": 0.7627,
+      "step": 12382
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.631509244529767,
+      "learning_rate": 1.6617375987011186e-05,
+      "loss": 0.7684,
+      "step": 12383
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.896627006045198,
+      "learning_rate": 1.6616803888761105e-05,
+      "loss": 0.7549,
+      "step": 12384
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.388590993259857,
+      "learning_rate": 1.6616231751985893e-05,
+      "loss": 0.8863,
+      "step": 12385
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6704745208828036,
+      "learning_rate": 1.6615659576688886e-05,
+      "loss": 0.7118,
+      "step": 12386
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.79907637248468,
+      "learning_rate": 1.6615087362873415e-05,
+      "loss": 0.8341,
+      "step": 12387
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6073507560749296,
+      "learning_rate": 1.6614515110542807e-05,
+      "loss": 0.5725,
+      "step": 12388
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3060897003954355,
+      "learning_rate": 1.6613942819700402e-05,
+      "loss": 0.7166,
+      "step": 12389
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6533313504124927,
+      "learning_rate": 1.661337049034953e-05,
+      "loss": 0.7026,
+      "step": 12390
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4627950988228353,
+      "learning_rate": 1.661279812249352e-05,
+      "loss": 0.716,
+      "step": 12391
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6012283099006641,
+      "learning_rate": 1.6612225716135703e-05,
+      "loss": 0.6198,
+      "step": 12392
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5415523348295577,
+      "learning_rate": 1.6611653271279415e-05,
+      "loss": 0.77,
+      "step": 12393
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4579358958660174,
+      "learning_rate": 1.661108078792799e-05,
+      "loss": 0.7669,
+      "step": 12394
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.390933460584663,
+      "learning_rate": 1.661050826608476e-05,
+      "loss": 0.7162,
+      "step": 12395
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3262925916692054,
+      "learning_rate": 1.6609935705753057e-05,
+      "loss": 0.7752,
+      "step": 12396
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1771207966726158,
+      "learning_rate": 1.6609363106936215e-05,
+      "loss": 0.7267,
+      "step": 12397
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.690270363714725,
+      "learning_rate": 1.660879046963757e-05,
+      "loss": 0.7111,
+      "step": 12398
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4520690837624353,
+      "learning_rate": 1.6608217793860455e-05,
+      "loss": 0.8363,
+      "step": 12399
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.629040179114772,
+      "learning_rate": 1.66076450796082e-05,
+      "loss": 0.6325,
+      "step": 12400
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3021946811365195,
+      "learning_rate": 1.660707232688415e-05,
+      "loss": 0.7914,
+      "step": 12401
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4368566107490346,
+      "learning_rate": 1.6606499535691623e-05,
+      "loss": 0.7117,
+      "step": 12402
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.656746349236372,
+      "learning_rate": 1.6605926706033973e-05,
+      "loss": 0.7511,
+      "step": 12403
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6557147040675115,
+      "learning_rate": 1.6605353837914522e-05,
+      "loss": 0.5851,
+      "step": 12404
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.587915759797511,
+      "learning_rate": 1.660478093133661e-05,
+      "loss": 0.8091,
+      "step": 12405
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5328931159828598,
+      "learning_rate": 1.6604207986303577e-05,
+      "loss": 0.7103,
+      "step": 12406
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8423411719478304,
+      "learning_rate": 1.660363500281875e-05,
+      "loss": 0.7749,
+      "step": 12407
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.704638826325992,
+      "learning_rate": 1.660306198088547e-05,
+      "loss": 0.8145,
+      "step": 12408
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3575485606046913,
+      "learning_rate": 1.6602488920507075e-05,
+      "loss": 0.7967,
+      "step": 12409
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6771495613973644,
+      "learning_rate": 1.6601915821686895e-05,
+      "loss": 0.6596,
+      "step": 12410
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.274791934810601,
+      "learning_rate": 1.6601342684428274e-05,
+      "loss": 0.8436,
+      "step": 12411
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2850110864765667,
+      "learning_rate": 1.6600769508734545e-05,
+      "loss": 0.744,
+      "step": 12412
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6628479529618854,
+      "learning_rate": 1.6600196294609047e-05,
+      "loss": 0.8721,
+      "step": 12413
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2817743199424971,
+      "learning_rate": 1.6599623042055114e-05,
+      "loss": 0.8007,
+      "step": 12414
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3625142881999777,
+      "learning_rate": 1.659904975107609e-05,
+      "loss": 0.7858,
+      "step": 12415
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7213149409737312,
+      "learning_rate": 1.6598476421675303e-05,
+      "loss": 0.7666,
+      "step": 12416
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5216392875333524,
+      "learning_rate": 1.6597903053856104e-05,
+      "loss": 0.7974,
+      "step": 12417
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.569975902517442,
+      "learning_rate": 1.6597329647621822e-05,
+      "loss": 0.7831,
+      "step": 12418
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2745379586565673,
+      "learning_rate": 1.65967562029758e-05,
+      "loss": 0.9764,
+      "step": 12419
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.699206710401447,
+      "learning_rate": 1.659618271992137e-05,
+      "loss": 0.8313,
+      "step": 12420
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6226506024350134,
+      "learning_rate": 1.6595609198461878e-05,
+      "loss": 0.8883,
+      "step": 12421
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6433645520035631,
+      "learning_rate": 1.6595035638600662e-05,
+      "loss": 0.5441,
+      "step": 12422
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3678123812518685,
+      "learning_rate": 1.659446204034106e-05,
+      "loss": 0.718,
+      "step": 12423
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.490346279768887,
+      "learning_rate": 1.6593888403686417e-05,
+      "loss": 0.6615,
+      "step": 12424
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3058083150133537,
+      "learning_rate": 1.6593314728640063e-05,
+      "loss": 0.8247,
+      "step": 12425
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5268627986741166,
+      "learning_rate": 1.6592741015205343e-05,
+      "loss": 0.7466,
+      "step": 12426
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5606852265591444,
+      "learning_rate": 1.65921672633856e-05,
+      "loss": 0.769,
+      "step": 12427
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.455613674205242,
+      "learning_rate": 1.659159347318417e-05,
+      "loss": 0.6874,
+      "step": 12428
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7888336103708642,
+      "learning_rate": 1.65910196446044e-05,
+      "loss": 0.8121,
+      "step": 12429
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.27435677906508,
+      "learning_rate": 1.659044577764962e-05,
+      "loss": 0.7074,
+      "step": 12430
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2827886688983139,
+      "learning_rate": 1.6589871872323185e-05,
+      "loss": 0.6786,
+      "step": 12431
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4111353940799884,
+      "learning_rate": 1.6589297928628423e-05,
+      "loss": 0.8386,
+      "step": 12432
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.001021490814709,
+      "learning_rate": 1.658872394656869e-05,
+      "loss": 0.8735,
+      "step": 12433
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5997507916488325,
+      "learning_rate": 1.658814992614731e-05,
+      "loss": 0.6576,
+      "step": 12434
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7077766511013521,
+      "learning_rate": 1.6587575867367645e-05,
+      "loss": 0.6662,
+      "step": 12435
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9235386438227258,
+      "learning_rate": 1.6587001770233024e-05,
+      "loss": 0.8237,
+      "step": 12436
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7689426787150835,
+      "learning_rate": 1.6586427634746794e-05,
+      "loss": 0.752,
+      "step": 12437
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5263621953454267,
+      "learning_rate": 1.6585853460912295e-05,
+      "loss": 0.6894,
+      "step": 12438
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.512088418830585,
+      "learning_rate": 1.6585279248732873e-05,
+      "loss": 0.724,
+      "step": 12439
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7011527439057517,
+      "learning_rate": 1.658470499821187e-05,
+      "loss": 0.7169,
+      "step": 12440
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.654691911398465,
+      "learning_rate": 1.658413070935263e-05,
+      "loss": 0.7206,
+      "step": 12441
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.601716290998158,
+      "learning_rate": 1.65835563821585e-05,
+      "loss": 0.5175,
+      "step": 12442
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5130876762388887,
+      "learning_rate": 1.6582982016632818e-05,
+      "loss": 0.8228,
+      "step": 12443
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5302795900030675,
+      "learning_rate": 1.658240761277893e-05,
+      "loss": 0.8128,
+      "step": 12444
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5743742448855753,
+      "learning_rate": 1.6581833170600182e-05,
+      "loss": 0.8955,
+      "step": 12445
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9631803417605238,
+      "learning_rate": 1.6581258690099916e-05,
+      "loss": 0.7772,
+      "step": 12446
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8165393915781725,
+      "learning_rate": 1.6580684171281478e-05,
+      "loss": 0.647,
+      "step": 12447
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7820781503653433,
+      "learning_rate": 1.6580109614148213e-05,
+      "loss": 0.7841,
+      "step": 12448
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3061720198080664,
+      "learning_rate": 1.6579535018703472e-05,
+      "loss": 0.7785,
+      "step": 12449
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2346732837337178,
+      "learning_rate": 1.6578960384950588e-05,
+      "loss": 0.8178,
+      "step": 12450
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4032091545292114,
+      "learning_rate": 1.657838571289292e-05,
+      "loss": 0.7378,
+      "step": 12451
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4478306829674135,
+      "learning_rate": 1.6577811002533803e-05,
+      "loss": 0.8022,
+      "step": 12452
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5265622292175085,
+      "learning_rate": 1.657723625387659e-05,
+      "loss": 0.7459,
+      "step": 12453
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4843358635644825,
+      "learning_rate": 1.6576661466924625e-05,
+      "loss": 0.7958,
+      "step": 12454
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.661001900719258,
+      "learning_rate": 1.657608664168125e-05,
+      "loss": 0.8556,
+      "step": 12455
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.698016979184315,
+      "learning_rate": 1.6575511778149823e-05,
+      "loss": 0.7643,
+      "step": 12456
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5833318524564933,
+      "learning_rate": 1.657493687633368e-05,
+      "loss": 0.5388,
+      "step": 12457
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3640380988495102,
+      "learning_rate": 1.6574361936236177e-05,
+      "loss": 0.8967,
+      "step": 12458
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.440980455876584,
+      "learning_rate": 1.6573786957860658e-05,
+      "loss": 0.7058,
+      "step": 12459
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4205798882634482,
+      "learning_rate": 1.6573211941210467e-05,
+      "loss": 0.8783,
+      "step": 12460
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2829837159922617,
+      "learning_rate": 1.6572636886288957e-05,
+      "loss": 0.7275,
+      "step": 12461
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3578538068834556,
+      "learning_rate": 1.6572061793099473e-05,
+      "loss": 0.7526,
+      "step": 12462
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6236094470522848,
+      "learning_rate": 1.6571486661645367e-05,
+      "loss": 0.7748,
+      "step": 12463
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6080553154748861,
+      "learning_rate": 1.6570911491929984e-05,
+      "loss": 0.5584,
+      "step": 12464
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.632686397494034,
+      "learning_rate": 1.657033628395667e-05,
+      "loss": 0.8979,
+      "step": 12465
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4483832252577051,
+      "learning_rate": 1.656976103772878e-05,
+      "loss": 0.8216,
+      "step": 12466
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5993331904386472,
+      "learning_rate": 1.6569185753249662e-05,
+      "loss": 0.681,
+      "step": 12467
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5289293060563198,
+      "learning_rate": 1.656861043052267e-05,
+      "loss": 0.6648,
+      "step": 12468
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5195893622491041,
+      "learning_rate": 1.656803506955114e-05,
+      "loss": 0.6725,
+      "step": 12469
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3783866027948337,
+      "learning_rate": 1.6567459670338434e-05,
+      "loss": 0.7331,
+      "step": 12470
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.880054848690108,
+      "learning_rate": 1.65668842328879e-05,
+      "loss": 0.6607,
+      "step": 12471
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5349261429101764,
+      "learning_rate": 1.6566308757202883e-05,
+      "loss": 0.6953,
+      "step": 12472
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8494053212863402,
+      "learning_rate": 1.656573324328674e-05,
+      "loss": 0.855,
+      "step": 12473
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7057491113116465,
+      "learning_rate": 1.656515769114282e-05,
+      "loss": 0.6973,
+      "step": 12474
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5838028092919886,
+      "learning_rate": 1.6564582100774472e-05,
+      "loss": 0.5582,
+      "step": 12475
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.680924860519229,
+      "learning_rate": 1.656400647218505e-05,
+      "loss": 0.7485,
+      "step": 12476
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6656011678747138,
+      "learning_rate": 1.6563430805377903e-05,
+      "loss": 0.6485,
+      "step": 12477
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7566394748904266,
+      "learning_rate": 1.6562855100356386e-05,
+      "loss": 0.7587,
+      "step": 12478
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.818805391457439,
+      "learning_rate": 1.656227935712385e-05,
+      "loss": 0.7396,
+      "step": 12479
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4045074185517772,
+      "learning_rate": 1.6561703575683645e-05,
+      "loss": 0.8192,
+      "step": 12480
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6986177732391938,
+      "learning_rate": 1.6561127756039122e-05,
+      "loss": 0.7817,
+      "step": 12481
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.963065765991116,
+      "learning_rate": 1.6560551898193637e-05,
+      "loss": 0.6914,
+      "step": 12482
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.376173887442856,
+      "learning_rate": 1.655997600215054e-05,
+      "loss": 0.6689,
+      "step": 12483
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.753222211845242,
+      "learning_rate": 1.6559400067913187e-05,
+      "loss": 0.783,
+      "step": 12484
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5311970327068745,
+      "learning_rate": 1.655882409548493e-05,
+      "loss": 0.7119,
+      "step": 12485
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5810114080818338,
+      "learning_rate": 1.6558248084869125e-05,
+      "loss": 0.8737,
+      "step": 12486
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3721301420198095,
+      "learning_rate": 1.6557672036069122e-05,
+      "loss": 0.8328,
+      "step": 12487
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7265726126164587,
+      "learning_rate": 1.6557095949088275e-05,
+      "loss": 0.7462,
+      "step": 12488
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5100491100661644,
+      "learning_rate": 1.655651982392994e-05,
+      "loss": 0.7683,
+      "step": 12489
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.416346509992339,
+      "learning_rate": 1.655594366059747e-05,
+      "loss": 0.7198,
+      "step": 12490
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.536282228422358,
+      "learning_rate": 1.6555367459094223e-05,
+      "loss": 0.8619,
+      "step": 12491
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.496556136220742,
+      "learning_rate": 1.6554791219423548e-05,
+      "loss": 0.735,
+      "step": 12492
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6749905040819666,
+      "learning_rate": 1.6554214941588807e-05,
+      "loss": 0.7484,
+      "step": 12493
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3523983973903249,
+      "learning_rate": 1.655363862559335e-05,
+      "loss": 0.7278,
+      "step": 12494
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3418697013231045,
+      "learning_rate": 1.6553062271440536e-05,
+      "loss": 0.621,
+      "step": 12495
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7829552846227958,
+      "learning_rate": 1.6552485879133714e-05,
+      "loss": 0.6589,
+      "step": 12496
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4477693887400558,
+      "learning_rate": 1.655190944867625e-05,
+      "loss": 0.7506,
+      "step": 12497
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6734260621868038,
+      "learning_rate": 1.6551332980071492e-05,
+      "loss": 0.7828,
+      "step": 12498
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7883858561523975,
+      "learning_rate": 1.6550756473322797e-05,
+      "loss": 0.7986,
+      "step": 12499
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5886561281343397,
+      "learning_rate": 1.6550179928433527e-05,
+      "loss": 0.7829,
+      "step": 12500
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3976761951335774,
+      "learning_rate": 1.6549603345407035e-05,
+      "loss": 0.7997,
+      "step": 12501
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.017207514322362,
+      "learning_rate": 1.654902672424668e-05,
+      "loss": 0.6401,
+      "step": 12502
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8196221088469047,
+      "learning_rate": 1.6548450064955815e-05,
+      "loss": 0.8292,
+      "step": 12503
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4854520487896312,
+      "learning_rate": 1.6547873367537805e-05,
+      "loss": 0.7619,
+      "step": 12504
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8412516887304364,
+      "learning_rate": 1.6547296631996e-05,
+      "loss": 0.7874,
+      "step": 12505
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5357543356631285,
+      "learning_rate": 1.654671985833376e-05,
+      "loss": 0.7215,
+      "step": 12506
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4554590558657303,
+      "learning_rate": 1.6546143046554447e-05,
+      "loss": 0.7748,
+      "step": 12507
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5823281933249371,
+      "learning_rate": 1.6545566196661416e-05,
+      "loss": 0.7738,
+      "step": 12508
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6200458979769838,
+      "learning_rate": 1.6544989308658025e-05,
+      "loss": 0.758,
+      "step": 12509
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8493347339270996,
+      "learning_rate": 1.6544412382547638e-05,
+      "loss": 0.7013,
+      "step": 12510
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.768276553121525,
+      "learning_rate": 1.6543835418333606e-05,
+      "loss": 0.7456,
+      "step": 12511
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7815491851990957,
+      "learning_rate": 1.6543258416019293e-05,
+      "loss": 0.8302,
+      "step": 12512
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8170601374587674,
+      "learning_rate": 1.654268137560806e-05,
+      "loss": 0.723,
+      "step": 12513
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4876471607802577,
+      "learning_rate": 1.6542104297103264e-05,
+      "loss": 0.7943,
+      "step": 12514
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4909249247121623,
+      "learning_rate": 1.6541527180508265e-05,
+      "loss": 0.7199,
+      "step": 12515
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4302146546512382,
+      "learning_rate": 1.6540950025826424e-05,
+      "loss": 0.7558,
+      "step": 12516
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4112570540367162,
+      "learning_rate": 1.65403728330611e-05,
+      "loss": 0.8276,
+      "step": 12517
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6885364635237239,
+      "learning_rate": 1.6539795602215655e-05,
+      "loss": 0.8612,
+      "step": 12518
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6447659356661712,
+      "learning_rate": 1.6539218333293453e-05,
+      "loss": 0.726,
+      "step": 12519
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7304955755822875,
+      "learning_rate": 1.653864102629785e-05,
+      "loss": 0.872,
+      "step": 12520
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3107682609075109,
+      "learning_rate": 1.6538063681232203e-05,
+      "loss": 0.6582,
+      "step": 12521
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.447675656393989,
+      "learning_rate": 1.6537486298099884e-05,
+      "loss": 0.7036,
+      "step": 12522
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3833653386536349,
+      "learning_rate": 1.653690887690425e-05,
+      "loss": 0.8047,
+      "step": 12523
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3963426326296748,
+      "learning_rate": 1.6536331417648664e-05,
+      "loss": 0.6905,
+      "step": 12524
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5414664754474448,
+      "learning_rate": 1.653575392033649e-05,
+      "loss": 0.6976,
+      "step": 12525
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.426158760745678,
+      "learning_rate": 1.6535176384971082e-05,
+      "loss": 0.6635,
+      "step": 12526
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.562721597717574,
+      "learning_rate": 1.653459881155581e-05,
+      "loss": 0.6763,
+      "step": 12527
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7166739850736374,
+      "learning_rate": 1.6534021200094033e-05,
+      "loss": 0.7865,
+      "step": 12528
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8753232335136103,
+      "learning_rate": 1.6533443550589116e-05,
+      "loss": 0.7802,
+      "step": 12529
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7299471013089773,
+      "learning_rate": 1.6532865863044424e-05,
+      "loss": 0.8557,
+      "step": 12530
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5992002726205674,
+      "learning_rate": 1.653228813746332e-05,
+      "loss": 0.5228,
+      "step": 12531
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4738735614025609,
+      "learning_rate": 1.6531710373849166e-05,
+      "loss": 0.6655,
+      "step": 12532
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3114046718428298,
+      "learning_rate": 1.6531132572205324e-05,
+      "loss": 0.8489,
+      "step": 12533
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1868395566859467,
+      "learning_rate": 1.653055473253516e-05,
+      "loss": 0.6447,
+      "step": 12534
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5800905059376837,
+      "learning_rate": 1.6529976854842042e-05,
+      "loss": 0.7147,
+      "step": 12535
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6341503339773349,
+      "learning_rate": 1.6529398939129328e-05,
+      "loss": 0.5431,
+      "step": 12536
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6185481122779715,
+      "learning_rate": 1.652882098540039e-05,
+      "loss": 0.6791,
+      "step": 12537
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6481869063892396,
+      "learning_rate": 1.6528242993658586e-05,
+      "loss": 0.871,
+      "step": 12538
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.321182242385398,
+      "learning_rate": 1.6527664963907288e-05,
+      "loss": 0.7287,
+      "step": 12539
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6088753225292307,
+      "learning_rate": 1.652708689614985e-05,
+      "loss": 0.5298,
+      "step": 12540
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4326489797650077,
+      "learning_rate": 1.6526508790389655e-05,
+      "loss": 0.9463,
+      "step": 12541
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5908523484299468,
+      "learning_rate": 1.6525930646630057e-05,
+      "loss": 0.7095,
+      "step": 12542
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4441980612651402,
+      "learning_rate": 1.652535246487442e-05,
+      "loss": 0.7262,
+      "step": 12543
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5839564750309028,
+      "learning_rate": 1.652477424512612e-05,
+      "loss": 0.6798,
+      "step": 12544
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3128101281034492,
+      "learning_rate": 1.652419598738852e-05,
+      "loss": 0.6361,
+      "step": 12545
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5396075458769838,
+      "learning_rate": 1.6523617691664982e-05,
+      "loss": 0.8191,
+      "step": 12546
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5837997829357215,
+      "learning_rate": 1.6523039357958876e-05,
+      "loss": 0.7351,
+      "step": 12547
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.203487612011781,
+      "learning_rate": 1.652246098627357e-05,
+      "loss": 0.8237,
+      "step": 12548
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6912924973435788,
+      "learning_rate": 1.6521882576612433e-05,
+      "loss": 0.8274,
+      "step": 12549
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6587702310547723,
+      "learning_rate": 1.6521304128978826e-05,
+      "loss": 0.6826,
+      "step": 12550
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5405235124323489,
+      "learning_rate": 1.652072564337613e-05,
+      "loss": 0.7073,
+      "step": 12551
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5618744411095021,
+      "learning_rate": 1.6520147119807698e-05,
+      "loss": 0.5564,
+      "step": 12552
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4930753933581935,
+      "learning_rate": 1.6519568558276908e-05,
+      "loss": 0.8365,
+      "step": 12553
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5174173455536721,
+      "learning_rate": 1.6518989958787126e-05,
+      "loss": 0.7581,
+      "step": 12554
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4388192586238677,
+      "learning_rate": 1.651841132134172e-05,
+      "loss": 0.7919,
+      "step": 12555
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6742529303501048,
+      "learning_rate": 1.651783264594406e-05,
+      "loss": 0.7958,
+      "step": 12556
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5243019435196792,
+      "learning_rate": 1.6517253932597515e-05,
+      "loss": 0.8694,
+      "step": 12557
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.348767193312487,
+      "learning_rate": 1.6516675181305456e-05,
+      "loss": 0.906,
+      "step": 12558
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5686100876611118,
+      "learning_rate": 1.6516096392071248e-05,
+      "loss": 0.7061,
+      "step": 12559
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.419898660440833,
+      "learning_rate": 1.651551756489826e-05,
+      "loss": 0.8454,
+      "step": 12560
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3382179185568412,
+      "learning_rate": 1.6514938699789877e-05,
+      "loss": 0.6887,
+      "step": 12561
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.732856972160893,
+      "learning_rate": 1.651435979674945e-05,
+      "loss": 0.7371,
+      "step": 12562
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3725286953231832,
+      "learning_rate": 1.6513780855780358e-05,
+      "loss": 0.7509,
+      "step": 12563
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4533268860948456,
+      "learning_rate": 1.6513201876885972e-05,
+      "loss": 0.7674,
+      "step": 12564
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8347604881932684,
+      "learning_rate": 1.6512622860069667e-05,
+      "loss": 0.81,
+      "step": 12565
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.485835924646397,
+      "learning_rate": 1.6512043805334808e-05,
+      "loss": 0.7281,
+      "step": 12566
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5923981697012117,
+      "learning_rate": 1.6511464712684764e-05,
+      "loss": 0.7339,
+      "step": 12567
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3742797507817548,
+      "learning_rate": 1.6510885582122912e-05,
+      "loss": 0.781,
+      "step": 12568
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3280798678561638,
+      "learning_rate": 1.651030641365263e-05,
+      "loss": 0.8336,
+      "step": 12569
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3063117079018212,
+      "learning_rate": 1.6509727207277273e-05,
+      "loss": 0.7813,
+      "step": 12570
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2387265768449174,
+      "learning_rate": 1.650914796300023e-05,
+      "loss": 0.7966,
+      "step": 12571
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5994053491678943,
+      "learning_rate": 1.6508568680824865e-05,
+      "loss": 0.5448,
+      "step": 12572
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5212015984999099,
+      "learning_rate": 1.6507989360754553e-05,
+      "loss": 0.7571,
+      "step": 12573
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6671757380778043,
+      "learning_rate": 1.6507410002792665e-05,
+      "loss": 0.7137,
+      "step": 12574
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.486637214102326,
+      "learning_rate": 1.6506830606942578e-05,
+      "loss": 0.7254,
+      "step": 12575
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5399913336812574,
+      "learning_rate": 1.650625117320766e-05,
+      "loss": 0.8485,
+      "step": 12576
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4068301148899351,
+      "learning_rate": 1.650567170159129e-05,
+      "loss": 0.7641,
+      "step": 12577
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4910576055581832,
+      "learning_rate": 1.6505092192096837e-05,
+      "loss": 0.8576,
+      "step": 12578
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.037793076597071,
+      "learning_rate": 1.650451264472768e-05,
+      "loss": 0.8151,
+      "step": 12579
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.646824481211278,
+      "learning_rate": 1.6503933059487195e-05,
+      "loss": 0.6856,
+      "step": 12580
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3396334794844595,
+      "learning_rate": 1.6503353436378747e-05,
+      "loss": 0.699,
+      "step": 12581
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.319451211718508,
+      "learning_rate": 1.6502773775405717e-05,
+      "loss": 0.7661,
+      "step": 12582
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6593635098379351,
+      "learning_rate": 1.650219407657148e-05,
+      "loss": 0.753,
+      "step": 12583
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.727248762149179,
+      "learning_rate": 1.6501614339879413e-05,
+      "loss": 0.823,
+      "step": 12584
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2784734626287815,
+      "learning_rate": 1.650103456533289e-05,
+      "loss": 0.8103,
+      "step": 12585
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5939433429973724,
+      "learning_rate": 1.650045475293528e-05,
+      "loss": 0.5478,
+      "step": 12586
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5824941293157229,
+      "learning_rate": 1.6499874902689968e-05,
+      "loss": 0.7739,
+      "step": 12587
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7340036873170055,
+      "learning_rate": 1.6499295014600327e-05,
+      "loss": 0.8269,
+      "step": 12588
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4136323677218983,
+      "learning_rate": 1.649871508866973e-05,
+      "loss": 0.7421,
+      "step": 12589
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.504889925752166,
+      "learning_rate": 1.6498135124901557e-05,
+      "loss": 0.8143,
+      "step": 12590
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4087525985614886,
+      "learning_rate": 1.6497555123299184e-05,
+      "loss": 0.702,
+      "step": 12591
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.611141821194816,
+      "learning_rate": 1.649697508386599e-05,
+      "loss": 0.7183,
+      "step": 12592
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.365905141312535,
+      "learning_rate": 1.6496395006605346e-05,
+      "loss": 0.7759,
+      "step": 12593
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9869445986755079,
+      "learning_rate": 1.6495814891520636e-05,
+      "loss": 0.6651,
+      "step": 12594
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1779608596500997,
+      "learning_rate": 1.6495234738615238e-05,
+      "loss": 0.833,
+      "step": 12595
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.814374227522818,
+      "learning_rate": 1.6494654547892522e-05,
+      "loss": 0.6507,
+      "step": 12596
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4787825647129824,
+      "learning_rate": 1.6494074319355874e-05,
+      "loss": 0.8341,
+      "step": 12597
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.371556007963213,
+      "learning_rate": 1.649349405300867e-05,
+      "loss": 0.7643,
+      "step": 12598
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7919173833438302,
+      "learning_rate": 1.6492913748854285e-05,
+      "loss": 0.9536,
+      "step": 12599
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4336473978965474,
+      "learning_rate": 1.6492333406896103e-05,
+      "loss": 0.747,
+      "step": 12600
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7735426558211078,
+      "learning_rate": 1.6491753027137497e-05,
+      "loss": 0.7793,
+      "step": 12601
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.751275867106916,
+      "learning_rate": 1.6491172609581856e-05,
+      "loss": 0.9754,
+      "step": 12602
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.3613220360498355,
+      "learning_rate": 1.649059215423255e-05,
+      "loss": 0.7071,
+      "step": 12603
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2937813706573333,
+      "learning_rate": 1.649001166109296e-05,
+      "loss": 0.8741,
+      "step": 12604
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9737918617200225,
+      "learning_rate": 1.648943113016647e-05,
+      "loss": 0.8517,
+      "step": 12605
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2400435155989984,
+      "learning_rate": 1.648885056145645e-05,
+      "loss": 0.7122,
+      "step": 12606
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2490096504827701,
+      "learning_rate": 1.6488269954966295e-05,
+      "loss": 0.7909,
+      "step": 12607
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7493856394871357,
+      "learning_rate": 1.6487689310699377e-05,
+      "loss": 0.6871,
+      "step": 12608
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3885090360944499,
+      "learning_rate": 1.648710862865908e-05,
+      "loss": 0.7958,
+      "step": 12609
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6538810451723265,
+      "learning_rate": 1.648652790884878e-05,
+      "loss": 0.7569,
+      "step": 12610
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3715946660821279,
+      "learning_rate": 1.648594715127186e-05,
+      "loss": 0.7302,
+      "step": 12611
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4332208503406727,
+      "learning_rate": 1.64853663559317e-05,
+      "loss": 0.6353,
+      "step": 12612
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4945101403183734,
+      "learning_rate": 1.648478552283169e-05,
+      "loss": 0.7163,
+      "step": 12613
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3252021667089566,
+      "learning_rate": 1.6484204651975203e-05,
+      "loss": 0.6804,
+      "step": 12614
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5961009358151292,
+      "learning_rate": 1.6483623743365624e-05,
+      "loss": 0.5342,
+      "step": 12615
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5726419046533138,
+      "learning_rate": 1.6483042797006334e-05,
+      "loss": 0.7752,
+      "step": 12616
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5031897449494873,
+      "learning_rate": 1.6482461812900718e-05,
+      "loss": 0.8441,
+      "step": 12617
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4490062596279782,
+      "learning_rate": 1.6481880791052158e-05,
+      "loss": 0.775,
+      "step": 12618
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.577360423420799,
+      "learning_rate": 1.648129973146403e-05,
+      "loss": 0.8357,
+      "step": 12619
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4916617082657342,
+      "learning_rate": 1.6480718634139726e-05,
+      "loss": 0.8217,
+      "step": 12620
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6036460587324254,
+      "learning_rate": 1.648013749908263e-05,
+      "loss": 0.577,
+      "step": 12621
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.555491544060486,
+      "learning_rate": 1.6479556326296116e-05,
+      "loss": 0.5175,
+      "step": 12622
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.503537652716524,
+      "learning_rate": 1.6478975115783577e-05,
+      "loss": 0.5916,
+      "step": 12623
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1129094378782645,
+      "learning_rate": 1.6478393867548396e-05,
+      "loss": 0.7211,
+      "step": 12624
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5798338279598427,
+      "learning_rate": 1.647781258159395e-05,
+      "loss": 0.5673,
+      "step": 12625
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0715422598641386,
+      "learning_rate": 1.647723125792363e-05,
+      "loss": 0.7451,
+      "step": 12626
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5543668262317072,
+      "learning_rate": 1.647664989654082e-05,
+      "loss": 0.7324,
+      "step": 12627
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.541941906255393,
+      "learning_rate": 1.6476068497448903e-05,
+      "loss": 0.7536,
+      "step": 12628
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.658895813644724,
+      "learning_rate": 1.6475487060651265e-05,
+      "loss": 0.8664,
+      "step": 12629
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.502873938636823,
+      "learning_rate": 1.6474905586151293e-05,
+      "loss": 0.683,
+      "step": 12630
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7081517018323353,
+      "learning_rate": 1.647432407395237e-05,
+      "loss": 0.6225,
+      "step": 12631
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.431788221585897,
+      "learning_rate": 1.647374252405788e-05,
+      "loss": 0.8161,
+      "step": 12632
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8720346845572595,
+      "learning_rate": 1.6473160936471214e-05,
+      "loss": 0.83,
+      "step": 12633
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8656248959096455,
+      "learning_rate": 1.6472579311195752e-05,
+      "loss": 0.7427,
+      "step": 12634
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5921347923091558,
+      "learning_rate": 1.6471997648234886e-05,
+      "loss": 0.531,
+      "step": 12635
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2969708211157376,
+      "learning_rate": 1.6471415947592e-05,
+      "loss": 0.7485,
+      "step": 12636
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6660473552444608,
+      "learning_rate": 1.647083420927049e-05,
+      "loss": 0.8427,
+      "step": 12637
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7271397013222487,
+      "learning_rate": 1.6470252433273723e-05,
+      "loss": 0.7455,
+      "step": 12638
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5657847210234663,
+      "learning_rate": 1.6469670619605103e-05,
+      "loss": 0.7961,
+      "step": 12639
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5980822042730178,
+      "learning_rate": 1.6469088768268013e-05,
+      "loss": 0.7357,
+      "step": 12640
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.447208645457654,
+      "learning_rate": 1.646850687926584e-05,
+      "loss": 0.897,
+      "step": 12641
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3767265891850269,
+      "learning_rate": 1.646792495260197e-05,
+      "loss": 0.7726,
+      "step": 12642
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.502875035824736,
+      "learning_rate": 1.6467342988279793e-05,
+      "loss": 0.7962,
+      "step": 12643
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.329064803235716,
+      "learning_rate": 1.64667609863027e-05,
+      "loss": 0.6582,
+      "step": 12644
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3173598315719084,
+      "learning_rate": 1.646617894667408e-05,
+      "loss": 0.763,
+      "step": 12645
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4542545996499494,
+      "learning_rate": 1.646559686939731e-05,
+      "loss": 0.7503,
+      "step": 12646
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.693709890438935,
+      "learning_rate": 1.6465014754475794e-05,
+      "loss": 0.7209,
+      "step": 12647
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.565613915323042,
+      "learning_rate": 1.6464432601912914e-05,
+      "loss": 0.5943,
+      "step": 12648
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3720049052794787,
+      "learning_rate": 1.6463850411712063e-05,
+      "loss": 0.6233,
+      "step": 12649
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8569779753663107,
+      "learning_rate": 1.6463268183876627e-05,
+      "loss": 0.7528,
+      "step": 12650
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7263322823434668,
+      "learning_rate": 1.6462685918409998e-05,
+      "loss": 0.697,
+      "step": 12651
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6291930648748167,
+      "learning_rate": 1.6462103615315563e-05,
+      "loss": 0.7859,
+      "step": 12652
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5330929910703275,
+      "learning_rate": 1.6461521274596717e-05,
+      "loss": 0.8173,
+      "step": 12653
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3719206962387795,
+      "learning_rate": 1.6460938896256848e-05,
+      "loss": 0.8294,
+      "step": 12654
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4344288184995293,
+      "learning_rate": 1.6460356480299345e-05,
+      "loss": 0.864,
+      "step": 12655
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5830242977435437,
+      "learning_rate": 1.6459774026727606e-05,
+      "loss": 0.4902,
+      "step": 12656
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.484800637857784,
+      "learning_rate": 1.6459191535545013e-05,
+      "loss": 0.7857,
+      "step": 12657
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5226191223607755,
+      "learning_rate": 1.6458609006754966e-05,
+      "loss": 0.7821,
+      "step": 12658
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5292682720106952,
+      "learning_rate": 1.6458026440360847e-05,
+      "loss": 0.6801,
+      "step": 12659
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7163768373459896,
+      "learning_rate": 1.6457443836366062e-05,
+      "loss": 0.7697,
+      "step": 12660
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7631803154007044,
+      "learning_rate": 1.6456861194773988e-05,
+      "loss": 0.7811,
+      "step": 12661
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.691789787027461,
+      "learning_rate": 1.6456278515588023e-05,
+      "loss": 0.9918,
+      "step": 12662
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4382629551153268,
+      "learning_rate": 1.6455695798811567e-05,
+      "loss": 0.7881,
+      "step": 12663
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3341931228009387,
+      "learning_rate": 1.6455113044448e-05,
+      "loss": 0.6924,
+      "step": 12664
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6514399893206455,
+      "learning_rate": 1.6454530252500723e-05,
+      "loss": 0.8498,
+      "step": 12665
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6227262797508155,
+      "learning_rate": 1.645394742297313e-05,
+      "loss": 0.5455,
+      "step": 12666
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.574314964181123,
+      "learning_rate": 1.645336455586861e-05,
+      "loss": 0.6837,
+      "step": 12667
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7352391849113604,
+      "learning_rate": 1.645278165119056e-05,
+      "loss": 0.8016,
+      "step": 12668
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.427145156515446,
+      "learning_rate": 1.6452198708942368e-05,
+      "loss": 0.6293,
+      "step": 12669
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3365351885341092,
+      "learning_rate": 1.6451615729127436e-05,
+      "loss": 0.7482,
+      "step": 12670
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6113100576106526,
+      "learning_rate": 1.6451032711749154e-05,
+      "loss": 0.5852,
+      "step": 12671
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3896053142736118,
+      "learning_rate": 1.6450449656810914e-05,
+      "loss": 0.6902,
+      "step": 12672
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6085692980633106,
+      "learning_rate": 1.6449866564316118e-05,
+      "loss": 0.5663,
+      "step": 12673
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0484471086855076,
+      "learning_rate": 1.6449283434268153e-05,
+      "loss": 0.7234,
+      "step": 12674
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0915901927166876,
+      "learning_rate": 1.644870026667042e-05,
+      "loss": 0.7011,
+      "step": 12675
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6291739812246588,
+      "learning_rate": 1.6448117061526314e-05,
+      "loss": 0.5456,
+      "step": 12676
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5476381657751637,
+      "learning_rate": 1.644753381883923e-05,
+      "loss": 0.6123,
+      "step": 12677
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3976560270664498,
+      "learning_rate": 1.644695053861256e-05,
+      "loss": 0.8555,
+      "step": 12678
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.816236273722308,
+      "learning_rate": 1.6446367220849704e-05,
+      "loss": 0.7881,
+      "step": 12679
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5122782006101032,
+      "learning_rate": 1.6445783865554056e-05,
+      "loss": 0.8304,
+      "step": 12680
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.36046885171472,
+      "learning_rate": 1.6445200472729016e-05,
+      "loss": 0.7593,
+      "step": 12681
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5254435793352354,
+      "learning_rate": 1.6444617042377978e-05,
+      "loss": 0.8133,
+      "step": 12682
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8823020075626802,
+      "learning_rate": 1.6444033574504336e-05,
+      "loss": 0.7383,
+      "step": 12683
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5361890663943452,
+      "learning_rate": 1.644345006911149e-05,
+      "loss": 0.8082,
+      "step": 12684
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3832331531620696,
+      "learning_rate": 1.6442866526202842e-05,
+      "loss": 0.8177,
+      "step": 12685
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.422180532022622,
+      "learning_rate": 1.6442282945781784e-05,
+      "loss": 0.6686,
+      "step": 12686
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.357460539628638,
+      "learning_rate": 1.6441699327851715e-05,
+      "loss": 0.6454,
+      "step": 12687
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3824900622942953,
+      "learning_rate": 1.6441115672416033e-05,
+      "loss": 0.9411,
+      "step": 12688
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0903812226448886,
+      "learning_rate": 1.6440531979478137e-05,
+      "loss": 0.6542,
+      "step": 12689
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3765043006135629,
+      "learning_rate": 1.643994824904142e-05,
+      "loss": 0.7776,
+      "step": 12690
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6492974437160912,
+      "learning_rate": 1.6439364481109292e-05,
+      "loss": 0.7549,
+      "step": 12691
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4466311452499292,
+      "learning_rate": 1.643878067568514e-05,
+      "loss": 0.873,
+      "step": 12692
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5022863573867113,
+      "learning_rate": 1.6438196832772372e-05,
+      "loss": 0.7751,
+      "step": 12693
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4668098996285115,
+      "learning_rate": 1.6437612952374382e-05,
+      "loss": 0.741,
+      "step": 12694
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7325856682294427,
+      "learning_rate": 1.6437029034494574e-05,
+      "loss": 0.7349,
+      "step": 12695
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6657567087259928,
+      "learning_rate": 1.6436445079136344e-05,
+      "loss": 0.694,
+      "step": 12696
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5187974961364201,
+      "learning_rate": 1.6435861086303092e-05,
+      "loss": 0.767,
+      "step": 12697
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5424944909289515,
+      "learning_rate": 1.643527705599822e-05,
+      "loss": 0.7359,
+      "step": 12698
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4223620446935075,
+      "learning_rate": 1.6434692988225125e-05,
+      "loss": 0.7173,
+      "step": 12699
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5031161451629567,
+      "learning_rate": 1.6434108882987213e-05,
+      "loss": 0.9028,
+      "step": 12700
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.560981343123061,
+      "learning_rate": 1.643352474028788e-05,
+      "loss": 0.7559,
+      "step": 12701
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4319334428794697,
+      "learning_rate": 1.643294056013053e-05,
+      "loss": 0.8835,
+      "step": 12702
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5850709719342324,
+      "learning_rate": 1.6432356342518564e-05,
+      "loss": 0.7568,
+      "step": 12703
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5242087987689072,
+      "learning_rate": 1.6431772087455384e-05,
+      "loss": 0.8192,
+      "step": 12704
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.445051006585381,
+      "learning_rate": 1.643118779494439e-05,
+      "loss": 0.8105,
+      "step": 12705
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.397771513496922,
+      "learning_rate": 1.6430603464988982e-05,
+      "loss": 0.7042,
+      "step": 12706
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3964013778695707,
+      "learning_rate": 1.6430019097592567e-05,
+      "loss": 0.6864,
+      "step": 12707
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.455341247245783,
+      "learning_rate": 1.6429434692758542e-05,
+      "loss": 0.8401,
+      "step": 12708
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3235159768045826,
+      "learning_rate": 1.6428850250490317e-05,
+      "loss": 0.6868,
+      "step": 12709
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7633031806618014,
+      "learning_rate": 1.642826577079129e-05,
+      "loss": 0.7574,
+      "step": 12710
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7436111287291016,
+      "learning_rate": 1.6427681253664863e-05,
+      "loss": 0.7581,
+      "step": 12711
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4806439087347574,
+      "learning_rate": 1.642709669911444e-05,
+      "loss": 0.6199,
+      "step": 12712
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5322600069507923,
+      "learning_rate": 1.642651210714343e-05,
+      "loss": 0.7366,
+      "step": 12713
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4610550809090561,
+      "learning_rate": 1.6425927477755227e-05,
+      "loss": 0.6627,
+      "step": 12714
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4314297600831307,
+      "learning_rate": 1.642534281095324e-05,
+      "loss": 0.742,
+      "step": 12715
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5595160704715563,
+      "learning_rate": 1.6424758106740874e-05,
+      "loss": 0.8088,
+      "step": 12716
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4661147116543753,
+      "learning_rate": 1.642417336512153e-05,
+      "loss": 0.7682,
+      "step": 12717
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6543982862678043,
+      "learning_rate": 1.642358858609862e-05,
+      "loss": 0.4826,
+      "step": 12718
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6667731288917877,
+      "learning_rate": 1.642300376967554e-05,
+      "loss": 0.8241,
+      "step": 12719
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3513370866477186,
+      "learning_rate": 1.64224189158557e-05,
+      "loss": 0.7183,
+      "step": 12720
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6618706863636789,
+      "learning_rate": 1.64218340246425e-05,
+      "loss": 0.5095,
+      "step": 12721
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5024153562955385,
+      "learning_rate": 1.6421249096039352e-05,
+      "loss": 0.7447,
+      "step": 12722
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.346901395502714,
+      "learning_rate": 1.6420664130049663e-05,
+      "loss": 0.7806,
+      "step": 12723
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4333609129776255,
+      "learning_rate": 1.6420079126676828e-05,
+      "loss": 0.581,
+      "step": 12724
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4114799948446992,
+      "learning_rate": 1.6419494085924267e-05,
+      "loss": 0.6891,
+      "step": 12725
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6154378203964758,
+      "learning_rate": 1.6418909007795374e-05,
+      "loss": 0.5779,
+      "step": 12726
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5391221399573602,
+      "learning_rate": 1.641832389229356e-05,
+      "loss": 0.8063,
+      "step": 12727
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4162818048899848,
+      "learning_rate": 1.6417738739422234e-05,
+      "loss": 0.7988,
+      "step": 12728
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2429336223678928,
+      "learning_rate": 1.64171535491848e-05,
+      "loss": 0.7498,
+      "step": 12729
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5340127132831782,
+      "learning_rate": 1.6416568321584668e-05,
+      "loss": 0.7446,
+      "step": 12730
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5758097462522551,
+      "learning_rate": 1.641598305662524e-05,
+      "loss": 0.5943,
+      "step": 12731
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5049219284003827,
+      "learning_rate": 1.641539775430993e-05,
+      "loss": 0.6883,
+      "step": 12732
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.426332721052975,
+      "learning_rate": 1.6414812414642144e-05,
+      "loss": 0.7719,
+      "step": 12733
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6000462860627572,
+      "learning_rate": 1.6414227037625288e-05,
+      "loss": 0.5557,
+      "step": 12734
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.345436243848167,
+      "learning_rate": 1.6413641623262774e-05,
+      "loss": 0.7611,
+      "step": 12735
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4498433796877417,
+      "learning_rate": 1.6413056171558006e-05,
+      "loss": 0.7654,
+      "step": 12736
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4085762044172356,
+      "learning_rate": 1.6412470682514396e-05,
+      "loss": 0.7073,
+      "step": 12737
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4419899799741307,
+      "learning_rate": 1.641188515613535e-05,
+      "loss": 0.7559,
+      "step": 12738
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.413601619141124,
+      "learning_rate": 1.6411299592424276e-05,
+      "loss": 0.771,
+      "step": 12739
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7803705277454325,
+      "learning_rate": 1.641071399138459e-05,
+      "loss": 0.7351,
+      "step": 12740
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5327742792803283,
+      "learning_rate": 1.6410128353019697e-05,
+      "loss": 0.7202,
+      "step": 12741
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4801554549973226,
+      "learning_rate": 1.6409542677333007e-05,
+      "loss": 0.7206,
+      "step": 12742
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.361695508366503,
+      "learning_rate": 1.640895696432793e-05,
+      "loss": 0.7344,
+      "step": 12743
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6990249702280555,
+      "learning_rate": 1.6408371214007876e-05,
+      "loss": 0.7825,
+      "step": 12744
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.331665606200579,
+      "learning_rate": 1.6407785426376258e-05,
+      "loss": 0.7093,
+      "step": 12745
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.5195919055669993,
+      "learning_rate": 1.640719960143648e-05,
+      "loss": 0.7628,
+      "step": 12746
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4680709393905402,
+      "learning_rate": 1.6406613739191964e-05,
+      "loss": 0.8045,
+      "step": 12747
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5604710065511407,
+      "learning_rate": 1.640602783964611e-05,
+      "loss": 0.799,
+      "step": 12748
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.6789400336053366,
+      "learning_rate": 1.6405441902802334e-05,
+      "loss": 0.7748,
+      "step": 12749
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5691064622200217,
+      "learning_rate": 1.640485592866405e-05,
+      "loss": 0.7449,
+      "step": 12750
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6161618718084773,
+      "learning_rate": 1.6404269917234662e-05,
+      "loss": 0.5424,
+      "step": 12751
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5887676125615936,
+      "learning_rate": 1.6403683868517593e-05,
+      "loss": 0.5979,
+      "step": 12752
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3127843348679897,
+      "learning_rate": 1.6403097782516244e-05,
+      "loss": 0.8392,
+      "step": 12753
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5460178544363963,
+      "learning_rate": 1.6402511659234038e-05,
+      "loss": 0.7154,
+      "step": 12754
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3270329951021755,
+      "learning_rate": 1.640192549867438e-05,
+      "loss": 0.6975,
+      "step": 12755
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7546653809550794,
+      "learning_rate": 1.6401339300840683e-05,
+      "loss": 0.7311,
+      "step": 12756
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8823135875860588,
+      "learning_rate": 1.640075306573636e-05,
+      "loss": 0.7411,
+      "step": 12757
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6086513176944914,
+      "learning_rate": 1.6400166793364828e-05,
+      "loss": 0.7747,
+      "step": 12758
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5734599738562682,
+      "learning_rate": 1.63995804837295e-05,
+      "loss": 0.8652,
+      "step": 12759
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5789890861975475,
+      "learning_rate": 1.639899413683379e-05,
+      "loss": 0.6515,
+      "step": 12760
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8301688308435253,
+      "learning_rate": 1.6398407752681108e-05,
+      "loss": 0.7983,
+      "step": 12761
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4040179653353346,
+      "learning_rate": 1.6397821331274872e-05,
+      "loss": 0.825,
+      "step": 12762
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9645277900576397,
+      "learning_rate": 1.639723487261849e-05,
+      "loss": 0.7358,
+      "step": 12763
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5480603503708714,
+      "learning_rate": 1.639664837671539e-05,
+      "loss": 0.7982,
+      "step": 12764
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.568977794941444,
+      "learning_rate": 1.6396061843568973e-05,
+      "loss": 0.829,
+      "step": 12765
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3460894166990267,
+      "learning_rate": 1.6395475273182655e-05,
+      "loss": 0.9028,
+      "step": 12766
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6871272642194235,
+      "learning_rate": 1.6394888665559857e-05,
+      "loss": 0.7656,
+      "step": 12767
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4191750355566815,
+      "learning_rate": 1.6394302020703995e-05,
+      "loss": 0.7034,
+      "step": 12768
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4216394058139663,
+      "learning_rate": 1.6393715338618483e-05,
+      "loss": 0.8034,
+      "step": 12769
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6587168915492432,
+      "learning_rate": 1.6393128619306734e-05,
+      "loss": 0.841,
+      "step": 12770
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7406302186199343,
+      "learning_rate": 1.6392541862772166e-05,
+      "loss": 0.6833,
+      "step": 12771
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5754948684505176,
+      "learning_rate": 1.6391955069018198e-05,
+      "loss": 0.5676,
+      "step": 12772
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.474216727652815,
+      "learning_rate": 1.639136823804824e-05,
+      "loss": 0.6897,
+      "step": 12773
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.544351687117115,
+      "learning_rate": 1.6390781369865715e-05,
+      "loss": 0.7685,
+      "step": 12774
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5229824525155162,
+      "learning_rate": 1.6390194464474038e-05,
+      "loss": 0.7364,
+      "step": 12775
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4890368613412897,
+      "learning_rate": 1.6389607521876625e-05,
+      "loss": 0.7158,
+      "step": 12776
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4883751420318296,
+      "learning_rate": 1.6389020542076898e-05,
+      "loss": 0.7198,
+      "step": 12777
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5720434387254966,
+      "learning_rate": 1.6388433525078266e-05,
+      "loss": 0.5394,
+      "step": 12778
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4130708388048838,
+      "learning_rate": 1.6387846470884154e-05,
+      "loss": 0.7579,
+      "step": 12779
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.633263969238192,
+      "learning_rate": 1.6387259379497973e-05,
+      "loss": 0.7346,
+      "step": 12780
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3686520697769995,
+      "learning_rate": 1.638667225092315e-05,
+      "loss": 0.8557,
+      "step": 12781
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4340807368017277,
+      "learning_rate": 1.6386085085163096e-05,
+      "loss": 0.7925,
+      "step": 12782
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.943460749133858,
+      "learning_rate": 1.6385497882221236e-05,
+      "loss": 0.7263,
+      "step": 12783
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.681616810828777,
+      "learning_rate": 1.6384910642100983e-05,
+      "loss": 0.6254,
+      "step": 12784
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.573134157267289,
+      "learning_rate": 1.6384323364805762e-05,
+      "loss": 0.5366,
+      "step": 12785
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9806685144074108,
+      "learning_rate": 1.6383736050338984e-05,
+      "loss": 0.7466,
+      "step": 12786
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.660733509399951,
+      "learning_rate": 1.638314869870408e-05,
+      "loss": 0.7726,
+      "step": 12787
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6785033652136774,
+      "learning_rate": 1.6382561309904457e-05,
+      "loss": 0.8382,
+      "step": 12788
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.16534043757918,
+      "learning_rate": 1.6381973883943546e-05,
+      "loss": 0.7837,
+      "step": 12789
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4253040752634847,
+      "learning_rate": 1.6381386420824763e-05,
+      "loss": 0.748,
+      "step": 12790
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4274616459293676,
+      "learning_rate": 1.6380798920551526e-05,
+      "loss": 0.6673,
+      "step": 12791
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.8004965560949975,
+      "learning_rate": 1.6380211383127255e-05,
+      "loss": 0.8094,
+      "step": 12792
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8756941062556982,
+      "learning_rate": 1.637962380855538e-05,
+      "loss": 0.6778,
+      "step": 12793
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.036555599181812,
+      "learning_rate": 1.637903619683931e-05,
+      "loss": 0.678,
+      "step": 12794
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4538366879587,
+      "learning_rate": 1.6378448547982473e-05,
+      "loss": 0.6998,
+      "step": 12795
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5571206271260094,
+      "learning_rate": 1.6377860861988288e-05,
+      "loss": 0.8713,
+      "step": 12796
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5889083943272586,
+      "learning_rate": 1.6377273138860182e-05,
+      "loss": 0.9389,
+      "step": 12797
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.495220723684675,
+      "learning_rate": 1.637668537860157e-05,
+      "loss": 0.7719,
+      "step": 12798
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7844041722984794,
+      "learning_rate": 1.6376097581215874e-05,
+      "loss": 0.8,
+      "step": 12799
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.402013648072166,
+      "learning_rate": 1.6375509746706523e-05,
+      "loss": 0.7739,
+      "step": 12800
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7200511144088362,
+      "learning_rate": 1.6374921875076933e-05,
+      "loss": 0.7117,
+      "step": 12801
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5574892649337693,
+      "learning_rate": 1.6374333966330535e-05,
+      "loss": 0.7138,
+      "step": 12802
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4326406880433489,
+      "learning_rate": 1.637374602047074e-05,
+      "loss": 0.7692,
+      "step": 12803
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7235306612447392,
+      "learning_rate": 1.6373158037500984e-05,
+      "loss": 0.7792,
+      "step": 12804
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9574427759639486,
+      "learning_rate": 1.6372570017424682e-05,
+      "loss": 0.8013,
+      "step": 12805
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3910918942319142,
+      "learning_rate": 1.6371981960245257e-05,
+      "loss": 0.9126,
+      "step": 12806
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5957191280214656,
+      "learning_rate": 1.6371393865966137e-05,
+      "loss": 0.57,
+      "step": 12807
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3617967455512618,
+      "learning_rate": 1.6370805734590746e-05,
+      "loss": 0.7704,
+      "step": 12808
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2786117664916987,
+      "learning_rate": 1.6370217566122505e-05,
+      "loss": 0.7477,
+      "step": 12809
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3219113058429293,
+      "learning_rate": 1.6369629360564845e-05,
+      "loss": 0.6712,
+      "step": 12810
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6180330267786953,
+      "learning_rate": 1.6369041117921183e-05,
+      "loss": 0.7128,
+      "step": 12811
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3111213349188102,
+      "learning_rate": 1.6368452838194947e-05,
+      "loss": 0.7294,
+      "step": 12812
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4581425430435648,
+      "learning_rate": 1.6367864521389563e-05,
+      "loss": 0.8832,
+      "step": 12813
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5954457123218244,
+      "learning_rate": 1.6367276167508456e-05,
+      "loss": 0.8201,
+      "step": 12814
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.332529928059204,
+      "learning_rate": 1.636668777655505e-05,
+      "loss": 0.6407,
+      "step": 12815
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.426340206507392,
+      "learning_rate": 1.6366099348532777e-05,
+      "loss": 0.7423,
+      "step": 12816
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3617478913701946,
+      "learning_rate": 1.636551088344505e-05,
+      "loss": 0.7465,
+      "step": 12817
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4690863167471344,
+      "learning_rate": 1.636492238129531e-05,
+      "loss": 0.6764,
+      "step": 12818
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4975371695652067,
+      "learning_rate": 1.636433384208697e-05,
+      "loss": 0.7553,
+      "step": 12819
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5168814938097372,
+      "learning_rate": 1.636374526582347e-05,
+      "loss": 0.7008,
+      "step": 12820
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7604702433328248,
+      "learning_rate": 1.636315665250823e-05,
+      "loss": 0.8109,
+      "step": 12821
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5451608487629134,
+      "learning_rate": 1.636256800214467e-05,
+      "loss": 0.7236,
+      "step": 12822
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.547310597323582,
+      "learning_rate": 1.6361979314736233e-05,
+      "loss": 0.8231,
+      "step": 12823
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6402061454415766,
+      "learning_rate": 1.6361390590286336e-05,
+      "loss": 0.7281,
+      "step": 12824
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4682061814534486,
+      "learning_rate": 1.636080182879841e-05,
+      "loss": 0.8114,
+      "step": 12825
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9628463429051242,
+      "learning_rate": 1.636021303027588e-05,
+      "loss": 0.7081,
+      "step": 12826
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6254203578404374,
+      "learning_rate": 1.6359624194722176e-05,
+      "loss": 0.8222,
+      "step": 12827
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4792685685754485,
+      "learning_rate": 1.6359035322140727e-05,
+      "loss": 0.7059,
+      "step": 12828
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4619175146281294,
+      "learning_rate": 1.635844641253496e-05,
+      "loss": 0.6379,
+      "step": 12829
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.821523508528754,
+      "learning_rate": 1.6357857465908308e-05,
+      "loss": 0.6744,
+      "step": 12830
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.924773644781739,
+      "learning_rate": 1.6357268482264195e-05,
+      "loss": 0.749,
+      "step": 12831
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.719323955248161,
+      "learning_rate": 1.6356679461606052e-05,
+      "loss": 0.7263,
+      "step": 12832
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.389347466773859,
+      "learning_rate": 1.6356090403937307e-05,
+      "loss": 0.7795,
+      "step": 12833
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3935711836665545,
+      "learning_rate": 1.6355501309261398e-05,
+      "loss": 0.7066,
+      "step": 12834
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.526800517577715,
+      "learning_rate": 1.6354912177581743e-05,
+      "loss": 0.8339,
+      "step": 12835
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.714355310689883,
+      "learning_rate": 1.6354323008901774e-05,
+      "loss": 0.8337,
+      "step": 12836
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5025033941311228,
+      "learning_rate": 1.635373380322493e-05,
+      "loss": 0.8116,
+      "step": 12837
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5063348608462752,
+      "learning_rate": 1.6353144560554636e-05,
+      "loss": 0.8252,
+      "step": 12838
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.565205337025998,
+      "learning_rate": 1.6352555280894323e-05,
+      "loss": 0.7458,
+      "step": 12839
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.704145046155156,
+      "learning_rate": 1.635196596424742e-05,
+      "loss": 0.7769,
+      "step": 12840
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.164588234001027,
+      "learning_rate": 1.635137661061736e-05,
+      "loss": 0.7723,
+      "step": 12841
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4870752706172217,
+      "learning_rate": 1.6350787220007576e-05,
+      "loss": 0.9574,
+      "step": 12842
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5514087665546863,
+      "learning_rate": 1.6350197792421497e-05,
+      "loss": 0.8814,
+      "step": 12843
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2884406617608972,
+      "learning_rate": 1.634960832786256e-05,
+      "loss": 0.71,
+      "step": 12844
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6017626278526382,
+      "learning_rate": 1.6349018826334188e-05,
+      "loss": 0.5344,
+      "step": 12845
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8707527747949089,
+      "learning_rate": 1.634842928783982e-05,
+      "loss": 0.822,
+      "step": 12846
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.585486395183106,
+      "learning_rate": 1.6347839712382887e-05,
+      "loss": 0.7424,
+      "step": 12847
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.40720521620236,
+      "learning_rate": 1.6347250099966823e-05,
+      "loss": 0.783,
+      "step": 12848
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4476771101617758,
+      "learning_rate": 1.6346660450595058e-05,
+      "loss": 0.8228,
+      "step": 12849
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5236149675556596,
+      "learning_rate": 1.6346070764271025e-05,
+      "loss": 0.843,
+      "step": 12850
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9074897735559817,
+      "learning_rate": 1.6345481040998163e-05,
+      "loss": 0.6497,
+      "step": 12851
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6088896254461136,
+      "learning_rate": 1.6344891280779894e-05,
+      "loss": 0.7647,
+      "step": 12852
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1038958789109428,
+      "learning_rate": 1.6344301483619666e-05,
+      "loss": 0.7402,
+      "step": 12853
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6732444412702722,
+      "learning_rate": 1.6343711649520904e-05,
+      "loss": 0.5528,
+      "step": 12854
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8081159952247712,
+      "learning_rate": 1.6343121778487047e-05,
+      "loss": 0.745,
+      "step": 12855
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3969085707764317,
+      "learning_rate": 1.6342531870521523e-05,
+      "loss": 0.8143,
+      "step": 12856
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4079747700087428,
+      "learning_rate": 1.634194192562777e-05,
+      "loss": 0.7482,
+      "step": 12857
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5883582849520606,
+      "learning_rate": 1.6341351943809228e-05,
+      "loss": 0.582,
+      "step": 12858
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6481401416154668,
+      "learning_rate": 1.634076192506932e-05,
+      "loss": 0.7711,
+      "step": 12859
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.487026383431134,
+      "learning_rate": 1.6340171869411497e-05,
+      "loss": 0.8304,
+      "step": 12860
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.04777823430677,
+      "learning_rate": 1.633958177683918e-05,
+      "loss": 0.8426,
+      "step": 12861
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3823815271437732,
+      "learning_rate": 1.6338991647355812e-05,
+      "loss": 0.809,
+      "step": 12862
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6255623047912462,
+      "learning_rate": 1.6338401480964832e-05,
+      "loss": 0.7967,
+      "step": 12863
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7397695808326807,
+      "learning_rate": 1.633781127766967e-05,
+      "loss": 0.9341,
+      "step": 12864
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.495078157672717,
+      "learning_rate": 1.633722103747376e-05,
+      "loss": 0.7579,
+      "step": 12865
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.235768677691476,
+      "learning_rate": 1.6336630760380546e-05,
+      "loss": 0.7281,
+      "step": 12866
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3977178564903467,
+      "learning_rate": 1.633604044639346e-05,
+      "loss": 0.8626,
+      "step": 12867
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4675266054714515,
+      "learning_rate": 1.6335450095515944e-05,
+      "loss": 0.786,
+      "step": 12868
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5033873005310885,
+      "learning_rate": 1.633485970775143e-05,
+      "loss": 0.8556,
+      "step": 12869
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5498903804413426,
+      "learning_rate": 1.6334269283103353e-05,
+      "loss": 0.7216,
+      "step": 12870
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7847676653826463,
+      "learning_rate": 1.633367882157516e-05,
+      "loss": 0.7683,
+      "step": 12871
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5687642375915867,
+      "learning_rate": 1.6333088323170285e-05,
+      "loss": 0.815,
+      "step": 12872
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7761260845872786,
+      "learning_rate": 1.633249778789216e-05,
+      "loss": 0.6206,
+      "step": 12873
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.126090835838544,
+      "learning_rate": 1.633190721574423e-05,
+      "loss": 0.739,
+      "step": 12874
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5127129038486038,
+      "learning_rate": 1.633131660672993e-05,
+      "loss": 0.8081,
+      "step": 12875
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.195087627120659,
+      "learning_rate": 1.6330725960852706e-05,
+      "loss": 0.7246,
+      "step": 12876
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9154960058540746,
+      "learning_rate": 1.6330135278115988e-05,
+      "loss": 0.6842,
+      "step": 12877
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.852002477097668,
+      "learning_rate": 1.6329544558523212e-05,
+      "loss": 0.8138,
+      "step": 12878
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6469268692445103,
+      "learning_rate": 1.6328953802077833e-05,
+      "loss": 0.8419,
+      "step": 12879
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5987831787806506,
+      "learning_rate": 1.6328363008783274e-05,
+      "loss": 0.7573,
+      "step": 12880
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.694876162620586,
+      "learning_rate": 1.6327772178642986e-05,
+      "loss": 0.713,
+      "step": 12881
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6489183619994245,
+      "learning_rate": 1.6327181311660406e-05,
+      "loss": 0.7729,
+      "step": 12882
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6288681223491204,
+      "learning_rate": 1.632659040783897e-05,
+      "loss": 0.5759,
+      "step": 12883
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4316283399258427,
+      "learning_rate": 1.6325999467182123e-05,
+      "loss": 0.8912,
+      "step": 12884
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.861055186260439,
+      "learning_rate": 1.6325408489693308e-05,
+      "loss": 0.767,
+      "step": 12885
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.607030446151334,
+      "learning_rate": 1.6324817475375957e-05,
+      "loss": 0.7534,
+      "step": 12886
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5920528956823893,
+      "learning_rate": 1.6324226424233517e-05,
+      "loss": 0.5577,
+      "step": 12887
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3897532510481836,
+      "learning_rate": 1.632363533626943e-05,
+      "loss": 0.805,
+      "step": 12888
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3518629491023202,
+      "learning_rate": 1.6323044211487135e-05,
+      "loss": 0.7507,
+      "step": 12889
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6154069278606005,
+      "learning_rate": 1.6322453049890078e-05,
+      "loss": 0.8555,
+      "step": 12890
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3798674039067287,
+      "learning_rate": 1.6321861851481693e-05,
+      "loss": 0.6906,
+      "step": 12891
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6316813331786338,
+      "learning_rate": 1.632127061626543e-05,
+      "loss": 0.5077,
+      "step": 12892
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3896563249879041,
+      "learning_rate": 1.6320679344244727e-05,
+      "loss": 0.8698,
+      "step": 12893
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.73260044436184,
+      "learning_rate": 1.6320088035423027e-05,
+      "loss": 0.8059,
+      "step": 12894
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7785577400584176,
+      "learning_rate": 1.6319496689803775e-05,
+      "loss": 0.7272,
+      "step": 12895
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.431575859185316,
+      "learning_rate": 1.6318905307390412e-05,
+      "loss": 0.7092,
+      "step": 12896
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5503912026622344,
+      "learning_rate": 1.6318313888186383e-05,
+      "loss": 0.7841,
+      "step": 12897
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5387923010779334,
+      "learning_rate": 1.6317722432195127e-05,
+      "loss": 0.8226,
+      "step": 12898
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3454322610685179,
+      "learning_rate": 1.6317130939420093e-05,
+      "loss": 0.7372,
+      "step": 12899
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6118757634603652,
+      "learning_rate": 1.6316539409864724e-05,
+      "loss": 0.5583,
+      "step": 12900
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.447615512184417,
+      "learning_rate": 1.631594784353246e-05,
+      "loss": 0.7753,
+      "step": 12901
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5545361799645587,
+      "learning_rate": 1.631535624042675e-05,
+      "loss": 0.8197,
+      "step": 12902
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7163294609634023,
+      "learning_rate": 1.6314764600551035e-05,
+      "loss": 0.7204,
+      "step": 12903
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.494155527158204,
+      "learning_rate": 1.6314172923908764e-05,
+      "loss": 0.8344,
+      "step": 12904
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.268416921632584,
+      "learning_rate": 1.6313581210503375e-05,
+      "loss": 0.7713,
+      "step": 12905
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4701968143924522,
+      "learning_rate": 1.631298946033832e-05,
+      "loss": 0.6001,
+      "step": 12906
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6173426151516814,
+      "learning_rate": 1.6312397673417043e-05,
+      "loss": 0.7422,
+      "step": 12907
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4117729984154064,
+      "learning_rate": 1.6311805849742984e-05,
+      "loss": 0.6441,
+      "step": 12908
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4924141124565455,
+      "learning_rate": 1.6311213989319598e-05,
+      "loss": 0.6828,
+      "step": 12909
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6423275414697451,
+      "learning_rate": 1.6310622092150323e-05,
+      "loss": 0.7633,
+      "step": 12910
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9227109154744784,
+      "learning_rate": 1.631003015823861e-05,
+      "loss": 0.8338,
+      "step": 12911
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.352235445989348,
+      "learning_rate": 1.63094381875879e-05,
+      "loss": 0.6679,
+      "step": 12912
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.360714156312499,
+      "learning_rate": 1.630884618020165e-05,
+      "loss": 0.7704,
+      "step": 12913
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8273533247122185,
+      "learning_rate": 1.6308254136083295e-05,
+      "loss": 0.8608,
+      "step": 12914
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3938478045754803,
+      "learning_rate": 1.6307662055236288e-05,
+      "loss": 0.8582,
+      "step": 12915
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.415693361436606,
+      "learning_rate": 1.6307069937664077e-05,
+      "loss": 0.6887,
+      "step": 12916
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6034748748075752,
+      "learning_rate": 1.6306477783370108e-05,
+      "loss": 0.5659,
+      "step": 12917
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2704506882614905,
+      "learning_rate": 1.6305885592357826e-05,
+      "loss": 0.75,
+      "step": 12918
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5022669850751567,
+      "learning_rate": 1.6305293364630684e-05,
+      "loss": 0.8465,
+      "step": 12919
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6865558865328267,
+      "learning_rate": 1.630470110019213e-05,
+      "loss": 0.7686,
+      "step": 12920
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3705744856102953,
+      "learning_rate": 1.6304108799045607e-05,
+      "loss": 0.9213,
+      "step": 12921
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.373964720933417,
+      "learning_rate": 1.6303516461194564e-05,
+      "loss": 0.828,
+      "step": 12922
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4293876763540923,
+      "learning_rate": 1.6302924086642456e-05,
+      "loss": 0.6922,
+      "step": 12923
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.483741328808615,
+      "learning_rate": 1.6302331675392728e-05,
+      "loss": 0.8226,
+      "step": 12924
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3762541745854524,
+      "learning_rate": 1.630173922744883e-05,
+      "loss": 0.8879,
+      "step": 12925
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6562698577364582,
+      "learning_rate": 1.6301146742814215e-05,
+      "loss": 0.5512,
+      "step": 12926
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4131407370899522,
+      "learning_rate": 1.630055422149232e-05,
+      "loss": 0.7322,
+      "step": 12927
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3445550608737062,
+      "learning_rate": 1.6299961663486613e-05,
+      "loss": 0.7746,
+      "step": 12928
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7281028977140125,
+      "learning_rate": 1.629936906880053e-05,
+      "loss": 0.7461,
+      "step": 12929
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5631861810471188,
+      "learning_rate": 1.6298776437437526e-05,
+      "loss": 0.5252,
+      "step": 12930
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6819278185410877,
+      "learning_rate": 1.629818376940105e-05,
+      "loss": 0.818,
+      "step": 12931
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9610203967220063,
+      "learning_rate": 1.6297591064694557e-05,
+      "loss": 0.7593,
+      "step": 12932
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.578291242282148,
+      "learning_rate": 1.629699832332149e-05,
+      "loss": 0.8155,
+      "step": 12933
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2426757716769554,
+      "learning_rate": 1.629640554528531e-05,
+      "loss": 0.7665,
+      "step": 12934
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5887752019017314,
+      "learning_rate": 1.629581273058946e-05,
+      "loss": 0.7672,
+      "step": 12935
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.419010158051553,
+      "learning_rate": 1.62952198792374e-05,
+      "loss": 0.8218,
+      "step": 12936
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5998206566638998,
+      "learning_rate": 1.6294626991232574e-05,
+      "loss": 0.5181,
+      "step": 12937
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3481980522442347,
+      "learning_rate": 1.6294034066578436e-05,
+      "loss": 0.7635,
+      "step": 12938
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7011215875509937,
+      "learning_rate": 1.629344110527844e-05,
+      "loss": 0.8076,
+      "step": 12939
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4699494603750438,
+      "learning_rate": 1.6292848107336038e-05,
+      "loss": 0.7758,
+      "step": 12940
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5046232929082874,
+      "learning_rate": 1.6292255072754682e-05,
+      "loss": 0.6901,
+      "step": 12941
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8219168131602934,
+      "learning_rate": 1.6291662001537827e-05,
+      "loss": 0.7286,
+      "step": 12942
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7721934431713329,
+      "learning_rate": 1.629106889368892e-05,
+      "loss": 0.7674,
+      "step": 12943
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7546413064779378,
+      "learning_rate": 1.6290475749211417e-05,
+      "loss": 0.6942,
+      "step": 12944
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5598591877798764,
+      "learning_rate": 1.6289882568108778e-05,
+      "loss": 0.8552,
+      "step": 12945
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.614314233700518,
+      "learning_rate": 1.628928935038445e-05,
+      "loss": 0.538,
+      "step": 12946
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5957506987461041,
+      "learning_rate": 1.6288696096041886e-05,
+      "loss": 0.5321,
+      "step": 12947
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.411968799825816,
+      "learning_rate": 1.628810280508454e-05,
+      "loss": 0.6156,
+      "step": 12948
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6560530688320603,
+      "learning_rate": 1.6287509477515872e-05,
+      "loss": 0.9303,
+      "step": 12949
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6415410210593977,
+      "learning_rate": 1.6286916113339333e-05,
+      "loss": 0.7489,
+      "step": 12950
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5704289888644204,
+      "learning_rate": 1.628632271255838e-05,
+      "loss": 0.5171,
+      "step": 12951
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5838154997644451,
+      "learning_rate": 1.6285729275176463e-05,
+      "loss": 0.7003,
+      "step": 12952
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.608497841305053,
+      "learning_rate": 1.628513580119704e-05,
+      "loss": 0.8488,
+      "step": 12953
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6358448773148322,
+      "learning_rate": 1.6284542290623568e-05,
+      "loss": 0.8052,
+      "step": 12954
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.514043941002455,
+      "learning_rate": 1.62839487434595e-05,
+      "loss": 0.8033,
+      "step": 12955
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8088423942897165,
+      "learning_rate": 1.6283355159708297e-05,
+      "loss": 0.7217,
+      "step": 12956
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3931495369248443,
+      "learning_rate": 1.6282761539373402e-05,
+      "loss": 0.8734,
+      "step": 12957
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.638758098034413,
+      "learning_rate": 1.628216788245829e-05,
+      "loss": 0.7946,
+      "step": 12958
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.7887222579012354,
+      "learning_rate": 1.6281574188966402e-05,
+      "loss": 0.7184,
+      "step": 12959
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3413980966634127,
+      "learning_rate": 1.6280980458901205e-05,
+      "loss": 0.8275,
+      "step": 12960
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7413013359770253,
+      "learning_rate": 1.6280386692266142e-05,
+      "loss": 0.755,
+      "step": 12961
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6066506502138311,
+      "learning_rate": 1.6279792889064688e-05,
+      "loss": 0.7164,
+      "step": 12962
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5077878926354138,
+      "learning_rate": 1.6279199049300288e-05,
+      "loss": 0.737,
+      "step": 12963
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4978098856717157,
+      "learning_rate": 1.6278605172976404e-05,
+      "loss": 0.6653,
+      "step": 12964
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9981965177524705,
+      "learning_rate": 1.6278011260096495e-05,
+      "loss": 0.8145,
+      "step": 12965
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7717209411290709,
+      "learning_rate": 1.6277417310664015e-05,
+      "loss": 0.8612,
+      "step": 12966
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5639102057876346,
+      "learning_rate": 1.6276823324682426e-05,
+      "loss": 0.6913,
+      "step": 12967
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.132687964972093,
+      "learning_rate": 1.6276229302155183e-05,
+      "loss": 0.7024,
+      "step": 12968
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.450338191010445,
+      "learning_rate": 1.6275635243085743e-05,
+      "loss": 0.6447,
+      "step": 12969
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4830023147278761,
+      "learning_rate": 1.6275041147477572e-05,
+      "loss": 0.7437,
+      "step": 12970
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0592891911324798,
+      "learning_rate": 1.6274447015334122e-05,
+      "loss": 0.7834,
+      "step": 12971
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.6175600146401994,
+      "learning_rate": 1.6273852846658856e-05,
+      "loss": 0.7539,
+      "step": 12972
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6590311301296996,
+      "learning_rate": 1.6273258641455233e-05,
+      "loss": 0.7971,
+      "step": 12973
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.831382090236183,
+      "learning_rate": 1.6272664399726712e-05,
+      "loss": 0.8113,
+      "step": 12974
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6603737757853216,
+      "learning_rate": 1.6272070121476753e-05,
+      "loss": 0.6739,
+      "step": 12975
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6729735721510188,
+      "learning_rate": 1.6271475806708815e-05,
+      "loss": 0.6255,
+      "step": 12976
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.90577749126797,
+      "learning_rate": 1.6270881455426358e-05,
+      "loss": 0.691,
+      "step": 12977
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6368158132325743,
+      "learning_rate": 1.6270287067632847e-05,
+      "loss": 0.7396,
+      "step": 12978
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.928584012566237,
+      "learning_rate": 1.626969264333174e-05,
+      "loss": 0.7367,
+      "step": 12979
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.822827641061125,
+      "learning_rate": 1.6269098182526494e-05,
+      "loss": 0.8656,
+      "step": 12980
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3528551781915716,
+      "learning_rate": 1.6268503685220577e-05,
+      "loss": 0.6288,
+      "step": 12981
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5907821716654156,
+      "learning_rate": 1.6267909151417444e-05,
+      "loss": 0.6387,
+      "step": 12982
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5400361112357388,
+      "learning_rate": 1.6267314581120564e-05,
+      "loss": 0.8471,
+      "step": 12983
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8867322625899032,
+      "learning_rate": 1.6266719974333387e-05,
+      "loss": 0.7902,
+      "step": 12984
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.60076849264135,
+      "learning_rate": 1.6266125331059387e-05,
+      "loss": 0.5438,
+      "step": 12985
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4730794112127332,
+      "learning_rate": 1.626553065130202e-05,
+      "loss": 0.7867,
+      "step": 12986
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3914152801944433,
+      "learning_rate": 1.6264935935064753e-05,
+      "loss": 0.7572,
+      "step": 12987
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3251144143608753,
+      "learning_rate": 1.6264341182351043e-05,
+      "loss": 0.7527,
+      "step": 12988
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2737701447277887,
+      "learning_rate": 1.6263746393164356e-05,
+      "loss": 0.7376,
+      "step": 12989
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5044203326626115,
+      "learning_rate": 1.6263151567508154e-05,
+      "loss": 0.767,
+      "step": 12990
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.401109549957667,
+      "learning_rate": 1.62625567053859e-05,
+      "loss": 0.7353,
+      "step": 12991
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.992345053009323,
+      "learning_rate": 1.6261961806801062e-05,
+      "loss": 0.8555,
+      "step": 12992
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5840149391915863,
+      "learning_rate": 1.6261366871757097e-05,
+      "loss": 0.696,
+      "step": 12993
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5553809046959806,
+      "learning_rate": 1.626077190025747e-05,
+      "loss": 0.6732,
+      "step": 12994
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4151859217728036,
+      "learning_rate": 1.6260176892305654e-05,
+      "loss": 0.6856,
+      "step": 12995
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.59201900934848,
+      "learning_rate": 1.6259581847905097e-05,
+      "loss": 0.7834,
+      "step": 12996
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4291253990568422,
+      "learning_rate": 1.6258986767059278e-05,
+      "loss": 0.7503,
+      "step": 12997
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7776718724615717,
+      "learning_rate": 1.6258391649771654e-05,
+      "loss": 0.9538,
+      "step": 12998
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.553656659226794,
+      "learning_rate": 1.6257796496045694e-05,
+      "loss": 0.761,
+      "step": 12999
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4574770831890973,
+      "learning_rate": 1.625720130588486e-05,
+      "loss": 0.6755,
+      "step": 13000
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9338336530845357,
+      "learning_rate": 1.625660607929262e-05,
+      "loss": 0.5964,
+      "step": 13001
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6109654393961375,
+      "learning_rate": 1.625601081627244e-05,
+      "loss": 0.5472,
+      "step": 13002
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5554543793304745,
+      "learning_rate": 1.6255415516827783e-05,
+      "loss": 0.7985,
+      "step": 13003
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.701265281057504,
+      "learning_rate": 1.6254820180962115e-05,
+      "loss": 0.7677,
+      "step": 13004
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4067514479599257,
+      "learning_rate": 1.6254224808678906e-05,
+      "loss": 0.6996,
+      "step": 13005
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.922305413231585,
+      "learning_rate": 1.625362939998162e-05,
+      "loss": 0.8317,
+      "step": 13006
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4370218019754697,
+      "learning_rate": 1.6253033954873725e-05,
+      "loss": 0.7781,
+      "step": 13007
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4177462983611444,
+      "learning_rate": 1.6252438473358682e-05,
+      "loss": 0.8278,
+      "step": 13008
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3225755523210927,
+      "learning_rate": 1.625184295543997e-05,
+      "loss": 0.8661,
+      "step": 13009
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2051205280308377,
+      "learning_rate": 1.625124740112104e-05,
+      "loss": 0.6779,
+      "step": 13010
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.385199968017699,
+      "learning_rate": 1.6250651810405375e-05,
+      "loss": 0.7984,
+      "step": 13011
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.787825029748526,
+      "learning_rate": 1.6250056183296434e-05,
+      "loss": 0.7571,
+      "step": 13012
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4788803090639848,
+      "learning_rate": 1.6249460519797687e-05,
+      "loss": 0.7151,
+      "step": 13013
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4484756841610846,
+      "learning_rate": 1.6248864819912603e-05,
+      "loss": 0.8289,
+      "step": 13014
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5981472745728191,
+      "learning_rate": 1.624826908364465e-05,
+      "loss": 0.8348,
+      "step": 13015
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.468063136579706,
+      "learning_rate": 1.6247673310997294e-05,
+      "loss": 0.7341,
+      "step": 13016
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.592769733967852,
+      "learning_rate": 1.6247077501974008e-05,
+      "loss": 0.81,
+      "step": 13017
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9146357637302787,
+      "learning_rate": 1.624648165657826e-05,
+      "loss": 0.7203,
+      "step": 13018
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.78752302525656,
+      "learning_rate": 1.6245885774813514e-05,
+      "loss": 0.8131,
+      "step": 13019
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6673073585642295,
+      "learning_rate": 1.624528985668325e-05,
+      "loss": 0.9134,
+      "step": 13020
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4520958316902732,
+      "learning_rate": 1.6244693902190928e-05,
+      "loss": 0.7781,
+      "step": 13021
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5333013943634268,
+      "learning_rate": 1.6244097911340018e-05,
+      "loss": 0.7605,
+      "step": 13022
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4639994156949534,
+      "learning_rate": 1.6243501884133997e-05,
+      "loss": 0.7055,
+      "step": 13023
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5784772608906648,
+      "learning_rate": 1.6242905820576327e-05,
+      "loss": 0.7995,
+      "step": 13024
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3619337521609827,
+      "learning_rate": 1.6242309720670492e-05,
+      "loss": 0.754,
+      "step": 13025
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.375086131739517,
+      "learning_rate": 1.6241713584419946e-05,
+      "loss": 0.8692,
+      "step": 13026
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.52510204936667,
+      "learning_rate": 1.6241117411828167e-05,
+      "loss": 0.712,
+      "step": 13027
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4712073797728684,
+      "learning_rate": 1.624052120289863e-05,
+      "loss": 0.7962,
+      "step": 13028
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4376066004528192,
+      "learning_rate": 1.6239924957634802e-05,
+      "loss": 0.6913,
+      "step": 13029
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4274748100224965,
+      "learning_rate": 1.6239328676040155e-05,
+      "loss": 0.7892,
+      "step": 13030
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9972599512158378,
+      "learning_rate": 1.623873235811816e-05,
+      "loss": 0.7795,
+      "step": 13031
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.791666674640665,
+      "learning_rate": 1.623813600387229e-05,
+      "loss": 0.7648,
+      "step": 13032
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2775136739208353,
+      "learning_rate": 1.6237539613306018e-05,
+      "loss": 0.6807,
+      "step": 13033
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.396810310149913,
+      "learning_rate": 1.6236943186422815e-05,
+      "loss": 0.7485,
+      "step": 13034
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5824635599244615,
+      "learning_rate": 1.6236346723226154e-05,
+      "loss": 0.5391,
+      "step": 13035
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.498777966522817,
+      "learning_rate": 1.623575022371951e-05,
+      "loss": 0.6855,
+      "step": 13036
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5117737710551216,
+      "learning_rate": 1.6235153687906354e-05,
+      "loss": 0.7628,
+      "step": 13037
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.617920388901224,
+      "learning_rate": 1.6234557115790157e-05,
+      "loss": 0.7052,
+      "step": 13038
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3016436084116814,
+      "learning_rate": 1.6233960507374397e-05,
+      "loss": 0.7103,
+      "step": 13039
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2896230188487379,
+      "learning_rate": 1.6233363862662543e-05,
+      "loss": 0.6978,
+      "step": 13040
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6229754306081683,
+      "learning_rate": 1.6232767181658072e-05,
+      "loss": 0.7039,
+      "step": 13041
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.645136426799134,
+      "learning_rate": 1.623217046436446e-05,
+      "loss": 0.7209,
+      "step": 13042
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.364205804962815,
+      "learning_rate": 1.6231573710785172e-05,
+      "loss": 0.6753,
+      "step": 13043
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.31987823892394,
+      "learning_rate": 1.6230976920923696e-05,
+      "loss": 0.5677,
+      "step": 13044
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9251071119677345,
+      "learning_rate": 1.6230380094783497e-05,
+      "loss": 0.6916,
+      "step": 13045
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4435415703225989,
+      "learning_rate": 1.622978323236805e-05,
+      "loss": 0.7571,
+      "step": 13046
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4951970074151735,
+      "learning_rate": 1.6229186333680832e-05,
+      "loss": 0.6912,
+      "step": 13047
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4413192171405027,
+      "learning_rate": 1.6228589398725326e-05,
+      "loss": 0.7794,
+      "step": 13048
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8501490682464232,
+      "learning_rate": 1.6227992427504995e-05,
+      "loss": 0.7185,
+      "step": 13049
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.636770955700696,
+      "learning_rate": 1.6227395420023323e-05,
+      "loss": 0.5708,
+      "step": 13050
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3585677522707176,
+      "learning_rate": 1.6226798376283782e-05,
+      "loss": 0.6395,
+      "step": 13051
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4272779339187551,
+      "learning_rate": 1.6226201296289848e-05,
+      "loss": 0.8355,
+      "step": 13052
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2447645124482913,
+      "learning_rate": 1.6225604180045e-05,
+      "loss": 0.7949,
+      "step": 13053
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6185397639040588,
+      "learning_rate": 1.6225007027552715e-05,
+      "loss": 0.7747,
+      "step": 13054
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4441981982298542,
+      "learning_rate": 1.6224409838816465e-05,
+      "loss": 0.767,
+      "step": 13055
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4752758135015631,
+      "learning_rate": 1.6223812613839734e-05,
+      "loss": 0.6996,
+      "step": 13056
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5582503325617796,
+      "learning_rate": 1.6223215352625994e-05,
+      "loss": 0.7616,
+      "step": 13057
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.470807579291445,
+      "learning_rate": 1.6222618055178723e-05,
+      "loss": 0.8354,
+      "step": 13058
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8358436277748278,
+      "learning_rate": 1.62220207215014e-05,
+      "loss": 0.715,
+      "step": 13059
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.863312766954567,
+      "learning_rate": 1.6221423351597505e-05,
+      "loss": 0.8457,
+      "step": 13060
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.1999039534519103,
+      "learning_rate": 1.622082594547051e-05,
+      "loss": 0.66,
+      "step": 13061
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5323498064437926,
+      "learning_rate": 1.62202285031239e-05,
+      "loss": 0.7552,
+      "step": 13062
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.763743760103,
+      "learning_rate": 1.6219631024561148e-05,
+      "loss": 0.6958,
+      "step": 13063
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8157162990273101,
+      "learning_rate": 1.6219033509785738e-05,
+      "loss": 0.6378,
+      "step": 13064
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.470740505008013,
+      "learning_rate": 1.6218435958801143e-05,
+      "loss": 0.7657,
+      "step": 13065
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.358505604159448,
+      "learning_rate": 1.6217838371610847e-05,
+      "loss": 0.7439,
+      "step": 13066
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5893354991643281,
+      "learning_rate": 1.6217240748218325e-05,
+      "loss": 0.5419,
+      "step": 13067
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5810605019610628,
+      "learning_rate": 1.621664308862706e-05,
+      "loss": 0.6269,
+      "step": 13068
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.530685991457203,
+      "learning_rate": 1.621604539284053e-05,
+      "loss": 0.7793,
+      "step": 13069
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0443241446491487,
+      "learning_rate": 1.6215447660862216e-05,
+      "loss": 0.7453,
+      "step": 13070
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4798334253703724,
+      "learning_rate": 1.6214849892695602e-05,
+      "loss": 0.8642,
+      "step": 13071
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.372542274851192,
+      "learning_rate": 1.6214252088344164e-05,
+      "loss": 0.728,
+      "step": 13072
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7048492319312099,
+      "learning_rate": 1.621365424781138e-05,
+      "loss": 0.6992,
+      "step": 13073
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.655752625351833,
+      "learning_rate": 1.6213056371100732e-05,
+      "loss": 0.7118,
+      "step": 13074
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.56219575302162,
+      "learning_rate": 1.6212458458215706e-05,
+      "loss": 0.7162,
+      "step": 13075
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4524288511612067,
+      "learning_rate": 1.621186050915978e-05,
+      "loss": 0.7647,
+      "step": 13076
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7935812690883772,
+      "learning_rate": 1.6211262523936432e-05,
+      "loss": 0.851,
+      "step": 13077
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4123622007008165,
+      "learning_rate": 1.6210664502549152e-05,
+      "loss": 0.781,
+      "step": 13078
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.683334805651365,
+      "learning_rate": 1.6210066445001413e-05,
+      "loss": 0.8338,
+      "step": 13079
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3676635787884093,
+      "learning_rate": 1.6209468351296703e-05,
+      "loss": 0.7745,
+      "step": 13080
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4755146977171358,
+      "learning_rate": 1.6208870221438503e-05,
+      "loss": 0.7069,
+      "step": 13081
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6568018372734876,
+      "learning_rate": 1.6208272055430292e-05,
+      "loss": 0.6723,
+      "step": 13082
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.37725970962143,
+      "learning_rate": 1.620767385327556e-05,
+      "loss": 0.7891,
+      "step": 13083
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5431430568239788,
+      "learning_rate": 1.620707561497778e-05,
+      "loss": 0.7698,
+      "step": 13084
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5484191999156383,
+      "learning_rate": 1.6206477340540444e-05,
+      "loss": 0.5842,
+      "step": 13085
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6180600229005021,
+      "learning_rate": 1.6205879029967032e-05,
+      "loss": 0.5562,
+      "step": 13086
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5610035882706044,
+      "learning_rate": 1.620528068326103e-05,
+      "loss": 0.6845,
+      "step": 13087
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4085670927369385,
+      "learning_rate": 1.6204682300425915e-05,
+      "loss": 0.706,
+      "step": 13088
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.041544463243894,
+      "learning_rate": 1.6204083881465175e-05,
+      "loss": 0.7443,
+      "step": 13089
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7165507685989565,
+      "learning_rate": 1.62034854263823e-05,
+      "loss": 0.8092,
+      "step": 13090
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7550328919462876,
+      "learning_rate": 1.620288693518076e-05,
+      "loss": 0.7213,
+      "step": 13091
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5392780584908043,
+      "learning_rate": 1.6202288407864056e-05,
+      "loss": 0.6235,
+      "step": 13092
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5647850398516034,
+      "learning_rate": 1.620168984443566e-05,
+      "loss": 0.7816,
+      "step": 13093
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4625453273475755,
+      "learning_rate": 1.6201091244899067e-05,
+      "loss": 0.6986,
+      "step": 13094
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8076807527717287,
+      "learning_rate": 1.6200492609257758e-05,
+      "loss": 0.917,
+      "step": 13095
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6874760619943385,
+      "learning_rate": 1.6199893937515214e-05,
+      "loss": 0.7929,
+      "step": 13096
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7076698614037693,
+      "learning_rate": 1.6199295229674927e-05,
+      "loss": 0.815,
+      "step": 13097
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6407760116710062,
+      "learning_rate": 1.6198696485740378e-05,
+      "loss": 0.7463,
+      "step": 13098
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4357289297360074,
+      "learning_rate": 1.619809770571506e-05,
+      "loss": 0.7718,
+      "step": 13099
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6203359706540363,
+      "learning_rate": 1.619749888960245e-05,
+      "loss": 0.8491,
+      "step": 13100
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.50315324614303,
+      "learning_rate": 1.6196900037406043e-05,
+      "loss": 0.8722,
+      "step": 13101
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7928441956326837,
+      "learning_rate": 1.619630114912932e-05,
+      "loss": 0.7445,
+      "step": 13102
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4482804404560026,
+      "learning_rate": 1.6195702224775768e-05,
+      "loss": 0.7496,
+      "step": 13103
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6147252564131311,
+      "learning_rate": 1.619510326434888e-05,
+      "loss": 0.4916,
+      "step": 13104
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4359806877893588,
+      "learning_rate": 1.619450426785214e-05,
+      "loss": 0.6401,
+      "step": 13105
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5004910236430409,
+      "learning_rate": 1.619390523528903e-05,
+      "loss": 0.8043,
+      "step": 13106
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3572210238336577,
+      "learning_rate": 1.619330616666305e-05,
+      "loss": 0.7869,
+      "step": 13107
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3827528258758763,
+      "learning_rate": 1.6192707061977676e-05,
+      "loss": 0.7628,
+      "step": 13108
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3458870286642044,
+      "learning_rate": 1.61921079212364e-05,
+      "loss": 0.7299,
+      "step": 13109
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2944316864449297,
+      "learning_rate": 1.6191508744442713e-05,
+      "loss": 0.6736,
+      "step": 13110
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.321741171250473,
+      "learning_rate": 1.6190909531600103e-05,
+      "loss": 0.8329,
+      "step": 13111
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6009095839744145,
+      "learning_rate": 1.6190310282712056e-05,
+      "loss": 0.7076,
+      "step": 13112
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7577411330242738,
+      "learning_rate": 1.6189710997782063e-05,
+      "loss": 0.8458,
+      "step": 13113
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6234011583315313,
+      "learning_rate": 1.6189111676813613e-05,
+      "loss": 0.7466,
+      "step": 13114
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5672399414815321,
+      "learning_rate": 1.6188512319810197e-05,
+      "loss": 0.8362,
+      "step": 13115
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.6740606873283026,
+      "learning_rate": 1.6187912926775303e-05,
+      "loss": 0.8363,
+      "step": 13116
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3838290175150714,
+      "learning_rate": 1.618731349771242e-05,
+      "loss": 0.7272,
+      "step": 13117
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5006699583108307,
+      "learning_rate": 1.6186714032625036e-05,
+      "loss": 0.8088,
+      "step": 13118
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5140320033873393,
+      "learning_rate": 1.6186114531516646e-05,
+      "loss": 0.7811,
+      "step": 13119
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7591892999281025,
+      "learning_rate": 1.6185514994390738e-05,
+      "loss": 0.7685,
+      "step": 13120
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.19268636535355,
+      "learning_rate": 1.6184915421250805e-05,
+      "loss": 0.7135,
+      "step": 13121
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6398077622313356,
+      "learning_rate": 1.6184315812100337e-05,
+      "loss": 0.7187,
+      "step": 13122
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.483330801667095,
+      "learning_rate": 1.6183716166942824e-05,
+      "loss": 0.784,
+      "step": 13123
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.450648166927659,
+      "learning_rate": 1.6183116485781755e-05,
+      "loss": 0.8033,
+      "step": 13124
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4698557600771744,
+      "learning_rate": 1.6182516768620627e-05,
+      "loss": 0.8402,
+      "step": 13125
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.625612785741676,
+      "learning_rate": 1.618191701546293e-05,
+      "loss": 0.4979,
+      "step": 13126
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7843883499851878,
+      "learning_rate": 1.6181317226312154e-05,
+      "loss": 0.6616,
+      "step": 13127
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7915034285111542,
+      "learning_rate": 1.618071740117179e-05,
+      "loss": 0.701,
+      "step": 13128
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2751625492547727,
+      "learning_rate": 1.6180117540045338e-05,
+      "loss": 0.7549,
+      "step": 13129
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6205030373587911,
+      "learning_rate": 1.617951764293628e-05,
+      "loss": 0.5874,
+      "step": 13130
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6219029180692932,
+      "learning_rate": 1.6178917709848113e-05,
+      "loss": 0.7081,
+      "step": 13131
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5205044547105475,
+      "learning_rate": 1.6178317740784333e-05,
+      "loss": 0.7821,
+      "step": 13132
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3762690307100396,
+      "learning_rate": 1.6177717735748433e-05,
+      "loss": 0.8139,
+      "step": 13133
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8411870388834424,
+      "learning_rate": 1.61771176947439e-05,
+      "loss": 0.7459,
+      "step": 13134
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4911156310186886,
+      "learning_rate": 1.6176517617774234e-05,
+      "loss": 0.8422,
+      "step": 13135
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6048157806977104,
+      "learning_rate": 1.617591750484293e-05,
+      "loss": 0.483,
+      "step": 13136
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3257229287962122,
+      "learning_rate": 1.6175317355953477e-05,
+      "loss": 0.7796,
+      "step": 13137
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4505078766500412,
+      "learning_rate": 1.6174717171109372e-05,
+      "loss": 0.7764,
+      "step": 13138
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4663947672579645,
+      "learning_rate": 1.6174116950314106e-05,
+      "loss": 0.7582,
+      "step": 13139
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.683529899215045,
+      "learning_rate": 1.6173516693571178e-05,
+      "loss": 0.6952,
+      "step": 13140
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5580605320650407,
+      "learning_rate": 1.617291640088408e-05,
+      "loss": 0.8099,
+      "step": 13141
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3873343649284788,
+      "learning_rate": 1.617231607225631e-05,
+      "loss": 0.7607,
+      "step": 13142
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5742735078767374,
+      "learning_rate": 1.6171715707691366e-05,
+      "loss": 0.556,
+      "step": 13143
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.669778004304319,
+      "learning_rate": 1.6171115307192734e-05,
+      "loss": 0.736,
+      "step": 13144
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5830410864050294,
+      "learning_rate": 1.6170514870763912e-05,
+      "loss": 0.5192,
+      "step": 13145
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.625207073222468,
+      "learning_rate": 1.6169914398408405e-05,
+      "loss": 0.7809,
+      "step": 13146
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4584212234656633,
+      "learning_rate": 1.6169313890129702e-05,
+      "loss": 0.8326,
+      "step": 13147
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4363688999421094,
+      "learning_rate": 1.61687133459313e-05,
+      "loss": 0.755,
+      "step": 13148
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4196262006578217,
+      "learning_rate": 1.6168112765816692e-05,
+      "loss": 0.7557,
+      "step": 13149
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.377594774650822,
+      "learning_rate": 1.616751214978938e-05,
+      "loss": 0.7507,
+      "step": 13150
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6716108119328528,
+      "learning_rate": 1.6166911497852864e-05,
+      "loss": 0.7502,
+      "step": 13151
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7627617705416838,
+      "learning_rate": 1.6166310810010633e-05,
+      "loss": 0.8085,
+      "step": 13152
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4461248024717328,
+      "learning_rate": 1.616571008626619e-05,
+      "loss": 0.7968,
+      "step": 13153
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6639547806739015,
+      "learning_rate": 1.616510932662303e-05,
+      "loss": 0.6017,
+      "step": 13154
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6142677431333332,
+      "learning_rate": 1.6164508531084653e-05,
+      "loss": 0.8526,
+      "step": 13155
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.5723555223156174,
+      "learning_rate": 1.616390769965455e-05,
+      "loss": 0.6682,
+      "step": 13156
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4935798113004228,
+      "learning_rate": 1.6163306832336232e-05,
+      "loss": 0.8084,
+      "step": 13157
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.324609354958795,
+      "learning_rate": 1.616270592913319e-05,
+      "loss": 0.8433,
+      "step": 13158
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4002045858707544,
+      "learning_rate": 1.616210499004892e-05,
+      "loss": 0.7702,
+      "step": 13159
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.286083279701672,
+      "learning_rate": 1.6161504015086923e-05,
+      "loss": 0.7462,
+      "step": 13160
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4990971358026002,
+      "learning_rate": 1.61609030042507e-05,
+      "loss": 0.8123,
+      "step": 13161
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5111712223146239,
+      "learning_rate": 1.616030195754375e-05,
+      "loss": 0.7543,
+      "step": 13162
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5558025272163318,
+      "learning_rate": 1.615970087496957e-05,
+      "loss": 0.7901,
+      "step": 13163
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5905715367677098,
+      "learning_rate": 1.6159099756531668e-05,
+      "loss": 0.843,
+      "step": 13164
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2957749752512466,
+      "learning_rate": 1.615849860223353e-05,
+      "loss": 0.7016,
+      "step": 13165
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4797696614480076,
+      "learning_rate": 1.6157897412078668e-05,
+      "loss": 0.7338,
+      "step": 13166
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7970810853487584,
+      "learning_rate": 1.615729618607057e-05,
+      "loss": 0.7976,
+      "step": 13167
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.620223497401968,
+      "learning_rate": 1.615669492421275e-05,
+      "loss": 0.685,
+      "step": 13168
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4336293700727674,
+      "learning_rate": 1.6156093626508706e-05,
+      "loss": 0.8055,
+      "step": 13169
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5137689539999963,
+      "learning_rate": 1.615549229296193e-05,
+      "loss": 0.7197,
+      "step": 13170
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6371704168787662,
+      "learning_rate": 1.6154890923575933e-05,
+      "loss": 0.7962,
+      "step": 13171
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6732893357301462,
+      "learning_rate": 1.6154289518354208e-05,
+      "loss": 0.7856,
+      "step": 13172
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3358367933508226,
+      "learning_rate": 1.6153688077300265e-05,
+      "loss": 0.6812,
+      "step": 13173
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.689569704970571,
+      "learning_rate": 1.6153086600417602e-05,
+      "loss": 0.7817,
+      "step": 13174
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.455303829994584,
+      "learning_rate": 1.615248508770972e-05,
+      "loss": 0.8887,
+      "step": 13175
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6270014193759708,
+      "learning_rate": 1.615188353918012e-05,
+      "loss": 0.8546,
+      "step": 13176
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5136781170354854,
+      "learning_rate": 1.6151281954832307e-05,
+      "loss": 0.8157,
+      "step": 13177
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3100917234610001,
+      "learning_rate": 1.6150680334669785e-05,
+      "loss": 0.7737,
+      "step": 13178
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5149796987829554,
+      "learning_rate": 1.615007867869605e-05,
+      "loss": 0.8129,
+      "step": 13179
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6192259352886555,
+      "learning_rate": 1.6149476986914617e-05,
+      "loss": 0.6363,
+      "step": 13180
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.461870817864426,
+      "learning_rate": 1.6148875259328978e-05,
+      "loss": 0.7774,
+      "step": 13181
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5307650306637695,
+      "learning_rate": 1.614827349594264e-05,
+      "loss": 0.7688,
+      "step": 13182
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.610058442075863,
+      "learning_rate": 1.614767169675911e-05,
+      "loss": 0.5537,
+      "step": 13183
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.704161976799504,
+      "learning_rate": 1.6147069861781887e-05,
+      "loss": 0.6193,
+      "step": 13184
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6179564674471365,
+      "learning_rate": 1.614646799101448e-05,
+      "loss": 0.6873,
+      "step": 13185
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5425148862275675,
+      "learning_rate": 1.6145866084460384e-05,
+      "loss": 0.6679,
+      "step": 13186
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.408554974466309,
+      "learning_rate": 1.6145264142123114e-05,
+      "loss": 0.7862,
+      "step": 13187
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4072375961720094,
+      "learning_rate": 1.6144662164006172e-05,
+      "loss": 0.7468,
+      "step": 13188
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.890972291202765,
+      "learning_rate": 1.614406015011306e-05,
+      "loss": 0.8274,
+      "step": 13189
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3550611780160307,
+      "learning_rate": 1.6143458100447283e-05,
+      "loss": 0.713,
+      "step": 13190
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5468443609148026,
+      "learning_rate": 1.614285601501235e-05,
+      "loss": 0.8538,
+      "step": 13191
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.398806886983958,
+      "learning_rate": 1.6142253893811766e-05,
+      "loss": 0.7044,
+      "step": 13192
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.729788778626376,
+      "learning_rate": 1.6141651736849032e-05,
+      "loss": 0.7262,
+      "step": 13193
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2451503973472375,
+      "learning_rate": 1.6141049544127657e-05,
+      "loss": 0.7738,
+      "step": 13194
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5729685570918674,
+      "learning_rate": 1.6140447315651152e-05,
+      "loss": 0.7451,
+      "step": 13195
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9490230849393007,
+      "learning_rate": 1.6139845051423012e-05,
+      "loss": 0.8484,
+      "step": 13196
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5741683004094256,
+      "learning_rate": 1.6139242751446754e-05,
+      "loss": 0.531,
+      "step": 13197
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3699200011137511,
+      "learning_rate": 1.6138640415725882e-05,
+      "loss": 0.7005,
+      "step": 13198
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5026169346779845,
+      "learning_rate": 1.61380380442639e-05,
+      "loss": 0.848,
+      "step": 13199
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.390894504142265,
+      "learning_rate": 1.613743563706432e-05,
+      "loss": 0.8257,
+      "step": 13200
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.546472768750255,
+      "learning_rate": 1.6136833194130642e-05,
+      "loss": 0.764,
+      "step": 13201
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4740768217204785,
+      "learning_rate": 1.6136230715466384e-05,
+      "loss": 0.7857,
+      "step": 13202
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.662093777972451,
+      "learning_rate": 1.6135628201075042e-05,
+      "loss": 0.789,
+      "step": 13203
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.525195475552928,
+      "learning_rate": 1.6135025650960137e-05,
+      "loss": 0.7897,
+      "step": 13204
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4379894617554299,
+      "learning_rate": 1.6134423065125164e-05,
+      "loss": 0.6508,
+      "step": 13205
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4443310197961041,
+      "learning_rate": 1.613382044357364e-05,
+      "loss": 0.7135,
+      "step": 13206
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6441355434777432,
+      "learning_rate": 1.613321778630907e-05,
+      "loss": 0.7665,
+      "step": 13207
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1388597102450797,
+      "learning_rate": 1.613261509333497e-05,
+      "loss": 0.8564,
+      "step": 13208
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3556337954550621,
+      "learning_rate": 1.613201236465484e-05,
+      "loss": 0.8629,
+      "step": 13209
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.572731108686768,
+      "learning_rate": 1.6131409600272186e-05,
+      "loss": 0.7595,
+      "step": 13210
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5699375940272218,
+      "learning_rate": 1.6130806800190532e-05,
+      "loss": 0.9121,
+      "step": 13211
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4591435501757033,
+      "learning_rate": 1.613020396441338e-05,
+      "loss": 0.6975,
+      "step": 13212
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6168469699198561,
+      "learning_rate": 1.612960109294424e-05,
+      "loss": 0.5245,
+      "step": 13213
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5998604667867082,
+      "learning_rate": 1.6128998185786616e-05,
+      "loss": 0.879,
+      "step": 13214
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3707652047949679,
+      "learning_rate": 1.6128395242944027e-05,
+      "loss": 0.6802,
+      "step": 13215
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5024852039646803,
+      "learning_rate": 1.6127792264419983e-05,
+      "loss": 0.8535,
+      "step": 13216
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.454656981060575,
+      "learning_rate": 1.6127189250217988e-05,
+      "loss": 0.7524,
+      "step": 13217
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2184358208894122,
+      "learning_rate": 1.612658620034156e-05,
+      "loss": 0.8039,
+      "step": 13218
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3180829276765,
+      "learning_rate": 1.612598311479421e-05,
+      "loss": 0.7064,
+      "step": 13219
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3144977447624413,
+      "learning_rate": 1.6125379993579448e-05,
+      "loss": 0.7255,
+      "step": 13220
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2659996522742112,
+      "learning_rate": 1.6124776836700778e-05,
+      "loss": 0.7241,
+      "step": 13221
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6213636478841789,
+      "learning_rate": 1.612417364416172e-05,
+      "loss": 0.6413,
+      "step": 13222
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2704174297623543,
+      "learning_rate": 1.6123570415965785e-05,
+      "loss": 0.7548,
+      "step": 13223
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.56002320469982,
+      "learning_rate": 1.6122967152116483e-05,
+      "loss": 0.8683,
+      "step": 13224
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3240028971316633,
+      "learning_rate": 1.6122363852617333e-05,
+      "loss": 0.721,
+      "step": 13225
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2779253381176197,
+      "learning_rate": 1.6121760517471837e-05,
+      "loss": 0.6442,
+      "step": 13226
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.598362598130779,
+      "learning_rate": 1.6121157146683514e-05,
+      "loss": 0.7676,
+      "step": 13227
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4264011863856567,
+      "learning_rate": 1.6120553740255877e-05,
+      "loss": 0.6234,
+      "step": 13228
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3086618857317638,
+      "learning_rate": 1.6119950298192438e-05,
+      "loss": 0.7087,
+      "step": 13229
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6640365653324978,
+      "learning_rate": 1.6119346820496712e-05,
+      "loss": 0.8327,
+      "step": 13230
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6341873264794375,
+      "learning_rate": 1.6118743307172212e-05,
+      "loss": 0.5353,
+      "step": 13231
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3876421110315058,
+      "learning_rate": 1.611813975822245e-05,
+      "loss": 0.7184,
+      "step": 13232
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8218381267731336,
+      "learning_rate": 1.611753617365094e-05,
+      "loss": 0.6232,
+      "step": 13233
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0924467059894694,
+      "learning_rate": 1.61169325534612e-05,
+      "loss": 0.771,
+      "step": 13234
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2498862033948863,
+      "learning_rate": 1.611632889765674e-05,
+      "loss": 0.6669,
+      "step": 13235
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4934346388713053,
+      "learning_rate": 1.6115725206241078e-05,
+      "loss": 0.9274,
+      "step": 13236
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5523447504390424,
+      "learning_rate": 1.6115121479217724e-05,
+      "loss": 0.7828,
+      "step": 13237
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5524275053257641,
+      "learning_rate": 1.6114517716590205e-05,
+      "loss": 0.8132,
+      "step": 13238
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3908268450721157,
+      "learning_rate": 1.6113913918362023e-05,
+      "loss": 0.736,
+      "step": 13239
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.518583321349304,
+      "learning_rate": 1.6113310084536698e-05,
+      "loss": 0.7066,
+      "step": 13240
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5230317088092602,
+      "learning_rate": 1.6112706215117748e-05,
+      "loss": 0.8358,
+      "step": 13241
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.616998623257855,
+      "learning_rate": 1.6112102310108686e-05,
+      "loss": 0.8934,
+      "step": 13242
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5752348485628962,
+      "learning_rate": 1.611149836951303e-05,
+      "loss": 0.7613,
+      "step": 13243
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.598965936321831,
+      "learning_rate": 1.6110894393334293e-05,
+      "loss": 0.5547,
+      "step": 13244
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5686998673266122,
+      "learning_rate": 1.6110290381575996e-05,
+      "loss": 0.7308,
+      "step": 13245
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4350310343175712,
+      "learning_rate": 1.6109686334241655e-05,
+      "loss": 0.745,
+      "step": 13246
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4085968623735046,
+      "learning_rate": 1.6109082251334786e-05,
+      "loss": 0.7284,
+      "step": 13247
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4765272366018347,
+      "learning_rate": 1.6108478132858903e-05,
+      "loss": 0.6734,
+      "step": 13248
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3889032779733292,
+      "learning_rate": 1.610787397881753e-05,
+      "loss": 0.7445,
+      "step": 13249
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.971880952577478,
+      "learning_rate": 1.610726978921418e-05,
+      "loss": 0.6765,
+      "step": 13250
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6405116039477257,
+      "learning_rate": 1.610666556405237e-05,
+      "loss": 0.7048,
+      "step": 13251
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4585701882137339,
+      "learning_rate": 1.6106061303335622e-05,
+      "loss": 0.9305,
+      "step": 13252
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4160032589553322,
+      "learning_rate": 1.610545700706745e-05,
+      "loss": 0.7534,
+      "step": 13253
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3180537047057341,
+      "learning_rate": 1.6104852675251373e-05,
+      "loss": 0.7399,
+      "step": 13254
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0067604406410555,
+      "learning_rate": 1.6104248307890915e-05,
+      "loss": 0.7423,
+      "step": 13255
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.52324992007562,
+      "learning_rate": 1.6103643904989588e-05,
+      "loss": 0.7185,
+      "step": 13256
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6458309640597695,
+      "learning_rate": 1.6103039466550916e-05,
+      "loss": 0.7726,
+      "step": 13257
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4815955192673131,
+      "learning_rate": 1.6102434992578415e-05,
+      "loss": 0.7469,
+      "step": 13258
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5749799470641093,
+      "learning_rate": 1.6101830483075603e-05,
+      "loss": 0.8453,
+      "step": 13259
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6099186036060305,
+      "learning_rate": 1.6101225938046007e-05,
+      "loss": 0.7298,
+      "step": 13260
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3683258753298682,
+      "learning_rate": 1.610062135749314e-05,
+      "loss": 0.8348,
+      "step": 13261
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.638041205549888,
+      "learning_rate": 1.610001674142052e-05,
+      "loss": 0.7994,
+      "step": 13262
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.497208527774591,
+      "learning_rate": 1.6099412089831676e-05,
+      "loss": 0.73,
+      "step": 13263
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4296725870965952,
+      "learning_rate": 1.6098807402730124e-05,
+      "loss": 0.6766,
+      "step": 13264
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.352607263223157,
+      "learning_rate": 1.6098202680119382e-05,
+      "loss": 0.7799,
+      "step": 13265
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6376284205499139,
+      "learning_rate": 1.6097597922002974e-05,
+      "loss": 0.7696,
+      "step": 13266
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4551740593535545,
+      "learning_rate": 1.609699312838442e-05,
+      "loss": 0.7524,
+      "step": 13267
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3702838896914447,
+      "learning_rate": 1.609638829926724e-05,
+      "loss": 0.7918,
+      "step": 13268
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3444771022338113,
+      "learning_rate": 1.6095783434654963e-05,
+      "loss": 0.7517,
+      "step": 13269
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5618084342219,
+      "learning_rate": 1.60951785345511e-05,
+      "loss": 0.6626,
+      "step": 13270
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5600546898927334,
+      "learning_rate": 1.609457359895918e-05,
+      "loss": 0.6851,
+      "step": 13271
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4043780740848528,
+      "learning_rate": 1.6093968627882723e-05,
+      "loss": 0.8071,
+      "step": 13272
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5005979616970493,
+      "learning_rate": 1.609336362132525e-05,
+      "loss": 0.8481,
+      "step": 13273
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4582482000732473,
+      "learning_rate": 1.6092758579290286e-05,
+      "loss": 0.6847,
+      "step": 13274
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.507853984715673,
+      "learning_rate": 1.6092153501781352e-05,
+      "loss": 0.6939,
+      "step": 13275
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6143062555796321,
+      "learning_rate": 1.6091548388801975e-05,
+      "loss": 0.5552,
+      "step": 13276
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5375840561285885,
+      "learning_rate": 1.609094324035567e-05,
+      "loss": 0.818,
+      "step": 13277
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6402786029515481,
+      "learning_rate": 1.6090338056445965e-05,
+      "loss": 0.7087,
+      "step": 13278
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.323715530897464,
+      "learning_rate": 1.6089732837076387e-05,
+      "loss": 0.7695,
+      "step": 13279
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3850667498342772,
+      "learning_rate": 1.6089127582250453e-05,
+      "loss": 0.8111,
+      "step": 13280
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.07323263818502,
+      "learning_rate": 1.6088522291971695e-05,
+      "loss": 0.8162,
+      "step": 13281
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4861467458789444,
+      "learning_rate": 1.608791696624363e-05,
+      "loss": 0.7354,
+      "step": 13282
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.098879403687733,
+      "learning_rate": 1.608731160506978e-05,
+      "loss": 0.8396,
+      "step": 13283
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6085863498183813,
+      "learning_rate": 1.6086706208453683e-05,
+      "loss": 0.5373,
+      "step": 13284
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3720935343261227,
+      "learning_rate": 1.608610077639885e-05,
+      "loss": 0.7556,
+      "step": 13285
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3193842401030003,
+      "learning_rate": 1.6085495308908813e-05,
+      "loss": 0.6535,
+      "step": 13286
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6030124508991684,
+      "learning_rate": 1.60848898059871e-05,
+      "loss": 0.5512,
+      "step": 13287
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6269274386088092,
+      "learning_rate": 1.6084284267637228e-05,
+      "loss": 0.5889,
+      "step": 13288
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6085471412393557,
+      "learning_rate": 1.6083678693862725e-05,
+      "loss": 0.5412,
+      "step": 13289
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5399379006840066,
+      "learning_rate": 1.608307308466712e-05,
+      "loss": 0.7918,
+      "step": 13290
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.614831744668689,
+      "learning_rate": 1.608246744005394e-05,
+      "loss": 0.8046,
+      "step": 13291
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5493618651362325,
+      "learning_rate": 1.6081861760026708e-05,
+      "loss": 0.5085,
+      "step": 13292
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6432159123766858,
+      "learning_rate": 1.608125604458895e-05,
+      "loss": 0.8215,
+      "step": 13293
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6485689740228775,
+      "learning_rate": 1.6080650293744192e-05,
+      "loss": 0.7585,
+      "step": 13294
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.80153827055769,
+      "learning_rate": 1.6080044507495967e-05,
+      "loss": 0.7583,
+      "step": 13295
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7103946907673957,
+      "learning_rate": 1.6079438685847796e-05,
+      "loss": 0.8368,
+      "step": 13296
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.465770321418603,
+      "learning_rate": 1.607883282880321e-05,
+      "loss": 0.8353,
+      "step": 13297
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4052024765030264,
+      "learning_rate": 1.607822693636573e-05,
+      "loss": 0.6936,
+      "step": 13298
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1241562673536163,
+      "learning_rate": 1.607762100853889e-05,
+      "loss": 0.7258,
+      "step": 13299
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6440370646998568,
+      "learning_rate": 1.607701504532622e-05,
+      "loss": 0.5554,
+      "step": 13300
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7084768870205014,
+      "learning_rate": 1.6076409046731243e-05,
+      "loss": 0.765,
+      "step": 13301
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.512112405444306,
+      "learning_rate": 1.607580301275749e-05,
+      "loss": 0.7498,
+      "step": 13302
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6344717202737609,
+      "learning_rate": 1.6075196943408485e-05,
+      "loss": 0.7394,
+      "step": 13303
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6087503718699983,
+      "learning_rate": 1.607459083868776e-05,
+      "loss": 0.5963,
+      "step": 13304
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.965435662632365,
+      "learning_rate": 1.6073984698598845e-05,
+      "loss": 0.7097,
+      "step": 13305
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3922317013260455,
+      "learning_rate": 1.6073378523145272e-05,
+      "loss": 0.7454,
+      "step": 13306
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5156797683340328,
+      "learning_rate": 1.607277231233056e-05,
+      "loss": 0.8398,
+      "step": 13307
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7073042838961425,
+      "learning_rate": 1.607216606615825e-05,
+      "loss": 0.8446,
+      "step": 13308
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.454810687634541,
+      "learning_rate": 1.6071559784631862e-05,
+      "loss": 0.6256,
+      "step": 13309
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.728642802801332,
+      "learning_rate": 1.6070953467754935e-05,
+      "loss": 0.8106,
+      "step": 13310
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2629478961010483,
+      "learning_rate": 1.6070347115530994e-05,
+      "loss": 0.7282,
+      "step": 13311
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.403749489244893,
+      "learning_rate": 1.606974072796357e-05,
+      "loss": 0.635,
+      "step": 13312
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4893208871542705,
+      "learning_rate": 1.6069134305056193e-05,
+      "loss": 0.7491,
+      "step": 13313
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5412246375818175,
+      "learning_rate": 1.6068527846812398e-05,
+      "loss": 0.711,
+      "step": 13314
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3279136382155097,
+      "learning_rate": 1.6067921353235707e-05,
+      "loss": 0.7682,
+      "step": 13315
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.381011069151267,
+      "learning_rate": 1.606731482432966e-05,
+      "loss": 0.9203,
+      "step": 13316
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5174224409683985,
+      "learning_rate": 1.6066708260097782e-05,
+      "loss": 0.8167,
+      "step": 13317
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5230558890113484,
+      "learning_rate": 1.6066101660543612e-05,
+      "loss": 0.8428,
+      "step": 13318
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.443051437601332,
+      "learning_rate": 1.6065495025670675e-05,
+      "loss": 0.7042,
+      "step": 13319
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4468888308075194,
+      "learning_rate": 1.606488835548251e-05,
+      "loss": 0.8205,
+      "step": 13320
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.591561025881568,
+      "learning_rate": 1.6064281649982637e-05,
+      "loss": 0.869,
+      "step": 13321
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6115259864731721,
+      "learning_rate": 1.6063674909174604e-05,
+      "loss": 0.8177,
+      "step": 13322
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4080173869717945,
+      "learning_rate": 1.606306813306193e-05,
+      "loss": 0.7184,
+      "step": 13323
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6469701993811123,
+      "learning_rate": 1.606246132164816e-05,
+      "loss": 0.7368,
+      "step": 13324
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4188485285160581,
+      "learning_rate": 1.606185447493682e-05,
+      "loss": 0.9081,
+      "step": 13325
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6642964263318307,
+      "learning_rate": 1.606124759293144e-05,
+      "loss": 0.7265,
+      "step": 13326
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4131761024148326,
+      "learning_rate": 1.606064067563556e-05,
+      "loss": 0.6846,
+      "step": 13327
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4394545406803827,
+      "learning_rate": 1.6060033723052707e-05,
+      "loss": 0.8914,
+      "step": 13328
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.651461060596462,
+      "learning_rate": 1.6059426735186426e-05,
+      "loss": 0.7151,
+      "step": 13329
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.35343079568654,
+      "learning_rate": 1.6058819712040237e-05,
+      "loss": 0.8288,
+      "step": 13330
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3702403973253943,
+      "learning_rate": 1.6058212653617688e-05,
+      "loss": 0.7837,
+      "step": 13331
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5047877518863388,
+      "learning_rate": 1.60576055599223e-05,
+      "loss": 0.7459,
+      "step": 13332
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3604322257849366,
+      "learning_rate": 1.605699843095762e-05,
+      "loss": 0.6888,
+      "step": 13333
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.57972904795826,
+      "learning_rate": 1.6056391266727178e-05,
+      "loss": 0.6698,
+      "step": 13334
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6026921930734759,
+      "learning_rate": 1.6055784067234507e-05,
+      "loss": 0.7443,
+      "step": 13335
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4445703774944143,
+      "learning_rate": 1.6055176832483143e-05,
+      "loss": 0.7084,
+      "step": 13336
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4291901660648125,
+      "learning_rate": 1.6054569562476622e-05,
+      "loss": 0.8538,
+      "step": 13337
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6191332098181394,
+      "learning_rate": 1.6053962257218483e-05,
+      "loss": 0.5563,
+      "step": 13338
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6267974474458524,
+      "learning_rate": 1.6053354916712258e-05,
+      "loss": 0.5462,
+      "step": 13339
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3506852521457644,
+      "learning_rate": 1.6052747540961483e-05,
+      "loss": 0.6864,
+      "step": 13340
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4324487234415966,
+      "learning_rate": 1.60521401299697e-05,
+      "loss": 0.7338,
+      "step": 13341
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5240957340453434,
+      "learning_rate": 1.6051532683740433e-05,
+      "loss": 0.791,
+      "step": 13342
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7002311324942632,
+      "learning_rate": 1.6050925202277232e-05,
+      "loss": 0.7342,
+      "step": 13343
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2981503748694057,
+      "learning_rate": 1.6050317685583627e-05,
+      "loss": 0.8039,
+      "step": 13344
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.494143609865174,
+      "learning_rate": 1.604971013366316e-05,
+      "loss": 0.6192,
+      "step": 13345
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3901829861012602,
+      "learning_rate": 1.6049102546519363e-05,
+      "loss": 0.7025,
+      "step": 13346
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4109999865994811,
+      "learning_rate": 1.6048494924155776e-05,
+      "loss": 0.6623,
+      "step": 13347
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.514465689135407,
+      "learning_rate": 1.6047887266575936e-05,
+      "loss": 0.8242,
+      "step": 13348
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6396463094370752,
+      "learning_rate": 1.6047279573783382e-05,
+      "loss": 0.7714,
+      "step": 13349
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7225494986931507,
+      "learning_rate": 1.6046671845781653e-05,
+      "loss": 0.7648,
+      "step": 13350
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5081535697251753,
+      "learning_rate": 1.6046064082574284e-05,
+      "loss": 0.7797,
+      "step": 13351
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5462115232465374,
+      "learning_rate": 1.604545628416482e-05,
+      "loss": 0.824,
+      "step": 13352
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3160722653585204,
+      "learning_rate": 1.6044848450556788e-05,
+      "loss": 0.678,
+      "step": 13353
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5761197896020356,
+      "learning_rate": 1.6044240581753742e-05,
+      "loss": 0.518,
+      "step": 13354
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.325688759942333,
+      "learning_rate": 1.604363267775921e-05,
+      "loss": 0.7012,
+      "step": 13355
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4967776353690114,
+      "learning_rate": 1.604302473857674e-05,
+      "loss": 0.6336,
+      "step": 13356
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4146308790040825,
+      "learning_rate": 1.604241676420986e-05,
+      "loss": 0.6166,
+      "step": 13357
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5106771812931596,
+      "learning_rate": 1.604180875466212e-05,
+      "loss": 0.7215,
+      "step": 13358
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4104193543183086,
+      "learning_rate": 1.6041200709937058e-05,
+      "loss": 0.7257,
+      "step": 13359
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.5201930546883853,
+      "learning_rate": 1.604059263003821e-05,
+      "loss": 0.8952,
+      "step": 13360
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.003338181550245,
+      "learning_rate": 1.603998451496912e-05,
+      "loss": 0.7676,
+      "step": 13361
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3115688230553775,
+      "learning_rate": 1.6039376364733328e-05,
+      "loss": 0.7058,
+      "step": 13362
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.422160849604243,
+      "learning_rate": 1.6038768179334373e-05,
+      "loss": 0.6179,
+      "step": 13363
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.681749942474651,
+      "learning_rate": 1.60381599587758e-05,
+      "loss": 0.6536,
+      "step": 13364
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5647121908186417,
+      "learning_rate": 1.6037551703061148e-05,
+      "loss": 0.5362,
+      "step": 13365
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9956402908009734,
+      "learning_rate": 1.603694341219396e-05,
+      "loss": 0.8632,
+      "step": 13366
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.213319012333713,
+      "learning_rate": 1.6036335086177772e-05,
+      "loss": 0.7623,
+      "step": 13367
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3162082677248543,
+      "learning_rate": 1.6035726725016134e-05,
+      "loss": 0.6371,
+      "step": 13368
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8057358025985735,
+      "learning_rate": 1.6035118328712583e-05,
+      "loss": 0.7566,
+      "step": 13369
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4302968575172261,
+      "learning_rate": 1.603450989727066e-05,
+      "loss": 0.8337,
+      "step": 13370
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5867513340502155,
+      "learning_rate": 1.603390143069391e-05,
+      "loss": 0.559,
+      "step": 13371
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3988408960432799,
+      "learning_rate": 1.603329292898588e-05,
+      "loss": 0.715,
+      "step": 13372
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5877839738168695,
+      "learning_rate": 1.6032684392150108e-05,
+      "loss": 0.7701,
+      "step": 13373
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5577617864586604,
+      "learning_rate": 1.6032075820190134e-05,
+      "loss": 0.5236,
+      "step": 13374
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6451291027257346,
+      "learning_rate": 1.6031467213109505e-05,
+      "loss": 0.7696,
+      "step": 13375
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4293896684877776,
+      "learning_rate": 1.6030858570911768e-05,
+      "loss": 0.6439,
+      "step": 13376
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4720769607548,
+      "learning_rate": 1.603024989360046e-05,
+      "loss": 0.6575,
+      "step": 13377
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.527454299248059,
+      "learning_rate": 1.602964118117913e-05,
+      "loss": 0.8602,
+      "step": 13378
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4093336483786443,
+      "learning_rate": 1.6029032433651317e-05,
+      "loss": 0.6133,
+      "step": 13379
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.517129323495992,
+      "learning_rate": 1.6028423651020572e-05,
+      "loss": 0.8801,
+      "step": 13380
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5892691317047718,
+      "learning_rate": 1.6027814833290435e-05,
+      "loss": 0.6927,
+      "step": 13381
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4895557077679689,
+      "learning_rate": 1.602720598046445e-05,
+      "loss": 0.6038,
+      "step": 13382
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6849447090617105,
+      "learning_rate": 1.6026597092546164e-05,
+      "loss": 0.7823,
+      "step": 13383
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4794850532592638,
+      "learning_rate": 1.6025988169539122e-05,
+      "loss": 0.7648,
+      "step": 13384
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3723762028369735,
+      "learning_rate": 1.602537921144687e-05,
+      "loss": 0.8109,
+      "step": 13385
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3456385337070427,
+      "learning_rate": 1.602477021827295e-05,
+      "loss": 0.9297,
+      "step": 13386
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7861389447287364,
+      "learning_rate": 1.6024161190020913e-05,
+      "loss": 0.8417,
+      "step": 13387
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4274309209246494,
+      "learning_rate": 1.60235521266943e-05,
+      "loss": 0.752,
+      "step": 13388
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4738280244920001,
+      "learning_rate": 1.6022943028296663e-05,
+      "loss": 0.7753,
+      "step": 13389
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4271542782325954,
+      "learning_rate": 1.6022333894831543e-05,
+      "loss": 0.7582,
+      "step": 13390
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.657007448698884,
+      "learning_rate": 1.6021724726302487e-05,
+      "loss": 0.7515,
+      "step": 13391
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6212082730935035,
+      "learning_rate": 1.6021115522713047e-05,
+      "loss": 0.5445,
+      "step": 13392
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.453680160797972,
+      "learning_rate": 1.6020506284066763e-05,
+      "loss": 0.8398,
+      "step": 13393
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9751604630600927,
+      "learning_rate": 1.6019897010367186e-05,
+      "loss": 0.7001,
+      "step": 13394
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8772389821492492,
+      "learning_rate": 1.6019287701617863e-05,
+      "loss": 0.9173,
+      "step": 13395
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6643382744119397,
+      "learning_rate": 1.6018678357822343e-05,
+      "loss": 0.6158,
+      "step": 13396
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.37059854359435,
+      "learning_rate": 1.601806897898417e-05,
+      "loss": 0.8054,
+      "step": 13397
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6071047990860265,
+      "learning_rate": 1.6017459565106896e-05,
+      "loss": 0.5193,
+      "step": 13398
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6148467072751103,
+      "learning_rate": 1.601685011619406e-05,
+      "loss": 0.7469,
+      "step": 13399
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3485274146048103,
+      "learning_rate": 1.6016240632249224e-05,
+      "loss": 0.6964,
+      "step": 13400
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5925055511071907,
+      "learning_rate": 1.601563111327593e-05,
+      "loss": 0.5496,
+      "step": 13401
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.377721294484408,
+      "learning_rate": 1.601502155927773e-05,
+      "loss": 0.7718,
+      "step": 13402
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.840014763588732,
+      "learning_rate": 1.6014411970258166e-05,
+      "loss": 0.8106,
+      "step": 13403
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.019553186148306,
+      "learning_rate": 1.601380234622079e-05,
+      "loss": 0.7808,
+      "step": 13404
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5801188636013109,
+      "learning_rate": 1.601319268716916e-05,
+      "loss": 0.4907,
+      "step": 13405
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7327883540345106,
+      "learning_rate": 1.6012582993106812e-05,
+      "loss": 0.8585,
+      "step": 13406
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5778761009305249,
+      "learning_rate": 1.6011973264037305e-05,
+      "loss": 0.5538,
+      "step": 13407
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6160492378351291,
+      "learning_rate": 1.6011363499964185e-05,
+      "loss": 0.5111,
+      "step": 13408
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3474458521797725,
+      "learning_rate": 1.6010753700891003e-05,
+      "loss": 0.6158,
+      "step": 13409
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1428237188315182,
+      "learning_rate": 1.601014386682131e-05,
+      "loss": 0.7124,
+      "step": 13410
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3489997534607094,
+      "learning_rate": 1.600953399775866e-05,
+      "loss": 0.794,
+      "step": 13411
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3291304003901732,
+      "learning_rate": 1.60089240937066e-05,
+      "loss": 0.7702,
+      "step": 13412
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3866475164328986,
+      "learning_rate": 1.6008314154668676e-05,
+      "loss": 0.7072,
+      "step": 13413
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3538549596994076,
+      "learning_rate": 1.6007704180648448e-05,
+      "loss": 0.7099,
+      "step": 13414
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.538091304018483,
+      "learning_rate": 1.6007094171649463e-05,
+      "loss": 0.6985,
+      "step": 13415
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.359046535715991,
+      "learning_rate": 1.6006484127675278e-05,
+      "loss": 0.6959,
+      "step": 13416
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8129054531497195,
+      "learning_rate": 1.6005874048729437e-05,
+      "loss": 0.8903,
+      "step": 13417
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3102898579315372,
+      "learning_rate": 1.6005263934815498e-05,
+      "loss": 0.7455,
+      "step": 13418
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2750729339451585,
+      "learning_rate": 1.600465378593701e-05,
+      "loss": 0.7253,
+      "step": 13419
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2726983266662253,
+      "learning_rate": 1.6004043602097527e-05,
+      "loss": 0.7802,
+      "step": 13420
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4530785648013294,
+      "learning_rate": 1.6003433383300603e-05,
+      "loss": 0.7992,
+      "step": 13421
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.125626477718878,
+      "learning_rate": 1.6002823129549784e-05,
+      "loss": 0.6837,
+      "step": 13422
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6288902938066843,
+      "learning_rate": 1.600221284084863e-05,
+      "loss": 0.7183,
+      "step": 13423
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6734624992827045,
+      "learning_rate": 1.6001602517200698e-05,
+      "loss": 0.7386,
+      "step": 13424
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6154015207581445,
+      "learning_rate": 1.600099215860953e-05,
+      "loss": 0.6823,
+      "step": 13425
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5244891251416817,
+      "learning_rate": 1.600038176507869e-05,
+      "loss": 0.6756,
+      "step": 13426
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.432031933586745,
+      "learning_rate": 1.5999771336611726e-05,
+      "loss": 0.6436,
+      "step": 13427
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.543101289759046,
+      "learning_rate": 1.5999160873212193e-05,
+      "loss": 0.7302,
+      "step": 13428
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9939632603612514,
+      "learning_rate": 1.599855037488365e-05,
+      "loss": 0.7069,
+      "step": 13429
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.711700760110335,
+      "learning_rate": 1.599793984162964e-05,
+      "loss": 0.8119,
+      "step": 13430
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6227792881666745,
+      "learning_rate": 1.599732927345373e-05,
+      "loss": 0.8817,
+      "step": 13431
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3132935502623628,
+      "learning_rate": 1.599671867035947e-05,
+      "loss": 0.7621,
+      "step": 13432
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1782061302550733,
+      "learning_rate": 1.5996108032350414e-05,
+      "loss": 0.7752,
+      "step": 13433
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.0680093726506077,
+      "learning_rate": 1.599549735943012e-05,
+      "loss": 0.7213,
+      "step": 13434
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4958807672982954,
+      "learning_rate": 1.599488665160214e-05,
+      "loss": 0.8307,
+      "step": 13435
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9682579074129878,
+      "learning_rate": 1.5994275908870037e-05,
+      "loss": 0.8574,
+      "step": 13436
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4388483505443537,
+      "learning_rate": 1.599366513123736e-05,
+      "loss": 0.8247,
+      "step": 13437
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.544980422248415,
+      "learning_rate": 1.5993054318707665e-05,
+      "loss": 0.793,
+      "step": 13438
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3027442849493995,
+      "learning_rate": 1.599244347128451e-05,
+      "loss": 0.7269,
+      "step": 13439
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3955605221255392,
+      "learning_rate": 1.5991832588971454e-05,
+      "loss": 0.7501,
+      "step": 13440
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6279897847135534,
+      "learning_rate": 1.5991221671772053e-05,
+      "loss": 0.7115,
+      "step": 13441
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5325112834087826,
+      "learning_rate": 1.5990610719689858e-05,
+      "loss": 0.7456,
+      "step": 13442
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6367795265382616,
+      "learning_rate": 1.5989999732728434e-05,
+      "loss": 0.8092,
+      "step": 13443
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6464486619471022,
+      "learning_rate": 1.5989388710891334e-05,
+      "loss": 0.6568,
+      "step": 13444
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7525366208461375,
+      "learning_rate": 1.5988777654182118e-05,
+      "loss": 0.7514,
+      "step": 13445
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.434599095067002,
+      "learning_rate": 1.5988166562604343e-05,
+      "loss": 0.6989,
+      "step": 13446
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3733142177952953,
+      "learning_rate": 1.5987555436161564e-05,
+      "loss": 0.6645,
+      "step": 13447
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.677159625509599,
+      "learning_rate": 1.5986944274857344e-05,
+      "loss": 0.7218,
+      "step": 13448
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.587961439484971,
+      "learning_rate": 1.598633307869524e-05,
+      "loss": 0.8833,
+      "step": 13449
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6235448754887316,
+      "learning_rate": 1.5985721847678806e-05,
+      "loss": 0.991,
+      "step": 13450
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4403086716321274,
+      "learning_rate": 1.59851105818116e-05,
+      "loss": 0.7565,
+      "step": 13451
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4992757388434659,
+      "learning_rate": 1.5984499281097194e-05,
+      "loss": 0.9021,
+      "step": 13452
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.627290247707706,
+      "learning_rate": 1.5983887945539138e-05,
+      "loss": 0.8295,
+      "step": 13453
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.278532056412145,
+      "learning_rate": 1.5983276575140986e-05,
+      "loss": 0.8061,
+      "step": 13454
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4401796049091495,
+      "learning_rate": 1.598266516990631e-05,
+      "loss": 0.7534,
+      "step": 13455
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.557277082738889,
+      "learning_rate": 1.5982053729838657e-05,
+      "loss": 0.6277,
+      "step": 13456
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2510206621353768,
+      "learning_rate": 1.59814422549416e-05,
+      "loss": 0.7417,
+      "step": 13457
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3716901175848804,
+      "learning_rate": 1.5980830745218686e-05,
+      "loss": 0.7352,
+      "step": 13458
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.658665501553308,
+      "learning_rate": 1.5980219200673488e-05,
+      "loss": 0.8254,
+      "step": 13459
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5911107318047861,
+      "learning_rate": 1.5979607621309557e-05,
+      "loss": 0.7116,
+      "step": 13460
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7430022736858966,
+      "learning_rate": 1.597899600713046e-05,
+      "loss": 0.7242,
+      "step": 13461
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4188701595645712,
+      "learning_rate": 1.5978384358139752e-05,
+      "loss": 0.762,
+      "step": 13462
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8756755961274436,
+      "learning_rate": 1.5977772674340996e-05,
+      "loss": 0.8225,
+      "step": 13463
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4344093519197396,
+      "learning_rate": 1.5977160955737758e-05,
+      "loss": 0.7137,
+      "step": 13464
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5774702212992885,
+      "learning_rate": 1.59765492023336e-05,
+      "loss": 0.5248,
+      "step": 13465
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4711106470790478,
+      "learning_rate": 1.5975937414132076e-05,
+      "loss": 0.7705,
+      "step": 13466
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7201667976156314,
+      "learning_rate": 1.5975325591136753e-05,
+      "loss": 0.8686,
+      "step": 13467
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3575844300958837,
+      "learning_rate": 1.5974713733351192e-05,
+      "loss": 0.6771,
+      "step": 13468
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6003138138176163,
+      "learning_rate": 1.5974101840778954e-05,
+      "loss": 0.5865,
+      "step": 13469
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2450194223673776,
+      "learning_rate": 1.597348991342361e-05,
+      "loss": 0.8423,
+      "step": 13470
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3428620432966636,
+      "learning_rate": 1.597287795128871e-05,
+      "loss": 0.7908,
+      "step": 13471
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4698071960357435,
+      "learning_rate": 1.597226595437783e-05,
+      "loss": 0.7835,
+      "step": 13472
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4089191411561575,
+      "learning_rate": 1.597165392269452e-05,
+      "loss": 0.7235,
+      "step": 13473
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4635703432307414,
+      "learning_rate": 1.5971041856242356e-05,
+      "loss": 0.7671,
+      "step": 13474
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5719961849763899,
+      "learning_rate": 1.5970429755024893e-05,
+      "loss": 0.5271,
+      "step": 13475
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7495170106546274,
+      "learning_rate": 1.59698176190457e-05,
+      "loss": 0.8153,
+      "step": 13476
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4584900174458946,
+      "learning_rate": 1.5969205448308335e-05,
+      "loss": 0.7645,
+      "step": 13477
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6178730617100405,
+      "learning_rate": 1.596859324281637e-05,
+      "loss": 0.5478,
+      "step": 13478
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.421826261688076,
+      "learning_rate": 1.5967981002573362e-05,
+      "loss": 0.8287,
+      "step": 13479
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5911944364518915,
+      "learning_rate": 1.5967368727582882e-05,
+      "loss": 0.5349,
+      "step": 13480
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6152185621598945,
+      "learning_rate": 1.5966756417848492e-05,
+      "loss": 0.8421,
+      "step": 13481
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.800816163995357,
+      "learning_rate": 1.5966144073373755e-05,
+      "loss": 0.7287,
+      "step": 13482
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6214615716540552,
+      "learning_rate": 1.596553169416224e-05,
+      "loss": 0.5489,
+      "step": 13483
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5943521580453808,
+      "learning_rate": 1.596491928021751e-05,
+      "loss": 0.6249,
+      "step": 13484
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3241813422493247,
+      "learning_rate": 1.5964306831543132e-05,
+      "loss": 0.7572,
+      "step": 13485
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.381331942857178,
+      "learning_rate": 1.5963694348142672e-05,
+      "loss": 0.7215,
+      "step": 13486
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6156094899064072,
+      "learning_rate": 1.5963081830019692e-05,
+      "loss": 0.5797,
+      "step": 13487
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4859595148230447,
+      "learning_rate": 1.5962469277177766e-05,
+      "loss": 0.7014,
+      "step": 13488
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7058613454332856,
+      "learning_rate": 1.5961856689620453e-05,
+      "loss": 0.7193,
+      "step": 13489
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3476072968095523,
+      "learning_rate": 1.5961244067351326e-05,
+      "loss": 0.6964,
+      "step": 13490
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.585427961714725,
+      "learning_rate": 1.5960631410373947e-05,
+      "loss": 0.549,
+      "step": 13491
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3848126154243765,
+      "learning_rate": 1.5960018718691886e-05,
+      "loss": 0.8206,
+      "step": 13492
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9445258643219479,
+      "learning_rate": 1.5959405992308705e-05,
+      "loss": 0.6858,
+      "step": 13493
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.567889198663053,
+      "learning_rate": 1.595879323122798e-05,
+      "loss": 0.8387,
+      "step": 13494
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.597068123084024,
+      "learning_rate": 1.5958180435453273e-05,
+      "loss": 0.8379,
+      "step": 13495
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3515159998251125,
+      "learning_rate": 1.5957567604988154e-05,
+      "loss": 0.7049,
+      "step": 13496
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8726231824050956,
+      "learning_rate": 1.595695473983619e-05,
+      "loss": 0.752,
+      "step": 13497
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.444305918229566,
+      "learning_rate": 1.5956341840000948e-05,
+      "loss": 0.7586,
+      "step": 13498
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.464682338980913,
+      "learning_rate": 1.5955728905486e-05,
+      "loss": 0.7649,
+      "step": 13499
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2993151676666832,
+      "learning_rate": 1.595511593629491e-05,
+      "loss": 0.7567,
+      "step": 13500
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6288000194186334,
+      "learning_rate": 1.595450293243125e-05,
+      "loss": 0.7453,
+      "step": 13501
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3202694187506034,
+      "learning_rate": 1.5953889893898592e-05,
+      "loss": 0.8386,
+      "step": 13502
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.844303703501721,
+      "learning_rate": 1.5953276820700502e-05,
+      "loss": 0.7029,
+      "step": 13503
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3685569679607932,
+      "learning_rate": 1.5952663712840542e-05,
+      "loss": 0.7391,
+      "step": 13504
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3410128259799752,
+      "learning_rate": 1.59520505703223e-05,
+      "loss": 0.7577,
+      "step": 13505
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4073696803585332,
+      "learning_rate": 1.595143739314933e-05,
+      "loss": 0.8831,
+      "step": 13506
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4822063868588136,
+      "learning_rate": 1.5950824181325206e-05,
+      "loss": 0.8788,
+      "step": 13507
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5548896907144214,
+      "learning_rate": 1.5950210934853502e-05,
+      "loss": 0.8379,
+      "step": 13508
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5285803952232664,
+      "learning_rate": 1.5949597653737785e-05,
+      "loss": 0.7212,
+      "step": 13509
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.464544317526743,
+      "learning_rate": 1.594898433798163e-05,
+      "loss": 0.885,
+      "step": 13510
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5684915656837197,
+      "learning_rate": 1.59483709875886e-05,
+      "loss": 0.6604,
+      "step": 13511
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9724431920633227,
+      "learning_rate": 1.5947757602562273e-05,
+      "loss": 0.7808,
+      "step": 13512
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5886210133999152,
+      "learning_rate": 1.5947144182906217e-05,
+      "loss": 0.8068,
+      "step": 13513
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4714664871952094,
+      "learning_rate": 1.5946530728624005e-05,
+      "loss": 0.7406,
+      "step": 13514
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5350285934115286,
+      "learning_rate": 1.594591723971921e-05,
+      "loss": 0.6838,
+      "step": 13515
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3248290392173472,
+      "learning_rate": 1.5945303716195403e-05,
+      "loss": 0.8527,
+      "step": 13516
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4681084108010913,
+      "learning_rate": 1.5944690158056154e-05,
+      "loss": 0.703,
+      "step": 13517
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.448033065803373,
+      "learning_rate": 1.5944076565305032e-05,
+      "loss": 0.8966,
+      "step": 13518
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5494183231773577,
+      "learning_rate": 1.5943462937945624e-05,
+      "loss": 0.8327,
+      "step": 13519
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7900083894284862,
+      "learning_rate": 1.5942849275981488e-05,
+      "loss": 0.7496,
+      "step": 13520
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.648211077230572,
+      "learning_rate": 1.59422355794162e-05,
+      "loss": 0.783,
+      "step": 13521
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3942843453386236,
+      "learning_rate": 1.5941621848253343e-05,
+      "loss": 0.7449,
+      "step": 13522
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3128597419416237,
+      "learning_rate": 1.5941008082496473e-05,
+      "loss": 0.7265,
+      "step": 13523
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5616193881892377,
+      "learning_rate": 1.594039428214918e-05,
+      "loss": 0.7167,
+      "step": 13524
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5133418851076952,
+      "learning_rate": 1.5939780447215024e-05,
+      "loss": 0.6995,
+      "step": 13525
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3618919195654382,
+      "learning_rate": 1.5939166577697593e-05,
+      "loss": 0.6986,
+      "step": 13526
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8957782348606003,
+      "learning_rate": 1.593855267360045e-05,
+      "loss": 0.7247,
+      "step": 13527
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1885027631543377,
+      "learning_rate": 1.5937938734927174e-05,
+      "loss": 0.6454,
+      "step": 13528
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5440237900487292,
+      "learning_rate": 1.593732476168134e-05,
+      "loss": 0.7712,
+      "step": 13529
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3620722454034986,
+      "learning_rate": 1.5936710753866516e-05,
+      "loss": 0.7281,
+      "step": 13530
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6496748054831502,
+      "learning_rate": 1.5936096711486287e-05,
+      "loss": 0.5639,
+      "step": 13531
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6144480023487717,
+      "learning_rate": 1.5935482634544227e-05,
+      "loss": 0.5277,
+      "step": 13532
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7085468590332038,
+      "learning_rate": 1.5934868523043903e-05,
+      "loss": 0.6367,
+      "step": 13533
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4600123957648428,
+      "learning_rate": 1.5934254376988895e-05,
+      "loss": 0.7211,
+      "step": 13534
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.498089735469114,
+      "learning_rate": 1.5933640196382782e-05,
+      "loss": 0.788,
+      "step": 13535
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5509253803014387,
+      "learning_rate": 1.5933025981229136e-05,
+      "loss": 0.8014,
+      "step": 13536
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4220672384704192,
+      "learning_rate": 1.5932411731531535e-05,
+      "loss": 0.7346,
+      "step": 13537
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.677416753822153,
+      "learning_rate": 1.5931797447293553e-05,
+      "loss": 0.5395,
+      "step": 13538
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5663242522300237,
+      "learning_rate": 1.593118312851877e-05,
+      "loss": 0.51,
+      "step": 13539
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9231003703060374,
+      "learning_rate": 1.5930568775210762e-05,
+      "loss": 0.7615,
+      "step": 13540
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6503037777952527,
+      "learning_rate": 1.5929954387373103e-05,
+      "loss": 0.7161,
+      "step": 13541
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.80021340234055,
+      "learning_rate": 1.5929339965009373e-05,
+      "loss": 0.803,
+      "step": 13542
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3454679133720804,
+      "learning_rate": 1.592872550812315e-05,
+      "loss": 0.6117,
+      "step": 13543
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.585302861893721,
+      "learning_rate": 1.5928111016718005e-05,
+      "loss": 0.5064,
+      "step": 13544
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4102192990359448,
+      "learning_rate": 1.5927496490797526e-05,
+      "loss": 0.8099,
+      "step": 13545
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5752973040528249,
+      "learning_rate": 1.5926881930365285e-05,
+      "loss": 0.7554,
+      "step": 13546
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4921695455530757,
+      "learning_rate": 1.5926267335424857e-05,
+      "loss": 0.5986,
+      "step": 13547
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5722181473842479,
+      "learning_rate": 1.5925652705979828e-05,
+      "loss": 0.5615,
+      "step": 13548
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0008622680130355,
+      "learning_rate": 1.592503804203377e-05,
+      "loss": 0.7722,
+      "step": 13549
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4310983550163066,
+      "learning_rate": 1.592442334359027e-05,
+      "loss": 0.6819,
+      "step": 13550
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.729115055442251,
+      "learning_rate": 1.59238086106529e-05,
+      "loss": 0.701,
+      "step": 13551
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8096906913665791,
+      "learning_rate": 1.592319384322524e-05,
+      "loss": 0.8016,
+      "step": 13552
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.676060359059377,
+      "learning_rate": 1.5922579041310866e-05,
+      "loss": 0.7172,
+      "step": 13553
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.848148543858676,
+      "learning_rate": 1.5921964204913366e-05,
+      "loss": 0.8467,
+      "step": 13554
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6964799751620876,
+      "learning_rate": 1.5921349334036316e-05,
+      "loss": 0.7321,
+      "step": 13555
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3160170033049394,
+      "learning_rate": 1.5920734428683295e-05,
+      "loss": 0.7347,
+      "step": 13556
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5015353124643251,
+      "learning_rate": 1.5920119488857885e-05,
+      "loss": 0.797,
+      "step": 13557
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2940160347734417,
+      "learning_rate": 1.591950451456366e-05,
+      "loss": 0.6759,
+      "step": 13558
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5583722653905494,
+      "learning_rate": 1.591888950580421e-05,
+      "loss": 0.8053,
+      "step": 13559
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6802325396067814,
+      "learning_rate": 1.591827446258311e-05,
+      "loss": 0.734,
+      "step": 13560
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4051908048491,
+      "learning_rate": 1.5917659384903944e-05,
+      "loss": 0.741,
+      "step": 13561
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4946842307402601,
+      "learning_rate": 1.5917044272770288e-05,
+      "loss": 0.74,
+      "step": 13562
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9603806908493184,
+      "learning_rate": 1.5916429126185732e-05,
+      "loss": 0.7055,
+      "step": 13563
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6477613322209863,
+      "learning_rate": 1.5915813945153847e-05,
+      "loss": 0.7153,
+      "step": 13564
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6375924562485065,
+      "learning_rate": 1.5915198729678224e-05,
+      "loss": 0.708,
+      "step": 13565
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5020628060593737,
+      "learning_rate": 1.591458347976244e-05,
+      "loss": 0.8123,
+      "step": 13566
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.015045959044575,
+      "learning_rate": 1.5913968195410077e-05,
+      "loss": 0.774,
+      "step": 13567
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.461141526586634,
+      "learning_rate": 1.5913352876624722e-05,
+      "loss": 0.8077,
+      "step": 13568
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.312919865360434,
+      "learning_rate": 1.591273752340995e-05,
+      "loss": 0.6908,
+      "step": 13569
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3998078322186474,
+      "learning_rate": 1.5912122135769356e-05,
+      "loss": 0.6973,
+      "step": 13570
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.387801065933276,
+      "learning_rate": 1.5911506713706507e-05,
+      "loss": 0.6764,
+      "step": 13571
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3547641111988244,
+      "learning_rate": 1.5910891257225e-05,
+      "loss": 0.7505,
+      "step": 13572
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4357209976155123,
+      "learning_rate": 1.5910275766328408e-05,
+      "loss": 0.7682,
+      "step": 13573
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4110351152078144,
+      "learning_rate": 1.5909660241020323e-05,
+      "loss": 0.8868,
+      "step": 13574
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6042599202444113,
+      "learning_rate": 1.5909044681304324e-05,
+      "loss": 0.5519,
+      "step": 13575
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4069235764010202,
+      "learning_rate": 1.5908429087183998e-05,
+      "loss": 0.6566,
+      "step": 13576
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.832755737603838,
+      "learning_rate": 1.5907813458662923e-05,
+      "loss": 0.8815,
+      "step": 13577
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0441689554739466,
+      "learning_rate": 1.590719779574469e-05,
+      "loss": 0.7808,
+      "step": 13578
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6078881733528625,
+      "learning_rate": 1.590658209843288e-05,
+      "loss": 0.7278,
+      "step": 13579
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5682646328944291,
+      "learning_rate": 1.590596636673108e-05,
+      "loss": 0.5331,
+      "step": 13580
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9025566380537633,
+      "learning_rate": 1.5905350600642876e-05,
+      "loss": 0.772,
+      "step": 13581
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5815656513656786,
+      "learning_rate": 1.590473480017185e-05,
+      "loss": 0.5176,
+      "step": 13582
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.603180898212935,
+      "learning_rate": 1.5904118965321585e-05,
+      "loss": 0.7865,
+      "step": 13583
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3758451753328294,
+      "learning_rate": 1.5903503096095674e-05,
+      "loss": 0.7403,
+      "step": 13584
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7265112582255422,
+      "learning_rate": 1.5902887192497698e-05,
+      "loss": 0.7021,
+      "step": 13585
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5727403386900163,
+      "learning_rate": 1.5902271254531245e-05,
+      "loss": 0.5128,
+      "step": 13586
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6006978285268083,
+      "learning_rate": 1.59016552821999e-05,
+      "loss": 0.7569,
+      "step": 13587
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.518594461407732,
+      "learning_rate": 1.5901039275507247e-05,
+      "loss": 0.713,
+      "step": 13588
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.701559431210841,
+      "learning_rate": 1.5900423234456877e-05,
+      "loss": 0.6604,
+      "step": 13589
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3521549404875703,
+      "learning_rate": 1.5899807159052373e-05,
+      "loss": 0.8639,
+      "step": 13590
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5916760744181502,
+      "learning_rate": 1.5899191049297328e-05,
+      "loss": 0.726,
+      "step": 13591
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7177079234861006,
+      "learning_rate": 1.589857490519532e-05,
+      "loss": 0.7119,
+      "step": 13592
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.087247904592742,
+      "learning_rate": 1.5897958726749947e-05,
+      "loss": 0.7177,
+      "step": 13593
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6992532296440346,
+      "learning_rate": 1.589734251396479e-05,
+      "loss": 0.7596,
+      "step": 13594
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4294544980704833,
+      "learning_rate": 1.5896726266843433e-05,
+      "loss": 0.8328,
+      "step": 13595
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6093614960734406,
+      "learning_rate": 1.589610998538947e-05,
+      "loss": 0.6925,
+      "step": 13596
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4941458146271618,
+      "learning_rate": 1.589549366960649e-05,
+      "loss": 0.6721,
+      "step": 13597
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4488002193913727,
+      "learning_rate": 1.589487731949808e-05,
+      "loss": 0.8761,
+      "step": 13598
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.494986604050906,
+      "learning_rate": 1.5894260935067828e-05,
+      "loss": 0.7517,
+      "step": 13599
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.609735004123298,
+      "learning_rate": 1.589364451631932e-05,
+      "loss": 0.5624,
+      "step": 13600
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.8708620714474526,
+      "learning_rate": 1.589302806325615e-05,
+      "loss": 0.7987,
+      "step": 13601
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3337907127167181,
+      "learning_rate": 1.5892411575881906e-05,
+      "loss": 0.6437,
+      "step": 13602
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4722831883918293,
+      "learning_rate": 1.5891795054200173e-05,
+      "loss": 0.8858,
+      "step": 13603
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.437943842821365,
+      "learning_rate": 1.5891178498214545e-05,
+      "loss": 0.9199,
+      "step": 13604
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5964896071810653,
+      "learning_rate": 1.5890561907928612e-05,
+      "loss": 0.7542,
+      "step": 13605
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5073494120152344,
+      "learning_rate": 1.5889945283345963e-05,
+      "loss": 0.7333,
+      "step": 13606
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3320563276862762,
+      "learning_rate": 1.5889328624470186e-05,
+      "loss": 0.7989,
+      "step": 13607
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4919374748311969,
+      "learning_rate": 1.5888711931304874e-05,
+      "loss": 0.7997,
+      "step": 13608
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7078760162631708,
+      "learning_rate": 1.5888095203853614e-05,
+      "loss": 0.7425,
+      "step": 13609
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4137574888874116,
+      "learning_rate": 1.5887478442120007e-05,
+      "loss": 0.8384,
+      "step": 13610
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5853091743094189,
+      "learning_rate": 1.588686164610763e-05,
+      "loss": 0.5355,
+      "step": 13611
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.722321298776749,
+      "learning_rate": 1.5886244815820082e-05,
+      "loss": 0.7445,
+      "step": 13612
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6298702878686373,
+      "learning_rate": 1.5885627951260954e-05,
+      "loss": 0.5798,
+      "step": 13613
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0585416300618995,
+      "learning_rate": 1.5885011052433834e-05,
+      "loss": 0.7273,
+      "step": 13614
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.335986648506517,
+      "learning_rate": 1.588439411934232e-05,
+      "loss": 0.7532,
+      "step": 13615
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6163254584463917,
+      "learning_rate": 1.5883777151990002e-05,
+      "loss": 0.5825,
+      "step": 13616
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6599856920669616,
+      "learning_rate": 1.588316015038047e-05,
+      "loss": 0.7794,
+      "step": 13617
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5490505682174966,
+      "learning_rate": 1.5882543114517313e-05,
+      "loss": 0.7203,
+      "step": 13618
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4698627563476296,
+      "learning_rate": 1.588192604440413e-05,
+      "loss": 0.8085,
+      "step": 13619
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8004061919963648,
+      "learning_rate": 1.588130894004451e-05,
+      "loss": 0.6943,
+      "step": 13620
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9484820866087524,
+      "learning_rate": 1.5880691801442047e-05,
+      "loss": 0.7199,
+      "step": 13621
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.436117633911365,
+      "learning_rate": 1.5880074628600338e-05,
+      "loss": 0.7188,
+      "step": 13622
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4046608490029775,
+      "learning_rate": 1.587945742152297e-05,
+      "loss": 0.6512,
+      "step": 13623
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4419060927991494,
+      "learning_rate": 1.587884018021354e-05,
+      "loss": 0.8311,
+      "step": 13624
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4501650166879136,
+      "learning_rate": 1.5878222904675643e-05,
+      "loss": 0.7506,
+      "step": 13625
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.441601732397454,
+      "learning_rate": 1.587760559491287e-05,
+      "loss": 0.7755,
+      "step": 13626
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.412219367471062,
+      "learning_rate": 1.5876988250928812e-05,
+      "loss": 0.7474,
+      "step": 13627
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.896555224378116,
+      "learning_rate": 1.5876370872727072e-05,
+      "loss": 0.7249,
+      "step": 13628
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4205820011671906,
+      "learning_rate": 1.5875753460311245e-05,
+      "loss": 0.5765,
+      "step": 13629
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8599561702263996,
+      "learning_rate": 1.5875136013684917e-05,
+      "loss": 0.6569,
+      "step": 13630
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4329246139114662,
+      "learning_rate": 1.5874518532851688e-05,
+      "loss": 0.6731,
+      "step": 13631
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5416045423868945,
+      "learning_rate": 1.5873901017815146e-05,
+      "loss": 0.8113,
+      "step": 13632
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5762966828310296,
+      "learning_rate": 1.58732834685789e-05,
+      "loss": 0.7729,
+      "step": 13633
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4891858303250574,
+      "learning_rate": 1.5872665885146535e-05,
+      "loss": 0.718,
+      "step": 13634
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4708727665986248,
+      "learning_rate": 1.587204826752165e-05,
+      "loss": 0.679,
+      "step": 13635
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.450840374724348,
+      "learning_rate": 1.5871430615707843e-05,
+      "loss": 0.8859,
+      "step": 13636
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5853721705684494,
+      "learning_rate": 1.5870812929708706e-05,
+      "loss": 0.6899,
+      "step": 13637
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.007476717747872,
+      "learning_rate": 1.587019520952784e-05,
+      "loss": 0.7829,
+      "step": 13638
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.353445524187487,
+      "learning_rate": 1.5869577455168833e-05,
+      "loss": 0.717,
+      "step": 13639
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6197455479491389,
+      "learning_rate": 1.5868959666635297e-05,
+      "loss": 0.5017,
+      "step": 13640
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.9848798204498026,
+      "learning_rate": 1.5868341843930812e-05,
+      "loss": 0.7731,
+      "step": 13641
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.365412967848017,
+      "learning_rate": 1.5867723987058986e-05,
+      "loss": 0.8444,
+      "step": 13642
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.33219861719356,
+      "learning_rate": 1.586710609602341e-05,
+      "loss": 0.8158,
+      "step": 13643
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7020407008322846,
+      "learning_rate": 1.586648817082769e-05,
+      "loss": 0.7797,
+      "step": 13644
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5827583371724491,
+      "learning_rate": 1.5865870211475417e-05,
+      "loss": 0.5648,
+      "step": 13645
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3481177901305244,
+      "learning_rate": 1.5865252217970187e-05,
+      "loss": 0.6757,
+      "step": 13646
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7203090904671392,
+      "learning_rate": 1.5864634190315608e-05,
+      "loss": 0.778,
+      "step": 13647
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.207125334049291,
+      "learning_rate": 1.5864016128515267e-05,
+      "loss": 0.7487,
+      "step": 13648
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5205596118247326,
+      "learning_rate": 1.586339803257277e-05,
+      "loss": 0.7664,
+      "step": 13649
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6227351079588774,
+      "learning_rate": 1.586277990249171e-05,
+      "loss": 0.7158,
+      "step": 13650
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5895692417039171,
+      "learning_rate": 1.586216173827569e-05,
+      "loss": 0.8398,
+      "step": 13651
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7431867687081513,
+      "learning_rate": 1.5861543539928313e-05,
+      "loss": 0.7018,
+      "step": 13652
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5861827854796318,
+      "learning_rate": 1.586092530745317e-05,
+      "loss": 0.7703,
+      "step": 13653
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.368979155827721,
+      "learning_rate": 1.5860307040853867e-05,
+      "loss": 0.7064,
+      "step": 13654
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.11622498748113,
+      "learning_rate": 1.5859688740133998e-05,
+      "loss": 0.8568,
+      "step": 13655
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6371354541596306,
+      "learning_rate": 1.5859070405297168e-05,
+      "loss": 0.8257,
+      "step": 13656
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6098209528477759,
+      "learning_rate": 1.5858452036346973e-05,
+      "loss": 0.5234,
+      "step": 13657
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5669087708723826,
+      "learning_rate": 1.5857833633287015e-05,
+      "loss": 0.9183,
+      "step": 13658
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5565364637444776,
+      "learning_rate": 1.58572151961209e-05,
+      "loss": 0.6507,
+      "step": 13659
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8243607283227399,
+      "learning_rate": 1.585659672485222e-05,
+      "loss": 0.7195,
+      "step": 13660
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.542705338070292,
+      "learning_rate": 1.5855978219484582e-05,
+      "loss": 0.8674,
+      "step": 13661
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5710972399788588,
+      "learning_rate": 1.5855359680021583e-05,
+      "loss": 0.7723,
+      "step": 13662
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.508442620620337,
+      "learning_rate": 1.5854741106466825e-05,
+      "loss": 0.8175,
+      "step": 13663
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.00552897865688,
+      "learning_rate": 1.585412249882391e-05,
+      "loss": 0.8114,
+      "step": 13664
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5312017918464316,
+      "learning_rate": 1.5853503857096446e-05,
+      "loss": 0.8066,
+      "step": 13665
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6726576734820586,
+      "learning_rate": 1.5852885181288023e-05,
+      "loss": 0.7194,
+      "step": 13666
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9008178366894504,
+      "learning_rate": 1.5852266471402256e-05,
+      "loss": 0.8013,
+      "step": 13667
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2465838317735753,
+      "learning_rate": 1.5851647727442735e-05,
+      "loss": 0.7808,
+      "step": 13668
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.604241596752338,
+      "learning_rate": 1.585102894941307e-05,
+      "loss": 0.7342,
+      "step": 13669
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4885195090612682,
+      "learning_rate": 1.5850410137316865e-05,
+      "loss": 0.6256,
+      "step": 13670
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.564231090778683,
+      "learning_rate": 1.5849791291157717e-05,
+      "loss": 0.7546,
+      "step": 13671
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4362371197334978,
+      "learning_rate": 1.584917241093923e-05,
+      "loss": 0.6507,
+      "step": 13672
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8242456054041303,
+      "learning_rate": 1.5848553496665013e-05,
+      "loss": 0.7844,
+      "step": 13673
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8265856623006376,
+      "learning_rate": 1.5847934548338665e-05,
+      "loss": 0.7592,
+      "step": 13674
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3961845506242432,
+      "learning_rate": 1.584731556596379e-05,
+      "loss": 0.8232,
+      "step": 13675
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5733925074593845,
+      "learning_rate": 1.5846696549543993e-05,
+      "loss": 0.5227,
+      "step": 13676
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6132096238680034,
+      "learning_rate": 1.5846077499082883e-05,
+      "loss": 0.7652,
+      "step": 13677
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5003544224396304,
+      "learning_rate": 1.5845458414584048e-05,
+      "loss": 0.7959,
+      "step": 13678
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3478302507482103,
+      "learning_rate": 1.5844839296051115e-05,
+      "loss": 0.8618,
+      "step": 13679
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6352493850294392,
+      "learning_rate": 1.5844220143487667e-05,
+      "loss": 0.7672,
+      "step": 13680
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.976447588873052,
+      "learning_rate": 1.5843600956897324e-05,
+      "loss": 0.7196,
+      "step": 13681
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7783647047145128,
+      "learning_rate": 1.5842981736283686e-05,
+      "loss": 0.6286,
+      "step": 13682
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7840771107239988,
+      "learning_rate": 1.5842362481650356e-05,
+      "loss": 0.7356,
+      "step": 13683
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5092181108720604,
+      "learning_rate": 1.5841743193000945e-05,
+      "loss": 0.6743,
+      "step": 13684
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7572359797330426,
+      "learning_rate": 1.5841123870339055e-05,
+      "loss": 0.8806,
+      "step": 13685
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9178589877131311,
+      "learning_rate": 1.5840504513668294e-05,
+      "loss": 0.7218,
+      "step": 13686
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5677338288251124,
+      "learning_rate": 1.5839885122992263e-05,
+      "loss": 0.7663,
+      "step": 13687
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9868337559452256,
+      "learning_rate": 1.5839265698314575e-05,
+      "loss": 0.69,
+      "step": 13688
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2114663413996314,
+      "learning_rate": 1.583864623963883e-05,
+      "loss": 0.6918,
+      "step": 13689
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5032890926219415,
+      "learning_rate": 1.5838026746968643e-05,
+      "loss": 0.6556,
+      "step": 13690
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6229066481369063,
+      "learning_rate": 1.5837407220307612e-05,
+      "loss": 0.7691,
+      "step": 13691
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5665002788095257,
+      "learning_rate": 1.583678765965935e-05,
+      "loss": 0.8528,
+      "step": 13692
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6148321102075907,
+      "learning_rate": 1.5836168065027463e-05,
+      "loss": 0.7789,
+      "step": 13693
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6017062832366027,
+      "learning_rate": 1.5835548436415558e-05,
+      "loss": 0.8391,
+      "step": 13694
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4019782989501555,
+      "learning_rate": 1.583492877382724e-05,
+      "loss": 0.7293,
+      "step": 13695
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9718206164798948,
+      "learning_rate": 1.583430907726612e-05,
+      "loss": 0.8265,
+      "step": 13696
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3550002478217011,
+      "learning_rate": 1.5833689346735807e-05,
+      "loss": 0.8014,
+      "step": 13697
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3818665588679453,
+      "learning_rate": 1.5833069582239905e-05,
+      "loss": 0.7894,
+      "step": 13698
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2559867606692037,
+      "learning_rate": 1.5832449783782027e-05,
+      "loss": 0.7359,
+      "step": 13699
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7523325607610607,
+      "learning_rate": 1.583182995136578e-05,
+      "loss": 0.6904,
+      "step": 13700
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.247393322193455,
+      "learning_rate": 1.5831210084994775e-05,
+      "loss": 0.7022,
+      "step": 13701
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5374144077054863,
+      "learning_rate": 1.5830590184672613e-05,
+      "loss": 0.9028,
+      "step": 13702
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.496338610898674,
+      "learning_rate": 1.5829970250402916e-05,
+      "loss": 0.725,
+      "step": 13703
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.520733876382885,
+      "learning_rate": 1.5829350282189278e-05,
+      "loss": 0.7591,
+      "step": 13704
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.52223628246941,
+      "learning_rate": 1.582873028003532e-05,
+      "loss": 0.9245,
+      "step": 13705
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3370637677500112,
+      "learning_rate": 1.582811024394465e-05,
+      "loss": 0.7484,
+      "step": 13706
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5697829974718833,
+      "learning_rate": 1.5827490173920882e-05,
+      "loss": 0.8114,
+      "step": 13707
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4880629886359251,
+      "learning_rate": 1.5826870069967615e-05,
+      "loss": 0.7706,
+      "step": 13708
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.525215880759838,
+      "learning_rate": 1.5826249932088467e-05,
+      "loss": 0.8209,
+      "step": 13709
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.681031931208449,
+      "learning_rate": 1.5825629760287052e-05,
+      "loss": 0.7017,
+      "step": 13710
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5652295161045962,
+      "learning_rate": 1.582500955456697e-05,
+      "loss": 0.6497,
+      "step": 13711
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5234889340525124,
+      "learning_rate": 1.582438931493184e-05,
+      "loss": 0.753,
+      "step": 13712
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2366330247894441,
+      "learning_rate": 1.5823769041385272e-05,
+      "loss": 0.7286,
+      "step": 13713
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.727196714756874,
+      "learning_rate": 1.5823148733930876e-05,
+      "loss": 0.6784,
+      "step": 13714
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4619966111292189,
+      "learning_rate": 1.5822528392572266e-05,
+      "loss": 0.7361,
+      "step": 13715
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.978444891247859,
+      "learning_rate": 1.5821908017313052e-05,
+      "loss": 0.634,
+      "step": 13716
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5392225091538496,
+      "learning_rate": 1.5821287608156847e-05,
+      "loss": 0.7412,
+      "step": 13717
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5441088676345878,
+      "learning_rate": 1.582066716510726e-05,
+      "loss": 0.6561,
+      "step": 13718
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.10888825072958,
+      "learning_rate": 1.582004668816791e-05,
+      "loss": 0.7748,
+      "step": 13719
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5599542631334512,
+      "learning_rate": 1.58194261773424e-05,
+      "loss": 0.5224,
+      "step": 13720
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6216271830548874,
+      "learning_rate": 1.581880563263435e-05,
+      "loss": 0.5521,
+      "step": 13721
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4453399618643308,
+      "learning_rate": 1.5818185054047376e-05,
+      "loss": 0.8156,
+      "step": 13722
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6084989269972466,
+      "learning_rate": 1.5817564441585083e-05,
+      "loss": 0.5941,
+      "step": 13723
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.480824089907059,
+      "learning_rate": 1.581694379525109e-05,
+      "loss": 0.7304,
+      "step": 13724
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4981050468187445,
+      "learning_rate": 1.5816323115049005e-05,
+      "loss": 0.6297,
+      "step": 13725
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1685430245051807,
+      "learning_rate": 1.5815702400982447e-05,
+      "loss": 0.7259,
+      "step": 13726
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5473494342334553,
+      "learning_rate": 1.5815081653055028e-05,
+      "loss": 0.7316,
+      "step": 13727
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4191544466075035,
+      "learning_rate": 1.5814460871270364e-05,
+      "loss": 0.7309,
+      "step": 13728
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4911070411697276,
+      "learning_rate": 1.5813840055632067e-05,
+      "loss": 0.7039,
+      "step": 13729
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9512626403141542,
+      "learning_rate": 1.5813219206143755e-05,
+      "loss": 0.7424,
+      "step": 13730
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4623612281069631,
+      "learning_rate": 1.5812598322809036e-05,
+      "loss": 0.7362,
+      "step": 13731
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.294124815420405,
+      "learning_rate": 1.581197740563153e-05,
+      "loss": 0.7058,
+      "step": 13732
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3975066968943732,
+      "learning_rate": 1.5811356454614853e-05,
+      "loss": 0.8129,
+      "step": 13733
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6366319365558377,
+      "learning_rate": 1.5810735469762623e-05,
+      "loss": 0.5579,
+      "step": 13734
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6043544438512345,
+      "learning_rate": 1.5810114451078446e-05,
+      "loss": 0.7755,
+      "step": 13735
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6036020176057153,
+      "learning_rate": 1.5809493398565945e-05,
+      "loss": 0.5942,
+      "step": 13736
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6380212925347514,
+      "learning_rate": 1.5808872312228735e-05,
+      "loss": 0.571,
+      "step": 13737
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6357551809762754,
+      "learning_rate": 1.5808251192070428e-05,
+      "loss": 0.8171,
+      "step": 13738
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3661932459472088,
+      "learning_rate": 1.5807630038094646e-05,
+      "loss": 0.6694,
+      "step": 13739
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.633645077377181,
+      "learning_rate": 1.5807008850305003e-05,
+      "loss": 0.7812,
+      "step": 13740
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5939158582950235,
+      "learning_rate": 1.580638762870512e-05,
+      "loss": 0.7368,
+      "step": 13741
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6228045524425206,
+      "learning_rate": 1.5805766373298605e-05,
+      "loss": 0.5008,
+      "step": 13742
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3213695480485652,
+      "learning_rate": 1.580514508408908e-05,
+      "loss": 0.7461,
+      "step": 13743
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.680106396529121,
+      "learning_rate": 1.580452376108016e-05,
+      "loss": 0.7594,
+      "step": 13744
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6684409033361036,
+      "learning_rate": 1.580390240427547e-05,
+      "loss": 0.7398,
+      "step": 13745
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5527036566271115,
+      "learning_rate": 1.5803281013678623e-05,
+      "loss": 0.9462,
+      "step": 13746
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6419466884393075,
+      "learning_rate": 1.5802659589293234e-05,
+      "loss": 0.7207,
+      "step": 13747
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.361900654467458,
+      "learning_rate": 1.5802038131122926e-05,
+      "loss": 0.7196,
+      "step": 13748
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5455226108783535,
+      "learning_rate": 1.5801416639171314e-05,
+      "loss": 0.7513,
+      "step": 13749
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3213954581901386,
+      "learning_rate": 1.5800795113442017e-05,
+      "loss": 0.7077,
+      "step": 13750
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4614639201574204,
+      "learning_rate": 1.5800173553938654e-05,
+      "loss": 0.7147,
+      "step": 13751
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.430199274973592,
+      "learning_rate": 1.5799551960664846e-05,
+      "loss": 0.6807,
+      "step": 13752
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3931410496768877,
+      "learning_rate": 1.579893033362421e-05,
+      "loss": 0.6876,
+      "step": 13753
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5056854545091671,
+      "learning_rate": 1.5798308672820363e-05,
+      "loss": 0.7234,
+      "step": 13754
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4235496934333756,
+      "learning_rate": 1.579768697825693e-05,
+      "loss": 0.6698,
+      "step": 13755
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6107347175412868,
+      "learning_rate": 1.5797065249937528e-05,
+      "loss": 0.576,
+      "step": 13756
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3826683886069118,
+      "learning_rate": 1.5796443487865774e-05,
+      "loss": 0.8388,
+      "step": 13757
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4326430521272002,
+      "learning_rate": 1.5795821692045294e-05,
+      "loss": 0.7547,
+      "step": 13758
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5043103791420631,
+      "learning_rate": 1.5795199862479706e-05,
+      "loss": 0.7863,
+      "step": 13759
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4589528684285382,
+      "learning_rate": 1.5794577999172628e-05,
+      "loss": 0.7456,
+      "step": 13760
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8912883475860323,
+      "learning_rate": 1.5793956102127682e-05,
+      "loss": 0.7993,
+      "step": 13761
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5701221621229091,
+      "learning_rate": 1.579333417134849e-05,
+      "loss": 0.8919,
+      "step": 13762
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5022818155873474,
+      "learning_rate": 1.579271220683867e-05,
+      "loss": 0.7281,
+      "step": 13763
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4955804128335204,
+      "learning_rate": 1.5792090208601847e-05,
+      "loss": 0.6924,
+      "step": 13764
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.362512987666945,
+      "learning_rate": 1.5791468176641642e-05,
+      "loss": 0.9421,
+      "step": 13765
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.372942993485862,
+      "learning_rate": 1.5790846110961676e-05,
+      "loss": 0.7715,
+      "step": 13766
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.44611665918134,
+      "learning_rate": 1.5790224011565566e-05,
+      "loss": 0.6884,
+      "step": 13767
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7061549154382405,
+      "learning_rate": 1.5789601878456946e-05,
+      "loss": 0.6787,
+      "step": 13768
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3572184478540277,
+      "learning_rate": 1.5788979711639427e-05,
+      "loss": 0.6957,
+      "step": 13769
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4747605432423618,
+      "learning_rate": 1.5788357511116633e-05,
+      "loss": 0.7558,
+      "step": 13770
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3341693862072226,
+      "learning_rate": 1.5787735276892194e-05,
+      "loss": 0.7483,
+      "step": 13771
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4949487371921328,
+      "learning_rate": 1.5787113008969725e-05,
+      "loss": 0.7469,
+      "step": 13772
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.068758158229426,
+      "learning_rate": 1.5786490707352852e-05,
+      "loss": 0.7779,
+      "step": 13773
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1871794864849214,
+      "learning_rate": 1.57858683720452e-05,
+      "loss": 0.761,
+      "step": 13774
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9624643341226804,
+      "learning_rate": 1.578524600305039e-05,
+      "loss": 0.7948,
+      "step": 13775
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.331937706172651,
+      "learning_rate": 1.578462360037204e-05,
+      "loss": 0.7998,
+      "step": 13776
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4667908121395146,
+      "learning_rate": 1.5784001164013788e-05,
+      "loss": 0.7935,
+      "step": 13777
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5514843644023548,
+      "learning_rate": 1.578337869397925e-05,
+      "loss": 0.7729,
+      "step": 13778
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.418953435978605,
+      "learning_rate": 1.5782756190272047e-05,
+      "loss": 0.7182,
+      "step": 13779
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.930997181140475,
+      "learning_rate": 1.5782133652895804e-05,
+      "loss": 0.6699,
+      "step": 13780
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7060692045128314,
+      "learning_rate": 1.5781511081854152e-05,
+      "loss": 0.6819,
+      "step": 13781
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.53480841724564,
+      "learning_rate": 1.5780888477150712e-05,
+      "loss": 0.7492,
+      "step": 13782
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2725413980620175,
+      "learning_rate": 1.5780265838789107e-05,
+      "loss": 0.7129,
+      "step": 13783
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3436686125476078,
+      "learning_rate": 1.5779643166772966e-05,
+      "loss": 0.6307,
+      "step": 13784
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.929432393446764,
+      "learning_rate": 1.5779020461105913e-05,
+      "loss": 0.7755,
+      "step": 13785
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5412539822357985,
+      "learning_rate": 1.5778397721791574e-05,
+      "loss": 0.6591,
+      "step": 13786
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5284014292410337,
+      "learning_rate": 1.5777774948833574e-05,
+      "loss": 0.8401,
+      "step": 13787
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5018575839994932,
+      "learning_rate": 1.5777152142235535e-05,
+      "loss": 0.7955,
+      "step": 13788
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3319581890778196,
+      "learning_rate": 1.5776529302001093e-05,
+      "loss": 0.6476,
+      "step": 13789
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4133339431344623,
+      "learning_rate": 1.5775906428133866e-05,
+      "loss": 0.8383,
+      "step": 13790
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3804275158252979,
+      "learning_rate": 1.5775283520637483e-05,
+      "loss": 0.8057,
+      "step": 13791
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3099606982495824,
+      "learning_rate": 1.577466057951557e-05,
+      "loss": 0.7147,
+      "step": 13792
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.879353877477045,
+      "learning_rate": 1.5774037604771756e-05,
+      "loss": 0.7422,
+      "step": 13793
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.531644021584246,
+      "learning_rate": 1.5773414596409666e-05,
+      "loss": 0.7589,
+      "step": 13794
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.637170841809069,
+      "learning_rate": 1.577279155443293e-05,
+      "loss": 0.8015,
+      "step": 13795
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.542278261304922,
+      "learning_rate": 1.5772168478845174e-05,
+      "loss": 0.6805,
+      "step": 13796
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.660586752394914,
+      "learning_rate": 1.5771545369650026e-05,
+      "loss": 0.5156,
+      "step": 13797
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5588026843169156,
+      "learning_rate": 1.577092222685111e-05,
+      "loss": 0.8449,
+      "step": 13798
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.394175944921519,
+      "learning_rate": 1.577029905045206e-05,
+      "loss": 0.6557,
+      "step": 13799
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.470581539134103,
+      "learning_rate": 1.5769675840456503e-05,
+      "loss": 0.7201,
+      "step": 13800
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5496565113821044,
+      "learning_rate": 1.5769052596868065e-05,
+      "loss": 0.8053,
+      "step": 13801
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4404928086480973,
+      "learning_rate": 1.5768429319690376e-05,
+      "loss": 0.77,
+      "step": 13802
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5863276761443776,
+      "learning_rate": 1.5767806008927068e-05,
+      "loss": 0.7281,
+      "step": 13803
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9152454494064335,
+      "learning_rate": 1.5767182664581767e-05,
+      "loss": 0.7485,
+      "step": 13804
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8232555006691882,
+      "learning_rate": 1.57665592866581e-05,
+      "loss": 0.8425,
+      "step": 13805
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5967027070453239,
+      "learning_rate": 1.5765935875159703e-05,
+      "loss": 0.5701,
+      "step": 13806
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6097974858171218,
+      "learning_rate": 1.5765312430090197e-05,
+      "loss": 0.5227,
+      "step": 13807
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5886782152233259,
+      "learning_rate": 1.5764688951453216e-05,
+      "loss": 0.6559,
+      "step": 13808
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3006458418953035,
+      "learning_rate": 1.57640654392524e-05,
+      "loss": 0.6924,
+      "step": 13809
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9792372906740687,
+      "learning_rate": 1.5763441893491364e-05,
+      "loss": 0.7844,
+      "step": 13810
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4724598334901642,
+      "learning_rate": 1.5762818314173743e-05,
+      "loss": 0.6545,
+      "step": 13811
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.298433273144213,
+      "learning_rate": 1.576219470130317e-05,
+      "loss": 0.7473,
+      "step": 13812
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4512538031021376,
+      "learning_rate": 1.576157105488328e-05,
+      "loss": 0.8488,
+      "step": 13813
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7684975595735535,
+      "learning_rate": 1.5760947374917693e-05,
+      "loss": 0.7042,
+      "step": 13814
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5705879094181685,
+      "learning_rate": 1.576032366141005e-05,
+      "loss": 0.5259,
+      "step": 13815
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2871136090596276,
+      "learning_rate": 1.5759699914363977e-05,
+      "loss": 0.823,
+      "step": 13816
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6687509438452826,
+      "learning_rate": 1.575907613378311e-05,
+      "loss": 0.653,
+      "step": 13817
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6610980781520857,
+      "learning_rate": 1.5758452319671075e-05,
+      "loss": 0.7515,
+      "step": 13818
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4121871905532002,
+      "learning_rate": 1.5757828472031512e-05,
+      "loss": 0.7551,
+      "step": 13819
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5096105077987891,
+      "learning_rate": 1.5757204590868042e-05,
+      "loss": 0.6274,
+      "step": 13820
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.454392215876855,
+      "learning_rate": 1.575658067618431e-05,
+      "loss": 0.7266,
+      "step": 13821
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6325238811004028,
+      "learning_rate": 1.5755956727983943e-05,
+      "loss": 0.6534,
+      "step": 13822
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1018668960830493,
+      "learning_rate": 1.5755332746270574e-05,
+      "loss": 0.7825,
+      "step": 13823
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3461470941046634,
+      "learning_rate": 1.5754708731047833e-05,
+      "loss": 0.7397,
+      "step": 13824
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3992437467541743,
+      "learning_rate": 1.5754084682319356e-05,
+      "loss": 0.801,
+      "step": 13825
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.712590960953921,
+      "learning_rate": 1.5753460600088777e-05,
+      "loss": 0.7565,
+      "step": 13826
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5284414398487345,
+      "learning_rate": 1.5752836484359726e-05,
+      "loss": 0.7779,
+      "step": 13827
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7507079143531725,
+      "learning_rate": 1.575221233513584e-05,
+      "loss": 0.7795,
+      "step": 13828
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4210895717887448,
+      "learning_rate": 1.5751588152420757e-05,
+      "loss": 0.7569,
+      "step": 13829
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9574690308123386,
+      "learning_rate": 1.5750963936218104e-05,
+      "loss": 0.8599,
+      "step": 13830
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6205937709505411,
+      "learning_rate": 1.575033968653152e-05,
+      "loss": 0.8104,
+      "step": 13831
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3255020074317034,
+      "learning_rate": 1.5749715403364635e-05,
+      "loss": 0.6858,
+      "step": 13832
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.432956401166676,
+      "learning_rate": 1.574909108672109e-05,
+      "loss": 0.6613,
+      "step": 13833
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4137251339885535,
+      "learning_rate": 1.5748466736604514e-05,
+      "loss": 0.7635,
+      "step": 13834
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.655297111451651,
+      "learning_rate": 1.574784235301855e-05,
+      "loss": 0.7516,
+      "step": 13835
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.578101434314941,
+      "learning_rate": 1.574721793596682e-05,
+      "loss": 0.5218,
+      "step": 13836
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8858403581777383,
+      "learning_rate": 1.5746593485452973e-05,
+      "loss": 0.6453,
+      "step": 13837
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3390213155877795,
+      "learning_rate": 1.5745969001480637e-05,
+      "loss": 0.6673,
+      "step": 13838
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4434003880689008,
+      "learning_rate": 1.574534448405345e-05,
+      "loss": 0.645,
+      "step": 13839
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.693060644260869,
+      "learning_rate": 1.5744719933175053e-05,
+      "loss": 0.7093,
+      "step": 13840
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5615496081276614,
+      "learning_rate": 1.5744095348849073e-05,
+      "loss": 0.6288,
+      "step": 13841
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5738006445160824,
+      "learning_rate": 1.5743470731079152e-05,
+      "loss": 0.8004,
+      "step": 13842
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3194754413914451,
+      "learning_rate": 1.5742846079868926e-05,
+      "loss": 0.6475,
+      "step": 13843
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6430831250640985,
+      "learning_rate": 1.5742221395222033e-05,
+      "loss": 0.7368,
+      "step": 13844
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6389290823805596,
+      "learning_rate": 1.574159667714211e-05,
+      "loss": 0.547,
+      "step": 13845
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.257330178262883,
+      "learning_rate": 1.574097192563279e-05,
+      "loss": 0.7347,
+      "step": 13846
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4034612165614737,
+      "learning_rate": 1.574034714069772e-05,
+      "loss": 0.8667,
+      "step": 13847
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.423488135744855,
+      "learning_rate": 1.5739722322340524e-05,
+      "loss": 0.6692,
+      "step": 13848
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6931918485980912,
+      "learning_rate": 1.5739097470564853e-05,
+      "loss": 0.9033,
+      "step": 13849
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.303290573124369,
+      "learning_rate": 1.5738472585374334e-05,
+      "loss": 0.8043,
+      "step": 13850
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3652845097184727,
+      "learning_rate": 1.5737847666772615e-05,
+      "loss": 0.8511,
+      "step": 13851
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5850017320391683,
+      "learning_rate": 1.573722271476333e-05,
+      "loss": 0.8103,
+      "step": 13852
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3375191780586493,
+      "learning_rate": 1.573659772935012e-05,
+      "loss": 0.6024,
+      "step": 13853
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7265717877137114,
+      "learning_rate": 1.5735972710536623e-05,
+      "loss": 0.6955,
+      "step": 13854
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3595647535578987,
+      "learning_rate": 1.5735347658326472e-05,
+      "loss": 0.7037,
+      "step": 13855
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6068298427660772,
+      "learning_rate": 1.5734722572723313e-05,
+      "loss": 0.6843,
+      "step": 13856
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4653617667740628,
+      "learning_rate": 1.573409745373078e-05,
+      "loss": 0.5911,
+      "step": 13857
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.716919450271619,
+      "learning_rate": 1.573347230135252e-05,
+      "loss": 0.7982,
+      "step": 13858
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4445359984426416,
+      "learning_rate": 1.573284711559217e-05,
+      "loss": 0.7138,
+      "step": 13859
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.427485384442732,
+      "learning_rate": 1.5732221896453368e-05,
+      "loss": 0.7853,
+      "step": 13860
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4679112362266824,
+      "learning_rate": 1.5731596643939755e-05,
+      "loss": 0.8127,
+      "step": 13861
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3236429972420023,
+      "learning_rate": 1.5730971358054972e-05,
+      "loss": 0.7141,
+      "step": 13862
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4894684687916346,
+      "learning_rate": 1.5730346038802657e-05,
+      "loss": 0.742,
+      "step": 13863
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3787643153611717,
+      "learning_rate": 1.5729720686186457e-05,
+      "loss": 0.7454,
+      "step": 13864
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5038403284790136,
+      "learning_rate": 1.5729095300210006e-05,
+      "loss": 0.8323,
+      "step": 13865
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.384281245725252,
+      "learning_rate": 1.572846988087695e-05,
+      "loss": 0.8945,
+      "step": 13866
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6331405483244331,
+      "learning_rate": 1.572784442819093e-05,
+      "loss": 0.5883,
+      "step": 13867
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5457577320734281,
+      "learning_rate": 1.5727218942155588e-05,
+      "loss": 0.7329,
+      "step": 13868
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6273502079980733,
+      "learning_rate": 1.5726593422774558e-05,
+      "loss": 0.6762,
+      "step": 13869
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5636039796596788,
+      "learning_rate": 1.572596787005149e-05,
+      "loss": 0.907,
+      "step": 13870
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4407484642268382,
+      "learning_rate": 1.5725342283990027e-05,
+      "loss": 0.6756,
+      "step": 13871
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3241418449178806,
+      "learning_rate": 1.572471666459381e-05,
+      "loss": 0.755,
+      "step": 13872
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6228472250947085,
+      "learning_rate": 1.5724091011866474e-05,
+      "loss": 0.5768,
+      "step": 13873
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6248556846611336,
+      "learning_rate": 1.572346532581167e-05,
+      "loss": 0.5687,
+      "step": 13874
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6018594821190577,
+      "learning_rate": 1.5722839606433043e-05,
+      "loss": 0.6226,
+      "step": 13875
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.325516423772817,
+      "learning_rate": 1.5722213853734227e-05,
+      "loss": 0.7165,
+      "step": 13876
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3263163527611614,
+      "learning_rate": 1.572158806771887e-05,
+      "loss": 0.6935,
+      "step": 13877
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6870342344828955,
+      "learning_rate": 1.572096224839062e-05,
+      "loss": 0.6693,
+      "step": 13878
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.844129201154196,
+      "learning_rate": 1.5720336395753117e-05,
+      "loss": 0.8621,
+      "step": 13879
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.534541235897263,
+      "learning_rate": 1.571971050981e-05,
+      "loss": 0.7073,
+      "step": 13880
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3747736550092866,
+      "learning_rate": 1.571908459056492e-05,
+      "loss": 0.8,
+      "step": 13881
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.640438790367557,
+      "learning_rate": 1.571845863802152e-05,
+      "loss": 0.6836,
+      "step": 13882
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7156189611909964,
+      "learning_rate": 1.571783265218344e-05,
+      "loss": 0.8186,
+      "step": 13883
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3728027966614813,
+      "learning_rate": 1.5717206633054328e-05,
+      "loss": 0.7343,
+      "step": 13884
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5186356747997194,
+      "learning_rate": 1.5716580580637833e-05,
+      "loss": 0.7076,
+      "step": 13885
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3194722884103522,
+      "learning_rate": 1.5715954494937595e-05,
+      "loss": 0.7126,
+      "step": 13886
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3926864503582101,
+      "learning_rate": 1.5715328375957256e-05,
+      "loss": 0.718,
+      "step": 13887
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6099814347363164,
+      "learning_rate": 1.5714702223700472e-05,
+      "loss": 0.5547,
+      "step": 13888
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8944823657259766,
+      "learning_rate": 1.571407603817088e-05,
+      "loss": 0.6596,
+      "step": 13889
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.584418329551435,
+      "learning_rate": 1.571344981937213e-05,
+      "loss": 0.8009,
+      "step": 13890
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2709213971002484,
+      "learning_rate": 1.571282356730786e-05,
+      "loss": 0.7487,
+      "step": 13891
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3051004998396876,
+      "learning_rate": 1.5712197281981727e-05,
+      "loss": 0.729,
+      "step": 13892
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5422531954773055,
+      "learning_rate": 1.5711570963397376e-05,
+      "loss": 0.7676,
+      "step": 13893
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8240867710256046,
+      "learning_rate": 1.5710944611558445e-05,
+      "loss": 0.7636,
+      "step": 13894
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0181542075340735,
+      "learning_rate": 1.5710318226468587e-05,
+      "loss": 0.8302,
+      "step": 13895
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4586519039915844,
+      "learning_rate": 1.570969180813145e-05,
+      "loss": 0.7468,
+      "step": 13896
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5025466506324894,
+      "learning_rate": 1.570906535655068e-05,
+      "loss": 0.7619,
+      "step": 13897
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4782850232988798,
+      "learning_rate": 1.5708438871729925e-05,
+      "loss": 0.9316,
+      "step": 13898
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5283065103442381,
+      "learning_rate": 1.570781235367283e-05,
+      "loss": 0.6591,
+      "step": 13899
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4071768514309482,
+      "learning_rate": 1.5707185802383045e-05,
+      "loss": 0.6899,
+      "step": 13900
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3587670243058847,
+      "learning_rate": 1.5706559217864214e-05,
+      "loss": 0.8585,
+      "step": 13901
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.503773376581411,
+      "learning_rate": 1.5705932600119993e-05,
+      "loss": 0.8859,
+      "step": 13902
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5951714327495874,
+      "learning_rate": 1.5705305949154025e-05,
+      "loss": 0.5555,
+      "step": 13903
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4337266537722693,
+      "learning_rate": 1.570467926496996e-05,
+      "loss": 0.7425,
+      "step": 13904
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5199672458284097,
+      "learning_rate": 1.5704052547571445e-05,
+      "loss": 0.7997,
+      "step": 13905
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4198434356727523,
+      "learning_rate": 1.570342579696213e-05,
+      "loss": 0.7049,
+      "step": 13906
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6130196323603891,
+      "learning_rate": 1.5702799013145665e-05,
+      "loss": 0.5301,
+      "step": 13907
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3938260382306369,
+      "learning_rate": 1.57021721961257e-05,
+      "loss": 0.6227,
+      "step": 13908
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5123214304363275,
+      "learning_rate": 1.570154534590588e-05,
+      "loss": 0.8053,
+      "step": 13909
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5183497098897756,
+      "learning_rate": 1.570091846248986e-05,
+      "loss": 0.8181,
+      "step": 13910
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9050889431716738,
+      "learning_rate": 1.570029154588129e-05,
+      "loss": 0.8173,
+      "step": 13911
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6268827030313661,
+      "learning_rate": 1.5699664596083814e-05,
+      "loss": 0.5414,
+      "step": 13912
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.090734204572233,
+      "learning_rate": 1.5699037613101087e-05,
+      "loss": 0.602,
+      "step": 13913
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5596847628879555,
+      "learning_rate": 1.569841059693676e-05,
+      "loss": 0.9576,
+      "step": 13914
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4186277621425027,
+      "learning_rate": 1.5697783547594482e-05,
+      "loss": 0.7062,
+      "step": 13915
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5131424829217515,
+      "learning_rate": 1.5697156465077903e-05,
+      "loss": 0.6988,
+      "step": 13916
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.349229854219815,
+      "learning_rate": 1.5696529349390676e-05,
+      "loss": 0.7538,
+      "step": 13917
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5634431935762196,
+      "learning_rate": 1.569590220053645e-05,
+      "loss": 0.8015,
+      "step": 13918
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6469375939613715,
+      "learning_rate": 1.5695275018518885e-05,
+      "loss": 0.8717,
+      "step": 13919
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3995678923231152,
+      "learning_rate": 1.5694647803341618e-05,
+      "loss": 0.6505,
+      "step": 13920
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.696145975429518,
+      "learning_rate": 1.5694020555008314e-05,
+      "loss": 0.6855,
+      "step": 13921
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6515733351059232,
+      "learning_rate": 1.5693393273522618e-05,
+      "loss": 0.7615,
+      "step": 13922
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.385689077830168,
+      "learning_rate": 1.5692765958888182e-05,
+      "loss": 0.7065,
+      "step": 13923
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7181772668694175,
+      "learning_rate": 1.569213861110866e-05,
+      "loss": 0.7961,
+      "step": 13924
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.544276474688856,
+      "learning_rate": 1.5691511230187706e-05,
+      "loss": 0.7113,
+      "step": 13925
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0359502964626173,
+      "learning_rate": 1.5690883816128973e-05,
+      "loss": 0.7592,
+      "step": 13926
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4491225163409382,
+      "learning_rate": 1.569025636893611e-05,
+      "loss": 0.8091,
+      "step": 13927
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3510419159421798,
+      "learning_rate": 1.5689628888612774e-05,
+      "loss": 0.6702,
+      "step": 13928
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5584037018784405,
+      "learning_rate": 1.568900137516262e-05,
+      "loss": 0.6886,
+      "step": 13929
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2732553611084436,
+      "learning_rate": 1.5688373828589296e-05,
+      "loss": 0.7488,
+      "step": 13930
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6525060992302253,
+      "learning_rate": 1.568774624889646e-05,
+      "loss": 0.5564,
+      "step": 13931
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3961220349142012,
+      "learning_rate": 1.5687118636087764e-05,
+      "loss": 0.7701,
+      "step": 13932
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2624126865138194,
+      "learning_rate": 1.5686490990166864e-05,
+      "loss": 0.6951,
+      "step": 13933
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3712908894083118,
+      "learning_rate": 1.568586331113741e-05,
+      "loss": 0.7546,
+      "step": 13934
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3923517729085921,
+      "learning_rate": 1.5685235599003065e-05,
+      "loss": 0.682,
+      "step": 13935
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6925105022680358,
+      "learning_rate": 1.5684607853767474e-05,
+      "loss": 0.7175,
+      "step": 13936
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6193358262577231,
+      "learning_rate": 1.5683980075434297e-05,
+      "loss": 0.5742,
+      "step": 13937
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4272038147825268,
+      "learning_rate": 1.5683352264007188e-05,
+      "loss": 0.7717,
+      "step": 13938
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.42792072083802,
+      "learning_rate": 1.5682724419489805e-05,
+      "loss": 0.8017,
+      "step": 13939
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.723443857115787,
+      "learning_rate": 1.5682096541885797e-05,
+      "loss": 0.8094,
+      "step": 13940
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5312572830962725,
+      "learning_rate": 1.568146863119883e-05,
+      "loss": 0.8203,
+      "step": 13941
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4002630898282793,
+      "learning_rate": 1.568084068743255e-05,
+      "loss": 0.6465,
+      "step": 13942
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4709402797264,
+      "learning_rate": 1.568021271059062e-05,
+      "loss": 0.846,
+      "step": 13943
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5433013312750496,
+      "learning_rate": 1.567958470067669e-05,
+      "loss": 0.852,
+      "step": 13944
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5705451696307706,
+      "learning_rate": 1.567895665769442e-05,
+      "loss": 0.6979,
+      "step": 13945
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.5599795448337557,
+      "learning_rate": 1.5678328581647467e-05,
+      "loss": 0.7338,
+      "step": 13946
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4446145798067167,
+      "learning_rate": 1.567770047253949e-05,
+      "loss": 0.736,
+      "step": 13947
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.312257500991238,
+      "learning_rate": 1.5677072330374138e-05,
+      "loss": 0.821,
+      "step": 13948
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2806148442458682,
+      "learning_rate": 1.567644415515508e-05,
+      "loss": 0.7291,
+      "step": 13949
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3835926305580162,
+      "learning_rate": 1.567581594688596e-05,
+      "loss": 0.7776,
+      "step": 13950
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4794835498175254,
+      "learning_rate": 1.567518770557045e-05,
+      "loss": 0.8009,
+      "step": 13951
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.474622813476976,
+      "learning_rate": 1.5674559431212193e-05,
+      "loss": 0.7191,
+      "step": 13952
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.734095995350619,
+      "learning_rate": 1.5673931123814862e-05,
+      "loss": 0.8178,
+      "step": 13953
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7181832023992993,
+      "learning_rate": 1.5673302783382104e-05,
+      "loss": 0.8662,
+      "step": 13954
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5160446658056872,
+      "learning_rate": 1.5672674409917575e-05,
+      "loss": 0.6954,
+      "step": 13955
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6381848297953028,
+      "learning_rate": 1.5672046003424948e-05,
+      "loss": 0.6951,
+      "step": 13956
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.533185662818209,
+      "learning_rate": 1.5671417563907872e-05,
+      "loss": 0.8267,
+      "step": 13957
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5059168312159479,
+      "learning_rate": 1.5670789091370005e-05,
+      "loss": 0.7432,
+      "step": 13958
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5981132615607199,
+      "learning_rate": 1.567016058581501e-05,
+      "loss": 0.7572,
+      "step": 13959
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.490434607121477,
+      "learning_rate": 1.5669532047246546e-05,
+      "loss": 0.7979,
+      "step": 13960
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6620347050836024,
+      "learning_rate": 1.566890347566827e-05,
+      "loss": 0.9214,
+      "step": 13961
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7245902003390656,
+      "learning_rate": 1.5668274871083844e-05,
+      "loss": 0.7041,
+      "step": 13962
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.438971007538669,
+      "learning_rate": 1.5667646233496925e-05,
+      "loss": 0.8067,
+      "step": 13963
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.7880651387678177,
+      "learning_rate": 1.566701756291118e-05,
+      "loss": 0.8591,
+      "step": 13964
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3817966296574942,
+      "learning_rate": 1.566638885933026e-05,
+      "loss": 0.7369,
+      "step": 13965
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3254255820832368,
+      "learning_rate": 1.566576012275783e-05,
+      "loss": 0.8139,
+      "step": 13966
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3125358928697275,
+      "learning_rate": 1.5665131353197555e-05,
+      "loss": 0.7771,
+      "step": 13967
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5217003805764113,
+      "learning_rate": 1.5664502550653088e-05,
+      "loss": 0.6743,
+      "step": 13968
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3750977455253603,
+      "learning_rate": 1.566387371512809e-05,
+      "loss": 0.667,
+      "step": 13969
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4918006575189942,
+      "learning_rate": 1.566324484662624e-05,
+      "loss": 0.8564,
+      "step": 13970
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6160331587691588,
+      "learning_rate": 1.5662615945151173e-05,
+      "loss": 0.8278,
+      "step": 13971
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7498255885339107,
+      "learning_rate": 1.566198701070657e-05,
+      "loss": 0.8121,
+      "step": 13972
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.568929845435121,
+      "learning_rate": 1.566135804329608e-05,
+      "loss": 0.714,
+      "step": 13973
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.231774500009965,
+      "learning_rate": 1.5660729042923374e-05,
+      "loss": 0.7468,
+      "step": 13974
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.60389105322489,
+      "learning_rate": 1.566010000959211e-05,
+      "loss": 0.7035,
+      "step": 13975
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8045611255439982,
+      "learning_rate": 1.5659470943305956e-05,
+      "loss": 0.8781,
+      "step": 13976
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.488199734064819,
+      "learning_rate": 1.5658841844068568e-05,
+      "loss": 0.8009,
+      "step": 13977
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6029886680571326,
+      "learning_rate": 1.565821271188361e-05,
+      "loss": 0.8942,
+      "step": 13978
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5093948113525157,
+      "learning_rate": 1.5657583546754744e-05,
+      "loss": 0.9315,
+      "step": 13979
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5124768228603258,
+      "learning_rate": 1.565695434868564e-05,
+      "loss": 0.8217,
+      "step": 13980
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.461913418106757,
+      "learning_rate": 1.5656325117679957e-05,
+      "loss": 0.8609,
+      "step": 13981
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.503113246503224,
+      "learning_rate": 1.5655695853741353e-05,
+      "loss": 0.7217,
+      "step": 13982
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3969254129263726,
+      "learning_rate": 1.5655066556873502e-05,
+      "loss": 0.8828,
+      "step": 13983
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7058526953006847,
+      "learning_rate": 1.565443722708006e-05,
+      "loss": 0.6364,
+      "step": 13984
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9505281880637524,
+      "learning_rate": 1.5653807864364696e-05,
+      "loss": 0.7075,
+      "step": 13985
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.497521740494836,
+      "learning_rate": 1.5653178468731076e-05,
+      "loss": 0.7473,
+      "step": 13986
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3626563804563538,
+      "learning_rate": 1.5652549040182857e-05,
+      "loss": 0.8137,
+      "step": 13987
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4157248435323755,
+      "learning_rate": 1.5651919578723708e-05,
+      "loss": 0.6853,
+      "step": 13988
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.003751322607319,
+      "learning_rate": 1.5651290084357295e-05,
+      "loss": 0.8565,
+      "step": 13989
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5058878372631372,
+      "learning_rate": 1.5650660557087284e-05,
+      "loss": 0.7441,
+      "step": 13990
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7698892610065171,
+      "learning_rate": 1.5650030996917334e-05,
+      "loss": 0.8795,
+      "step": 13991
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.456338319368894,
+      "learning_rate": 1.5649401403851115e-05,
+      "loss": 0.6771,
+      "step": 13992
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5907292124368169,
+      "learning_rate": 1.5648771777892296e-05,
+      "loss": 0.5193,
+      "step": 13993
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4899422294584015,
+      "learning_rate": 1.564814211904454e-05,
+      "loss": 0.7885,
+      "step": 13994
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9136086283130822,
+      "learning_rate": 1.5647512427311508e-05,
+      "loss": 0.778,
+      "step": 13995
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5201131176177247,
+      "learning_rate": 1.564688270269687e-05,
+      "loss": 0.691,
+      "step": 13996
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6283253503396036,
+      "learning_rate": 1.5646252945204298e-05,
+      "loss": 0.5158,
+      "step": 13997
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4421894110431936,
+      "learning_rate": 1.5645623154837454e-05,
+      "loss": 0.8006,
+      "step": 13998
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7793883566917132,
+      "learning_rate": 1.56449933316e-05,
+      "loss": 0.7689,
+      "step": 13999
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3258230075341906,
+      "learning_rate": 1.5644363475495614e-05,
+      "loss": 0.8081,
+      "step": 14000
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5720686205868929,
+      "learning_rate": 1.564373358652795e-05,
+      "loss": 0.504,
+      "step": 14001
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6117859002318519,
+      "learning_rate": 1.5643103664700687e-05,
+      "loss": 0.5,
+      "step": 14002
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5302029675634135,
+      "learning_rate": 1.564247371001748e-05,
+      "loss": 0.8059,
+      "step": 14003
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8973585337505185,
+      "learning_rate": 1.5641843722482015e-05,
+      "loss": 0.636,
+      "step": 14004
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.551755487448227,
+      "learning_rate": 1.5641213702097947e-05,
+      "loss": 0.7781,
+      "step": 14005
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4767209053535926,
+      "learning_rate": 1.5640583648868943e-05,
+      "loss": 0.6952,
+      "step": 14006
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.428294901530772,
+      "learning_rate": 1.5639953562798676e-05,
+      "loss": 0.8789,
+      "step": 14007
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9306751606298564,
+      "learning_rate": 1.563932344389082e-05,
+      "loss": 0.7878,
+      "step": 14008
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6047488088519994,
+      "learning_rate": 1.563869329214903e-05,
+      "loss": 0.7757,
+      "step": 14009
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4896791190609902,
+      "learning_rate": 1.5638063107576983e-05,
+      "loss": 0.6806,
+      "step": 14010
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4543471819638132,
+      "learning_rate": 1.5637432890178354e-05,
+      "loss": 0.7417,
+      "step": 14011
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7219375125719714,
+      "learning_rate": 1.5636802639956798e-05,
+      "loss": 0.8837,
+      "step": 14012
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6762212126811746,
+      "learning_rate": 1.5636172356915995e-05,
+      "loss": 0.784,
+      "step": 14013
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5442678116435515,
+      "learning_rate": 1.5635542041059615e-05,
+      "loss": 0.5838,
+      "step": 14014
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2105020276962044,
+      "learning_rate": 1.5634911692391323e-05,
+      "loss": 0.7553,
+      "step": 14015
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.59616189663655,
+      "learning_rate": 1.5634281310914792e-05,
+      "loss": 0.7491,
+      "step": 14016
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.581339055723967,
+      "learning_rate": 1.563365089663369e-05,
+      "loss": 0.8062,
+      "step": 14017
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3930369543393688,
+      "learning_rate": 1.5633020449551694e-05,
+      "loss": 0.7261,
+      "step": 14018
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4334769816841846,
+      "learning_rate": 1.5632389969672462e-05,
+      "loss": 0.7392,
+      "step": 14019
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.424019207898748,
+      "learning_rate": 1.5631759456999675e-05,
+      "loss": 0.7309,
+      "step": 14020
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9981374374859115,
+      "learning_rate": 1.5631128911537007e-05,
+      "loss": 0.7367,
+      "step": 14021
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.500341538145777,
+      "learning_rate": 1.5630498333288115e-05,
+      "loss": 0.8919,
+      "step": 14022
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.009720941945043,
+      "learning_rate": 1.5629867722256684e-05,
+      "loss": 0.8032,
+      "step": 14023
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6032003334362498,
+      "learning_rate": 1.5629237078446378e-05,
+      "loss": 0.5317,
+      "step": 14024
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4575843709751704,
+      "learning_rate": 1.5628606401860875e-05,
+      "loss": 0.6973,
+      "step": 14025
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.363289950083517,
+      "learning_rate": 1.562797569250384e-05,
+      "loss": 0.8152,
+      "step": 14026
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4762137681283016,
+      "learning_rate": 1.562734495037895e-05,
+      "loss": 0.6188,
+      "step": 14027
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4714143022298245,
+      "learning_rate": 1.5626714175489877e-05,
+      "loss": 0.8313,
+      "step": 14028
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.66740411485695,
+      "learning_rate": 1.5626083367840296e-05,
+      "loss": 0.7935,
+      "step": 14029
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6145175922422838,
+      "learning_rate": 1.562545252743387e-05,
+      "loss": 0.568,
+      "step": 14030
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.537007839716178,
+      "learning_rate": 1.562482165427428e-05,
+      "loss": 0.7049,
+      "step": 14031
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.214759919552835,
+      "learning_rate": 1.56241907483652e-05,
+      "loss": 0.6655,
+      "step": 14032
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.185244158751765,
+      "learning_rate": 1.56235598097103e-05,
+      "loss": 0.7699,
+      "step": 14033
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4869475640484375,
+      "learning_rate": 1.5622928838313254e-05,
+      "loss": 0.7697,
+      "step": 14034
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.495456685402603,
+      "learning_rate": 1.5622297834177738e-05,
+      "loss": 0.7188,
+      "step": 14035
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4781787783000815,
+      "learning_rate": 1.562166679730742e-05,
+      "loss": 0.8434,
+      "step": 14036
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.51922544453725,
+      "learning_rate": 1.5621035727705982e-05,
+      "loss": 0.7063,
+      "step": 14037
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4399117498749168,
+      "learning_rate": 1.562040462537709e-05,
+      "loss": 0.7353,
+      "step": 14038
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.641258892014925,
+      "learning_rate": 1.5619773490324427e-05,
+      "loss": 0.5214,
+      "step": 14039
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7057196032547062,
+      "learning_rate": 1.561914232255166e-05,
+      "loss": 0.7267,
+      "step": 14040
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.154205736773252,
+      "learning_rate": 1.5618511122062475e-05,
+      "loss": 0.7107,
+      "step": 14041
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2520042789904102,
+      "learning_rate": 1.5617879888860533e-05,
+      "loss": 0.7721,
+      "step": 14042
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6621054086686184,
+      "learning_rate": 1.5617248622949515e-05,
+      "loss": 0.7705,
+      "step": 14043
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3625259174408506,
+      "learning_rate": 1.56166173243331e-05,
+      "loss": 0.699,
+      "step": 14044
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.848127513730617,
+      "learning_rate": 1.5615985993014956e-05,
+      "loss": 0.7362,
+      "step": 14045
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5830994890293439,
+      "learning_rate": 1.561535462899877e-05,
+      "loss": 0.4795,
+      "step": 14046
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4440130057791734,
+      "learning_rate": 1.561472323228821e-05,
+      "loss": 0.7142,
+      "step": 14047
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5883532366158887,
+      "learning_rate": 1.561409180288695e-05,
+      "loss": 0.5373,
+      "step": 14048
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6094240976368517,
+      "learning_rate": 1.5613460340798674e-05,
+      "loss": 0.8156,
+      "step": 14049
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2592202866565245,
+      "learning_rate": 1.5612828846027052e-05,
+      "loss": 0.7463,
+      "step": 14050
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5815681204653639,
+      "learning_rate": 1.5612197318575767e-05,
+      "loss": 0.523,
+      "step": 14051
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.725244802353335,
+      "learning_rate": 1.561156575844849e-05,
+      "loss": 0.8115,
+      "step": 14052
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6329343292650071,
+      "learning_rate": 1.56109341656489e-05,
+      "loss": 0.693,
+      "step": 14053
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.819379424863428,
+      "learning_rate": 1.5610302540180674e-05,
+      "loss": 0.6762,
+      "step": 14054
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7720575668463667,
+      "learning_rate": 1.5609670882047492e-05,
+      "loss": 0.7982,
+      "step": 14055
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3020504985434358,
+      "learning_rate": 1.5609039191253032e-05,
+      "loss": 0.6194,
+      "step": 14056
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.119225207226515,
+      "learning_rate": 1.560840746780097e-05,
+      "loss": 0.8022,
+      "step": 14057
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4783239310432355,
+      "learning_rate": 1.560777571169498e-05,
+      "loss": 0.7461,
+      "step": 14058
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6502652801808082,
+      "learning_rate": 1.5607143922938747e-05,
+      "loss": 0.7383,
+      "step": 14059
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.340310559215552,
+      "learning_rate": 1.5606512101535947e-05,
+      "loss": 0.7431,
+      "step": 14060
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4209193266237616,
+      "learning_rate": 1.560588024749026e-05,
+      "loss": 0.7097,
+      "step": 14061
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4906901211496213,
+      "learning_rate": 1.5605248360805358e-05,
+      "loss": 0.66,
+      "step": 14062
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3842638623068482,
+      "learning_rate": 1.560461644148493e-05,
+      "loss": 0.7672,
+      "step": 14063
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.3367492372304137,
+      "learning_rate": 1.560398448953265e-05,
+      "loss": 0.6623,
+      "step": 14064
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.368165036318723,
+      "learning_rate": 1.56033525049522e-05,
+      "loss": 0.7432,
+      "step": 14065
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4964051861497263,
+      "learning_rate": 1.560272048774725e-05,
+      "loss": 0.9077,
+      "step": 14066
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2350290183657513,
+      "learning_rate": 1.5602088437921497e-05,
+      "loss": 0.6757,
+      "step": 14067
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4268362736511873,
+      "learning_rate": 1.5601456355478606e-05,
+      "loss": 0.6923,
+      "step": 14068
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4270052807632885,
+      "learning_rate": 1.5600824240422266e-05,
+      "loss": 0.8676,
+      "step": 14069
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3516088285526755,
+      "learning_rate": 1.5600192092756154e-05,
+      "loss": 0.7652,
+      "step": 14070
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4531967409812152,
+      "learning_rate": 1.5599559912483945e-05,
+      "loss": 0.7431,
+      "step": 14071
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4432282125634226,
+      "learning_rate": 1.559892769960933e-05,
+      "loss": 0.7668,
+      "step": 14072
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5568528138619571,
+      "learning_rate": 1.5598295454135985e-05,
+      "loss": 0.5224,
+      "step": 14073
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5819912519463398,
+      "learning_rate": 1.5597663176067596e-05,
+      "loss": 0.5231,
+      "step": 14074
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5957542983484502,
+      "learning_rate": 1.5597030865407828e-05,
+      "loss": 0.5756,
+      "step": 14075
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4442094608453195,
+      "learning_rate": 1.5596398522160385e-05,
+      "loss": 0.8233,
+      "step": 14076
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4076442390061699,
+      "learning_rate": 1.5595766146328934e-05,
+      "loss": 0.8565,
+      "step": 14077
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6895989968328218,
+      "learning_rate": 1.5595133737917163e-05,
+      "loss": 0.8304,
+      "step": 14078
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9526145934222423,
+      "learning_rate": 1.559450129692875e-05,
+      "loss": 0.7236,
+      "step": 14079
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4382881592473495,
+      "learning_rate": 1.5593868823367378e-05,
+      "loss": 0.845,
+      "step": 14080
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3318663065690168,
+      "learning_rate": 1.5593236317236733e-05,
+      "loss": 0.6121,
+      "step": 14081
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.559297063949878,
+      "learning_rate": 1.5592603778540493e-05,
+      "loss": 0.7179,
+      "step": 14082
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.8594022442983995,
+      "learning_rate": 1.5591971207282345e-05,
+      "loss": 0.7107,
+      "step": 14083
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.523174124761123,
+      "learning_rate": 1.5591338603465972e-05,
+      "loss": 0.8641,
+      "step": 14084
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8014537939390731,
+      "learning_rate": 1.5590705967095053e-05,
+      "loss": 0.8045,
+      "step": 14085
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5919623037851656,
+      "learning_rate": 1.5590073298173275e-05,
+      "loss": 0.6673,
+      "step": 14086
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3121968086285691,
+      "learning_rate": 1.558944059670432e-05,
+      "loss": 0.6441,
+      "step": 14087
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3190347261732438,
+      "learning_rate": 1.5588807862691874e-05,
+      "loss": 0.7088,
+      "step": 14088
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6492368699355286,
+      "learning_rate": 1.5588175096139614e-05,
+      "loss": 0.6484,
+      "step": 14089
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4516254529504304,
+      "learning_rate": 1.5587542297051233e-05,
+      "loss": 0.821,
+      "step": 14090
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5520469976525735,
+      "learning_rate": 1.5586909465430414e-05,
+      "loss": 0.7103,
+      "step": 14091
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2189503707386928,
+      "learning_rate": 1.5586276601280833e-05,
+      "loss": 0.5987,
+      "step": 14092
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.627688177362545,
+      "learning_rate": 1.558564370460619e-05,
+      "loss": 0.7932,
+      "step": 14093
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4000342400270307,
+      "learning_rate": 1.5585010775410153e-05,
+      "loss": 0.7781,
+      "step": 14094
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3536630014913114,
+      "learning_rate": 1.5584377813696417e-05,
+      "loss": 0.7181,
+      "step": 14095
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1768993831565457,
+      "learning_rate": 1.5583744819468667e-05,
+      "loss": 0.6702,
+      "step": 14096
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4222592264462255,
+      "learning_rate": 1.5583111792730587e-05,
+      "loss": 0.854,
+      "step": 14097
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2438710176993393,
+      "learning_rate": 1.558247873348586e-05,
+      "loss": 0.7037,
+      "step": 14098
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2837506792748938,
+      "learning_rate": 1.558184564173818e-05,
+      "loss": 0.766,
+      "step": 14099
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3555330034006767,
+      "learning_rate": 1.5581212517491224e-05,
+      "loss": 0.7952,
+      "step": 14100
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8342545953237899,
+      "learning_rate": 1.5580579360748676e-05,
+      "loss": 0.7592,
+      "step": 14101
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4011592355458469,
+      "learning_rate": 1.5579946171514237e-05,
+      "loss": 0.7107,
+      "step": 14102
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5464249958352325,
+      "learning_rate": 1.557931294979158e-05,
+      "loss": 0.6613,
+      "step": 14103
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5986584253465613,
+      "learning_rate": 1.5578679695584398e-05,
+      "loss": 0.5229,
+      "step": 14104
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6442423787595974,
+      "learning_rate": 1.5578046408896378e-05,
+      "loss": 0.7078,
+      "step": 14105
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.450070632255361,
+      "learning_rate": 1.5577413089731204e-05,
+      "loss": 0.7035,
+      "step": 14106
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9619715312240613,
+      "learning_rate": 1.5576779738092566e-05,
+      "loss": 0.7936,
+      "step": 14107
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4952050222819142,
+      "learning_rate": 1.557614635398415e-05,
+      "loss": 0.8863,
+      "step": 14108
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5509761238124546,
+      "learning_rate": 1.5575512937409648e-05,
+      "loss": 0.698,
+      "step": 14109
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5060396306900843,
+      "learning_rate": 1.5574879488372738e-05,
+      "loss": 0.722,
+      "step": 14110
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9094898912584009,
+      "learning_rate": 1.5574246006877118e-05,
+      "loss": 0.799,
+      "step": 14111
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.322387873625728,
+      "learning_rate": 1.5573612492926472e-05,
+      "loss": 0.7437,
+      "step": 14112
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.370715696960744,
+      "learning_rate": 1.5572978946524493e-05,
+      "loss": 0.721,
+      "step": 14113
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3218129828176841,
+      "learning_rate": 1.5572345367674864e-05,
+      "loss": 0.7371,
+      "step": 14114
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2636959500263618,
+      "learning_rate": 1.5571711756381274e-05,
+      "loss": 0.8076,
+      "step": 14115
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.393804629725648,
+      "learning_rate": 1.557107811264742e-05,
+      "loss": 0.7275,
+      "step": 14116
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5147374085386751,
+      "learning_rate": 1.557044443647698e-05,
+      "loss": 0.6309,
+      "step": 14117
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6134540144096698,
+      "learning_rate": 1.5569810727873652e-05,
+      "loss": 0.7728,
+      "step": 14118
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5976771172640514,
+      "learning_rate": 1.556917698684112e-05,
+      "loss": 0.7595,
+      "step": 14119
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.669985091247827,
+      "learning_rate": 1.5568543213383076e-05,
+      "loss": 0.7311,
+      "step": 14120
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8506337805921462,
+      "learning_rate": 1.5567909407503213e-05,
+      "loss": 0.8823,
+      "step": 14121
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6346285949835283,
+      "learning_rate": 1.5567275569205216e-05,
+      "loss": 0.5358,
+      "step": 14122
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.853317582321337,
+      "learning_rate": 1.556664169849278e-05,
+      "loss": 0.7307,
+      "step": 14123
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6513873213161319,
+      "learning_rate": 1.5566007795369594e-05,
+      "loss": 0.5738,
+      "step": 14124
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2826544916403906,
+      "learning_rate": 1.5565373859839348e-05,
+      "loss": 0.7082,
+      "step": 14125
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3964489265841193,
+      "learning_rate": 1.5564739891905734e-05,
+      "loss": 0.6836,
+      "step": 14126
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4794837334815454,
+      "learning_rate": 1.556410589157244e-05,
+      "loss": 0.7275,
+      "step": 14127
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5610018953275898,
+      "learning_rate": 1.5563471858843163e-05,
+      "loss": 0.5022,
+      "step": 14128
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3475789027632967,
+      "learning_rate": 1.556283779372159e-05,
+      "loss": 0.7273,
+      "step": 14129
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.58634100804473,
+      "learning_rate": 1.5562203696211416e-05,
+      "loss": 0.7893,
+      "step": 14130
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.574417811634108,
+      "learning_rate": 1.5561569566316327e-05,
+      "loss": 0.5098,
+      "step": 14131
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9157674532600955,
+      "learning_rate": 1.5560935404040023e-05,
+      "loss": 0.7957,
+      "step": 14132
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.42483716103954,
+      "learning_rate": 1.556030120938619e-05,
+      "loss": 0.7575,
+      "step": 14133
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4650546940374516,
+      "learning_rate": 1.5559666982358525e-05,
+      "loss": 0.7418,
+      "step": 14134
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5935393861829554,
+      "learning_rate": 1.555903272296072e-05,
+      "loss": 0.733,
+      "step": 14135
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7886114557499266,
+      "learning_rate": 1.5558398431196465e-05,
+      "loss": 0.7031,
+      "step": 14136
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4562224209099353,
+      "learning_rate": 1.5557764107069454e-05,
+      "loss": 0.734,
+      "step": 14137
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7904741460174476,
+      "learning_rate": 1.555712975058338e-05,
+      "loss": 0.7987,
+      "step": 14138
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6682560847223058,
+      "learning_rate": 1.5556495361741936e-05,
+      "loss": 0.7956,
+      "step": 14139
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4390923251570382,
+      "learning_rate": 1.555586094054882e-05,
+      "loss": 0.7166,
+      "step": 14140
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0272841048275376,
+      "learning_rate": 1.5555226487007723e-05,
+      "loss": 0.8143,
+      "step": 14141
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5797849998718994,
+      "learning_rate": 1.5554592001122335e-05,
+      "loss": 0.7337,
+      "step": 14142
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.381503475588396,
+      "learning_rate": 1.5553957482896356e-05,
+      "loss": 0.8662,
+      "step": 14143
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.440494931857834,
+      "learning_rate": 1.5553322932333477e-05,
+      "loss": 0.7565,
+      "step": 14144
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6474180028528118,
+      "learning_rate": 1.5552688349437396e-05,
+      "loss": 0.7308,
+      "step": 14145
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6038277329518165,
+      "learning_rate": 1.5552053734211805e-05,
+      "loss": 0.7363,
+      "step": 14146
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8493640111692558,
+      "learning_rate": 1.55514190866604e-05,
+      "loss": 0.7628,
+      "step": 14147
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5848543812334874,
+      "learning_rate": 1.5550784406786875e-05,
+      "loss": 0.5264,
+      "step": 14148
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5838345397691267,
+      "learning_rate": 1.5550149694594924e-05,
+      "loss": 0.7084,
+      "step": 14149
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.70167874060838,
+      "learning_rate": 1.5549514950088243e-05,
+      "loss": 0.747,
+      "step": 14150
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7360266151593375,
+      "learning_rate": 1.554888017327053e-05,
+      "loss": 0.8083,
+      "step": 14151
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2374441974783006,
+      "learning_rate": 1.5548245364145486e-05,
+      "loss": 0.6816,
+      "step": 14152
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5152531068800288,
+      "learning_rate": 1.5547610522716796e-05,
+      "loss": 0.5959,
+      "step": 14153
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0040290041047686,
+      "learning_rate": 1.5546975648988157e-05,
+      "loss": 0.7152,
+      "step": 14154
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7870591954250359,
+      "learning_rate": 1.5546340742963278e-05,
+      "loss": 0.7604,
+      "step": 14155
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4954996656928548,
+      "learning_rate": 1.554570580464584e-05,
+      "loss": 0.863,
+      "step": 14156
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4911551675000525,
+      "learning_rate": 1.5545070834039556e-05,
+      "loss": 0.6852,
+      "step": 14157
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.352630214855919,
+      "learning_rate": 1.5544435831148107e-05,
+      "loss": 0.8133,
+      "step": 14158
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3333543470367637,
+      "learning_rate": 1.55438007959752e-05,
+      "loss": 0.6584,
+      "step": 14159
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.261993939500595,
+      "learning_rate": 1.554316572852453e-05,
+      "loss": 0.6185,
+      "step": 14160
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5916388185585528,
+      "learning_rate": 1.554253062879979e-05,
+      "loss": 0.5591,
+      "step": 14161
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2347556771386985,
+      "learning_rate": 1.5541895496804686e-05,
+      "loss": 0.6995,
+      "step": 14162
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3792331683144237,
+      "learning_rate": 1.5541260332542906e-05,
+      "loss": 0.8192,
+      "step": 14163
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8726945512677051,
+      "learning_rate": 1.554062513601816e-05,
+      "loss": 0.7694,
+      "step": 14164
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5590577872595848,
+      "learning_rate": 1.553998990723414e-05,
+      "loss": 0.6736,
+      "step": 14165
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.463080465870108,
+      "learning_rate": 1.5539354646194544e-05,
+      "loss": 0.7094,
+      "step": 14166
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6384475632331639,
+      "learning_rate": 1.553871935290307e-05,
+      "loss": 0.6204,
+      "step": 14167
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5129791061318234,
+      "learning_rate": 1.553808402736342e-05,
+      "loss": 0.7602,
+      "step": 14168
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.468442945527901,
+      "learning_rate": 1.553744866957929e-05,
+      "loss": 0.6622,
+      "step": 14169
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3803273729663323,
+      "learning_rate": 1.553681327955438e-05,
+      "loss": 0.8094,
+      "step": 14170
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.581828391405181,
+      "learning_rate": 1.5536177857292392e-05,
+      "loss": 0.6905,
+      "step": 14171
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4481028429419325,
+      "learning_rate": 1.5535542402797027e-05,
+      "loss": 0.7245,
+      "step": 14172
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5435644503177544,
+      "learning_rate": 1.5534906916071976e-05,
+      "loss": 0.917,
+      "step": 14173
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5791767765664936,
+      "learning_rate": 1.5534271397120947e-05,
+      "loss": 0.5311,
+      "step": 14174
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2947817651067952,
+      "learning_rate": 1.553363584594764e-05,
+      "loss": 0.7526,
+      "step": 14175
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.456185189559372,
+      "learning_rate": 1.553300026255575e-05,
+      "loss": 0.7329,
+      "step": 14176
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3394724664059294,
+      "learning_rate": 1.553236464694898e-05,
+      "loss": 0.7983,
+      "step": 14177
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.350780819207275,
+      "learning_rate": 1.5531728999131034e-05,
+      "loss": 0.7905,
+      "step": 14178
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.593219285237734,
+      "learning_rate": 1.553109331910561e-05,
+      "loss": 0.5143,
+      "step": 14179
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4884976957442024,
+      "learning_rate": 1.553045760687641e-05,
+      "loss": 0.6939,
+      "step": 14180
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6081575221213933,
+      "learning_rate": 1.5529821862447133e-05,
+      "loss": 0.5745,
+      "step": 14181
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.527298492954542,
+      "learning_rate": 1.5529186085821485e-05,
+      "loss": 0.7152,
+      "step": 14182
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.610682876059462,
+      "learning_rate": 1.552855027700316e-05,
+      "loss": 0.563,
+      "step": 14183
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3560384624541726,
+      "learning_rate": 1.552791443599587e-05,
+      "loss": 0.8115,
+      "step": 14184
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7479881250348117,
+      "learning_rate": 1.5527278562803313e-05,
+      "loss": 0.8378,
+      "step": 14185
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5101925720852358,
+      "learning_rate": 1.552664265742919e-05,
+      "loss": 0.7677,
+      "step": 14186
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5466222547361017,
+      "learning_rate": 1.5526006719877198e-05,
+      "loss": 0.8397,
+      "step": 14187
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5811206794816428,
+      "learning_rate": 1.552537075015105e-05,
+      "loss": 0.6515,
+      "step": 14188
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3311731323584608,
+      "learning_rate": 1.5524734748254442e-05,
+      "loss": 0.7386,
+      "step": 14189
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.357252223436673,
+      "learning_rate": 1.552409871419108e-05,
+      "loss": 0.6791,
+      "step": 14190
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6063273302579413,
+      "learning_rate": 1.5523462647964667e-05,
+      "loss": 0.5219,
+      "step": 14191
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5973780929711434,
+      "learning_rate": 1.5522826549578908e-05,
+      "loss": 0.5015,
+      "step": 14192
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6087040747105552,
+      "learning_rate": 1.55221904190375e-05,
+      "loss": 0.5128,
+      "step": 14193
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.543636746789529,
+      "learning_rate": 1.5521554256344152e-05,
+      "loss": 0.6776,
+      "step": 14194
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8415525808157918,
+      "learning_rate": 1.552091806150257e-05,
+      "loss": 0.7492,
+      "step": 14195
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.051213695480827,
+      "learning_rate": 1.5520281834516452e-05,
+      "loss": 0.8428,
+      "step": 14196
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8868345835053835,
+      "learning_rate": 1.5519645575389508e-05,
+      "loss": 0.8458,
+      "step": 14197
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.434236815499778,
+      "learning_rate": 1.5519009284125438e-05,
+      "loss": 0.6986,
+      "step": 14198
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5842385259687107,
+      "learning_rate": 1.5518372960727947e-05,
+      "loss": 0.5609,
+      "step": 14199
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4418599084480035,
+      "learning_rate": 1.5517736605200745e-05,
+      "loss": 0.7477,
+      "step": 14200
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4736084416024051,
+      "learning_rate": 1.5517100217547532e-05,
+      "loss": 0.6533,
+      "step": 14201
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4697790015725565,
+      "learning_rate": 1.5516463797772017e-05,
+      "loss": 0.6697,
+      "step": 14202
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6022162665343214,
+      "learning_rate": 1.55158273458779e-05,
+      "loss": 0.863,
+      "step": 14203
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3303845806411436,
+      "learning_rate": 1.551519086186889e-05,
+      "loss": 0.7525,
+      "step": 14204
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4008936851946148,
+      "learning_rate": 1.5514554345748694e-05,
+      "loss": 0.7579,
+      "step": 14205
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4105799988021632,
+      "learning_rate": 1.551391779752102e-05,
+      "loss": 0.8269,
+      "step": 14206
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.468127463685003,
+      "learning_rate": 1.5513281217189565e-05,
+      "loss": 0.7865,
+      "step": 14207
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7105492257673265,
+      "learning_rate": 1.5512644604758046e-05,
+      "loss": 0.7815,
+      "step": 14208
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3129105056355854,
+      "learning_rate": 1.551200796023016e-05,
+      "loss": 0.7237,
+      "step": 14209
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3392265368066687,
+      "learning_rate": 1.5511371283609622e-05,
+      "loss": 0.7415,
+      "step": 14210
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.404459213405265,
+      "learning_rate": 1.5510734574900137e-05,
+      "loss": 0.7976,
+      "step": 14211
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5487738000790157,
+      "learning_rate": 1.5510097834105406e-05,
+      "loss": 0.7745,
+      "step": 14212
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8097004816443385,
+      "learning_rate": 1.5509461061229147e-05,
+      "loss": 0.8382,
+      "step": 14213
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.234625555714181,
+      "learning_rate": 1.5508824256275055e-05,
+      "loss": 0.8546,
+      "step": 14214
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6322137538320298,
+      "learning_rate": 1.550818741924685e-05,
+      "loss": 0.7628,
+      "step": 14215
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7940886729688876,
+      "learning_rate": 1.550755055014823e-05,
+      "loss": 0.7384,
+      "step": 14216
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.369885811377006,
+      "learning_rate": 1.550691364898291e-05,
+      "loss": 0.8209,
+      "step": 14217
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3749804182001903,
+      "learning_rate": 1.5506276715754592e-05,
+      "loss": 0.7501,
+      "step": 14218
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3882174578636854,
+      "learning_rate": 1.5505639750466986e-05,
+      "loss": 0.8849,
+      "step": 14219
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3238061821714888,
+      "learning_rate": 1.5505002753123807e-05,
+      "loss": 0.8814,
+      "step": 14220
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8217510691908951,
+      "learning_rate": 1.550436572372876e-05,
+      "loss": 0.8712,
+      "step": 14221
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4028956299041202,
+      "learning_rate": 1.550372866228555e-05,
+      "loss": 0.5851,
+      "step": 14222
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4351459157308961,
+      "learning_rate": 1.550309156879789e-05,
+      "loss": 0.8055,
+      "step": 14223
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3460384354906614,
+      "learning_rate": 1.5502454443269485e-05,
+      "loss": 0.7821,
+      "step": 14224
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6054191469775871,
+      "learning_rate": 1.5501817285704055e-05,
+      "loss": 0.6003,
+      "step": 14225
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6034206416862844,
+      "learning_rate": 1.55011800961053e-05,
+      "loss": 0.8099,
+      "step": 14226
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3741710827589952,
+      "learning_rate": 1.550054287447693e-05,
+      "loss": 0.7213,
+      "step": 14227
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4253833537142189,
+      "learning_rate": 1.5499905620822662e-05,
+      "loss": 0.6653,
+      "step": 14228
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4315747406323822,
+      "learning_rate": 1.54992683351462e-05,
+      "loss": 0.6848,
+      "step": 14229
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5526142898081703,
+      "learning_rate": 1.5498631017451256e-05,
+      "loss": 0.6968,
+      "step": 14230
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.392417176313697,
+      "learning_rate": 1.5497993667741545e-05,
+      "loss": 0.6655,
+      "step": 14231
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.392980886158587,
+      "learning_rate": 1.549735628602077e-05,
+      "loss": 0.8065,
+      "step": 14232
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3227892086265063,
+      "learning_rate": 1.5496718872292648e-05,
+      "loss": 0.6883,
+      "step": 14233
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.315621212109912,
+      "learning_rate": 1.549608142656089e-05,
+      "loss": 0.8367,
+      "step": 14234
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2494732774136141,
+      "learning_rate": 1.5495443948829203e-05,
+      "loss": 0.7313,
+      "step": 14235
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4877951201103163,
+      "learning_rate": 1.5494806439101306e-05,
+      "loss": 0.7476,
+      "step": 14236
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4947257117873445,
+      "learning_rate": 1.5494168897380905e-05,
+      "loss": 0.7692,
+      "step": 14237
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6402906973182325,
+      "learning_rate": 1.5493531323671707e-05,
+      "loss": 0.6579,
+      "step": 14238
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.519685703871433,
+      "learning_rate": 1.5492893717977438e-05,
+      "loss": 0.7252,
+      "step": 14239
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5837136272334729,
+      "learning_rate": 1.54922560803018e-05,
+      "loss": 0.7052,
+      "step": 14240
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4285169252946186,
+      "learning_rate": 1.5491618410648507e-05,
+      "loss": 0.6898,
+      "step": 14241
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5773685270015971,
+      "learning_rate": 1.549098070902127e-05,
+      "loss": 0.833,
+      "step": 14242
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3079681443990638,
+      "learning_rate": 1.5490342975423814e-05,
+      "loss": 0.7753,
+      "step": 14243
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.506648268063969,
+      "learning_rate": 1.5489705209859838e-05,
+      "loss": 0.8083,
+      "step": 14244
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3839059130614029,
+      "learning_rate": 1.548906741233306e-05,
+      "loss": 0.7841,
+      "step": 14245
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3369173237495757,
+      "learning_rate": 1.5488429582847194e-05,
+      "loss": 0.7033,
+      "step": 14246
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3511093922381179,
+      "learning_rate": 1.5487791721405953e-05,
+      "loss": 0.6661,
+      "step": 14247
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8350120450431218,
+      "learning_rate": 1.548715382801305e-05,
+      "loss": 0.8307,
+      "step": 14248
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.927684705649594,
+      "learning_rate": 1.54865159026722e-05,
+      "loss": 0.8038,
+      "step": 14249
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5391303541186188,
+      "learning_rate": 1.548587794538712e-05,
+      "loss": 0.7301,
+      "step": 14250
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.717213939941973,
+      "learning_rate": 1.5485239956161523e-05,
+      "loss": 0.7548,
+      "step": 14251
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6070700887664975,
+      "learning_rate": 1.548460193499912e-05,
+      "loss": 0.5655,
+      "step": 14252
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4539039639496965,
+      "learning_rate": 1.548396388190363e-05,
+      "loss": 0.8031,
+      "step": 14253
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3783399974513253,
+      "learning_rate": 1.5483325796878764e-05,
+      "loss": 0.7181,
+      "step": 14254
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5236659696801047,
+      "learning_rate": 1.548268767992824e-05,
+      "loss": 0.7398,
+      "step": 14255
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4248261152801625,
+      "learning_rate": 1.5482049531055772e-05,
+      "loss": 0.6652,
+      "step": 14256
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.375875402655836,
+      "learning_rate": 1.5481411350265075e-05,
+      "loss": 0.8088,
+      "step": 14257
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4277827107560188,
+      "learning_rate": 1.5480773137559867e-05,
+      "loss": 0.8114,
+      "step": 14258
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5274774287743087,
+      "learning_rate": 1.5480134892943865e-05,
+      "loss": 0.7383,
+      "step": 14259
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6080140040689656,
+      "learning_rate": 1.5479496616420782e-05,
+      "loss": 0.5408,
+      "step": 14260
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6724008383849613,
+      "learning_rate": 1.5478858307994333e-05,
+      "loss": 0.6671,
+      "step": 14261
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5419062685780989,
+      "learning_rate": 1.5478219967668236e-05,
+      "loss": 0.6749,
+      "step": 14262
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4125141696796462,
+      "learning_rate": 1.547758159544621e-05,
+      "loss": 0.7692,
+      "step": 14263
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4128876001648576,
+      "learning_rate": 1.5476943191331965e-05,
+      "loss": 0.7281,
+      "step": 14264
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3232090155158007,
+      "learning_rate": 1.5476304755329227e-05,
+      "loss": 0.8131,
+      "step": 14265
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8069833743935981,
+      "learning_rate": 1.547566628744171e-05,
+      "loss": 0.7277,
+      "step": 14266
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.433611540843234,
+      "learning_rate": 1.5475027787673126e-05,
+      "loss": 0.8277,
+      "step": 14267
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.054084600935078,
+      "learning_rate": 1.54743892560272e-05,
+      "loss": 0.9252,
+      "step": 14268
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3700354965591228,
+      "learning_rate": 1.5473750692507647e-05,
+      "loss": 0.7595,
+      "step": 14269
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5756247694787879,
+      "learning_rate": 1.547311209711818e-05,
+      "loss": 0.5635,
+      "step": 14270
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6383539062884902,
+      "learning_rate": 1.547247346986253e-05,
+      "loss": 0.6598,
+      "step": 14271
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4629383313341602,
+      "learning_rate": 1.54718348107444e-05,
+      "loss": 0.7339,
+      "step": 14272
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5360622021842805,
+      "learning_rate": 1.5471196119767515e-05,
+      "loss": 0.8693,
+      "step": 14273
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.892607327383021,
+      "learning_rate": 1.5470557396935597e-05,
+      "loss": 0.6685,
+      "step": 14274
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3690243432668368,
+      "learning_rate": 1.5469918642252363e-05,
+      "loss": 0.7755,
+      "step": 14275
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3844698172198808,
+      "learning_rate": 1.5469279855721527e-05,
+      "loss": 0.8087,
+      "step": 14276
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6401273565707972,
+      "learning_rate": 1.5468641037346816e-05,
+      "loss": 0.7728,
+      "step": 14277
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4286551345865957,
+      "learning_rate": 1.5468002187131944e-05,
+      "loss": 0.7169,
+      "step": 14278
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4344865554258515,
+      "learning_rate": 1.546736330508063e-05,
+      "loss": 0.7509,
+      "step": 14279
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5709558461830977,
+      "learning_rate": 1.5466724391196598e-05,
+      "loss": 0.5804,
+      "step": 14280
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6035178581942726,
+      "learning_rate": 1.5466085445483568e-05,
+      "loss": 0.7994,
+      "step": 14281
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3600429375573133,
+      "learning_rate": 1.5465446467945254e-05,
+      "loss": 0.8401,
+      "step": 14282
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.612779592609318,
+      "learning_rate": 1.5464807458585383e-05,
+      "loss": 0.5445,
+      "step": 14283
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8239050446650338,
+      "learning_rate": 1.546416841740767e-05,
+      "loss": 0.8172,
+      "step": 14284
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3789363728643311,
+      "learning_rate": 1.5463529344415842e-05,
+      "loss": 0.7004,
+      "step": 14285
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4544325044088064,
+      "learning_rate": 1.5462890239613613e-05,
+      "loss": 0.5543,
+      "step": 14286
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5833803760557247,
+      "learning_rate": 1.546225110300471e-05,
+      "loss": 0.5411,
+      "step": 14287
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5091163977200437,
+      "learning_rate": 1.5461611934592852e-05,
+      "loss": 0.915,
+      "step": 14288
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8228057536499238,
+      "learning_rate": 1.546097273438176e-05,
+      "loss": 0.7636,
+      "step": 14289
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8518796522587369,
+      "learning_rate": 1.5460333502375155e-05,
+      "loss": 0.9025,
+      "step": 14290
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5807651432247944,
+      "learning_rate": 1.5459694238576757e-05,
+      "loss": 0.8723,
+      "step": 14291
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5707738574289708,
+      "learning_rate": 1.5459054942990298e-05,
+      "loss": 0.5292,
+      "step": 14292
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.344812966566552,
+      "learning_rate": 1.5458415615619486e-05,
+      "loss": 0.7632,
+      "step": 14293
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.634093199526444,
+      "learning_rate": 1.545777625646805e-05,
+      "loss": 0.7679,
+      "step": 14294
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5517833684280682,
+      "learning_rate": 1.5457136865539716e-05,
+      "loss": 0.7729,
+      "step": 14295
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.370042155114391,
+      "learning_rate": 1.5456497442838202e-05,
+      "loss": 0.7317,
+      "step": 14296
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2358276863326287,
+      "learning_rate": 1.5455857988367236e-05,
+      "loss": 0.7773,
+      "step": 14297
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.697687013881028,
+      "learning_rate": 1.5455218502130533e-05,
+      "loss": 0.7882,
+      "step": 14298
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.615900216483074,
+      "learning_rate": 1.5454578984131822e-05,
+      "loss": 0.5515,
+      "step": 14299
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.435389642433202,
+      "learning_rate": 1.5453939434374824e-05,
+      "loss": 0.7226,
+      "step": 14300
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4014279234933904,
+      "learning_rate": 1.5453299852863265e-05,
+      "loss": 0.7426,
+      "step": 14301
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9338416823450626,
+      "learning_rate": 1.5452660239600873e-05,
+      "loss": 0.7088,
+      "step": 14302
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4113526001850725,
+      "learning_rate": 1.545202059459136e-05,
+      "loss": 0.8195,
+      "step": 14303
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4855317041267297,
+      "learning_rate": 1.5451380917838462e-05,
+      "loss": 0.7381,
+      "step": 14304
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.567101278284583,
+      "learning_rate": 1.5450741209345894e-05,
+      "loss": 0.7646,
+      "step": 14305
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3373418403048334,
+      "learning_rate": 1.5450101469117386e-05,
+      "loss": 0.7474,
+      "step": 14306
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5812637947273505,
+      "learning_rate": 1.544946169715666e-05,
+      "loss": 0.7624,
+      "step": 14307
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4234345329045979,
+      "learning_rate": 1.5448821893467444e-05,
+      "loss": 0.7411,
+      "step": 14308
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6051365026890336,
+      "learning_rate": 1.5448182058053466e-05,
+      "loss": 0.7494,
+      "step": 14309
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.382919473508427,
+      "learning_rate": 1.5447542190918444e-05,
+      "loss": 0.6973,
+      "step": 14310
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6848674699875472,
+      "learning_rate": 1.5446902292066106e-05,
+      "loss": 0.6709,
+      "step": 14311
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5256646757895513,
+      "learning_rate": 1.5446262361500178e-05,
+      "loss": 0.8167,
+      "step": 14312
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5437603853191808,
+      "learning_rate": 1.544562239922439e-05,
+      "loss": 0.7164,
+      "step": 14313
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4248127511706012,
+      "learning_rate": 1.544498240524246e-05,
+      "loss": 0.7325,
+      "step": 14314
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.632130317808135,
+      "learning_rate": 1.544434237955812e-05,
+      "loss": 0.7194,
+      "step": 14315
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.367185005926514,
+      "learning_rate": 1.5443702322175095e-05,
+      "loss": 0.735,
+      "step": 14316
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4280419077166762,
+      "learning_rate": 1.544306223309711e-05,
+      "loss": 0.8439,
+      "step": 14317
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3551589468845335,
+      "learning_rate": 1.54424221123279e-05,
+      "loss": 0.8336,
+      "step": 14318
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4168435812818656,
+      "learning_rate": 1.544178195987118e-05,
+      "loss": 0.792,
+      "step": 14319
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3536120506132687,
+      "learning_rate": 1.5441141775730685e-05,
+      "loss": 0.8288,
+      "step": 14320
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5329186415713887,
+      "learning_rate": 1.5440501559910135e-05,
+      "loss": 0.7172,
+      "step": 14321
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7848062960503495,
+      "learning_rate": 1.5439861312413267e-05,
+      "loss": 0.7335,
+      "step": 14322
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1400800875230126,
+      "learning_rate": 1.5439221033243804e-05,
+      "loss": 0.7816,
+      "step": 14323
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7526492758736172,
+      "learning_rate": 1.5438580722405473e-05,
+      "loss": 0.6529,
+      "step": 14324
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8283879874783382,
+      "learning_rate": 1.5437940379902e-05,
+      "loss": 0.7471,
+      "step": 14325
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5279271940451193,
+      "learning_rate": 1.5437300005737122e-05,
+      "loss": 0.8584,
+      "step": 14326
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3910237438130981,
+      "learning_rate": 1.543665959991456e-05,
+      "loss": 0.8323,
+      "step": 14327
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9596500564244659,
+      "learning_rate": 1.5436019162438044e-05,
+      "loss": 0.8873,
+      "step": 14328
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3349545670149268,
+      "learning_rate": 1.5435378693311305e-05,
+      "loss": 0.764,
+      "step": 14329
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6554703166643007,
+      "learning_rate": 1.5434738192538067e-05,
+      "loss": 0.5584,
+      "step": 14330
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.42147750079927,
+      "learning_rate": 1.5434097660122066e-05,
+      "loss": 0.6506,
+      "step": 14331
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8286266851475048,
+      "learning_rate": 1.5433457096067027e-05,
+      "loss": 0.8492,
+      "step": 14332
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4921000307603167,
+      "learning_rate": 1.543281650037668e-05,
+      "loss": 0.7189,
+      "step": 14333
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8708233169576216,
+      "learning_rate": 1.5432175873054758e-05,
+      "loss": 0.7541,
+      "step": 14334
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.512749486322409,
+      "learning_rate": 1.543153521410498e-05,
+      "loss": 0.8257,
+      "step": 14335
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3861495747721957,
+      "learning_rate": 1.5430894523531094e-05,
+      "loss": 0.6937,
+      "step": 14336
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.212380568420674,
+      "learning_rate": 1.5430253801336816e-05,
+      "loss": 0.8211,
+      "step": 14337
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4824643113192557,
+      "learning_rate": 1.5429613047525885e-05,
+      "loss": 0.7448,
+      "step": 14338
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6273246890481776,
+      "learning_rate": 1.5428972262102027e-05,
+      "loss": 0.8726,
+      "step": 14339
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4884873826714893,
+      "learning_rate": 1.542833144506897e-05,
+      "loss": 0.6957,
+      "step": 14340
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.777047977868821,
+      "learning_rate": 1.542769059643045e-05,
+      "loss": 0.707,
+      "step": 14341
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.609079582339739,
+      "learning_rate": 1.5427049716190196e-05,
+      "loss": 0.6777,
+      "step": 14342
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4162092145178182,
+      "learning_rate": 1.5426408804351943e-05,
+      "loss": 0.8132,
+      "step": 14343
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4152998572145845,
+      "learning_rate": 1.542576786091942e-05,
+      "loss": 0.7791,
+      "step": 14344
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3687344088475082,
+      "learning_rate": 1.5425126885896357e-05,
+      "loss": 0.8228,
+      "step": 14345
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4722237094247668,
+      "learning_rate": 1.5424485879286488e-05,
+      "loss": 0.7383,
+      "step": 14346
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3600561401788853,
+      "learning_rate": 1.5423844841093545e-05,
+      "loss": 0.7162,
+      "step": 14347
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.481927904932346,
+      "learning_rate": 1.5423203771321263e-05,
+      "loss": 0.6399,
+      "step": 14348
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3299771854083997,
+      "learning_rate": 1.5422562669973367e-05,
+      "loss": 0.7941,
+      "step": 14349
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5948791468558705,
+      "learning_rate": 1.5421921537053596e-05,
+      "loss": 0.5733,
+      "step": 14350
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.367172997037991,
+      "learning_rate": 1.5421280372565683e-05,
+      "loss": 0.648,
+      "step": 14351
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5788841122351325,
+      "learning_rate": 1.5420639176513358e-05,
+      "loss": 0.7348,
+      "step": 14352
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5005535055462416,
+      "learning_rate": 1.5419997948900358e-05,
+      "loss": 0.9097,
+      "step": 14353
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5544870954816101,
+      "learning_rate": 1.5419356689730408e-05,
+      "loss": 0.7842,
+      "step": 14354
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4578933463888073,
+      "learning_rate": 1.5418715399007253e-05,
+      "loss": 0.659,
+      "step": 14355
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.569941557258749,
+      "learning_rate": 1.541807407673462e-05,
+      "loss": 0.5214,
+      "step": 14356
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.93165896352644,
+      "learning_rate": 1.5417432722916244e-05,
+      "loss": 0.8107,
+      "step": 14357
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4379272960565035,
+      "learning_rate": 1.541679133755586e-05,
+      "loss": 0.7255,
+      "step": 14358
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5419805464311194,
+      "learning_rate": 1.54161499206572e-05,
+      "loss": 0.6683,
+      "step": 14359
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.340117824228686,
+      "learning_rate": 1.5415508472224002e-05,
+      "loss": 0.6975,
+      "step": 14360
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.686569046326418,
+      "learning_rate": 1.541486699226e-05,
+      "loss": 0.96,
+      "step": 14361
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4096411144545693,
+      "learning_rate": 1.541422548076893e-05,
+      "loss": 0.7715,
+      "step": 14362
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5415206856716865,
+      "learning_rate": 1.5413583937754517e-05,
+      "loss": 0.4842,
+      "step": 14363
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3703388552162712,
+      "learning_rate": 1.5412942363220512e-05,
+      "loss": 0.7543,
+      "step": 14364
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6871053458795635,
+      "learning_rate": 1.541230075717064e-05,
+      "loss": 0.7092,
+      "step": 14365
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.047713853773502,
+      "learning_rate": 1.5411659119608643e-05,
+      "loss": 0.6569,
+      "step": 14366
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5082338145045506,
+      "learning_rate": 1.541101745053825e-05,
+      "loss": 0.791,
+      "step": 14367
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3134628235796828,
+      "learning_rate": 1.5410375749963197e-05,
+      "loss": 0.7889,
+      "step": 14368
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4666386358255854,
+      "learning_rate": 1.540973401788723e-05,
+      "loss": 0.7659,
+      "step": 14369
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3601551890287782,
+      "learning_rate": 1.5409092254314076e-05,
+      "loss": 0.7327,
+      "step": 14370
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3458330501491527,
+      "learning_rate": 1.5408450459247474e-05,
+      "loss": 0.7732,
+      "step": 14371
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4413299935619661,
+      "learning_rate": 1.5407808632691162e-05,
+      "loss": 0.8612,
+      "step": 14372
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3198269385977695,
+      "learning_rate": 1.5407166774648877e-05,
+      "loss": 0.6861,
+      "step": 14373
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7877657533645146,
+      "learning_rate": 1.5406524885124352e-05,
+      "loss": 0.7898,
+      "step": 14374
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6442045063072814,
+      "learning_rate": 1.540588296412133e-05,
+      "loss": 0.8567,
+      "step": 14375
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5593255719182526,
+      "learning_rate": 1.5405241011643544e-05,
+      "loss": 0.7944,
+      "step": 14376
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3727923798865966,
+      "learning_rate": 1.5404599027694734e-05,
+      "loss": 0.7236,
+      "step": 14377
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6853794118950156,
+      "learning_rate": 1.540395701227864e-05,
+      "loss": 0.6409,
+      "step": 14378
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4011099009201249,
+      "learning_rate": 1.5403314965398995e-05,
+      "loss": 0.6305,
+      "step": 14379
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6803375395079492,
+      "learning_rate": 1.540267288705954e-05,
+      "loss": 0.8428,
+      "step": 14380
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4209406993400533,
+      "learning_rate": 1.5402030777264014e-05,
+      "loss": 0.7608,
+      "step": 14381
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1159405639084032,
+      "learning_rate": 1.5401388636016154e-05,
+      "loss": 0.7703,
+      "step": 14382
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4810680991909906,
+      "learning_rate": 1.54007464633197e-05,
+      "loss": 0.8176,
+      "step": 14383
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6405388831952497,
+      "learning_rate": 1.5400104259178386e-05,
+      "loss": 0.5853,
+      "step": 14384
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.095903161528534,
+      "learning_rate": 1.539946202359596e-05,
+      "loss": 0.8047,
+      "step": 14385
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2844669943047173,
+      "learning_rate": 1.5398819756576155e-05,
+      "loss": 0.7923,
+      "step": 14386
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5780989852342863,
+      "learning_rate": 1.539817745812271e-05,
+      "loss": 0.5038,
+      "step": 14387
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6322261652580268,
+      "learning_rate": 1.539753512823937e-05,
+      "loss": 0.5167,
+      "step": 14388
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.595028918401784,
+      "learning_rate": 1.539689276692987e-05,
+      "loss": 0.7212,
+      "step": 14389
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2218668831285606,
+      "learning_rate": 1.5396250374197953e-05,
+      "loss": 0.5659,
+      "step": 14390
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6919518637541826,
+      "learning_rate": 1.5395607950047357e-05,
+      "loss": 0.8148,
+      "step": 14391
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.593754447488227,
+      "learning_rate": 1.5394965494481824e-05,
+      "loss": 0.8004,
+      "step": 14392
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6277270369553887,
+      "learning_rate": 1.5394323007505094e-05,
+      "loss": 0.5579,
+      "step": 14393
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.623910832823765,
+      "learning_rate": 1.5393680489120907e-05,
+      "loss": 0.7283,
+      "step": 14394
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5983296344043326,
+      "learning_rate": 1.5393037939333004e-05,
+      "loss": 0.5339,
+      "step": 14395
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5791976623617865,
+      "learning_rate": 1.5392395358145125e-05,
+      "loss": 0.803,
+      "step": 14396
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4262536795912915,
+      "learning_rate": 1.5391752745561022e-05,
+      "loss": 0.71,
+      "step": 14397
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4015601197085839,
+      "learning_rate": 1.539111010158442e-05,
+      "loss": 0.7311,
+      "step": 14398
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6349945304750912,
+      "learning_rate": 1.539046742621907e-05,
+      "loss": 0.7543,
+      "step": 14399
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4786094571246575,
+      "learning_rate": 1.538982471946871e-05,
+      "loss": 0.7788,
+      "step": 14400
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5918384295108978,
+      "learning_rate": 1.5389181981337085e-05,
+      "loss": 0.4885,
+      "step": 14401
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3722619833969678,
+      "learning_rate": 1.5388539211827934e-05,
+      "loss": 0.8103,
+      "step": 14402
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5653560443826426,
+      "learning_rate": 1.538789641094501e-05,
+      "loss": 0.5212,
+      "step": 14403
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4941182619674964,
+      "learning_rate": 1.538725357869204e-05,
+      "loss": 0.7833,
+      "step": 14404
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.496026871276713,
+      "learning_rate": 1.5386610715072773e-05,
+      "loss": 0.8308,
+      "step": 14405
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.975280560327141,
+      "learning_rate": 1.5385967820090958e-05,
+      "loss": 0.669,
+      "step": 14406
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4432720929479215,
+      "learning_rate": 1.538532489375033e-05,
+      "loss": 0.6381,
+      "step": 14407
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0292271808858664,
+      "learning_rate": 1.5384681936054636e-05,
+      "loss": 0.7439,
+      "step": 14408
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4266157061905227,
+      "learning_rate": 1.5384038947007617e-05,
+      "loss": 0.7975,
+      "step": 14409
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4058417831056826,
+      "learning_rate": 1.5383395926613026e-05,
+      "loss": 0.7141,
+      "step": 14410
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.409054976466379,
+      "learning_rate": 1.5382752874874595e-05,
+      "loss": 0.8298,
+      "step": 14411
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.284776363511737,
+      "learning_rate": 1.5382109791796073e-05,
+      "loss": 0.7396,
+      "step": 14412
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4245420496715466,
+      "learning_rate": 1.53814666773812e-05,
+      "loss": 0.7869,
+      "step": 14413
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.472288963173806,
+      "learning_rate": 1.5380823531633727e-05,
+      "loss": 0.7028,
+      "step": 14414
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2485489297475,
+      "learning_rate": 1.5380180354557398e-05,
+      "loss": 0.7355,
+      "step": 14415
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.703524588511103,
+      "learning_rate": 1.5379537146155954e-05,
+      "loss": 0.7699,
+      "step": 14416
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.6423966547298527,
+      "learning_rate": 1.5378893906433142e-05,
+      "loss": 0.7923,
+      "step": 14417
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3967809521837797,
+      "learning_rate": 1.5378250635392707e-05,
+      "loss": 0.7704,
+      "step": 14418
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7241827427059064,
+      "learning_rate": 1.5377607333038392e-05,
+      "loss": 0.7759,
+      "step": 14419
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4297968073505452,
+      "learning_rate": 1.5376963999373945e-05,
+      "loss": 0.7629,
+      "step": 14420
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5551882202591922,
+      "learning_rate": 1.5376320634403112e-05,
+      "loss": 0.8324,
+      "step": 14421
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3364695188995521,
+      "learning_rate": 1.537567723812964e-05,
+      "loss": 0.7238,
+      "step": 14422
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7940209912164593,
+      "learning_rate": 1.537503381055727e-05,
+      "loss": 0.74,
+      "step": 14423
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5359622174327079,
+      "learning_rate": 1.5374390351689752e-05,
+      "loss": 0.6436,
+      "step": 14424
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.374807597794679,
+      "learning_rate": 1.5373746861530835e-05,
+      "loss": 0.7747,
+      "step": 14425
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5993729858178836,
+      "learning_rate": 1.5373103340084262e-05,
+      "loss": 0.552,
+      "step": 14426
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9390947375981726,
+      "learning_rate": 1.537245978735378e-05,
+      "loss": 0.8371,
+      "step": 14427
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4401619097260818,
+      "learning_rate": 1.5371816203343132e-05,
+      "loss": 0.7728,
+      "step": 14428
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5855772806648085,
+      "learning_rate": 1.5371172588056076e-05,
+      "loss": 0.7897,
+      "step": 14429
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.626964725110122,
+      "learning_rate": 1.5370528941496346e-05,
+      "loss": 0.5415,
+      "step": 14430
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.153423056616578,
+      "learning_rate": 1.53698852636677e-05,
+      "loss": 0.6137,
+      "step": 14431
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3270674023798652,
+      "learning_rate": 1.536924155457388e-05,
+      "loss": 0.8003,
+      "step": 14432
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6168191896849606,
+      "learning_rate": 1.536859781421864e-05,
+      "loss": 0.5409,
+      "step": 14433
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6803414303715773,
+      "learning_rate": 1.536795404260572e-05,
+      "loss": 0.7196,
+      "step": 14434
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.3714787483298565,
+      "learning_rate": 1.536731023973887e-05,
+      "loss": 0.7159,
+      "step": 14435
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4828473319965128,
+      "learning_rate": 1.5366666405621846e-05,
+      "loss": 0.7639,
+      "step": 14436
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6672883459365723,
+      "learning_rate": 1.5366022540258388e-05,
+      "loss": 0.6153,
+      "step": 14437
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.460215947994561,
+      "learning_rate": 1.536537864365225e-05,
+      "loss": 0.7633,
+      "step": 14438
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.642212069664598,
+      "learning_rate": 1.5364734715807176e-05,
+      "loss": 0.7763,
+      "step": 14439
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.620027345669164,
+      "learning_rate": 1.536409075672692e-05,
+      "loss": 0.5669,
+      "step": 14440
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5855011925383373,
+      "learning_rate": 1.5363446766415226e-05,
+      "loss": 0.5407,
+      "step": 14441
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4180992909296932,
+      "learning_rate": 1.536280274487585e-05,
+      "loss": 0.7606,
+      "step": 14442
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8864374718908237,
+      "learning_rate": 1.536215869211254e-05,
+      "loss": 0.7749,
+      "step": 14443
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.8424511663623,
+      "learning_rate": 1.5361514608129043e-05,
+      "loss": 0.7552,
+      "step": 14444
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.456291825206911,
+      "learning_rate": 1.536087049292911e-05,
+      "loss": 0.6935,
+      "step": 14445
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6845049162105767,
+      "learning_rate": 1.5360226346516493e-05,
+      "loss": 0.8135,
+      "step": 14446
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4206951333262168,
+      "learning_rate": 1.535958216889494e-05,
+      "loss": 0.769,
+      "step": 14447
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4174566321018733,
+      "learning_rate": 1.5358937960068204e-05,
+      "loss": 0.7294,
+      "step": 14448
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3610480760212567,
+      "learning_rate": 1.535829372004003e-05,
+      "loss": 0.7327,
+      "step": 14449
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3567529939746503,
+      "learning_rate": 1.5357649448814177e-05,
+      "loss": 0.7013,
+      "step": 14450
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.670277880909844,
+      "learning_rate": 1.5357005146394394e-05,
+      "loss": 0.799,
+      "step": 14451
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4772479766251299,
+      "learning_rate": 1.535636081278443e-05,
+      "loss": 0.6438,
+      "step": 14452
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.404630066207849,
+      "learning_rate": 1.5355716447988038e-05,
+      "loss": 0.7561,
+      "step": 14453
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.536370640162751,
+      "learning_rate": 1.535507205200897e-05,
+      "loss": 0.7703,
+      "step": 14454
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6178421842449238,
+      "learning_rate": 1.5354427624850978e-05,
+      "loss": 0.7118,
+      "step": 14455
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5417688099919595,
+      "learning_rate": 1.5353783166517806e-05,
+      "loss": 0.7993,
+      "step": 14456
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5935616879774637,
+      "learning_rate": 1.535313867701322e-05,
+      "loss": 0.6557,
+      "step": 14457
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3053083196524566,
+      "learning_rate": 1.5352494156340963e-05,
+      "loss": 0.884,
+      "step": 14458
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3303066087632944,
+      "learning_rate": 1.5351849604504794e-05,
+      "loss": 0.7182,
+      "step": 14459
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3576090329139798,
+      "learning_rate": 1.5351205021508457e-05,
+      "loss": 0.8157,
+      "step": 14460
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6184349915180944,
+      "learning_rate": 1.5350560407355714e-05,
+      "loss": 0.8306,
+      "step": 14461
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6116322421319826,
+      "learning_rate": 1.5349915762050316e-05,
+      "loss": 0.8578,
+      "step": 14462
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5374366707815634,
+      "learning_rate": 1.5349271085596013e-05,
+      "loss": 0.7546,
+      "step": 14463
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6094055829231784,
+      "learning_rate": 1.534862637799656e-05,
+      "loss": 0.7602,
+      "step": 14464
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.688798746128266,
+      "learning_rate": 1.534798163925571e-05,
+      "loss": 0.8787,
+      "step": 14465
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7910996768705505,
+      "learning_rate": 1.5347336869377215e-05,
+      "loss": 0.7136,
+      "step": 14466
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4209314529177586,
+      "learning_rate": 1.5346692068364834e-05,
+      "loss": 0.6157,
+      "step": 14467
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6701505977290212,
+      "learning_rate": 1.534604723622232e-05,
+      "loss": 0.7782,
+      "step": 14468
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6362534607377806,
+      "learning_rate": 1.5345402372953427e-05,
+      "loss": 0.7822,
+      "step": 14469
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6075639499761862,
+      "learning_rate": 1.534475747856191e-05,
+      "loss": 0.5185,
+      "step": 14470
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5633800669458695,
+      "learning_rate": 1.534411255305152e-05,
+      "loss": 0.7471,
+      "step": 14471
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6341495062427165,
+      "learning_rate": 1.5343467596426015e-05,
+      "loss": 0.7726,
+      "step": 14472
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6496777173691572,
+      "learning_rate": 1.5342822608689153e-05,
+      "loss": 0.8122,
+      "step": 14473
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6677159114369862,
+      "learning_rate": 1.5342177589844682e-05,
+      "loss": 0.7552,
+      "step": 14474
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.3439938661803916,
+      "learning_rate": 1.5341532539896364e-05,
+      "loss": 0.7613,
+      "step": 14475
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5810225492267775,
+      "learning_rate": 1.5340887458847954e-05,
+      "loss": 0.53,
+      "step": 14476
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5959972381297584,
+      "learning_rate": 1.5340242346703204e-05,
+      "loss": 0.5652,
+      "step": 14477
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6501499337951004,
+      "learning_rate": 1.5339597203465876e-05,
+      "loss": 0.756,
+      "step": 14478
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.38518228049893,
+      "learning_rate": 1.5338952029139718e-05,
+      "loss": 0.748,
+      "step": 14479
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.256590998841142,
+      "learning_rate": 1.5338306823728494e-05,
+      "loss": 0.6967,
+      "step": 14480
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3755992176088723,
+      "learning_rate": 1.5337661587235952e-05,
+      "loss": 0.7003,
+      "step": 14481
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4605271862300255,
+      "learning_rate": 1.533701631966586e-05,
+      "loss": 0.7354,
+      "step": 14482
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5010846914554616,
+      "learning_rate": 1.533637102102197e-05,
+      "loss": 0.7515,
+      "step": 14483
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3809238866056237,
+      "learning_rate": 1.5335725691308035e-05,
+      "loss": 0.7519,
+      "step": 14484
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4646623197593571,
+      "learning_rate": 1.5335080330527818e-05,
+      "loss": 0.7066,
+      "step": 14485
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2671518354282285,
+      "learning_rate": 1.5334434938685073e-05,
+      "loss": 0.8029,
+      "step": 14486
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3253511553061224,
+      "learning_rate": 1.5333789515783558e-05,
+      "loss": 0.7027,
+      "step": 14487
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.692913807319437,
+      "learning_rate": 1.5333144061827033e-05,
+      "loss": 0.7624,
+      "step": 14488
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5645824374484214,
+      "learning_rate": 1.5332498576819256e-05,
+      "loss": 0.7698,
+      "step": 14489
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.407456182945072,
+      "learning_rate": 1.5331853060763985e-05,
+      "loss": 0.7831,
+      "step": 14490
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.482877040662219,
+      "learning_rate": 1.5331207513664973e-05,
+      "loss": 0.7574,
+      "step": 14491
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5677893630617785,
+      "learning_rate": 1.5330561935525988e-05,
+      "loss": 0.8078,
+      "step": 14492
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4100305915153117,
+      "learning_rate": 1.532991632635078e-05,
+      "loss": 0.8824,
+      "step": 14493
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0077977587692404,
+      "learning_rate": 1.5329270686143112e-05,
+      "loss": 0.8095,
+      "step": 14494
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6275796344652974,
+      "learning_rate": 1.5328625014906746e-05,
+      "loss": 0.5534,
+      "step": 14495
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.748850459120851,
+      "learning_rate": 1.5327979312645433e-05,
+      "loss": 0.8861,
+      "step": 14496
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.440401515503354,
+      "learning_rate": 1.5327333579362945e-05,
+      "loss": 0.6961,
+      "step": 14497
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8340245787333038,
+      "learning_rate": 1.5326687815063025e-05,
+      "loss": 0.7735,
+      "step": 14498
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6978121488801696,
+      "learning_rate": 1.532604201974945e-05,
+      "loss": 0.7205,
+      "step": 14499
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3852983492561712,
+      "learning_rate": 1.5325396193425967e-05,
+      "loss": 0.7563,
+      "step": 14500
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.644028063688218,
+      "learning_rate": 1.5324750336096347e-05,
+      "loss": 0.5491,
+      "step": 14501
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6054646989976238,
+      "learning_rate": 1.5324104447764342e-05,
+      "loss": 0.5523,
+      "step": 14502
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5120020752482364,
+      "learning_rate": 1.5323458528433713e-05,
+      "loss": 0.8257,
+      "step": 14503
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3530381074112101,
+      "learning_rate": 1.5322812578108224e-05,
+      "loss": 0.698,
+      "step": 14504
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6211956813346295,
+      "learning_rate": 1.5322166596791636e-05,
+      "loss": 0.5261,
+      "step": 14505
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6029382849023526,
+      "learning_rate": 1.532152058448771e-05,
+      "loss": 0.5028,
+      "step": 14506
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5352308327845983,
+      "learning_rate": 1.5320874541200202e-05,
+      "loss": 0.7858,
+      "step": 14507
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4482331264603705,
+      "learning_rate": 1.5320228466932885e-05,
+      "loss": 0.6621,
+      "step": 14508
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8283083663323245,
+      "learning_rate": 1.5319582361689508e-05,
+      "loss": 0.8866,
+      "step": 14509
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4723552924217804,
+      "learning_rate": 1.531893622547384e-05,
+      "loss": 0.7575,
+      "step": 14510
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9028717482765998,
+      "learning_rate": 1.531829005828964e-05,
+      "loss": 0.8946,
+      "step": 14511
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3392642339541103,
+      "learning_rate": 1.5317643860140675e-05,
+      "loss": 0.8554,
+      "step": 14512
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8011782213251841,
+      "learning_rate": 1.53169976310307e-05,
+      "loss": 0.7832,
+      "step": 14513
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.588999356438269,
+      "learning_rate": 1.5316351370963484e-05,
+      "loss": 0.7191,
+      "step": 14514
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.503607699420497,
+      "learning_rate": 1.5315705079942787e-05,
+      "loss": 0.7783,
+      "step": 14515
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9164163530223164,
+      "learning_rate": 1.5315058757972373e-05,
+      "loss": 0.8835,
+      "step": 14516
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4375454330946014,
+      "learning_rate": 1.5314412405056e-05,
+      "loss": 0.692,
+      "step": 14517
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5620566473919626,
+      "learning_rate": 1.531376602119744e-05,
+      "loss": 0.7932,
+      "step": 14518
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7401546916997401,
+      "learning_rate": 1.531311960640045e-05,
+      "loss": 0.8232,
+      "step": 14519
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.348911466519407,
+      "learning_rate": 1.5312473160668796e-05,
+      "loss": 0.8748,
+      "step": 14520
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.683809201736246,
+      "learning_rate": 1.5311826684006243e-05,
+      "loss": 0.7331,
+      "step": 14521
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3553552910416,
+      "learning_rate": 1.5311180176416548e-05,
+      "loss": 0.7956,
+      "step": 14522
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3397984852802616,
+      "learning_rate": 1.5310533637903483e-05,
+      "loss": 0.8598,
+      "step": 14523
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3558527422382358,
+      "learning_rate": 1.530988706847081e-05,
+      "loss": 0.7842,
+      "step": 14524
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6289331950520083,
+      "learning_rate": 1.5309240468122294e-05,
+      "loss": 0.7552,
+      "step": 14525
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4650125165518206,
+      "learning_rate": 1.53085938368617e-05,
+      "loss": 0.7531,
+      "step": 14526
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1664502236765024,
+      "learning_rate": 1.5307947174692795e-05,
+      "loss": 0.672,
+      "step": 14527
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7426410290073893,
+      "learning_rate": 1.5307300481619334e-05,
+      "loss": 0.6867,
+      "step": 14528
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3889080169289503,
+      "learning_rate": 1.5306653757645095e-05,
+      "loss": 0.7497,
+      "step": 14529
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.598397827703057,
+      "learning_rate": 1.5306007002773834e-05,
+      "loss": 0.7643,
+      "step": 14530
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6087957918625795,
+      "learning_rate": 1.5305360217009324e-05,
+      "loss": 0.5228,
+      "step": 14531
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.811218475737073,
+      "learning_rate": 1.530471340035532e-05,
+      "loss": 0.7311,
+      "step": 14532
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.444969405781478,
+      "learning_rate": 1.5304066552815604e-05,
+      "loss": 0.7606,
+      "step": 14533
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3746219020209627,
+      "learning_rate": 1.530341967439393e-05,
+      "loss": 0.6388,
+      "step": 14534
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4459870803024528,
+      "learning_rate": 1.5302772765094065e-05,
+      "loss": 0.7278,
+      "step": 14535
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2508887171286105,
+      "learning_rate": 1.530212582491978e-05,
+      "loss": 0.8513,
+      "step": 14536
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3646553953703704,
+      "learning_rate": 1.5301478853874838e-05,
+      "loss": 0.81,
+      "step": 14537
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.5466930970177115,
+      "learning_rate": 1.5300831851963013e-05,
+      "loss": 0.6867,
+      "step": 14538
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.369442737560788,
+      "learning_rate": 1.530018481918806e-05,
+      "loss": 0.7686,
+      "step": 14539
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.627073045550169,
+      "learning_rate": 1.529953775555376e-05,
+      "loss": 0.5725,
+      "step": 14540
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5760358072021374,
+      "learning_rate": 1.529889066106387e-05,
+      "loss": 0.5179,
+      "step": 14541
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4283473654549763,
+      "learning_rate": 1.5298243535722162e-05,
+      "loss": 0.7067,
+      "step": 14542
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.890856587152815,
+      "learning_rate": 1.52975963795324e-05,
+      "loss": 0.6829,
+      "step": 14543
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4770819245850177,
+      "learning_rate": 1.5296949192498358e-05,
+      "loss": 0.6816,
+      "step": 14544
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4633528260122393,
+      "learning_rate": 1.5296301974623794e-05,
+      "loss": 0.6954,
+      "step": 14545
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3479728031952163,
+      "learning_rate": 1.529565472591249e-05,
+      "loss": 0.7292,
+      "step": 14546
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4970926848176624,
+      "learning_rate": 1.5295007446368207e-05,
+      "loss": 0.7192,
+      "step": 14547
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4239474527450733,
+      "learning_rate": 1.5294360135994712e-05,
+      "loss": 0.7949,
+      "step": 14548
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3898357819464109,
+      "learning_rate": 1.5293712794795777e-05,
+      "loss": 0.9054,
+      "step": 14549
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5329828342775491,
+      "learning_rate": 1.529306542277517e-05,
+      "loss": 0.6711,
+      "step": 14550
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.557065355084592,
+      "learning_rate": 1.529241801993666e-05,
+      "loss": 0.7402,
+      "step": 14551
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3473284692261271,
+      "learning_rate": 1.5291770586284018e-05,
+      "loss": 0.8431,
+      "step": 14552
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5148961989803675,
+      "learning_rate": 1.5291123121821013e-05,
+      "loss": 0.6681,
+      "step": 14553
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8385341184250525,
+      "learning_rate": 1.529047562655141e-05,
+      "loss": 0.753,
+      "step": 14554
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3886600183615885,
+      "learning_rate": 1.528982810047899e-05,
+      "loss": 0.7072,
+      "step": 14555
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6087185650240635,
+      "learning_rate": 1.528918054360751e-05,
+      "loss": 0.53,
+      "step": 14556
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4698870966552084,
+      "learning_rate": 1.528853295594075e-05,
+      "loss": 0.6866,
+      "step": 14557
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8689489708808313,
+      "learning_rate": 1.5287885337482472e-05,
+      "loss": 0.807,
+      "step": 14558
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2608745539722452,
+      "learning_rate": 1.5287237688236455e-05,
+      "loss": 0.8604,
+      "step": 14559
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5735937649136194,
+      "learning_rate": 1.5286590008206465e-05,
+      "loss": 0.744,
+      "step": 14560
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5600508623129441,
+      "learning_rate": 1.5285942297396277e-05,
+      "loss": 0.5134,
+      "step": 14561
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.855267407275299,
+      "learning_rate": 1.5285294555809658e-05,
+      "loss": 0.6657,
+      "step": 14562
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.376005350385648,
+      "learning_rate": 1.528464678345038e-05,
+      "loss": 0.6502,
+      "step": 14563
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6056132359344062,
+      "learning_rate": 1.5283998980322214e-05,
+      "loss": 0.7749,
+      "step": 14564
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.460105530530274,
+      "learning_rate": 1.5283351146428933e-05,
+      "loss": 0.7667,
+      "step": 14565
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4075717934593266,
+      "learning_rate": 1.5282703281774312e-05,
+      "loss": 0.7449,
+      "step": 14566
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7463936075881457,
+      "learning_rate": 1.5282055386362116e-05,
+      "loss": 0.8392,
+      "step": 14567
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4211390596649143,
+      "learning_rate": 1.5281407460196126e-05,
+      "loss": 0.6893,
+      "step": 14568
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6748882170257509,
+      "learning_rate": 1.528075950328011e-05,
+      "loss": 0.6344,
+      "step": 14569
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3885329172658512,
+      "learning_rate": 1.5280111515617835e-05,
+      "loss": 0.8025,
+      "step": 14570
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7117938507054755,
+      "learning_rate": 1.527946349721308e-05,
+      "loss": 0.7393,
+      "step": 14571
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.302982390245907,
+      "learning_rate": 1.5278815448069622e-05,
+      "loss": 0.7509,
+      "step": 14572
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.413727358268799,
+      "learning_rate": 1.5278167368191224e-05,
+      "loss": 0.7893,
+      "step": 14573
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9275045518482283,
+      "learning_rate": 1.5277519257581668e-05,
+      "loss": 0.7792,
+      "step": 14574
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.40808161070804,
+      "learning_rate": 1.527687111624472e-05,
+      "loss": 0.8326,
+      "step": 14575
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5261021961486736,
+      "learning_rate": 1.527622294418416e-05,
+      "loss": 0.6109,
+      "step": 14576
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.377300615102298,
+      "learning_rate": 1.5275574741403763e-05,
+      "loss": 0.7141,
+      "step": 14577
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6123586137651498,
+      "learning_rate": 1.52749265079073e-05,
+      "loss": 0.7578,
+      "step": 14578
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3893477688312261,
+      "learning_rate": 1.527427824369854e-05,
+      "loss": 0.7328,
+      "step": 14579
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3519072250979547,
+      "learning_rate": 1.5273629948781264e-05,
+      "loss": 0.8725,
+      "step": 14580
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.1710573399328377,
+      "learning_rate": 1.527298162315924e-05,
+      "loss": 0.7425,
+      "step": 14581
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5053671109345275,
+      "learning_rate": 1.5272333266836256e-05,
+      "loss": 0.7081,
+      "step": 14582
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5030144164643124,
+      "learning_rate": 1.5271684879816074e-05,
+      "loss": 0.9643,
+      "step": 14583
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4546127702688407,
+      "learning_rate": 1.5271036462102472e-05,
+      "loss": 0.7285,
+      "step": 14584
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7210016108766832,
+      "learning_rate": 1.5270388013699232e-05,
+      "loss": 0.7396,
+      "step": 14585
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7074063248543005,
+      "learning_rate": 1.526973953461012e-05,
+      "loss": 0.7485,
+      "step": 14586
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6073511643115803,
+      "learning_rate": 1.526909102483892e-05,
+      "loss": 0.7395,
+      "step": 14587
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3304726530979512,
+      "learning_rate": 1.52684424843894e-05,
+      "loss": 0.8159,
+      "step": 14588
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4134581709904022,
+      "learning_rate": 1.5267793913265344e-05,
+      "loss": 0.7546,
+      "step": 14589
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.767594278525864,
+      "learning_rate": 1.526714531147052e-05,
+      "loss": 0.8016,
+      "step": 14590
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5245647112369796,
+      "learning_rate": 1.526649667900871e-05,
+      "loss": 0.7078,
+      "step": 14591
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5085684707252534,
+      "learning_rate": 1.5265848015883687e-05,
+      "loss": 0.7291,
+      "step": 14592
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0766256224990203,
+      "learning_rate": 1.5265199322099235e-05,
+      "loss": 0.743,
+      "step": 14593
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6053574119381138,
+      "learning_rate": 1.5264550597659122e-05,
+      "loss": 0.524,
+      "step": 14594
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.480617291044587,
+      "learning_rate": 1.526390184256713e-05,
+      "loss": 0.8433,
+      "step": 14595
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5608911470039127,
+      "learning_rate": 1.5263253056827032e-05,
+      "loss": 0.5143,
+      "step": 14596
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3104230235429268,
+      "learning_rate": 1.526260424044261e-05,
+      "loss": 0.8552,
+      "step": 14597
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5559311110144054,
+      "learning_rate": 1.5261955393417643e-05,
+      "loss": 0.6982,
+      "step": 14598
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5927004488346421,
+      "learning_rate": 1.5261306515755904e-05,
+      "loss": 0.7838,
+      "step": 14599
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8171994907624511,
+      "learning_rate": 1.5260657607461167e-05,
+      "loss": 0.8567,
+      "step": 14600
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4233631243159548,
+      "learning_rate": 1.5260008668537225e-05,
+      "loss": 0.7392,
+      "step": 14601
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5002655154452236,
+      "learning_rate": 1.525935969898784e-05,
+      "loss": 0.6301,
+      "step": 14602
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7996714830474603,
+      "learning_rate": 1.52587106988168e-05,
+      "loss": 0.6907,
+      "step": 14603
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5940977509996435,
+      "learning_rate": 1.5258061668027884e-05,
+      "loss": 0.7073,
+      "step": 14604
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5511916138369912,
+      "learning_rate": 1.5257412606624868e-05,
+      "loss": 0.7789,
+      "step": 14605
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.433600212085121,
+      "learning_rate": 1.525676351461153e-05,
+      "loss": 0.795,
+      "step": 14606
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8553780652930492,
+      "learning_rate": 1.525611439199165e-05,
+      "loss": 0.7331,
+      "step": 14607
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.54471930362296,
+      "learning_rate": 1.5255465238769009e-05,
+      "loss": 0.6552,
+      "step": 14608
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.322470319173549,
+      "learning_rate": 1.5254816054947385e-05,
+      "loss": 0.7656,
+      "step": 14609
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.342194374764081,
+      "learning_rate": 1.525416684053056e-05,
+      "loss": 0.6255,
+      "step": 14610
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2378143189915924,
+      "learning_rate": 1.5253517595522308e-05,
+      "loss": 0.8806,
+      "step": 14611
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.501471905546329,
+      "learning_rate": 1.5252868319926418e-05,
+      "loss": 0.6752,
+      "step": 14612
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3221761949899653,
+      "learning_rate": 1.5252219013746663e-05,
+      "loss": 0.7796,
+      "step": 14613
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5172146128571318,
+      "learning_rate": 1.5251569676986826e-05,
+      "loss": 0.7868,
+      "step": 14614
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4450879021346412,
+      "learning_rate": 1.5250920309650688e-05,
+      "loss": 0.7582,
+      "step": 14615
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.801673789644837,
+      "learning_rate": 1.5250270911742031e-05,
+      "loss": 0.7535,
+      "step": 14616
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.441695673798254,
+      "learning_rate": 1.5249621483264632e-05,
+      "loss": 0.7359,
+      "step": 14617
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.8743597376854364,
+      "learning_rate": 1.5248972024222277e-05,
+      "loss": 0.8419,
+      "step": 14618
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5901729192048484,
+      "learning_rate": 1.5248322534618742e-05,
+      "loss": 0.6537,
+      "step": 14619
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9086571405602477,
+      "learning_rate": 1.5247673014457815e-05,
+      "loss": 0.6375,
+      "step": 14620
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3853719693220983,
+      "learning_rate": 1.5247023463743274e-05,
+      "loss": 0.7892,
+      "step": 14621
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.356894502824097,
+      "learning_rate": 1.5246373882478899e-05,
+      "loss": 0.894,
+      "step": 14622
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5731456217374853,
+      "learning_rate": 1.5245724270668476e-05,
+      "loss": 0.8683,
+      "step": 14623
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3826753058852077,
+      "learning_rate": 1.5245074628315786e-05,
+      "loss": 0.6968,
+      "step": 14624
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4490158278862164,
+      "learning_rate": 1.5244424955424611e-05,
+      "loss": 0.7971,
+      "step": 14625
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4101227693918548,
+      "learning_rate": 1.5243775251998729e-05,
+      "loss": 0.8581,
+      "step": 14626
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4827167816012201,
+      "learning_rate": 1.5243125518041935e-05,
+      "loss": 0.7228,
+      "step": 14627
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5532985987025305,
+      "learning_rate": 1.5242475753558e-05,
+      "loss": 0.6905,
+      "step": 14628
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4237500662190041,
+      "learning_rate": 1.5241825958550712e-05,
+      "loss": 0.7593,
+      "step": 14629
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.411826480733602,
+      "learning_rate": 1.524117613302385e-05,
+      "loss": 0.7233,
+      "step": 14630
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.63441692487757,
+      "learning_rate": 1.5240526276981207e-05,
+      "loss": 0.7883,
+      "step": 14631
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3623570242436986,
+      "learning_rate": 1.5239876390426558e-05,
+      "loss": 0.6875,
+      "step": 14632
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4024453724966661,
+      "learning_rate": 1.523922647336369e-05,
+      "loss": 0.7427,
+      "step": 14633
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.608383247669401,
+      "learning_rate": 1.523857652579639e-05,
+      "loss": 0.6034,
+      "step": 14634
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5820712898930324,
+      "learning_rate": 1.5237926547728432e-05,
+      "loss": 0.5242,
+      "step": 14635
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6775166861369943,
+      "learning_rate": 1.5237276539163613e-05,
+      "loss": 0.7349,
+      "step": 14636
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6986675109611165,
+      "learning_rate": 1.5236626500105706e-05,
+      "loss": 0.726,
+      "step": 14637
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6133616121545672,
+      "learning_rate": 1.523597643055851e-05,
+      "loss": 0.5546,
+      "step": 14638
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7006097735836307,
+      "learning_rate": 1.5235326330525794e-05,
+      "loss": 0.7657,
+      "step": 14639
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6721433946834265,
+      "learning_rate": 1.5234676200011356e-05,
+      "loss": 0.6974,
+      "step": 14640
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4805432727730277,
+      "learning_rate": 1.5234026039018974e-05,
+      "loss": 0.7531,
+      "step": 14641
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6440543099363562,
+      "learning_rate": 1.5233375847552436e-05,
+      "loss": 0.5336,
+      "step": 14642
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4507183833963835,
+      "learning_rate": 1.5232725625615525e-05,
+      "loss": 0.7956,
+      "step": 14643
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9218237371260154,
+      "learning_rate": 1.523207537321203e-05,
+      "loss": 0.7537,
+      "step": 14644
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5043399255172285,
+      "learning_rate": 1.5231425090345737e-05,
+      "loss": 0.712,
+      "step": 14645
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.5659142248893576,
+      "learning_rate": 1.5230774777020428e-05,
+      "loss": 0.7666,
+      "step": 14646
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6364965257929918,
+      "learning_rate": 1.5230124433239894e-05,
+      "loss": 0.8426,
+      "step": 14647
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4777509333221595,
+      "learning_rate": 1.5229474059007921e-05,
+      "loss": 0.8026,
+      "step": 14648
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6099317384128169,
+      "learning_rate": 1.5228823654328292e-05,
+      "loss": 0.5582,
+      "step": 14649
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6262995490403291,
+      "learning_rate": 1.5228173219204795e-05,
+      "loss": 0.715,
+      "step": 14650
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6588630291303516,
+      "learning_rate": 1.5227522753641222e-05,
+      "loss": 0.5501,
+      "step": 14651
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7269853303687361,
+      "learning_rate": 1.5226872257641354e-05,
+      "loss": 0.7545,
+      "step": 14652
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6263391644361416,
+      "learning_rate": 1.5226221731208984e-05,
+      "loss": 0.5333,
+      "step": 14653
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7556805131115976,
+      "learning_rate": 1.5225571174347893e-05,
+      "loss": 0.7131,
+      "step": 14654
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4701359983497677,
+      "learning_rate": 1.5224920587061873e-05,
+      "loss": 0.8618,
+      "step": 14655
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3453071261857557,
+      "learning_rate": 1.5224269969354713e-05,
+      "loss": 0.71,
+      "step": 14656
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3166552462559902,
+      "learning_rate": 1.52236193212302e-05,
+      "loss": 0.8378,
+      "step": 14657
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3806738675418144,
+      "learning_rate": 1.5222968642692118e-05,
+      "loss": 0.7862,
+      "step": 14658
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0792827317713165,
+      "learning_rate": 1.5222317933744262e-05,
+      "loss": 0.7366,
+      "step": 14659
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3834715545740137,
+      "learning_rate": 1.5221667194390412e-05,
+      "loss": 0.8538,
+      "step": 14660
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5331401919665228,
+      "learning_rate": 1.522101642463437e-05,
+      "loss": 0.6381,
+      "step": 14661
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8442735930106815,
+      "learning_rate": 1.5220365624479914e-05,
+      "loss": 0.7503,
+      "step": 14662
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5685737447381183,
+      "learning_rate": 1.5219714793930838e-05,
+      "loss": 0.6469,
+      "step": 14663
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5248354560901745,
+      "learning_rate": 1.5219063932990927e-05,
+      "loss": 0.7562,
+      "step": 14664
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6471805417334513,
+      "learning_rate": 1.5218413041663976e-05,
+      "loss": 0.5677,
+      "step": 14665
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6188623000626643,
+      "learning_rate": 1.521776211995377e-05,
+      "loss": 0.801,
+      "step": 14666
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4884381065806442,
+      "learning_rate": 1.5217111167864102e-05,
+      "loss": 0.6975,
+      "step": 14667
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4352755201508984,
+      "learning_rate": 1.5216460185398761e-05,
+      "loss": 0.8075,
+      "step": 14668
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5212919902422861,
+      "learning_rate": 1.521580917256154e-05,
+      "loss": 0.6874,
+      "step": 14669
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7101569672082584,
+      "learning_rate": 1.5215158129356222e-05,
+      "loss": 0.6661,
+      "step": 14670
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6283655394727765,
+      "learning_rate": 1.5214507055786605e-05,
+      "loss": 0.7116,
+      "step": 14671
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.579775573727699,
+      "learning_rate": 1.5213855951856475e-05,
+      "loss": 0.6616,
+      "step": 14672
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.553602930379796,
+      "learning_rate": 1.5213204817569627e-05,
+      "loss": 0.5037,
+      "step": 14673
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3912711658208352,
+      "learning_rate": 1.521255365292985e-05,
+      "loss": 0.8122,
+      "step": 14674
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3174993750181305,
+      "learning_rate": 1.5211902457940934e-05,
+      "loss": 0.7766,
+      "step": 14675
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3162925480137286,
+      "learning_rate": 1.5211251232606674e-05,
+      "loss": 0.7726,
+      "step": 14676
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2498720087940733,
+      "learning_rate": 1.5210599976930857e-05,
+      "loss": 0.6591,
+      "step": 14677
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5937519045516105,
+      "learning_rate": 1.5209948690917282e-05,
+      "loss": 0.601,
+      "step": 14678
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5242294370307986,
+      "learning_rate": 1.5209297374569732e-05,
+      "loss": 0.5448,
+      "step": 14679
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.746277269249224,
+      "learning_rate": 1.5208646027892005e-05,
+      "loss": 0.8285,
+      "step": 14680
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4494998134099903,
+      "learning_rate": 1.520799465088789e-05,
+      "loss": 0.6892,
+      "step": 14681
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7976997988647154,
+      "learning_rate": 1.5207343243561182e-05,
+      "loss": 0.5887,
+      "step": 14682
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5780665442162192,
+      "learning_rate": 1.5206691805915674e-05,
+      "loss": 0.7707,
+      "step": 14683
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7738600034502354,
+      "learning_rate": 1.5206040337955157e-05,
+      "loss": 0.7947,
+      "step": 14684
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.828501783075208,
+      "learning_rate": 1.5205388839683428e-05,
+      "loss": 0.8067,
+      "step": 14685
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7018180235787084,
+      "learning_rate": 1.5204737311104271e-05,
+      "loss": 0.7906,
+      "step": 14686
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5719401093810836,
+      "learning_rate": 1.5204085752221491e-05,
+      "loss": 0.5366,
+      "step": 14687
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3096347969153046,
+      "learning_rate": 1.5203434163038872e-05,
+      "loss": 0.7229,
+      "step": 14688
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5560195546833306,
+      "learning_rate": 1.5202782543560214e-05,
+      "loss": 0.7899,
+      "step": 14689
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.689597241816149,
+      "learning_rate": 1.520213089378931e-05,
+      "loss": 0.652,
+      "step": 14690
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.381376750890112,
+      "learning_rate": 1.5201479213729954e-05,
+      "loss": 0.7725,
+      "step": 14691
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5071731336386351,
+      "learning_rate": 1.5200827503385935e-05,
+      "loss": 0.6882,
+      "step": 14692
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6546923290677715,
+      "learning_rate": 1.5200175762761055e-05,
+      "loss": 0.771,
+      "step": 14693
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2213895572907838,
+      "learning_rate": 1.5199523991859107e-05,
+      "loss": 0.7699,
+      "step": 14694
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7741422380779257,
+      "learning_rate": 1.519887219068388e-05,
+      "loss": 0.7589,
+      "step": 14695
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4370243245808803,
+      "learning_rate": 1.5198220359239173e-05,
+      "loss": 0.7339,
+      "step": 14696
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7337973561431808,
+      "learning_rate": 1.5197568497528785e-05,
+      "loss": 0.848,
+      "step": 14697
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2709211918036332,
+      "learning_rate": 1.5196916605556504e-05,
+      "loss": 0.7716,
+      "step": 14698
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6124862430532139,
+      "learning_rate": 1.5196264683326131e-05,
+      "loss": 0.7921,
+      "step": 14699
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8269667088974282,
+      "learning_rate": 1.519561273084146e-05,
+      "loss": 0.7837,
+      "step": 14700
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6070975042399085,
+      "learning_rate": 1.5194960748106287e-05,
+      "loss": 0.5926,
+      "step": 14701
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5877616485622357,
+      "learning_rate": 1.5194308735124406e-05,
+      "loss": 0.7406,
+      "step": 14702
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2631259851807821,
+      "learning_rate": 1.5193656691899618e-05,
+      "loss": 0.7867,
+      "step": 14703
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8893157967771061,
+      "learning_rate": 1.5193004618435713e-05,
+      "loss": 0.714,
+      "step": 14704
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3239087680823325,
+      "learning_rate": 1.5192352514736495e-05,
+      "loss": 0.7702,
+      "step": 14705
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6145474107482197,
+      "learning_rate": 1.5191700380805754e-05,
+      "loss": 0.5724,
+      "step": 14706
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8019518218123876,
+      "learning_rate": 1.5191048216647291e-05,
+      "loss": 0.7208,
+      "step": 14707
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7580110050661688,
+      "learning_rate": 1.5190396022264903e-05,
+      "loss": 0.773,
+      "step": 14708
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4471034321960665,
+      "learning_rate": 1.5189743797662382e-05,
+      "loss": 0.8037,
+      "step": 14709
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4262530868132248,
+      "learning_rate": 1.5189091542843533e-05,
+      "loss": 0.8593,
+      "step": 14710
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4300462231295519,
+      "learning_rate": 1.5188439257812148e-05,
+      "loss": 0.7892,
+      "step": 14711
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7828063071554583,
+      "learning_rate": 1.5187786942572032e-05,
+      "loss": 0.7733,
+      "step": 14712
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5385892481835548,
+      "learning_rate": 1.518713459712697e-05,
+      "loss": 0.8194,
+      "step": 14713
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7041346492977842,
+      "learning_rate": 1.5186482221480774e-05,
+      "loss": 0.6238,
+      "step": 14714
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4790516458876464,
+      "learning_rate": 1.5185829815637236e-05,
+      "loss": 0.7677,
+      "step": 14715
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.309756651564651,
+      "learning_rate": 1.5185177379600153e-05,
+      "loss": 0.8507,
+      "step": 14716
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.455406420120726,
+      "learning_rate": 1.5184524913373326e-05,
+      "loss": 0.8069,
+      "step": 14717
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5763906103083056,
+      "learning_rate": 1.5183872416960554e-05,
+      "loss": 0.4961,
+      "step": 14718
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6216283185406626,
+      "learning_rate": 1.5183219890365638e-05,
+      "loss": 0.743,
+      "step": 14719
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4736053067438306,
+      "learning_rate": 1.5182567333592368e-05,
+      "loss": 0.7169,
+      "step": 14720
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5435675443742305,
+      "learning_rate": 1.5181914746644557e-05,
+      "loss": 0.6291,
+      "step": 14721
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3313680790291935,
+      "learning_rate": 1.5181262129525996e-05,
+      "loss": 0.8335,
+      "step": 14722
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.37915754142286,
+      "learning_rate": 1.5180609482240485e-05,
+      "loss": 0.7021,
+      "step": 14723
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3904439639707422,
+      "learning_rate": 1.5179956804791826e-05,
+      "loss": 0.8034,
+      "step": 14724
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4303597165087618,
+      "learning_rate": 1.517930409718382e-05,
+      "loss": 0.8088,
+      "step": 14725
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3090856852697423,
+      "learning_rate": 1.5178651359420264e-05,
+      "loss": 0.8675,
+      "step": 14726
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.568983821390816,
+      "learning_rate": 1.5177998591504963e-05,
+      "loss": 0.733,
+      "step": 14727
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4811024566328492,
+      "learning_rate": 1.517734579344171e-05,
+      "loss": 0.6938,
+      "step": 14728
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.627634221433271,
+      "learning_rate": 1.5176692965234316e-05,
+      "loss": 0.7955,
+      "step": 14729
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.460514194099493,
+      "learning_rate": 1.5176040106886574e-05,
+      "loss": 0.7592,
+      "step": 14730
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1778098049486174,
+      "learning_rate": 1.5175387218402287e-05,
+      "loss": 0.7093,
+      "step": 14731
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3442355856689152,
+      "learning_rate": 1.5174734299785259e-05,
+      "loss": 0.7291,
+      "step": 14732
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4418290182118476,
+      "learning_rate": 1.5174081351039287e-05,
+      "loss": 0.6911,
+      "step": 14733
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8701224888587615,
+      "learning_rate": 1.5173428372168177e-05,
+      "loss": 0.8004,
+      "step": 14734
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1921098729173987,
+      "learning_rate": 1.5172775363175732e-05,
+      "loss": 0.7093,
+      "step": 14735
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6064974075294581,
+      "learning_rate": 1.5172122324065748e-05,
+      "loss": 0.5259,
+      "step": 14736
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3635277010954439,
+      "learning_rate": 1.5171469254842031e-05,
+      "loss": 0.8499,
+      "step": 14737
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7536715856000284,
+      "learning_rate": 1.517081615550838e-05,
+      "loss": 0.7976,
+      "step": 14738
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4654662812456487,
+      "learning_rate": 1.5170163026068602e-05,
+      "loss": 0.7973,
+      "step": 14739
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3531071587909103,
+      "learning_rate": 1.5169509866526501e-05,
+      "loss": 0.6662,
+      "step": 14740
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6441732692268243,
+      "learning_rate": 1.5168856676885874e-05,
+      "loss": 0.8556,
+      "step": 14741
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5910373538807276,
+      "learning_rate": 1.5168203457150527e-05,
+      "loss": 0.8505,
+      "step": 14742
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5805284818103889,
+      "learning_rate": 1.5167550207324263e-05,
+      "loss": 0.7099,
+      "step": 14743
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.048979975146188,
+      "learning_rate": 1.516689692741089e-05,
+      "loss": 0.6449,
+      "step": 14744
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4802861227249369,
+      "learning_rate": 1.5166243617414202e-05,
+      "loss": 0.7014,
+      "step": 14745
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3765199446145358,
+      "learning_rate": 1.5165590277338011e-05,
+      "loss": 0.7913,
+      "step": 14746
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3433380939402113,
+      "learning_rate": 1.5164936907186116e-05,
+      "loss": 0.7032,
+      "step": 14747
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3459716933879948,
+      "learning_rate": 1.5164283506962326e-05,
+      "loss": 0.7454,
+      "step": 14748
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3315651417172798,
+      "learning_rate": 1.516363007667044e-05,
+      "loss": 0.7326,
+      "step": 14749
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3549762078703653,
+      "learning_rate": 1.5162976616314264e-05,
+      "loss": 0.7439,
+      "step": 14750
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6894249443521496,
+      "learning_rate": 1.5162323125897604e-05,
+      "loss": 0.7832,
+      "step": 14751
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5336157900303484,
+      "learning_rate": 1.5161669605424268e-05,
+      "loss": 0.7335,
+      "step": 14752
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3089335779649194,
+      "learning_rate": 1.5161016054898055e-05,
+      "loss": 0.6602,
+      "step": 14753
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5368328638640745,
+      "learning_rate": 1.5160362474322771e-05,
+      "loss": 0.8806,
+      "step": 14754
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.500891757954649,
+      "learning_rate": 1.5159708863702225e-05,
+      "loss": 0.807,
+      "step": 14755
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7000255291275415,
+      "learning_rate": 1.515905522304022e-05,
+      "loss": 0.7846,
+      "step": 14756
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.379769741233586,
+      "learning_rate": 1.5158401552340562e-05,
+      "loss": 0.7015,
+      "step": 14757
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2927094558830867,
+      "learning_rate": 1.5157747851607055e-05,
+      "loss": 0.725,
+      "step": 14758
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.466942327735152,
+      "learning_rate": 1.5157094120843512e-05,
+      "loss": 0.7069,
+      "step": 14759
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5884041107526568,
+      "learning_rate": 1.515644036005373e-05,
+      "loss": 0.8516,
+      "step": 14760
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5445108042761078,
+      "learning_rate": 1.5155786569241521e-05,
+      "loss": 0.6415,
+      "step": 14761
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4073579965371414,
+      "learning_rate": 1.515513274841069e-05,
+      "loss": 0.767,
+      "step": 14762
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4923689853019293,
+      "learning_rate": 1.5154478897565046e-05,
+      "loss": 0.6755,
+      "step": 14763
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6694017661004292,
+      "learning_rate": 1.515382501670839e-05,
+      "loss": 0.7454,
+      "step": 14764
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7232176536614034,
+      "learning_rate": 1.5153171105844537e-05,
+      "loss": 0.7853,
+      "step": 14765
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5721356640290496,
+      "learning_rate": 1.5152517164977289e-05,
+      "loss": 0.7164,
+      "step": 14766
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3213383892026387,
+      "learning_rate": 1.5151863194110455e-05,
+      "loss": 0.8148,
+      "step": 14767
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5137572633593048,
+      "learning_rate": 1.515120919324784e-05,
+      "loss": 0.7505,
+      "step": 14768
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2980340467260396,
+      "learning_rate": 1.5150555162393254e-05,
+      "loss": 0.8794,
+      "step": 14769
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.404287944782384,
+      "learning_rate": 1.514990110155051e-05,
+      "loss": 0.822,
+      "step": 14770
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3374450304616932,
+      "learning_rate": 1.5149247010723409e-05,
+      "loss": 0.6236,
+      "step": 14771
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3016603744193485,
+      "learning_rate": 1.5148592889915763e-05,
+      "loss": 0.7075,
+      "step": 14772
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2136637071492,
+      "learning_rate": 1.5147938739131375e-05,
+      "loss": 0.7827,
+      "step": 14773
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9555241809112849,
+      "learning_rate": 1.5147284558374062e-05,
+      "loss": 0.7511,
+      "step": 14774
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2854805977477821,
+      "learning_rate": 1.5146630347647628e-05,
+      "loss": 0.8212,
+      "step": 14775
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.371524855284163,
+      "learning_rate": 1.5145976106955882e-05,
+      "loss": 0.7756,
+      "step": 14776
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4086662628030655,
+      "learning_rate": 1.5145321836302635e-05,
+      "loss": 0.8693,
+      "step": 14777
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9538867921837026,
+      "learning_rate": 1.5144667535691694e-05,
+      "loss": 0.6754,
+      "step": 14778
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7027923283670736,
+      "learning_rate": 1.5144013205126868e-05,
+      "loss": 0.786,
+      "step": 14779
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4155131963143446,
+      "learning_rate": 1.5143358844611973e-05,
+      "loss": 0.7978,
+      "step": 14780
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5899919391852478,
+      "learning_rate": 1.5142704454150812e-05,
+      "loss": 0.5367,
+      "step": 14781
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.390523465578154,
+      "learning_rate": 1.5142050033747197e-05,
+      "loss": 0.8323,
+      "step": 14782
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5076048332890966,
+      "learning_rate": 1.5141395583404939e-05,
+      "loss": 0.6779,
+      "step": 14783
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6145033051330416,
+      "learning_rate": 1.5140741103127848e-05,
+      "loss": 0.5231,
+      "step": 14784
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.443752157863923,
+      "learning_rate": 1.5140086592919735e-05,
+      "loss": 0.7733,
+      "step": 14785
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4322428056278529,
+      "learning_rate": 1.513943205278441e-05,
+      "loss": 0.7616,
+      "step": 14786
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5772916238432343,
+      "learning_rate": 1.5138777482725688e-05,
+      "loss": 0.5277,
+      "step": 14787
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3559897208699183,
+      "learning_rate": 1.5138122882747373e-05,
+      "loss": 0.8061,
+      "step": 14788
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6238838836442164,
+      "learning_rate": 1.513746825285328e-05,
+      "loss": 0.5268,
+      "step": 14789
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5434381515754316,
+      "learning_rate": 1.5136813593047222e-05,
+      "loss": 0.7465,
+      "step": 14790
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.387887603255532,
+      "learning_rate": 1.5136158903333007e-05,
+      "loss": 0.7748,
+      "step": 14791
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7030409065358705,
+      "learning_rate": 1.5135504183714448e-05,
+      "loss": 0.955,
+      "step": 14792
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.496103874351567,
+      "learning_rate": 1.5134849434195362e-05,
+      "loss": 0.7246,
+      "step": 14793
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5019073432651304,
+      "learning_rate": 1.5134194654779555e-05,
+      "loss": 0.6809,
+      "step": 14794
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5781300932296999,
+      "learning_rate": 1.5133539845470839e-05,
+      "loss": 0.5068,
+      "step": 14795
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3242124935414834,
+      "learning_rate": 1.5132885006273032e-05,
+      "loss": 0.817,
+      "step": 14796
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3960038879580732,
+      "learning_rate": 1.5132230137189941e-05,
+      "loss": 0.7467,
+      "step": 14797
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3955316090031444,
+      "learning_rate": 1.5131575238225385e-05,
+      "loss": 0.7523,
+      "step": 14798
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3419437108708074,
+      "learning_rate": 1.5130920309383167e-05,
+      "loss": 0.7838,
+      "step": 14799
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9226264695480424,
+      "learning_rate": 1.5130265350667111e-05,
+      "loss": 0.8187,
+      "step": 14800
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.318120224087967,
+      "learning_rate": 1.5129610362081027e-05,
+      "loss": 0.5828,
+      "step": 14801
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3277446301411928,
+      "learning_rate": 1.5128955343628724e-05,
+      "loss": 0.7721,
+      "step": 14802
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7802687041230976,
+      "learning_rate": 1.5128300295314021e-05,
+      "loss": 0.6059,
+      "step": 14803
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.610790161057972,
+      "learning_rate": 1.512764521714073e-05,
+      "loss": 0.7076,
+      "step": 14804
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4486135039955195,
+      "learning_rate": 1.5126990109112665e-05,
+      "loss": 0.5967,
+      "step": 14805
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1145286110913224,
+      "learning_rate": 1.512633497123364e-05,
+      "loss": 0.8455,
+      "step": 14806
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4503391612360024,
+      "learning_rate": 1.5125679803507472e-05,
+      "loss": 0.8507,
+      "step": 14807
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8464074722355088,
+      "learning_rate": 1.512502460593797e-05,
+      "loss": 0.6838,
+      "step": 14808
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4869016417765386,
+      "learning_rate": 1.5124369378528954e-05,
+      "loss": 0.7035,
+      "step": 14809
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3543704883855145,
+      "learning_rate": 1.512371412128424e-05,
+      "loss": 0.8891,
+      "step": 14810
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.657370215822465,
+      "learning_rate": 1.5123058834207633e-05,
+      "loss": 0.7641,
+      "step": 14811
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.9602890090055394,
+      "learning_rate": 1.5122403517302964e-05,
+      "loss": 0.7564,
+      "step": 14812
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4909066095309549,
+      "learning_rate": 1.512174817057403e-05,
+      "loss": 0.6806,
+      "step": 14813
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6652690488361002,
+      "learning_rate": 1.5121092794024668e-05,
+      "loss": 0.7235,
+      "step": 14814
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.507066347925417,
+      "learning_rate": 1.5120437387658675e-05,
+      "loss": 0.8142,
+      "step": 14815
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2452059665402504,
+      "learning_rate": 1.5119781951479877e-05,
+      "loss": 0.8374,
+      "step": 14816
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7384222813920847,
+      "learning_rate": 1.5119126485492086e-05,
+      "loss": 0.667,
+      "step": 14817
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6228150283717115,
+      "learning_rate": 1.5118470989699119e-05,
+      "loss": 0.5325,
+      "step": 14818
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.566156880640562,
+      "learning_rate": 1.5117815464104797e-05,
+      "loss": 0.7019,
+      "step": 14819
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6720361517470845,
+      "learning_rate": 1.5117159908712929e-05,
+      "loss": 0.7445,
+      "step": 14820
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4922938804932815,
+      "learning_rate": 1.511650432352734e-05,
+      "loss": 0.6216,
+      "step": 14821
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.767305947664144,
+      "learning_rate": 1.511584870855184e-05,
+      "loss": 0.7532,
+      "step": 14822
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9040140094266227,
+      "learning_rate": 1.5115193063790253e-05,
+      "loss": 0.8219,
+      "step": 14823
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5639418534759044,
+      "learning_rate": 1.5114537389246388e-05,
+      "loss": 0.7915,
+      "step": 14824
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4359484439777026,
+      "learning_rate": 1.511388168492407e-05,
+      "loss": 0.8834,
+      "step": 14825
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5661478374586133,
+      "learning_rate": 1.511322595082711e-05,
+      "loss": 0.6667,
+      "step": 14826
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6151436583801005,
+      "learning_rate": 1.5112570186959334e-05,
+      "loss": 0.5864,
+      "step": 14827
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3783261390245714,
+      "learning_rate": 1.5111914393324553e-05,
+      "loss": 0.7764,
+      "step": 14828
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3930252567410935,
+      "learning_rate": 1.511125856992659e-05,
+      "loss": 0.6745,
+      "step": 14829
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.325596346298081,
+      "learning_rate": 1.511060271676926e-05,
+      "loss": 0.7965,
+      "step": 14830
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4835611452054087,
+      "learning_rate": 1.5109946833856386e-05,
+      "loss": 0.7005,
+      "step": 14831
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8282925559426741,
+      "learning_rate": 1.510929092119178e-05,
+      "loss": 0.6739,
+      "step": 14832
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5913448794374432,
+      "learning_rate": 1.5108634978779267e-05,
+      "loss": 0.5387,
+      "step": 14833
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.883875946162356,
+      "learning_rate": 1.5107979006622662e-05,
+      "loss": 0.7363,
+      "step": 14834
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4489576473203376,
+      "learning_rate": 1.5107323004725786e-05,
+      "loss": 0.7924,
+      "step": 14835
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6441980249441742,
+      "learning_rate": 1.510666697309246e-05,
+      "loss": 0.5876,
+      "step": 14836
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1391891492166675,
+      "learning_rate": 1.5106010911726501e-05,
+      "loss": 0.7489,
+      "step": 14837
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3478871368770722,
+      "learning_rate": 1.5105354820631732e-05,
+      "loss": 0.8089,
+      "step": 14838
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4012950235238357,
+      "learning_rate": 1.5104698699811968e-05,
+      "loss": 0.7929,
+      "step": 14839
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.100609029212801,
+      "learning_rate": 1.5104042549271033e-05,
+      "loss": 0.6401,
+      "step": 14840
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4723483970256253,
+      "learning_rate": 1.5103386369012746e-05,
+      "loss": 0.7921,
+      "step": 14841
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.377964842703214,
+      "learning_rate": 1.510273015904093e-05,
+      "loss": 0.7498,
+      "step": 14842
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6432124784572268,
+      "learning_rate": 1.5102073919359402e-05,
+      "loss": 0.7865,
+      "step": 14843
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.272641723705343,
+      "learning_rate": 1.5101417649971986e-05,
+      "loss": 0.7794,
+      "step": 14844
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7955567170602198,
+      "learning_rate": 1.51007613508825e-05,
+      "loss": 0.892,
+      "step": 14845
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4909702202292432,
+      "learning_rate": 1.5100105022094766e-05,
+      "loss": 0.6761,
+      "step": 14846
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5837101683564079,
+      "learning_rate": 1.5099448663612605e-05,
+      "loss": 0.6841,
+      "step": 14847
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2458547237537205,
+      "learning_rate": 1.5098792275439842e-05,
+      "loss": 0.703,
+      "step": 14848
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6517059008869543,
+      "learning_rate": 1.5098135857580296e-05,
+      "loss": 0.878,
+      "step": 14849
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6040722170346493,
+      "learning_rate": 1.5097479410037787e-05,
+      "loss": 0.709,
+      "step": 14850
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5979378698501916,
+      "learning_rate": 1.509682293281614e-05,
+      "loss": 0.7809,
+      "step": 14851
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3578521126772345,
+      "learning_rate": 1.5096166425919176e-05,
+      "loss": 0.7322,
+      "step": 14852
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3600779662587226,
+      "learning_rate": 1.5095509889350716e-05,
+      "loss": 0.6175,
+      "step": 14853
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4329910165905302,
+      "learning_rate": 1.5094853323114587e-05,
+      "loss": 0.7325,
+      "step": 14854
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4916570972112326,
+      "learning_rate": 1.5094196727214607e-05,
+      "loss": 0.7605,
+      "step": 14855
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6073928504535389,
+      "learning_rate": 1.5093540101654602e-05,
+      "loss": 0.8904,
+      "step": 14856
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2906749134815372,
+      "learning_rate": 1.5092883446438396e-05,
+      "loss": 0.7501,
+      "step": 14857
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9729371121730952,
+      "learning_rate": 1.5092226761569805e-05,
+      "loss": 0.6713,
+      "step": 14858
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4610693669944863,
+      "learning_rate": 1.5091570047052666e-05,
+      "loss": 0.8501,
+      "step": 14859
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5094106629909232,
+      "learning_rate": 1.5090913302890786e-05,
+      "loss": 0.8026,
+      "step": 14860
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3218723756131636,
+      "learning_rate": 1.5090256529088e-05,
+      "loss": 0.7085,
+      "step": 14861
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4621561223750326,
+      "learning_rate": 1.5089599725648128e-05,
+      "loss": 0.8205,
+      "step": 14862
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3411102422308483,
+      "learning_rate": 1.5088942892574995e-05,
+      "loss": 0.7637,
+      "step": 14863
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.238126332540756,
+      "learning_rate": 1.5088286029872427e-05,
+      "loss": 0.7162,
+      "step": 14864
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.600243075162475,
+      "learning_rate": 1.5087629137544246e-05,
+      "loss": 0.6834,
+      "step": 14865
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5456663186360173,
+      "learning_rate": 1.508697221559428e-05,
+      "loss": 0.851,
+      "step": 14866
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5395368176029518,
+      "learning_rate": 1.5086315264026345e-05,
+      "loss": 0.7831,
+      "step": 14867
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4966023175582197,
+      "learning_rate": 1.5085658282844278e-05,
+      "loss": 0.8624,
+      "step": 14868
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.529802210084469,
+      "learning_rate": 1.5085001272051893e-05,
+      "loss": 0.7444,
+      "step": 14869
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4852844426951606,
+      "learning_rate": 1.5084344231653027e-05,
+      "loss": 0.8394,
+      "step": 14870
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.804368974148332,
+      "learning_rate": 1.5083687161651495e-05,
+      "loss": 0.74,
+      "step": 14871
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7580211222383124,
+      "learning_rate": 1.508303006205113e-05,
+      "loss": 0.6375,
+      "step": 14872
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3097160256267129,
+      "learning_rate": 1.5082372932855753e-05,
+      "loss": 0.7842,
+      "step": 14873
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6105591172413312,
+      "learning_rate": 1.5081715774069191e-05,
+      "loss": 0.5516,
+      "step": 14874
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.326485188999451,
+      "learning_rate": 1.5081058585695272e-05,
+      "loss": 0.8961,
+      "step": 14875
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.5860285600005346,
+      "learning_rate": 1.508040136773782e-05,
+      "loss": 0.6362,
+      "step": 14876
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8684470127524526,
+      "learning_rate": 1.5079744120200663e-05,
+      "loss": 0.8425,
+      "step": 14877
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4234913605583643,
+      "learning_rate": 1.5079086843087626e-05,
+      "loss": 0.7936,
+      "step": 14878
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5835294176992015,
+      "learning_rate": 1.507842953640254e-05,
+      "loss": 0.5561,
+      "step": 14879
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5037316544643415,
+      "learning_rate": 1.507777220014923e-05,
+      "loss": 0.7364,
+      "step": 14880
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4287099636787373,
+      "learning_rate": 1.5077114834331524e-05,
+      "loss": 0.6857,
+      "step": 14881
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6761848841452434,
+      "learning_rate": 1.5076457438953242e-05,
+      "loss": 0.7523,
+      "step": 14882
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4823777264131772,
+      "learning_rate": 1.5075800014018223e-05,
+      "loss": 0.6433,
+      "step": 14883
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4829246582815927,
+      "learning_rate": 1.5075142559530287e-05,
+      "loss": 0.6467,
+      "step": 14884
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4416131787984618,
+      "learning_rate": 1.5074485075493263e-05,
+      "loss": 0.8002,
+      "step": 14885
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.443552639195534,
+      "learning_rate": 1.5073827561910983e-05,
+      "loss": 0.6963,
+      "step": 14886
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.433226999234487,
+      "learning_rate": 1.5073170018787272e-05,
+      "loss": 0.7975,
+      "step": 14887
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6110157655873432,
+      "learning_rate": 1.5072512446125958e-05,
+      "loss": 0.521,
+      "step": 14888
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5971448659107514,
+      "learning_rate": 1.5071854843930873e-05,
+      "loss": 0.5172,
+      "step": 14889
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3731531185534676,
+      "learning_rate": 1.5071197212205842e-05,
+      "loss": 0.721,
+      "step": 14890
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.325691875542542,
+      "learning_rate": 1.5070539550954696e-05,
+      "loss": 0.7149,
+      "step": 14891
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7043309406529281,
+      "learning_rate": 1.5069881860181261e-05,
+      "loss": 0.7701,
+      "step": 14892
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.395651184858871,
+      "learning_rate": 1.5069224139889372e-05,
+      "loss": 0.7863,
+      "step": 14893
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.575158924312455,
+      "learning_rate": 1.506856639008285e-05,
+      "loss": 0.667,
+      "step": 14894
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5891119158961624,
+      "learning_rate": 1.5067908610765536e-05,
+      "loss": 0.5502,
+      "step": 14895
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4374688077654194,
+      "learning_rate": 1.506725080194125e-05,
+      "loss": 0.8591,
+      "step": 14896
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3326066387019468,
+      "learning_rate": 1.5066592963613826e-05,
+      "loss": 0.7551,
+      "step": 14897
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8815380151687238,
+      "learning_rate": 1.5065935095787093e-05,
+      "loss": 0.6831,
+      "step": 14898
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.460822525676194,
+      "learning_rate": 1.5065277198464885e-05,
+      "loss": 0.8032,
+      "step": 14899
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4113269630479675,
+      "learning_rate": 1.5064619271651027e-05,
+      "loss": 0.7283,
+      "step": 14900
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6754770081251953,
+      "learning_rate": 1.5063961315349352e-05,
+      "loss": 0.7979,
+      "step": 14901
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3343571157682168,
+      "learning_rate": 1.5063303329563693e-05,
+      "loss": 0.7339,
+      "step": 14902
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.962334967447218,
+      "learning_rate": 1.5062645314297877e-05,
+      "loss": 0.6963,
+      "step": 14903
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.541668130982987,
+      "learning_rate": 1.5061987269555735e-05,
+      "loss": 0.7376,
+      "step": 14904
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6584336161315,
+      "learning_rate": 1.5061329195341105e-05,
+      "loss": 0.7189,
+      "step": 14905
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6173954118323952,
+      "learning_rate": 1.5060671091657813e-05,
+      "loss": 0.8192,
+      "step": 14906
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.445691343856006,
+      "learning_rate": 1.506001295850969e-05,
+      "loss": 0.837,
+      "step": 14907
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9715277174021073,
+      "learning_rate": 1.5059354795900571e-05,
+      "loss": 0.6648,
+      "step": 14908
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5947647945796999,
+      "learning_rate": 1.5058696603834287e-05,
+      "loss": 0.744,
+      "step": 14909
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3045407737564367,
+      "learning_rate": 1.5058038382314673e-05,
+      "loss": 0.6952,
+      "step": 14910
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9512783884792162,
+      "learning_rate": 1.505738013134555e-05,
+      "loss": 0.6623,
+      "step": 14911
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6480100373600748,
+      "learning_rate": 1.5056721850930766e-05,
+      "loss": 0.5931,
+      "step": 14912
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5255407562935086,
+      "learning_rate": 1.505606354107414e-05,
+      "loss": 0.68,
+      "step": 14913
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6263588128103914,
+      "learning_rate": 1.5055405201779518e-05,
+      "loss": 0.5976,
+      "step": 14914
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6155485478441454,
+      "learning_rate": 1.5054746833050721e-05,
+      "loss": 0.7123,
+      "step": 14915
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.586714839185792,
+      "learning_rate": 1.505408843489159e-05,
+      "loss": 0.7933,
+      "step": 14916
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5165128761018614,
+      "learning_rate": 1.5053430007305958e-05,
+      "loss": 0.7154,
+      "step": 14917
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.315920152954313,
+      "learning_rate": 1.5052771550297651e-05,
+      "loss": 0.762,
+      "step": 14918
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7711487026245163,
+      "learning_rate": 1.5052113063870514e-05,
+      "loss": 0.7096,
+      "step": 14919
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.652772125332942,
+      "learning_rate": 1.5051454548028373e-05,
+      "loss": 0.7182,
+      "step": 14920
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5871765587922808,
+      "learning_rate": 1.5050796002775065e-05,
+      "loss": 0.6543,
+      "step": 14921
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7018928794220667,
+      "learning_rate": 1.5050137428114422e-05,
+      "loss": 0.7221,
+      "step": 14922
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5712196151269038,
+      "learning_rate": 1.5049478824050282e-05,
+      "loss": 0.8313,
+      "step": 14923
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6081062812732758,
+      "learning_rate": 1.5048820190586478e-05,
+      "loss": 0.8848,
+      "step": 14924
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.420755263883127,
+      "learning_rate": 1.5048161527726843e-05,
+      "loss": 0.7163,
+      "step": 14925
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3133744040431778,
+      "learning_rate": 1.5047502835475211e-05,
+      "loss": 0.775,
+      "step": 14926
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4346342781376158,
+      "learning_rate": 1.5046844113835426e-05,
+      "loss": 0.8038,
+      "step": 14927
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.966057399898517,
+      "learning_rate": 1.504618536281131e-05,
+      "loss": 0.8125,
+      "step": 14928
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4516677497175339,
+      "learning_rate": 1.5045526582406708e-05,
+      "loss": 0.7376,
+      "step": 14929
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5426067888862134,
+      "learning_rate": 1.5044867772625455e-05,
+      "loss": 0.8211,
+      "step": 14930
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.410535602067234,
+      "learning_rate": 1.5044208933471383e-05,
+      "loss": 0.7818,
+      "step": 14931
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4730838484414965,
+      "learning_rate": 1.5043550064948327e-05,
+      "loss": 0.6275,
+      "step": 14932
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5756678057543982,
+      "learning_rate": 1.504289116706013e-05,
+      "loss": 0.7997,
+      "step": 14933
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6888300709010626,
+      "learning_rate": 1.5042232239810618e-05,
+      "loss": 0.7591,
+      "step": 14934
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4079841405867037,
+      "learning_rate": 1.5041573283203641e-05,
+      "loss": 0.812,
+      "step": 14935
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6049819705524242,
+      "learning_rate": 1.5040914297243023e-05,
+      "loss": 0.8012,
+      "step": 14936
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6216344937689433,
+      "learning_rate": 1.5040255281932608e-05,
+      "loss": 0.782,
+      "step": 14937
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3439519028215505,
+      "learning_rate": 1.5039596237276234e-05,
+      "loss": 0.6588,
+      "step": 14938
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.939133300352506,
+      "learning_rate": 1.5038937163277729e-05,
+      "loss": 0.6801,
+      "step": 14939
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6106559441478852,
+      "learning_rate": 1.503827805994094e-05,
+      "loss": 0.5835,
+      "step": 14940
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8561241447301333,
+      "learning_rate": 1.50376189272697e-05,
+      "loss": 0.7564,
+      "step": 14941
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7067465074502746,
+      "learning_rate": 1.5036959765267852e-05,
+      "loss": 0.7305,
+      "step": 14942
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5946853876054348,
+      "learning_rate": 1.5036300573939224e-05,
+      "loss": 0.6906,
+      "step": 14943
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3642192136478537,
+      "learning_rate": 1.5035641353287662e-05,
+      "loss": 0.7374,
+      "step": 14944
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.281857939084795,
+      "learning_rate": 1.5034982103317001e-05,
+      "loss": 0.6875,
+      "step": 14945
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4250600722308668,
+      "learning_rate": 1.5034322824031083e-05,
+      "loss": 0.8329,
+      "step": 14946
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7146419621913165,
+      "learning_rate": 1.5033663515433738e-05,
+      "loss": 0.732,
+      "step": 14947
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2849175303680673,
+      "learning_rate": 1.5033004177528816e-05,
+      "loss": 0.758,
+      "step": 14948
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.522919754832281,
+      "learning_rate": 1.5032344810320148e-05,
+      "loss": 0.6727,
+      "step": 14949
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3905405347859678,
+      "learning_rate": 1.5031685413811574e-05,
+      "loss": 0.7652,
+      "step": 14950
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4466838353432212,
+      "learning_rate": 1.5031025988006935e-05,
+      "loss": 0.8634,
+      "step": 14951
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3900443560758238,
+      "learning_rate": 1.503036653291007e-05,
+      "loss": 0.8673,
+      "step": 14952
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4622407555066794,
+      "learning_rate": 1.502970704852482e-05,
+      "loss": 0.727,
+      "step": 14953
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6613101944239164,
+      "learning_rate": 1.502904753485502e-05,
+      "loss": 0.759,
+      "step": 14954
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6359834960791448,
+      "learning_rate": 1.5028387991904517e-05,
+      "loss": 0.8644,
+      "step": 14955
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6241083554204918,
+      "learning_rate": 1.5027728419677146e-05,
+      "loss": 0.7948,
+      "step": 14956
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.795603237100402,
+      "learning_rate": 1.5027068818176747e-05,
+      "loss": 0.6513,
+      "step": 14957
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3009820208959721,
+      "learning_rate": 1.502640918740716e-05,
+      "loss": 0.8056,
+      "step": 14958
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7085100741589836,
+      "learning_rate": 1.502574952737223e-05,
+      "loss": 0.7431,
+      "step": 14959
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5113676581159166,
+      "learning_rate": 1.5025089838075797e-05,
+      "loss": 0.7706,
+      "step": 14960
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8733495559330255,
+      "learning_rate": 1.50244301195217e-05,
+      "loss": 0.7723,
+      "step": 14961
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4340175650032563,
+      "learning_rate": 1.5023770371713778e-05,
+      "loss": 0.8871,
+      "step": 14962
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2618766311536886,
+      "learning_rate": 1.5023110594655875e-05,
+      "loss": 0.7055,
+      "step": 14963
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.691667222363052,
+      "learning_rate": 1.5022450788351828e-05,
+      "loss": 0.7711,
+      "step": 14964
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6089756353374287,
+      "learning_rate": 1.5021790952805489e-05,
+      "loss": 0.574,
+      "step": 14965
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4844975387683703,
+      "learning_rate": 1.5021131088020688e-05,
+      "loss": 0.7294,
+      "step": 14966
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8855871993389217,
+      "learning_rate": 1.5020471194001274e-05,
+      "loss": 0.6481,
+      "step": 14967
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4031660536508428,
+      "learning_rate": 1.501981127075109e-05,
+      "loss": 0.7948,
+      "step": 14968
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7955687122578712,
+      "learning_rate": 1.501915131827397e-05,
+      "loss": 0.7385,
+      "step": 14969
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5234423357247975,
+      "learning_rate": 1.5018491336573768e-05,
+      "loss": 0.6719,
+      "step": 14970
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5961175862599268,
+      "learning_rate": 1.5017831325654316e-05,
+      "loss": 0.7668,
+      "step": 14971
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3675867315120527,
+      "learning_rate": 1.5017171285519463e-05,
+      "loss": 0.7617,
+      "step": 14972
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6689344037331704,
+      "learning_rate": 1.5016511216173052e-05,
+      "loss": 0.6355,
+      "step": 14973
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5625412037378719,
+      "learning_rate": 1.5015851117618923e-05,
+      "loss": 0.8255,
+      "step": 14974
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0879283370868915,
+      "learning_rate": 1.501519098986092e-05,
+      "loss": 0.8738,
+      "step": 14975
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5162181005432473,
+      "learning_rate": 1.5014530832902886e-05,
+      "loss": 0.7091,
+      "step": 14976
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.527434867292081,
+      "learning_rate": 1.501387064674867e-05,
+      "loss": 0.7971,
+      "step": 14977
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.526358443896864,
+      "learning_rate": 1.501321043140211e-05,
+      "loss": 0.7832,
+      "step": 14978
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5114539390437436,
+      "learning_rate": 1.5012550186867051e-05,
+      "loss": 0.8279,
+      "step": 14979
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.556107492768596,
+      "learning_rate": 1.501188991314734e-05,
+      "loss": 0.6765,
+      "step": 14980
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4909134333953455,
+      "learning_rate": 1.5011229610246815e-05,
+      "loss": 0.7623,
+      "step": 14981
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.296886236756519,
+      "learning_rate": 1.5010569278169327e-05,
+      "loss": 0.8663,
+      "step": 14982
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6696048565934212,
+      "learning_rate": 1.5009908916918719e-05,
+      "loss": 0.6746,
+      "step": 14983
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4297706439854363,
+      "learning_rate": 1.5009248526498834e-05,
+      "loss": 0.8675,
+      "step": 14984
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4074259020514555,
+      "learning_rate": 1.500858810691352e-05,
+      "loss": 0.7463,
+      "step": 14985
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.370197430292973,
+      "learning_rate": 1.5007927658166619e-05,
+      "loss": 0.7564,
+      "step": 14986
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4461467334698113,
+      "learning_rate": 1.5007267180261975e-05,
+      "loss": 0.7023,
+      "step": 14987
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.297915768647508,
+      "learning_rate": 1.500660667320344e-05,
+      "loss": 0.6507,
+      "step": 14988
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4610376044438933,
+      "learning_rate": 1.5005946136994857e-05,
+      "loss": 0.6794,
+      "step": 14989
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5767479547698124,
+      "learning_rate": 1.5005285571640067e-05,
+      "loss": 0.6659,
+      "step": 14990
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.340048153674773,
+      "learning_rate": 1.5004624977142923e-05,
+      "loss": 0.6291,
+      "step": 14991
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4610330927820347,
+      "learning_rate": 1.5003964353507265e-05,
+      "loss": 0.7617,
+      "step": 14992
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.423220616012327,
+      "learning_rate": 1.5003303700736944e-05,
+      "loss": 0.7505,
+      "step": 14993
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0566925729758627,
+      "learning_rate": 1.5002643018835806e-05,
+      "loss": 0.811,
+      "step": 14994
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6059789004047084,
+      "learning_rate": 1.5001982307807694e-05,
+      "loss": 0.5956,
+      "step": 14995
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3810661171446812,
+      "learning_rate": 1.5001321567656459e-05,
+      "loss": 0.7099,
+      "step": 14996
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5825270942709034,
+      "learning_rate": 1.5000660798385943e-05,
+      "loss": 0.5005,
+      "step": 14997
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.474514843404626,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 0.7802,
+      "step": 14998
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3766441828871334,
+      "learning_rate": 1.4999339172502477e-05,
+      "loss": 0.6859,
+      "step": 14999
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3031252175712287,
+      "learning_rate": 1.4998678315897213e-05,
+      "loss": 0.6852,
+      "step": 15000
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5984135451958416,
+      "learning_rate": 1.4998017430188063e-05,
+      "loss": 0.5521,
+      "step": 15001
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.905538531039444,
+      "learning_rate": 1.4997356515378877e-05,
+      "loss": 0.7881,
+      "step": 15002
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7401796942546657,
+      "learning_rate": 1.4996695571473495e-05,
+      "loss": 0.6989,
+      "step": 15003
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4292971165657622,
+      "learning_rate": 1.4996034598475769e-05,
+      "loss": 0.748,
+      "step": 15004
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6433804413849766,
+      "learning_rate": 1.4995373596389549e-05,
+      "loss": 0.5748,
+      "step": 15005
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5209972405697216,
+      "learning_rate": 1.499471256521868e-05,
+      "loss": 0.6333,
+      "step": 15006
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5729120166196315,
+      "learning_rate": 1.4994051504967018e-05,
+      "loss": 0.7701,
+      "step": 15007
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3466685649528864,
+      "learning_rate": 1.4993390415638405e-05,
+      "loss": 0.6829,
+      "step": 15008
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.457058779385549,
+      "learning_rate": 1.4992729297236693e-05,
+      "loss": 0.6161,
+      "step": 15009
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5229338986739815,
+      "learning_rate": 1.4992068149765726e-05,
+      "loss": 0.7292,
+      "step": 15010
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6192991455642504,
+      "learning_rate": 1.499140697322936e-05,
+      "loss": 0.5583,
+      "step": 15011
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.444928035373445,
+      "learning_rate": 1.4990745767631445e-05,
+      "loss": 0.7785,
+      "step": 15012
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2994139368661604,
+      "learning_rate": 1.4990084532975826e-05,
+      "loss": 0.7615,
+      "step": 15013
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3487399592675764,
+      "learning_rate": 1.4989423269266357e-05,
+      "loss": 0.863,
+      "step": 15014
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3203369972486754,
+      "learning_rate": 1.4988761976506885e-05,
+      "loss": 0.7759,
+      "step": 15015
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4415264296997772,
+      "learning_rate": 1.4988100654701258e-05,
+      "loss": 0.7889,
+      "step": 15016
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7264666345276203,
+      "learning_rate": 1.4987439303853336e-05,
+      "loss": 0.7,
+      "step": 15017
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.759912621356738,
+      "learning_rate": 1.4986777923966958e-05,
+      "loss": 0.8266,
+      "step": 15018
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9294921353970689,
+      "learning_rate": 1.4986116515045982e-05,
+      "loss": 0.7355,
+      "step": 15019
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4145714150362607,
+      "learning_rate": 1.4985455077094258e-05,
+      "loss": 0.9052,
+      "step": 15020
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8263560211424847,
+      "learning_rate": 1.4984793610115635e-05,
+      "loss": 0.7474,
+      "step": 15021
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5751840912659109,
+      "learning_rate": 1.4984132114113965e-05,
+      "loss": 0.5154,
+      "step": 15022
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8036789288440507,
+      "learning_rate": 1.4983470589093103e-05,
+      "loss": 0.6471,
+      "step": 15023
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7397143188727773,
+      "learning_rate": 1.4982809035056894e-05,
+      "loss": 0.7936,
+      "step": 15024
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.2210950059421553,
+      "learning_rate": 1.4982147452009197e-05,
+      "loss": 0.8524,
+      "step": 15025
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4754127730159057,
+      "learning_rate": 1.4981485839953855e-05,
+      "loss": 0.7623,
+      "step": 15026
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4099602934212594,
+      "learning_rate": 1.498082419889473e-05,
+      "loss": 0.577,
+      "step": 15027
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4491277965830809,
+      "learning_rate": 1.4980162528835666e-05,
+      "loss": 0.7656,
+      "step": 15028
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6446728108052522,
+      "learning_rate": 1.4979500829780522e-05,
+      "loss": 0.7739,
+      "step": 15029
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3341747792050989,
+      "learning_rate": 1.4978839101733144e-05,
+      "loss": 0.7942,
+      "step": 15030
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9152876480536003,
+      "learning_rate": 1.4978177344697394e-05,
+      "loss": 0.8795,
+      "step": 15031
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5879871853276348,
+      "learning_rate": 1.4977515558677116e-05,
+      "loss": 0.7285,
+      "step": 15032
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3606195656828752,
+      "learning_rate": 1.4976853743676165e-05,
+      "loss": 0.8645,
+      "step": 15033
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2702723480385993,
+      "learning_rate": 1.4976191899698398e-05,
+      "loss": 0.7591,
+      "step": 15034
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5066234786176271,
+      "learning_rate": 1.4975530026747667e-05,
+      "loss": 0.8142,
+      "step": 15035
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6360979223506524,
+      "learning_rate": 1.4974868124827823e-05,
+      "loss": 0.7865,
+      "step": 15036
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6061100747539048,
+      "learning_rate": 1.4974206193942724e-05,
+      "loss": 0.5046,
+      "step": 15037
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6002756141279508,
+      "learning_rate": 1.497354423409622e-05,
+      "loss": 0.7257,
+      "step": 15038
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.313944280219214,
+      "learning_rate": 1.4972882245292165e-05,
+      "loss": 0.6629,
+      "step": 15039
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6431797417777003,
+      "learning_rate": 1.4972220227534422e-05,
+      "loss": 0.7565,
+      "step": 15040
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.068253617671802,
+      "learning_rate": 1.4971558180826832e-05,
+      "loss": 0.6687,
+      "step": 15041
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4170056645349256,
+      "learning_rate": 1.497089610517326e-05,
+      "loss": 0.8215,
+      "step": 15042
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5150558399362697,
+      "learning_rate": 1.4970234000577555e-05,
+      "loss": 0.8809,
+      "step": 15043
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3827428596374471,
+      "learning_rate": 1.4969571867043576e-05,
+      "loss": 0.8109,
+      "step": 15044
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.40253650477363,
+      "learning_rate": 1.4968909704575171e-05,
+      "loss": 0.7829,
+      "step": 15045
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6080089859585045,
+      "learning_rate": 1.4968247513176209e-05,
+      "loss": 0.7219,
+      "step": 15046
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3994520133296264,
+      "learning_rate": 1.4967585292850531e-05,
+      "loss": 0.874,
+      "step": 15047
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5807871044388053,
+      "learning_rate": 1.4966923043602002e-05,
+      "loss": 0.5204,
+      "step": 15048
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5447547051814599,
+      "learning_rate": 1.4966260765434472e-05,
+      "loss": 0.7139,
+      "step": 15049
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3635786700319494,
+      "learning_rate": 1.4965598458351797e-05,
+      "loss": 0.7624,
+      "step": 15050
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5130177739382542,
+      "learning_rate": 1.4964936122357843e-05,
+      "loss": 0.8071,
+      "step": 15051
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.523824136586111,
+      "learning_rate": 1.4964273757456451e-05,
+      "loss": 0.7947,
+      "step": 15052
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.530152665564888,
+      "learning_rate": 1.4963611363651494e-05,
+      "loss": 0.8148,
+      "step": 15053
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.2439314048310512,
+      "learning_rate": 1.4962948940946814e-05,
+      "loss": 0.7122,
+      "step": 15054
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4563006900070388,
+      "learning_rate": 1.4962286489346276e-05,
+      "loss": 0.724,
+      "step": 15055
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3266168641877658,
+      "learning_rate": 1.4961624008853735e-05,
+      "loss": 0.7575,
+      "step": 15056
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.54858871919235,
+      "learning_rate": 1.4960961499473048e-05,
+      "loss": 0.7303,
+      "step": 15057
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.524414228770603,
+      "learning_rate": 1.4960298961208073e-05,
+      "loss": 0.7556,
+      "step": 15058
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3607993449457834,
+      "learning_rate": 1.4959636394062666e-05,
+      "loss": 0.7005,
+      "step": 15059
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8689080780147465,
+      "learning_rate": 1.4958973798040685e-05,
+      "loss": 0.7978,
+      "step": 15060
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6053761100725628,
+      "learning_rate": 1.4958311173145991e-05,
+      "loss": 0.5702,
+      "step": 15061
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4559936673154972,
+      "learning_rate": 1.495764851938244e-05,
+      "loss": 0.8215,
+      "step": 15062
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5382142658286466,
+      "learning_rate": 1.4956985836753887e-05,
+      "loss": 0.7416,
+      "step": 15063
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6609900850742949,
+      "learning_rate": 1.4956323125264193e-05,
+      "loss": 0.5806,
+      "step": 15064
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6126118230042482,
+      "learning_rate": 1.4955660384917218e-05,
+      "loss": 0.5207,
+      "step": 15065
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8147831990460368,
+      "learning_rate": 1.495499761571682e-05,
+      "loss": 0.9741,
+      "step": 15066
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4108174563376226,
+      "learning_rate": 1.4954334817666855e-05,
+      "loss": 0.8815,
+      "step": 15067
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6276210623480118,
+      "learning_rate": 1.4953671990771189e-05,
+      "loss": 0.6958,
+      "step": 15068
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5671729399739636,
+      "learning_rate": 1.4953009135033671e-05,
+      "loss": 0.7243,
+      "step": 15069
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.407124690783056,
+      "learning_rate": 1.495234625045817e-05,
+      "loss": 0.6788,
+      "step": 15070
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.693709195702094,
+      "learning_rate": 1.4951683337048536e-05,
+      "loss": 0.7689,
+      "step": 15071
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6314104119634685,
+      "learning_rate": 1.495102039480864e-05,
+      "loss": 0.7128,
+      "step": 15072
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5615961409493115,
+      "learning_rate": 1.495035742374233e-05,
+      "loss": 0.6855,
+      "step": 15073
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2925330547393998,
+      "learning_rate": 1.4949694423853478e-05,
+      "loss": 0.796,
+      "step": 15074
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5068568046499988,
+      "learning_rate": 1.4949031395145932e-05,
+      "loss": 0.7508,
+      "step": 15075
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.443565439488281,
+      "learning_rate": 1.4948368337623562e-05,
+      "loss": 0.7327,
+      "step": 15076
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6227186715170183,
+      "learning_rate": 1.4947705251290223e-05,
+      "loss": 0.8028,
+      "step": 15077
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2875322217294236,
+      "learning_rate": 1.4947042136149778e-05,
+      "loss": 0.7604,
+      "step": 15078
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6589975801856243,
+      "learning_rate": 1.494637899220609e-05,
+      "loss": 0.7589,
+      "step": 15079
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7883604009843692,
+      "learning_rate": 1.4945715819463013e-05,
+      "loss": 0.7775,
+      "step": 15080
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4456879186204157,
+      "learning_rate": 1.4945052617924415e-05,
+      "loss": 0.7844,
+      "step": 15081
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3178130960281502,
+      "learning_rate": 1.4944389387594156e-05,
+      "loss": 0.7662,
+      "step": 15082
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6863365658781493,
+      "learning_rate": 1.4943726128476097e-05,
+      "loss": 0.746,
+      "step": 15083
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5132852321043029,
+      "learning_rate": 1.4943062840574096e-05,
+      "loss": 0.7561,
+      "step": 15084
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.454902988973245,
+      "learning_rate": 1.4942399523892021e-05,
+      "loss": 0.7667,
+      "step": 15085
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6425105047166595,
+      "learning_rate": 1.494173617843373e-05,
+      "loss": 0.7151,
+      "step": 15086
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.638673329786559,
+      "learning_rate": 1.4941072804203089e-05,
+      "loss": 0.7317,
+      "step": 15087
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6508459365044463,
+      "learning_rate": 1.4940409401203955e-05,
+      "loss": 0.825,
+      "step": 15088
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4912056355266583,
+      "learning_rate": 1.4939745969440194e-05,
+      "loss": 0.6861,
+      "step": 15089
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5986523178873085,
+      "learning_rate": 1.4939082508915668e-05,
+      "loss": 0.5256,
+      "step": 15090
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.59050503159041,
+      "learning_rate": 1.4938419019634243e-05,
+      "loss": 0.5188,
+      "step": 15091
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3736039249606729,
+      "learning_rate": 1.4937755501599774e-05,
+      "loss": 0.7416,
+      "step": 15092
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5720282285257531,
+      "learning_rate": 1.4937091954816131e-05,
+      "loss": 0.5626,
+      "step": 15093
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4884516853253618,
+      "learning_rate": 1.4936428379287176e-05,
+      "loss": 0.7279,
+      "step": 15094
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5742137965485725,
+      "learning_rate": 1.4935764775016775e-05,
+      "loss": 0.6987,
+      "step": 15095
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4309413947270164,
+      "learning_rate": 1.4935101142008783e-05,
+      "loss": 0.6593,
+      "step": 15096
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3258920255647142,
+      "learning_rate": 1.4934437480267075e-05,
+      "loss": 0.8463,
+      "step": 15097
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.331500269376508,
+      "learning_rate": 1.4933773789795508e-05,
+      "loss": 0.7614,
+      "step": 15098
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4951763634145168,
+      "learning_rate": 1.4933110070597945e-05,
+      "loss": 0.7242,
+      "step": 15099
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.36647623745033,
+      "learning_rate": 1.4932446322678255e-05,
+      "loss": 0.7057,
+      "step": 15100
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4311815351880306,
+      "learning_rate": 1.4931782546040302e-05,
+      "loss": 0.8296,
+      "step": 15101
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5081841077268028,
+      "learning_rate": 1.4931118740687952e-05,
+      "loss": 0.7077,
+      "step": 15102
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8321070337685206,
+      "learning_rate": 1.4930454906625062e-05,
+      "loss": 0.7408,
+      "step": 15103
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3507752961335524,
+      "learning_rate": 1.4929791043855506e-05,
+      "loss": 0.8549,
+      "step": 15104
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6460997902812213,
+      "learning_rate": 1.4929127152383148e-05,
+      "loss": 0.8321,
+      "step": 15105
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.706390703384272,
+      "learning_rate": 1.4928463232211848e-05,
+      "loss": 0.8018,
+      "step": 15106
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5208914199623023,
+      "learning_rate": 1.4927799283345474e-05,
+      "loss": 0.7407,
+      "step": 15107
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7957306805273645,
+      "learning_rate": 1.4927135305787895e-05,
+      "loss": 0.7743,
+      "step": 15108
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8520298790261334,
+      "learning_rate": 1.492647129954297e-05,
+      "loss": 0.7826,
+      "step": 15109
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3469520934270376,
+      "learning_rate": 1.4925807264614573e-05,
+      "loss": 0.7691,
+      "step": 15110
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.488381127147286,
+      "learning_rate": 1.4925143201006566e-05,
+      "loss": 0.743,
+      "step": 15111
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.380851033380644,
+      "learning_rate": 1.4924479108722816e-05,
+      "loss": 0.8415,
+      "step": 15112
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4334158950398448,
+      "learning_rate": 1.4923814987767191e-05,
+      "loss": 0.7397,
+      "step": 15113
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5186480400700766,
+      "learning_rate": 1.4923150838143555e-05,
+      "loss": 0.7616,
+      "step": 15114
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4005788787733153,
+      "learning_rate": 1.4922486659855774e-05,
+      "loss": 0.8865,
+      "step": 15115
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3998740405610957,
+      "learning_rate": 1.492182245290772e-05,
+      "loss": 0.6992,
+      "step": 15116
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3551033219175497,
+      "learning_rate": 1.4921158217303257e-05,
+      "loss": 0.6642,
+      "step": 15117
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.043064995408625,
+      "learning_rate": 1.4920493953046254e-05,
+      "loss": 0.7493,
+      "step": 15118
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4609309810162927,
+      "learning_rate": 1.4919829660140577e-05,
+      "loss": 0.7366,
+      "step": 15119
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.033927044098323,
+      "learning_rate": 1.491916533859009e-05,
+      "loss": 0.7525,
+      "step": 15120
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5498516381787681,
+      "learning_rate": 1.491850098839867e-05,
+      "loss": 0.7207,
+      "step": 15121
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4908478885774994,
+      "learning_rate": 1.4917836609570176e-05,
+      "loss": 0.7347,
+      "step": 15122
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6356491883991402,
+      "learning_rate": 1.4917172202108485e-05,
+      "loss": 0.5556,
+      "step": 15123
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4252339737655253,
+      "learning_rate": 1.4916507766017455e-05,
+      "loss": 0.6588,
+      "step": 15124
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5774452995033978,
+      "learning_rate": 1.4915843301300968e-05,
+      "loss": 0.6999,
+      "step": 15125
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.571021391943215,
+      "learning_rate": 1.491517880796288e-05,
+      "loss": 0.6774,
+      "step": 15126
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3452277527019274,
+      "learning_rate": 1.4914514286007064e-05,
+      "loss": 0.7194,
+      "step": 15127
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5556486649688155,
+      "learning_rate": 1.4913849735437392e-05,
+      "loss": 0.7558,
+      "step": 15128
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.407148983789262,
+      "learning_rate": 1.491318515625773e-05,
+      "loss": 0.8179,
+      "step": 15129
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6952912083901945,
+      "learning_rate": 1.491252054847195e-05,
+      "loss": 0.7311,
+      "step": 15130
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6193531031807893,
+      "learning_rate": 1.4911855912083916e-05,
+      "loss": 0.7097,
+      "step": 15131
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5910256478486945,
+      "learning_rate": 1.4911191247097504e-05,
+      "loss": 0.5262,
+      "step": 15132
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.311869215701882,
+      "learning_rate": 1.4910526553516585e-05,
+      "loss": 0.7696,
+      "step": 15133
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.438691850374456,
+      "learning_rate": 1.4909861831345022e-05,
+      "loss": 0.6964,
+      "step": 15134
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9962018324715847,
+      "learning_rate": 1.490919708058669e-05,
+      "loss": 0.6732,
+      "step": 15135
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6420178056763806,
+      "learning_rate": 1.4908532301245459e-05,
+      "loss": 0.8088,
+      "step": 15136
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.499039855519202,
+      "learning_rate": 1.4907867493325199e-05,
+      "loss": 0.7433,
+      "step": 15137
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.8488855508084607,
+      "learning_rate": 1.490720265682978e-05,
+      "loss": 0.713,
+      "step": 15138
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7240713424480816,
+      "learning_rate": 1.4906537791763073e-05,
+      "loss": 0.7544,
+      "step": 15139
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5572491520248015,
+      "learning_rate": 1.490587289812895e-05,
+      "loss": 0.6659,
+      "step": 15140
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6619744509853818,
+      "learning_rate": 1.4905207975931282e-05,
+      "loss": 0.5872,
+      "step": 15141
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7411777758743312,
+      "learning_rate": 1.4904543025173943e-05,
+      "loss": 0.8261,
+      "step": 15142
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3102217881118556,
+      "learning_rate": 1.4903878045860798e-05,
+      "loss": 0.7434,
+      "step": 15143
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.673764377461802,
+      "learning_rate": 1.4903213037995725e-05,
+      "loss": 0.7719,
+      "step": 15144
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3398503146079395,
+      "learning_rate": 1.4902548001582591e-05,
+      "loss": 0.8172,
+      "step": 15145
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4686381352791127,
+      "learning_rate": 1.4901882936625272e-05,
+      "loss": 0.8348,
+      "step": 15146
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.395391177154831,
+      "learning_rate": 1.4901217843127636e-05,
+      "loss": 0.7493,
+      "step": 15147
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7397671911904695,
+      "learning_rate": 1.4900552721093559e-05,
+      "loss": 0.6742,
+      "step": 15148
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3996842721427372,
+      "learning_rate": 1.4899887570526916e-05,
+      "loss": 0.6845,
+      "step": 15149
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.480526040681431,
+      "learning_rate": 1.4899222391431572e-05,
+      "loss": 0.6696,
+      "step": 15150
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5710786676709918,
+      "learning_rate": 1.4898557183811409e-05,
+      "loss": 0.7534,
+      "step": 15151
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8945167383794645,
+      "learning_rate": 1.489789194767029e-05,
+      "loss": 0.7891,
+      "step": 15152
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5661556715026455,
+      "learning_rate": 1.4897226683012096e-05,
+      "loss": 0.6623,
+      "step": 15153
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.327901823452926,
+      "learning_rate": 1.4896561389840696e-05,
+      "loss": 0.6263,
+      "step": 15154
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.7801143920437563,
+      "learning_rate": 1.4895896068159967e-05,
+      "loss": 0.8453,
+      "step": 15155
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4356531145741631,
+      "learning_rate": 1.4895230717973783e-05,
+      "loss": 0.7899,
+      "step": 15156
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4283193732591046,
+      "learning_rate": 1.4894565339286014e-05,
+      "loss": 0.6798,
+      "step": 15157
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.430278104631709,
+      "learning_rate": 1.4893899932100534e-05,
+      "loss": 0.8526,
+      "step": 15158
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5757951427293636,
+      "learning_rate": 1.4893234496421222e-05,
+      "loss": 0.5397,
+      "step": 15159
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6475303680814546,
+      "learning_rate": 1.4892569032251949e-05,
+      "loss": 0.8048,
+      "step": 15160
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.3397697168731444,
+      "learning_rate": 1.4891903539596588e-05,
+      "loss": 0.7945,
+      "step": 15161
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4675213352256955,
+      "learning_rate": 1.489123801845902e-05,
+      "loss": 0.854,
+      "step": 15162
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.608116763119394,
+      "learning_rate": 1.4890572468843112e-05,
+      "loss": 0.4886,
+      "step": 15163
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5722086374657314,
+      "learning_rate": 1.4889906890752745e-05,
+      "loss": 0.5667,
+      "step": 15164
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8238729109953047,
+      "learning_rate": 1.488924128419179e-05,
+      "loss": 0.8472,
+      "step": 15165
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4905957447106466,
+      "learning_rate": 1.4888575649164125e-05,
+      "loss": 0.7103,
+      "step": 15166
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6116089839433525,
+      "learning_rate": 1.4887909985673626e-05,
+      "loss": 0.7655,
+      "step": 15167
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5993912528037024,
+      "learning_rate": 1.4887244293724167e-05,
+      "loss": 0.5488,
+      "step": 15168
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9921428312086735,
+      "learning_rate": 1.4886578573319623e-05,
+      "loss": 0.7312,
+      "step": 15169
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4040259565334723,
+      "learning_rate": 1.4885912824463875e-05,
+      "loss": 0.6695,
+      "step": 15170
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6546273728058694,
+      "learning_rate": 1.4885247047160792e-05,
+      "loss": 0.738,
+      "step": 15171
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7398816636487904,
+      "learning_rate": 1.4884581241414257e-05,
+      "loss": 0.8108,
+      "step": 15172
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6044602223497249,
+      "learning_rate": 1.488391540722814e-05,
+      "loss": 0.7438,
+      "step": 15173
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5679186482354686,
+      "learning_rate": 1.4883249544606326e-05,
+      "loss": 0.7228,
+      "step": 15174
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4319835693286846,
+      "learning_rate": 1.4882583653552683e-05,
+      "loss": 0.7857,
+      "step": 15175
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5680064058492189,
+      "learning_rate": 1.4881917734071094e-05,
+      "loss": 0.5153,
+      "step": 15176
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.531744838430909,
+      "learning_rate": 1.4881251786165436e-05,
+      "loss": 0.7517,
+      "step": 15177
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0257521831853245,
+      "learning_rate": 1.488058580983958e-05,
+      "loss": 0.6736,
+      "step": 15178
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4766347865328526,
+      "learning_rate": 1.4879919805097412e-05,
+      "loss": 0.7089,
+      "step": 15179
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7424841502898136,
+      "learning_rate": 1.4879253771942805e-05,
+      "loss": 0.6984,
+      "step": 15180
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8870419250372883,
+      "learning_rate": 1.4878587710379639e-05,
+      "loss": 0.8595,
+      "step": 15181
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4505602880262536,
+      "learning_rate": 1.4877921620411789e-05,
+      "loss": 0.7048,
+      "step": 15182
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8046321489491854,
+      "learning_rate": 1.4877255502043136e-05,
+      "loss": 0.7922,
+      "step": 15183
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7955765075952914,
+      "learning_rate": 1.4876589355277555e-05,
+      "loss": 0.8287,
+      "step": 15184
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7374719303849027,
+      "learning_rate": 1.487592318011893e-05,
+      "loss": 0.7466,
+      "step": 15185
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.190472171509929,
+      "learning_rate": 1.4875256976571135e-05,
+      "loss": 0.7899,
+      "step": 15186
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7861916788890662,
+      "learning_rate": 1.487459074463805e-05,
+      "loss": 0.7145,
+      "step": 15187
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4398436668987362,
+      "learning_rate": 1.4873924484323557e-05,
+      "loss": 0.6713,
+      "step": 15188
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6147831589100874,
+      "learning_rate": 1.487325819563153e-05,
+      "loss": 0.556,
+      "step": 15189
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4630768498986657,
+      "learning_rate": 1.4872591878565852e-05,
+      "loss": 0.7287,
+      "step": 15190
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3754925587494164,
+      "learning_rate": 1.4871925533130402e-05,
+      "loss": 0.7133,
+      "step": 15191
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4992172162752633,
+      "learning_rate": 1.4871259159329056e-05,
+      "loss": 0.6648,
+      "step": 15192
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4495973076947073,
+      "learning_rate": 1.48705927571657e-05,
+      "loss": 0.7157,
+      "step": 15193
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.35549475604504,
+      "learning_rate": 1.4869926326644208e-05,
+      "loss": 0.6452,
+      "step": 15194
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5986452705311561,
+      "learning_rate": 1.486925986776847e-05,
+      "loss": 0.5458,
+      "step": 15195
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.545122507275835,
+      "learning_rate": 1.4868593380542351e-05,
+      "loss": 0.7115,
+      "step": 15196
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6176674206053844,
+      "learning_rate": 1.4867926864969744e-05,
+      "loss": 0.5752,
+      "step": 15197
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5639327424066,
+      "learning_rate": 1.4867260321054526e-05,
+      "loss": 0.5116,
+      "step": 15198
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4041172817677954,
+      "learning_rate": 1.4866593748800578e-05,
+      "loss": 0.7771,
+      "step": 15199
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5057685834692105,
+      "learning_rate": 1.4865927148211779e-05,
+      "loss": 0.7489,
+      "step": 15200
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.798810441886304,
+      "learning_rate": 1.486526051929201e-05,
+      "loss": 0.8114,
+      "step": 15201
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4369490270581176,
+      "learning_rate": 1.4864593862045159e-05,
+      "loss": 0.6814,
+      "step": 15202
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.247282452090205,
+      "learning_rate": 1.4863927176475097e-05,
+      "loss": 0.6704,
+      "step": 15203
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4460994358009895,
+      "learning_rate": 1.4863260462585715e-05,
+      "loss": 0.8438,
+      "step": 15204
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.573306677173679,
+      "learning_rate": 1.4862593720380886e-05,
+      "loss": 0.7963,
+      "step": 15205
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3578514056105726,
+      "learning_rate": 1.4861926949864502e-05,
+      "loss": 0.7895,
+      "step": 15206
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.951481272054683,
+      "learning_rate": 1.486126015104044e-05,
+      "loss": 0.7037,
+      "step": 15207
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.5590496970593417,
+      "learning_rate": 1.486059332391258e-05,
+      "loss": 0.8424,
+      "step": 15208
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5396957956686235,
+      "learning_rate": 1.4859926468484807e-05,
+      "loss": 0.8383,
+      "step": 15209
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4320203209333053,
+      "learning_rate": 1.4859259584761002e-05,
+      "loss": 0.6559,
+      "step": 15210
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6408205473984045,
+      "learning_rate": 1.4858592672745053e-05,
+      "loss": 0.7995,
+      "step": 15211
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6553654497448829,
+      "learning_rate": 1.4857925732440839e-05,
+      "loss": 0.7405,
+      "step": 15212
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3144994753406882,
+      "learning_rate": 1.4857258763852241e-05,
+      "loss": 0.8871,
+      "step": 15213
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5316696442417133,
+      "learning_rate": 1.4856591766983145e-05,
+      "loss": 0.798,
+      "step": 15214
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5508510134529898,
+      "learning_rate": 1.4855924741837435e-05,
+      "loss": 0.686,
+      "step": 15215
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3500545455053656,
+      "learning_rate": 1.4855257688418991e-05,
+      "loss": 0.728,
+      "step": 15216
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.614321651868798,
+      "learning_rate": 1.4854590606731702e-05,
+      "loss": 0.8026,
+      "step": 15217
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4042832441432747,
+      "learning_rate": 1.485392349677945e-05,
+      "loss": 0.7621,
+      "step": 15218
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4867570312237348,
+      "learning_rate": 1.485325635856612e-05,
+      "loss": 0.7054,
+      "step": 15219
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3187160708986276,
+      "learning_rate": 1.4852589192095592e-05,
+      "loss": 0.6797,
+      "step": 15220
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4663012530216613,
+      "learning_rate": 1.4851921997371756e-05,
+      "loss": 0.7934,
+      "step": 15221
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.734424858484805,
+      "learning_rate": 1.4851254774398492e-05,
+      "loss": 0.8441,
+      "step": 15222
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.813980523931483,
+      "learning_rate": 1.4850587523179689e-05,
+      "loss": 0.7388,
+      "step": 15223
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.237088563799239,
+      "learning_rate": 1.4849920243719227e-05,
+      "loss": 0.7903,
+      "step": 15224
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.513184378275017,
+      "learning_rate": 1.4849252936020998e-05,
+      "loss": 0.7853,
+      "step": 15225
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4001026732467359,
+      "learning_rate": 1.4848585600088876e-05,
+      "loss": 0.7052,
+      "step": 15226
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7521029501846637,
+      "learning_rate": 1.484791823592676e-05,
+      "loss": 0.7336,
+      "step": 15227
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.926825108731355,
+      "learning_rate": 1.4847250843538527e-05,
+      "loss": 0.6829,
+      "step": 15228
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7763259613412556,
+      "learning_rate": 1.4846583422928067e-05,
+      "loss": 0.8161,
+      "step": 15229
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5856372544161954,
+      "learning_rate": 1.484591597409926e-05,
+      "loss": 0.718,
+      "step": 15230
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5945963055292398,
+      "learning_rate": 1.4845248497055997e-05,
+      "loss": 0.5614,
+      "step": 15231
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6471633214353376,
+      "learning_rate": 1.4844580991802166e-05,
+      "loss": 0.7288,
+      "step": 15232
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0904240173625417,
+      "learning_rate": 1.4843913458341646e-05,
+      "loss": 0.8277,
+      "step": 15233
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6069484063664322,
+      "learning_rate": 1.4843245896678333e-05,
+      "loss": 0.5011,
+      "step": 15234
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.39744110705851,
+      "learning_rate": 1.4842578306816108e-05,
+      "loss": 0.7426,
+      "step": 15235
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6012039473390405,
+      "learning_rate": 1.4841910688758858e-05,
+      "loss": 0.7946,
+      "step": 15236
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2364988360886964,
+      "learning_rate": 1.4841243042510472e-05,
+      "loss": 0.7741,
+      "step": 15237
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4231658288744127,
+      "learning_rate": 1.4840575368074835e-05,
+      "loss": 0.7346,
+      "step": 15238
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5230893646071715,
+      "learning_rate": 1.4839907665455837e-05,
+      "loss": 0.8721,
+      "step": 15239
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.555965108676689,
+      "learning_rate": 1.4839239934657364e-05,
+      "loss": 0.7173,
+      "step": 15240
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4570265380794212,
+      "learning_rate": 1.4838572175683304e-05,
+      "loss": 0.6497,
+      "step": 15241
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8073920781763702,
+      "learning_rate": 1.4837904388537545e-05,
+      "loss": 0.7206,
+      "step": 15242
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4012354473117579,
+      "learning_rate": 1.4837236573223973e-05,
+      "loss": 0.8034,
+      "step": 15243
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.422385140210773,
+      "learning_rate": 1.483656872974648e-05,
+      "loss": 0.625,
+      "step": 15244
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.439332680229901,
+      "learning_rate": 1.4835900858108951e-05,
+      "loss": 0.8525,
+      "step": 15245
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5071469107143167,
+      "learning_rate": 1.4835232958315278e-05,
+      "loss": 0.7769,
+      "step": 15246
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7335426034463408,
+      "learning_rate": 1.4834565030369348e-05,
+      "loss": 0.7863,
+      "step": 15247
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.634434198452841,
+      "learning_rate": 1.4833897074275049e-05,
+      "loss": 0.7954,
+      "step": 15248
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8270236269966065,
+      "learning_rate": 1.4833229090036273e-05,
+      "loss": 0.9209,
+      "step": 15249
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7326129942020048,
+      "learning_rate": 1.4832561077656905e-05,
+      "loss": 0.744,
+      "step": 15250
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3415291826701277,
+      "learning_rate": 1.4831893037140836e-05,
+      "loss": 0.8478,
+      "step": 15251
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4700987217542139,
+      "learning_rate": 1.4831224968491955e-05,
+      "loss": 0.7277,
+      "step": 15252
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.372343814593351,
+      "learning_rate": 1.4830556871714157e-05,
+      "loss": 0.6799,
+      "step": 15253
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5215218199713327,
+      "learning_rate": 1.4829888746811323e-05,
+      "loss": 0.6621,
+      "step": 15254
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.377429902906974,
+      "learning_rate": 1.482922059378735e-05,
+      "loss": 0.7687,
+      "step": 15255
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3725731221515436,
+      "learning_rate": 1.4828552412646126e-05,
+      "loss": 0.7738,
+      "step": 15256
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9595105399120298,
+      "learning_rate": 1.482788420339154e-05,
+      "loss": 0.7774,
+      "step": 15257
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.424012402675065,
+      "learning_rate": 1.4827215966027483e-05,
+      "loss": 0.79,
+      "step": 15258
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6717528726202409,
+      "learning_rate": 1.482654770055785e-05,
+      "loss": 0.752,
+      "step": 15259
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6417910974313492,
+      "learning_rate": 1.4825879406986525e-05,
+      "loss": 0.7572,
+      "step": 15260
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3011675884634446,
+      "learning_rate": 1.4825211085317404e-05,
+      "loss": 0.7383,
+      "step": 15261
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4168695350882243,
+      "learning_rate": 1.4824542735554375e-05,
+      "loss": 0.7496,
+      "step": 15262
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8970173108952495,
+      "learning_rate": 1.482387435770133e-05,
+      "loss": 0.7969,
+      "step": 15263
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6767914664218073,
+      "learning_rate": 1.4823205951762164e-05,
+      "loss": 0.7772,
+      "step": 15264
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1779720839327315,
+      "learning_rate": 1.4822537517740763e-05,
+      "loss": 0.8115,
+      "step": 15265
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5233730778231798,
+      "learning_rate": 1.4821869055641023e-05,
+      "loss": 0.7539,
+      "step": 15266
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.659173440893403,
+      "learning_rate": 1.4821200565466835e-05,
+      "loss": 0.8075,
+      "step": 15267
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.593639098595439,
+      "learning_rate": 1.4820532047222091e-05,
+      "loss": 0.7595,
+      "step": 15268
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2883356715970606,
+      "learning_rate": 1.4819863500910683e-05,
+      "loss": 0.8017,
+      "step": 15269
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5124774949621054,
+      "learning_rate": 1.4819194926536504e-05,
+      "loss": 0.7184,
+      "step": 15270
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.589763637224613,
+      "learning_rate": 1.4818526324103445e-05,
+      "loss": 0.6528,
+      "step": 15271
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0456463977184542,
+      "learning_rate": 1.4817857693615406e-05,
+      "loss": 0.9195,
+      "step": 15272
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.415521054325461,
+      "learning_rate": 1.4817189035076265e-05,
+      "loss": 0.6682,
+      "step": 15273
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.607211003784027,
+      "learning_rate": 1.4816520348489932e-05,
+      "loss": 0.5521,
+      "step": 15274
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4638381865608843,
+      "learning_rate": 1.4815851633860288e-05,
+      "loss": 0.7124,
+      "step": 15275
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8218765673011599,
+      "learning_rate": 1.4815182891191234e-05,
+      "loss": 0.6826,
+      "step": 15276
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.532563887145021,
+      "learning_rate": 1.4814514120486657e-05,
+      "loss": 0.8413,
+      "step": 15277
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5459290571627098,
+      "learning_rate": 1.481384532175046e-05,
+      "loss": 0.7334,
+      "step": 15278
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.534143676481629,
+      "learning_rate": 1.481317649498653e-05,
+      "loss": 0.749,
+      "step": 15279
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3355821940039885,
+      "learning_rate": 1.481250764019876e-05,
+      "loss": 0.9101,
+      "step": 15280
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0307325038943285,
+      "learning_rate": 1.4811838757391049e-05,
+      "loss": 0.8158,
+      "step": 15281
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.37597556079049,
+      "learning_rate": 1.4811169846567286e-05,
+      "loss": 0.7696,
+      "step": 15282
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6809104816138063,
+      "learning_rate": 1.4810500907731374e-05,
+      "loss": 0.6399,
+      "step": 15283
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9603367122668887,
+      "learning_rate": 1.4809831940887196e-05,
+      "loss": 0.7562,
+      "step": 15284
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4327919961914575,
+      "learning_rate": 1.4809162946038659e-05,
+      "loss": 0.6923,
+      "step": 15285
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.399031351593125,
+      "learning_rate": 1.4808493923189652e-05,
+      "loss": 0.7495,
+      "step": 15286
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3821916615597796,
+      "learning_rate": 1.4807824872344072e-05,
+      "loss": 0.7827,
+      "step": 15287
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3660054994282813,
+      "learning_rate": 1.4807155793505811e-05,
+      "loss": 0.7706,
+      "step": 15288
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9539765749201943,
+      "learning_rate": 1.480648668667877e-05,
+      "loss": 0.641,
+      "step": 15289
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3601003161717902,
+      "learning_rate": 1.4805817551866839e-05,
+      "loss": 0.869,
+      "step": 15290
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9486177730161547,
+      "learning_rate": 1.4805148389073918e-05,
+      "loss": 0.7637,
+      "step": 15291
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5500098080525893,
+      "learning_rate": 1.48044791983039e-05,
+      "loss": 0.6665,
+      "step": 15292
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.479152211355805,
+      "learning_rate": 1.4803809979560688e-05,
+      "loss": 0.8011,
+      "step": 15293
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6319017326833287,
+      "learning_rate": 1.480314073284817e-05,
+      "loss": 0.8035,
+      "step": 15294
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1302302961297954,
+      "learning_rate": 1.4802471458170245e-05,
+      "loss": 0.6666,
+      "step": 15295
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.459582090156023,
+      "learning_rate": 1.480180215553081e-05,
+      "loss": 0.8829,
+      "step": 15296
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4093077966354999,
+      "learning_rate": 1.4801132824933767e-05,
+      "loss": 0.7137,
+      "step": 15297
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9145719254363358,
+      "learning_rate": 1.4800463466383004e-05,
+      "loss": 0.7994,
+      "step": 15298
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5798773565073179,
+      "learning_rate": 1.4799794079882426e-05,
+      "loss": 0.7396,
+      "step": 15299
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6025669308252368,
+      "learning_rate": 1.4799124665435927e-05,
+      "loss": 0.7677,
+      "step": 15300
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.702279366044118,
+      "learning_rate": 1.4798455223047401e-05,
+      "loss": 0.697,
+      "step": 15301
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.483221906800703,
+      "learning_rate": 1.4797785752720755e-05,
+      "loss": 0.7458,
+      "step": 15302
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4660106069080523,
+      "learning_rate": 1.4797116254459878e-05,
+      "loss": 0.6991,
+      "step": 15303
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3342284773209716,
+      "learning_rate": 1.4796446728268674e-05,
+      "loss": 0.7101,
+      "step": 15304
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7595344483746966,
+      "learning_rate": 1.4795777174151033e-05,
+      "loss": 0.8231,
+      "step": 15305
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4886960023251974,
+      "learning_rate": 1.4795107592110866e-05,
+      "loss": 0.8173,
+      "step": 15306
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4812218130151702,
+      "learning_rate": 1.4794437982152061e-05,
+      "loss": 0.5866,
+      "step": 15307
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6762523045423479,
+      "learning_rate": 1.4793768344278521e-05,
+      "loss": 0.8809,
+      "step": 15308
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4200384604858771,
+      "learning_rate": 1.4793098678494145e-05,
+      "loss": 0.7531,
+      "step": 15309
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.742317616824915,
+      "learning_rate": 1.479242898480283e-05,
+      "loss": 0.79,
+      "step": 15310
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.520714689852106,
+      "learning_rate": 1.4791759263208474e-05,
+      "loss": 0.8515,
+      "step": 15311
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5171414135622578,
+      "learning_rate": 1.479108951371498e-05,
+      "loss": 0.731,
+      "step": 15312
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6019313146087049,
+      "learning_rate": 1.4790419736326248e-05,
+      "loss": 0.8136,
+      "step": 15313
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.59069798813789,
+      "learning_rate": 1.4789749931046173e-05,
+      "loss": 0.5333,
+      "step": 15314
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9430492370878223,
+      "learning_rate": 1.4789080097878659e-05,
+      "loss": 0.7232,
+      "step": 15315
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.218985300357846,
+      "learning_rate": 1.4788410236827605e-05,
+      "loss": 0.7068,
+      "step": 15316
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2591987755592904,
+      "learning_rate": 1.4787740347896908e-05,
+      "loss": 0.8048,
+      "step": 15317
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.324834377911341,
+      "learning_rate": 1.4787070431090474e-05,
+      "loss": 0.7148,
+      "step": 15318
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5139411177289674,
+      "learning_rate": 1.4786400486412198e-05,
+      "loss": 0.6268,
+      "step": 15319
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.396489908415412,
+      "learning_rate": 1.4785730513865983e-05,
+      "loss": 0.7119,
+      "step": 15320
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.785822487386636,
+      "learning_rate": 1.4785060513455732e-05,
+      "loss": 0.6492,
+      "step": 15321
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0740822643429944,
+      "learning_rate": 1.4784390485185342e-05,
+      "loss": 0.7867,
+      "step": 15322
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3751820162940287,
+      "learning_rate": 1.4783720429058719e-05,
+      "loss": 0.7273,
+      "step": 15323
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4084959896524243,
+      "learning_rate": 1.4783050345079755e-05,
+      "loss": 0.8393,
+      "step": 15324
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5699120268209688,
+      "learning_rate": 1.4782380233252365e-05,
+      "loss": 0.5943,
+      "step": 15325
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4297435561635905,
+      "learning_rate": 1.4781710093580437e-05,
+      "loss": 0.6648,
+      "step": 15326
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.461617912342446,
+      "learning_rate": 1.4781039926067884e-05,
+      "loss": 0.7996,
+      "step": 15327
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7856549756906694,
+      "learning_rate": 1.4780369730718595e-05,
+      "loss": 0.7698,
+      "step": 15328
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5597167782145644,
+      "learning_rate": 1.477969950753649e-05,
+      "loss": 0.6571,
+      "step": 15329
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4508375946189376,
+      "learning_rate": 1.4779029256525457e-05,
+      "loss": 0.6763,
+      "step": 15330
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3609797723982748,
+      "learning_rate": 1.47783589776894e-05,
+      "loss": 0.6872,
+      "step": 15331
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3089728688841884,
+      "learning_rate": 1.4777688671032226e-05,
+      "loss": 0.783,
+      "step": 15332
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3031110425770764,
+      "learning_rate": 1.4777018336557835e-05,
+      "loss": 0.7511,
+      "step": 15333
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6555001750770246,
+      "learning_rate": 1.4776347974270135e-05,
+      "loss": 0.7189,
+      "step": 15334
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4954178332338068,
+      "learning_rate": 1.477567758417302e-05,
+      "loss": 0.7157,
+      "step": 15335
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.6892151614084994,
+      "learning_rate": 1.4775007166270402e-05,
+      "loss": 0.8101,
+      "step": 15336
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3514919950179904,
+      "learning_rate": 1.4774336720566178e-05,
+      "loss": 0.7068,
+      "step": 15337
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8151874911478092,
+      "learning_rate": 1.4773666247064255e-05,
+      "loss": 0.7579,
+      "step": 15338
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5827802515687875,
+      "learning_rate": 1.4772995745768538e-05,
+      "loss": 0.5581,
+      "step": 15339
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.246792193403192,
+      "learning_rate": 1.4772325216682926e-05,
+      "loss": 0.7246,
+      "step": 15340
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5788424724855759,
+      "learning_rate": 1.4771654659811326e-05,
+      "loss": 0.7568,
+      "step": 15341
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8697758766699721,
+      "learning_rate": 1.4770984075157641e-05,
+      "loss": 0.8048,
+      "step": 15342
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3557051084665346,
+      "learning_rate": 1.4770313462725777e-05,
+      "loss": 0.7191,
+      "step": 15343
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3962170506437934,
+      "learning_rate": 1.476964282251964e-05,
+      "loss": 0.7162,
+      "step": 15344
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5518862416815933,
+      "learning_rate": 1.4768972154543129e-05,
+      "loss": 0.7077,
+      "step": 15345
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4705083212400822,
+      "learning_rate": 1.4768301458800154e-05,
+      "loss": 0.6772,
+      "step": 15346
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3018291767098158,
+      "learning_rate": 1.4767630735294618e-05,
+      "loss": 0.5936,
+      "step": 15347
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5913089093125774,
+      "learning_rate": 1.4766959984030425e-05,
+      "loss": 0.7143,
+      "step": 15348
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5798680745204,
+      "learning_rate": 1.4766289205011483e-05,
+      "loss": 0.645,
+      "step": 15349
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.300656593341598,
+      "learning_rate": 1.4765618398241697e-05,
+      "loss": 0.7748,
+      "step": 15350
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6625683272676506,
+      "learning_rate": 1.4764947563724974e-05,
+      "loss": 0.7702,
+      "step": 15351
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6392316774935607,
+      "learning_rate": 1.4764276701465212e-05,
+      "loss": 0.7672,
+      "step": 15352
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5226082698414862,
+      "learning_rate": 1.4763605811466328e-05,
+      "loss": 0.7465,
+      "step": 15353
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8106925249943264,
+      "learning_rate": 1.4762934893732218e-05,
+      "loss": 0.5642,
+      "step": 15354
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5149266269605257,
+      "learning_rate": 1.47622639482668e-05,
+      "loss": 0.65,
+      "step": 15355
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6454614495916167,
+      "learning_rate": 1.4761592975073965e-05,
+      "loss": 0.5191,
+      "step": 15356
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6916105024168124,
+      "learning_rate": 1.4760921974157636e-05,
+      "loss": 0.5387,
+      "step": 15357
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4309087195425594,
+      "learning_rate": 1.4760250945521708e-05,
+      "loss": 0.6732,
+      "step": 15358
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2727189137308004,
+      "learning_rate": 1.4759579889170093e-05,
+      "loss": 0.8291,
+      "step": 15359
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.300558884681668,
+      "learning_rate": 1.4758908805106696e-05,
+      "loss": 0.7911,
+      "step": 15360
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9388464938175318,
+      "learning_rate": 1.4758237693335424e-05,
+      "loss": 0.8759,
+      "step": 15361
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0392168283836876,
+      "learning_rate": 1.475756655386019e-05,
+      "loss": 0.7825,
+      "step": 15362
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.855460493263897,
+      "learning_rate": 1.4756895386684893e-05,
+      "loss": 0.655,
+      "step": 15363
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5676954474577902,
+      "learning_rate": 1.4756224191813449e-05,
+      "loss": 0.52,
+      "step": 15364
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4181939676760205,
+      "learning_rate": 1.4755552969249758e-05,
+      "loss": 0.9059,
+      "step": 15365
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6651143241258635,
+      "learning_rate": 1.4754881718997735e-05,
+      "loss": 0.7874,
+      "step": 15366
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4495850258275067,
+      "learning_rate": 1.4754210441061283e-05,
+      "loss": 0.7068,
+      "step": 15367
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5743418455841531,
+      "learning_rate": 1.4753539135444314e-05,
+      "loss": 0.7724,
+      "step": 15368
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.341809321164276,
+      "learning_rate": 1.4752867802150735e-05,
+      "loss": 0.798,
+      "step": 15369
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.523737794653212,
+      "learning_rate": 1.4752196441184454e-05,
+      "loss": 0.7305,
+      "step": 15370
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3729533829703207,
+      "learning_rate": 1.4751525052549383e-05,
+      "loss": 0.6825,
+      "step": 15371
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5644754288190925,
+      "learning_rate": 1.475085363624943e-05,
+      "loss": 0.7003,
+      "step": 15372
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.511475142196721,
+      "learning_rate": 1.4750182192288497e-05,
+      "loss": 0.7755,
+      "step": 15373
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0532187397887744,
+      "learning_rate": 1.4749510720670506e-05,
+      "loss": 0.8134,
+      "step": 15374
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.579598603240891,
+      "learning_rate": 1.4748839221399354e-05,
+      "loss": 0.8089,
+      "step": 15375
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6199047421500168,
+      "learning_rate": 1.4748167694478963e-05,
+      "loss": 0.7988,
+      "step": 15376
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7292438895996065,
+      "learning_rate": 1.4747496139913229e-05,
+      "loss": 0.6446,
+      "step": 15377
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.724749534858197,
+      "learning_rate": 1.4746824557706076e-05,
+      "loss": 0.7492,
+      "step": 15378
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6388952046329592,
+      "learning_rate": 1.4746152947861405e-05,
+      "loss": 0.5709,
+      "step": 15379
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.735295473505431,
+      "learning_rate": 1.4745481310383127e-05,
+      "loss": 0.7066,
+      "step": 15380
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5870065101324804,
+      "learning_rate": 1.4744809645275157e-05,
+      "loss": 0.731,
+      "step": 15381
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.407851082304171,
+      "learning_rate": 1.4744137952541402e-05,
+      "loss": 0.6935,
+      "step": 15382
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.594356314273652,
+      "learning_rate": 1.4743466232185775e-05,
+      "loss": 0.5531,
+      "step": 15383
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3742014054755767,
+      "learning_rate": 1.4742794484212184e-05,
+      "loss": 0.8456,
+      "step": 15384
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.477857904034656,
+      "learning_rate": 1.4742122708624544e-05,
+      "loss": 0.7688,
+      "step": 15385
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4615922443759648,
+      "learning_rate": 1.4741450905426763e-05,
+      "loss": 0.5819,
+      "step": 15386
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4360244587977622,
+      "learning_rate": 1.4740779074622756e-05,
+      "loss": 0.7448,
+      "step": 15387
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5436623620893368,
+      "learning_rate": 1.4740107216216428e-05,
+      "loss": 0.6414,
+      "step": 15388
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4725509551840774,
+      "learning_rate": 1.4739435330211697e-05,
+      "loss": 0.7427,
+      "step": 15389
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5039655580743316,
+      "learning_rate": 1.4738763416612476e-05,
+      "loss": 0.759,
+      "step": 15390
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.559427879129628,
+      "learning_rate": 1.473809147542267e-05,
+      "loss": 0.7882,
+      "step": 15391
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6573150933693117,
+      "learning_rate": 1.4737419506646197e-05,
+      "loss": 0.7432,
+      "step": 15392
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6039316741364986,
+      "learning_rate": 1.4736747510286968e-05,
+      "loss": 0.752,
+      "step": 15393
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.536609960397284,
+      "learning_rate": 1.4736075486348896e-05,
+      "loss": 0.7497,
+      "step": 15394
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4321636409080063,
+      "learning_rate": 1.4735403434835891e-05,
+      "loss": 0.8394,
+      "step": 15395
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4166290914172746,
+      "learning_rate": 1.473473135575187e-05,
+      "loss": 0.7826,
+      "step": 15396
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7718287912901531,
+      "learning_rate": 1.4734059249100743e-05,
+      "loss": 0.7187,
+      "step": 15397
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7771581666778276,
+      "learning_rate": 1.4733387114886424e-05,
+      "loss": 0.6926,
+      "step": 15398
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3442805498049453,
+      "learning_rate": 1.4732714953112828e-05,
+      "loss": 0.6899,
+      "step": 15399
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6538308679806992,
+      "learning_rate": 1.4732042763783866e-05,
+      "loss": 0.6102,
+      "step": 15400
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.10453448150054,
+      "learning_rate": 1.4731370546903454e-05,
+      "loss": 0.7829,
+      "step": 15401
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6842938142509907,
+      "learning_rate": 1.4730698302475505e-05,
+      "loss": 0.7951,
+      "step": 15402
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3559143713715474,
+      "learning_rate": 1.473002603050393e-05,
+      "loss": 0.7013,
+      "step": 15403
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.434271604128991,
+      "learning_rate": 1.472935373099265e-05,
+      "loss": 0.7244,
+      "step": 15404
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4626242211593843,
+      "learning_rate": 1.472868140394557e-05,
+      "loss": 0.7191,
+      "step": 15405
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6214380726985302,
+      "learning_rate": 1.4728009049366617e-05,
+      "loss": 0.5426,
+      "step": 15406
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3305690869033129,
+      "learning_rate": 1.472733666725969e-05,
+      "loss": 0.7962,
+      "step": 15407
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4288149351930826,
+      "learning_rate": 1.472666425762872e-05,
+      "loss": 0.7752,
+      "step": 15408
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6249641197911845,
+      "learning_rate": 1.4725991820477612e-05,
+      "loss": 0.5729,
+      "step": 15409
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0236634032041714,
+      "learning_rate": 1.4725319355810282e-05,
+      "loss": 0.7521,
+      "step": 15410
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4355859499149772,
+      "learning_rate": 1.4724646863630646e-05,
+      "loss": 0.6629,
+      "step": 15411
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2278868679730606,
+      "learning_rate": 1.472397434394262e-05,
+      "loss": 0.7119,
+      "step": 15412
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5447527102715135,
+      "learning_rate": 1.4723301796750122e-05,
+      "loss": 0.824,
+      "step": 15413
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5191490980770241,
+      "learning_rate": 1.4722629222057063e-05,
+      "loss": 0.7362,
+      "step": 15414
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4975142584709722,
+      "learning_rate": 1.4721956619867364e-05,
+      "loss": 0.882,
+      "step": 15415
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4612946040304147,
+      "learning_rate": 1.4721283990184938e-05,
+      "loss": 0.8328,
+      "step": 15416
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.256245487591775,
+      "learning_rate": 1.4720611333013701e-05,
+      "loss": 0.8373,
+      "step": 15417
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.412262475647355,
+      "learning_rate": 1.471993864835757e-05,
+      "loss": 0.781,
+      "step": 15418
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3620128161904315,
+      "learning_rate": 1.4719265936220464e-05,
+      "loss": 0.7246,
+      "step": 15419
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.31385566916557,
+      "learning_rate": 1.4718593196606294e-05,
+      "loss": 0.6798,
+      "step": 15420
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5916563401320365,
+      "learning_rate": 1.4717920429518984e-05,
+      "loss": 0.7973,
+      "step": 15421
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.54340489848978,
+      "learning_rate": 1.4717247634962446e-05,
+      "loss": 0.7688,
+      "step": 15422
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3798062056762423,
+      "learning_rate": 1.4716574812940603e-05,
+      "loss": 0.6342,
+      "step": 15423
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7088625087202458,
+      "learning_rate": 1.4715901963457362e-05,
+      "loss": 0.7124,
+      "step": 15424
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9797427332507884,
+      "learning_rate": 1.4715229086516649e-05,
+      "loss": 0.7578,
+      "step": 15425
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6359850784461218,
+      "learning_rate": 1.471455618212238e-05,
+      "loss": 0.8294,
+      "step": 15426
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.624725284340679,
+      "learning_rate": 1.4713883250278471e-05,
+      "loss": 0.6442,
+      "step": 15427
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.290708921825923,
+      "learning_rate": 1.4713210290988842e-05,
+      "loss": 0.7942,
+      "step": 15428
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6006407295021209,
+      "learning_rate": 1.471253730425741e-05,
+      "loss": 0.5753,
+      "step": 15429
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5494118300846749,
+      "learning_rate": 1.4711864290088099e-05,
+      "loss": 0.6798,
+      "step": 15430
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5433102067786704,
+      "learning_rate": 1.4711191248484812e-05,
+      "loss": 0.7291,
+      "step": 15431
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.316716594021252,
+      "learning_rate": 1.4710518179451487e-05,
+      "loss": 0.7038,
+      "step": 15432
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3068890817268253,
+      "learning_rate": 1.470984508299203e-05,
+      "loss": 0.7137,
+      "step": 15433
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3354451179659366,
+      "learning_rate": 1.4709171959110367e-05,
+      "loss": 0.868,
+      "step": 15434
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2089709973483427,
+      "learning_rate": 1.470849880781041e-05,
+      "loss": 0.7465,
+      "step": 15435
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6216918304967587,
+      "learning_rate": 1.4707825629096085e-05,
+      "loss": 0.8714,
+      "step": 15436
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5389153533027802,
+      "learning_rate": 1.4707152422971307e-05,
+      "loss": 0.7074,
+      "step": 15437
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.434775047463362,
+      "learning_rate": 1.4706479189439998e-05,
+      "loss": 0.7447,
+      "step": 15438
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.27022651078529,
+      "learning_rate": 1.4705805928506076e-05,
+      "loss": 0.6748,
+      "step": 15439
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.390442910988789,
+      "learning_rate": 1.4705132640173463e-05,
+      "loss": 0.786,
+      "step": 15440
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1824513502971654,
+      "learning_rate": 1.4704459324446078e-05,
+      "loss": 0.8015,
+      "step": 15441
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5572908508195156,
+      "learning_rate": 1.4703785981327841e-05,
+      "loss": 0.5035,
+      "step": 15442
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6009429131577381,
+      "learning_rate": 1.4703112610822674e-05,
+      "loss": 0.5005,
+      "step": 15443
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5734701931820394,
+      "learning_rate": 1.4702439212934496e-05,
+      "loss": 0.5003,
+      "step": 15444
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5999958424207406,
+      "learning_rate": 1.4701765787667228e-05,
+      "loss": 0.8079,
+      "step": 15445
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2656544901237265,
+      "learning_rate": 1.4701092335024792e-05,
+      "loss": 0.8379,
+      "step": 15446
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.361769041329442,
+      "learning_rate": 1.4700418855011105e-05,
+      "loss": 0.6889,
+      "step": 15447
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.4223048233565425,
+      "learning_rate": 1.4699745347630095e-05,
+      "loss": 0.6703,
+      "step": 15448
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.451417398191424,
+      "learning_rate": 1.4699071812885677e-05,
+      "loss": 0.7178,
+      "step": 15449
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.748978880105901,
+      "learning_rate": 1.4698398250781776e-05,
+      "loss": 0.7838,
+      "step": 15450
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3485493440352714,
+      "learning_rate": 1.4697724661322312e-05,
+      "loss": 0.8089,
+      "step": 15451
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4801736191661237,
+      "learning_rate": 1.469705104451121e-05,
+      "loss": 0.6314,
+      "step": 15452
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5864631116270559,
+      "learning_rate": 1.469637740035239e-05,
+      "loss": 0.5536,
+      "step": 15453
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5963487538427454,
+      "learning_rate": 1.4695703728849771e-05,
+      "loss": 0.5165,
+      "step": 15454
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3883288106596259,
+      "learning_rate": 1.4695030030007282e-05,
+      "loss": 0.7841,
+      "step": 15455
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4408845287019874,
+      "learning_rate": 1.4694356303828837e-05,
+      "loss": 0.8027,
+      "step": 15456
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.414123248753446,
+      "learning_rate": 1.4693682550318368e-05,
+      "loss": 0.6782,
+      "step": 15457
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9423131733081551,
+      "learning_rate": 1.4693008769479788e-05,
+      "loss": 0.7878,
+      "step": 15458
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7836129490664256,
+      "learning_rate": 1.4692334961317033e-05,
+      "loss": 0.7389,
+      "step": 15459
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0275557642564164,
+      "learning_rate": 1.4691661125834012e-05,
+      "loss": 0.7708,
+      "step": 15460
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6039164390296363,
+      "learning_rate": 1.4690987263034655e-05,
+      "loss": 0.7294,
+      "step": 15461
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.412832189566493,
+      "learning_rate": 1.4690313372922887e-05,
+      "loss": 0.7535,
+      "step": 15462
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5852939604699035,
+      "learning_rate": 1.4689639455502629e-05,
+      "loss": 0.6994,
+      "step": 15463
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4470222165354634,
+      "learning_rate": 1.4688965510777803e-05,
+      "loss": 0.7672,
+      "step": 15464
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3610064908705635,
+      "learning_rate": 1.468829153875234e-05,
+      "loss": 0.7344,
+      "step": 15465
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6360437023295239,
+      "learning_rate": 1.4687617539430154e-05,
+      "loss": 0.5312,
+      "step": 15466
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3660272214809843,
+      "learning_rate": 1.4686943512815179e-05,
+      "loss": 0.735,
+      "step": 15467
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5016391236256326,
+      "learning_rate": 1.4686269458911333e-05,
+      "loss": 0.8117,
+      "step": 15468
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4959897355301657,
+      "learning_rate": 1.4685595377722544e-05,
+      "loss": 0.7138,
+      "step": 15469
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3772004214236953,
+      "learning_rate": 1.4684921269252734e-05,
+      "loss": 0.7341,
+      "step": 15470
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4899381094699686,
+      "learning_rate": 1.468424713350583e-05,
+      "loss": 0.6419,
+      "step": 15471
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7422602700215621,
+      "learning_rate": 1.4683572970485756e-05,
+      "loss": 0.714,
+      "step": 15472
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.387249547151763,
+      "learning_rate": 1.4682898780196434e-05,
+      "loss": 0.7048,
+      "step": 15473
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5939888424032322,
+      "learning_rate": 1.4682224562641798e-05,
+      "loss": 0.6385,
+      "step": 15474
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5771548662058585,
+      "learning_rate": 1.4681550317825766e-05,
+      "loss": 0.517,
+      "step": 15475
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5206698186919916,
+      "learning_rate": 1.4680876045752266e-05,
+      "loss": 0.7346,
+      "step": 15476
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6735349079298734,
+      "learning_rate": 1.4680201746425221e-05,
+      "loss": 0.7548,
+      "step": 15477
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5534930919960397,
+      "learning_rate": 1.4679527419848563e-05,
+      "loss": 0.8052,
+      "step": 15478
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.459579891665184,
+      "learning_rate": 1.4678853066026216e-05,
+      "loss": 0.7319,
+      "step": 15479
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5674876216653686,
+      "learning_rate": 1.4678178684962101e-05,
+      "loss": 0.8875,
+      "step": 15480
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3809407785241266,
+      "learning_rate": 1.4677504276660153e-05,
+      "loss": 0.629,
+      "step": 15481
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2987781383505723,
+      "learning_rate": 1.4676829841124288e-05,
+      "loss": 0.7584,
+      "step": 15482
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5110940720400372,
+      "learning_rate": 1.4676155378358447e-05,
+      "loss": 0.8375,
+      "step": 15483
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3734360964486962,
+      "learning_rate": 1.4675480888366545e-05,
+      "loss": 0.7148,
+      "step": 15484
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8735981598322509,
+      "learning_rate": 1.4674806371152514e-05,
+      "loss": 0.7806,
+      "step": 15485
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6035183196929691,
+      "learning_rate": 1.4674131826720278e-05,
+      "loss": 0.5698,
+      "step": 15486
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0889459286978025,
+      "learning_rate": 1.4673457255073769e-05,
+      "loss": 0.8252,
+      "step": 15487
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6043746477871585,
+      "learning_rate": 1.4672782656216911e-05,
+      "loss": 0.8097,
+      "step": 15488
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.268107814334448,
+      "learning_rate": 1.4672108030153634e-05,
+      "loss": 0.7227,
+      "step": 15489
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4221155135165773,
+      "learning_rate": 1.4671433376887864e-05,
+      "loss": 0.8234,
+      "step": 15490
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0425275451329092,
+      "learning_rate": 1.467075869642353e-05,
+      "loss": 0.7379,
+      "step": 15491
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.744447902802504,
+      "learning_rate": 1.4670083988764559e-05,
+      "loss": 0.8146,
+      "step": 15492
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5602932295549297,
+      "learning_rate": 1.4669409253914882e-05,
+      "loss": 0.6879,
+      "step": 15493
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3559683575554653,
+      "learning_rate": 1.4668734491878426e-05,
+      "loss": 0.7558,
+      "step": 15494
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5378369283440112,
+      "learning_rate": 1.4668059702659121e-05,
+      "loss": 0.8941,
+      "step": 15495
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0548721376582693,
+      "learning_rate": 1.4667384886260892e-05,
+      "loss": 0.7595,
+      "step": 15496
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1045907988409924,
+      "learning_rate": 1.466671004268767e-05,
+      "loss": 0.7702,
+      "step": 15497
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.348005303887969,
+      "learning_rate": 1.4666035171943388e-05,
+      "loss": 0.7451,
+      "step": 15498
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0767126282901507,
+      "learning_rate": 1.466536027403197e-05,
+      "loss": 0.7517,
+      "step": 15499
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.51500536666294,
+      "learning_rate": 1.4664685348957348e-05,
+      "loss": 0.7278,
+      "step": 15500
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.596379193840752,
+      "learning_rate": 1.4664010396723449e-05,
+      "loss": 0.7222,
+      "step": 15501
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2043991497224034,
+      "learning_rate": 1.466333541733421e-05,
+      "loss": 0.7825,
+      "step": 15502
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.603839152671876,
+      "learning_rate": 1.466266041079355e-05,
+      "loss": 0.5405,
+      "step": 15503
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.400285092912031,
+      "learning_rate": 1.466198537710541e-05,
+      "loss": 0.7774,
+      "step": 15504
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3116450625499185,
+      "learning_rate": 1.4661310316273709e-05,
+      "loss": 0.6468,
+      "step": 15505
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.390313952132271,
+      "learning_rate": 1.4660635228302388e-05,
+      "loss": 0.6499,
+      "step": 15506
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.365570439571303,
+      "learning_rate": 1.4659960113195371e-05,
+      "loss": 0.6769,
+      "step": 15507
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6963335632690835,
+      "learning_rate": 1.4659284970956593e-05,
+      "loss": 0.7055,
+      "step": 15508
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6148482930519927,
+      "learning_rate": 1.4658609801589982e-05,
+      "loss": 0.56,
+      "step": 15509
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6121856427556036,
+      "learning_rate": 1.465793460509947e-05,
+      "loss": 0.7961,
+      "step": 15510
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4375529397734839,
+      "learning_rate": 1.4657259381488988e-05,
+      "loss": 0.7916,
+      "step": 15511
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3651183832391076,
+      "learning_rate": 1.4656584130762467e-05,
+      "loss": 0.7099,
+      "step": 15512
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6113582832230688,
+      "learning_rate": 1.4655908852923841e-05,
+      "loss": 0.7772,
+      "step": 15513
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7051312184103626,
+      "learning_rate": 1.4655233547977037e-05,
+      "loss": 0.7451,
+      "step": 15514
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.380012324731544,
+      "learning_rate": 1.465455821592599e-05,
+      "loss": 0.7921,
+      "step": 15515
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5192499896429157,
+      "learning_rate": 1.465388285677463e-05,
+      "loss": 0.8019,
+      "step": 15516
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3190235838889606,
+      "learning_rate": 1.4653207470526893e-05,
+      "loss": 0.8237,
+      "step": 15517
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.587974026735041,
+      "learning_rate": 1.465253205718671e-05,
+      "loss": 0.7182,
+      "step": 15518
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2894882472901168,
+      "learning_rate": 1.4651856616758009e-05,
+      "loss": 0.65,
+      "step": 15519
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.396328771075864,
+      "learning_rate": 1.4651181149244728e-05,
+      "loss": 0.7079,
+      "step": 15520
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2804011827940098,
+      "learning_rate": 1.46505056546508e-05,
+      "loss": 0.7219,
+      "step": 15521
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7938809489715466,
+      "learning_rate": 1.4649830132980151e-05,
+      "loss": 0.6403,
+      "step": 15522
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4946912832537225,
+      "learning_rate": 1.4649154584236723e-05,
+      "loss": 0.7601,
+      "step": 15523
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5686253500967282,
+      "learning_rate": 1.464847900842444e-05,
+      "loss": 0.5101,
+      "step": 15524
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4703382355975827,
+      "learning_rate": 1.4647803405547248e-05,
+      "loss": 0.8348,
+      "step": 15525
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6059963867053055,
+      "learning_rate": 1.4647127775609067e-05,
+      "loss": 0.7452,
+      "step": 15526
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7554850715016409,
+      "learning_rate": 1.4646452118613838e-05,
+      "loss": 0.5342,
+      "step": 15527
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3554426807260571,
+      "learning_rate": 1.4645776434565492e-05,
+      "loss": 0.7005,
+      "step": 15528
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8566513609458515,
+      "learning_rate": 1.4645100723467969e-05,
+      "loss": 0.7617,
+      "step": 15529
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7112187620189323,
+      "learning_rate": 1.4644424985325198e-05,
+      "loss": 0.6567,
+      "step": 15530
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2348469226257,
+      "learning_rate": 1.4643749220141113e-05,
+      "loss": 0.7448,
+      "step": 15531
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5331986609188166,
+      "learning_rate": 1.4643073427919652e-05,
+      "loss": 0.571,
+      "step": 15532
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3844890404015844,
+      "learning_rate": 1.4642397608664744e-05,
+      "loss": 0.7661,
+      "step": 15533
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.505168009520566,
+      "learning_rate": 1.4641721762380332e-05,
+      "loss": 0.8241,
+      "step": 15534
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7028837483391817,
+      "learning_rate": 1.4641045889070343e-05,
+      "loss": 0.8738,
+      "step": 15535
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6328326084584803,
+      "learning_rate": 1.4640369988738717e-05,
+      "loss": 0.7389,
+      "step": 15536
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5505482483929067,
+      "learning_rate": 1.4639694061389384e-05,
+      "loss": 0.672,
+      "step": 15537
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4180305072668635,
+      "learning_rate": 1.4639018107026288e-05,
+      "loss": 0.6954,
+      "step": 15538
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5084171933408903,
+      "learning_rate": 1.4638342125653358e-05,
+      "loss": 0.9666,
+      "step": 15539
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5787052566073351,
+      "learning_rate": 1.4637666117274531e-05,
+      "loss": 0.5399,
+      "step": 15540
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4305595753123235,
+      "learning_rate": 1.4636990081893746e-05,
+      "loss": 0.9132,
+      "step": 15541
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4280987905012035,
+      "learning_rate": 1.4636314019514935e-05,
+      "loss": 0.7579,
+      "step": 15542
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6413138051378553,
+      "learning_rate": 1.4635637930142035e-05,
+      "loss": 0.912,
+      "step": 15543
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.939777104864941,
+      "learning_rate": 1.4634961813778982e-05,
+      "loss": 0.8272,
+      "step": 15544
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7004240295979958,
+      "learning_rate": 1.4634285670429717e-05,
+      "loss": 0.8375,
+      "step": 15545
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4633434944972337,
+      "learning_rate": 1.4633609500098174e-05,
+      "loss": 0.7363,
+      "step": 15546
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.684018564193405,
+      "learning_rate": 1.4632933302788289e-05,
+      "loss": 0.7086,
+      "step": 15547
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4006085395955035,
+      "learning_rate": 1.4632257078503997e-05,
+      "loss": 0.7687,
+      "step": 15548
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8085733689426298,
+      "learning_rate": 1.4631580827249242e-05,
+      "loss": 0.7184,
+      "step": 15549
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.894958184899455,
+      "learning_rate": 1.4630904549027954e-05,
+      "loss": 0.6773,
+      "step": 15550
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6003426835752441,
+      "learning_rate": 1.4630228243844073e-05,
+      "loss": 0.5486,
+      "step": 15551
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.39339941554798,
+      "learning_rate": 1.462955191170154e-05,
+      "loss": 0.5858,
+      "step": 15552
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.97484250697233,
+      "learning_rate": 1.4628875552604291e-05,
+      "loss": 0.7742,
+      "step": 15553
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.791457177678496,
+      "learning_rate": 1.4628199166556258e-05,
+      "loss": 0.6718,
+      "step": 15554
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5738387082461998,
+      "learning_rate": 1.462752275356139e-05,
+      "loss": 0.651,
+      "step": 15555
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.3184828456058675,
+      "learning_rate": 1.4626846313623615e-05,
+      "loss": 0.7815,
+      "step": 15556
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5960600720769478,
+      "learning_rate": 1.4626169846746878e-05,
+      "loss": 0.7618,
+      "step": 15557
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5390589170040563,
+      "learning_rate": 1.4625493352935114e-05,
+      "loss": 0.8482,
+      "step": 15558
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.13758463359842,
+      "learning_rate": 1.4624816832192266e-05,
+      "loss": 0.7575,
+      "step": 15559
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.641598356974321,
+      "learning_rate": 1.4624140284522267e-05,
+      "loss": 0.7901,
+      "step": 15560
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.028033662633851,
+      "learning_rate": 1.462346370992906e-05,
+      "loss": 0.7062,
+      "step": 15561
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3465103976907518,
+      "learning_rate": 1.4622787108416585e-05,
+      "loss": 0.7093,
+      "step": 15562
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6196812414035581,
+      "learning_rate": 1.4622110479988778e-05,
+      "loss": 0.5593,
+      "step": 15563
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4154956715146922,
+      "learning_rate": 1.4621433824649582e-05,
+      "loss": 0.8133,
+      "step": 15564
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2241719566628584,
+      "learning_rate": 1.4620757142402933e-05,
+      "loss": 0.7141,
+      "step": 15565
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3736340261497824,
+      "learning_rate": 1.4620080433252776e-05,
+      "loss": 0.7061,
+      "step": 15566
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.408866804445006,
+      "learning_rate": 1.4619403697203045e-05,
+      "loss": 0.7095,
+      "step": 15567
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6234423203922814,
+      "learning_rate": 1.4618726934257687e-05,
+      "loss": 0.6487,
+      "step": 15568
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5205938123829417,
+      "learning_rate": 1.4618050144420637e-05,
+      "loss": 0.9252,
+      "step": 15569
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.430652729313394,
+      "learning_rate": 1.4617373327695837e-05,
+      "loss": 0.704,
+      "step": 15570
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6699942847428093,
+      "learning_rate": 1.4616696484087228e-05,
+      "loss": 0.7053,
+      "step": 15571
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3528855739455357,
+      "learning_rate": 1.4616019613598748e-05,
+      "loss": 0.8203,
+      "step": 15572
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5977738486085332,
+      "learning_rate": 1.4615342716234342e-05,
+      "loss": 0.5329,
+      "step": 15573
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.394644132245806,
+      "learning_rate": 1.461466579199795e-05,
+      "loss": 0.6817,
+      "step": 15574
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4796428103402932,
+      "learning_rate": 1.4613988840893512e-05,
+      "loss": 0.8734,
+      "step": 15575
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2530334849971643,
+      "learning_rate": 1.461331186292497e-05,
+      "loss": 0.6568,
+      "step": 15576
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4411354342068252,
+      "learning_rate": 1.4612634858096266e-05,
+      "loss": 0.7706,
+      "step": 15577
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7576280146719014,
+      "learning_rate": 1.4611957826411343e-05,
+      "loss": 0.7916,
+      "step": 15578
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3800968418332045,
+      "learning_rate": 1.461128076787414e-05,
+      "loss": 0.6942,
+      "step": 15579
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9613114943530632,
+      "learning_rate": 1.46106036824886e-05,
+      "loss": 0.8184,
+      "step": 15580
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.71210333110895,
+      "learning_rate": 1.4609926570258665e-05,
+      "loss": 0.6575,
+      "step": 15581
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2622207647887076,
+      "learning_rate": 1.460924943118828e-05,
+      "loss": 0.7484,
+      "step": 15582
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3071083586953098,
+      "learning_rate": 1.460857226528139e-05,
+      "loss": 0.751,
+      "step": 15583
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6244315080157696,
+      "learning_rate": 1.4607895072541925e-05,
+      "loss": 0.7415,
+      "step": 15584
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7854182871864823,
+      "learning_rate": 1.4607217852973841e-05,
+      "loss": 0.6541,
+      "step": 15585
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4013019774941466,
+      "learning_rate": 1.460654060658107e-05,
+      "loss": 0.7456,
+      "step": 15586
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6204610607346162,
+      "learning_rate": 1.4605863333367567e-05,
+      "loss": 0.6138,
+      "step": 15587
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4271438144283584,
+      "learning_rate": 1.4605186033337266e-05,
+      "loss": 0.8349,
+      "step": 15588
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4262001204521233,
+      "learning_rate": 1.4604508706494119e-05,
+      "loss": 0.8461,
+      "step": 15589
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4635881749137698,
+      "learning_rate": 1.4603831352842061e-05,
+      "loss": 0.8557,
+      "step": 15590
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6420102410846024,
+      "learning_rate": 1.460315397238504e-05,
+      "loss": 0.545,
+      "step": 15591
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5183605695102704,
+      "learning_rate": 1.4602476565127e-05,
+      "loss": 0.81,
+      "step": 15592
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5767519638155425,
+      "learning_rate": 1.4601799131071883e-05,
+      "loss": 0.5166,
+      "step": 15593
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5214305100119179,
+      "learning_rate": 1.4601121670223633e-05,
+      "loss": 0.7508,
+      "step": 15594
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4912421035823742,
+      "learning_rate": 1.4600444182586197e-05,
+      "loss": 0.8556,
+      "step": 15595
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3299478637422073,
+      "learning_rate": 1.4599766668163522e-05,
+      "loss": 0.7836,
+      "step": 15596
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.578652413211658,
+      "learning_rate": 1.4599089126959545e-05,
+      "loss": 0.5274,
+      "step": 15597
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4889796212451085,
+      "learning_rate": 1.4598411558978217e-05,
+      "loss": 0.7529,
+      "step": 15598
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3462849353251327,
+      "learning_rate": 1.4597733964223479e-05,
+      "loss": 0.8068,
+      "step": 15599
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4939732357106617,
+      "learning_rate": 1.459705634269928e-05,
+      "loss": 0.7088,
+      "step": 15600
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5661230072363624,
+      "learning_rate": 1.4596378694409565e-05,
+      "loss": 0.5064,
+      "step": 15601
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5992627863455201,
+      "learning_rate": 1.4595701019358275e-05,
+      "loss": 0.5077,
+      "step": 15602
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6604237850334884,
+      "learning_rate": 1.4595023317549362e-05,
+      "loss": 0.7263,
+      "step": 15603
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3871805665877281,
+      "learning_rate": 1.4594345588986767e-05,
+      "loss": 0.8801,
+      "step": 15604
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.445262381698168,
+      "learning_rate": 1.4593667833674437e-05,
+      "loss": 0.7071,
+      "step": 15605
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.748391984575736,
+      "learning_rate": 1.4592990051616318e-05,
+      "loss": 0.7681,
+      "step": 15606
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6828516340540318,
+      "learning_rate": 1.4592312242816355e-05,
+      "loss": 0.7376,
+      "step": 15607
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0319155122787143,
+      "learning_rate": 1.45916344072785e-05,
+      "loss": 0.6867,
+      "step": 15608
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.587584961803875,
+      "learning_rate": 1.4590956545006693e-05,
+      "loss": 0.8023,
+      "step": 15609
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5519806302302417,
+      "learning_rate": 1.4590278656004888e-05,
+      "loss": 0.6924,
+      "step": 15610
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4807501763251065,
+      "learning_rate": 1.4589600740277023e-05,
+      "loss": 0.6579,
+      "step": 15611
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.433444004856382,
+      "learning_rate": 1.4588922797827052e-05,
+      "loss": 0.7603,
+      "step": 15612
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.789523934425429,
+      "learning_rate": 1.4588244828658918e-05,
+      "loss": 0.7817,
+      "step": 15613
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.608694611018738,
+      "learning_rate": 1.4587566832776569e-05,
+      "loss": 0.4852,
+      "step": 15614
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6050275477322968,
+      "learning_rate": 1.4586888810183956e-05,
+      "loss": 0.5773,
+      "step": 15615
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5993384142462052,
+      "learning_rate": 1.4586210760885023e-05,
+      "loss": 0.5224,
+      "step": 15616
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.396485018990525,
+      "learning_rate": 1.4585532684883718e-05,
+      "loss": 0.6343,
+      "step": 15617
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5274182048859744,
+      "learning_rate": 1.4584854582183992e-05,
+      "loss": 0.7781,
+      "step": 15618
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9457845689583972,
+      "learning_rate": 1.4584176452789791e-05,
+      "loss": 0.7372,
+      "step": 15619
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.481802511564908,
+      "learning_rate": 1.4583498296705063e-05,
+      "loss": 0.8568,
+      "step": 15620
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5977609086849383,
+      "learning_rate": 1.4582820113933755e-05,
+      "loss": 0.8639,
+      "step": 15621
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5832439813858052,
+      "learning_rate": 1.4582141904479819e-05,
+      "loss": 0.6827,
+      "step": 15622
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8695655750243623,
+      "learning_rate": 1.4581463668347201e-05,
+      "loss": 0.7755,
+      "step": 15623
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3671825883449302,
+      "learning_rate": 1.4580785405539853e-05,
+      "loss": 0.7193,
+      "step": 15624
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4038818312904462,
+      "learning_rate": 1.4580107116061722e-05,
+      "loss": 0.8074,
+      "step": 15625
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3425914868819233,
+      "learning_rate": 1.4579428799916754e-05,
+      "loss": 0.7453,
+      "step": 15626
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8394851905441607,
+      "learning_rate": 1.4578750457108904e-05,
+      "loss": 0.7255,
+      "step": 15627
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3336233146685865,
+      "learning_rate": 1.4578072087642118e-05,
+      "loss": 0.6935,
+      "step": 15628
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.427758459568157,
+      "learning_rate": 1.4577393691520349e-05,
+      "loss": 0.829,
+      "step": 15629
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5665141554203716,
+      "learning_rate": 1.4576715268747545e-05,
+      "loss": 0.8093,
+      "step": 15630
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9301780995270115,
+      "learning_rate": 1.4576036819327654e-05,
+      "loss": 0.7385,
+      "step": 15631
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.490774081349512,
+      "learning_rate": 1.4575358343264628e-05,
+      "loss": 0.8109,
+      "step": 15632
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5631771537827348,
+      "learning_rate": 1.4574679840562417e-05,
+      "loss": 0.7249,
+      "step": 15633
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7289895971035223,
+      "learning_rate": 1.4574001311224976e-05,
+      "loss": 0.6827,
+      "step": 15634
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7330911449097692,
+      "learning_rate": 1.4573322755256245e-05,
+      "loss": 0.7517,
+      "step": 15635
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.4078720341624416,
+      "learning_rate": 1.4572644172660186e-05,
+      "loss": 0.8429,
+      "step": 15636
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2808442667389899,
+      "learning_rate": 1.457196556344074e-05,
+      "loss": 0.7434,
+      "step": 15637
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.733108603666001,
+      "learning_rate": 1.4571286927601868e-05,
+      "loss": 0.7609,
+      "step": 15638
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6002944111787313,
+      "learning_rate": 1.4570608265147512e-05,
+      "loss": 0.4943,
+      "step": 15639
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5374837079445054,
+      "learning_rate": 1.4569929576081633e-05,
+      "loss": 0.7199,
+      "step": 15640
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.31198716943678,
+      "learning_rate": 1.4569250860408173e-05,
+      "loss": 0.7057,
+      "step": 15641
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4485104592557336,
+      "learning_rate": 1.4568572118131087e-05,
+      "loss": 0.7236,
+      "step": 15642
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3750679430762058,
+      "learning_rate": 1.456789334925433e-05,
+      "loss": 0.6423,
+      "step": 15643
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5653587055453517,
+      "learning_rate": 1.456721455378185e-05,
+      "loss": 0.5046,
+      "step": 15644
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2136990936641998,
+      "learning_rate": 1.4566535731717603e-05,
+      "loss": 0.7596,
+      "step": 15645
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6133125707641842,
+      "learning_rate": 1.456585688306554e-05,
+      "loss": 0.7444,
+      "step": 15646
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4059236693400172,
+      "learning_rate": 1.456517800782961e-05,
+      "loss": 0.825,
+      "step": 15647
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3860034874789056,
+      "learning_rate": 1.4564499106013768e-05,
+      "loss": 0.6839,
+      "step": 15648
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4611294421673693,
+      "learning_rate": 1.4563820177621968e-05,
+      "loss": 0.7856,
+      "step": 15649
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5526945080246328,
+      "learning_rate": 1.4563141222658163e-05,
+      "loss": 0.7151,
+      "step": 15650
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4133887949142587,
+      "learning_rate": 1.4562462241126305e-05,
+      "loss": 0.775,
+      "step": 15651
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4839359795727678,
+      "learning_rate": 1.4561783233030346e-05,
+      "loss": 0.7388,
+      "step": 15652
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6375880366377584,
+      "learning_rate": 1.4561104198374243e-05,
+      "loss": 0.7676,
+      "step": 15653
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4248654483497505,
+      "learning_rate": 1.4560425137161946e-05,
+      "loss": 0.7269,
+      "step": 15654
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7784767783897542,
+      "learning_rate": 1.4559746049397411e-05,
+      "loss": 0.7729,
+      "step": 15655
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8713878475076045,
+      "learning_rate": 1.455906693508459e-05,
+      "loss": 0.744,
+      "step": 15656
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3156996869517112,
+      "learning_rate": 1.4558387794227438e-05,
+      "loss": 0.6778,
+      "step": 15657
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9004599546881267,
+      "learning_rate": 1.4557708626829907e-05,
+      "loss": 0.6915,
+      "step": 15658
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5715507202186314,
+      "learning_rate": 1.455702943289596e-05,
+      "loss": 0.5155,
+      "step": 15659
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.453965652634711,
+      "learning_rate": 1.4556350212429538e-05,
+      "loss": 0.7317,
+      "step": 15660
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8311740074460652,
+      "learning_rate": 1.4555670965434602e-05,
+      "loss": 0.629,
+      "step": 15661
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6505578616580967,
+      "learning_rate": 1.4554991691915115e-05,
+      "loss": 0.8605,
+      "step": 15662
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3982225062836933,
+      "learning_rate": 1.4554312391875016e-05,
+      "loss": 0.7535,
+      "step": 15663
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.866260527714061,
+      "learning_rate": 1.4553633065318274e-05,
+      "loss": 0.8798,
+      "step": 15664
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.4585852543958078,
+      "learning_rate": 1.4552953712248834e-05,
+      "loss": 0.8535,
+      "step": 15665
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8071721302854624,
+      "learning_rate": 1.4552274332670662e-05,
+      "loss": 0.7156,
+      "step": 15666
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6012676839020227,
+      "learning_rate": 1.4551594926587703e-05,
+      "loss": 0.5654,
+      "step": 15667
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6828091109360194,
+      "learning_rate": 1.4550915494003919e-05,
+      "loss": 0.5366,
+      "step": 15668
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2509561707984704,
+      "learning_rate": 1.4550236034923263e-05,
+      "loss": 0.8053,
+      "step": 15669
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.9579601840894587,
+      "learning_rate": 1.4549556549349694e-05,
+      "loss": 0.6506,
+      "step": 15670
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4152065821127617,
+      "learning_rate": 1.4548877037287166e-05,
+      "loss": 0.7479,
+      "step": 15671
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6581149781114393,
+      "learning_rate": 1.4548197498739636e-05,
+      "loss": 0.7445,
+      "step": 15672
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6340472843299254,
+      "learning_rate": 1.4547517933711058e-05,
+      "loss": 0.6908,
+      "step": 15673
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.354647499619189,
+      "learning_rate": 1.4546838342205392e-05,
+      "loss": 0.7887,
+      "step": 15674
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4113514589952803,
+      "learning_rate": 1.4546158724226594e-05,
+      "loss": 0.6095,
+      "step": 15675
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6133093227892574,
+      "learning_rate": 1.4545479079778622e-05,
+      "loss": 0.5308,
+      "step": 15676
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4870038673750607,
+      "learning_rate": 1.4544799408865428e-05,
+      "loss": 0.7472,
+      "step": 15677
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4468400380592352,
+      "learning_rate": 1.4544119711490976e-05,
+      "loss": 0.7829,
+      "step": 15678
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5707344121957065,
+      "learning_rate": 1.4543439987659221e-05,
+      "loss": 0.5327,
+      "step": 15679
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5409744272996087,
+      "learning_rate": 1.4542760237374119e-05,
+      "loss": 0.8187,
+      "step": 15680
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2575418558690101,
+      "learning_rate": 1.4542080460639627e-05,
+      "loss": 0.7749,
+      "step": 15681
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1582262556249243,
+      "learning_rate": 1.4541400657459705e-05,
+      "loss": 0.6102,
+      "step": 15682
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5345003977553278,
+      "learning_rate": 1.4540720827838313e-05,
+      "loss": 0.6254,
+      "step": 15683
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4710458909009474,
+      "learning_rate": 1.4540040971779403e-05,
+      "loss": 0.6763,
+      "step": 15684
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.840019631434515,
+      "learning_rate": 1.4539361089286942e-05,
+      "loss": 0.7772,
+      "step": 15685
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2960166346332314,
+      "learning_rate": 1.4538681180364878e-05,
+      "loss": 0.739,
+      "step": 15686
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6328275106406853,
+      "learning_rate": 1.4538001245017182e-05,
+      "loss": 0.8092,
+      "step": 15687
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.428117238437652,
+      "learning_rate": 1.4537321283247799e-05,
+      "loss": 0.6966,
+      "step": 15688
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5937698869090444,
+      "learning_rate": 1.4536641295060701e-05,
+      "loss": 0.7322,
+      "step": 15689
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7761138768221558,
+      "learning_rate": 1.4535961280459838e-05,
+      "loss": 0.6919,
+      "step": 15690
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3292771944410178,
+      "learning_rate": 1.4535281239449173e-05,
+      "loss": 0.8904,
+      "step": 15691
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5803832111245177,
+      "learning_rate": 1.4534601172032663e-05,
+      "loss": 0.6846,
+      "step": 15692
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4192985555360356,
+      "learning_rate": 1.4533921078214271e-05,
+      "loss": 0.7711,
+      "step": 15693
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7609725767847753,
+      "learning_rate": 1.4533240957997956e-05,
+      "loss": 0.866,
+      "step": 15694
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6271027990987965,
+      "learning_rate": 1.4532560811387674e-05,
+      "loss": 0.7481,
+      "step": 15695
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5510508856853897,
+      "learning_rate": 1.453188063838739e-05,
+      "loss": 0.7493,
+      "step": 15696
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.622864223729575,
+      "learning_rate": 1.4531200439001063e-05,
+      "loss": 0.5726,
+      "step": 15697
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8186207237622085,
+      "learning_rate": 1.453052021323265e-05,
+      "loss": 0.8874,
+      "step": 15698
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4805155808076067,
+      "learning_rate": 1.4529839961086116e-05,
+      "loss": 0.7458,
+      "step": 15699
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6580870219868,
+      "learning_rate": 1.452915968256542e-05,
+      "loss": 0.7475,
+      "step": 15700
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5274864943577777,
+      "learning_rate": 1.4528479377674521e-05,
+      "loss": 0.7211,
+      "step": 15701
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.777838639724679,
+      "learning_rate": 1.452779904641738e-05,
+      "loss": 0.8502,
+      "step": 15702
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3868176911880452,
+      "learning_rate": 1.4527118688797963e-05,
+      "loss": 0.663,
+      "step": 15703
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4453708631460922,
+      "learning_rate": 1.4526438304820225e-05,
+      "loss": 0.7569,
+      "step": 15704
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5341095438827772,
+      "learning_rate": 1.452575789448813e-05,
+      "loss": 0.8612,
+      "step": 15705
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0201334741940533,
+      "learning_rate": 1.4525077457805646e-05,
+      "loss": 0.6519,
+      "step": 15706
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4444220327854633,
+      "learning_rate": 1.452439699477672e-05,
+      "loss": 0.7129,
+      "step": 15707
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3773489322654624,
+      "learning_rate": 1.4523716505405327e-05,
+      "loss": 0.7206,
+      "step": 15708
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.568442749679774,
+      "learning_rate": 1.4523035989695422e-05,
+      "loss": 0.9023,
+      "step": 15709
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7349419878593966,
+      "learning_rate": 1.452235544765097e-05,
+      "loss": 0.767,
+      "step": 15710
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6824322584416695,
+      "learning_rate": 1.4521674879275933e-05,
+      "loss": 0.7992,
+      "step": 15711
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4368400819884726,
+      "learning_rate": 1.4520994284574275e-05,
+      "loss": 0.7085,
+      "step": 15712
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.465042521024242,
+      "learning_rate": 1.4520313663549957e-05,
+      "loss": 0.7462,
+      "step": 15713
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3456724228184578,
+      "learning_rate": 1.4519633016206938e-05,
+      "loss": 0.799,
+      "step": 15714
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5615184309580867,
+      "learning_rate": 1.4518952342549189e-05,
+      "loss": 0.7889,
+      "step": 15715
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6031145108598229,
+      "learning_rate": 1.4518271642580663e-05,
+      "loss": 0.5448,
+      "step": 15716
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4080215721914713,
+      "learning_rate": 1.4517590916305335e-05,
+      "loss": 0.716,
+      "step": 15717
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.845149985753595,
+      "learning_rate": 1.4516910163727158e-05,
+      "loss": 0.8037,
+      "step": 15718
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3919061601696558,
+      "learning_rate": 1.4516229384850105e-05,
+      "loss": 0.8367,
+      "step": 15719
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7704242152833467,
+      "learning_rate": 1.451554857967813e-05,
+      "loss": 0.7345,
+      "step": 15720
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4539967388600707,
+      "learning_rate": 1.4514867748215204e-05,
+      "loss": 0.7344,
+      "step": 15721
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2715884336076813,
+      "learning_rate": 1.4514186890465289e-05,
+      "loss": 0.738,
+      "step": 15722
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.183041491145077,
+      "learning_rate": 1.4513506006432347e-05,
+      "loss": 0.6767,
+      "step": 15723
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5359549302793567,
+      "learning_rate": 1.4512825096120346e-05,
+      "loss": 0.8211,
+      "step": 15724
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.402298398744793,
+      "learning_rate": 1.4512144159533248e-05,
+      "loss": 0.7327,
+      "step": 15725
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2853426204185654,
+      "learning_rate": 1.4511463196675018e-05,
+      "loss": 0.5931,
+      "step": 15726
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.616422631951114,
+      "learning_rate": 1.4510782207549619e-05,
+      "loss": 0.7384,
+      "step": 15727
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4782051103123706,
+      "learning_rate": 1.451010119216102e-05,
+      "loss": 0.7863,
+      "step": 15728
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.156557302113867,
+      "learning_rate": 1.4509420150513183e-05,
+      "loss": 0.6436,
+      "step": 15729
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6581397380661933,
+      "learning_rate": 1.4508739082610075e-05,
+      "loss": 0.6658,
+      "step": 15730
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.398663469725649,
+      "learning_rate": 1.4508057988455659e-05,
+      "loss": 0.6767,
+      "step": 15731
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2014198186056135,
+      "learning_rate": 1.4507376868053903e-05,
+      "loss": 0.8327,
+      "step": 15732
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4807087095670173,
+      "learning_rate": 1.4506695721408775e-05,
+      "loss": 0.6001,
+      "step": 15733
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.9746577336496203,
+      "learning_rate": 1.4506014548524237e-05,
+      "loss": 0.6876,
+      "step": 15734
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2971415318146011,
+      "learning_rate": 1.4505333349404252e-05,
+      "loss": 0.7236,
+      "step": 15735
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5242489447013676,
+      "learning_rate": 1.4504652124052792e-05,
+      "loss": 0.7222,
+      "step": 15736
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5010622406747889,
+      "learning_rate": 1.450397087247382e-05,
+      "loss": 0.6306,
+      "step": 15737
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.652306849733301,
+      "learning_rate": 1.4503289594671306e-05,
+      "loss": 0.8384,
+      "step": 15738
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4339699942714428,
+      "learning_rate": 1.450260829064921e-05,
+      "loss": 0.6597,
+      "step": 15739
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1093377804894256,
+      "learning_rate": 1.4501926960411508e-05,
+      "loss": 0.644,
+      "step": 15740
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.80115666307938,
+      "learning_rate": 1.450124560396216e-05,
+      "loss": 0.7019,
+      "step": 15741
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8925742680404027,
+      "learning_rate": 1.4500564221305133e-05,
+      "loss": 0.6942,
+      "step": 15742
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3761167490473154,
+      "learning_rate": 1.44998828124444e-05,
+      "loss": 0.6732,
+      "step": 15743
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8606720094975393,
+      "learning_rate": 1.4499201377383923e-05,
+      "loss": 0.7217,
+      "step": 15744
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.690193375599363,
+      "learning_rate": 1.449851991612767e-05,
+      "loss": 0.6357,
+      "step": 15745
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4561634037725564,
+      "learning_rate": 1.449783842867961e-05,
+      "loss": 0.7453,
+      "step": 15746
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3755438042152064,
+      "learning_rate": 1.4497156915043712e-05,
+      "loss": 0.7195,
+      "step": 15747
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3424307737624994,
+      "learning_rate": 1.4496475375223941e-05,
+      "loss": 0.8296,
+      "step": 15748
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0779615130768025,
+      "learning_rate": 1.4495793809224268e-05,
+      "loss": 0.7519,
+      "step": 15749
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4998876228451627,
+      "learning_rate": 1.449511221704866e-05,
+      "loss": 0.818,
+      "step": 15750
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.532076933071302,
+      "learning_rate": 1.4494430598701087e-05,
+      "loss": 0.7125,
+      "step": 15751
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4082287916029033,
+      "learning_rate": 1.4493748954185513e-05,
+      "loss": 0.754,
+      "step": 15752
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.676597834393941,
+      "learning_rate": 1.449306728350591e-05,
+      "loss": 0.7378,
+      "step": 15753
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7126845892472373,
+      "learning_rate": 1.4492385586666249e-05,
+      "loss": 0.8267,
+      "step": 15754
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5565633490907742,
+      "learning_rate": 1.4491703863670496e-05,
+      "loss": 0.7502,
+      "step": 15755
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6949703820803037,
+      "learning_rate": 1.4491022114522621e-05,
+      "loss": 0.6521,
+      "step": 15756
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.597193101795541,
+      "learning_rate": 1.4490340339226596e-05,
+      "loss": 0.7494,
+      "step": 15757
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8751529935804847,
+      "learning_rate": 1.4489658537786385e-05,
+      "loss": 0.6809,
+      "step": 15758
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.452119904973275,
+      "learning_rate": 1.448897671020596e-05,
+      "loss": 0.8092,
+      "step": 15759
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5026928960855197,
+      "learning_rate": 1.4488294856489291e-05,
+      "loss": 0.7541,
+      "step": 15760
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4839550562160242,
+      "learning_rate": 1.448761297664035e-05,
+      "loss": 0.835,
+      "step": 15761
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3117577259937798,
+      "learning_rate": 1.4486931070663107e-05,
+      "loss": 0.6497,
+      "step": 15762
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3732016424025333,
+      "learning_rate": 1.4486249138561528e-05,
+      "loss": 0.7493,
+      "step": 15763
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.616148251465999,
+      "learning_rate": 1.448556718033959e-05,
+      "loss": 0.7454,
+      "step": 15764
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6825399150840528,
+      "learning_rate": 1.4484885196001255e-05,
+      "loss": 0.6895,
+      "step": 15765
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6121756494421272,
+      "learning_rate": 1.4484203185550503e-05,
+      "loss": 0.5937,
+      "step": 15766
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5946580786506779,
+      "learning_rate": 1.4483521148991295e-05,
+      "loss": 0.7155,
+      "step": 15767
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5459803245260502,
+      "learning_rate": 1.4482839086327614e-05,
+      "loss": 0.7089,
+      "step": 15768
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5681801311742997,
+      "learning_rate": 1.4482156997563417e-05,
+      "loss": 0.4979,
+      "step": 15769
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6429853560121209,
+      "learning_rate": 1.4481474882702688e-05,
+      "loss": 0.5991,
+      "step": 15770
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5591174577619632,
+      "learning_rate": 1.4480792741749395e-05,
+      "loss": 0.5235,
+      "step": 15771
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.388698524580367,
+      "learning_rate": 1.4480110574707504e-05,
+      "loss": 0.819,
+      "step": 15772
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.696915900607432,
+      "learning_rate": 1.4479428381580993e-05,
+      "loss": 0.8618,
+      "step": 15773
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5496150069458052,
+      "learning_rate": 1.447874616237383e-05,
+      "loss": 0.7709,
+      "step": 15774
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.184035155722803,
+      "learning_rate": 1.447806391708999e-05,
+      "loss": 0.7196,
+      "step": 15775
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4112828982435481,
+      "learning_rate": 1.4477381645733446e-05,
+      "loss": 0.6181,
+      "step": 15776
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4424667484097602,
+      "learning_rate": 1.4476699348308165e-05,
+      "loss": 0.8076,
+      "step": 15777
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6329912065720591,
+      "learning_rate": 1.4476017024818126e-05,
+      "loss": 0.55,
+      "step": 15778
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4391363457839248,
+      "learning_rate": 1.4475334675267296e-05,
+      "loss": 0.7316,
+      "step": 15779
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5223285057419407,
+      "learning_rate": 1.4474652299659653e-05,
+      "loss": 0.6798,
+      "step": 15780
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.363239139991703,
+      "learning_rate": 1.4473969897999168e-05,
+      "loss": 0.7901,
+      "step": 15781
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5706110740297681,
+      "learning_rate": 1.4473287470289811e-05,
+      "loss": 0.4879,
+      "step": 15782
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.007291869736348,
+      "learning_rate": 1.447260501653556e-05,
+      "loss": 0.6805,
+      "step": 15783
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.374572270346579,
+      "learning_rate": 1.4471922536740385e-05,
+      "loss": 0.7411,
+      "step": 15784
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3050962393489092,
+      "learning_rate": 1.4471240030908265e-05,
+      "loss": 0.8044,
+      "step": 15785
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4767055332113836,
+      "learning_rate": 1.4470557499043167e-05,
+      "loss": 0.7428,
+      "step": 15786
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3251484560189277,
+      "learning_rate": 1.4469874941149069e-05,
+      "loss": 0.6318,
+      "step": 15787
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.545257927755271,
+      "learning_rate": 1.4469192357229942e-05,
+      "loss": 0.7997,
+      "step": 15788
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2515912222739263,
+      "learning_rate": 1.4468509747289764e-05,
+      "loss": 0.6913,
+      "step": 15789
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5307599438614419,
+      "learning_rate": 1.4467827111332506e-05,
+      "loss": 0.663,
+      "step": 15790
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6483605555219373,
+      "learning_rate": 1.4467144449362147e-05,
+      "loss": 0.5439,
+      "step": 15791
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4222866711354907,
+      "learning_rate": 1.4466461761382657e-05,
+      "loss": 0.6457,
+      "step": 15792
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5821042426114018,
+      "learning_rate": 1.4465779047398014e-05,
+      "loss": 0.5087,
+      "step": 15793
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3930724957609972,
+      "learning_rate": 1.446509630741219e-05,
+      "loss": 0.7661,
+      "step": 15794
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4550903034080593,
+      "learning_rate": 1.4464413541429161e-05,
+      "loss": 0.7684,
+      "step": 15795
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9071322980952443,
+      "learning_rate": 1.4463730749452904e-05,
+      "loss": 0.7758,
+      "step": 15796
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6081192037893924,
+      "learning_rate": 1.4463047931487391e-05,
+      "loss": 0.5321,
+      "step": 15797
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3739734420825973,
+      "learning_rate": 1.4462365087536603e-05,
+      "loss": 0.7898,
+      "step": 15798
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5288460367827428,
+      "learning_rate": 1.446168221760451e-05,
+      "loss": 0.7713,
+      "step": 15799
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5899288263428633,
+      "learning_rate": 1.4460999321695095e-05,
+      "loss": 0.7871,
+      "step": 15800
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6098041486070896,
+      "learning_rate": 1.4460316399812326e-05,
+      "loss": 0.5216,
+      "step": 15801
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4282447303080938,
+      "learning_rate": 1.4459633451960184e-05,
+      "loss": 0.7107,
+      "step": 15802
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.703454398753846,
+      "learning_rate": 1.4458950478142643e-05,
+      "loss": 0.745,
+      "step": 15803
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6094788142985499,
+      "learning_rate": 1.4458267478363683e-05,
+      "loss": 0.754,
+      "step": 15804
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4912180535828052,
+      "learning_rate": 1.4457584452627276e-05,
+      "loss": 0.7679,
+      "step": 15805
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5635849704670762,
+      "learning_rate": 1.4456901400937404e-05,
+      "loss": 0.4841,
+      "step": 15806
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4166269292914504,
+      "learning_rate": 1.4456218323298037e-05,
+      "loss": 0.679,
+      "step": 15807
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5271394933855962,
+      "learning_rate": 1.4455535219713157e-05,
+      "loss": 0.6776,
+      "step": 15808
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5298376928946071,
+      "learning_rate": 1.4454852090186742e-05,
+      "loss": 0.7129,
+      "step": 15809
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.598138156235524,
+      "learning_rate": 1.4454168934722765e-05,
+      "loss": 0.8403,
+      "step": 15810
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4631726557961307,
+      "learning_rate": 1.445348575332521e-05,
+      "loss": 0.7711,
+      "step": 15811
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5845463165381802,
+      "learning_rate": 1.4452802545998048e-05,
+      "loss": 0.5281,
+      "step": 15812
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5883389892358575,
+      "learning_rate": 1.4452119312745266e-05,
+      "loss": 0.5611,
+      "step": 15813
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3449928622161018,
+      "learning_rate": 1.4451436053570826e-05,
+      "loss": 0.6922,
+      "step": 15814
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9577820893337212,
+      "learning_rate": 1.4450752768478725e-05,
+      "loss": 0.7444,
+      "step": 15815
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6753507130334935,
+      "learning_rate": 1.4450069457472926e-05,
+      "loss": 0.6983,
+      "step": 15816
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6655789522987419,
+      "learning_rate": 1.4449386120557418e-05,
+      "loss": 0.7121,
+      "step": 15817
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3774625013681492,
+      "learning_rate": 1.4448702757736172e-05,
+      "loss": 0.6869,
+      "step": 15818
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.426937025037468,
+      "learning_rate": 1.4448019369013173e-05,
+      "loss": 0.6976,
+      "step": 15819
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5558558170086731,
+      "learning_rate": 1.4447335954392393e-05,
+      "loss": 0.7738,
+      "step": 15820
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5970508616886943,
+      "learning_rate": 1.4446652513877818e-05,
+      "loss": 0.548,
+      "step": 15821
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6433571236852904,
+      "learning_rate": 1.4445969047473427e-05,
+      "loss": 0.5584,
+      "step": 15822
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2741049920211016,
+      "learning_rate": 1.4445285555183193e-05,
+      "loss": 0.706,
+      "step": 15823
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5559179243662185,
+      "learning_rate": 1.4444602037011099e-05,
+      "loss": 0.6677,
+      "step": 15824
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.614747545244478,
+      "learning_rate": 1.4443918492961127e-05,
+      "loss": 0.5607,
+      "step": 15825
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4259044745837122,
+      "learning_rate": 1.4443234923037253e-05,
+      "loss": 0.8081,
+      "step": 15826
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0794576800495816,
+      "learning_rate": 1.4442551327243456e-05,
+      "loss": 0.7077,
+      "step": 15827
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7910062579300796,
+      "learning_rate": 1.4441867705583723e-05,
+      "loss": 0.7473,
+      "step": 15828
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.165187479491333,
+      "learning_rate": 1.4441184058062028e-05,
+      "loss": 0.8352,
+      "step": 15829
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.481681910108049,
+      "learning_rate": 1.4440500384682353e-05,
+      "loss": 0.6824,
+      "step": 15830
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8326725468989116,
+      "learning_rate": 1.443981668544868e-05,
+      "loss": 0.7309,
+      "step": 15831
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4725573522573394,
+      "learning_rate": 1.4439132960364988e-05,
+      "loss": 0.8929,
+      "step": 15832
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4759809851106924,
+      "learning_rate": 1.4438449209435259e-05,
+      "loss": 0.7954,
+      "step": 15833
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3613466570332842,
+      "learning_rate": 1.4437765432663472e-05,
+      "loss": 0.6716,
+      "step": 15834
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6638146938296192,
+      "learning_rate": 1.4437081630053614e-05,
+      "loss": 0.727,
+      "step": 15835
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4698598088413874,
+      "learning_rate": 1.4436397801609658e-05,
+      "loss": 0.6299,
+      "step": 15836
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7400904156419714,
+      "learning_rate": 1.443571394733559e-05,
+      "loss": 0.6567,
+      "step": 15837
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.99280046422679,
+      "learning_rate": 1.4435030067235394e-05,
+      "loss": 0.7074,
+      "step": 15838
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6066283592096858,
+      "learning_rate": 1.4434346161313041e-05,
+      "loss": 0.5168,
+      "step": 15839
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4964697690419522,
+      "learning_rate": 1.4433662229572529e-05,
+      "loss": 0.7022,
+      "step": 15840
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.400325942906611,
+      "learning_rate": 1.4432978272017828e-05,
+      "loss": 0.6666,
+      "step": 15841
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.126888944602642,
+      "learning_rate": 1.4432294288652926e-05,
+      "loss": 0.7585,
+      "step": 15842
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5842019639076979,
+      "learning_rate": 1.4431610279481802e-05,
+      "loss": 0.6915,
+      "step": 15843
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.3333683203077165,
+      "learning_rate": 1.4430926244508441e-05,
+      "loss": 0.7039,
+      "step": 15844
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5411626839830816,
+      "learning_rate": 1.4430242183736824e-05,
+      "loss": 0.7024,
+      "step": 15845
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4058923626446436,
+      "learning_rate": 1.4429558097170932e-05,
+      "loss": 0.6795,
+      "step": 15846
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.361379187970044,
+      "learning_rate": 1.4428873984814755e-05,
+      "loss": 0.7277,
+      "step": 15847
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4454129870180312,
+      "learning_rate": 1.4428189846672267e-05,
+      "loss": 0.66,
+      "step": 15848
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.340374323563311,
+      "learning_rate": 1.442750568274746e-05,
+      "loss": 0.8326,
+      "step": 15849
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.401884721536637,
+      "learning_rate": 1.4426821493044312e-05,
+      "loss": 0.7561,
+      "step": 15850
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6308930391685417,
+      "learning_rate": 1.4426137277566805e-05,
+      "loss": 0.755,
+      "step": 15851
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8450292842300613,
+      "learning_rate": 1.4425453036318929e-05,
+      "loss": 0.8146,
+      "step": 15852
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2878100076928263,
+      "learning_rate": 1.4424768769304663e-05,
+      "loss": 0.7531,
+      "step": 15853
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4682999327885264,
+      "learning_rate": 1.4424084476527991e-05,
+      "loss": 0.7756,
+      "step": 15854
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.41019667627085,
+      "learning_rate": 1.44234001579929e-05,
+      "loss": 0.7396,
+      "step": 15855
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6764520997799206,
+      "learning_rate": 1.4422715813703372e-05,
+      "loss": 0.6362,
+      "step": 15856
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4392942743654684,
+      "learning_rate": 1.4422031443663391e-05,
+      "loss": 0.7136,
+      "step": 15857
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6552298101580738,
+      "learning_rate": 1.4421347047876945e-05,
+      "loss": 0.7358,
+      "step": 15858
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5052082994136409,
+      "learning_rate": 1.4420662626348017e-05,
+      "loss": 0.8657,
+      "step": 15859
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.647208881735496,
+      "learning_rate": 1.441997817908059e-05,
+      "loss": 0.7745,
+      "step": 15860
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4576238793032197,
+      "learning_rate": 1.4419293706078654e-05,
+      "loss": 0.7645,
+      "step": 15861
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1416578618599296,
+      "learning_rate": 1.4418609207346189e-05,
+      "loss": 0.6146,
+      "step": 15862
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.564896972485912,
+      "learning_rate": 1.4417924682887181e-05,
+      "loss": 0.8569,
+      "step": 15863
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.458429243643312,
+      "learning_rate": 1.4417240132705618e-05,
+      "loss": 0.7529,
+      "step": 15864
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6587999068159298,
+      "learning_rate": 1.4416555556805482e-05,
+      "loss": 0.6754,
+      "step": 15865
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5951088132652884,
+      "learning_rate": 1.4415870955190767e-05,
+      "loss": 0.5303,
+      "step": 15866
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3290602706533254,
+      "learning_rate": 1.441518632786545e-05,
+      "loss": 0.8833,
+      "step": 15867
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5375583942127498,
+      "learning_rate": 1.441450167483352e-05,
+      "loss": 0.7408,
+      "step": 15868
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6219751907879992,
+      "learning_rate": 1.4413816996098962e-05,
+      "loss": 0.5069,
+      "step": 15869
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7780525346092557,
+      "learning_rate": 1.4413132291665768e-05,
+      "loss": 0.685,
+      "step": 15870
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8399916980210957,
+      "learning_rate": 1.4412447561537919e-05,
+      "loss": 0.8309,
+      "step": 15871
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.76174989895835,
+      "learning_rate": 1.4411762805719402e-05,
+      "loss": 0.7939,
+      "step": 15872
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2871786104223513,
+      "learning_rate": 1.4411078024214207e-05,
+      "loss": 0.7253,
+      "step": 15873
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6270602384533477,
+      "learning_rate": 1.4410393217026317e-05,
+      "loss": 0.6546,
+      "step": 15874
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5186412011583388,
+      "learning_rate": 1.4409708384159724e-05,
+      "loss": 0.8821,
+      "step": 15875
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4313794110548836,
+      "learning_rate": 1.4409023525618411e-05,
+      "loss": 0.7977,
+      "step": 15876
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2642590644929639,
+      "learning_rate": 1.4408338641406368e-05,
+      "loss": 0.7147,
+      "step": 15877
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6654142760693553,
+      "learning_rate": 1.440765373152758e-05,
+      "loss": 0.8156,
+      "step": 15878
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5242776008365584,
+      "learning_rate": 1.4406968795986038e-05,
+      "loss": 0.6646,
+      "step": 15879
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.39222259795418,
+      "learning_rate": 1.4406283834785728e-05,
+      "loss": 0.7147,
+      "step": 15880
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2490837150988972,
+      "learning_rate": 1.4405598847930639e-05,
+      "loss": 0.6205,
+      "step": 15881
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4002365069749603,
+      "learning_rate": 1.440491383542476e-05,
+      "loss": 0.7163,
+      "step": 15882
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1024529659320654,
+      "learning_rate": 1.4404228797272076e-05,
+      "loss": 0.8385,
+      "step": 15883
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.812098843018179,
+      "learning_rate": 1.4403543733476578e-05,
+      "loss": 0.7651,
+      "step": 15884
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6206901936891922,
+      "learning_rate": 1.4402858644042254e-05,
+      "loss": 0.5506,
+      "step": 15885
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4956591560523418,
+      "learning_rate": 1.4402173528973094e-05,
+      "loss": 0.7955,
+      "step": 15886
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.479743785551026,
+      "learning_rate": 1.4401488388273086e-05,
+      "loss": 0.7346,
+      "step": 15887
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3730092272406207,
+      "learning_rate": 1.4400803221946218e-05,
+      "loss": 0.7939,
+      "step": 15888
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5454675199196064,
+      "learning_rate": 1.4400118029996481e-05,
+      "loss": 0.6184,
+      "step": 15889
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.5278164053553343,
+      "learning_rate": 1.4399432812427862e-05,
+      "loss": 0.6898,
+      "step": 15890
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5337807734673166,
+      "learning_rate": 1.4398747569244355e-05,
+      "loss": 0.6889,
+      "step": 15891
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.497762259758015,
+      "learning_rate": 1.4398062300449947e-05,
+      "loss": 0.7262,
+      "step": 15892
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6731858854357613,
+      "learning_rate": 1.4397377006048628e-05,
+      "loss": 0.7672,
+      "step": 15893
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5508226954308229,
+      "learning_rate": 1.4396691686044386e-05,
+      "loss": 0.6733,
+      "step": 15894
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5859431857095385,
+      "learning_rate": 1.4396006340441214e-05,
+      "loss": 0.7888,
+      "step": 15895
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5454566198109267,
+      "learning_rate": 1.43953209692431e-05,
+      "loss": 0.8958,
+      "step": 15896
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5303768094687524,
+      "learning_rate": 1.4394635572454036e-05,
+      "loss": 0.8128,
+      "step": 15897
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7805000919994791,
+      "learning_rate": 1.4393950150078016e-05,
+      "loss": 0.7227,
+      "step": 15898
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5959317728187923,
+      "learning_rate": 1.4393264702119023e-05,
+      "loss": 0.5321,
+      "step": 15899
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4885962081741657,
+      "learning_rate": 1.4392579228581054e-05,
+      "loss": 0.7586,
+      "step": 15900
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.552726452660293,
+      "learning_rate": 1.4391893729468099e-05,
+      "loss": 0.8093,
+      "step": 15901
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.495472110225854,
+      "learning_rate": 1.4391208204784146e-05,
+      "loss": 0.7483,
+      "step": 15902
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6720050502139532,
+      "learning_rate": 1.439052265453319e-05,
+      "loss": 0.705,
+      "step": 15903
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.483441354361962,
+      "learning_rate": 1.438983707871922e-05,
+      "loss": 0.703,
+      "step": 15904
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8749683249239384,
+      "learning_rate": 1.4389151477346231e-05,
+      "loss": 0.648,
+      "step": 15905
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4650016530781538,
+      "learning_rate": 1.438846585041821e-05,
+      "loss": 0.6331,
+      "step": 15906
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4324204150418354,
+      "learning_rate": 1.4387780197939152e-05,
+      "loss": 0.8157,
+      "step": 15907
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.38667541116333,
+      "learning_rate": 1.4387094519913048e-05,
+      "loss": 0.6788,
+      "step": 15908
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4831804815873355,
+      "learning_rate": 1.438640881634389e-05,
+      "loss": 0.8176,
+      "step": 15909
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3242005435569737,
+      "learning_rate": 1.4385723087235674e-05,
+      "loss": 0.7196,
+      "step": 15910
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8610543039651957,
+      "learning_rate": 1.438503733259239e-05,
+      "loss": 0.7269,
+      "step": 15911
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.3186583882530893,
+      "learning_rate": 1.4384351552418027e-05,
+      "loss": 0.7363,
+      "step": 15912
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0660615169079293,
+      "learning_rate": 1.4383665746716582e-05,
+      "loss": 0.7404,
+      "step": 15913
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6068348733566443,
+      "learning_rate": 1.4382979915492051e-05,
+      "loss": 0.496,
+      "step": 15914
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7688422099413352,
+      "learning_rate": 1.4382294058748421e-05,
+      "loss": 0.7517,
+      "step": 15915
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5700027140083632,
+      "learning_rate": 1.4381608176489686e-05,
+      "loss": 0.5808,
+      "step": 15916
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3321731475383405,
+      "learning_rate": 1.438092226871984e-05,
+      "loss": 0.7584,
+      "step": 15917
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5998740681986885,
+      "learning_rate": 1.438023633544288e-05,
+      "loss": 0.7152,
+      "step": 15918
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4843914778184606,
+      "learning_rate": 1.4379550376662799e-05,
+      "loss": 0.6915,
+      "step": 15919
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5711953183877836,
+      "learning_rate": 1.4378864392383585e-05,
+      "loss": 0.6618,
+      "step": 15920
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.396118100460585,
+      "learning_rate": 1.437817838260924e-05,
+      "loss": 0.7905,
+      "step": 15921
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5985062575624487,
+      "learning_rate": 1.4377492347343753e-05,
+      "loss": 0.541,
+      "step": 15922
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6067875720895133,
+      "learning_rate": 1.4376806286591118e-05,
+      "loss": 0.5642,
+      "step": 15923
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.086073269325558,
+      "learning_rate": 1.4376120200355334e-05,
+      "loss": 0.6432,
+      "step": 15924
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9062240855823704,
+      "learning_rate": 1.437543408864039e-05,
+      "loss": 0.8701,
+      "step": 15925
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.888149305130757,
+      "learning_rate": 1.4374747951450285e-05,
+      "loss": 0.6687,
+      "step": 15926
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5844657712878851,
+      "learning_rate": 1.4374061788789014e-05,
+      "loss": 0.578,
+      "step": 15927
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.519334183806057,
+      "learning_rate": 1.4373375600660567e-05,
+      "loss": 0.8346,
+      "step": 15928
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5202353939031092,
+      "learning_rate": 1.4372689387068943e-05,
+      "loss": 0.7802,
+      "step": 15929
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5953240244704733,
+      "learning_rate": 1.4372003148018139e-05,
+      "loss": 0.537,
+      "step": 15930
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5981364404364694,
+      "learning_rate": 1.4371316883512147e-05,
+      "loss": 0.5462,
+      "step": 15931
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5445341171322515,
+      "learning_rate": 1.4370630593554965e-05,
+      "loss": 0.8107,
+      "step": 15932
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6055536412640135,
+      "learning_rate": 1.4369944278150587e-05,
+      "loss": 0.7639,
+      "step": 15933
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4761237850358817,
+      "learning_rate": 1.4369257937303012e-05,
+      "loss": 0.6782,
+      "step": 15934
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4329263465324875,
+      "learning_rate": 1.4368571571016232e-05,
+      "loss": 0.8026,
+      "step": 15935
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4117172859572673,
+      "learning_rate": 1.4367885179294248e-05,
+      "loss": 0.7004,
+      "step": 15936
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4034428865149475,
+      "learning_rate": 1.436719876214105e-05,
+      "loss": 0.829,
+      "step": 15937
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5888395122307533,
+      "learning_rate": 1.4366512319560642e-05,
+      "loss": 0.8349,
+      "step": 15938
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4041981465283966,
+      "learning_rate": 1.4365825851557012e-05,
+      "loss": 0.7279,
+      "step": 15939
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5814008566903905,
+      "learning_rate": 1.4365139358134166e-05,
+      "loss": 0.5191,
+      "step": 15940
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5715416090112673,
+      "learning_rate": 1.4364452839296092e-05,
+      "loss": 0.5768,
+      "step": 15941
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3995535739950664,
+      "learning_rate": 1.4363766295046797e-05,
+      "loss": 0.8597,
+      "step": 15942
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.715300488378783,
+      "learning_rate": 1.436307972539027e-05,
+      "loss": 0.742,
+      "step": 15943
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3355870505491523,
+      "learning_rate": 1.4362393130330512e-05,
+      "loss": 0.669,
+      "step": 15944
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7351929992279442,
+      "learning_rate": 1.4361706509871523e-05,
+      "loss": 0.6755,
+      "step": 15945
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.599172641127179,
+      "learning_rate": 1.4361019864017293e-05,
+      "loss": 0.6956,
+      "step": 15946
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5403724967501635,
+      "learning_rate": 1.436033319277183e-05,
+      "loss": 0.7334,
+      "step": 15947
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6958209959855812,
+      "learning_rate": 1.4359646496139122e-05,
+      "loss": 0.7362,
+      "step": 15948
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.648483071658447,
+      "learning_rate": 1.4358959774123174e-05,
+      "loss": 0.813,
+      "step": 15949
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4644089844950674,
+      "learning_rate": 1.435827302672798e-05,
+      "loss": 0.6348,
+      "step": 15950
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.429590999409494,
+      "learning_rate": 1.4357586253957545e-05,
+      "loss": 0.7988,
+      "step": 15951
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5893427906143336,
+      "learning_rate": 1.435689945581586e-05,
+      "loss": 0.5019,
+      "step": 15952
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5487656348461214,
+      "learning_rate": 1.4356212632306927e-05,
+      "loss": 0.6808,
+      "step": 15953
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.476975596411885,
+      "learning_rate": 1.4355525783434746e-05,
+      "loss": 0.7164,
+      "step": 15954
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0229034341173904,
+      "learning_rate": 1.4354838909203315e-05,
+      "loss": 0.7505,
+      "step": 15955
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.666258324161367,
+      "learning_rate": 1.4354152009616634e-05,
+      "loss": 0.7865,
+      "step": 15956
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.422209033232901,
+      "learning_rate": 1.4353465084678697e-05,
+      "loss": 0.6969,
+      "step": 15957
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4650528445916713,
+      "learning_rate": 1.4352778134393513e-05,
+      "loss": 0.6636,
+      "step": 15958
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6007054214151917,
+      "learning_rate": 1.4352091158765074e-05,
+      "loss": 0.7989,
+      "step": 15959
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6136939889356814,
+      "learning_rate": 1.4351404157797383e-05,
+      "loss": 0.9422,
+      "step": 15960
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3646262749972333,
+      "learning_rate": 1.4350717131494438e-05,
+      "loss": 0.7594,
+      "step": 15961
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.4471191511094803,
+      "learning_rate": 1.4350030079860245e-05,
+      "loss": 0.7538,
+      "step": 15962
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8547772617299618,
+      "learning_rate": 1.4349343002898795e-05,
+      "loss": 0.789,
+      "step": 15963
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4617923110013864,
+      "learning_rate": 1.4348655900614094e-05,
+      "loss": 0.857,
+      "step": 15964
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4699355925307809,
+      "learning_rate": 1.4347968773010143e-05,
+      "loss": 0.727,
+      "step": 15965
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5337860660745035,
+      "learning_rate": 1.4347281620090942e-05,
+      "loss": 0.6965,
+      "step": 15966
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5944564893816031,
+      "learning_rate": 1.434659444186049e-05,
+      "loss": 0.7369,
+      "step": 15967
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6514734307275185,
+      "learning_rate": 1.434590723832279e-05,
+      "loss": 0.8172,
+      "step": 15968
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4199221410174725,
+      "learning_rate": 1.4345220009481837e-05,
+      "loss": 0.7533,
+      "step": 15969
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6709844118616832,
+      "learning_rate": 1.4344532755341645e-05,
+      "loss": 0.7795,
+      "step": 15970
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7876494074982885,
+      "learning_rate": 1.4343845475906201e-05,
+      "loss": 0.72,
+      "step": 15971
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3847570526023494,
+      "learning_rate": 1.434315817117952e-05,
+      "loss": 0.77,
+      "step": 15972
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3677416242735176,
+      "learning_rate": 1.4342470841165592e-05,
+      "loss": 0.6965,
+      "step": 15973
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5997523934267364,
+      "learning_rate": 1.4341783485868426e-05,
+      "loss": 0.7492,
+      "step": 15974
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4280888213821579,
+      "learning_rate": 1.4341096105292021e-05,
+      "loss": 0.73,
+      "step": 15975
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4626824871768942,
+      "learning_rate": 1.434040869944038e-05,
+      "loss": 0.5922,
+      "step": 15976
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3937452853790557,
+      "learning_rate": 1.4339721268317508e-05,
+      "loss": 0.7902,
+      "step": 15977
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6855154965661594,
+      "learning_rate": 1.43390338119274e-05,
+      "loss": 0.7655,
+      "step": 15978
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.485418082243302,
+      "learning_rate": 1.4338346330274067e-05,
+      "loss": 0.7104,
+      "step": 15979
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.383978045407972,
+      "learning_rate": 1.4337658823361509e-05,
+      "loss": 0.6938,
+      "step": 15980
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2986408770452522,
+      "learning_rate": 1.4336971291193726e-05,
+      "loss": 0.6613,
+      "step": 15981
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.575987158125168,
+      "learning_rate": 1.4336283733774722e-05,
+      "loss": 0.6822,
+      "step": 15982
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3404234901758914,
+      "learning_rate": 1.4335596151108504e-05,
+      "loss": 0.6746,
+      "step": 15983
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.530929011395292,
+      "learning_rate": 1.433490854319907e-05,
+      "loss": 0.8027,
+      "step": 15984
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6037727712168344,
+      "learning_rate": 1.4334220910050429e-05,
+      "loss": 0.7802,
+      "step": 15985
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.250747259561582,
+      "learning_rate": 1.433353325166658e-05,
+      "loss": 0.7026,
+      "step": 15986
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4244345953544644,
+      "learning_rate": 1.4332845568051532e-05,
+      "loss": 0.7161,
+      "step": 15987
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6531597701271271,
+      "learning_rate": 1.4332157859209278e-05,
+      "loss": 0.6057,
+      "step": 15988
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.732090583555195,
+      "learning_rate": 1.4331470125143835e-05,
+      "loss": 0.7618,
+      "step": 15989
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3252415182077792,
+      "learning_rate": 1.43307823658592e-05,
+      "loss": 0.7388,
+      "step": 15990
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7002040140197991,
+      "learning_rate": 1.433009458135938e-05,
+      "loss": 0.6227,
+      "step": 15991
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5870714972603295,
+      "learning_rate": 1.4329406771648375e-05,
+      "loss": 0.765,
+      "step": 15992
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4804357895521647,
+      "learning_rate": 1.4328718936730197e-05,
+      "loss": 0.764,
+      "step": 15993
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5991463601868114,
+      "learning_rate": 1.4328031076608847e-05,
+      "loss": 0.6895,
+      "step": 15994
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.599226079614886,
+      "learning_rate": 1.4327343191288325e-05,
+      "loss": 0.5166,
+      "step": 15995
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.722133269134823,
+      "learning_rate": 1.4326655280772648e-05,
+      "loss": 0.7546,
+      "step": 15996
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.398085236427764,
+      "learning_rate": 1.4325967345065808e-05,
+      "loss": 0.8532,
+      "step": 15997
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7605797404480534,
+      "learning_rate": 1.432527938417182e-05,
+      "loss": 0.7079,
+      "step": 15998
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.682914990923032,
+      "learning_rate": 1.4324591398094683e-05,
+      "loss": 0.7122,
+      "step": 15999
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3434751616790812,
+      "learning_rate": 1.4323903386838409e-05,
+      "loss": 0.7032,
+      "step": 16000
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5988876767501792,
+      "learning_rate": 1.4323215350406998e-05,
+      "loss": 0.5329,
+      "step": 16001
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7257046112340269,
+      "learning_rate": 1.4322527288804461e-05,
+      "loss": 0.7105,
+      "step": 16002
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3016381507819679,
+      "learning_rate": 1.4321839202034799e-05,
+      "loss": 0.7679,
+      "step": 16003
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4400761708660217,
+      "learning_rate": 1.4321151090102024e-05,
+      "loss": 0.7915,
+      "step": 16004
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3331807090999912,
+      "learning_rate": 1.4320462953010137e-05,
+      "loss": 0.7226,
+      "step": 16005
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3759250440650426,
+      "learning_rate": 1.4319774790763148e-05,
+      "loss": 0.7394,
+      "step": 16006
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3503323245418917,
+      "learning_rate": 1.4319086603365062e-05,
+      "loss": 0.8193,
+      "step": 16007
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3495007259477028,
+      "learning_rate": 1.4318398390819885e-05,
+      "loss": 0.6578,
+      "step": 16008
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5574442057813821,
+      "learning_rate": 1.4317710153131628e-05,
+      "loss": 0.5285,
+      "step": 16009
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.487033441486703,
+      "learning_rate": 1.4317021890304294e-05,
+      "loss": 0.8096,
+      "step": 16010
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.209003265143538,
+      "learning_rate": 1.4316333602341892e-05,
+      "loss": 0.7159,
+      "step": 16011
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5126417229732405,
+      "learning_rate": 1.431564528924843e-05,
+      "loss": 0.8819,
+      "step": 16012
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7446633071411128,
+      "learning_rate": 1.4314956951027914e-05,
+      "loss": 0.7635,
+      "step": 16013
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5142635407154037,
+      "learning_rate": 1.4314268587684353e-05,
+      "loss": 0.7267,
+      "step": 16014
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4066154160948798,
+      "learning_rate": 1.4313580199221754e-05,
+      "loss": 0.7299,
+      "step": 16015
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3855289519429965,
+      "learning_rate": 1.4312891785644124e-05,
+      "loss": 0.8836,
+      "step": 16016
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.693780536095302,
+      "learning_rate": 1.4312203346955478e-05,
+      "loss": 0.6897,
+      "step": 16017
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5998744964205412,
+      "learning_rate": 1.4311514883159812e-05,
+      "loss": 0.5382,
+      "step": 16018
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2520086577227654,
+      "learning_rate": 1.4310826394261148e-05,
+      "loss": 0.7169,
+      "step": 16019
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5941509639964073,
+      "learning_rate": 1.4310137880263481e-05,
+      "loss": 0.8066,
+      "step": 16020
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6129680413051073,
+      "learning_rate": 1.4309449341170832e-05,
+      "loss": 0.7233,
+      "step": 16021
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.510841047555686,
+      "learning_rate": 1.43087607769872e-05,
+      "loss": 0.751,
+      "step": 16022
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5787923757013261,
+      "learning_rate": 1.4308072187716604e-05,
+      "loss": 0.5237,
+      "step": 16023
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.457309739281546,
+      "learning_rate": 1.4307383573363044e-05,
+      "loss": 0.6673,
+      "step": 16024
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6780223283755773,
+      "learning_rate": 1.4306694933930537e-05,
+      "loss": 0.6694,
+      "step": 16025
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5008571880893455,
+      "learning_rate": 1.4306006269423082e-05,
+      "loss": 0.841,
+      "step": 16026
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6008446092395564,
+      "learning_rate": 1.4305317579844701e-05,
+      "loss": 0.5122,
+      "step": 16027
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6460437249077065,
+      "learning_rate": 1.4304628865199397e-05,
+      "loss": 0.8115,
+      "step": 16028
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2357762059107877,
+      "learning_rate": 1.4303940125491175e-05,
+      "loss": 0.7705,
+      "step": 16029
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3711922469371527,
+      "learning_rate": 1.430325136072406e-05,
+      "loss": 0.8684,
+      "step": 16030
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.603037795056382,
+      "learning_rate": 1.4302562570902048e-05,
+      "loss": 0.5284,
+      "step": 16031
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3041284309463876,
+      "learning_rate": 1.4301873756029156e-05,
+      "loss": 0.6474,
+      "step": 16032
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2336718292796016,
+      "learning_rate": 1.4301184916109395e-05,
+      "loss": 0.6835,
+      "step": 16033
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9428173807065567,
+      "learning_rate": 1.430049605114677e-05,
+      "loss": 0.8348,
+      "step": 16034
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.183397710506628,
+      "learning_rate": 1.4299807161145297e-05,
+      "loss": 0.7137,
+      "step": 16035
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3677520407240025,
+      "learning_rate": 1.4299118246108985e-05,
+      "loss": 0.7988,
+      "step": 16036
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5950197159108936,
+      "learning_rate": 1.4298429306041846e-05,
+      "loss": 0.8155,
+      "step": 16037
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.304367515442397,
+      "learning_rate": 1.4297740340947891e-05,
+      "loss": 0.6216,
+      "step": 16038
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5698231460699625,
+      "learning_rate": 1.429705135083113e-05,
+      "loss": 0.7894,
+      "step": 16039
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4872364971762762,
+      "learning_rate": 1.4296362335695578e-05,
+      "loss": 0.6617,
+      "step": 16040
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.1509519945519116,
+      "learning_rate": 1.429567329554524e-05,
+      "loss": 0.7052,
+      "step": 16041
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7015572457446149,
+      "learning_rate": 1.4294984230384135e-05,
+      "loss": 0.821,
+      "step": 16042
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5699212942723957,
+      "learning_rate": 1.429429514021627e-05,
+      "loss": 0.5307,
+      "step": 16043
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.147409914193048,
+      "learning_rate": 1.4293606025045663e-05,
+      "loss": 0.6542,
+      "step": 16044
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4811785900798986,
+      "learning_rate": 1.429291688487632e-05,
+      "loss": 0.6766,
+      "step": 16045
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4715593327125034,
+      "learning_rate": 1.4292227719712254e-05,
+      "loss": 0.6364,
+      "step": 16046
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9331803310504276,
+      "learning_rate": 1.4291538529557485e-05,
+      "loss": 0.7948,
+      "step": 16047
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8298638129125262,
+      "learning_rate": 1.4290849314416014e-05,
+      "loss": 0.6887,
+      "step": 16048
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4941432469397957,
+      "learning_rate": 1.4290160074291864e-05,
+      "loss": 0.851,
+      "step": 16049
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.421966093806039,
+      "learning_rate": 1.4289470809189039e-05,
+      "loss": 0.6716,
+      "step": 16050
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.473162347942591,
+      "learning_rate": 1.4288781519111563e-05,
+      "loss": 0.7779,
+      "step": 16051
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.633203100874446,
+      "learning_rate": 1.428809220406344e-05,
+      "loss": 0.8074,
+      "step": 16052
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.299592419875621,
+      "learning_rate": 1.4287402864048686e-05,
+      "loss": 0.7301,
+      "step": 16053
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6349917196562678,
+      "learning_rate": 1.4286713499071313e-05,
+      "loss": 0.8584,
+      "step": 16054
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3354775404847228,
+      "learning_rate": 1.428602410913534e-05,
+      "loss": 0.86,
+      "step": 16055
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3872201118168068,
+      "learning_rate": 1.4285334694244778e-05,
+      "loss": 0.7975,
+      "step": 16056
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4557457725977012,
+      "learning_rate": 1.4284645254403642e-05,
+      "loss": 0.8357,
+      "step": 16057
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.2791642003353862,
+      "learning_rate": 1.4283955789615942e-05,
+      "loss": 0.707,
+      "step": 16058
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5583012112673646,
+      "learning_rate": 1.4283266299885696e-05,
+      "loss": 0.7485,
+      "step": 16059
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.071406643811092,
+      "learning_rate": 1.4282576785216917e-05,
+      "loss": 0.7764,
+      "step": 16060
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.29133658050814,
+      "learning_rate": 1.4281887245613621e-05,
+      "loss": 0.8228,
+      "step": 16061
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8821950683093194,
+      "learning_rate": 1.4281197681079821e-05,
+      "loss": 0.7637,
+      "step": 16062
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.421583927491668,
+      "learning_rate": 1.4280508091619534e-05,
+      "loss": 0.5947,
+      "step": 16063
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3991735936452652,
+      "learning_rate": 1.4279818477236773e-05,
+      "loss": 0.7746,
+      "step": 16064
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6708904239437226,
+      "learning_rate": 1.4279128837935554e-05,
+      "loss": 0.7324,
+      "step": 16065
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6269367442034108,
+      "learning_rate": 1.4278439173719892e-05,
+      "loss": 0.7552,
+      "step": 16066
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.490276305149873,
+      "learning_rate": 1.4277749484593801e-05,
+      "loss": 0.7827,
+      "step": 16067
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5987387667858217,
+      "learning_rate": 1.4277059770561304e-05,
+      "loss": 0.8411,
+      "step": 16068
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4034758784708314,
+      "learning_rate": 1.4276370031626403e-05,
+      "loss": 0.8633,
+      "step": 16069
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3819498388193687,
+      "learning_rate": 1.4275680267793125e-05,
+      "loss": 0.6767,
+      "step": 16070
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4337599258126494,
+      "learning_rate": 1.4274990479065483e-05,
+      "loss": 0.7588,
+      "step": 16071
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4075058330438728,
+      "learning_rate": 1.4274300665447496e-05,
+      "loss": 0.7057,
+      "step": 16072
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4716955966728713,
+      "learning_rate": 1.4273610826943171e-05,
+      "loss": 0.7835,
+      "step": 16073
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5301566870033536,
+      "learning_rate": 1.4272920963556537e-05,
+      "loss": 0.5844,
+      "step": 16074
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.2207458338335817,
+      "learning_rate": 1.4272231075291603e-05,
+      "loss": 0.9053,
+      "step": 16075
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.465187339690325,
+      "learning_rate": 1.4271541162152385e-05,
+      "loss": 0.7869,
+      "step": 16076
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6218424722562955,
+      "learning_rate": 1.4270851224142904e-05,
+      "loss": 0.5653,
+      "step": 16077
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3878375429502223,
+      "learning_rate": 1.427016126126717e-05,
+      "loss": 0.7854,
+      "step": 16078
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6030630920360522,
+      "learning_rate": 1.4269471273529212e-05,
+      "loss": 0.5757,
+      "step": 16079
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5716100739581218,
+      "learning_rate": 1.4268781260933034e-05,
+      "loss": 0.7707,
+      "step": 16080
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3637315424912244,
+      "learning_rate": 1.4268091223482667e-05,
+      "loss": 0.6568,
+      "step": 16081
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4081218987631474,
+      "learning_rate": 1.4267401161182117e-05,
+      "loss": 0.714,
+      "step": 16082
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6040886431822234,
+      "learning_rate": 1.4266711074035408e-05,
+      "loss": 0.7929,
+      "step": 16083
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.836226710416059,
+      "learning_rate": 1.4266020962046555e-05,
+      "loss": 0.7044,
+      "step": 16084
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.182712210699509,
+      "learning_rate": 1.4265330825219578e-05,
+      "loss": 0.673,
+      "step": 16085
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.376433223419772,
+      "learning_rate": 1.4264640663558496e-05,
+      "loss": 0.769,
+      "step": 16086
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5667728568210653,
+      "learning_rate": 1.4263950477067326e-05,
+      "loss": 0.5297,
+      "step": 16087
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.459440651804257,
+      "learning_rate": 1.4263260265750085e-05,
+      "loss": 0.7951,
+      "step": 16088
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.604999100016584,
+      "learning_rate": 1.4262570029610792e-05,
+      "loss": 0.8406,
+      "step": 16089
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4473611704244267,
+      "learning_rate": 1.4261879768653469e-05,
+      "loss": 0.9144,
+      "step": 16090
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3801690165562375,
+      "learning_rate": 1.4261189482882131e-05,
+      "loss": 0.7632,
+      "step": 16091
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.435484843369511,
+      "learning_rate": 1.4260499172300798e-05,
+      "loss": 0.7668,
+      "step": 16092
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5405314863239183,
+      "learning_rate": 1.425980883691349e-05,
+      "loss": 0.7694,
+      "step": 16093
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.3843430852593337,
+      "learning_rate": 1.425911847672423e-05,
+      "loss": 0.7743,
+      "step": 16094
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6365899449489645,
+      "learning_rate": 1.4258428091737031e-05,
+      "loss": 0.7989,
+      "step": 16095
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2813373438760818,
+      "learning_rate": 1.425773768195592e-05,
+      "loss": 0.7006,
+      "step": 16096
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4802429744224366,
+      "learning_rate": 1.4257047247384904e-05,
+      "loss": 0.7952,
+      "step": 16097
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5166961092725457,
+      "learning_rate": 1.4256356788028018e-05,
+      "loss": 0.8912,
+      "step": 16098
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5102137745434763,
+      "learning_rate": 1.425566630388927e-05,
+      "loss": 0.6933,
+      "step": 16099
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6172847337404352,
+      "learning_rate": 1.425497579497269e-05,
+      "loss": 0.6483,
+      "step": 16100
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.606043087069201,
+      "learning_rate": 1.425428526128229e-05,
+      "loss": 0.7564,
+      "step": 16101
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4476309945864572,
+      "learning_rate": 1.42535947028221e-05,
+      "loss": 0.794,
+      "step": 16102
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5351387775686562,
+      "learning_rate": 1.4252904119596129e-05,
+      "loss": 0.6696,
+      "step": 16103
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3271105242377566,
+      "learning_rate": 1.4252213511608406e-05,
+      "loss": 0.7265,
+      "step": 16104
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4102546144797976,
+      "learning_rate": 1.425152287886295e-05,
+      "loss": 0.7594,
+      "step": 16105
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3574807244775458,
+      "learning_rate": 1.4250832221363782e-05,
+      "loss": 0.7512,
+      "step": 16106
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.505753433265249,
+      "learning_rate": 1.425014153911492e-05,
+      "loss": 0.8061,
+      "step": 16107
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2870907910827694,
+      "learning_rate": 1.4249450832120393e-05,
+      "loss": 0.6845,
+      "step": 16108
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3867600372184876,
+      "learning_rate": 1.4248760100384216e-05,
+      "loss": 0.6912,
+      "step": 16109
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3131511323972753,
+      "learning_rate": 1.4248069343910412e-05,
+      "loss": 0.7372,
+      "step": 16110
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4031100487465038,
+      "learning_rate": 1.4247378562703004e-05,
+      "loss": 0.8012,
+      "step": 16111
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9106635752351415,
+      "learning_rate": 1.4246687756766013e-05,
+      "loss": 0.8322,
+      "step": 16112
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.467369173246092,
+      "learning_rate": 1.424599692610346e-05,
+      "loss": 0.9295,
+      "step": 16113
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.461200489434112,
+      "learning_rate": 1.424530607071937e-05,
+      "loss": 0.7769,
+      "step": 16114
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3806803947219481,
+      "learning_rate": 1.4244615190617765e-05,
+      "loss": 0.7211,
+      "step": 16115
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.584168935597618,
+      "learning_rate": 1.4243924285802665e-05,
+      "loss": 0.7575,
+      "step": 16116
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4143384129494692,
+      "learning_rate": 1.4243233356278097e-05,
+      "loss": 0.7215,
+      "step": 16117
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.491182875607309,
+      "learning_rate": 1.4242542402048076e-05,
+      "loss": 0.7585,
+      "step": 16118
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.384129523823234,
+      "learning_rate": 1.4241851423116634e-05,
+      "loss": 0.7862,
+      "step": 16119
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.904648044405524,
+      "learning_rate": 1.4241160419487789e-05,
+      "loss": 0.7482,
+      "step": 16120
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5492208474959488,
+      "learning_rate": 1.4240469391165569e-05,
+      "loss": 0.7361,
+      "step": 16121
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5442514125357885,
+      "learning_rate": 1.4239778338153984e-05,
+      "loss": 0.738,
+      "step": 16122
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.734993158734338,
+      "learning_rate": 1.4239087260457076e-05,
+      "loss": 0.6856,
+      "step": 16123
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4892912198544797,
+      "learning_rate": 1.4238396158078859e-05,
+      "loss": 0.7772,
+      "step": 16124
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.196326402431949,
+      "learning_rate": 1.4237705031023354e-05,
+      "loss": 0.7028,
+      "step": 16125
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6560724883121607,
+      "learning_rate": 1.423701387929459e-05,
+      "loss": 0.9485,
+      "step": 16126
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.88835773164118,
+      "learning_rate": 1.4236322702896588e-05,
+      "loss": 0.66,
+      "step": 16127
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.889256312817766,
+      "learning_rate": 1.4235631501833378e-05,
+      "loss": 0.7097,
+      "step": 16128
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4671334867429908,
+      "learning_rate": 1.4234940276108977e-05,
+      "loss": 0.8236,
+      "step": 16129
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.672673549445992,
+      "learning_rate": 1.4234249025727419e-05,
+      "loss": 0.7229,
+      "step": 16130
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1180141753817674,
+      "learning_rate": 1.4233557750692715e-05,
+      "loss": 0.6165,
+      "step": 16131
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6582054738180028,
+      "learning_rate": 1.4232866451008902e-05,
+      "loss": 0.7448,
+      "step": 16132
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5320018793196477,
+      "learning_rate": 1.4232175126679999e-05,
+      "loss": 0.6954,
+      "step": 16133
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5024484199491785,
+      "learning_rate": 1.423148377771003e-05,
+      "loss": 0.7521,
+      "step": 16134
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5914948912046213,
+      "learning_rate": 1.4230792404103026e-05,
+      "loss": 0.5159,
+      "step": 16135
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.878504885759681,
+      "learning_rate": 1.4230101005863008e-05,
+      "loss": 0.8221,
+      "step": 16136
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7847623459369473,
+      "learning_rate": 1.4229409582994005e-05,
+      "loss": 0.604,
+      "step": 16137
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5405649081228237,
+      "learning_rate": 1.422871813550004e-05,
+      "loss": 0.7085,
+      "step": 16138
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6229449422260516,
+      "learning_rate": 1.4228026663385136e-05,
+      "loss": 0.6754,
+      "step": 16139
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.723399766966532,
+      "learning_rate": 1.4227335166653325e-05,
+      "loss": 0.7701,
+      "step": 16140
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5752658096339737,
+      "learning_rate": 1.4226643645308628e-05,
+      "loss": 0.5214,
+      "step": 16141
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.50364545438627,
+      "learning_rate": 1.4225952099355075e-05,
+      "loss": 0.6825,
+      "step": 16142
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6357066096521848,
+      "learning_rate": 1.4225260528796692e-05,
+      "loss": 0.746,
+      "step": 16143
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5710971053796527,
+      "learning_rate": 1.42245689336375e-05,
+      "loss": 0.7483,
+      "step": 16144
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2852691371903782,
+      "learning_rate": 1.4223877313881535e-05,
+      "loss": 0.6607,
+      "step": 16145
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9809760261322644,
+      "learning_rate": 1.422318566953282e-05,
+      "loss": 0.7182,
+      "step": 16146
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3053836618605943,
+      "learning_rate": 1.422249400059538e-05,
+      "loss": 0.6822,
+      "step": 16147
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9496503631875597,
+      "learning_rate": 1.422180230707324e-05,
+      "loss": 0.7799,
+      "step": 16148
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4993730583296425,
+      "learning_rate": 1.4221110588970432e-05,
+      "loss": 0.8787,
+      "step": 16149
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.656028139472972,
+      "learning_rate": 1.4220418846290981e-05,
+      "loss": 0.748,
+      "step": 16150
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.125845558611564,
+      "learning_rate": 1.4219727079038919e-05,
+      "loss": 0.7604,
+      "step": 16151
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5843554839943373,
+      "learning_rate": 1.4219035287218264e-05,
+      "loss": 0.5212,
+      "step": 16152
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4544513770702823,
+      "learning_rate": 1.4218343470833055e-05,
+      "loss": 0.7025,
+      "step": 16153
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2494481805047983,
+      "learning_rate": 1.4217651629887312e-05,
+      "loss": 0.6844,
+      "step": 16154
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.343304529217663,
+      "learning_rate": 1.4216959764385067e-05,
+      "loss": 0.669,
+      "step": 16155
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5922510944223451,
+      "learning_rate": 1.4216267874330347e-05,
+      "loss": 0.5726,
+      "step": 16156
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0934475513601645,
+      "learning_rate": 1.421557595972718e-05,
+      "loss": 0.7688,
+      "step": 16157
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.673363776948235,
+      "learning_rate": 1.4214884020579595e-05,
+      "loss": 0.8668,
+      "step": 16158
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4684707984837386,
+      "learning_rate": 1.4214192056891622e-05,
+      "loss": 0.6486,
+      "step": 16159
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2661670799880662,
+      "learning_rate": 1.421350006866729e-05,
+      "loss": 0.7123,
+      "step": 16160
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4811093798328632,
+      "learning_rate": 1.4212808055910623e-05,
+      "loss": 0.7236,
+      "step": 16161
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6304192048329829,
+      "learning_rate": 1.4212116018625658e-05,
+      "loss": 0.8035,
+      "step": 16162
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6014865016842176,
+      "learning_rate": 1.4211423956816415e-05,
+      "loss": 0.5814,
+      "step": 16163
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2926693055504812,
+      "learning_rate": 1.4210731870486931e-05,
+      "loss": 0.7503,
+      "step": 16164
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.785039966267082,
+      "learning_rate": 1.4210039759641232e-05,
+      "loss": 0.702,
+      "step": 16165
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.501419671898829,
+      "learning_rate": 1.4209347624283352e-05,
+      "loss": 0.6287,
+      "step": 16166
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5842746470010791,
+      "learning_rate": 1.4208655464417313e-05,
+      "loss": 0.7948,
+      "step": 16167
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7044559574106941,
+      "learning_rate": 1.4207963280047155e-05,
+      "loss": 0.8161,
+      "step": 16168
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.497335568531372,
+      "learning_rate": 1.4207271071176896e-05,
+      "loss": 0.78,
+      "step": 16169
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4941543240713908,
+      "learning_rate": 1.4206578837810578e-05,
+      "loss": 0.7359,
+      "step": 16170
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6061826640982156,
+      "learning_rate": 1.4205886579952221e-05,
+      "loss": 0.503,
+      "step": 16171
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3765907451077886,
+      "learning_rate": 1.4205194297605867e-05,
+      "loss": 0.7856,
+      "step": 16172
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3967675197306582,
+      "learning_rate": 1.4204501990775535e-05,
+      "loss": 0.7905,
+      "step": 16173
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2590298105793882,
+      "learning_rate": 1.4203809659465266e-05,
+      "loss": 0.7005,
+      "step": 16174
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5647934709067812,
+      "learning_rate": 1.4203117303679083e-05,
+      "loss": 0.6294,
+      "step": 16175
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.532758330019235,
+      "learning_rate": 1.4202424923421018e-05,
+      "loss": 0.7019,
+      "step": 16176
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5876572216190715,
+      "learning_rate": 1.420173251869511e-05,
+      "loss": 0.5689,
+      "step": 16177
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6885635120961795,
+      "learning_rate": 1.4201040089505383e-05,
+      "loss": 0.7788,
+      "step": 16178
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4580442666522302,
+      "learning_rate": 1.420034763585587e-05,
+      "loss": 0.6063,
+      "step": 16179
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7960272812824167,
+      "learning_rate": 1.4199655157750602e-05,
+      "loss": 0.8648,
+      "step": 16180
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4725830713855161,
+      "learning_rate": 1.4198962655193617e-05,
+      "loss": 0.862,
+      "step": 16181
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4872019601510533,
+      "learning_rate": 1.4198270128188938e-05,
+      "loss": 0.7377,
+      "step": 16182
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5584379199364036,
+      "learning_rate": 1.4197577576740602e-05,
+      "loss": 0.7896,
+      "step": 16183
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.0740776861471097,
+      "learning_rate": 1.419688500085264e-05,
+      "loss": 0.7886,
+      "step": 16184
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7209126343522103,
+      "learning_rate": 1.4196192400529088e-05,
+      "loss": 0.6072,
+      "step": 16185
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6331472405490841,
+      "learning_rate": 1.4195499775773972e-05,
+      "loss": 0.7565,
+      "step": 16186
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.095630521785416,
+      "learning_rate": 1.4194807126591328e-05,
+      "loss": 0.6188,
+      "step": 16187
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2970695969330728,
+      "learning_rate": 1.4194114452985189e-05,
+      "loss": 0.6788,
+      "step": 16188
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.702430742398258,
+      "learning_rate": 1.4193421754959586e-05,
+      "loss": 0.6687,
+      "step": 16189
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4657013741905582,
+      "learning_rate": 1.4192729032518557e-05,
+      "loss": 0.6954,
+      "step": 16190
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4951893095219049,
+      "learning_rate": 1.4192036285666133e-05,
+      "loss": 0.7939,
+      "step": 16191
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4265780605147322,
+      "learning_rate": 1.4191343514406345e-05,
+      "loss": 0.8052,
+      "step": 16192
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3718365288722123,
+      "learning_rate": 1.4190650718743229e-05,
+      "loss": 0.7903,
+      "step": 16193
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4099127768849367,
+      "learning_rate": 1.4189957898680815e-05,
+      "loss": 0.763,
+      "step": 16194
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5128713088711727,
+      "learning_rate": 1.4189265054223142e-05,
+      "loss": 0.8406,
+      "step": 16195
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3539758160714437,
+      "learning_rate": 1.4188572185374241e-05,
+      "loss": 0.7176,
+      "step": 16196
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.275118972888534,
+      "learning_rate": 1.4187879292138144e-05,
+      "loss": 0.7459,
+      "step": 16197
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6846075155980667,
+      "learning_rate": 1.4187186374518893e-05,
+      "loss": 0.7281,
+      "step": 16198
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3093229840395593,
+      "learning_rate": 1.4186493432520513e-05,
+      "loss": 0.642,
+      "step": 16199
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5104540754733147,
+      "learning_rate": 1.4185800466147048e-05,
+      "loss": 0.7579,
+      "step": 16200
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8262547016405448,
+      "learning_rate": 1.4185107475402522e-05,
+      "loss": 0.6026,
+      "step": 16201
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7203840032443918,
+      "learning_rate": 1.4184414460290979e-05,
+      "loss": 0.7151,
+      "step": 16202
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3534045512347554,
+      "learning_rate": 1.4183721420816448e-05,
+      "loss": 0.6572,
+      "step": 16203
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5391651885979063,
+      "learning_rate": 1.4183028356982969e-05,
+      "loss": 0.7926,
+      "step": 16204
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5411132528708464,
+      "learning_rate": 1.4182335268794571e-05,
+      "loss": 0.694,
+      "step": 16205
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6018666648298328,
+      "learning_rate": 1.4181642156255295e-05,
+      "loss": 0.8153,
+      "step": 16206
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3657777274233789,
+      "learning_rate": 1.4180949019369174e-05,
+      "loss": 0.643,
+      "step": 16207
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4562414615264645,
+      "learning_rate": 1.4180255858140243e-05,
+      "loss": 0.7348,
+      "step": 16208
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8325622054711985,
+      "learning_rate": 1.417956267257254e-05,
+      "loss": 0.6696,
+      "step": 16209
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3576492051246738,
+      "learning_rate": 1.4178869462670102e-05,
+      "loss": 0.7321,
+      "step": 16210
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.553724363875251,
+      "learning_rate": 1.417817622843696e-05,
+      "loss": 0.754,
+      "step": 16211
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5653171868017997,
+      "learning_rate": 1.4177482969877155e-05,
+      "loss": 0.7482,
+      "step": 16212
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6391749590674226,
+      "learning_rate": 1.4176789686994718e-05,
+      "loss": 0.5046,
+      "step": 16213
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4086956677587237,
+      "learning_rate": 1.4176096379793694e-05,
+      "loss": 0.7538,
+      "step": 16214
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6368473883231145,
+      "learning_rate": 1.4175403048278113e-05,
+      "loss": 0.8292,
+      "step": 16215
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5816079998804984,
+      "learning_rate": 1.4174709692452014e-05,
+      "loss": 0.6751,
+      "step": 16216
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4860262120841274,
+      "learning_rate": 1.4174016312319432e-05,
+      "loss": 0.7704,
+      "step": 16217
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4991381353292996,
+      "learning_rate": 1.4173322907884406e-05,
+      "loss": 0.6273,
+      "step": 16218
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3345958320583788,
+      "learning_rate": 1.4172629479150975e-05,
+      "loss": 0.6863,
+      "step": 16219
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5691685957361942,
+      "learning_rate": 1.417193602612317e-05,
+      "loss": 0.5719,
+      "step": 16220
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4083045498005984,
+      "learning_rate": 1.4171242548805036e-05,
+      "loss": 0.7569,
+      "step": 16221
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4696477860585877,
+      "learning_rate": 1.4170549047200604e-05,
+      "loss": 0.7167,
+      "step": 16222
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5010225443069112,
+      "learning_rate": 1.416985552131392e-05,
+      "loss": 0.6739,
+      "step": 16223
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6237279981685101,
+      "learning_rate": 1.416916197114901e-05,
+      "loss": 0.549,
+      "step": 16224
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.459229627490039,
+      "learning_rate": 1.4168468396709925e-05,
+      "loss": 0.7509,
+      "step": 16225
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5951135825406977,
+      "learning_rate": 1.4167774798000697e-05,
+      "loss": 0.5204,
+      "step": 16226
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4070075996425018,
+      "learning_rate": 1.416708117502536e-05,
+      "loss": 0.6902,
+      "step": 16227
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.771034383396039,
+      "learning_rate": 1.4166387527787964e-05,
+      "loss": 0.7666,
+      "step": 16228
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4878462988495933,
+      "learning_rate": 1.4165693856292535e-05,
+      "loss": 0.6601,
+      "step": 16229
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3508733766559367,
+      "learning_rate": 1.416500016054312e-05,
+      "loss": 0.7116,
+      "step": 16230
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3207321534114183,
+      "learning_rate": 1.4164306440543754e-05,
+      "loss": 0.8004,
+      "step": 16231
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4845458458432397,
+      "learning_rate": 1.4163612696298481e-05,
+      "loss": 0.749,
+      "step": 16232
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5306750240675124,
+      "learning_rate": 1.4162918927811334e-05,
+      "loss": 0.7308,
+      "step": 16233
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4058466072586198,
+      "learning_rate": 1.4162225135086354e-05,
+      "loss": 0.6073,
+      "step": 16234
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.472417927877805,
+      "learning_rate": 1.4161531318127584e-05,
+      "loss": 0.7164,
+      "step": 16235
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3297595729466778,
+      "learning_rate": 1.416083747693906e-05,
+      "loss": 0.7097,
+      "step": 16236
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3995637041887172,
+      "learning_rate": 1.4160143611524821e-05,
+      "loss": 0.7895,
+      "step": 16237
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6955315049325537,
+      "learning_rate": 1.4159449721888911e-05,
+      "loss": 0.729,
+      "step": 16238
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7929829883600943,
+      "learning_rate": 1.4158755808035367e-05,
+      "loss": 0.6766,
+      "step": 16239
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5947515788580654,
+      "learning_rate": 1.4158061869968232e-05,
+      "loss": 0.5276,
+      "step": 16240
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7846434196050511,
+      "learning_rate": 1.4157367907691542e-05,
+      "loss": 0.7146,
+      "step": 16241
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.385209490287038,
+      "learning_rate": 1.4156673921209341e-05,
+      "loss": 0.7029,
+      "step": 16242
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.444569605418989,
+      "learning_rate": 1.4155979910525668e-05,
+      "loss": 0.7954,
+      "step": 16243
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4522608897543268,
+      "learning_rate": 1.4155285875644564e-05,
+      "loss": 0.769,
+      "step": 16244
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6453044952743199,
+      "learning_rate": 1.4154591816570072e-05,
+      "loss": 0.798,
+      "step": 16245
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8014953586809055,
+      "learning_rate": 1.4153897733306228e-05,
+      "loss": 0.7885,
+      "step": 16246
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2944294182141847,
+      "learning_rate": 1.415320362585708e-05,
+      "loss": 0.6888,
+      "step": 16247
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7246067141872026,
+      "learning_rate": 1.4152509494226662e-05,
+      "loss": 0.8872,
+      "step": 16248
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4929092776366661,
+      "learning_rate": 1.4151815338419022e-05,
+      "loss": 0.5861,
+      "step": 16249
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4301831522743012,
+      "learning_rate": 1.4151121158438195e-05,
+      "loss": 0.7226,
+      "step": 16250
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.548552573977935,
+      "learning_rate": 1.415042695428823e-05,
+      "loss": 0.7481,
+      "step": 16251
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4474642223355112,
+      "learning_rate": 1.4149732725973162e-05,
+      "loss": 0.6278,
+      "step": 16252
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4157114833711653,
+      "learning_rate": 1.414903847349704e-05,
+      "loss": 0.7553,
+      "step": 16253
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7547529511403155,
+      "learning_rate": 1.41483441968639e-05,
+      "loss": 0.7743,
+      "step": 16254
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5813766232299218,
+      "learning_rate": 1.4147649896077789e-05,
+      "loss": 0.6204,
+      "step": 16255
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9041615973085542,
+      "learning_rate": 1.4146955571142745e-05,
+      "loss": 0.8314,
+      "step": 16256
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4768497805144378,
+      "learning_rate": 1.4146261222062813e-05,
+      "loss": 0.781,
+      "step": 16257
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.3482536757374644,
+      "learning_rate": 1.4145566848842036e-05,
+      "loss": 0.7006,
+      "step": 16258
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4730336530142796,
+      "learning_rate": 1.4144872451484455e-05,
+      "loss": 0.8092,
+      "step": 16259
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6749333258830839,
+      "learning_rate": 1.4144178029994118e-05,
+      "loss": 0.7619,
+      "step": 16260
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5146422805697888,
+      "learning_rate": 1.414348358437506e-05,
+      "loss": 0.8586,
+      "step": 16261
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3416108365706052,
+      "learning_rate": 1.414278911463133e-05,
+      "loss": 0.8643,
+      "step": 16262
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.439453490199715,
+      "learning_rate": 1.4142094620766972e-05,
+      "loss": 0.8327,
+      "step": 16263
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5417692835569181,
+      "learning_rate": 1.4141400102786027e-05,
+      "loss": 0.7537,
+      "step": 16264
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5808115896359027,
+      "learning_rate": 1.4140705560692538e-05,
+      "loss": 0.5364,
+      "step": 16265
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5754371926200967,
+      "learning_rate": 1.4140010994490552e-05,
+      "loss": 0.5423,
+      "step": 16266
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3972815022853398,
+      "learning_rate": 1.4139316404184111e-05,
+      "loss": 0.5991,
+      "step": 16267
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.703551144777405,
+      "learning_rate": 1.4138621789777258e-05,
+      "loss": 0.7879,
+      "step": 16268
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5961870660055655,
+      "learning_rate": 1.413792715127404e-05,
+      "loss": 0.5143,
+      "step": 16269
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0852251878019277,
+      "learning_rate": 1.4137232488678503e-05,
+      "loss": 0.7953,
+      "step": 16270
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.789631185258019,
+      "learning_rate": 1.4136537801994683e-05,
+      "loss": 0.8521,
+      "step": 16271
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.338879957473392,
+      "learning_rate": 1.4135843091226635e-05,
+      "loss": 0.718,
+      "step": 16272
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6509551402846849,
+      "learning_rate": 1.4135148356378393e-05,
+      "loss": 0.4769,
+      "step": 16273
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6015263455132773,
+      "learning_rate": 1.4134453597454015e-05,
+      "loss": 0.8061,
+      "step": 16274
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4116181596665616,
+      "learning_rate": 1.4133758814457534e-05,
+      "loss": 0.7437,
+      "step": 16275
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.920047761010938,
+      "learning_rate": 1.4133064007393005e-05,
+      "loss": 0.7331,
+      "step": 16276
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.425309392634583,
+      "learning_rate": 1.4132369176264467e-05,
+      "loss": 0.7341,
+      "step": 16277
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2002493478712521,
+      "learning_rate": 1.4131674321075966e-05,
+      "loss": 0.6252,
+      "step": 16278
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5400631024681661,
+      "learning_rate": 1.4130979441831552e-05,
+      "loss": 0.6842,
+      "step": 16279
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3932692713891903,
+      "learning_rate": 1.4130284538535261e-05,
+      "loss": 0.611,
+      "step": 16280
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4207980527066002,
+      "learning_rate": 1.4129589611191154e-05,
+      "loss": 0.7901,
+      "step": 16281
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7368945568301237,
+      "learning_rate": 1.4128894659803263e-05,
+      "loss": 0.8161,
+      "step": 16282
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.51250516733134,
+      "learning_rate": 1.4128199684375645e-05,
+      "loss": 0.6739,
+      "step": 16283
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5350005275173517,
+      "learning_rate": 1.412750468491234e-05,
+      "loss": 0.8222,
+      "step": 16284
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3649084293514229,
+      "learning_rate": 1.4126809661417396e-05,
+      "loss": 0.9068,
+      "step": 16285
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6389301407817056,
+      "learning_rate": 1.4126114613894857e-05,
+      "loss": 0.7932,
+      "step": 16286
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5255999899865038,
+      "learning_rate": 1.4125419542348775e-05,
+      "loss": 0.7108,
+      "step": 16287
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4469419953904148,
+      "learning_rate": 1.4124724446783194e-05,
+      "loss": 0.7576,
+      "step": 16288
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.2481761605536947,
+      "learning_rate": 1.412402932720216e-05,
+      "loss": 0.7752,
+      "step": 16289
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4689482192882934,
+      "learning_rate": 1.412333418360972e-05,
+      "loss": 0.7631,
+      "step": 16290
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5736584238208797,
+      "learning_rate": 1.4122639016009927e-05,
+      "loss": 0.5822,
+      "step": 16291
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4181075407713708,
+      "learning_rate": 1.4121943824406822e-05,
+      "loss": 0.818,
+      "step": 16292
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4535455747031198,
+      "learning_rate": 1.4121248608804456e-05,
+      "loss": 0.8693,
+      "step": 16293
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5825081699718097,
+      "learning_rate": 1.4120553369206876e-05,
+      "loss": 0.5105,
+      "step": 16294
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3632113014183418,
+      "learning_rate": 1.4119858105618129e-05,
+      "loss": 0.7021,
+      "step": 16295
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4684806166355757,
+      "learning_rate": 1.4119162818042265e-05,
+      "loss": 0.7879,
+      "step": 16296
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7086898474728567,
+      "learning_rate": 1.411846750648333e-05,
+      "loss": 0.7654,
+      "step": 16297
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5878846197382044,
+      "learning_rate": 1.4117772170945375e-05,
+      "loss": 0.7271,
+      "step": 16298
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7036587756002999,
+      "learning_rate": 1.4117076811432444e-05,
+      "loss": 0.6786,
+      "step": 16299
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6058651064954248,
+      "learning_rate": 1.4116381427948591e-05,
+      "loss": 0.6628,
+      "step": 16300
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.62970092647383,
+      "learning_rate": 1.4115686020497858e-05,
+      "loss": 0.8142,
+      "step": 16301
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5244089646524515,
+      "learning_rate": 1.4114990589084304e-05,
+      "loss": 0.7062,
+      "step": 16302
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4432044455527877,
+      "learning_rate": 1.4114295133711966e-05,
+      "loss": 0.7029,
+      "step": 16303
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7452155278484933,
+      "learning_rate": 1.4113599654384904e-05,
+      "loss": 0.6334,
+      "step": 16304
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5728952692348925,
+      "learning_rate": 1.411290415110716e-05,
+      "loss": 0.7376,
+      "step": 16305
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.8252752410383275,
+      "learning_rate": 1.4112208623882787e-05,
+      "loss": 0.7677,
+      "step": 16306
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6933091671390599,
+      "learning_rate": 1.4111513072715833e-05,
+      "loss": 0.6845,
+      "step": 16307
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5634316704088083,
+      "learning_rate": 1.4110817497610347e-05,
+      "loss": 0.8275,
+      "step": 16308
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3813060267717236,
+      "learning_rate": 1.4110121898570382e-05,
+      "loss": 0.6295,
+      "step": 16309
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2951060178571343,
+      "learning_rate": 1.4109426275599983e-05,
+      "loss": 0.6929,
+      "step": 16310
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.538095927527697,
+      "learning_rate": 1.4108730628703207e-05,
+      "loss": 0.6713,
+      "step": 16311
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.721952384454086,
+      "learning_rate": 1.4108034957884096e-05,
+      "loss": 0.7438,
+      "step": 16312
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.525783679060158,
+      "learning_rate": 1.4107339263146709e-05,
+      "loss": 0.7222,
+      "step": 16313
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.2504590805523907,
+      "learning_rate": 1.4106643544495092e-05,
+      "loss": 0.7138,
+      "step": 16314
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5954028866014228,
+      "learning_rate": 1.4105947801933295e-05,
+      "loss": 0.5502,
+      "step": 16315
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3970537826076361,
+      "learning_rate": 1.4105252035465368e-05,
+      "loss": 0.7226,
+      "step": 16316
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.59199754309081,
+      "learning_rate": 1.4104556245095368e-05,
+      "loss": 0.738,
+      "step": 16317
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5479584601715857,
+      "learning_rate": 1.4103860430827342e-05,
+      "loss": 0.8178,
+      "step": 16318
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.494149023251368,
+      "learning_rate": 1.4103164592665337e-05,
+      "loss": 0.8261,
+      "step": 16319
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.317373107157913,
+      "learning_rate": 1.4102468730613412e-05,
+      "loss": 0.7551,
+      "step": 16320
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.567013788491904,
+      "learning_rate": 1.4101772844675617e-05,
+      "loss": 0.7033,
+      "step": 16321
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4547492538900686,
+      "learning_rate": 1.4101076934855997e-05,
+      "loss": 0.9108,
+      "step": 16322
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2362296183502435,
+      "learning_rate": 1.4100381001158612e-05,
+      "loss": 0.7811,
+      "step": 16323
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.675005899051784,
+      "learning_rate": 1.409968504358751e-05,
+      "loss": 0.7819,
+      "step": 16324
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.354059888438249,
+      "learning_rate": 1.4098989062146741e-05,
+      "loss": 0.7575,
+      "step": 16325
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6428243113323755,
+      "learning_rate": 1.4098293056840364e-05,
+      "loss": 0.779,
+      "step": 16326
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5800320801648878,
+      "learning_rate": 1.4097597027672425e-05,
+      "loss": 0.5177,
+      "step": 16327
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6221045282977702,
+      "learning_rate": 1.409690097464698e-05,
+      "loss": 0.5703,
+      "step": 16328
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4834054826227927,
+      "learning_rate": 1.409620489776808e-05,
+      "loss": 0.7091,
+      "step": 16329
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.575707452706526,
+      "learning_rate": 1.4095508797039781e-05,
+      "loss": 0.7139,
+      "step": 16330
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4562618204548294,
+      "learning_rate": 1.4094812672466127e-05,
+      "loss": 0.675,
+      "step": 16331
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.666890135925625,
+      "learning_rate": 1.4094116524051182e-05,
+      "loss": 0.7797,
+      "step": 16332
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3800419053574169,
+      "learning_rate": 1.4093420351798991e-05,
+      "loss": 0.6463,
+      "step": 16333
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4550996016451145,
+      "learning_rate": 1.4092724155713614e-05,
+      "loss": 0.7588,
+      "step": 16334
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5431987842316564,
+      "learning_rate": 1.4092027935799102e-05,
+      "loss": 0.7865,
+      "step": 16335
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5760641677487237,
+      "learning_rate": 1.4091331692059506e-05,
+      "loss": 0.7174,
+      "step": 16336
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.553280538870317,
+      "learning_rate": 1.4090635424498882e-05,
+      "loss": 0.815,
+      "step": 16337
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2702206990288223,
+      "learning_rate": 1.4089939133121282e-05,
+      "loss": 0.8099,
+      "step": 16338
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7941449452031704,
+      "learning_rate": 1.4089242817930764e-05,
+      "loss": 0.8032,
+      "step": 16339
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3281641242960487,
+      "learning_rate": 1.4088546478931378e-05,
+      "loss": 0.6731,
+      "step": 16340
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5604803173763322,
+      "learning_rate": 1.408785011612718e-05,
+      "loss": 0.756,
+      "step": 16341
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5897115984381518,
+      "learning_rate": 1.4087153729522224e-05,
+      "loss": 0.6892,
+      "step": 16342
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9917515001632715,
+      "learning_rate": 1.4086457319120566e-05,
+      "loss": 0.6888,
+      "step": 16343
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4120216033816655,
+      "learning_rate": 1.408576088492626e-05,
+      "loss": 0.7392,
+      "step": 16344
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.842451442659336,
+      "learning_rate": 1.408506442694336e-05,
+      "loss": 0.6987,
+      "step": 16345
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7578526792056919,
+      "learning_rate": 1.408436794517592e-05,
+      "loss": 0.7149,
+      "step": 16346
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.004252227550618,
+      "learning_rate": 1.4083671439628e-05,
+      "loss": 0.8676,
+      "step": 16347
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3120080948846329,
+      "learning_rate": 1.4082974910303648e-05,
+      "loss": 0.8022,
+      "step": 16348
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6067582922124619,
+      "learning_rate": 1.4082278357206928e-05,
+      "loss": 0.5464,
+      "step": 16349
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6476334124510286,
+      "learning_rate": 1.4081581780341887e-05,
+      "loss": 0.9245,
+      "step": 16350
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6187034214995533,
+      "learning_rate": 1.4080885179712588e-05,
+      "loss": 0.7332,
+      "step": 16351
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2455123888459283,
+      "learning_rate": 1.4080188555323079e-05,
+      "loss": 0.7726,
+      "step": 16352
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.507373264812781,
+      "learning_rate": 1.4079491907177428e-05,
+      "loss": 0.6784,
+      "step": 16353
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6239001838633302,
+      "learning_rate": 1.4078795235279677e-05,
+      "loss": 0.5717,
+      "step": 16354
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6702496496428498,
+      "learning_rate": 1.407809853963389e-05,
+      "loss": 0.7823,
+      "step": 16355
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7244150896104569,
+      "learning_rate": 1.4077401820244125e-05,
+      "loss": 0.7472,
+      "step": 16356
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4234748361334018,
+      "learning_rate": 1.4076705077114432e-05,
+      "loss": 0.8174,
+      "step": 16357
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5363325442097644,
+      "learning_rate": 1.4076008310248874e-05,
+      "loss": 0.7153,
+      "step": 16358
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4577818375825036,
+      "learning_rate": 1.4075311519651502e-05,
+      "loss": 0.796,
+      "step": 16359
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6988207837477067,
+      "learning_rate": 1.4074614705326379e-05,
+      "loss": 0.7384,
+      "step": 16360
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.9061259992094106,
+      "learning_rate": 1.4073917867277557e-05,
+      "loss": 0.7363,
+      "step": 16361
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5729601712427872,
+      "learning_rate": 1.40732210055091e-05,
+      "loss": 0.6281,
+      "step": 16362
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4574229351579242,
+      "learning_rate": 1.4072524120025057e-05,
+      "loss": 0.7416,
+      "step": 16363
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7338444451971993,
+      "learning_rate": 1.4071827210829493e-05,
+      "loss": 0.778,
+      "step": 16364
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3230268412916626,
+      "learning_rate": 1.4071130277926457e-05,
+      "loss": 0.5797,
+      "step": 16365
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6749399269945244,
+      "learning_rate": 1.4070433321320012e-05,
+      "loss": 0.8078,
+      "step": 16366
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4675830825020573,
+      "learning_rate": 1.4069736341014218e-05,
+      "loss": 0.6645,
+      "step": 16367
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3313546350476415,
+      "learning_rate": 1.406903933701313e-05,
+      "loss": 0.7462,
+      "step": 16368
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.458821640715921,
+      "learning_rate": 1.4068342309320807e-05,
+      "loss": 0.6898,
+      "step": 16369
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3810825209647017,
+      "learning_rate": 1.4067645257941308e-05,
+      "loss": 0.7837,
+      "step": 16370
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8692416609863136,
+      "learning_rate": 1.4066948182878686e-05,
+      "loss": 0.8098,
+      "step": 16371
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5499860112078045,
+      "learning_rate": 1.4066251084137006e-05,
+      "loss": 0.7474,
+      "step": 16372
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7854375803573135,
+      "learning_rate": 1.4065553961720327e-05,
+      "loss": 0.772,
+      "step": 16373
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6087447529855371,
+      "learning_rate": 1.4064856815632703e-05,
+      "loss": 0.529,
+      "step": 16374
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5803694698064938,
+      "learning_rate": 1.4064159645878197e-05,
+      "loss": 0.5112,
+      "step": 16375
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.81003625182319,
+      "learning_rate": 1.4063462452460868e-05,
+      "loss": 0.8339,
+      "step": 16376
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4317992154851396,
+      "learning_rate": 1.4062765235384772e-05,
+      "loss": 0.7674,
+      "step": 16377
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4614387309738242,
+      "learning_rate": 1.4062067994653971e-05,
+      "loss": 0.6262,
+      "step": 16378
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8268850073942418,
+      "learning_rate": 1.4061370730272527e-05,
+      "loss": 0.7214,
+      "step": 16379
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.302850956234198,
+      "learning_rate": 1.4060673442244493e-05,
+      "loss": 0.7424,
+      "step": 16380
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6944886718210623,
+      "learning_rate": 1.4059976130573936e-05,
+      "loss": 0.7572,
+      "step": 16381
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4337389158326894,
+      "learning_rate": 1.4059278795264908e-05,
+      "loss": 0.8753,
+      "step": 16382
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2850822343444757,
+      "learning_rate": 1.4058581436321478e-05,
+      "loss": 0.6997,
+      "step": 16383
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5876461903296692,
+      "learning_rate": 1.4057884053747699e-05,
+      "loss": 0.7331,
+      "step": 16384
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3856354574509222,
+      "learning_rate": 1.4057186647547638e-05,
+      "loss": 0.7942,
+      "step": 16385
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.5454392047960575,
+      "learning_rate": 1.405648921772535e-05,
+      "loss": 0.6709,
+      "step": 16386
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8459690492322454,
+      "learning_rate": 1.4055791764284897e-05,
+      "loss": 0.6381,
+      "step": 16387
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7566190299473963,
+      "learning_rate": 1.405509428723034e-05,
+      "loss": 0.6923,
+      "step": 16388
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3943781923165244,
+      "learning_rate": 1.405439678656574e-05,
+      "loss": 0.6753,
+      "step": 16389
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.91311246459296,
+      "learning_rate": 1.405369926229516e-05,
+      "loss": 0.7559,
+      "step": 16390
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3251110546573566,
+      "learning_rate": 1.4053001714422658e-05,
+      "loss": 0.6798,
+      "step": 16391
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5104450790643644,
+      "learning_rate": 1.4052304142952297e-05,
+      "loss": 0.6764,
+      "step": 16392
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8664861750475912,
+      "learning_rate": 1.405160654788814e-05,
+      "loss": 0.5942,
+      "step": 16393
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.1832406237465665,
+      "learning_rate": 1.4050908929234245e-05,
+      "loss": 0.6632,
+      "step": 16394
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3534749152109915,
+      "learning_rate": 1.4050211286994678e-05,
+      "loss": 0.8247,
+      "step": 16395
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3781789903579966,
+      "learning_rate": 1.4049513621173496e-05,
+      "loss": 0.7933,
+      "step": 16396
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.571562012950439,
+      "learning_rate": 1.4048815931774764e-05,
+      "loss": 0.8165,
+      "step": 16397
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0881632886186465,
+      "learning_rate": 1.4048118218802545e-05,
+      "loss": 0.7875,
+      "step": 16398
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3945150685598713,
+      "learning_rate": 1.40474204822609e-05,
+      "loss": 0.6888,
+      "step": 16399
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3787924173572559,
+      "learning_rate": 1.4046722722153891e-05,
+      "loss": 0.7056,
+      "step": 16400
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5214814486979689,
+      "learning_rate": 1.4046024938485581e-05,
+      "loss": 0.7961,
+      "step": 16401
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5765663194841968,
+      "learning_rate": 1.4045327131260033e-05,
+      "loss": 0.811,
+      "step": 16402
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2456429269606024,
+      "learning_rate": 1.404462930048131e-05,
+      "loss": 0.6247,
+      "step": 16403
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6106023457417317,
+      "learning_rate": 1.4043931446153476e-05,
+      "loss": 0.5267,
+      "step": 16404
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.588496439309879,
+      "learning_rate": 1.404323356828059e-05,
+      "loss": 0.5023,
+      "step": 16405
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3830531892960691,
+      "learning_rate": 1.404253566686672e-05,
+      "loss": 0.7724,
+      "step": 16406
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3946763352262646,
+      "learning_rate": 1.404183774191593e-05,
+      "loss": 0.7276,
+      "step": 16407
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6878264936197491,
+      "learning_rate": 1.4041139793432274e-05,
+      "loss": 0.7357,
+      "step": 16408
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.566108560750406,
+      "learning_rate": 1.4040441821419829e-05,
+      "loss": 0.4999,
+      "step": 16409
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.485833800976516,
+      "learning_rate": 1.4039743825882648e-05,
+      "loss": 0.7349,
+      "step": 16410
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.310995450419754,
+      "learning_rate": 1.4039045806824802e-05,
+      "loss": 0.6528,
+      "step": 16411
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3436831812611734,
+      "learning_rate": 1.4038347764250353e-05,
+      "loss": 0.731,
+      "step": 16412
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9435290142535373,
+      "learning_rate": 1.4037649698163364e-05,
+      "loss": 0.7998,
+      "step": 16413
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.477578214294802,
+      "learning_rate": 1.4036951608567901e-05,
+      "loss": 0.6651,
+      "step": 16414
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.271018750829483,
+      "learning_rate": 1.4036253495468026e-05,
+      "loss": 0.7116,
+      "step": 16415
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7226389247365415,
+      "learning_rate": 1.4035555358867807e-05,
+      "loss": 0.7584,
+      "step": 16416
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3837800812770515,
+      "learning_rate": 1.4034857198771308e-05,
+      "loss": 0.6557,
+      "step": 16417
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.975189275437894,
+      "learning_rate": 1.4034159015182591e-05,
+      "loss": 0.7664,
+      "step": 16418
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9576184036407291,
+      "learning_rate": 1.4033460808105723e-05,
+      "loss": 0.6172,
+      "step": 16419
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7344115199619698,
+      "learning_rate": 1.403276257754477e-05,
+      "loss": 0.8168,
+      "step": 16420
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.369705293836321,
+      "learning_rate": 1.4032064323503796e-05,
+      "loss": 0.6837,
+      "step": 16421
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3230934797306422,
+      "learning_rate": 1.4031366045986867e-05,
+      "loss": 0.8434,
+      "step": 16422
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.221614206874439,
+      "learning_rate": 1.4030667744998048e-05,
+      "loss": 0.7117,
+      "step": 16423
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2772084382615958,
+      "learning_rate": 1.4029969420541406e-05,
+      "loss": 0.7134,
+      "step": 16424
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0558592419235375,
+      "learning_rate": 1.4029271072621007e-05,
+      "loss": 0.7263,
+      "step": 16425
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.535136159975149,
+      "learning_rate": 1.4028572701240915e-05,
+      "loss": 0.9326,
+      "step": 16426
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5802454399175105,
+      "learning_rate": 1.4027874306405196e-05,
+      "loss": 0.8417,
+      "step": 16427
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.425729412388216,
+      "learning_rate": 1.4027175888117921e-05,
+      "loss": 0.6938,
+      "step": 16428
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3130219321848018,
+      "learning_rate": 1.4026477446383149e-05,
+      "loss": 0.7479,
+      "step": 16429
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5287033164624972,
+      "learning_rate": 1.4025778981204954e-05,
+      "loss": 0.7415,
+      "step": 16430
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6552958181015993,
+      "learning_rate": 1.4025080492587397e-05,
+      "loss": 0.6076,
+      "step": 16431
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4132223321513773,
+      "learning_rate": 1.4024381980534549e-05,
+      "loss": 0.8658,
+      "step": 16432
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6711320035135233,
+      "learning_rate": 1.4023683445050473e-05,
+      "loss": 0.6298,
+      "step": 16433
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4541480342059672,
+      "learning_rate": 1.4022984886139241e-05,
+      "loss": 0.7864,
+      "step": 16434
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6319084733405171,
+      "learning_rate": 1.4022286303804912e-05,
+      "loss": 0.6709,
+      "step": 16435
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6014342294659487,
+      "learning_rate": 1.4021587698051567e-05,
+      "loss": 0.5256,
+      "step": 16436
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7540881345468495,
+      "learning_rate": 1.4020889068883258e-05,
+      "loss": 0.7749,
+      "step": 16437
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4072178836688238,
+      "learning_rate": 1.4020190416304062e-05,
+      "loss": 0.7465,
+      "step": 16438
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6874661979898586,
+      "learning_rate": 1.4019491740318045e-05,
+      "loss": 0.8322,
+      "step": 16439
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.726318646549497,
+      "learning_rate": 1.4018793040929275e-05,
+      "loss": 0.6931,
+      "step": 16440
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3846492664748524,
+      "learning_rate": 1.401809431814182e-05,
+      "loss": 0.7892,
+      "step": 16441
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3274882661017569,
+      "learning_rate": 1.4017395571959745e-05,
+      "loss": 0.5968,
+      "step": 16442
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7487525674033892,
+      "learning_rate": 1.4016696802387121e-05,
+      "loss": 0.7533,
+      "step": 16443
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.604849369318979,
+      "learning_rate": 1.401599800942802e-05,
+      "loss": 0.5594,
+      "step": 16444
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6480117288319178,
+      "learning_rate": 1.4015299193086507e-05,
+      "loss": 0.6489,
+      "step": 16445
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.406817498754297,
+      "learning_rate": 1.4014600353366648e-05,
+      "loss": 0.8235,
+      "step": 16446
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7187232862351973,
+      "learning_rate": 1.4013901490272517e-05,
+      "loss": 0.8098,
+      "step": 16447
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8177767765507986,
+      "learning_rate": 1.401320260380818e-05,
+      "loss": 0.7975,
+      "step": 16448
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.380770739342446,
+      "learning_rate": 1.4012503693977705e-05,
+      "loss": 0.709,
+      "step": 16449
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4688850808199472,
+      "learning_rate": 1.4011804760785165e-05,
+      "loss": 0.7455,
+      "step": 16450
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.254213490213645,
+      "learning_rate": 1.401110580423463e-05,
+      "loss": 0.6644,
+      "step": 16451
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5143391525457575,
+      "learning_rate": 1.4010406824330162e-05,
+      "loss": 0.7216,
+      "step": 16452
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5462975629700495,
+      "learning_rate": 1.4009707821075843e-05,
+      "loss": 0.8062,
+      "step": 16453
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.620537843936099,
+      "learning_rate": 1.4009008794475726e-05,
+      "loss": 0.5617,
+      "step": 16454
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7314327321174419,
+      "learning_rate": 1.40083097445339e-05,
+      "loss": 0.6401,
+      "step": 16455
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6049756252278314,
+      "learning_rate": 1.4007610671254417e-05,
+      "loss": 0.5549,
+      "step": 16456
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6462948820258718,
+      "learning_rate": 1.4006911574641363e-05,
+      "loss": 0.8236,
+      "step": 16457
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.313720576715738,
+      "learning_rate": 1.4006212454698798e-05,
+      "loss": 0.6969,
+      "step": 16458
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4831885818356991,
+      "learning_rate": 1.4005513311430795e-05,
+      "loss": 0.716,
+      "step": 16459
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5336241124865322,
+      "learning_rate": 1.4004814144841428e-05,
+      "loss": 0.7304,
+      "step": 16460
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4626093131786835,
+      "learning_rate": 1.4004114954934762e-05,
+      "loss": 0.8921,
+      "step": 16461
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5105881272213646,
+      "learning_rate": 1.4003415741714874e-05,
+      "loss": 0.7439,
+      "step": 16462
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5734292438897801,
+      "learning_rate": 1.4002716505185828e-05,
+      "loss": 0.5236,
+      "step": 16463
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3600578933329726,
+      "learning_rate": 1.4002017245351705e-05,
+      "loss": 0.8233,
+      "step": 16464
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6178201671808682,
+      "learning_rate": 1.4001317962216568e-05,
+      "loss": 0.5885,
+      "step": 16465
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0322585236963264,
+      "learning_rate": 1.400061865578449e-05,
+      "loss": 0.7577,
+      "step": 16466
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6887560010627607,
+      "learning_rate": 1.3999919326059543e-05,
+      "loss": 0.7846,
+      "step": 16467
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4215223103720025,
+      "learning_rate": 1.3999219973045802e-05,
+      "loss": 0.6985,
+      "step": 16468
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5948970269529996,
+      "learning_rate": 1.3998520596747333e-05,
+      "loss": 0.9222,
+      "step": 16469
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9268988598410761,
+      "learning_rate": 1.3997821197168214e-05,
+      "loss": 0.6762,
+      "step": 16470
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.611387360922809,
+      "learning_rate": 1.3997121774312512e-05,
+      "loss": 0.5746,
+      "step": 16471
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4773882399626859,
+      "learning_rate": 1.3996422328184305e-05,
+      "loss": 0.7506,
+      "step": 16472
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.543243489579157,
+      "learning_rate": 1.3995722858787658e-05,
+      "loss": 0.7385,
+      "step": 16473
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.347847643012571,
+      "learning_rate": 1.399502336612665e-05,
+      "loss": 0.6841,
+      "step": 16474
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4691662597609974,
+      "learning_rate": 1.3994323850205349e-05,
+      "loss": 0.7083,
+      "step": 16475
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.243121295396624,
+      "learning_rate": 1.3993624311027833e-05,
+      "loss": 0.7512,
+      "step": 16476
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5713064253963838,
+      "learning_rate": 1.399292474859817e-05,
+      "loss": 0.8049,
+      "step": 16477
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.332635532827275,
+      "learning_rate": 1.3992225162920434e-05,
+      "loss": 0.6733,
+      "step": 16478
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5279943853713527,
+      "learning_rate": 1.3991525553998704e-05,
+      "loss": 0.8085,
+      "step": 16479
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4738289611954254,
+      "learning_rate": 1.399082592183704e-05,
+      "loss": 0.7552,
+      "step": 16480
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6910146899589331,
+      "learning_rate": 1.3990126266439532e-05,
+      "loss": 0.8082,
+      "step": 16481
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6072322127670198,
+      "learning_rate": 1.3989426587810241e-05,
+      "loss": 0.6269,
+      "step": 16482
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3017952909966946,
+      "learning_rate": 1.398872688595325e-05,
+      "loss": 0.6449,
+      "step": 16483
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.083805722083019,
+      "learning_rate": 1.3988027160872624e-05,
+      "loss": 0.6614,
+      "step": 16484
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4499080148304038,
+      "learning_rate": 1.3987327412572442e-05,
+      "loss": 0.7967,
+      "step": 16485
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5885136665648787,
+      "learning_rate": 1.3986627641056777e-05,
+      "loss": 0.535,
+      "step": 16486
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5058332147889495,
+      "learning_rate": 1.3985927846329705e-05,
+      "loss": 0.7943,
+      "step": 16487
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.411396976297817,
+      "learning_rate": 1.3985228028395298e-05,
+      "loss": 0.6799,
+      "step": 16488
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0953924498752974,
+      "learning_rate": 1.3984528187257636e-05,
+      "loss": 0.6868,
+      "step": 16489
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5676926015092052,
+      "learning_rate": 1.3983828322920786e-05,
+      "loss": 0.7263,
+      "step": 16490
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5213140675571737,
+      "learning_rate": 1.3983128435388824e-05,
+      "loss": 0.7239,
+      "step": 16491
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3313057089622888,
+      "learning_rate": 1.398242852466583e-05,
+      "loss": 0.7582,
+      "step": 16492
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.557393822510477,
+      "learning_rate": 1.3981728590755877e-05,
+      "loss": 0.6492,
+      "step": 16493
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7170224584537803,
+      "learning_rate": 1.3981028633663036e-05,
+      "loss": 0.8057,
+      "step": 16494
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5453753581957947,
+      "learning_rate": 1.3980328653391389e-05,
+      "loss": 0.7249,
+      "step": 16495
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3820159743014284,
+      "learning_rate": 1.3979628649945007e-05,
+      "loss": 0.5971,
+      "step": 16496
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.361966098112277,
+      "learning_rate": 1.3978928623327969e-05,
+      "loss": 0.6913,
+      "step": 16497
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5589086815646113,
+      "learning_rate": 1.3978228573544348e-05,
+      "loss": 0.4818,
+      "step": 16498
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.738948596142153,
+      "learning_rate": 1.3977528500598218e-05,
+      "loss": 0.7724,
+      "step": 16499
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0081807861389285,
+      "learning_rate": 1.3976828404493662e-05,
+      "loss": 0.6491,
+      "step": 16500
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3879118241837198,
+      "learning_rate": 1.3976128285234748e-05,
+      "loss": 0.7257,
+      "step": 16501
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.115102628919233,
+      "learning_rate": 1.3975428142825562e-05,
+      "loss": 0.8226,
+      "step": 16502
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4536642161404958,
+      "learning_rate": 1.3974727977270167e-05,
+      "loss": 0.8171,
+      "step": 16503
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9015754283424746,
+      "learning_rate": 1.3974027788572655e-05,
+      "loss": 0.7414,
+      "step": 16504
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4530244162244024,
+      "learning_rate": 1.397332757673709e-05,
+      "loss": 0.7936,
+      "step": 16505
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5981260986675032,
+      "learning_rate": 1.3972627341767555e-05,
+      "loss": 0.549,
+      "step": 16506
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9494738928425275,
+      "learning_rate": 1.3971927083668127e-05,
+      "loss": 0.817,
+      "step": 16507
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5817025031716742,
+      "learning_rate": 1.397122680244288e-05,
+      "loss": 0.5768,
+      "step": 16508
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3425583161107746,
+      "learning_rate": 1.3970526498095897e-05,
+      "loss": 0.6858,
+      "step": 16509
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.87489185308216,
+      "learning_rate": 1.3969826170631247e-05,
+      "loss": 0.7366,
+      "step": 16510
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5783423809551187,
+      "learning_rate": 1.3969125820053018e-05,
+      "loss": 0.7245,
+      "step": 16511
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7434695974740557,
+      "learning_rate": 1.3968425446365275e-05,
+      "loss": 0.7796,
+      "step": 16512
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4749873598870824,
+      "learning_rate": 1.3967725049572109e-05,
+      "loss": 0.6545,
+      "step": 16513
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9469503723135964,
+      "learning_rate": 1.3967024629677585e-05,
+      "loss": 0.7386,
+      "step": 16514
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6213041890809112,
+      "learning_rate": 1.3966324186685795e-05,
+      "loss": 0.5486,
+      "step": 16515
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4395431161729573,
+      "learning_rate": 1.3965623720600805e-05,
+      "loss": 0.7656,
+      "step": 16516
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4503381202644925,
+      "learning_rate": 1.3964923231426701e-05,
+      "loss": 0.825,
+      "step": 16517
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6860795985307302,
+      "learning_rate": 1.3964222719167557e-05,
+      "loss": 0.7158,
+      "step": 16518
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.442628205134697,
+      "learning_rate": 1.3963522183827454e-05,
+      "loss": 0.5523,
+      "step": 16519
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5114625506362132,
+      "learning_rate": 1.3962821625410468e-05,
+      "loss": 0.7222,
+      "step": 16520
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7021731807795388,
+      "learning_rate": 1.3962121043920682e-05,
+      "loss": 0.7637,
+      "step": 16521
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8359956184912636,
+      "learning_rate": 1.3961420439362173e-05,
+      "loss": 0.853,
+      "step": 16522
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.369320273060235,
+      "learning_rate": 1.396071981173902e-05,
+      "loss": 0.7062,
+      "step": 16523
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.319280626914183,
+      "learning_rate": 1.3960019161055297e-05,
+      "loss": 0.6669,
+      "step": 16524
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4042507153039547,
+      "learning_rate": 1.3959318487315096e-05,
+      "loss": 0.7129,
+      "step": 16525
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6716802309627642,
+      "learning_rate": 1.3958617790522486e-05,
+      "loss": 0.5773,
+      "step": 16526
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.614667158667298,
+      "learning_rate": 1.395791707068155e-05,
+      "loss": 0.6728,
+      "step": 16527
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.303110292474478,
+      "learning_rate": 1.3957216327796368e-05,
+      "loss": 0.8364,
+      "step": 16528
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5096199445859304,
+      "learning_rate": 1.395651556187102e-05,
+      "loss": 0.7793,
+      "step": 16529
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3452502415524972,
+      "learning_rate": 1.3955814772909588e-05,
+      "loss": 0.723,
+      "step": 16530
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7781269116365994,
+      "learning_rate": 1.3955113960916146e-05,
+      "loss": 0.7446,
+      "step": 16531
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4841875237593778,
+      "learning_rate": 1.3954413125894783e-05,
+      "loss": 0.6927,
+      "step": 16532
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4788110276438178,
+      "learning_rate": 1.3953712267849568e-05,
+      "loss": 0.757,
+      "step": 16533
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5080756188483475,
+      "learning_rate": 1.3953011386784597e-05,
+      "loss": 0.692,
+      "step": 16534
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4213665500925492,
+      "learning_rate": 1.3952310482703935e-05,
+      "loss": 0.7581,
+      "step": 16535
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6575293436639589,
+      "learning_rate": 1.3951609555611673e-05,
+      "loss": 0.6248,
+      "step": 16536
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.48110630139842,
+      "learning_rate": 1.3950908605511888e-05,
+      "loss": 0.6449,
+      "step": 16537
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3626276528755328,
+      "learning_rate": 1.3950207632408662e-05,
+      "loss": 0.6945,
+      "step": 16538
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6754810680430878,
+      "learning_rate": 1.394950663630608e-05,
+      "loss": 0.8508,
+      "step": 16539
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3060812610932493,
+      "learning_rate": 1.3948805617208216e-05,
+      "loss": 0.7153,
+      "step": 16540
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4245174963790892,
+      "learning_rate": 1.394810457511916e-05,
+      "loss": 0.8698,
+      "step": 16541
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6668458483915196,
+      "learning_rate": 1.3947403510042984e-05,
+      "loss": 0.7303,
+      "step": 16542
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.282135569630444,
+      "learning_rate": 1.3946702421983778e-05,
+      "loss": 0.6595,
+      "step": 16543
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4006561697730262,
+      "learning_rate": 1.394600131094562e-05,
+      "loss": 0.6805,
+      "step": 16544
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0914913339756764,
+      "learning_rate": 1.3945300176932594e-05,
+      "loss": 0.7466,
+      "step": 16545
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3124888238930397,
+      "learning_rate": 1.3944599019948779e-05,
+      "loss": 0.7114,
+      "step": 16546
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.18832505378585,
+      "learning_rate": 1.3943897839998265e-05,
+      "loss": 0.819,
+      "step": 16547
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5981945824832946,
+      "learning_rate": 1.3943196637085123e-05,
+      "loss": 0.7296,
+      "step": 16548
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.537767167285833,
+      "learning_rate": 1.3942495411213446e-05,
+      "loss": 0.8213,
+      "step": 16549
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5170939451463936,
+      "learning_rate": 1.394179416238731e-05,
+      "loss": 0.7204,
+      "step": 16550
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4951265410882701,
+      "learning_rate": 1.3941092890610803e-05,
+      "loss": 0.6248,
+      "step": 16551
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5561296418920956,
+      "learning_rate": 1.3940391595888e-05,
+      "loss": 0.8128,
+      "step": 16552
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3515528024531591,
+      "learning_rate": 1.3939690278222997e-05,
+      "loss": 0.7713,
+      "step": 16553
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.612158542273353,
+      "learning_rate": 1.3938988937619864e-05,
+      "loss": 0.8488,
+      "step": 16554
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6465992699254384,
+      "learning_rate": 1.3938287574082692e-05,
+      "loss": 0.8373,
+      "step": 16555
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5427001765915627,
+      "learning_rate": 1.3937586187615564e-05,
+      "loss": 0.4755,
+      "step": 16556
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6854538258774514,
+      "learning_rate": 1.3936884778222562e-05,
+      "loss": 0.712,
+      "step": 16557
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4953263660061153,
+      "learning_rate": 1.3936183345907772e-05,
+      "loss": 0.7235,
+      "step": 16558
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5400067800101225,
+      "learning_rate": 1.3935481890675273e-05,
+      "loss": 0.8472,
+      "step": 16559
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5567318465254987,
+      "learning_rate": 1.3934780412529158e-05,
+      "loss": 0.7221,
+      "step": 16560
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4534870384762248,
+      "learning_rate": 1.3934078911473499e-05,
+      "loss": 0.8689,
+      "step": 16561
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6726377070940652,
+      "learning_rate": 1.3933377387512393e-05,
+      "loss": 0.7927,
+      "step": 16562
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8300355918684863,
+      "learning_rate": 1.3932675840649915e-05,
+      "loss": 0.6872,
+      "step": 16563
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8352338986276586,
+      "learning_rate": 1.3931974270890154e-05,
+      "loss": 0.8526,
+      "step": 16564
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4412331841771426,
+      "learning_rate": 1.393127267823719e-05,
+      "loss": 0.6919,
+      "step": 16565
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2187904953700408,
+      "learning_rate": 1.393057106269512e-05,
+      "loss": 0.7594,
+      "step": 16566
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2792309491764637,
+      "learning_rate": 1.3929869424268016e-05,
+      "loss": 0.764,
+      "step": 16567
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.623250194060809,
+      "learning_rate": 1.392916776295997e-05,
+      "loss": 0.6449,
+      "step": 16568
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.599355819306418,
+      "learning_rate": 1.3928466078775063e-05,
+      "loss": 0.7512,
+      "step": 16569
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3880320363380276,
+      "learning_rate": 1.3927764371717386e-05,
+      "loss": 0.881,
+      "step": 16570
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.697763346325796,
+      "learning_rate": 1.3927062641791018e-05,
+      "loss": 0.8292,
+      "step": 16571
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.273374452077614,
+      "learning_rate": 1.3926360889000048e-05,
+      "loss": 0.7533,
+      "step": 16572
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7385977202650067,
+      "learning_rate": 1.3925659113348564e-05,
+      "loss": 0.7361,
+      "step": 16573
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5605177475272272,
+      "learning_rate": 1.3924957314840648e-05,
+      "loss": 0.705,
+      "step": 16574
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.653639026437602,
+      "learning_rate": 1.3924255493480387e-05,
+      "loss": 0.7847,
+      "step": 16575
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5500368787417869,
+      "learning_rate": 1.392355364927187e-05,
+      "loss": 0.7668,
+      "step": 16576
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.4028349356211276,
+      "learning_rate": 1.3922851782219182e-05,
+      "loss": 0.7793,
+      "step": 16577
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.022628411832623,
+      "learning_rate": 1.3922149892326409e-05,
+      "loss": 0.6601,
+      "step": 16578
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.489087226421851,
+      "learning_rate": 1.3921447979597636e-05,
+      "loss": 0.7028,
+      "step": 16579
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3846413284577455,
+      "learning_rate": 1.3920746044036953e-05,
+      "loss": 0.7639,
+      "step": 16580
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6075244453346329,
+      "learning_rate": 1.3920044085648446e-05,
+      "loss": 0.5212,
+      "step": 16581
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5790281046148688,
+      "learning_rate": 1.3919342104436198e-05,
+      "loss": 0.5424,
+      "step": 16582
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.725200102592657,
+      "learning_rate": 1.3918640100404301e-05,
+      "loss": 0.6137,
+      "step": 16583
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5421353873704136,
+      "learning_rate": 1.3917938073556841e-05,
+      "loss": 0.7341,
+      "step": 16584
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5048714699757124,
+      "learning_rate": 1.3917236023897906e-05,
+      "loss": 0.6459,
+      "step": 16585
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.538797165304881,
+      "learning_rate": 1.3916533951431579e-05,
+      "loss": 0.8473,
+      "step": 16586
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4373956650148019,
+      "learning_rate": 1.3915831856161956e-05,
+      "loss": 0.7566,
+      "step": 16587
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7389295985163324,
+      "learning_rate": 1.3915129738093119e-05,
+      "loss": 0.6363,
+      "step": 16588
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.047365317302713,
+      "learning_rate": 1.3914427597229156e-05,
+      "loss": 0.7696,
+      "step": 16589
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5107895674601013,
+      "learning_rate": 1.3913725433574157e-05,
+      "loss": 0.8495,
+      "step": 16590
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5707479868224845,
+      "learning_rate": 1.391302324713221e-05,
+      "loss": 0.6942,
+      "step": 16591
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5245811153528728,
+      "learning_rate": 1.39123210379074e-05,
+      "loss": 0.8314,
+      "step": 16592
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5733809176595233,
+      "learning_rate": 1.391161880590382e-05,
+      "loss": 0.7542,
+      "step": 16593
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5871612132071338,
+      "learning_rate": 1.3910916551125558e-05,
+      "loss": 0.9344,
+      "step": 16594
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3673510422346027,
+      "learning_rate": 1.3910214273576701e-05,
+      "loss": 0.8705,
+      "step": 16595
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5776821932508107,
+      "learning_rate": 1.3909511973261339e-05,
+      "loss": 0.5487,
+      "step": 16596
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.398520991169603,
+      "learning_rate": 1.3908809650183559e-05,
+      "loss": 0.6857,
+      "step": 16597
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5098658999402672,
+      "learning_rate": 1.3908107304347453e-05,
+      "loss": 0.699,
+      "step": 16598
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5855473096604613,
+      "learning_rate": 1.3907404935757107e-05,
+      "loss": 0.7748,
+      "step": 16599
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7521781910216436,
+      "learning_rate": 1.3906702544416614e-05,
+      "loss": 0.7389,
+      "step": 16600
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.705640193617158,
+      "learning_rate": 1.3906000130330061e-05,
+      "loss": 0.6612,
+      "step": 16601
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7683834849480278,
+      "learning_rate": 1.3905297693501543e-05,
+      "loss": 0.7404,
+      "step": 16602
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.289625719331213,
+      "learning_rate": 1.3904595233935139e-05,
+      "loss": 0.7719,
+      "step": 16603
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4276971472147528,
+      "learning_rate": 1.3903892751634949e-05,
+      "loss": 0.7524,
+      "step": 16604
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.705610529617307,
+      "learning_rate": 1.3903190246605059e-05,
+      "loss": 0.7383,
+      "step": 16605
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3342033006588065,
+      "learning_rate": 1.390248771884956e-05,
+      "loss": 0.7827,
+      "step": 16606
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.487656379228824,
+      "learning_rate": 1.3901785168372542e-05,
+      "loss": 0.6639,
+      "step": 16607
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9516206287394913,
+      "learning_rate": 1.3901082595178092e-05,
+      "loss": 0.8836,
+      "step": 16608
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5445147767135001,
+      "learning_rate": 1.390037999927031e-05,
+      "loss": 0.8286,
+      "step": 16609
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2786305483012725,
+      "learning_rate": 1.3899677380653276e-05,
+      "loss": 0.7748,
+      "step": 16610
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7925484842177948,
+      "learning_rate": 1.3898974739331091e-05,
+      "loss": 0.7589,
+      "step": 16611
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7162627837133446,
+      "learning_rate": 1.3898272075307833e-05,
+      "loss": 0.9032,
+      "step": 16612
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7982155231634993,
+      "learning_rate": 1.3897569388587609e-05,
+      "loss": 0.7509,
+      "step": 16613
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0328015127542973,
+      "learning_rate": 1.3896866679174495e-05,
+      "loss": 0.7252,
+      "step": 16614
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5453701767130548,
+      "learning_rate": 1.3896163947072595e-05,
+      "loss": 0.8093,
+      "step": 16615
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.5300364097818058,
+      "learning_rate": 1.3895461192285992e-05,
+      "loss": 0.8667,
+      "step": 16616
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4428000490231185,
+      "learning_rate": 1.3894758414818781e-05,
+      "loss": 0.741,
+      "step": 16617
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3691920422024182,
+      "learning_rate": 1.3894055614675052e-05,
+      "loss": 0.8172,
+      "step": 16618
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.803327234277151,
+      "learning_rate": 1.38933527918589e-05,
+      "loss": 0.8669,
+      "step": 16619
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7695438765509273,
+      "learning_rate": 1.3892649946374417e-05,
+      "loss": 0.8559,
+      "step": 16620
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7580247404114364,
+      "learning_rate": 1.3891947078225692e-05,
+      "loss": 0.8388,
+      "step": 16621
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5908839578824827,
+      "learning_rate": 1.389124418741682e-05,
+      "loss": 0.7765,
+      "step": 16622
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0709359578182513,
+      "learning_rate": 1.389054127395189e-05,
+      "loss": 0.7679,
+      "step": 16623
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.353847592997738,
+      "learning_rate": 1.3889838337834997e-05,
+      "loss": 0.8393,
+      "step": 16624
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0457559337141227,
+      "learning_rate": 1.3889135379070235e-05,
+      "loss": 0.6688,
+      "step": 16625
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2313110343488287,
+      "learning_rate": 1.3888432397661697e-05,
+      "loss": 0.6826,
+      "step": 16626
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6806834921511713,
+      "learning_rate": 1.3887729393613471e-05,
+      "loss": 0.7676,
+      "step": 16627
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5520430893578545,
+      "learning_rate": 1.3887026366929657e-05,
+      "loss": 0.7856,
+      "step": 16628
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.592451215023503,
+      "learning_rate": 1.3886323317614344e-05,
+      "loss": 0.6106,
+      "step": 16629
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.485509439563941,
+      "learning_rate": 1.3885620245671624e-05,
+      "loss": 0.7586,
+      "step": 16630
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5947393187848435,
+      "learning_rate": 1.3884917151105595e-05,
+      "loss": 0.5544,
+      "step": 16631
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5600456250592682,
+      "learning_rate": 1.388421403392035e-05,
+      "loss": 0.74,
+      "step": 16632
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.448049392386397,
+      "learning_rate": 1.3883510894119979e-05,
+      "loss": 0.6256,
+      "step": 16633
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3465394073725978,
+      "learning_rate": 1.3882807731708579e-05,
+      "loss": 0.8117,
+      "step": 16634
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5814178146804717,
+      "learning_rate": 1.3882104546690241e-05,
+      "loss": 0.5274,
+      "step": 16635
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8818700687680274,
+      "learning_rate": 1.3881401339069064e-05,
+      "loss": 0.7759,
+      "step": 16636
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6754289412225685,
+      "learning_rate": 1.3880698108849138e-05,
+      "loss": 0.8215,
+      "step": 16637
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6085901254749452,
+      "learning_rate": 1.387999485603456e-05,
+      "loss": 0.4906,
+      "step": 16638
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3548293677698966,
+      "learning_rate": 1.3879291580629421e-05,
+      "loss": 0.7682,
+      "step": 16639
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5280653814534995,
+      "learning_rate": 1.3878588282637822e-05,
+      "loss": 0.8317,
+      "step": 16640
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.857604030303687,
+      "learning_rate": 1.3877884962063853e-05,
+      "loss": 0.6215,
+      "step": 16641
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4647771134319694,
+      "learning_rate": 1.3877181618911606e-05,
+      "loss": 0.743,
+      "step": 16642
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8365045025993914,
+      "learning_rate": 1.3876478253185183e-05,
+      "loss": 0.7326,
+      "step": 16643
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5471646666722362,
+      "learning_rate": 1.3875774864888675e-05,
+      "loss": 0.5422,
+      "step": 16644
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4031132216931754,
+      "learning_rate": 1.3875071454026183e-05,
+      "loss": 0.6785,
+      "step": 16645
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6349343164992505,
+      "learning_rate": 1.3874368020601796e-05,
+      "loss": 0.7465,
+      "step": 16646
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3994333722543608,
+      "learning_rate": 1.3873664564619612e-05,
+      "loss": 0.7593,
+      "step": 16647
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5858461863402789,
+      "learning_rate": 1.3872961086083725e-05,
+      "loss": 0.6581,
+      "step": 16648
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3323591981706955,
+      "learning_rate": 1.3872257584998235e-05,
+      "loss": 0.7625,
+      "step": 16649
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.444152932503401,
+      "learning_rate": 1.3871554061367232e-05,
+      "loss": 0.8077,
+      "step": 16650
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5046044740975602,
+      "learning_rate": 1.3870850515194817e-05,
+      "loss": 0.8247,
+      "step": 16651
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3835152705867113,
+      "learning_rate": 1.3870146946485083e-05,
+      "loss": 0.8646,
+      "step": 16652
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.115495303514414,
+      "learning_rate": 1.386944335524213e-05,
+      "loss": 0.6701,
+      "step": 16653
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2953604032324162,
+      "learning_rate": 1.3868739741470055e-05,
+      "loss": 0.688,
+      "step": 16654
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.584358528006336,
+      "learning_rate": 1.386803610517295e-05,
+      "loss": 0.55,
+      "step": 16655
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4045743823294516,
+      "learning_rate": 1.3867332446354916e-05,
+      "loss": 0.7274,
+      "step": 16656
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.356861406669297,
+      "learning_rate": 1.3866628765020045e-05,
+      "loss": 0.8179,
+      "step": 16657
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4581202228959882,
+      "learning_rate": 1.3865925061172439e-05,
+      "loss": 0.7424,
+      "step": 16658
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6140367046367947,
+      "learning_rate": 1.3865221334816193e-05,
+      "loss": 0.6509,
+      "step": 16659
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4230776396126248,
+      "learning_rate": 1.3864517585955406e-05,
+      "loss": 0.718,
+      "step": 16660
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5736749879410062,
+      "learning_rate": 1.3863813814594172e-05,
+      "loss": 0.5246,
+      "step": 16661
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3950895621124912,
+      "learning_rate": 1.3863110020736592e-05,
+      "loss": 0.7281,
+      "step": 16662
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3622305238120098,
+      "learning_rate": 1.3862406204386761e-05,
+      "loss": 0.6897,
+      "step": 16663
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6873200007227407,
+      "learning_rate": 1.3861702365548781e-05,
+      "loss": 0.812,
+      "step": 16664
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6104227046204818,
+      "learning_rate": 1.3860998504226744e-05,
+      "loss": 0.7478,
+      "step": 16665
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8811261590495882,
+      "learning_rate": 1.3860294620424754e-05,
+      "loss": 0.8187,
+      "step": 16666
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4636017067254685,
+      "learning_rate": 1.3859590714146906e-05,
+      "loss": 0.7902,
+      "step": 16667
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2407735852273682,
+      "learning_rate": 1.38588867853973e-05,
+      "loss": 0.6541,
+      "step": 16668
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3775253461222057,
+      "learning_rate": 1.385818283418003e-05,
+      "loss": 0.6852,
+      "step": 16669
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5513774625578283,
+      "learning_rate": 1.3857478860499201e-05,
+      "loss": 0.7324,
+      "step": 16670
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.216950062646801,
+      "learning_rate": 1.3856774864358906e-05,
+      "loss": 0.6974,
+      "step": 16671
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7982266759726515,
+      "learning_rate": 1.385607084576325e-05,
+      "loss": 0.7092,
+      "step": 16672
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4530064319706835,
+      "learning_rate": 1.3855366804716326e-05,
+      "loss": 0.666,
+      "step": 16673
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6020525627156281,
+      "learning_rate": 1.3854662741222238e-05,
+      "loss": 0.5204,
+      "step": 16674
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4586815102779693,
+      "learning_rate": 1.385395865528508e-05,
+      "loss": 0.7155,
+      "step": 16675
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.243117892501683,
+      "learning_rate": 1.3853254546908958e-05,
+      "loss": 0.7224,
+      "step": 16676
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4459291080791976,
+      "learning_rate": 1.3852550416097968e-05,
+      "loss": 0.7173,
+      "step": 16677
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3256312430764863,
+      "learning_rate": 1.3851846262856208e-05,
+      "loss": 0.8203,
+      "step": 16678
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2917434664753937,
+      "learning_rate": 1.3851142087187778e-05,
+      "loss": 0.725,
+      "step": 16679
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4254256181152019,
+      "learning_rate": 1.3850437889096781e-05,
+      "loss": 0.6478,
+      "step": 16680
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4979240648922054,
+      "learning_rate": 1.384973366858732e-05,
+      "loss": 0.704,
+      "step": 16681
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5975465786261014,
+      "learning_rate": 1.3849029425663485e-05,
+      "loss": 0.7329,
+      "step": 16682
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.406540853895495,
+      "learning_rate": 1.3848325160329386e-05,
+      "loss": 0.7006,
+      "step": 16683
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4066102012370827,
+      "learning_rate": 1.3847620872589116e-05,
+      "loss": 0.6718,
+      "step": 16684
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3918569833497956,
+      "learning_rate": 1.3846916562446784e-05,
+      "loss": 0.7334,
+      "step": 16685
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5212157721011241,
+      "learning_rate": 1.384621222990648e-05,
+      "loss": 0.6091,
+      "step": 16686
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5754031451770266,
+      "learning_rate": 1.3845507874972314e-05,
+      "loss": 0.4791,
+      "step": 16687
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3711013363894242,
+      "learning_rate": 1.3844803497648382e-05,
+      "loss": 0.7227,
+      "step": 16688
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4835598535757448,
+      "learning_rate": 1.384409909793879e-05,
+      "loss": 0.6992,
+      "step": 16689
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3919683556215514,
+      "learning_rate": 1.3843394675847635e-05,
+      "loss": 0.7586,
+      "step": 16690
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6021095916172344,
+      "learning_rate": 1.3842690231379014e-05,
+      "loss": 0.5395,
+      "step": 16691
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5607233235564669,
+      "learning_rate": 1.384198576453704e-05,
+      "loss": 0.5647,
+      "step": 16692
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9208662604114848,
+      "learning_rate": 1.3841281275325807e-05,
+      "loss": 0.6279,
+      "step": 16693
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6796795670882014,
+      "learning_rate": 1.3840576763749419e-05,
+      "loss": 0.8128,
+      "step": 16694
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3942963775168258,
+      "learning_rate": 1.3839872229811974e-05,
+      "loss": 0.755,
+      "step": 16695
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5919834301904374,
+      "learning_rate": 1.3839167673517582e-05,
+      "loss": 0.528,
+      "step": 16696
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.081468402374323,
+      "learning_rate": 1.3838463094870338e-05,
+      "loss": 0.8216,
+      "step": 16697
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3789765004842187,
+      "learning_rate": 1.3837758493874347e-05,
+      "loss": 0.8019,
+      "step": 16698
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3839044954037794,
+      "learning_rate": 1.383705387053371e-05,
+      "loss": 0.6407,
+      "step": 16699
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5413834391397006,
+      "learning_rate": 1.383634922485253e-05,
+      "loss": 0.7449,
+      "step": 16700
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.419958930865559,
+      "learning_rate": 1.3835644556834913e-05,
+      "loss": 0.6533,
+      "step": 16701
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6494267189895848,
+      "learning_rate": 1.3834939866484956e-05,
+      "loss": 0.8189,
+      "step": 16702
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5338054541121107,
+      "learning_rate": 1.3834235153806766e-05,
+      "loss": 0.7026,
+      "step": 16703
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.545606668979648,
+      "learning_rate": 1.3833530418804446e-05,
+      "loss": 0.6642,
+      "step": 16704
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9136122907354534,
+      "learning_rate": 1.3832825661482097e-05,
+      "loss": 0.687,
+      "step": 16705
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.5702536186213227,
+      "learning_rate": 1.3832120881843825e-05,
+      "loss": 0.765,
+      "step": 16706
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3942260086518128,
+      "learning_rate": 1.383141607989373e-05,
+      "loss": 0.8232,
+      "step": 16707
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5976800342895414,
+      "learning_rate": 1.3830711255635919e-05,
+      "loss": 0.7757,
+      "step": 16708
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3208678420114348,
+      "learning_rate": 1.3830006409074495e-05,
+      "loss": 0.7585,
+      "step": 16709
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4901880693615872,
+      "learning_rate": 1.3829301540213557e-05,
+      "loss": 0.7002,
+      "step": 16710
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6271999333353346,
+      "learning_rate": 1.382859664905722e-05,
+      "loss": 0.6501,
+      "step": 16711
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.387057709046427,
+      "learning_rate": 1.3827891735609575e-05,
+      "loss": 0.6189,
+      "step": 16712
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5532484320280019,
+      "learning_rate": 1.3827186799874737e-05,
+      "loss": 0.5061,
+      "step": 16713
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4025385291009953,
+      "learning_rate": 1.3826481841856802e-05,
+      "loss": 0.7722,
+      "step": 16714
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4952099168766801,
+      "learning_rate": 1.382577686155988e-05,
+      "loss": 0.8278,
+      "step": 16715
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4763672885069627,
+      "learning_rate": 1.3825071858988073e-05,
+      "loss": 0.6608,
+      "step": 16716
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4771810381572412,
+      "learning_rate": 1.3824366834145488e-05,
+      "loss": 0.7983,
+      "step": 16717
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3786512967840487,
+      "learning_rate": 1.382366178703623e-05,
+      "loss": 0.7469,
+      "step": 16718
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.862707509782514,
+      "learning_rate": 1.38229567176644e-05,
+      "loss": 0.715,
+      "step": 16719
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8963735996323727,
+      "learning_rate": 1.3822251626034105e-05,
+      "loss": 0.7179,
+      "step": 16720
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.445566564900188,
+      "learning_rate": 1.3821546512149453e-05,
+      "loss": 0.7844,
+      "step": 16721
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6026394114960711,
+      "learning_rate": 1.3820841376014546e-05,
+      "loss": 0.8156,
+      "step": 16722
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4539163129139019,
+      "learning_rate": 1.3820136217633491e-05,
+      "loss": 0.6932,
+      "step": 16723
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1036818383958344,
+      "learning_rate": 1.3819431037010393e-05,
+      "loss": 0.7035,
+      "step": 16724
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.2865530815962916,
+      "learning_rate": 1.381872583414936e-05,
+      "loss": 0.7404,
+      "step": 16725
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.045028161814996,
+      "learning_rate": 1.3818020609054494e-05,
+      "loss": 0.6313,
+      "step": 16726
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.830376298620337,
+      "learning_rate": 1.3817315361729904e-05,
+      "loss": 0.6216,
+      "step": 16727
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5754140669392536,
+      "learning_rate": 1.3816610092179697e-05,
+      "loss": 0.7373,
+      "step": 16728
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8276058683007002,
+      "learning_rate": 1.3815904800407975e-05,
+      "loss": 0.9506,
+      "step": 16729
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.543984436836191,
+      "learning_rate": 1.3815199486418851e-05,
+      "loss": 0.7274,
+      "step": 16730
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.370424851675352,
+      "learning_rate": 1.3814494150216424e-05,
+      "loss": 0.8522,
+      "step": 16731
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2491465225414113,
+      "learning_rate": 1.3813788791804807e-05,
+      "loss": 0.6268,
+      "step": 16732
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.801904543759641,
+      "learning_rate": 1.3813083411188104e-05,
+      "loss": 0.8829,
+      "step": 16733
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7133298778490333,
+      "learning_rate": 1.3812378008370422e-05,
+      "loss": 0.6419,
+      "step": 16734
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.289146250520191,
+      "learning_rate": 1.3811672583355864e-05,
+      "loss": 0.7444,
+      "step": 16735
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.562519567430431,
+      "learning_rate": 1.3810967136148547e-05,
+      "loss": 0.7246,
+      "step": 16736
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5174952505811943,
+      "learning_rate": 1.3810261666752571e-05,
+      "loss": 0.8773,
+      "step": 16737
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4788958679900073,
+      "learning_rate": 1.3809556175172043e-05,
+      "loss": 0.7862,
+      "step": 16738
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5001772985552018,
+      "learning_rate": 1.3808850661411077e-05,
+      "loss": 0.7976,
+      "step": 16739
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6308381062197916,
+      "learning_rate": 1.3808145125473773e-05,
+      "loss": 0.7759,
+      "step": 16740
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5787156291181923,
+      "learning_rate": 1.3807439567364246e-05,
+      "loss": 0.7904,
+      "step": 16741
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5278574321896974,
+      "learning_rate": 1.3806733987086597e-05,
+      "loss": 0.8018,
+      "step": 16742
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.50031668766581,
+      "learning_rate": 1.380602838464494e-05,
+      "loss": 0.6978,
+      "step": 16743
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4908141548300566,
+      "learning_rate": 1.3805322760043378e-05,
+      "loss": 0.7697,
+      "step": 16744
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6699812852911475,
+      "learning_rate": 1.3804617113286027e-05,
+      "loss": 0.7421,
+      "step": 16745
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6161865172836197,
+      "learning_rate": 1.3803911444376984e-05,
+      "loss": 0.5338,
+      "step": 16746
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5554878564950636,
+      "learning_rate": 1.380320575332037e-05,
+      "loss": 0.6948,
+      "step": 16747
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3522344592245374,
+      "learning_rate": 1.3802500040120285e-05,
+      "loss": 0.7428,
+      "step": 16748
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7891442805349709,
+      "learning_rate": 1.3801794304780843e-05,
+      "loss": 0.7042,
+      "step": 16749
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.365694537442721,
+      "learning_rate": 1.3801088547306149e-05,
+      "loss": 0.7726,
+      "step": 16750
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3535431159870672,
+      "learning_rate": 1.3800382767700316e-05,
+      "loss": 0.8017,
+      "step": 16751
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4751831899670997,
+      "learning_rate": 1.3799676965967451e-05,
+      "loss": 0.8097,
+      "step": 16752
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8096323535670265,
+      "learning_rate": 1.3798971142111661e-05,
+      "loss": 0.776,
+      "step": 16753
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.381479432031139,
+      "learning_rate": 1.379826529613706e-05,
+      "loss": 0.639,
+      "step": 16754
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5400662895911301,
+      "learning_rate": 1.3797559428047755e-05,
+      "loss": 0.7346,
+      "step": 16755
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3373723214195585,
+      "learning_rate": 1.3796853537847859e-05,
+      "loss": 0.7499,
+      "step": 16756
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3565978838070367,
+      "learning_rate": 1.3796147625541479e-05,
+      "loss": 0.6804,
+      "step": 16757
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4842999417180487,
+      "learning_rate": 1.3795441691132726e-05,
+      "loss": 0.785,
+      "step": 16758
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5187229245085865,
+      "learning_rate": 1.3794735734625708e-05,
+      "loss": 0.6574,
+      "step": 16759
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3710723126759226,
+      "learning_rate": 1.3794029756024536e-05,
+      "loss": 0.8759,
+      "step": 16760
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3845640512432449,
+      "learning_rate": 1.3793323755333323e-05,
+      "loss": 0.7602,
+      "step": 16761
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5886747855572094,
+      "learning_rate": 1.3792617732556182e-05,
+      "loss": 0.7735,
+      "step": 16762
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5130255589860029,
+      "learning_rate": 1.3791911687697215e-05,
+      "loss": 0.7273,
+      "step": 16763
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5737775345847088,
+      "learning_rate": 1.3791205620760541e-05,
+      "loss": 0.5573,
+      "step": 16764
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4364121256969473,
+      "learning_rate": 1.3790499531750262e-05,
+      "loss": 0.7831,
+      "step": 16765
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5388968695786616,
+      "learning_rate": 1.3789793420670503e-05,
+      "loss": 0.5836,
+      "step": 16766
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8806270950041921,
+      "learning_rate": 1.378908728752536e-05,
+      "loss": 0.776,
+      "step": 16767
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9198194469382244,
+      "learning_rate": 1.3788381132318955e-05,
+      "loss": 0.7646,
+      "step": 16768
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.752552909084214,
+      "learning_rate": 1.3787674955055395e-05,
+      "loss": 0.6438,
+      "step": 16769
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2777292405456344,
+      "learning_rate": 1.3786968755738792e-05,
+      "loss": 0.7027,
+      "step": 16770
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7119044637876775,
+      "learning_rate": 1.3786262534373256e-05,
+      "loss": 0.7096,
+      "step": 16771
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7601709088431858,
+      "learning_rate": 1.3785556290962902e-05,
+      "loss": 0.6971,
+      "step": 16772
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4586894343121746,
+      "learning_rate": 1.3784850025511842e-05,
+      "loss": 0.7641,
+      "step": 16773
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3283008308611968,
+      "learning_rate": 1.3784143738024188e-05,
+      "loss": 0.6896,
+      "step": 16774
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3872438013107198,
+      "learning_rate": 1.3783437428504048e-05,
+      "loss": 0.7843,
+      "step": 16775
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.522469429150103,
+      "learning_rate": 1.3782731096955537e-05,
+      "loss": 0.7188,
+      "step": 16776
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.91377560892067,
+      "learning_rate": 1.378202474338277e-05,
+      "loss": 0.8414,
+      "step": 16777
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5736013996686826,
+      "learning_rate": 1.3781318367789858e-05,
+      "loss": 0.7261,
+      "step": 16778
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5330332996268423,
+      "learning_rate": 1.3780611970180913e-05,
+      "loss": 0.7422,
+      "step": 16779
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.35055222479128,
+      "learning_rate": 1.3779905550560047e-05,
+      "loss": 0.6479,
+      "step": 16780
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8554951788720349,
+      "learning_rate": 1.3779199108931374e-05,
+      "loss": 0.6393,
+      "step": 16781
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3741082417652815,
+      "learning_rate": 1.3778492645299008e-05,
+      "loss": 0.5609,
+      "step": 16782
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.9187090626418097,
+      "learning_rate": 1.3777786159667063e-05,
+      "loss": 0.8434,
+      "step": 16783
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.359555543239932,
+      "learning_rate": 1.3777079652039649e-05,
+      "loss": 0.7135,
+      "step": 16784
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2948985707676661,
+      "learning_rate": 1.3776373122420884e-05,
+      "loss": 0.7026,
+      "step": 16785
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.526691198928625,
+      "learning_rate": 1.3775666570814876e-05,
+      "loss": 0.6553,
+      "step": 16786
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.588148032222278,
+      "learning_rate": 1.3774959997225743e-05,
+      "loss": 0.5104,
+      "step": 16787
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.300173204151964,
+      "learning_rate": 1.3774253401657599e-05,
+      "loss": 0.7403,
+      "step": 16788
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2445993811283487,
+      "learning_rate": 1.3773546784114556e-05,
+      "loss": 0.7196,
+      "step": 16789
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5676547942460687,
+      "learning_rate": 1.377284014460073e-05,
+      "loss": 0.4757,
+      "step": 16790
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4718270294903373,
+      "learning_rate": 1.3772133483120229e-05,
+      "loss": 0.6724,
+      "step": 16791
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4056801628000308,
+      "learning_rate": 1.377142679967718e-05,
+      "loss": 0.7966,
+      "step": 16792
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5533222512846947,
+      "learning_rate": 1.3770720094275685e-05,
+      "loss": 0.7351,
+      "step": 16793
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4548017180494857,
+      "learning_rate": 1.3770013366919867e-05,
+      "loss": 0.7273,
+      "step": 16794
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4830880800426145,
+      "learning_rate": 1.3769306617613833e-05,
+      "loss": 0.7801,
+      "step": 16795
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5988582075474679,
+      "learning_rate": 1.3768599846361707e-05,
+      "loss": 0.5521,
+      "step": 16796
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7032505049684887,
+      "learning_rate": 1.37678930531676e-05,
+      "loss": 0.8144,
+      "step": 16797
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3029501638317413,
+      "learning_rate": 1.3767186238035624e-05,
+      "loss": 0.6622,
+      "step": 16798
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.444554257468355,
+      "learning_rate": 1.3766479400969899e-05,
+      "loss": 0.7961,
+      "step": 16799
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0230578234893706,
+      "learning_rate": 1.3765772541974536e-05,
+      "loss": 0.685,
+      "step": 16800
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3432999434274844,
+      "learning_rate": 1.3765065661053654e-05,
+      "loss": 0.7956,
+      "step": 16801
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.821193483169175,
+      "learning_rate": 1.376435875821137e-05,
+      "loss": 0.7062,
+      "step": 16802
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5745104834150638,
+      "learning_rate": 1.3763651833451795e-05,
+      "loss": 0.75,
+      "step": 16803
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4220527993842489,
+      "learning_rate": 1.3762944886779049e-05,
+      "loss": 0.6885,
+      "step": 16804
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4473818177004694,
+      "learning_rate": 1.3762237918197245e-05,
+      "loss": 0.8516,
+      "step": 16805
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.544497445926837,
+      "learning_rate": 1.37615309277105e-05,
+      "loss": 0.7942,
+      "step": 16806
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.33213985008624,
+      "learning_rate": 1.3760823915322934e-05,
+      "loss": 0.6648,
+      "step": 16807
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3596529031492672,
+      "learning_rate": 1.3760116881038658e-05,
+      "loss": 0.6843,
+      "step": 16808
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4777216221052207,
+      "learning_rate": 1.375940982486179e-05,
+      "loss": 0.7724,
+      "step": 16809
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.40561064854482,
+      "learning_rate": 1.3758702746796454e-05,
+      "loss": 0.7601,
+      "step": 16810
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2691220907269896,
+      "learning_rate": 1.3757995646846755e-05,
+      "loss": 0.7412,
+      "step": 16811
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.441328389723607,
+      "learning_rate": 1.3757288525016816e-05,
+      "loss": 0.8638,
+      "step": 16812
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.601937357443814,
+      "learning_rate": 1.375658138131076e-05,
+      "loss": 0.6613,
+      "step": 16813
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5450171147294083,
+      "learning_rate": 1.3755874215732687e-05,
+      "loss": 0.7402,
+      "step": 16814
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3538931345577814,
+      "learning_rate": 1.3755167028286733e-05,
+      "loss": 0.6993,
+      "step": 16815
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9350885576512893,
+      "learning_rate": 1.3754459818977007e-05,
+      "loss": 0.7845,
+      "step": 16816
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.793444648216533,
+      "learning_rate": 1.3753752587807627e-05,
+      "loss": 0.7935,
+      "step": 16817
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6568483780975847,
+      "learning_rate": 1.3753045334782707e-05,
+      "loss": 0.57,
+      "step": 16818
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6536139676596642,
+      "learning_rate": 1.3752338059906374e-05,
+      "loss": 0.837,
+      "step": 16819
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6231599037598217,
+      "learning_rate": 1.3751630763182738e-05,
+      "loss": 0.7506,
+      "step": 16820
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2861273232263588,
+      "learning_rate": 1.375092344461592e-05,
+      "loss": 0.7057,
+      "step": 16821
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7524967156725448,
+      "learning_rate": 1.375021610421004e-05,
+      "loss": 0.5712,
+      "step": 16822
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3601065410077233,
+      "learning_rate": 1.3749508741969213e-05,
+      "loss": 0.825,
+      "step": 16823
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4159374393467148,
+      "learning_rate": 1.3748801357897559e-05,
+      "loss": 0.7215,
+      "step": 16824
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.964611362806988,
+      "learning_rate": 1.3748093951999195e-05,
+      "loss": 0.8166,
+      "step": 16825
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3520358926854643,
+      "learning_rate": 1.3747386524278246e-05,
+      "loss": 0.802,
+      "step": 16826
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3201496030876365,
+      "learning_rate": 1.3746679074738822e-05,
+      "loss": 0.7304,
+      "step": 16827
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9335622078301873,
+      "learning_rate": 1.3745971603385049e-05,
+      "loss": 0.7826,
+      "step": 16828
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.8236894664447876,
+      "learning_rate": 1.3745264110221042e-05,
+      "loss": 0.6938,
+      "step": 16829
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.021895591495711,
+      "learning_rate": 1.3744556595250922e-05,
+      "loss": 0.8808,
+      "step": 16830
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.360498976674625,
+      "learning_rate": 1.3743849058478809e-05,
+      "loss": 0.7747,
+      "step": 16831
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0069160222182867,
+      "learning_rate": 1.374314149990882e-05,
+      "loss": 0.7213,
+      "step": 16832
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8734875258867036,
+      "learning_rate": 1.3742433919545076e-05,
+      "loss": 0.799,
+      "step": 16833
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5211470077973044,
+      "learning_rate": 1.3741726317391699e-05,
+      "loss": 0.8497,
+      "step": 16834
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4534350952974602,
+      "learning_rate": 1.3741018693452803e-05,
+      "loss": 0.8325,
+      "step": 16835
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.540051514566742,
+      "learning_rate": 1.3740311047732514e-05,
+      "loss": 0.7621,
+      "step": 16836
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.45181232005589,
+      "learning_rate": 1.373960338023495e-05,
+      "loss": 0.8294,
+      "step": 16837
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.41466110549502,
+      "learning_rate": 1.3738895690964231e-05,
+      "loss": 0.7983,
+      "step": 16838
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2759400995460162,
+      "learning_rate": 1.3738187979924478e-05,
+      "loss": 0.6967,
+      "step": 16839
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5512993854321446,
+      "learning_rate": 1.373748024711981e-05,
+      "loss": 0.7748,
+      "step": 16840
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4170294211615448,
+      "learning_rate": 1.3736772492554352e-05,
+      "loss": 0.7435,
+      "step": 16841
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3042344184774664,
+      "learning_rate": 1.3736064716232216e-05,
+      "loss": 0.6561,
+      "step": 16842
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.337409776802845,
+      "learning_rate": 1.3735356918157532e-05,
+      "loss": 0.7231,
+      "step": 16843
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.577044928452477,
+      "learning_rate": 1.3734649098334416e-05,
+      "loss": 0.4943,
+      "step": 16844
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3344065342503324,
+      "learning_rate": 1.3733941256766994e-05,
+      "loss": 0.714,
+      "step": 16845
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5629695473103578,
+      "learning_rate": 1.3733233393459377e-05,
+      "loss": 0.7618,
+      "step": 16846
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.896382619476104,
+      "learning_rate": 1.3732525508415702e-05,
+      "loss": 0.6926,
+      "step": 16847
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2445791149384489,
+      "learning_rate": 1.3731817601640077e-05,
+      "loss": 0.7745,
+      "step": 16848
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3961069995159885,
+      "learning_rate": 1.3731109673136627e-05,
+      "loss": 0.6627,
+      "step": 16849
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3909150737251603,
+      "learning_rate": 1.3730401722909479e-05,
+      "loss": 0.7625,
+      "step": 16850
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6020763885432534,
+      "learning_rate": 1.3729693750962747e-05,
+      "loss": 0.7261,
+      "step": 16851
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4011274461353942,
+      "learning_rate": 1.372898575730056e-05,
+      "loss": 0.7914,
+      "step": 16852
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4278445183309725,
+      "learning_rate": 1.3728277741927037e-05,
+      "loss": 0.7874,
+      "step": 16853
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7739215075114299,
+      "learning_rate": 1.37275697048463e-05,
+      "loss": 0.676,
+      "step": 16854
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4170011052880291,
+      "learning_rate": 1.3726861646062472e-05,
+      "loss": 0.7763,
+      "step": 16855
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8448479837042766,
+      "learning_rate": 1.3726153565579676e-05,
+      "loss": 0.7016,
+      "step": 16856
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.508363007524278,
+      "learning_rate": 1.3725445463402034e-05,
+      "loss": 0.7369,
+      "step": 16857
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.524179860905176,
+      "learning_rate": 1.3724737339533669e-05,
+      "loss": 0.7739,
+      "step": 16858
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6251121134633137,
+      "learning_rate": 1.3724029193978706e-05,
+      "loss": 0.6802,
+      "step": 16859
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2202166404625983,
+      "learning_rate": 1.3723321026741262e-05,
+      "loss": 0.5746,
+      "step": 16860
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.433147861708266,
+      "learning_rate": 1.3722612837825466e-05,
+      "loss": 0.7215,
+      "step": 16861
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8974331634802284,
+      "learning_rate": 1.3721904627235444e-05,
+      "loss": 0.7234,
+      "step": 16862
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5924214063826754,
+      "learning_rate": 1.3721196394975306e-05,
+      "loss": 0.6809,
+      "step": 16863
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3553082325733825,
+      "learning_rate": 1.3720488141049193e-05,
+      "loss": 0.8465,
+      "step": 16864
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3997863067951186,
+      "learning_rate": 1.3719779865461214e-05,
+      "loss": 0.6263,
+      "step": 16865
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9356010735837044,
+      "learning_rate": 1.3719071568215503e-05,
+      "loss": 0.8211,
+      "step": 16866
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6241429367288522,
+      "learning_rate": 1.3718363249316179e-05,
+      "loss": 0.5749,
+      "step": 16867
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3890123832067705,
+      "learning_rate": 1.3717654908767366e-05,
+      "loss": 0.7087,
+      "step": 16868
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3485759701267628,
+      "learning_rate": 1.3716946546573186e-05,
+      "loss": 0.7596,
+      "step": 16869
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5603007933943747,
+      "learning_rate": 1.3716238162737772e-05,
+      "loss": 0.7491,
+      "step": 16870
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5120228799242232,
+      "learning_rate": 1.3715529757265243e-05,
+      "loss": 0.6411,
+      "step": 16871
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2644851685270009,
+      "learning_rate": 1.371482133015972e-05,
+      "loss": 0.7354,
+      "step": 16872
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4794076221638714,
+      "learning_rate": 1.3714112881425332e-05,
+      "loss": 0.7067,
+      "step": 16873
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5439558161806373,
+      "learning_rate": 1.3713404411066203e-05,
+      "loss": 0.6377,
+      "step": 16874
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5716358566658797,
+      "learning_rate": 1.3712695919086458e-05,
+      "loss": 0.8476,
+      "step": 16875
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.449679014979394,
+      "learning_rate": 1.371198740549022e-05,
+      "loss": 0.809,
+      "step": 16876
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.996285850874445,
+      "learning_rate": 1.3711278870281622e-05,
+      "loss": 0.7654,
+      "step": 16877
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.621028229544824,
+      "learning_rate": 1.3710570313464778e-05,
+      "loss": 0.7345,
+      "step": 16878
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5729584395293255,
+      "learning_rate": 1.3709861735043818e-05,
+      "loss": 0.5898,
+      "step": 16879
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5889936218940065,
+      "learning_rate": 1.3709153135022873e-05,
+      "loss": 0.5169,
+      "step": 16880
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7112021631745553,
+      "learning_rate": 1.3708444513406062e-05,
+      "loss": 0.6774,
+      "step": 16881
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5992542372304839,
+      "learning_rate": 1.3707735870197514e-05,
+      "loss": 0.7521,
+      "step": 16882
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4772470811378267,
+      "learning_rate": 1.3707027205401353e-05,
+      "loss": 0.8207,
+      "step": 16883
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.717172820106436,
+      "learning_rate": 1.3706318519021704e-05,
+      "loss": 0.723,
+      "step": 16884
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6399286712557495,
+      "learning_rate": 1.37056098110627e-05,
+      "loss": 0.5484,
+      "step": 16885
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.398994146450334,
+      "learning_rate": 1.3704901081528457e-05,
+      "loss": 0.7603,
+      "step": 16886
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6818169116576072,
+      "learning_rate": 1.3704192330423109e-05,
+      "loss": 0.7096,
+      "step": 16887
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.421545999143561,
+      "learning_rate": 1.3703483557750783e-05,
+      "loss": 0.7425,
+      "step": 16888
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4675706949643557,
+      "learning_rate": 1.3702774763515601e-05,
+      "loss": 0.7585,
+      "step": 16889
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4240450418399577,
+      "learning_rate": 1.3702065947721691e-05,
+      "loss": 0.6821,
+      "step": 16890
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3270700248676524,
+      "learning_rate": 1.3701357110373184e-05,
+      "loss": 0.8113,
+      "step": 16891
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3708226753690262,
+      "learning_rate": 1.3700648251474204e-05,
+      "loss": 0.8522,
+      "step": 16892
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2810279342710673,
+      "learning_rate": 1.3699939371028873e-05,
+      "loss": 0.693,
+      "step": 16893
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9284630374021237,
+      "learning_rate": 1.3699230469041331e-05,
+      "loss": 0.731,
+      "step": 16894
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3640774719756117,
+      "learning_rate": 1.3698521545515693e-05,
+      "loss": 0.8553,
+      "step": 16895
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5911978225815306,
+      "learning_rate": 1.3697812600456094e-05,
+      "loss": 0.644,
+      "step": 16896
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6132690954594788,
+      "learning_rate": 1.3697103633866657e-05,
+      "loss": 0.6654,
+      "step": 16897
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5869130835102487,
+      "learning_rate": 1.3696394645751514e-05,
+      "loss": 0.5555,
+      "step": 16898
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.533738725848651,
+      "learning_rate": 1.369568563611479e-05,
+      "loss": 0.8602,
+      "step": 16899
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3997126478264517,
+      "learning_rate": 1.3694976604960615e-05,
+      "loss": 0.6715,
+      "step": 16900
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6613109182923524,
+      "learning_rate": 1.3694267552293117e-05,
+      "loss": 0.6401,
+      "step": 16901
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7689439827947666,
+      "learning_rate": 1.3693558478116425e-05,
+      "loss": 0.8602,
+      "step": 16902
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.586175102690102,
+      "learning_rate": 1.369284938243466e-05,
+      "loss": 0.5306,
+      "step": 16903
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3945568777478163,
+      "learning_rate": 1.3692140265251962e-05,
+      "loss": 0.7552,
+      "step": 16904
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.983062311889723,
+      "learning_rate": 1.3691431126572454e-05,
+      "loss": 0.7189,
+      "step": 16905
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3565721053178836,
+      "learning_rate": 1.3690721966400264e-05,
+      "loss": 0.7701,
+      "step": 16906
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2769636381990985,
+      "learning_rate": 1.3690012784739524e-05,
+      "loss": 0.8039,
+      "step": 16907
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2634370558275059,
+      "learning_rate": 1.368930358159436e-05,
+      "loss": 0.6417,
+      "step": 16908
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4570173619616924,
+      "learning_rate": 1.3688594356968903e-05,
+      "loss": 0.734,
+      "step": 16909
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.67239308610865,
+      "learning_rate": 1.368788511086728e-05,
+      "loss": 0.6466,
+      "step": 16910
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5529241076721092,
+      "learning_rate": 1.3687175843293625e-05,
+      "loss": 0.6767,
+      "step": 16911
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6772474890794047,
+      "learning_rate": 1.3686466554252062e-05,
+      "loss": 0.8126,
+      "step": 16912
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3679840455552037,
+      "learning_rate": 1.3685757243746728e-05,
+      "loss": 0.7436,
+      "step": 16913
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.317662432811865,
+      "learning_rate": 1.3685047911781744e-05,
+      "loss": 0.7562,
+      "step": 16914
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7887981574595664,
+      "learning_rate": 1.368433855836125e-05,
+      "loss": 0.8114,
+      "step": 16915
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5655104916627575,
+      "learning_rate": 1.3683629183489364e-05,
+      "loss": 0.5454,
+      "step": 16916
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.944906765098912,
+      "learning_rate": 1.3682919787170228e-05,
+      "loss": 0.6491,
+      "step": 16917
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4209319441017887,
+      "learning_rate": 1.3682210369407961e-05,
+      "loss": 0.7794,
+      "step": 16918
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5883688574119672,
+      "learning_rate": 1.3681500930206707e-05,
+      "loss": 0.7828,
+      "step": 16919
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3221674168348234,
+      "learning_rate": 1.3680791469570585e-05,
+      "loss": 0.7649,
+      "step": 16920
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2433741857692764,
+      "learning_rate": 1.3680081987503728e-05,
+      "loss": 0.7422,
+      "step": 16921
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.606364835091866,
+      "learning_rate": 1.3679372484010269e-05,
+      "loss": 0.842,
+      "step": 16922
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3550100566881187,
+      "learning_rate": 1.3678662959094338e-05,
+      "loss": 0.835,
+      "step": 16923
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.454089168964279,
+      "learning_rate": 1.3677953412760072e-05,
+      "loss": 0.7856,
+      "step": 16924
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.38691284588944,
+      "learning_rate": 1.3677243845011589e-05,
+      "loss": 0.7247,
+      "step": 16925
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3776549965943994,
+      "learning_rate": 1.3676534255853036e-05,
+      "loss": 0.6925,
+      "step": 16926
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3265328785096495,
+      "learning_rate": 1.3675824645288533e-05,
+      "loss": 0.6586,
+      "step": 16927
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6000945995763555,
+      "learning_rate": 1.3675115013322215e-05,
+      "loss": 0.5718,
+      "step": 16928
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.607161221171331,
+      "learning_rate": 1.3674405359958214e-05,
+      "loss": 0.5499,
+      "step": 16929
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4496392809076422,
+      "learning_rate": 1.367369568520066e-05,
+      "loss": 0.6771,
+      "step": 16930
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.356791193065891,
+      "learning_rate": 1.367298598905369e-05,
+      "loss": 0.8,
+      "step": 16931
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.589990325723563,
+      "learning_rate": 1.367227627152143e-05,
+      "loss": 0.5725,
+      "step": 16932
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2474109407146494,
+      "learning_rate": 1.3671566532608015e-05,
+      "loss": 0.6682,
+      "step": 16933
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6629795404337342,
+      "learning_rate": 1.367085677231758e-05,
+      "loss": 0.564,
+      "step": 16934
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5283135737420623,
+      "learning_rate": 1.3670146990654252e-05,
+      "loss": 0.7755,
+      "step": 16935
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7545404440214558,
+      "learning_rate": 1.3669437187622168e-05,
+      "loss": 0.7944,
+      "step": 16936
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9107771143565884,
+      "learning_rate": 1.3668727363225459e-05,
+      "loss": 0.8488,
+      "step": 16937
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7606092651000735,
+      "learning_rate": 1.3668017517468255e-05,
+      "loss": 0.7135,
+      "step": 16938
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5484083741915728,
+      "learning_rate": 1.3667307650354693e-05,
+      "loss": 0.8812,
+      "step": 16939
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.377083813268617,
+      "learning_rate": 1.3666597761888906e-05,
+      "loss": 0.7681,
+      "step": 16940
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.0335041843078674,
+      "learning_rate": 1.3665887852075026e-05,
+      "loss": 0.7584,
+      "step": 16941
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5355030034017667,
+      "learning_rate": 1.3665177920917185e-05,
+      "loss": 0.7179,
+      "step": 16942
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8089772726756188,
+      "learning_rate": 1.3664467968419522e-05,
+      "loss": 0.6246,
+      "step": 16943
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3223797574553764,
+      "learning_rate": 1.366375799458616e-05,
+      "loss": 0.7127,
+      "step": 16944
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5436945974017522,
+      "learning_rate": 1.3663047999421243e-05,
+      "loss": 0.7995,
+      "step": 16945
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7117318852354604,
+      "learning_rate": 1.36623379829289e-05,
+      "loss": 0.6983,
+      "step": 16946
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.678684679983646,
+      "learning_rate": 1.3661627945113268e-05,
+      "loss": 0.8882,
+      "step": 16947
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4619474991487664,
+      "learning_rate": 1.3660917885978473e-05,
+      "loss": 0.7797,
+      "step": 16948
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.346234000423813,
+      "learning_rate": 1.366020780552866e-05,
+      "loss": 0.7246,
+      "step": 16949
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5362423398222325,
+      "learning_rate": 1.3659497703767957e-05,
+      "loss": 0.7296,
+      "step": 16950
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.789451599283735,
+      "learning_rate": 1.36587875807005e-05,
+      "loss": 0.755,
+      "step": 16951
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2889045276430025,
+      "learning_rate": 1.3658077436330421e-05,
+      "loss": 0.687,
+      "step": 16952
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.297597704083487,
+      "learning_rate": 1.365736727066186e-05,
+      "loss": 0.6319,
+      "step": 16953
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4522235274802142,
+      "learning_rate": 1.3656657083698947e-05,
+      "loss": 0.6869,
+      "step": 16954
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4763354719694306,
+      "learning_rate": 1.365594687544582e-05,
+      "loss": 0.7263,
+      "step": 16955
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.734123385147771,
+      "learning_rate": 1.3655236645906613e-05,
+      "loss": 0.8384,
+      "step": 16956
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6117090944052377,
+      "learning_rate": 1.3654526395085461e-05,
+      "loss": 0.6518,
+      "step": 16957
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6138330881624872,
+      "learning_rate": 1.3653816122986499e-05,
+      "loss": 0.8356,
+      "step": 16958
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.459665589176874,
+      "learning_rate": 1.3653105829613862e-05,
+      "loss": 0.722,
+      "step": 16959
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6138493642866034,
+      "learning_rate": 1.3652395514971686e-05,
+      "loss": 0.7683,
+      "step": 16960
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5385367708202367,
+      "learning_rate": 1.365168517906411e-05,
+      "loss": 0.7633,
+      "step": 16961
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4360915935550083,
+      "learning_rate": 1.3650974821895262e-05,
+      "loss": 0.8469,
+      "step": 16962
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4446342540544121,
+      "learning_rate": 1.3650264443469285e-05,
+      "loss": 0.7104,
+      "step": 16963
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5782251030657957,
+      "learning_rate": 1.3649554043790316e-05,
+      "loss": 0.8119,
+      "step": 16964
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5038882317756377,
+      "learning_rate": 1.3648843622862483e-05,
+      "loss": 0.7816,
+      "step": 16965
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6672637267995898,
+      "learning_rate": 1.3648133180689932e-05,
+      "loss": 0.8007,
+      "step": 16966
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.522037242382821,
+      "learning_rate": 1.3647422717276788e-05,
+      "loss": 0.6776,
+      "step": 16967
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.620038878702996,
+      "learning_rate": 1.3646712232627201e-05,
+      "loss": 0.7809,
+      "step": 16968
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3893953922728242,
+      "learning_rate": 1.3646001726745298e-05,
+      "loss": 0.7856,
+      "step": 16969
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4965751553571198,
+      "learning_rate": 1.3645291199635218e-05,
+      "loss": 0.6637,
+      "step": 16970
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.1422384071444736,
+      "learning_rate": 1.36445806513011e-05,
+      "loss": 0.7844,
+      "step": 16971
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.704138151571202,
+      "learning_rate": 1.3643870081747076e-05,
+      "loss": 0.6166,
+      "step": 16972
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4965656922247357,
+      "learning_rate": 1.364315949097729e-05,
+      "loss": 0.6849,
+      "step": 16973
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.531405818826942,
+      "learning_rate": 1.3642448878995876e-05,
+      "loss": 0.7317,
+      "step": 16974
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6231634033986034,
+      "learning_rate": 1.3641738245806973e-05,
+      "loss": 0.5741,
+      "step": 16975
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4858167028156843,
+      "learning_rate": 1.364102759141471e-05,
+      "loss": 0.695,
+      "step": 16976
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3384420506626298,
+      "learning_rate": 1.3640316915823242e-05,
+      "loss": 0.7161,
+      "step": 16977
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.514197104508984,
+      "learning_rate": 1.363960621903669e-05,
+      "loss": 0.7681,
+      "step": 16978
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9739466748415522,
+      "learning_rate": 1.36388955010592e-05,
+      "loss": 0.7159,
+      "step": 16979
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.172013664373973,
+      "learning_rate": 1.3638184761894907e-05,
+      "loss": 0.7041,
+      "step": 16980
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3288905391958967,
+      "learning_rate": 1.363747400154795e-05,
+      "loss": 0.7727,
+      "step": 16981
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5910630560702437,
+      "learning_rate": 1.3636763220022468e-05,
+      "loss": 0.559,
+      "step": 16982
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.340596147586225,
+      "learning_rate": 1.3636052417322601e-05,
+      "loss": 0.8162,
+      "step": 16983
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6072374793967634,
+      "learning_rate": 1.3635341593452484e-05,
+      "loss": 0.7535,
+      "step": 16984
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5922834247259028,
+      "learning_rate": 1.3634630748416257e-05,
+      "loss": 0.7056,
+      "step": 16985
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5180609440850965,
+      "learning_rate": 1.3633919882218063e-05,
+      "loss": 0.8129,
+      "step": 16986
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7656447954305754,
+      "learning_rate": 1.3633208994862033e-05,
+      "loss": 0.6787,
+      "step": 16987
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5903412329306104,
+      "learning_rate": 1.363249808635231e-05,
+      "loss": 0.7109,
+      "step": 16988
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8040667440717242,
+      "learning_rate": 1.3631787156693035e-05,
+      "loss": 0.8471,
+      "step": 16989
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4966047650083831,
+      "learning_rate": 1.3631076205888344e-05,
+      "loss": 0.7053,
+      "step": 16990
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4393215167096205,
+      "learning_rate": 1.363036523394238e-05,
+      "loss": 0.9071,
+      "step": 16991
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3843185047601556,
+      "learning_rate": 1.3629654240859279e-05,
+      "loss": 0.7392,
+      "step": 16992
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5961189386948671,
+      "learning_rate": 1.362894322664318e-05,
+      "loss": 0.5147,
+      "step": 16993
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4325089300897236,
+      "learning_rate": 1.3628232191298227e-05,
+      "loss": 0.7468,
+      "step": 16994
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4567376120959064,
+      "learning_rate": 1.3627521134828553e-05,
+      "loss": 0.7737,
+      "step": 16995
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3307604810605285,
+      "learning_rate": 1.3626810057238311e-05,
+      "loss": 0.7209,
+      "step": 16996
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4170717860042452,
+      "learning_rate": 1.3626098958531624e-05,
+      "loss": 0.7173,
+      "step": 16997
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3244280795277807,
+      "learning_rate": 1.3625387838712647e-05,
+      "loss": 0.7175,
+      "step": 16998
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5858089665906715,
+      "learning_rate": 1.3624676697785511e-05,
+      "loss": 0.5827,
+      "step": 16999
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6465799854452337,
+      "learning_rate": 1.3623965535754363e-05,
+      "loss": 0.7382,
+      "step": 17000
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9275478341520815,
+      "learning_rate": 1.3623254352623336e-05,
+      "loss": 0.6899,
+      "step": 17001
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5930591597408597,
+      "learning_rate": 1.3622543148396577e-05,
+      "loss": 0.5494,
+      "step": 17002
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8803558248616525,
+      "learning_rate": 1.3621831923078224e-05,
+      "loss": 0.7182,
+      "step": 17003
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5444516159562491,
+      "learning_rate": 1.362112067667242e-05,
+      "loss": 0.7022,
+      "step": 17004
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4275848664105164,
+      "learning_rate": 1.3620409409183304e-05,
+      "loss": 0.8505,
+      "step": 17005
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0776914597961107,
+      "learning_rate": 1.3619698120615019e-05,
+      "loss": 0.7341,
+      "step": 17006
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5318095847675337,
+      "learning_rate": 1.3618986810971707e-05,
+      "loss": 0.7283,
+      "step": 17007
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.559549417442341,
+      "learning_rate": 1.3618275480257506e-05,
+      "loss": 0.5847,
+      "step": 17008
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5616042825285865,
+      "learning_rate": 1.361756412847656e-05,
+      "loss": 0.6214,
+      "step": 17009
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5716192459592185,
+      "learning_rate": 1.361685275563301e-05,
+      "loss": 0.5409,
+      "step": 17010
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.557086433445644,
+      "learning_rate": 1.3616141361730999e-05,
+      "loss": 0.8259,
+      "step": 17011
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6078523412203192,
+      "learning_rate": 1.3615429946774668e-05,
+      "loss": 0.5612,
+      "step": 17012
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.175192274928348,
+      "learning_rate": 1.3614718510768156e-05,
+      "loss": 0.7833,
+      "step": 17013
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9847179398705372,
+      "learning_rate": 1.3614007053715612e-05,
+      "loss": 0.7448,
+      "step": 17014
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5490794501181915,
+      "learning_rate": 1.3613295575621177e-05,
+      "loss": 0.6532,
+      "step": 17015
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.169949600559137,
+      "learning_rate": 1.3612584076488986e-05,
+      "loss": 0.8002,
+      "step": 17016
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5895551266603214,
+      "learning_rate": 1.361187255632319e-05,
+      "loss": 0.5953,
+      "step": 17017
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2836953807559248,
+      "learning_rate": 1.3611161015127926e-05,
+      "loss": 0.8277,
+      "step": 17018
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2832918125806507,
+      "learning_rate": 1.3610449452907345e-05,
+      "loss": 0.6865,
+      "step": 17019
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5087687128837382,
+      "learning_rate": 1.3609737869665579e-05,
+      "loss": 0.8769,
+      "step": 17020
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7537051281906215,
+      "learning_rate": 1.3609026265406775e-05,
+      "loss": 0.7136,
+      "step": 17021
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.503341791474258,
+      "learning_rate": 1.3608314640135084e-05,
+      "loss": 0.7451,
+      "step": 17022
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5500367280344651,
+      "learning_rate": 1.3607602993854637e-05,
+      "loss": 0.7249,
+      "step": 17023
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5833332540073733,
+      "learning_rate": 1.3606891326569584e-05,
+      "loss": 0.5932,
+      "step": 17024
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2504816685265825,
+      "learning_rate": 1.3606179638284067e-05,
+      "loss": 0.676,
+      "step": 17025
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.53946013296991,
+      "learning_rate": 1.3605467929002233e-05,
+      "loss": 0.8199,
+      "step": 17026
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4731590663942313,
+      "learning_rate": 1.3604756198728222e-05,
+      "loss": 0.657,
+      "step": 17027
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4690937459311293,
+      "learning_rate": 1.3604044447466182e-05,
+      "loss": 0.6348,
+      "step": 17028
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6276746865542178,
+      "learning_rate": 1.360333267522025e-05,
+      "loss": 0.7361,
+      "step": 17029
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5066855206747924,
+      "learning_rate": 1.3602620881994577e-05,
+      "loss": 0.7252,
+      "step": 17030
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6104423074855684,
+      "learning_rate": 1.3601909067793304e-05,
+      "loss": 0.5483,
+      "step": 17031
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8037420817390954,
+      "learning_rate": 1.3601197232620575e-05,
+      "loss": 0.631,
+      "step": 17032
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5588486402804895,
+      "learning_rate": 1.3600485376480536e-05,
+      "loss": 0.5089,
+      "step": 17033
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8281533417331775,
+      "learning_rate": 1.3599773499377333e-05,
+      "loss": 0.7759,
+      "step": 17034
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.390929418813302,
+      "learning_rate": 1.359906160131511e-05,
+      "loss": 0.6083,
+      "step": 17035
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3700906393743713,
+      "learning_rate": 1.3598349682298006e-05,
+      "loss": 0.7223,
+      "step": 17036
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.403134727326494,
+      "learning_rate": 1.3597637742330174e-05,
+      "loss": 0.6698,
+      "step": 17037
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7242036012060409,
+      "learning_rate": 1.3596925781415755e-05,
+      "loss": 0.7413,
+      "step": 17038
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.2479947781558542,
+      "learning_rate": 1.3596213799558898e-05,
+      "loss": 0.8472,
+      "step": 17039
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5516868170466767,
+      "learning_rate": 1.359550179676374e-05,
+      "loss": 0.5358,
+      "step": 17040
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.748689526589152,
+      "learning_rate": 1.3594789773034439e-05,
+      "loss": 0.7524,
+      "step": 17041
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5773918245794024,
+      "learning_rate": 1.3594077728375129e-05,
+      "loss": 0.541,
+      "step": 17042
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4104609295684325,
+      "learning_rate": 1.3593365662789966e-05,
+      "loss": 0.7243,
+      "step": 17043
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9969997357247464,
+      "learning_rate": 1.3592653576283083e-05,
+      "loss": 0.7619,
+      "step": 17044
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3240566434770205,
+      "learning_rate": 1.359194146885864e-05,
+      "loss": 0.83,
+      "step": 17045
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5084682332399164,
+      "learning_rate": 1.359122934052077e-05,
+      "loss": 0.7939,
+      "step": 17046
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7146207837030751,
+      "learning_rate": 1.3590517191273634e-05,
+      "loss": 0.7181,
+      "step": 17047
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7415459974772654,
+      "learning_rate": 1.3589805021121363e-05,
+      "loss": 0.7531,
+      "step": 17048
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.374242831449291,
+      "learning_rate": 1.3589092830068116e-05,
+      "loss": 0.7787,
+      "step": 17049
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5722163871311938,
+      "learning_rate": 1.3588380618118031e-05,
+      "loss": 0.489,
+      "step": 17050
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.050279882429695,
+      "learning_rate": 1.3587668385275258e-05,
+      "loss": 0.7635,
+      "step": 17051
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4770132056788636,
+      "learning_rate": 1.3586956131543944e-05,
+      "loss": 0.8493,
+      "step": 17052
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8424490400989164,
+      "learning_rate": 1.3586243856928238e-05,
+      "loss": 0.7065,
+      "step": 17053
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.601736303855957,
+      "learning_rate": 1.3585531561432284e-05,
+      "loss": 0.5509,
+      "step": 17054
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5758974547214144,
+      "learning_rate": 1.3584819245060228e-05,
+      "loss": 0.7958,
+      "step": 17055
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.380973709553144,
+      "learning_rate": 1.3584106907816222e-05,
+      "loss": 0.8236,
+      "step": 17056
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4577925064023534,
+      "learning_rate": 1.3583394549704404e-05,
+      "loss": 0.7654,
+      "step": 17057
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4796327844979478,
+      "learning_rate": 1.3582682170728939e-05,
+      "loss": 0.7321,
+      "step": 17058
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9840825207174033,
+      "learning_rate": 1.3581969770893959e-05,
+      "loss": 0.625,
+      "step": 17059
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3457711916639523,
+      "learning_rate": 1.3581257350203616e-05,
+      "loss": 0.7539,
+      "step": 17060
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.441642050947104,
+      "learning_rate": 1.3580544908662062e-05,
+      "loss": 0.7144,
+      "step": 17061
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.905293173729917,
+      "learning_rate": 1.3579832446273438e-05,
+      "loss": 0.7611,
+      "step": 17062
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0060437894310796,
+      "learning_rate": 1.35791199630419e-05,
+      "loss": 0.8439,
+      "step": 17063
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6015798821711151,
+      "learning_rate": 1.357840745897159e-05,
+      "loss": 0.5693,
+      "step": 17064
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6576167253401226,
+      "learning_rate": 1.3577694934066659e-05,
+      "loss": 0.756,
+      "step": 17065
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.40296230869304,
+      "learning_rate": 1.3576982388331258e-05,
+      "loss": 0.7037,
+      "step": 17066
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4616931651023393,
+      "learning_rate": 1.357626982176953e-05,
+      "loss": 0.822,
+      "step": 17067
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5028521949810616,
+      "learning_rate": 1.3575557234385628e-05,
+      "loss": 0.7225,
+      "step": 17068
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5476895267924438,
+      "learning_rate": 1.3574844626183698e-05,
+      "loss": 0.6224,
+      "step": 17069
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3957574672996242,
+      "learning_rate": 1.3574131997167891e-05,
+      "loss": 0.7691,
+      "step": 17070
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5882183731967472,
+      "learning_rate": 1.3573419347342358e-05,
+      "loss": 0.5423,
+      "step": 17071
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3642265584230424,
+      "learning_rate": 1.3572706676711247e-05,
+      "loss": 0.6674,
+      "step": 17072
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3451964344861012,
+      "learning_rate": 1.3571993985278706e-05,
+      "loss": 0.7972,
+      "step": 17073
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.451100646303076,
+      "learning_rate": 1.3571281273048885e-05,
+      "loss": 0.7857,
+      "step": 17074
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7419374210714906,
+      "learning_rate": 1.3570568540025933e-05,
+      "loss": 0.6791,
+      "step": 17075
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5849067317859429,
+      "learning_rate": 1.3569855786213998e-05,
+      "loss": 0.8248,
+      "step": 17076
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4346580732526366,
+      "learning_rate": 1.3569143011617238e-05,
+      "loss": 0.802,
+      "step": 17077
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.399396718345566,
+      "learning_rate": 1.3568430216239792e-05,
+      "loss": 0.719,
+      "step": 17078
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7017800223991804,
+      "learning_rate": 1.356771740008582e-05,
+      "loss": 0.8675,
+      "step": 17079
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.320964533417276,
+      "learning_rate": 1.3567004563159467e-05,
+      "loss": 0.6852,
+      "step": 17080
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.531063942975778,
+      "learning_rate": 1.3566291705464883e-05,
+      "loss": 0.8301,
+      "step": 17081
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0129709526862682,
+      "learning_rate": 1.3565578827006218e-05,
+      "loss": 0.6682,
+      "step": 17082
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6941521920365363,
+      "learning_rate": 1.3564865927787627e-05,
+      "loss": 0.8198,
+      "step": 17083
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.646975732822242,
+      "learning_rate": 1.3564153007813257e-05,
+      "loss": 0.8934,
+      "step": 17084
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7442823407409482,
+      "learning_rate": 1.356344006708726e-05,
+      "loss": 0.7364,
+      "step": 17085
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4156818664228528,
+      "learning_rate": 1.3562727105613784e-05,
+      "loss": 0.7203,
+      "step": 17086
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3823414336369406,
+      "learning_rate": 1.3562014123396988e-05,
+      "loss": 0.6919,
+      "step": 17087
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8639749356902768,
+      "learning_rate": 1.3561301120441015e-05,
+      "loss": 0.663,
+      "step": 17088
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5273294569755895,
+      "learning_rate": 1.3560588096750018e-05,
+      "loss": 0.7489,
+      "step": 17089
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3769591197135307,
+      "learning_rate": 1.355987505232815e-05,
+      "loss": 0.786,
+      "step": 17090
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.834386265064941,
+      "learning_rate": 1.3559161987179565e-05,
+      "loss": 0.9436,
+      "step": 17091
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4105055057256812,
+      "learning_rate": 1.3558448901308411e-05,
+      "loss": 0.7905,
+      "step": 17092
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.319934696296801,
+      "learning_rate": 1.355773579471884e-05,
+      "loss": 0.7775,
+      "step": 17093
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.268507836790399,
+      "learning_rate": 1.3557022667415009e-05,
+      "loss": 0.703,
+      "step": 17094
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.522359562232592,
+      "learning_rate": 1.3556309519401061e-05,
+      "loss": 0.7447,
+      "step": 17095
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4582285668310682,
+      "learning_rate": 1.3555596350681158e-05,
+      "loss": 0.8517,
+      "step": 17096
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.01668410820678,
+      "learning_rate": 1.355488316125944e-05,
+      "loss": 0.7859,
+      "step": 17097
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7288132726045822,
+      "learning_rate": 1.3554169951140073e-05,
+      "loss": 0.8473,
+      "step": 17098
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.794903644455501,
+      "learning_rate": 1.35534567203272e-05,
+      "loss": 0.734,
+      "step": 17099
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3582862482485458,
+      "learning_rate": 1.355274346882498e-05,
+      "loss": 0.6865,
+      "step": 17100
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.482808181374582,
+      "learning_rate": 1.355203019663756e-05,
+      "loss": 0.6536,
+      "step": 17101
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.354407682484506,
+      "learning_rate": 1.3551316903769097e-05,
+      "loss": 0.6471,
+      "step": 17102
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4106672515467968,
+      "learning_rate": 1.3550603590223744e-05,
+      "loss": 0.8575,
+      "step": 17103
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.407306560372551,
+      "learning_rate": 1.3549890256005649e-05,
+      "loss": 0.7421,
+      "step": 17104
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.700111830370939,
+      "learning_rate": 1.354917690111897e-05,
+      "loss": 0.7824,
+      "step": 17105
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3954393342072444,
+      "learning_rate": 1.3548463525567858e-05,
+      "loss": 0.7892,
+      "step": 17106
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.956540484467884,
+      "learning_rate": 1.3547750129356474e-05,
+      "loss": 0.7301,
+      "step": 17107
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4375437425738569,
+      "learning_rate": 1.3547036712488961e-05,
+      "loss": 0.7805,
+      "step": 17108
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6125565157218246,
+      "learning_rate": 1.3546323274969479e-05,
+      "loss": 0.5749,
+      "step": 17109
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6325223909519184,
+      "learning_rate": 1.354560981680218e-05,
+      "loss": 0.7554,
+      "step": 17110
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6164562451368911,
+      "learning_rate": 1.3544896337991219e-05,
+      "loss": 0.8864,
+      "step": 17111
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5643349965890878,
+      "learning_rate": 1.3544182838540745e-05,
+      "loss": 0.831,
+      "step": 17112
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5764238406421796,
+      "learning_rate": 1.354346931845492e-05,
+      "loss": 0.7532,
+      "step": 17113
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3262227777009925,
+      "learning_rate": 1.3542755777737892e-05,
+      "loss": 0.7866,
+      "step": 17114
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3512899452136586,
+      "learning_rate": 1.3542042216393825e-05,
+      "loss": 0.8279,
+      "step": 17115
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6730326040911454,
+      "learning_rate": 1.3541328634426858e-05,
+      "loss": 0.7897,
+      "step": 17116
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.637161605383174,
+      "learning_rate": 1.3540615031841163e-05,
+      "loss": 0.829,
+      "step": 17117
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2899011570421586,
+      "learning_rate": 1.353990140864088e-05,
+      "loss": 0.7258,
+      "step": 17118
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4313072542945324,
+      "learning_rate": 1.353918776483017e-05,
+      "loss": 0.6522,
+      "step": 17119
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4500460456160615,
+      "learning_rate": 1.353847410041319e-05,
+      "loss": 0.7069,
+      "step": 17120
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1664858954784874,
+      "learning_rate": 1.3537760415394095e-05,
+      "loss": 0.5907,
+      "step": 17121
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.606346908449786,
+      "learning_rate": 1.3537046709777037e-05,
+      "loss": 0.6874,
+      "step": 17122
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6155455402239844,
+      "learning_rate": 1.3536332983566177e-05,
+      "loss": 0.5287,
+      "step": 17123
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3721672640102653,
+      "learning_rate": 1.3535619236765665e-05,
+      "loss": 0.8011,
+      "step": 17124
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.480939127698756,
+      "learning_rate": 1.3534905469379654e-05,
+      "loss": 0.7499,
+      "step": 17125
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.512595837658051,
+      "learning_rate": 1.3534191681412309e-05,
+      "loss": 0.7286,
+      "step": 17126
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2997541266639776,
+      "learning_rate": 1.353347787286778e-05,
+      "loss": 0.7244,
+      "step": 17127
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0008086433446994,
+      "learning_rate": 1.3532764043750224e-05,
+      "loss": 0.8365,
+      "step": 17128
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.717509519702765,
+      "learning_rate": 1.3532050194063795e-05,
+      "loss": 0.7423,
+      "step": 17129
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5561718928285482,
+      "learning_rate": 1.3531336323812656e-05,
+      "loss": 0.7559,
+      "step": 17130
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9162244431495463,
+      "learning_rate": 1.3530622433000958e-05,
+      "loss": 0.7616,
+      "step": 17131
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6398115373742042,
+      "learning_rate": 1.3529908521632854e-05,
+      "loss": 0.8923,
+      "step": 17132
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.668613732962518,
+      "learning_rate": 1.352919458971251e-05,
+      "loss": 0.783,
+      "step": 17133
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4450615244109402,
+      "learning_rate": 1.3528480637244075e-05,
+      "loss": 0.7076,
+      "step": 17134
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3520438237405585,
+      "learning_rate": 1.3527766664231709e-05,
+      "loss": 0.7094,
+      "step": 17135
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4934171561503953,
+      "learning_rate": 1.352705267067957e-05,
+      "loss": 0.8072,
+      "step": 17136
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5794648430086006,
+      "learning_rate": 1.3526338656591813e-05,
+      "loss": 0.5327,
+      "step": 17137
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3644666271415504,
+      "learning_rate": 1.3525624621972595e-05,
+      "loss": 0.8548,
+      "step": 17138
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7236809197336775,
+      "learning_rate": 1.3524910566826075e-05,
+      "loss": 0.7037,
+      "step": 17139
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5320468413631425,
+      "learning_rate": 1.3524196491156409e-05,
+      "loss": 0.7881,
+      "step": 17140
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4089056736729406,
+      "learning_rate": 1.3523482394967756e-05,
+      "loss": 0.6102,
+      "step": 17141
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4327101249679195,
+      "learning_rate": 1.3522768278264273e-05,
+      "loss": 0.7757,
+      "step": 17142
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7755128924237584,
+      "learning_rate": 1.3522054141050118e-05,
+      "loss": 0.6574,
+      "step": 17143
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.336674128911178,
+      "learning_rate": 1.352133998332945e-05,
+      "loss": 0.7739,
+      "step": 17144
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.789918174979355,
+      "learning_rate": 1.3520625805106425e-05,
+      "loss": 0.7509,
+      "step": 17145
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7054176541624164,
+      "learning_rate": 1.3519911606385199e-05,
+      "loss": 0.7384,
+      "step": 17146
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4985455008344248,
+      "learning_rate": 1.3519197387169939e-05,
+      "loss": 0.6729,
+      "step": 17147
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4212002159788235,
+      "learning_rate": 1.3518483147464792e-05,
+      "loss": 0.6042,
+      "step": 17148
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4232843287517254,
+      "learning_rate": 1.3517768887273923e-05,
+      "loss": 0.8085,
+      "step": 17149
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4480748299927189,
+      "learning_rate": 1.3517054606601492e-05,
+      "loss": 0.7198,
+      "step": 17150
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.231858782655709,
+      "learning_rate": 1.3516340305451654e-05,
+      "loss": 0.7355,
+      "step": 17151
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6197224872502742,
+      "learning_rate": 1.3515625983828572e-05,
+      "loss": 0.6049,
+      "step": 17152
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6363465481911326,
+      "learning_rate": 1.35149116417364e-05,
+      "loss": 0.665,
+      "step": 17153
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3768527442658756,
+      "learning_rate": 1.35141972791793e-05,
+      "loss": 0.8379,
+      "step": 17154
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5008947743226004,
+      "learning_rate": 1.351348289616143e-05,
+      "loss": 0.5734,
+      "step": 17155
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2823669954848065,
+      "learning_rate": 1.3512768492686954e-05,
+      "loss": 0.8665,
+      "step": 17156
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5452657656519524,
+      "learning_rate": 1.351205406876002e-05,
+      "loss": 0.6566,
+      "step": 17157
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.640108749089623,
+      "learning_rate": 1.3511339624384805e-05,
+      "loss": 0.7799,
+      "step": 17158
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.482901289846135,
+      "learning_rate": 1.3510625159565453e-05,
+      "loss": 0.8345,
+      "step": 17159
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.394372434961813,
+      "learning_rate": 1.3509910674306133e-05,
+      "loss": 0.692,
+      "step": 17160
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3446508331642217,
+      "learning_rate": 1.3509196168610999e-05,
+      "loss": 0.7496,
+      "step": 17161
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3601033407856045,
+      "learning_rate": 1.3508481642484215e-05,
+      "loss": 0.908,
+      "step": 17162
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5956285639493725,
+      "learning_rate": 1.350776709592994e-05,
+      "loss": 0.5521,
+      "step": 17163
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9551259941114705,
+      "learning_rate": 1.3507052528952333e-05,
+      "loss": 0.8034,
+      "step": 17164
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4708611341442581,
+      "learning_rate": 1.3506337941555559e-05,
+      "loss": 0.6804,
+      "step": 17165
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2878064475365383,
+      "learning_rate": 1.3505623333743774e-05,
+      "loss": 0.9267,
+      "step": 17166
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6751858654300806,
+      "learning_rate": 1.350490870552114e-05,
+      "loss": 0.9116,
+      "step": 17167
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2873394517231624,
+      "learning_rate": 1.3504194056891817e-05,
+      "loss": 0.7455,
+      "step": 17168
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.080228175638314,
+      "learning_rate": 1.3503479387859967e-05,
+      "loss": 0.6016,
+      "step": 17169
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3390063204666283,
+      "learning_rate": 1.3502764698429753e-05,
+      "loss": 0.7724,
+      "step": 17170
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4201984352940762,
+      "learning_rate": 1.3502049988605332e-05,
+      "loss": 0.5622,
+      "step": 17171
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2450889278692368,
+      "learning_rate": 1.3501335258390868e-05,
+      "loss": 0.7668,
+      "step": 17172
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5509829964976878,
+      "learning_rate": 1.3500620507790523e-05,
+      "loss": 0.6999,
+      "step": 17173
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3958796759896697,
+      "learning_rate": 1.3499905736808455e-05,
+      "loss": 0.8789,
+      "step": 17174
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7042517679535782,
+      "learning_rate": 1.3499190945448828e-05,
+      "loss": 0.7409,
+      "step": 17175
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5563109833288713,
+      "learning_rate": 1.3498476133715803e-05,
+      "loss": 0.6834,
+      "step": 17176
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.49467700214059,
+      "learning_rate": 1.3497761301613543e-05,
+      "loss": 0.8381,
+      "step": 17177
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8260426447367328,
+      "learning_rate": 1.3497046449146209e-05,
+      "loss": 0.7449,
+      "step": 17178
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9164606310657109,
+      "learning_rate": 1.3496331576317964e-05,
+      "loss": 0.6949,
+      "step": 17179
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5843910515703598,
+      "learning_rate": 1.3495616683132969e-05,
+      "loss": 0.5496,
+      "step": 17180
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5809158967856871,
+      "learning_rate": 1.349490176959539e-05,
+      "loss": 0.5562,
+      "step": 17181
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3800501937146714,
+      "learning_rate": 1.3494186835709383e-05,
+      "loss": 0.753,
+      "step": 17182
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.3703571406878106,
+      "learning_rate": 1.3493471881479114e-05,
+      "loss": 0.694,
+      "step": 17183
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.339955211045547,
+      "learning_rate": 1.3492756906908748e-05,
+      "loss": 0.6963,
+      "step": 17184
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2887447692768568,
+      "learning_rate": 1.3492041912002442e-05,
+      "loss": 0.7281,
+      "step": 17185
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5355113578394661,
+      "learning_rate": 1.3491326896764366e-05,
+      "loss": 0.6787,
+      "step": 17186
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.162381546053479,
+      "learning_rate": 1.3490611861198676e-05,
+      "loss": 0.7983,
+      "step": 17187
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5836447500930222,
+      "learning_rate": 1.3489896805309541e-05,
+      "loss": 0.7699,
+      "step": 17188
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2855341788501766,
+      "learning_rate": 1.3489181729101122e-05,
+      "loss": 0.7395,
+      "step": 17189
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2822025804056376,
+      "learning_rate": 1.3488466632577581e-05,
+      "loss": 0.6241,
+      "step": 17190
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.546295754634466,
+      "learning_rate": 1.3487751515743084e-05,
+      "loss": 0.8041,
+      "step": 17191
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.335442771407888,
+      "learning_rate": 1.3487036378601794e-05,
+      "loss": 0.6621,
+      "step": 17192
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.0058375353276157,
+      "learning_rate": 1.3486321221157872e-05,
+      "loss": 0.6269,
+      "step": 17193
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.340033937138832,
+      "learning_rate": 1.3485606043415487e-05,
+      "loss": 0.7427,
+      "step": 17194
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5854092202888241,
+      "learning_rate": 1.3484890845378798e-05,
+      "loss": 0.626,
+      "step": 17195
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4842058315745164,
+      "learning_rate": 1.3484175627051974e-05,
+      "loss": 0.7057,
+      "step": 17196
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6686147538315772,
+      "learning_rate": 1.3483460388439172e-05,
+      "loss": 0.5543,
+      "step": 17197
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6236084694104964,
+      "learning_rate": 1.3482745129544564e-05,
+      "loss": 0.5395,
+      "step": 17198
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.044638689778593,
+      "learning_rate": 1.3482029850372309e-05,
+      "loss": 0.8407,
+      "step": 17199
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7966791465879293,
+      "learning_rate": 1.3481314550926576e-05,
+      "loss": 0.7506,
+      "step": 17200
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7731297787573892,
+      "learning_rate": 1.3480599231211524e-05,
+      "loss": 0.7551,
+      "step": 17201
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6100241493215215,
+      "learning_rate": 1.3479883891231326e-05,
+      "loss": 0.7126,
+      "step": 17202
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3996505864977637,
+      "learning_rate": 1.347916853099014e-05,
+      "loss": 0.6621,
+      "step": 17203
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4285668779108704,
+      "learning_rate": 1.3478453150492132e-05,
+      "loss": 0.7405,
+      "step": 17204
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3064967256725672,
+      "learning_rate": 1.3477737749741473e-05,
+      "loss": 0.6984,
+      "step": 17205
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2951882345067156,
+      "learning_rate": 1.3477022328742318e-05,
+      "loss": 0.7104,
+      "step": 17206
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6419165792199013,
+      "learning_rate": 1.3476306887498842e-05,
+      "loss": 0.79,
+      "step": 17207
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4076034971876894,
+      "learning_rate": 1.3475591426015204e-05,
+      "loss": 0.7776,
+      "step": 17208
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5059271444200177,
+      "learning_rate": 1.3474875944295576e-05,
+      "loss": 0.6233,
+      "step": 17209
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2862313564964125,
+      "learning_rate": 1.3474160442344118e-05,
+      "loss": 0.6648,
+      "step": 17210
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5535164196451255,
+      "learning_rate": 1.3473444920164997e-05,
+      "loss": 0.4929,
+      "step": 17211
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.392345763114352,
+      "learning_rate": 1.3472729377762381e-05,
+      "loss": 0.8015,
+      "step": 17212
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5024360423066707,
+      "learning_rate": 1.3472013815140437e-05,
+      "loss": 0.6802,
+      "step": 17213
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5420952074030279,
+      "learning_rate": 1.3471298232303326e-05,
+      "loss": 0.7663,
+      "step": 17214
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5614521671328463,
+      "learning_rate": 1.347058262925522e-05,
+      "loss": 0.7032,
+      "step": 17215
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.411170571681655,
+      "learning_rate": 1.346986700600028e-05,
+      "loss": 0.7448,
+      "step": 17216
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4073922550434943,
+      "learning_rate": 1.346915136254268e-05,
+      "loss": 0.8471,
+      "step": 17217
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7389157611043458,
+      "learning_rate": 1.3468435698886582e-05,
+      "loss": 0.8792,
+      "step": 17218
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5876797011889348,
+      "learning_rate": 1.3467720015036152e-05,
+      "loss": 0.5064,
+      "step": 17219
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6743063188005682,
+      "learning_rate": 1.3467004310995557e-05,
+      "loss": 0.5882,
+      "step": 17220
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4323206987461758,
+      "learning_rate": 1.3466288586768968e-05,
+      "loss": 0.6544,
+      "step": 17221
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4234079605062309,
+      "learning_rate": 1.3465572842360546e-05,
+      "loss": 0.837,
+      "step": 17222
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3422705353396431,
+      "learning_rate": 1.3464857077774465e-05,
+      "loss": 0.8601,
+      "step": 17223
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4653140193942225,
+      "learning_rate": 1.346414129301489e-05,
+      "loss": 0.6348,
+      "step": 17224
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5843641393484296,
+      "learning_rate": 1.3463425488085982e-05,
+      "loss": 0.7932,
+      "step": 17225
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4259341666403402,
+      "learning_rate": 1.346270966299192e-05,
+      "loss": 0.6923,
+      "step": 17226
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.3262795967861574,
+      "learning_rate": 1.3461993817736861e-05,
+      "loss": 0.8063,
+      "step": 17227
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5901576684886343,
+      "learning_rate": 1.3461277952324982e-05,
+      "loss": 0.7659,
+      "step": 17228
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2126910268267905,
+      "learning_rate": 1.3460562066760444e-05,
+      "loss": 0.7391,
+      "step": 17229
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5593802021434713,
+      "learning_rate": 1.3459846161047421e-05,
+      "loss": 0.6209,
+      "step": 17230
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5789838889443812,
+      "learning_rate": 1.3459130235190077e-05,
+      "loss": 0.9187,
+      "step": 17231
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8533507156829316,
+      "learning_rate": 1.3458414289192582e-05,
+      "loss": 0.722,
+      "step": 17232
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5744332038063956,
+      "learning_rate": 1.3457698323059102e-05,
+      "loss": 0.5291,
+      "step": 17233
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4439809497575589,
+      "learning_rate": 1.3456982336793811e-05,
+      "loss": 0.7815,
+      "step": 17234
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.625801429529567,
+      "learning_rate": 1.3456266330400873e-05,
+      "loss": 0.5543,
+      "step": 17235
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5126063166938934,
+      "learning_rate": 1.3455550303884458e-05,
+      "loss": 0.848,
+      "step": 17236
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.709308554249661,
+      "learning_rate": 1.3454834257248734e-05,
+      "loss": 0.6513,
+      "step": 17237
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.566461357407954,
+      "learning_rate": 1.3454118190497874e-05,
+      "loss": 0.7899,
+      "step": 17238
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3840418805941128,
+      "learning_rate": 1.3453402103636042e-05,
+      "loss": 0.8147,
+      "step": 17239
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.310559030926738,
+      "learning_rate": 1.3452685996667409e-05,
+      "loss": 0.7276,
+      "step": 17240
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.859608741221436,
+      "learning_rate": 1.3451969869596146e-05,
+      "loss": 0.6978,
+      "step": 17241
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6985356026394858,
+      "learning_rate": 1.3451253722426423e-05,
+      "loss": 0.7532,
+      "step": 17242
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4516122650626377,
+      "learning_rate": 1.3450537555162406e-05,
+      "loss": 0.7475,
+      "step": 17243
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5408395200873686,
+      "learning_rate": 1.3449821367808269e-05,
+      "loss": 0.7523,
+      "step": 17244
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.627779051831582,
+      "learning_rate": 1.3449105160368179e-05,
+      "loss": 0.532,
+      "step": 17245
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6217042357032299,
+      "learning_rate": 1.3448388932846306e-05,
+      "loss": 0.7479,
+      "step": 17246
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4320833755845268,
+      "learning_rate": 1.3447672685246822e-05,
+      "loss": 0.7391,
+      "step": 17247
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.770780856936095,
+      "learning_rate": 1.3446956417573894e-05,
+      "loss": 0.8525,
+      "step": 17248
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4452015400029392,
+      "learning_rate": 1.3446240129831699e-05,
+      "loss": 0.7024,
+      "step": 17249
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3153940481743727,
+      "learning_rate": 1.3445523822024398e-05,
+      "loss": 0.7239,
+      "step": 17250
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.240378963368642,
+      "learning_rate": 1.344480749415617e-05,
+      "loss": 0.6616,
+      "step": 17251
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8843208555882027,
+      "learning_rate": 1.3444091146231178e-05,
+      "loss": 0.7118,
+      "step": 17252
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.544894079045056,
+      "learning_rate": 1.3443374778253597e-05,
+      "loss": 0.7661,
+      "step": 17253
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.373407629677958,
+      "learning_rate": 1.3442658390227604e-05,
+      "loss": 0.797,
+      "step": 17254
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3650390383421664,
+      "learning_rate": 1.3441941982157358e-05,
+      "loss": 0.755,
+      "step": 17255
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.598257792597227,
+      "learning_rate": 1.3441225554047039e-05,
+      "loss": 0.6757,
+      "step": 17256
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5362157426219774,
+      "learning_rate": 1.3440509105900813e-05,
+      "loss": 0.7002,
+      "step": 17257
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8315352740300304,
+      "learning_rate": 1.3439792637722859e-05,
+      "loss": 0.7946,
+      "step": 17258
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5148803732923675,
+      "learning_rate": 1.3439076149517337e-05,
+      "loss": 0.7178,
+      "step": 17259
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2769288474924485,
+      "learning_rate": 1.343835964128843e-05,
+      "loss": 0.7446,
+      "step": 17260
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.401318151811503,
+      "learning_rate": 1.3437643113040302e-05,
+      "loss": 0.6623,
+      "step": 17261
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3856607932344407,
+      "learning_rate": 1.3436926564777127e-05,
+      "loss": 0.7722,
+      "step": 17262
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.015892962230916,
+      "learning_rate": 1.3436209996503078e-05,
+      "loss": 0.7018,
+      "step": 17263
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4496619463203677,
+      "learning_rate": 1.3435493408222326e-05,
+      "loss": 0.7807,
+      "step": 17264
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.636163590562691,
+      "learning_rate": 1.3434776799939045e-05,
+      "loss": 0.747,
+      "step": 17265
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7204538949072674,
+      "learning_rate": 1.3434060171657403e-05,
+      "loss": 0.8219,
+      "step": 17266
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2588104225726724,
+      "learning_rate": 1.3433343523381578e-05,
+      "loss": 0.7044,
+      "step": 17267
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.60712012983126,
+      "learning_rate": 1.343262685511574e-05,
+      "loss": 0.7943,
+      "step": 17268
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0847386681049596,
+      "learning_rate": 1.3431910166864062e-05,
+      "loss": 0.6609,
+      "step": 17269
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2057348967704116,
+      "learning_rate": 1.3431193458630716e-05,
+      "loss": 0.6318,
+      "step": 17270
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.74526935847281,
+      "learning_rate": 1.3430476730419874e-05,
+      "loss": 0.7847,
+      "step": 17271
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.414732725061573,
+      "learning_rate": 1.3429759982235712e-05,
+      "loss": 0.8442,
+      "step": 17272
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6297958367530507,
+      "learning_rate": 1.3429043214082403e-05,
+      "loss": 0.6706,
+      "step": 17273
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3949149292506142,
+      "learning_rate": 1.3428326425964115e-05,
+      "loss": 0.5578,
+      "step": 17274
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5047313873945218,
+      "learning_rate": 1.3427609617885029e-05,
+      "loss": 0.712,
+      "step": 17275
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6126909076345783,
+      "learning_rate": 1.3426892789849312e-05,
+      "loss": 0.6611,
+      "step": 17276
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4477657012676086,
+      "learning_rate": 1.3426175941861144e-05,
+      "loss": 0.7357,
+      "step": 17277
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9695339673962826,
+      "learning_rate": 1.3425459073924689e-05,
+      "loss": 0.6413,
+      "step": 17278
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8486252279586994,
+      "learning_rate": 1.3424742186044131e-05,
+      "loss": 0.7093,
+      "step": 17279
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2950274846343464,
+      "learning_rate": 1.3424025278223637e-05,
+      "loss": 0.6878,
+      "step": 17280
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3490023951004624,
+      "learning_rate": 1.3423308350467388e-05,
+      "loss": 0.7698,
+      "step": 17281
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3270072823130172,
+      "learning_rate": 1.342259140277955e-05,
+      "loss": 0.6866,
+      "step": 17282
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.413628297591397,
+      "learning_rate": 1.3421874435164303e-05,
+      "loss": 0.8219,
+      "step": 17283
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6713069066168416,
+      "learning_rate": 1.3421157447625822e-05,
+      "loss": 0.7274,
+      "step": 17284
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2761910449202998,
+      "learning_rate": 1.3420440440168275e-05,
+      "loss": 0.6389,
+      "step": 17285
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.515443469404745,
+      "learning_rate": 1.3419723412795843e-05,
+      "loss": 0.6237,
+      "step": 17286
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5846802283744956,
+      "learning_rate": 1.3419006365512697e-05,
+      "loss": 0.799,
+      "step": 17287
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.814418326789786,
+      "learning_rate": 1.3418289298323015e-05,
+      "loss": 0.7343,
+      "step": 17288
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2855666908189387,
+      "learning_rate": 1.3417572211230969e-05,
+      "loss": 0.6737,
+      "step": 17289
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6913750486336556,
+      "learning_rate": 1.3416855104240739e-05,
+      "loss": 0.7856,
+      "step": 17290
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5451567785349656,
+      "learning_rate": 1.3416137977356491e-05,
+      "loss": 0.9047,
+      "step": 17291
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.711805378837665,
+      "learning_rate": 1.341542083058241e-05,
+      "loss": 0.6959,
+      "step": 17292
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5891032284823251,
+      "learning_rate": 1.3414703663922668e-05,
+      "loss": 0.5444,
+      "step": 17293
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6364670659018056,
+      "learning_rate": 1.3413986477381439e-05,
+      "loss": 0.5602,
+      "step": 17294
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3878296453772367,
+      "learning_rate": 1.3413269270962902e-05,
+      "loss": 0.5994,
+      "step": 17295
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5689634392592409,
+      "learning_rate": 1.341255204467123e-05,
+      "loss": 0.6364,
+      "step": 17296
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4963663791275295,
+      "learning_rate": 1.3411834798510596e-05,
+      "loss": 0.7027,
+      "step": 17297
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5892430604954795,
+      "learning_rate": 1.3411117532485185e-05,
+      "loss": 0.7216,
+      "step": 17298
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.554946725378539,
+      "learning_rate": 1.3410400246599163e-05,
+      "loss": 0.8723,
+      "step": 17299
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.5378204664776165,
+      "learning_rate": 1.3409682940856714e-05,
+      "loss": 0.8349,
+      "step": 17300
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5074341059957064,
+      "learning_rate": 1.3408965615262008e-05,
+      "loss": 0.778,
+      "step": 17301
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8017023354142887,
+      "learning_rate": 1.3408248269819228e-05,
+      "loss": 0.6438,
+      "step": 17302
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.978251431860011,
+      "learning_rate": 1.3407530904532546e-05,
+      "loss": 0.5958,
+      "step": 17303
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8160709778845263,
+      "learning_rate": 1.3406813519406142e-05,
+      "loss": 0.8478,
+      "step": 17304
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5362249847749143,
+      "learning_rate": 1.3406096114444193e-05,
+      "loss": 0.8685,
+      "step": 17305
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7521461793562132,
+      "learning_rate": 1.340537868965087e-05,
+      "loss": 0.8539,
+      "step": 17306
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.559040555759222,
+      "learning_rate": 1.3404661245030357e-05,
+      "loss": 0.7822,
+      "step": 17307
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.541975793587526,
+      "learning_rate": 1.3403943780586826e-05,
+      "loss": 0.7311,
+      "step": 17308
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.425315424832317,
+      "learning_rate": 1.3403226296324458e-05,
+      "loss": 0.6009,
+      "step": 17309
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.630009087699835,
+      "learning_rate": 1.3402508792247427e-05,
+      "loss": 0.8446,
+      "step": 17310
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5889977401285672,
+      "learning_rate": 1.3401791268359916e-05,
+      "loss": 0.5157,
+      "step": 17311
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4612465712175629,
+      "learning_rate": 1.3401073724666098e-05,
+      "loss": 0.7423,
+      "step": 17312
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5940351300207017,
+      "learning_rate": 1.3400356161170152e-05,
+      "loss": 0.543,
+      "step": 17313
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5815748056169684,
+      "learning_rate": 1.3399638577876255e-05,
+      "loss": 0.5376,
+      "step": 17314
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5961098557319635,
+      "learning_rate": 1.3398920974788585e-05,
+      "loss": 0.8035,
+      "step": 17315
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5701073022090777,
+      "learning_rate": 1.3398203351911323e-05,
+      "loss": 0.7122,
+      "step": 17316
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8114743884388729,
+      "learning_rate": 1.3397485709248643e-05,
+      "loss": 0.8117,
+      "step": 17317
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.620495200661966,
+      "learning_rate": 1.3396768046804728e-05,
+      "loss": 0.6182,
+      "step": 17318
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6540965181974607,
+      "learning_rate": 1.3396050364583752e-05,
+      "loss": 0.7459,
+      "step": 17319
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5742531744951105,
+      "learning_rate": 1.3395332662589898e-05,
+      "loss": 0.8336,
+      "step": 17320
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6310320094851338,
+      "learning_rate": 1.339461494082734e-05,
+      "loss": 0.7233,
+      "step": 17321
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.323015440669712,
+      "learning_rate": 1.3393897199300262e-05,
+      "loss": 0.7712,
+      "step": 17322
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9860920833291802,
+      "learning_rate": 1.3393179438012836e-05,
+      "loss": 0.6827,
+      "step": 17323
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3312081299936884,
+      "learning_rate": 1.3392461656969246e-05,
+      "loss": 0.6948,
+      "step": 17324
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.055197828626444,
+      "learning_rate": 1.3391743856173673e-05,
+      "loss": 0.7552,
+      "step": 17325
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8732660098375655,
+      "learning_rate": 1.3391026035630294e-05,
+      "loss": 0.8151,
+      "step": 17326
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7316625868183322,
+      "learning_rate": 1.3390308195343282e-05,
+      "loss": 0.7536,
+      "step": 17327
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6644118161283008,
+      "learning_rate": 1.338959033531683e-05,
+      "loss": 0.6469,
+      "step": 17328
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0699656243646434,
+      "learning_rate": 1.3388872455555102e-05,
+      "loss": 0.7977,
+      "step": 17329
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3907801469865486,
+      "learning_rate": 1.3388154556062292e-05,
+      "loss": 0.6492,
+      "step": 17330
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7846683099804432,
+      "learning_rate": 1.338743663684257e-05,
+      "loss": 0.9579,
+      "step": 17331
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3884430024370555,
+      "learning_rate": 1.3386718697900123e-05,
+      "loss": 0.8637,
+      "step": 17332
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5248396839720597,
+      "learning_rate": 1.3386000739239125e-05,
+      "loss": 0.7038,
+      "step": 17333
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6100387441329106,
+      "learning_rate": 1.3385282760863758e-05,
+      "loss": 0.5071,
+      "step": 17334
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4980675319177,
+      "learning_rate": 1.3384564762778203e-05,
+      "loss": 0.685,
+      "step": 17335
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.60930204062175,
+      "learning_rate": 1.3383846744986642e-05,
+      "loss": 0.5493,
+      "step": 17336
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.61764557067844,
+      "learning_rate": 1.3383128707493253e-05,
+      "loss": 0.7701,
+      "step": 17337
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6036335129547927,
+      "learning_rate": 1.3382410650302218e-05,
+      "loss": 0.7684,
+      "step": 17338
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3782307614852105,
+      "learning_rate": 1.3381692573417718e-05,
+      "loss": 0.7712,
+      "step": 17339
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7062932277183087,
+      "learning_rate": 1.3380974476843932e-05,
+      "loss": 0.716,
+      "step": 17340
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5856469906791955,
+      "learning_rate": 1.338025636058504e-05,
+      "loss": 0.8097,
+      "step": 17341
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.4728073811385247,
+      "learning_rate": 1.3379538224645228e-05,
+      "loss": 0.9117,
+      "step": 17342
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5730822570090093,
+      "learning_rate": 1.3378820069028675e-05,
+      "loss": 0.5414,
+      "step": 17343
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3331220571106444,
+      "learning_rate": 1.3378101893739561e-05,
+      "loss": 0.6306,
+      "step": 17344
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5747823001556438,
+      "learning_rate": 1.3377383698782066e-05,
+      "loss": 0.7737,
+      "step": 17345
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.378746500213867,
+      "learning_rate": 1.3376665484160378e-05,
+      "loss": 0.592,
+      "step": 17346
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4837295647140956,
+      "learning_rate": 1.3375947249878675e-05,
+      "loss": 0.7503,
+      "step": 17347
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4567717840382064,
+      "learning_rate": 1.3375228995941135e-05,
+      "loss": 0.6242,
+      "step": 17348
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6204769626496816,
+      "learning_rate": 1.3374510722351945e-05,
+      "loss": 0.895,
+      "step": 17349
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4790690219366194,
+      "learning_rate": 1.3373792429115282e-05,
+      "loss": 0.7295,
+      "step": 17350
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.436736583152573,
+      "learning_rate": 1.3373074116235334e-05,
+      "loss": 0.638,
+      "step": 17351
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4594209793344852,
+      "learning_rate": 1.3372355783716282e-05,
+      "loss": 0.5952,
+      "step": 17352
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4027524287726134,
+      "learning_rate": 1.3371637431562303e-05,
+      "loss": 0.8104,
+      "step": 17353
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.5293641192033522,
+      "learning_rate": 1.337091905977759e-05,
+      "loss": 0.7232,
+      "step": 17354
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4490063935070223,
+      "learning_rate": 1.337020066836631e-05,
+      "loss": 0.7714,
+      "step": 17355
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5238125572285488,
+      "learning_rate": 1.3369482257332664e-05,
+      "loss": 0.624,
+      "step": 17356
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5059617404107781,
+      "learning_rate": 1.336876382668082e-05,
+      "loss": 0.784,
+      "step": 17357
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5846116190192794,
+      "learning_rate": 1.3368045376414968e-05,
+      "loss": 0.8205,
+      "step": 17358
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3259386535256212,
+      "learning_rate": 1.3367326906539286e-05,
+      "loss": 0.734,
+      "step": 17359
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.036976582324116,
+      "learning_rate": 1.3366608417057966e-05,
+      "loss": 0.6983,
+      "step": 17360
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3864640382898419,
+      "learning_rate": 1.3365889907975184e-05,
+      "loss": 0.7395,
+      "step": 17361
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.868984701916885,
+      "learning_rate": 1.3365171379295126e-05,
+      "loss": 0.7075,
+      "step": 17362
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6951186747909595,
+      "learning_rate": 1.3364452831021972e-05,
+      "loss": 0.7575,
+      "step": 17363
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4482521976897005,
+      "learning_rate": 1.3363734263159911e-05,
+      "loss": 0.7331,
+      "step": 17364
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5404073505250062,
+      "learning_rate": 1.3363015675713124e-05,
+      "loss": 0.8072,
+      "step": 17365
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7501868190576324,
+      "learning_rate": 1.3362297068685794e-05,
+      "loss": 0.7985,
+      "step": 17366
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3286801386411276,
+      "learning_rate": 1.3361578442082107e-05,
+      "loss": 0.8027,
+      "step": 17367
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4517566396913355,
+      "learning_rate": 1.3360859795906244e-05,
+      "loss": 0.655,
+      "step": 17368
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6083926985286866,
+      "learning_rate": 1.3360141130162394e-05,
+      "loss": 0.5769,
+      "step": 17369
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5494971970085472,
+      "learning_rate": 1.3359422444854738e-05,
+      "loss": 0.65,
+      "step": 17370
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4649319726881367,
+      "learning_rate": 1.3358703739987459e-05,
+      "loss": 0.7954,
+      "step": 17371
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3750640441014652,
+      "learning_rate": 1.3357985015564744e-05,
+      "loss": 0.8681,
+      "step": 17372
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3601450872842211,
+      "learning_rate": 1.335726627159078e-05,
+      "loss": 0.753,
+      "step": 17373
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7175872676029431,
+      "learning_rate": 1.3356547508069745e-05,
+      "loss": 0.9107,
+      "step": 17374
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3225409860197803,
+      "learning_rate": 1.335582872500583e-05,
+      "loss": 0.6247,
+      "step": 17375
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.092053963961904,
+      "learning_rate": 1.3355109922403218e-05,
+      "loss": 0.6689,
+      "step": 17376
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2397489717718395,
+      "learning_rate": 1.3354391100266096e-05,
+      "loss": 0.6794,
+      "step": 17377
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.611012509991988,
+      "learning_rate": 1.335367225859864e-05,
+      "loss": 0.7417,
+      "step": 17378
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5935708281544345,
+      "learning_rate": 1.3352953397405049e-05,
+      "loss": 0.7529,
+      "step": 17379
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.541107253189446,
+      "learning_rate": 1.3352234516689497e-05,
+      "loss": 0.7785,
+      "step": 17380
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3750124496309855,
+      "learning_rate": 1.3351515616456178e-05,
+      "loss": 0.714,
+      "step": 17381
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3542175599111634,
+      "learning_rate": 1.3350796696709272e-05,
+      "loss": 0.7939,
+      "step": 17382
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5087419459225904,
+      "learning_rate": 1.3350077757452969e-05,
+      "loss": 0.7645,
+      "step": 17383
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9174433609023398,
+      "learning_rate": 1.334935879869145e-05,
+      "loss": 0.6949,
+      "step": 17384
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2736755157046125,
+      "learning_rate": 1.3348639820428906e-05,
+      "loss": 0.6898,
+      "step": 17385
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.348296363535754,
+      "learning_rate": 1.3347920822669518e-05,
+      "loss": 0.669,
+      "step": 17386
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5082675686417695,
+      "learning_rate": 1.3347201805417476e-05,
+      "loss": 0.7741,
+      "step": 17387
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3558540504959806,
+      "learning_rate": 1.3346482768676968e-05,
+      "loss": 0.821,
+      "step": 17388
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4953367817868746,
+      "learning_rate": 1.3345763712452174e-05,
+      "loss": 0.7093,
+      "step": 17389
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2648767269866439,
+      "learning_rate": 1.3345044636747288e-05,
+      "loss": 0.7607,
+      "step": 17390
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5793325679259813,
+      "learning_rate": 1.3344325541566492e-05,
+      "loss": 0.7598,
+      "step": 17391
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3022497709625191,
+      "learning_rate": 1.334360642691397e-05,
+      "loss": 0.859,
+      "step": 17392
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7971613379751246,
+      "learning_rate": 1.3342887292793918e-05,
+      "loss": 0.7204,
+      "step": 17393
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.284813977378682,
+      "learning_rate": 1.3342168139210517e-05,
+      "loss": 0.7365,
+      "step": 17394
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4100751904614315,
+      "learning_rate": 1.3341448966167954e-05,
+      "loss": 0.655,
+      "step": 17395
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7410938753913079,
+      "learning_rate": 1.3340729773670416e-05,
+      "loss": 0.6699,
+      "step": 17396
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6245314499174995,
+      "learning_rate": 1.3340010561722093e-05,
+      "loss": 0.5748,
+      "step": 17397
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6206967931355045,
+      "learning_rate": 1.3339291330327172e-05,
+      "loss": 0.7093,
+      "step": 17398
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5876168146324192,
+      "learning_rate": 1.3338572079489835e-05,
+      "loss": 0.7868,
+      "step": 17399
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2873254053584249,
+      "learning_rate": 1.3337852809214278e-05,
+      "loss": 0.6957,
+      "step": 17400
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.785253979133884,
+      "learning_rate": 1.3337133519504685e-05,
+      "loss": 0.7373,
+      "step": 17401
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5303761505337554,
+      "learning_rate": 1.3336414210365245e-05,
+      "loss": 0.5926,
+      "step": 17402
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4974183741091824,
+      "learning_rate": 1.3335694881800145e-05,
+      "loss": 0.7709,
+      "step": 17403
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3014277503801621,
+      "learning_rate": 1.3334975533813572e-05,
+      "loss": 0.7885,
+      "step": 17404
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.448882074301712,
+      "learning_rate": 1.3334256166409717e-05,
+      "loss": 0.5354,
+      "step": 17405
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.383718392874425,
+      "learning_rate": 1.3333536779592766e-05,
+      "loss": 0.7018,
+      "step": 17406
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7722877272491147,
+      "learning_rate": 1.333281737336691e-05,
+      "loss": 0.7657,
+      "step": 17407
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6108541238744696,
+      "learning_rate": 1.3332097947736333e-05,
+      "loss": 0.5111,
+      "step": 17408
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5953428940905754,
+      "learning_rate": 1.333137850270523e-05,
+      "loss": 0.7635,
+      "step": 17409
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6623647933956762,
+      "learning_rate": 1.3330659038277785e-05,
+      "loss": 0.7572,
+      "step": 17410
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.598359365418742,
+      "learning_rate": 1.3329939554458188e-05,
+      "loss": 0.4981,
+      "step": 17411
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4252350997146335,
+      "learning_rate": 1.332922005125063e-05,
+      "loss": 0.6645,
+      "step": 17412
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5518047571996032,
+      "learning_rate": 1.3328500528659301e-05,
+      "loss": 0.6255,
+      "step": 17413
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5574540960197456,
+      "learning_rate": 1.3327780986688384e-05,
+      "loss": 0.676,
+      "step": 17414
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4465904259529867,
+      "learning_rate": 1.3327061425342073e-05,
+      "loss": 0.7332,
+      "step": 17415
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.096864856612223,
+      "learning_rate": 1.332634184462456e-05,
+      "loss": 0.7028,
+      "step": 17416
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5925166785774503,
+      "learning_rate": 1.3325622244540028e-05,
+      "loss": 0.7444,
+      "step": 17417
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.2569453298892586,
+      "learning_rate": 1.3324902625092672e-05,
+      "loss": 0.7142,
+      "step": 17418
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.929308600091344,
+      "learning_rate": 1.3324182986286683e-05,
+      "loss": 0.6788,
+      "step": 17419
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3573322883852623,
+      "learning_rate": 1.3323463328126245e-05,
+      "loss": 0.6825,
+      "step": 17420
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5207150782438692,
+      "learning_rate": 1.3322743650615552e-05,
+      "loss": 0.6172,
+      "step": 17421
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.877616578230605,
+      "learning_rate": 1.3322023953758794e-05,
+      "loss": 0.8228,
+      "step": 17422
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3162285995569363,
+      "learning_rate": 1.332130423756016e-05,
+      "loss": 0.8362,
+      "step": 17423
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9119037324759873,
+      "learning_rate": 1.3320584502023841e-05,
+      "loss": 0.8231,
+      "step": 17424
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3981764686883766,
+      "learning_rate": 1.3319864747154028e-05,
+      "loss": 0.6917,
+      "step": 17425
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4560671799363363,
+      "learning_rate": 1.3319144972954915e-05,
+      "loss": 0.7173,
+      "step": 17426
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.483276068060547,
+      "learning_rate": 1.3318425179430683e-05,
+      "loss": 0.6875,
+      "step": 17427
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5620397333991658,
+      "learning_rate": 1.3317705366585534e-05,
+      "loss": 0.7226,
+      "step": 17428
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6041313969337878,
+      "learning_rate": 1.331698553442365e-05,
+      "loss": 0.6053,
+      "step": 17429
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6567244687075169,
+      "learning_rate": 1.331626568294923e-05,
+      "loss": 0.5845,
+      "step": 17430
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7029222256411336,
+      "learning_rate": 1.3315545812166457e-05,
+      "loss": 0.7225,
+      "step": 17431
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4547427584214372,
+      "learning_rate": 1.3314825922079531e-05,
+      "loss": 0.8126,
+      "step": 17432
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.563944335747065,
+      "learning_rate": 1.3314106012692633e-05,
+      "loss": 0.7784,
+      "step": 17433
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6034677119598622,
+      "learning_rate": 1.331338608400997e-05,
+      "loss": 0.6803,
+      "step": 17434
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.381512090036679,
+      "learning_rate": 1.3312666136035716e-05,
+      "loss": 0.7579,
+      "step": 17435
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5986783718917246,
+      "learning_rate": 1.3311946168774072e-05,
+      "loss": 0.7263,
+      "step": 17436
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5297435687269192,
+      "learning_rate": 1.3311226182229233e-05,
+      "loss": 0.8329,
+      "step": 17437
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.667730608352258,
+      "learning_rate": 1.3310506176405383e-05,
+      "loss": 0.8717,
+      "step": 17438
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4944256240942508,
+      "learning_rate": 1.3309786151306721e-05,
+      "loss": 0.8514,
+      "step": 17439
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7520313899397852,
+      "learning_rate": 1.3309066106937432e-05,
+      "loss": 0.7857,
+      "step": 17440
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5270382832068286,
+      "learning_rate": 1.3308346043301717e-05,
+      "loss": 0.7407,
+      "step": 17441
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4863399993374227,
+      "learning_rate": 1.3307625960403763e-05,
+      "loss": 0.6158,
+      "step": 17442
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.265544682537873,
+      "learning_rate": 1.330690585824776e-05,
+      "loss": 0.7,
+      "step": 17443
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7981736088920768,
+      "learning_rate": 1.3306185736837908e-05,
+      "loss": 0.972,
+      "step": 17444
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5638670700207071,
+      "learning_rate": 1.3305465596178395e-05,
+      "loss": 0.6944,
+      "step": 17445
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2819675756579916,
+      "learning_rate": 1.3304745436273415e-05,
+      "loss": 0.712,
+      "step": 17446
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3674511066334034,
+      "learning_rate": 1.330402525712716e-05,
+      "loss": 0.7,
+      "step": 17447
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.504633319672061,
+      "learning_rate": 1.3303305058743823e-05,
+      "loss": 0.7785,
+      "step": 17448
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5548947082912499,
+      "learning_rate": 1.33025848411276e-05,
+      "loss": 0.7705,
+      "step": 17449
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6143151665283253,
+      "learning_rate": 1.330186460428268e-05,
+      "loss": 0.7109,
+      "step": 17450
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6345901687154086,
+      "learning_rate": 1.3301144348213262e-05,
+      "loss": 0.5295,
+      "step": 17451
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.591300930483627,
+      "learning_rate": 1.3300424072923534e-05,
+      "loss": 0.757,
+      "step": 17452
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.526991675939793,
+      "learning_rate": 1.3299703778417694e-05,
+      "loss": 0.6947,
+      "step": 17453
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4115576548598567,
+      "learning_rate": 1.3298983464699933e-05,
+      "loss": 0.7473,
+      "step": 17454
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4701764809795557,
+      "learning_rate": 1.3298263131774445e-05,
+      "loss": 0.7047,
+      "step": 17455
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4683337509505225,
+      "learning_rate": 1.3297542779645428e-05,
+      "loss": 0.7228,
+      "step": 17456
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5514183833505009,
+      "learning_rate": 1.3296822408317067e-05,
+      "loss": 0.7187,
+      "step": 17457
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3356728592291118,
+      "learning_rate": 1.3296102017793569e-05,
+      "loss": 0.747,
+      "step": 17458
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5251347438749805,
+      "learning_rate": 1.3295381608079116e-05,
+      "loss": 0.7286,
+      "step": 17459
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4149651095392994,
+      "learning_rate": 1.3294661179177912e-05,
+      "loss": 0.7062,
+      "step": 17460
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4802336363888382,
+      "learning_rate": 1.3293940731094143e-05,
+      "loss": 0.613,
+      "step": 17461
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.130096621625328,
+      "learning_rate": 1.3293220263832015e-05,
+      "loss": 0.8343,
+      "step": 17462
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.342345604644697,
+      "learning_rate": 1.3292499777395711e-05,
+      "loss": 0.6895,
+      "step": 17463
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5840731046243562,
+      "learning_rate": 1.3291779271789433e-05,
+      "loss": 0.5319,
+      "step": 17464
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.228209431993343,
+      "learning_rate": 1.3291058747017373e-05,
+      "loss": 0.6993,
+      "step": 17465
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3641383483294314,
+      "learning_rate": 1.3290338203083728e-05,
+      "loss": 0.6837,
+      "step": 17466
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5392566199819755,
+      "learning_rate": 1.328961763999269e-05,
+      "loss": 0.7602,
+      "step": 17467
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.523974161472889,
+      "learning_rate": 1.3288897057748458e-05,
+      "loss": 0.7246,
+      "step": 17468
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0963894476208424,
+      "learning_rate": 1.3288176456355226e-05,
+      "loss": 0.8606,
+      "step": 17469
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0048248320635897,
+      "learning_rate": 1.3287455835817189e-05,
+      "loss": 0.744,
+      "step": 17470
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.184306909213664,
+      "learning_rate": 1.3286735196138544e-05,
+      "loss": 0.6669,
+      "step": 17471
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9436760503240371,
+      "learning_rate": 1.3286014537323486e-05,
+      "loss": 0.6468,
+      "step": 17472
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8191769258509527,
+      "learning_rate": 1.3285293859376212e-05,
+      "loss": 0.7015,
+      "step": 17473
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3184423602441937,
+      "learning_rate": 1.3284573162300916e-05,
+      "loss": 0.8349,
+      "step": 17474
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.445533723022606,
+      "learning_rate": 1.3283852446101794e-05,
+      "loss": 0.7292,
+      "step": 17475
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3438045719039984,
+      "learning_rate": 1.3283131710783046e-05,
+      "loss": 0.8417,
+      "step": 17476
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5541601792772142,
+      "learning_rate": 1.3282410956348865e-05,
+      "loss": 0.7263,
+      "step": 17477
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4496432111179405,
+      "learning_rate": 1.3281690182803448e-05,
+      "loss": 0.7127,
+      "step": 17478
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.1444915547102394,
+      "learning_rate": 1.3280969390150993e-05,
+      "loss": 0.848,
+      "step": 17479
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8319254822776778,
+      "learning_rate": 1.328024857839569e-05,
+      "loss": 0.6741,
+      "step": 17480
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2654515585941786,
+      "learning_rate": 1.327952774754175e-05,
+      "loss": 0.6815,
+      "step": 17481
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6948020740459164,
+      "learning_rate": 1.3278806897593352e-05,
+      "loss": 0.7103,
+      "step": 17482
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.892574280850199,
+      "learning_rate": 1.3278086028554711e-05,
+      "loss": 0.8118,
+      "step": 17483
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4611095622417531,
+      "learning_rate": 1.327736514043001e-05,
+      "loss": 0.8827,
+      "step": 17484
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9260875732529548,
+      "learning_rate": 1.327664423322345e-05,
+      "loss": 0.7317,
+      "step": 17485
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3704331920695272,
+      "learning_rate": 1.3275923306939235e-05,
+      "loss": 0.7997,
+      "step": 17486
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3748772222744363,
+      "learning_rate": 1.3275202361581552e-05,
+      "loss": 0.597,
+      "step": 17487
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.1400001967494138,
+      "learning_rate": 1.327448139715461e-05,
+      "loss": 0.7238,
+      "step": 17488
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3787219605754673,
+      "learning_rate": 1.3273760413662596e-05,
+      "loss": 0.6944,
+      "step": 17489
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3983483740223839,
+      "learning_rate": 1.3273039411109714e-05,
+      "loss": 0.681,
+      "step": 17490
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.288750662347195,
+      "learning_rate": 1.3272318389500158e-05,
+      "loss": 0.7158,
+      "step": 17491
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5370061101574204,
+      "learning_rate": 1.3271597348838133e-05,
+      "loss": 0.7952,
+      "step": 17492
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3362655148074394,
+      "learning_rate": 1.327087628912783e-05,
+      "loss": 0.8617,
+      "step": 17493
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3698045166671318,
+      "learning_rate": 1.327015521037345e-05,
+      "loss": 0.7033,
+      "step": 17494
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4750115664494625,
+      "learning_rate": 1.326943411257919e-05,
+      "loss": 0.687,
+      "step": 17495
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.308557423289341,
+      "learning_rate": 1.3268712995749251e-05,
+      "loss": 0.7605,
+      "step": 17496
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3908757363859052,
+      "learning_rate": 1.3267991859887827e-05,
+      "loss": 0.6796,
+      "step": 17497
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5771230931615083,
+      "learning_rate": 1.3267270704999122e-05,
+      "loss": 0.7826,
+      "step": 17498
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3729583417610856,
+      "learning_rate": 1.3266549531087333e-05,
+      "loss": 0.7175,
+      "step": 17499
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6649031499727338,
+      "learning_rate": 1.3265828338156658e-05,
+      "loss": 0.7599,
+      "step": 17500
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4490201756612668,
+      "learning_rate": 1.3265107126211296e-05,
+      "loss": 0.776,
+      "step": 17501
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5575683305534753,
+      "learning_rate": 1.326438589525545e-05,
+      "loss": 0.7318,
+      "step": 17502
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6881994943707253,
+      "learning_rate": 1.326366464529331e-05,
+      "loss": 0.581,
+      "step": 17503
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.594968462553568,
+      "learning_rate": 1.3262943376329085e-05,
+      "loss": 0.4986,
+      "step": 17504
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4195261659918894,
+      "learning_rate": 1.326222208836697e-05,
+      "loss": 0.6899,
+      "step": 17505
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.538941786802445,
+      "learning_rate": 1.3261500781411165e-05,
+      "loss": 0.7005,
+      "step": 17506
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4454309623670174,
+      "learning_rate": 1.3260779455465871e-05,
+      "loss": 0.6702,
+      "step": 17507
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3078859810724714,
+      "learning_rate": 1.3260058110535283e-05,
+      "loss": 0.6734,
+      "step": 17508
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6650331009694409,
+      "learning_rate": 1.325933674662361e-05,
+      "loss": 0.5607,
+      "step": 17509
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5978374323998543,
+      "learning_rate": 1.3258615363735041e-05,
+      "loss": 0.6763,
+      "step": 17510
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8399154532514461,
+      "learning_rate": 1.3257893961873787e-05,
+      "loss": 0.7007,
+      "step": 17511
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7450194060759052,
+      "learning_rate": 1.3257172541044039e-05,
+      "loss": 0.6419,
+      "step": 17512
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5763311503576467,
+      "learning_rate": 1.3256451101250002e-05,
+      "loss": 0.61,
+      "step": 17513
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.490533064748877,
+      "learning_rate": 1.3255729642495878e-05,
+      "loss": 0.7942,
+      "step": 17514
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.238419459417811,
+      "learning_rate": 1.3255008164785863e-05,
+      "loss": 0.778,
+      "step": 17515
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5958328997356646,
+      "learning_rate": 1.325428666812416e-05,
+      "loss": 0.6943,
+      "step": 17516
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6814629817388453,
+      "learning_rate": 1.3253565152514971e-05,
+      "loss": 0.783,
+      "step": 17517
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4809919049215974,
+      "learning_rate": 1.3252843617962494e-05,
+      "loss": 0.8576,
+      "step": 17518
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.460039452248464,
+      "learning_rate": 1.3252122064470934e-05,
+      "loss": 0.7701,
+      "step": 17519
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8617093919059302,
+      "learning_rate": 1.3251400492044487e-05,
+      "loss": 0.6581,
+      "step": 17520
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5920543311244773,
+      "learning_rate": 1.3250678900687356e-05,
+      "loss": 0.7513,
+      "step": 17521
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8004999422423766,
+      "learning_rate": 1.3249957290403748e-05,
+      "loss": 0.6864,
+      "step": 17522
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.383558778386729,
+      "learning_rate": 1.3249235661197857e-05,
+      "loss": 0.7987,
+      "step": 17523
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.649052564792278,
+      "learning_rate": 1.3248514013073886e-05,
+      "loss": 0.7741,
+      "step": 17524
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4334373040184007,
+      "learning_rate": 1.3247792346036039e-05,
+      "loss": 0.7688,
+      "step": 17525
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5306972481815182,
+      "learning_rate": 1.3247070660088519e-05,
+      "loss": 0.7505,
+      "step": 17526
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4546347602884675,
+      "learning_rate": 1.3246348955235521e-05,
+      "loss": 0.7114,
+      "step": 17527
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.229997216559186,
+      "learning_rate": 1.3245627231481258e-05,
+      "loss": 0.648,
+      "step": 17528
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.713103359992121,
+      "learning_rate": 1.324490548882992e-05,
+      "loss": 0.642,
+      "step": 17529
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4011808892417363,
+      "learning_rate": 1.3244183727285717e-05,
+      "loss": 0.6925,
+      "step": 17530
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3719878864941473,
+      "learning_rate": 1.3243461946852849e-05,
+      "loss": 0.7979,
+      "step": 17531
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.44792540957276,
+      "learning_rate": 1.324274014753552e-05,
+      "loss": 0.7387,
+      "step": 17532
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.256230936733927,
+      "learning_rate": 1.3242018329337928e-05,
+      "loss": 0.6011,
+      "step": 17533
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.256114263533971,
+      "learning_rate": 1.3241296492264282e-05,
+      "loss": 0.7216,
+      "step": 17534
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5991565855644857,
+      "learning_rate": 1.324057463631878e-05,
+      "loss": 0.5338,
+      "step": 17535
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3531983606943514,
+      "learning_rate": 1.3239852761505627e-05,
+      "loss": 0.79,
+      "step": 17536
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4508223014815602,
+      "learning_rate": 1.3239130867829028e-05,
+      "loss": 0.7709,
+      "step": 17537
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.227088121031842,
+      "learning_rate": 1.3238408955293179e-05,
+      "loss": 0.7761,
+      "step": 17538
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5120781799183451,
+      "learning_rate": 1.3237687023902292e-05,
+      "loss": 0.6911,
+      "step": 17539
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5502803279706783,
+      "learning_rate": 1.3236965073660562e-05,
+      "loss": 0.7965,
+      "step": 17540
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3547089013323954,
+      "learning_rate": 1.3236243104572202e-05,
+      "loss": 0.8024,
+      "step": 17541
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3327732632530884,
+      "learning_rate": 1.3235521116641405e-05,
+      "loss": 0.6605,
+      "step": 17542
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5467285916461813,
+      "learning_rate": 1.3234799109872382e-05,
+      "loss": 0.7758,
+      "step": 17543
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5977601346128478,
+      "learning_rate": 1.3234077084269334e-05,
+      "loss": 0.7558,
+      "step": 17544
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6869708068592442,
+      "learning_rate": 1.3233355039836466e-05,
+      "loss": 0.7322,
+      "step": 17545
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3990382687295644,
+      "learning_rate": 1.323263297657798e-05,
+      "loss": 0.8665,
+      "step": 17546
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4369739222854103,
+      "learning_rate": 1.3231910894498082e-05,
+      "loss": 0.6588,
+      "step": 17547
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3719272505427673,
+      "learning_rate": 1.3231188793600976e-05,
+      "loss": 0.822,
+      "step": 17548
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5988940492571837,
+      "learning_rate": 1.3230466673890865e-05,
+      "loss": 0.5388,
+      "step": 17549
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8282366611158662,
+      "learning_rate": 1.3229744535371953e-05,
+      "loss": 0.7283,
+      "step": 17550
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.401839747570592,
+      "learning_rate": 1.3229022378048448e-05,
+      "loss": 0.7482,
+      "step": 17551
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4671102660969133,
+      "learning_rate": 1.322830020192455e-05,
+      "loss": 0.8826,
+      "step": 17552
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4194471317517647,
+      "learning_rate": 1.3227578007004471e-05,
+      "loss": 0.7073,
+      "step": 17553
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4727850007009626,
+      "learning_rate": 1.3226855793292406e-05,
+      "loss": 0.6276,
+      "step": 17554
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6629971209876968,
+      "learning_rate": 1.322613356079257e-05,
+      "loss": 0.7788,
+      "step": 17555
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3927734792621178,
+      "learning_rate": 1.3225411309509158e-05,
+      "loss": 0.7035,
+      "step": 17556
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.804810598277656,
+      "learning_rate": 1.322468903944638e-05,
+      "loss": 0.8969,
+      "step": 17557
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0831401744432383,
+      "learning_rate": 1.3223966750608447e-05,
+      "loss": 0.6772,
+      "step": 17558
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4502677799882568,
+      "learning_rate": 1.3223244442999554e-05,
+      "loss": 0.6551,
+      "step": 17559
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4942692927009293,
+      "learning_rate": 1.3222522116623915e-05,
+      "loss": 0.6817,
+      "step": 17560
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2925065671694678,
+      "learning_rate": 1.3221799771485728e-05,
+      "loss": 0.6427,
+      "step": 17561
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.430680967406413,
+      "learning_rate": 1.3221077407589207e-05,
+      "loss": 0.7381,
+      "step": 17562
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4823842424054088,
+      "learning_rate": 1.3220355024938549e-05,
+      "loss": 0.7692,
+      "step": 17563
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3970631351498901,
+      "learning_rate": 1.3219632623537968e-05,
+      "loss": 0.7693,
+      "step": 17564
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5448886339382786,
+      "learning_rate": 1.3218910203391666e-05,
+      "loss": 0.7239,
+      "step": 17565
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.510069673898658,
+      "learning_rate": 1.321818776450385e-05,
+      "loss": 0.8469,
+      "step": 17566
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.329435343682216,
+      "learning_rate": 1.3217465306878725e-05,
+      "loss": 0.6997,
+      "step": 17567
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6028964035035705,
+      "learning_rate": 1.3216742830520497e-05,
+      "loss": 0.5254,
+      "step": 17568
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3640658302228243,
+      "learning_rate": 1.3216020335433375e-05,
+      "loss": 0.604,
+      "step": 17569
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4756777256797908,
+      "learning_rate": 1.3215297821621565e-05,
+      "loss": 0.7267,
+      "step": 17570
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2500449540391982,
+      "learning_rate": 1.3214575289089272e-05,
+      "loss": 0.8058,
+      "step": 17571
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.588621861981486,
+      "learning_rate": 1.3213852737840705e-05,
+      "loss": 0.8149,
+      "step": 17572
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4548629230674504,
+      "learning_rate": 1.3213130167880071e-05,
+      "loss": 0.7648,
+      "step": 17573
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6141389753311163,
+      "learning_rate": 1.3212407579211575e-05,
+      "loss": 0.5515,
+      "step": 17574
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.018443771296303,
+      "learning_rate": 1.3211684971839422e-05,
+      "loss": 0.7085,
+      "step": 17575
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3551815897512045,
+      "learning_rate": 1.3210962345767826e-05,
+      "loss": 0.7823,
+      "step": 17576
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5197756878726165,
+      "learning_rate": 1.3210239701000989e-05,
+      "loss": 0.7463,
+      "step": 17577
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.298407416465532,
+      "learning_rate": 1.320951703754312e-05,
+      "loss": 0.7526,
+      "step": 17578
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5695580023518237,
+      "learning_rate": 1.3208794355398431e-05,
+      "loss": 0.5055,
+      "step": 17579
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.575904129826271,
+      "learning_rate": 1.320807165457112e-05,
+      "loss": 0.8925,
+      "step": 17580
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.228578762783878,
+      "learning_rate": 1.3207348935065403e-05,
+      "loss": 0.737,
+      "step": 17581
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6497918651218005,
+      "learning_rate": 1.3206626196885481e-05,
+      "loss": 0.576,
+      "step": 17582
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.82217623583638,
+      "learning_rate": 1.320590344003557e-05,
+      "loss": 0.811,
+      "step": 17583
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4490671988630293,
+      "learning_rate": 1.3205180664519872e-05,
+      "loss": 0.7359,
+      "step": 17584
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6108169214151423,
+      "learning_rate": 1.3204457870342599e-05,
+      "loss": 0.5489,
+      "step": 17585
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.2753540053575056,
+      "learning_rate": 1.320373505750796e-05,
+      "loss": 0.7003,
+      "step": 17586
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.665991747106869,
+      "learning_rate": 1.3203012226020156e-05,
+      "loss": 0.7625,
+      "step": 17587
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4964069163658391,
+      "learning_rate": 1.3202289375883405e-05,
+      "loss": 0.8099,
+      "step": 17588
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9790019554138687,
+      "learning_rate": 1.3201566507101908e-05,
+      "loss": 0.7375,
+      "step": 17589
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.757775501583728,
+      "learning_rate": 1.3200843619679882e-05,
+      "loss": 0.8422,
+      "step": 17590
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5690139144719186,
+      "learning_rate": 1.3200120713621525e-05,
+      "loss": 0.5433,
+      "step": 17591
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5305379502085676,
+      "learning_rate": 1.3199397788931057e-05,
+      "loss": 0.6511,
+      "step": 17592
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.106951105003864,
+      "learning_rate": 1.319867484561268e-05,
+      "loss": 0.8656,
+      "step": 17593
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3883895934290031,
+      "learning_rate": 1.3197951883670606e-05,
+      "loss": 0.7809,
+      "step": 17594
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4186306908371829,
+      "learning_rate": 1.3197228903109044e-05,
+      "loss": 0.7728,
+      "step": 17595
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3717289548457925,
+      "learning_rate": 1.3196505903932203e-05,
+      "loss": 0.7072,
+      "step": 17596
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.05491521535477,
+      "learning_rate": 1.3195782886144291e-05,
+      "loss": 0.7117,
+      "step": 17597
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3275296955242426,
+      "learning_rate": 1.3195059849749519e-05,
+      "loss": 0.8605,
+      "step": 17598
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6280841393413219,
+      "learning_rate": 1.3194336794752099e-05,
+      "loss": 0.5681,
+      "step": 17599
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.412373434305888,
+      "learning_rate": 1.3193613721156237e-05,
+      "loss": 0.6508,
+      "step": 17600
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6532336970600903,
+      "learning_rate": 1.3192890628966146e-05,
+      "loss": 0.7878,
+      "step": 17601
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4737856213670564,
+      "learning_rate": 1.3192167518186036e-05,
+      "loss": 0.8065,
+      "step": 17602
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6111678449882634,
+      "learning_rate": 1.3191444388820114e-05,
+      "loss": 0.6626,
+      "step": 17603
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3439907671772318,
+      "learning_rate": 1.3190721240872595e-05,
+      "loss": 0.8443,
+      "step": 17604
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.55863031331408,
+      "learning_rate": 1.3189998074347684e-05,
+      "loss": 0.7036,
+      "step": 17605
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.440283766948416,
+      "learning_rate": 1.3189274889249598e-05,
+      "loss": 0.7747,
+      "step": 17606
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6399116761556973,
+      "learning_rate": 1.3188551685582542e-05,
+      "loss": 0.7208,
+      "step": 17607
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5580905024393394,
+      "learning_rate": 1.3187828463350725e-05,
+      "loss": 0.703,
+      "step": 17608
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.585868887882613,
+      "learning_rate": 1.3187105222558368e-05,
+      "loss": 0.569,
+      "step": 17609
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7021537425188467,
+      "learning_rate": 1.3186381963209666e-05,
+      "loss": 0.7179,
+      "step": 17610
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4723616992698478,
+      "learning_rate": 1.3185658685308848e-05,
+      "loss": 0.7562,
+      "step": 17611
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5851544894365959,
+      "learning_rate": 1.3184935388860111e-05,
+      "loss": 0.5581,
+      "step": 17612
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3641515814418468,
+      "learning_rate": 1.3184212073867678e-05,
+      "loss": 0.6741,
+      "step": 17613
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2116513602603656,
+      "learning_rate": 1.3183488740335748e-05,
+      "loss": 0.7272,
+      "step": 17614
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4913064050030753,
+      "learning_rate": 1.3182765388268544e-05,
+      "loss": 0.7548,
+      "step": 17615
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.610370355854117,
+      "learning_rate": 1.318204201767027e-05,
+      "loss": 0.8173,
+      "step": 17616
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8646071795063326,
+      "learning_rate": 1.3181318628545138e-05,
+      "loss": 0.6977,
+      "step": 17617
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5621484486995059,
+      "learning_rate": 1.3180595220897365e-05,
+      "loss": 0.7541,
+      "step": 17618
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.3964176167830082,
+      "learning_rate": 1.3179871794731158e-05,
+      "loss": 0.7292,
+      "step": 17619
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.480940177090834,
+      "learning_rate": 1.3179148350050733e-05,
+      "loss": 0.8255,
+      "step": 17620
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4904574872215914,
+      "learning_rate": 1.3178424886860292e-05,
+      "loss": 0.6691,
+      "step": 17621
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7614053680368615,
+      "learning_rate": 1.3177701405164065e-05,
+      "loss": 0.6717,
+      "step": 17622
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4905046805749378,
+      "learning_rate": 1.3176977904966248e-05,
+      "loss": 0.607,
+      "step": 17623
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6011983426654457,
+      "learning_rate": 1.3176254386271063e-05,
+      "loss": 0.8335,
+      "step": 17624
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.581171834451053,
+      "learning_rate": 1.3175530849082718e-05,
+      "loss": 0.7688,
+      "step": 17625
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6494392841149423,
+      "learning_rate": 1.3174807293405427e-05,
+      "loss": 0.7073,
+      "step": 17626
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.543476914164166,
+      "learning_rate": 1.3174083719243402e-05,
+      "loss": 0.7133,
+      "step": 17627
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5638684611251193,
+      "learning_rate": 1.3173360126600858e-05,
+      "loss": 0.5504,
+      "step": 17628
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4301540933993382,
+      "learning_rate": 1.3172636515482007e-05,
+      "loss": 0.7876,
+      "step": 17629
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4720850288931082,
+      "learning_rate": 1.3171912885891063e-05,
+      "loss": 0.6889,
+      "step": 17630
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5233335830267252,
+      "learning_rate": 1.3171189237832234e-05,
+      "loss": 0.7724,
+      "step": 17631
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.428747626519302,
+      "learning_rate": 1.3170465571309738e-05,
+      "loss": 0.8587,
+      "step": 17632
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4251200154306534,
+      "learning_rate": 1.316974188632779e-05,
+      "loss": 0.6754,
+      "step": 17633
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4960405968083408,
+      "learning_rate": 1.31690181828906e-05,
+      "loss": 0.7706,
+      "step": 17634
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4224804466641428,
+      "learning_rate": 1.3168294461002385e-05,
+      "loss": 0.7171,
+      "step": 17635
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7506717584149627,
+      "learning_rate": 1.3167570720667353e-05,
+      "loss": 0.5499,
+      "step": 17636
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.470005658195349,
+      "learning_rate": 1.3166846961889725e-05,
+      "loss": 0.7777,
+      "step": 17637
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.098314050131732,
+      "learning_rate": 1.316612318467371e-05,
+      "loss": 0.7771,
+      "step": 17638
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9003977939075622,
+      "learning_rate": 1.3165399389023523e-05,
+      "loss": 0.8004,
+      "step": 17639
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.540700423100707,
+      "learning_rate": 1.3164675574943377e-05,
+      "loss": 0.7851,
+      "step": 17640
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6123242347546388,
+      "learning_rate": 1.3163951742437493e-05,
+      "loss": 0.5533,
+      "step": 17641
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5681460249710253,
+      "learning_rate": 1.3163227891510073e-05,
+      "loss": 0.6323,
+      "step": 17642
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.531487106415104,
+      "learning_rate": 1.3162504022165345e-05,
+      "loss": 0.7671,
+      "step": 17643
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.414242918830115,
+      "learning_rate": 1.3161780134407515e-05,
+      "loss": 0.8452,
+      "step": 17644
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5326065189712894,
+      "learning_rate": 1.3161056228240798e-05,
+      "loss": 0.6384,
+      "step": 17645
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5285470008120354,
+      "learning_rate": 1.3160332303669414e-05,
+      "loss": 0.7579,
+      "step": 17646
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.442219182092483,
+      "learning_rate": 1.3159608360697573e-05,
+      "loss": 0.7748,
+      "step": 17647
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4231721019582257,
+      "learning_rate": 1.3158884399329492e-05,
+      "loss": 0.7076,
+      "step": 17648
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2531494506815517,
+      "learning_rate": 1.3158160419569388e-05,
+      "loss": 0.6772,
+      "step": 17649
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.34567267058563,
+      "learning_rate": 1.3157436421421472e-05,
+      "loss": 0.7879,
+      "step": 17650
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9102703879305833,
+      "learning_rate": 1.3156712404889962e-05,
+      "loss": 0.812,
+      "step": 17651
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4345629958590393,
+      "learning_rate": 1.3155988369979071e-05,
+      "loss": 0.7356,
+      "step": 17652
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5358105541689557,
+      "learning_rate": 1.315526431669302e-05,
+      "loss": 0.7551,
+      "step": 17653
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.437403340469385,
+      "learning_rate": 1.315454024503602e-05,
+      "loss": 0.7761,
+      "step": 17654
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6171854327521946,
+      "learning_rate": 1.3153816155012287e-05,
+      "loss": 0.6353,
+      "step": 17655
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2555082037455618,
+      "learning_rate": 1.3153092046626039e-05,
+      "loss": 0.6614,
+      "step": 17656
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3121966472285522,
+      "learning_rate": 1.315236791988149e-05,
+      "loss": 0.6863,
+      "step": 17657
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8377654596934232,
+      "learning_rate": 1.315164377478286e-05,
+      "loss": 0.7361,
+      "step": 17658
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5819559790408118,
+      "learning_rate": 1.3150919611334358e-05,
+      "loss": 0.7885,
+      "step": 17659
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5589306134176104,
+      "learning_rate": 1.3150195429540207e-05,
+      "loss": 0.6984,
+      "step": 17660
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.062685901627534,
+      "learning_rate": 1.3149471229404617e-05,
+      "loss": 0.7675,
+      "step": 17661
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5400896813284506,
+      "learning_rate": 1.3148747010931814e-05,
+      "loss": 0.5718,
+      "step": 17662
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4701720520302966,
+      "learning_rate": 1.3148022774126003e-05,
+      "loss": 0.7183,
+      "step": 17663
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4218214950128765,
+      "learning_rate": 1.3147298518991413e-05,
+      "loss": 0.6864,
+      "step": 17664
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3984382302503604,
+      "learning_rate": 1.3146574245532254e-05,
+      "loss": 0.7639,
+      "step": 17665
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5990243441674208,
+      "learning_rate": 1.3145849953752739e-05,
+      "loss": 0.6988,
+      "step": 17666
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4202917569436375,
+      "learning_rate": 1.3145125643657093e-05,
+      "loss": 0.7191,
+      "step": 17667
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3657851712265348,
+      "learning_rate": 1.3144401315249529e-05,
+      "loss": 0.78,
+      "step": 17668
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5918386283388835,
+      "learning_rate": 1.3143676968534264e-05,
+      "loss": 0.531,
+      "step": 17669
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.592062235284803,
+      "learning_rate": 1.3142952603515515e-05,
+      "loss": 0.5126,
+      "step": 17670
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.851781996616129,
+      "learning_rate": 1.3142228220197503e-05,
+      "loss": 0.7537,
+      "step": 17671
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6044632127512264,
+      "learning_rate": 1.3141503818584444e-05,
+      "loss": 0.6861,
+      "step": 17672
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7445531835616068,
+      "learning_rate": 1.3140779398680557e-05,
+      "loss": 0.7574,
+      "step": 17673
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4571687312548356,
+      "learning_rate": 1.3140054960490053e-05,
+      "loss": 0.8079,
+      "step": 17674
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4424823271833729,
+      "learning_rate": 1.3139330504017159e-05,
+      "loss": 0.6902,
+      "step": 17675
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5923785630454057,
+      "learning_rate": 1.3138606029266087e-05,
+      "loss": 0.7171,
+      "step": 17676
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5885702349141627,
+      "learning_rate": 1.3137881536241058e-05,
+      "loss": 0.4877,
+      "step": 17677
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.91841334411301,
+      "learning_rate": 1.3137157024946288e-05,
+      "loss": 0.7514,
+      "step": 17678
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3511426184106687,
+      "learning_rate": 1.3136432495385997e-05,
+      "loss": 0.6639,
+      "step": 17679
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5345510351440084,
+      "learning_rate": 1.3135707947564402e-05,
+      "loss": 0.7572,
+      "step": 17680
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3434469990946432,
+      "learning_rate": 1.3134983381485724e-05,
+      "loss": 0.7645,
+      "step": 17681
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4872254116597488,
+      "learning_rate": 1.3134258797154183e-05,
+      "loss": 0.7773,
+      "step": 17682
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5767777497337354,
+      "learning_rate": 1.3133534194573992e-05,
+      "loss": 0.7834,
+      "step": 17683
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4226498436592891,
+      "learning_rate": 1.3132809573749371e-05,
+      "loss": 0.7783,
+      "step": 17684
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.693995201603151,
+      "learning_rate": 1.3132084934684544e-05,
+      "loss": 0.6478,
+      "step": 17685
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7005085549927057,
+      "learning_rate": 1.3131360277383725e-05,
+      "loss": 0.7933,
+      "step": 17686
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6223744178414825,
+      "learning_rate": 1.3130635601851136e-05,
+      "loss": 0.7591,
+      "step": 17687
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4950049419007052,
+      "learning_rate": 1.3129910908090997e-05,
+      "loss": 0.8048,
+      "step": 17688
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5467747910735186,
+      "learning_rate": 1.3129186196107522e-05,
+      "loss": 0.5705,
+      "step": 17689
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9003770121125925,
+      "learning_rate": 1.3128461465904938e-05,
+      "loss": 0.8368,
+      "step": 17690
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3414068246613793,
+      "learning_rate": 1.3127736717487459e-05,
+      "loss": 0.7884,
+      "step": 17691
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.385703223124979,
+      "learning_rate": 1.3127011950859307e-05,
+      "loss": 0.6495,
+      "step": 17692
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6113320479204296,
+      "learning_rate": 1.31262871660247e-05,
+      "loss": 0.5482,
+      "step": 17693
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6484625784947323,
+      "learning_rate": 1.3125562362987863e-05,
+      "loss": 0.5372,
+      "step": 17694
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5589487393647938,
+      "learning_rate": 1.3124837541753008e-05,
+      "loss": 0.6275,
+      "step": 17695
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4704370036347552,
+      "learning_rate": 1.3124112702324363e-05,
+      "loss": 0.7556,
+      "step": 17696
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.401331224306128,
+      "learning_rate": 1.3123387844706144e-05,
+      "loss": 0.7147,
+      "step": 17697
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5957261085717809,
+      "learning_rate": 1.312266296890257e-05,
+      "loss": 0.5442,
+      "step": 17698
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5934997204693957,
+      "learning_rate": 1.3121938074917866e-05,
+      "loss": 0.5313,
+      "step": 17699
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4969481340504875,
+      "learning_rate": 1.3121213162756247e-05,
+      "loss": 0.8146,
+      "step": 17700
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.415837113128405,
+      "learning_rate": 1.3120488232421939e-05,
+      "loss": 0.7661,
+      "step": 17701
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6364128128542745,
+      "learning_rate": 1.311976328391916e-05,
+      "loss": 0.5618,
+      "step": 17702
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2153314128606334,
+      "learning_rate": 1.3119038317252133e-05,
+      "loss": 0.7078,
+      "step": 17703
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6693225386271833,
+      "learning_rate": 1.3118313332425077e-05,
+      "loss": 0.7001,
+      "step": 17704
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4276744171760842,
+      "learning_rate": 1.3117588329442213e-05,
+      "loss": 0.7119,
+      "step": 17705
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4781791268343412,
+      "learning_rate": 1.3116863308307762e-05,
+      "loss": 0.8795,
+      "step": 17706
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.674440081825638,
+      "learning_rate": 1.3116138269025947e-05,
+      "loss": 0.7452,
+      "step": 17707
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3913984733174263,
+      "learning_rate": 1.3115413211600987e-05,
+      "loss": 0.7395,
+      "step": 17708
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6041821209508906,
+      "learning_rate": 1.3114688136037109e-05,
+      "loss": 0.5611,
+      "step": 17709
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8976001697391818,
+      "learning_rate": 1.3113963042338527e-05,
+      "loss": 0.7742,
+      "step": 17710
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.1408713698054918,
+      "learning_rate": 1.3113237930509469e-05,
+      "loss": 0.7393,
+      "step": 17711
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3042475472018322,
+      "learning_rate": 1.3112512800554149e-05,
+      "loss": 0.7468,
+      "step": 17712
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2334728310738328,
+      "learning_rate": 1.31117876524768e-05,
+      "loss": 0.7261,
+      "step": 17713
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5659659577827731,
+      "learning_rate": 1.3111062486281632e-05,
+      "loss": 0.7458,
+      "step": 17714
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7104004759956473,
+      "learning_rate": 1.3110337301972877e-05,
+      "loss": 0.6775,
+      "step": 17715
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2706726646559836,
+      "learning_rate": 1.3109612099554754e-05,
+      "loss": 0.7798,
+      "step": 17716
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5954257513428969,
+      "learning_rate": 1.3108886879031485e-05,
+      "loss": 0.7755,
+      "step": 17717
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6077543170706947,
+      "learning_rate": 1.310816164040729e-05,
+      "loss": 0.5547,
+      "step": 17718
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8278589241688705,
+      "learning_rate": 1.3107436383686394e-05,
+      "loss": 0.7398,
+      "step": 17719
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4196924414041026,
+      "learning_rate": 1.3106711108873022e-05,
+      "loss": 0.8791,
+      "step": 17720
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4353161360908555,
+      "learning_rate": 1.3105985815971392e-05,
+      "loss": 0.6336,
+      "step": 17721
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.811890623049277,
+      "learning_rate": 1.3105260504985732e-05,
+      "loss": 0.7538,
+      "step": 17722
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6089653397168948,
+      "learning_rate": 1.3104535175920261e-05,
+      "loss": 0.5052,
+      "step": 17723
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.766899363843895,
+      "learning_rate": 1.3103809828779202e-05,
+      "loss": 0.7682,
+      "step": 17724
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5361156544417516,
+      "learning_rate": 1.310308446356678e-05,
+      "loss": 0.681,
+      "step": 17725
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3415892817315171,
+      "learning_rate": 1.3102359080287219e-05,
+      "loss": 0.645,
+      "step": 17726
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5316928874697555,
+      "learning_rate": 1.3101633678944741e-05,
+      "loss": 0.8368,
+      "step": 17727
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8824663665409629,
+      "learning_rate": 1.3100908259543567e-05,
+      "loss": 0.8311,
+      "step": 17728
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4710891355166225,
+      "learning_rate": 1.3100182822087928e-05,
+      "loss": 0.6757,
+      "step": 17729
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6479113743904796,
+      "learning_rate": 1.309945736658204e-05,
+      "loss": 0.7378,
+      "step": 17730
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5774956522721953,
+      "learning_rate": 1.309873189303013e-05,
+      "loss": 0.5147,
+      "step": 17731
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4634438834563062,
+      "learning_rate": 1.3098006401436422e-05,
+      "loss": 0.77,
+      "step": 17732
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.741991617834813,
+      "learning_rate": 1.3097280891805141e-05,
+      "loss": 0.6642,
+      "step": 17733
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4814488384523752,
+      "learning_rate": 1.309655536414051e-05,
+      "loss": 0.5941,
+      "step": 17734
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.501450822210358,
+      "learning_rate": 1.3095829818446753e-05,
+      "loss": 0.7532,
+      "step": 17735
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3525576390422087,
+      "learning_rate": 1.3095104254728095e-05,
+      "loss": 0.7333,
+      "step": 17736
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6936929719207732,
+      "learning_rate": 1.3094378672988759e-05,
+      "loss": 0.6828,
+      "step": 17737
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6859127857089078,
+      "learning_rate": 1.3093653073232972e-05,
+      "loss": 0.6857,
+      "step": 17738
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4762574595102227,
+      "learning_rate": 1.3092927455464959e-05,
+      "loss": 0.7894,
+      "step": 17739
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2770819200387447,
+      "learning_rate": 1.3092201819688937e-05,
+      "loss": 0.846,
+      "step": 17740
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3038552672480117,
+      "learning_rate": 1.3091476165909142e-05,
+      "loss": 0.6738,
+      "step": 17741
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.471456514896925,
+      "learning_rate": 1.3090750494129792e-05,
+      "loss": 0.7104,
+      "step": 17742
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.452301838796093,
+      "learning_rate": 1.3090024804355117e-05,
+      "loss": 0.7327,
+      "step": 17743
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7759847272599163,
+      "learning_rate": 1.3089299096589336e-05,
+      "loss": 0.7775,
+      "step": 17744
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2055904327901499,
+      "learning_rate": 1.308857337083668e-05,
+      "loss": 0.8528,
+      "step": 17745
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0510763091817066,
+      "learning_rate": 1.3087847627101372e-05,
+      "loss": 0.6824,
+      "step": 17746
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.41476561908476,
+      "learning_rate": 1.3087121865387636e-05,
+      "loss": 0.6547,
+      "step": 17747
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3978492265085711,
+      "learning_rate": 1.30863960856997e-05,
+      "loss": 0.7345,
+      "step": 17748
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3843582330681599,
+      "learning_rate": 1.3085670288041787e-05,
+      "loss": 0.785,
+      "step": 17749
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7794242909869247,
+      "learning_rate": 1.3084944472418126e-05,
+      "loss": 0.7819,
+      "step": 17750
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5748605964878637,
+      "learning_rate": 1.3084218638832943e-05,
+      "loss": 0.7735,
+      "step": 17751
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3032073475590489,
+      "learning_rate": 1.3083492787290462e-05,
+      "loss": 0.8314,
+      "step": 17752
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3934312828782753,
+      "learning_rate": 1.308276691779491e-05,
+      "loss": 0.7636,
+      "step": 17753
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6642326575241533,
+      "learning_rate": 1.3082041030350512e-05,
+      "loss": 0.7756,
+      "step": 17754
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6146124351660545,
+      "learning_rate": 1.3081315124961496e-05,
+      "loss": 0.703,
+      "step": 17755
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5964352974650087,
+      "learning_rate": 1.3080589201632086e-05,
+      "loss": 0.5745,
+      "step": 17756
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.7819698032724616,
+      "learning_rate": 1.3079863260366512e-05,
+      "loss": 0.8161,
+      "step": 17757
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5528641274471124,
+      "learning_rate": 1.3079137301168998e-05,
+      "loss": 0.6673,
+      "step": 17758
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7401897422890127,
+      "learning_rate": 1.3078411324043772e-05,
+      "loss": 0.786,
+      "step": 17759
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5580969980215678,
+      "learning_rate": 1.3077685328995065e-05,
+      "loss": 0.6917,
+      "step": 17760
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4059227660369564,
+      "learning_rate": 1.3076959316027093e-05,
+      "loss": 0.7433,
+      "step": 17761
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.440124609935682,
+      "learning_rate": 1.3076233285144096e-05,
+      "loss": 0.7978,
+      "step": 17762
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.76163450571364,
+      "learning_rate": 1.307550723635029e-05,
+      "loss": 0.814,
+      "step": 17763
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4574334125040875,
+      "learning_rate": 1.3074781169649908e-05,
+      "loss": 0.6792,
+      "step": 17764
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6081899159610764,
+      "learning_rate": 1.3074055085047174e-05,
+      "loss": 0.76,
+      "step": 17765
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5062663287913525,
+      "learning_rate": 1.3073328982546325e-05,
+      "loss": 0.6551,
+      "step": 17766
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6045781983157759,
+      "learning_rate": 1.3072602862151577e-05,
+      "loss": 0.678,
+      "step": 17767
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5762735228370248,
+      "learning_rate": 1.307187672386716e-05,
+      "loss": 0.7037,
+      "step": 17768
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5159201762693648,
+      "learning_rate": 1.3071150567697313e-05,
+      "loss": 0.6251,
+      "step": 17769
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2557099400574214,
+      "learning_rate": 1.3070424393646246e-05,
+      "loss": 0.7972,
+      "step": 17770
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.525162488669887,
+      "learning_rate": 1.3069698201718202e-05,
+      "loss": 0.771,
+      "step": 17771
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2980452910329525,
+      "learning_rate": 1.3068971991917398e-05,
+      "loss": 0.6129,
+      "step": 17772
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8176120212955484,
+      "learning_rate": 1.3068245764248073e-05,
+      "loss": 0.6653,
+      "step": 17773
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4140943333484062,
+      "learning_rate": 1.3067519518714447e-05,
+      "loss": 0.7279,
+      "step": 17774
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3920100287309976,
+      "learning_rate": 1.306679325532075e-05,
+      "loss": 0.8667,
+      "step": 17775
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9850884434777212,
+      "learning_rate": 1.3066066974071213e-05,
+      "loss": 0.795,
+      "step": 17776
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.129940995800023,
+      "learning_rate": 1.3065340674970064e-05,
+      "loss": 0.654,
+      "step": 17777
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.673460172706017,
+      "learning_rate": 1.3064614358021532e-05,
+      "loss": 0.6548,
+      "step": 17778
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5000713508738786,
+      "learning_rate": 1.3063888023229842e-05,
+      "loss": 0.5965,
+      "step": 17779
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8100416509097361,
+      "learning_rate": 1.3063161670599227e-05,
+      "loss": 0.8743,
+      "step": 17780
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.46647953009624,
+      "learning_rate": 1.3062435300133917e-05,
+      "loss": 0.8629,
+      "step": 17781
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5909765830566681,
+      "learning_rate": 1.3061708911838138e-05,
+      "loss": 0.5185,
+      "step": 17782
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4256422609115271,
+      "learning_rate": 1.306098250571612e-05,
+      "loss": 0.683,
+      "step": 17783
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4445102762163464,
+      "learning_rate": 1.3060256081772092e-05,
+      "loss": 0.7393,
+      "step": 17784
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6655883279867735,
+      "learning_rate": 1.3059529640010284e-05,
+      "loss": 0.5517,
+      "step": 17785
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.415444649419648,
+      "learning_rate": 1.3058803180434927e-05,
+      "loss": 0.6879,
+      "step": 17786
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3854427796108522,
+      "learning_rate": 1.3058076703050249e-05,
+      "loss": 0.6879,
+      "step": 17787
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4659098693764183,
+      "learning_rate": 1.3057350207860484e-05,
+      "loss": 0.819,
+      "step": 17788
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2696609533860959,
+      "learning_rate": 1.305662369486985e-05,
+      "loss": 0.7311,
+      "step": 17789
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8474948197476788,
+      "learning_rate": 1.3055897164082594e-05,
+      "loss": 0.6166,
+      "step": 17790
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4279183682579384,
+      "learning_rate": 1.305517061550293e-05,
+      "loss": 0.8325,
+      "step": 17791
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8631656561233267,
+      "learning_rate": 1.3054444049135099e-05,
+      "loss": 0.7648,
+      "step": 17792
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6083999345386534,
+      "learning_rate": 1.3053717464983327e-05,
+      "loss": 0.5668,
+      "step": 17793
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.655851625969575,
+      "learning_rate": 1.3052990863051846e-05,
+      "loss": 0.7935,
+      "step": 17794
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3902581032399521,
+      "learning_rate": 1.3052264243344885e-05,
+      "loss": 0.7324,
+      "step": 17795
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5723170931200634,
+      "learning_rate": 1.3051537605866675e-05,
+      "loss": 0.7119,
+      "step": 17796
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.390387439618017,
+      "learning_rate": 1.305081095062145e-05,
+      "loss": 0.7347,
+      "step": 17797
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6497845648869576,
+      "learning_rate": 1.3050084277613433e-05,
+      "loss": 0.7909,
+      "step": 17798
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2926365532165711,
+      "learning_rate": 1.3049357586846862e-05,
+      "loss": 0.7698,
+      "step": 17799
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5298218558521922,
+      "learning_rate": 1.3048630878325966e-05,
+      "loss": 0.5359,
+      "step": 17800
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3037096230048435,
+      "learning_rate": 1.3047904152054974e-05,
+      "loss": 0.6714,
+      "step": 17801
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9058722293683754,
+      "learning_rate": 1.3047177408038119e-05,
+      "loss": 0.7368,
+      "step": 17802
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4691066004878959,
+      "learning_rate": 1.3046450646279633e-05,
+      "loss": 0.69,
+      "step": 17803
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4432686727441162,
+      "learning_rate": 1.3045723866783749e-05,
+      "loss": 0.5943,
+      "step": 17804
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6139437668614363,
+      "learning_rate": 1.3044997069554695e-05,
+      "loss": 0.6761,
+      "step": 17805
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5350974602482081,
+      "learning_rate": 1.3044270254596706e-05,
+      "loss": 0.6741,
+      "step": 17806
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6120097550838589,
+      "learning_rate": 1.3043543421914007e-05,
+      "loss": 0.515,
+      "step": 17807
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5493063943373664,
+      "learning_rate": 1.3042816571510836e-05,
+      "loss": 0.7714,
+      "step": 17808
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.501858581297139,
+      "learning_rate": 1.3042089703391425e-05,
+      "loss": 0.8309,
+      "step": 17809
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.415348861593346,
+      "learning_rate": 1.3041362817560007e-05,
+      "loss": 0.6126,
+      "step": 17810
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7021707364073422,
+      "learning_rate": 1.304063591402081e-05,
+      "loss": 0.6409,
+      "step": 17811
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3599822126983545,
+      "learning_rate": 1.3039908992778065e-05,
+      "loss": 0.7463,
+      "step": 17812
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8996813164261996,
+      "learning_rate": 1.3039182053836013e-05,
+      "loss": 0.606,
+      "step": 17813
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3613747851983813,
+      "learning_rate": 1.3038455097198875e-05,
+      "loss": 0.8083,
+      "step": 17814
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.458046881563633,
+      "learning_rate": 1.3037728122870891e-05,
+      "loss": 0.678,
+      "step": 17815
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.674593418005604,
+      "learning_rate": 1.3037001130856296e-05,
+      "loss": 0.7646,
+      "step": 17816
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7784173758759745,
+      "learning_rate": 1.3036274121159313e-05,
+      "loss": 0.8101,
+      "step": 17817
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5955109554755122,
+      "learning_rate": 1.3035547093784187e-05,
+      "loss": 0.5426,
+      "step": 17818
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.598836914957848,
+      "learning_rate": 1.3034820048735139e-05,
+      "loss": 0.6339,
+      "step": 17819
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4950983585525157,
+      "learning_rate": 1.3034092986016413e-05,
+      "loss": 0.7472,
+      "step": 17820
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.762518416760466,
+      "learning_rate": 1.3033365905632234e-05,
+      "loss": 0.8147,
+      "step": 17821
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4019518138665994,
+      "learning_rate": 1.3032638807586842e-05,
+      "loss": 0.7186,
+      "step": 17822
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4656750420612432,
+      "learning_rate": 1.303191169188446e-05,
+      "loss": 0.7238,
+      "step": 17823
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5855347659955736,
+      "learning_rate": 1.3031184558529335e-05,
+      "loss": 0.6688,
+      "step": 17824
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3595838544207572,
+      "learning_rate": 1.303045740752569e-05,
+      "loss": 0.7345,
+      "step": 17825
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6943424040161843,
+      "learning_rate": 1.3029730238877764e-05,
+      "loss": 0.7586,
+      "step": 17826
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.747427202994992,
+      "learning_rate": 1.3029003052589789e-05,
+      "loss": 0.7933,
+      "step": 17827
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3933576943330674,
+      "learning_rate": 1.3028275848665998e-05,
+      "loss": 0.8242,
+      "step": 17828
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3576592395428504,
+      "learning_rate": 1.3027548627110629e-05,
+      "loss": 0.8531,
+      "step": 17829
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3421541982836147,
+      "learning_rate": 1.3026821387927911e-05,
+      "loss": 0.8083,
+      "step": 17830
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.375750122235693,
+      "learning_rate": 1.302609413112208e-05,
+      "loss": 0.6929,
+      "step": 17831
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3389410480188975,
+      "learning_rate": 1.3025366856697372e-05,
+      "loss": 0.7617,
+      "step": 17832
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5667415670091545,
+      "learning_rate": 1.3024639564658022e-05,
+      "loss": 0.8067,
+      "step": 17833
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5105868645774687,
+      "learning_rate": 1.3023912255008261e-05,
+      "loss": 0.7712,
+      "step": 17834
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1079471086381742,
+      "learning_rate": 1.3023184927752326e-05,
+      "loss": 0.8543,
+      "step": 17835
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.405222503289619,
+      "learning_rate": 1.3022457582894452e-05,
+      "loss": 0.7719,
+      "step": 17836
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6262365285241271,
+      "learning_rate": 1.302173022043887e-05,
+      "loss": 0.7446,
+      "step": 17837
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4824518344438415,
+      "learning_rate": 1.302100284038982e-05,
+      "loss": 0.6455,
+      "step": 17838
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4878490013817456,
+      "learning_rate": 1.3020275442751538e-05,
+      "loss": 0.7206,
+      "step": 17839
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.267483488784569,
+      "learning_rate": 1.3019548027528253e-05,
+      "loss": 0.6821,
+      "step": 17840
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3881316019802823,
+      "learning_rate": 1.3018820594724204e-05,
+      "loss": 0.7409,
+      "step": 17841
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.401361699230632,
+      "learning_rate": 1.3018093144343624e-05,
+      "loss": 0.6523,
+      "step": 17842
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6211470772010743,
+      "learning_rate": 1.3017365676390755e-05,
+      "loss": 0.5921,
+      "step": 17843
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5536057386461307,
+      "learning_rate": 1.3016638190869822e-05,
+      "loss": 0.5306,
+      "step": 17844
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3234782307361188,
+      "learning_rate": 1.3015910687785072e-05,
+      "loss": 0.7691,
+      "step": 17845
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2734844230635083,
+      "learning_rate": 1.3015183167140731e-05,
+      "loss": 0.725,
+      "step": 17846
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6149642455225783,
+      "learning_rate": 1.3014455628941041e-05,
+      "loss": 0.7849,
+      "step": 17847
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5573042724259336,
+      "learning_rate": 1.3013728073190235e-05,
+      "loss": 0.8044,
+      "step": 17848
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5418408669057028,
+      "learning_rate": 1.3013000499892549e-05,
+      "loss": 0.5142,
+      "step": 17849
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9042226210508015,
+      "learning_rate": 1.3012272909052221e-05,
+      "loss": 0.7818,
+      "step": 17850
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3738345311251816,
+      "learning_rate": 1.3011545300673486e-05,
+      "loss": 0.6305,
+      "step": 17851
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4327160416127225,
+      "learning_rate": 1.3010817674760583e-05,
+      "loss": 0.677,
+      "step": 17852
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4130436262293495,
+      "learning_rate": 1.3010090031317746e-05,
+      "loss": 0.823,
+      "step": 17853
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.098351234751551,
+      "learning_rate": 1.300936237034921e-05,
+      "loss": 0.8083,
+      "step": 17854
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.429545884645769,
+      "learning_rate": 1.3008634691859214e-05,
+      "loss": 0.5888,
+      "step": 17855
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5151584752111564,
+      "learning_rate": 1.3007906995851994e-05,
+      "loss": 0.7361,
+      "step": 17856
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7122210823888566,
+      "learning_rate": 1.3007179282331788e-05,
+      "loss": 0.6794,
+      "step": 17857
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.653069732588966,
+      "learning_rate": 1.3006451551302833e-05,
+      "loss": 0.6407,
+      "step": 17858
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4049986078016177,
+      "learning_rate": 1.3005723802769363e-05,
+      "loss": 0.7385,
+      "step": 17859
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2906237688486977,
+      "learning_rate": 1.300499603673562e-05,
+      "loss": 0.7682,
+      "step": 17860
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6396683636926959,
+      "learning_rate": 1.3004268253205837e-05,
+      "loss": 0.6946,
+      "step": 17861
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7302310805080634,
+      "learning_rate": 1.3003540452184254e-05,
+      "loss": 0.7388,
+      "step": 17862
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4326131838053162,
+      "learning_rate": 1.3002812633675105e-05,
+      "loss": 0.8598,
+      "step": 17863
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5877619011509994,
+      "learning_rate": 1.3002084797682633e-05,
+      "loss": 0.8098,
+      "step": 17864
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3988706349287152,
+      "learning_rate": 1.300135694421107e-05,
+      "loss": 0.6605,
+      "step": 17865
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6058345417611544,
+      "learning_rate": 1.3000629073264658e-05,
+      "loss": 0.8583,
+      "step": 17866
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.477369578711821,
+      "learning_rate": 1.2999901184847636e-05,
+      "loss": 0.7821,
+      "step": 17867
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.389060229816294,
+      "learning_rate": 1.2999173278964237e-05,
+      "loss": 0.7373,
+      "step": 17868
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6308252418466709,
+      "learning_rate": 1.2998445355618704e-05,
+      "loss": 0.6952,
+      "step": 17869
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5511938307911513,
+      "learning_rate": 1.2997717414815269e-05,
+      "loss": 0.8411,
+      "step": 17870
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5995774349973046,
+      "learning_rate": 1.299698945655818e-05,
+      "loss": 0.6196,
+      "step": 17871
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.460457306685319,
+      "learning_rate": 1.2996261480851664e-05,
+      "loss": 0.7784,
+      "step": 17872
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4052815844144428,
+      "learning_rate": 1.299553348769997e-05,
+      "loss": 0.7169,
+      "step": 17873
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2667232635014467,
+      "learning_rate": 1.2994805477107323e-05,
+      "loss": 0.6395,
+      "step": 17874
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6617898472199724,
+      "learning_rate": 1.299407744907798e-05,
+      "loss": 0.7727,
+      "step": 17875
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5578680541360298,
+      "learning_rate": 1.2993349403616167e-05,
+      "loss": 0.5178,
+      "step": 17876
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.543089326126453,
+      "learning_rate": 1.2992621340726124e-05,
+      "loss": 0.6722,
+      "step": 17877
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5902180217233891,
+      "learning_rate": 1.2991893260412093e-05,
+      "loss": 0.5303,
+      "step": 17878
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3113655208204091,
+      "learning_rate": 1.2991165162678315e-05,
+      "loss": 0.735,
+      "step": 17879
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.577799392516058,
+      "learning_rate": 1.2990437047529022e-05,
+      "loss": 0.5098,
+      "step": 17880
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5933517342543753,
+      "learning_rate": 1.298970891496846e-05,
+      "loss": 0.724,
+      "step": 17881
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5513834509443996,
+      "learning_rate": 1.2988980765000867e-05,
+      "loss": 0.7137,
+      "step": 17882
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3841041003033718,
+      "learning_rate": 1.298825259763048e-05,
+      "loss": 0.8424,
+      "step": 17883
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5057450717397889,
+      "learning_rate": 1.298752441286154e-05,
+      "loss": 0.7217,
+      "step": 17884
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7101137328565146,
+      "learning_rate": 1.2986796210698288e-05,
+      "loss": 0.8304,
+      "step": 17885
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0311938704474954,
+      "learning_rate": 1.2986067991144963e-05,
+      "loss": 0.6993,
+      "step": 17886
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.518745015368286,
+      "learning_rate": 1.2985339754205806e-05,
+      "loss": 0.7081,
+      "step": 17887
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6407212163869456,
+      "learning_rate": 1.2984611499885053e-05,
+      "loss": 0.5704,
+      "step": 17888
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5632250575888271,
+      "learning_rate": 1.2983883228186949e-05,
+      "loss": 0.6662,
+      "step": 17889
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4701115974170418,
+      "learning_rate": 1.2983154939115734e-05,
+      "loss": 0.7684,
+      "step": 17890
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.107113827731134,
+      "learning_rate": 1.2982426632675641e-05,
+      "loss": 0.7273,
+      "step": 17891
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7684959511087797,
+      "learning_rate": 1.2981698308870923e-05,
+      "loss": 0.7741,
+      "step": 17892
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5162508178740124,
+      "learning_rate": 1.2980969967705806e-05,
+      "loss": 0.7701,
+      "step": 17893
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4677479947414165,
+      "learning_rate": 1.2980241609184546e-05,
+      "loss": 0.7374,
+      "step": 17894
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4486156413757207,
+      "learning_rate": 1.2979513233311367e-05,
+      "loss": 0.6642,
+      "step": 17895
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.428872583139412,
+      "learning_rate": 1.2978784840090526e-05,
+      "loss": 0.7604,
+      "step": 17896
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6226804709314226,
+      "learning_rate": 1.2978056429526254e-05,
+      "loss": 0.7521,
+      "step": 17897
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4634115645771018,
+      "learning_rate": 1.2977328001622793e-05,
+      "loss": 0.6919,
+      "step": 17898
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.043294280844989,
+      "learning_rate": 1.2976599556384385e-05,
+      "loss": 0.7684,
+      "step": 17899
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3712803748416686,
+      "learning_rate": 1.2975871093815275e-05,
+      "loss": 0.6969,
+      "step": 17900
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6287298878631462,
+      "learning_rate": 1.2975142613919702e-05,
+      "loss": 0.7829,
+      "step": 17901
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.441657232081373,
+      "learning_rate": 1.2974414116701903e-05,
+      "loss": 0.7259,
+      "step": 17902
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.598338884506056,
+      "learning_rate": 1.2973685602166129e-05,
+      "loss": 0.7861,
+      "step": 17903
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5606463134442718,
+      "learning_rate": 1.297295707031661e-05,
+      "loss": 0.5896,
+      "step": 17904
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8086113337594443,
+      "learning_rate": 1.2972228521157598e-05,
+      "loss": 0.6665,
+      "step": 17905
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.414418432460597,
+      "learning_rate": 1.2971499954693327e-05,
+      "loss": 0.7201,
+      "step": 17906
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9053745789102179,
+      "learning_rate": 1.2970771370928045e-05,
+      "loss": 0.801,
+      "step": 17907
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5799176453793818,
+      "learning_rate": 1.2970042769865992e-05,
+      "loss": 0.5651,
+      "step": 17908
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5560984614432858,
+      "learning_rate": 1.2969314151511407e-05,
+      "loss": 0.6366,
+      "step": 17909
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6130942990629487,
+      "learning_rate": 1.2968585515868537e-05,
+      "loss": 0.7526,
+      "step": 17910
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.0986059455617143,
+      "learning_rate": 1.2967856862941623e-05,
+      "loss": 0.7495,
+      "step": 17911
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5382056384964822,
+      "learning_rate": 1.2967128192734903e-05,
+      "loss": 0.6327,
+      "step": 17912
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7181803059637413,
+      "learning_rate": 1.2966399505252628e-05,
+      "loss": 0.8308,
+      "step": 17913
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4787823324188332,
+      "learning_rate": 1.2965670800499032e-05,
+      "loss": 0.7621,
+      "step": 17914
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5926305611062666,
+      "learning_rate": 1.2964942078478365e-05,
+      "loss": 0.7383,
+      "step": 17915
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5898236221565645,
+      "learning_rate": 1.2964213339194865e-05,
+      "loss": 0.7637,
+      "step": 17916
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7857742604964273,
+      "learning_rate": 1.2963484582652775e-05,
+      "loss": 0.8488,
+      "step": 17917
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4688771324435852,
+      "learning_rate": 1.2962755808856341e-05,
+      "loss": 0.7474,
+      "step": 17918
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5029036311269037,
+      "learning_rate": 1.2962027017809804e-05,
+      "loss": 0.7643,
+      "step": 17919
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5864043559781826,
+      "learning_rate": 1.296129820951741e-05,
+      "loss": 0.7584,
+      "step": 17920
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4055288319172785,
+      "learning_rate": 1.2960569383983397e-05,
+      "loss": 0.7944,
+      "step": 17921
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3188219758540551,
+      "learning_rate": 1.2959840541212017e-05,
+      "loss": 0.7315,
+      "step": 17922
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4372335375203746,
+      "learning_rate": 1.29591116812075e-05,
+      "loss": 0.7139,
+      "step": 17923
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3619567978610385,
+      "learning_rate": 1.2958382803974106e-05,
+      "loss": 0.7706,
+      "step": 17924
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.479895482583548,
+      "learning_rate": 1.2957653909516065e-05,
+      "loss": 0.8355,
+      "step": 17925
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5326092523834367,
+      "learning_rate": 1.2956924997837633e-05,
+      "loss": 0.6314,
+      "step": 17926
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.9066094160851637,
+      "learning_rate": 1.2956196068943043e-05,
+      "loss": 0.5746,
+      "step": 17927
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7249177765366435,
+      "learning_rate": 1.2955467122836545e-05,
+      "loss": 0.5924,
+      "step": 17928
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8183348710517184,
+      "learning_rate": 1.295473815952238e-05,
+      "loss": 0.8741,
+      "step": 17929
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6480594095195122,
+      "learning_rate": 1.2954009179004794e-05,
+      "loss": 0.7297,
+      "step": 17930
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5297717807988955,
+      "learning_rate": 1.2953280181288033e-05,
+      "loss": 0.6672,
+      "step": 17931
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5408870707953222,
+      "learning_rate": 1.2952551166376339e-05,
+      "loss": 0.8719,
+      "step": 17932
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5514487515199817,
+      "learning_rate": 1.2951822134273957e-05,
+      "loss": 0.8193,
+      "step": 17933
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6376000525302368,
+      "learning_rate": 1.2951093084985131e-05,
+      "loss": 0.5357,
+      "step": 17934
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4744970641743647,
+      "learning_rate": 1.295036401851411e-05,
+      "loss": 0.7384,
+      "step": 17935
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8009995969726385,
+      "learning_rate": 1.2949634934865134e-05,
+      "loss": 0.8344,
+      "step": 17936
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.689226964301675,
+      "learning_rate": 1.294890583404245e-05,
+      "loss": 0.7698,
+      "step": 17937
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4668436668760654,
+      "learning_rate": 1.29481767160503e-05,
+      "loss": 0.8156,
+      "step": 17938
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3993782705553475,
+      "learning_rate": 1.2947447580892933e-05,
+      "loss": 0.7621,
+      "step": 17939
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5415345884409162,
+      "learning_rate": 1.2946718428574595e-05,
+      "loss": 0.7755,
+      "step": 17940
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4000946572031454,
+      "learning_rate": 1.294598925909953e-05,
+      "loss": 0.7552,
+      "step": 17941
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4610260685922611,
+      "learning_rate": 1.294526007247198e-05,
+      "loss": 0.6927,
+      "step": 17942
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4340145828783961,
+      "learning_rate": 1.2944530868696194e-05,
+      "loss": 0.8086,
+      "step": 17943
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2255352380940723,
+      "learning_rate": 1.2943801647776418e-05,
+      "loss": 0.6917,
+      "step": 17944
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.586535915729907,
+      "learning_rate": 1.2943072409716897e-05,
+      "loss": 0.7072,
+      "step": 17945
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5381111023278564,
+      "learning_rate": 1.2942343154521875e-05,
+      "loss": 0.6639,
+      "step": 17946
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4658445502544102,
+      "learning_rate": 1.2941613882195601e-05,
+      "loss": 0.785,
+      "step": 17947
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.591649315070643,
+      "learning_rate": 1.294088459274232e-05,
+      "loss": 0.7007,
+      "step": 17948
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4408922281199203,
+      "learning_rate": 1.2940155286166279e-05,
+      "loss": 0.787,
+      "step": 17949
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4013824719582395,
+      "learning_rate": 1.2939425962471719e-05,
+      "loss": 0.7737,
+      "step": 17950
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4155234535289822,
+      "learning_rate": 1.293869662166289e-05,
+      "loss": 0.8801,
+      "step": 17951
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4223380871481885,
+      "learning_rate": 1.2937967263744044e-05,
+      "loss": 0.7914,
+      "step": 17952
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6470519306578815,
+      "learning_rate": 1.2937237888719418e-05,
+      "loss": 0.7393,
+      "step": 17953
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.739866511792358,
+      "learning_rate": 1.2936508496593268e-05,
+      "loss": 0.7323,
+      "step": 17954
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5016249411989897,
+      "learning_rate": 1.2935779087369832e-05,
+      "loss": 0.758,
+      "step": 17955
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3100683590226978,
+      "learning_rate": 1.2935049661053363e-05,
+      "loss": 0.7146,
+      "step": 17956
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3664277153362105,
+      "learning_rate": 1.2934320217648101e-05,
+      "loss": 0.8861,
+      "step": 17957
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3347416150501927,
+      "learning_rate": 1.2933590757158303e-05,
+      "loss": 0.7455,
+      "step": 17958
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2743681452823292,
+      "learning_rate": 1.293286127958821e-05,
+      "loss": 0.8206,
+      "step": 17959
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7945122459343903,
+      "learning_rate": 1.2932131784942069e-05,
+      "loss": 0.7928,
+      "step": 17960
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4518509600515388,
+      "learning_rate": 1.2931402273224128e-05,
+      "loss": 0.7489,
+      "step": 17961
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.721816248741482,
+      "learning_rate": 1.2930672744438634e-05,
+      "loss": 0.7407,
+      "step": 17962
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7650815556936787,
+      "learning_rate": 1.2929943198589836e-05,
+      "loss": 0.7628,
+      "step": 17963
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3456425992853394,
+      "learning_rate": 1.2929213635681983e-05,
+      "loss": 0.7062,
+      "step": 17964
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.45767571490665,
+      "learning_rate": 1.292848405571932e-05,
+      "loss": 0.7883,
+      "step": 17965
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2880859431599558,
+      "learning_rate": 1.2927754458706094e-05,
+      "loss": 0.6747,
+      "step": 17966
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3756699886917443,
+      "learning_rate": 1.2927024844646556e-05,
+      "loss": 0.7959,
+      "step": 17967
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7398019134324207,
+      "learning_rate": 1.2926295213544953e-05,
+      "loss": 0.7031,
+      "step": 17968
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9475008691126867,
+      "learning_rate": 1.2925565565405536e-05,
+      "loss": 0.7891,
+      "step": 17969
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.668647738293983,
+      "learning_rate": 1.2924835900232542e-05,
+      "loss": 0.6709,
+      "step": 17970
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9160775817277953,
+      "learning_rate": 1.2924106218030235e-05,
+      "loss": 0.6413,
+      "step": 17971
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.457598753853566,
+      "learning_rate": 1.292337651880285e-05,
+      "loss": 0.6612,
+      "step": 17972
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.366303149952499,
+      "learning_rate": 1.2922646802554649e-05,
+      "loss": 0.789,
+      "step": 17973
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.325710572469643,
+      "learning_rate": 1.2921917069289864e-05,
+      "loss": 0.6737,
+      "step": 17974
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4096299511066066,
+      "learning_rate": 1.2921187319012761e-05,
+      "loss": 0.7317,
+      "step": 17975
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4541983124308933,
+      "learning_rate": 1.2920457551727575e-05,
+      "loss": 0.716,
+      "step": 17976
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6104967058749287,
+      "learning_rate": 1.2919727767438562e-05,
+      "loss": 0.7213,
+      "step": 17977
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6442488872996321,
+      "learning_rate": 1.2918997966149969e-05,
+      "loss": 0.9102,
+      "step": 17978
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4507198390100926,
+      "learning_rate": 1.2918268147866047e-05,
+      "loss": 0.6224,
+      "step": 17979
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5489407090461698,
+      "learning_rate": 1.2917538312591043e-05,
+      "loss": 0.7048,
+      "step": 17980
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6534585720500423,
+      "learning_rate": 1.2916808460329206e-05,
+      "loss": 0.8202,
+      "step": 17981
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5712983019707103,
+      "learning_rate": 1.2916078591084788e-05,
+      "loss": 0.6696,
+      "step": 17982
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6669792917565112,
+      "learning_rate": 1.291534870486204e-05,
+      "loss": 0.7774,
+      "step": 17983
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5225714143137186,
+      "learning_rate": 1.2914618801665205e-05,
+      "loss": 0.7772,
+      "step": 17984
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4207768175484279,
+      "learning_rate": 1.2913888881498537e-05,
+      "loss": 0.793,
+      "step": 17985
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6031275237056615,
+      "learning_rate": 1.2913158944366286e-05,
+      "loss": 0.5563,
+      "step": 17986
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6125073482307334,
+      "learning_rate": 1.2912428990272704e-05,
+      "loss": 0.5245,
+      "step": 17987
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.059458694917052,
+      "learning_rate": 1.2911699019222035e-05,
+      "loss": 0.7462,
+      "step": 17988
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.930045220798693,
+      "learning_rate": 1.2910969031218532e-05,
+      "loss": 0.7344,
+      "step": 17989
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.543014389818123,
+      "learning_rate": 1.2910239026266446e-05,
+      "loss": 0.6968,
+      "step": 17990
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.383461808111236,
+      "learning_rate": 1.290950900437003e-05,
+      "loss": 0.7283,
+      "step": 17991
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5943638209040962,
+      "learning_rate": 1.2908778965533533e-05,
+      "loss": 0.8,
+      "step": 17992
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5608907144867058,
+      "learning_rate": 1.2908048909761197e-05,
+      "loss": 0.529,
+      "step": 17993
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5409640615774802,
+      "learning_rate": 1.2907318837057286e-05,
+      "loss": 0.7175,
+      "step": 17994
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5422643940427905,
+      "learning_rate": 1.2906588747426038e-05,
+      "loss": 0.7356,
+      "step": 17995
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3534637007465797,
+      "learning_rate": 1.2905858640871717e-05,
+      "loss": 0.6809,
+      "step": 17996
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.617205541529422,
+      "learning_rate": 1.2905128517398562e-05,
+      "loss": 0.5753,
+      "step": 17997
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4325256723408921,
+      "learning_rate": 1.2904398377010832e-05,
+      "loss": 0.6909,
+      "step": 17998
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5785983485473691,
+      "learning_rate": 1.2903668219712773e-05,
+      "loss": 0.7001,
+      "step": 17999
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.404527580685914,
+      "learning_rate": 1.2902938045508638e-05,
+      "loss": 0.7318,
+      "step": 18000
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6854032008958826,
+      "learning_rate": 1.2902207854402682e-05,
+      "loss": 0.6181,
+      "step": 18001
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4858696651255363,
+      "learning_rate": 1.2901477646399147e-05,
+      "loss": 0.6685,
+      "step": 18002
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4908566475121205,
+      "learning_rate": 1.2900747421502297e-05,
+      "loss": 0.737,
+      "step": 18003
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.376369652189608,
+      "learning_rate": 1.2900017179716373e-05,
+      "loss": 0.6785,
+      "step": 18004
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.429965891606164,
+      "learning_rate": 1.2899286921045633e-05,
+      "loss": 0.7671,
+      "step": 18005
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6662807364419217,
+      "learning_rate": 1.2898556645494327e-05,
+      "loss": 0.6851,
+      "step": 18006
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6125309287003071,
+      "learning_rate": 1.2897826353066705e-05,
+      "loss": 0.5406,
+      "step": 18007
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5671187569925932,
+      "learning_rate": 1.289709604376702e-05,
+      "loss": 0.4866,
+      "step": 18008
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4572163307615942,
+      "learning_rate": 1.2896365717599526e-05,
+      "loss": 0.7912,
+      "step": 18009
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5621533152208804,
+      "learning_rate": 1.2895635374568474e-05,
+      "loss": 0.7214,
+      "step": 18010
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6620495956055152,
+      "learning_rate": 1.2894905014678117e-05,
+      "loss": 0.858,
+      "step": 18011
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4490750542018327,
+      "learning_rate": 1.2894174637932703e-05,
+      "loss": 0.6757,
+      "step": 18012
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3617810730814481,
+      "learning_rate": 1.289344424433649e-05,
+      "loss": 0.7481,
+      "step": 18013
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5165572687608193,
+      "learning_rate": 1.2892713833893731e-05,
+      "loss": 0.7513,
+      "step": 18014
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7033646223272256,
+      "learning_rate": 1.2891983406608673e-05,
+      "loss": 0.6068,
+      "step": 18015
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5302843122945653,
+      "learning_rate": 1.2891252962485573e-05,
+      "loss": 0.9015,
+      "step": 18016
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6220239034479529,
+      "learning_rate": 1.2890522501528682e-05,
+      "loss": 0.7062,
+      "step": 18017
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.662802617407923,
+      "learning_rate": 1.2889792023742253e-05,
+      "loss": 0.7819,
+      "step": 18018
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3141121338883925,
+      "learning_rate": 1.2889061529130542e-05,
+      "loss": 0.8031,
+      "step": 18019
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.50298763353253,
+      "learning_rate": 1.2888331017697804e-05,
+      "loss": 0.6956,
+      "step": 18020
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9470665796358133,
+      "learning_rate": 1.2887600489448282e-05,
+      "loss": 0.7781,
+      "step": 18021
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3788073472035038,
+      "learning_rate": 1.2886869944386237e-05,
+      "loss": 0.7081,
+      "step": 18022
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.195365954340498,
+      "learning_rate": 1.2886139382515924e-05,
+      "loss": 0.6115,
+      "step": 18023
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5805663580248775,
+      "learning_rate": 1.2885408803841593e-05,
+      "loss": 0.5504,
+      "step": 18024
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.652453306838609,
+      "learning_rate": 1.2884678208367493e-05,
+      "loss": 0.7106,
+      "step": 18025
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.124035220271073,
+      "learning_rate": 1.288394759609789e-05,
+      "loss": 0.8423,
+      "step": 18026
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7722969901007823,
+      "learning_rate": 1.2883216967037029e-05,
+      "loss": 0.6342,
+      "step": 18027
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.282155047790296,
+      "learning_rate": 1.2882486321189165e-05,
+      "loss": 0.775,
+      "step": 18028
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5174070194828326,
+      "learning_rate": 1.2881755658558555e-05,
+      "loss": 0.6478,
+      "step": 18029
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.451783232308465,
+      "learning_rate": 1.2881024979149448e-05,
+      "loss": 0.77,
+      "step": 18030
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6347218086788033,
+      "learning_rate": 1.2880294282966104e-05,
+      "loss": 0.6933,
+      "step": 18031
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5655202832398386,
+      "learning_rate": 1.2879563570012773e-05,
+      "loss": 0.5281,
+      "step": 18032
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5358418065414092,
+      "learning_rate": 1.2878832840293713e-05,
+      "loss": 0.7445,
+      "step": 18033
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2844899966965952,
+      "learning_rate": 1.2878102093813176e-05,
+      "loss": 0.7247,
+      "step": 18034
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7047362921921723,
+      "learning_rate": 1.287737133057542e-05,
+      "loss": 0.7608,
+      "step": 18035
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6127538501960021,
+      "learning_rate": 1.2876640550584693e-05,
+      "loss": 0.5383,
+      "step": 18036
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3036576321661721,
+      "learning_rate": 1.2875909753845255e-05,
+      "loss": 0.628,
+      "step": 18037
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3831783493117435,
+      "learning_rate": 1.2875178940361361e-05,
+      "loss": 0.7755,
+      "step": 18038
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4757271853652465,
+      "learning_rate": 1.2874448110137263e-05,
+      "loss": 0.6849,
+      "step": 18039
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5279959728874588,
+      "learning_rate": 1.2873717263177221e-05,
+      "loss": 0.8019,
+      "step": 18040
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6566711566192505,
+      "learning_rate": 1.2872986399485487e-05,
+      "loss": 0.7508,
+      "step": 18041
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.476295927513513,
+      "learning_rate": 1.2872255519066314e-05,
+      "loss": 0.6223,
+      "step": 18042
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4584725562106593,
+      "learning_rate": 1.2871524621923963e-05,
+      "loss": 0.654,
+      "step": 18043
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.197735699290044,
+      "learning_rate": 1.2870793708062684e-05,
+      "loss": 0.7937,
+      "step": 18044
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5393294115295681,
+      "learning_rate": 1.2870062777486738e-05,
+      "loss": 0.8567,
+      "step": 18045
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6704993402765376,
+      "learning_rate": 1.2869331830200374e-05,
+      "loss": 0.7531,
+      "step": 18046
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.034212347342524,
+      "learning_rate": 1.2868600866207856e-05,
+      "loss": 0.7677,
+      "step": 18047
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6174801036213982,
+      "learning_rate": 1.2867869885513432e-05,
+      "loss": 0.6862,
+      "step": 18048
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4370713007419973,
+      "learning_rate": 1.2867138888121362e-05,
+      "loss": 0.8437,
+      "step": 18049
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5482016285637474,
+      "learning_rate": 1.2866407874035904e-05,
+      "loss": 0.8289,
+      "step": 18050
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3185259688165123,
+      "learning_rate": 1.2865676843261308e-05,
+      "loss": 0.8435,
+      "step": 18051
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4470291632872836,
+      "learning_rate": 1.2864945795801837e-05,
+      "loss": 0.7985,
+      "step": 18052
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.010859240935431,
+      "learning_rate": 1.2864214731661743e-05,
+      "loss": 0.7792,
+      "step": 18053
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7491531360423143,
+      "learning_rate": 1.2863483650845288e-05,
+      "loss": 0.6673,
+      "step": 18054
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5003229264297568,
+      "learning_rate": 1.2862752553356721e-05,
+      "loss": 0.6427,
+      "step": 18055
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4443379096305633,
+      "learning_rate": 1.2862021439200306e-05,
+      "loss": 0.7035,
+      "step": 18056
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6069067542831245,
+      "learning_rate": 1.2861290308380289e-05,
+      "loss": 0.5531,
+      "step": 18057
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5133796549894623,
+      "learning_rate": 1.286055916090094e-05,
+      "loss": 0.6909,
+      "step": 18058
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4454572141314699,
+      "learning_rate": 1.2859827996766508e-05,
+      "loss": 0.6694,
+      "step": 18059
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3825586136602095,
+      "learning_rate": 1.2859096815981254e-05,
+      "loss": 0.6429,
+      "step": 18060
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.371173859942639,
+      "learning_rate": 1.285836561854943e-05,
+      "loss": 0.7933,
+      "step": 18061
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.704756813921135,
+      "learning_rate": 1.2857634404475297e-05,
+      "loss": 0.7104,
+      "step": 18062
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5758254285506408,
+      "learning_rate": 1.2856903173763114e-05,
+      "loss": 0.7352,
+      "step": 18063
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8082153591203414,
+      "learning_rate": 1.2856171926417134e-05,
+      "loss": 0.6917,
+      "step": 18064
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.778286326030659,
+      "learning_rate": 1.2855440662441617e-05,
+      "loss": 0.6556,
+      "step": 18065
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3505138395664644,
+      "learning_rate": 1.285470938184082e-05,
+      "loss": 0.8108,
+      "step": 18066
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5461064501634754,
+      "learning_rate": 1.2853978084619004e-05,
+      "loss": 0.848,
+      "step": 18067
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9096952519297292,
+      "learning_rate": 1.285324677078042e-05,
+      "loss": 0.6853,
+      "step": 18068
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.966955049903765,
+      "learning_rate": 1.2852515440329331e-05,
+      "loss": 0.7427,
+      "step": 18069
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6337409517584454,
+      "learning_rate": 1.2851784093269997e-05,
+      "loss": 0.5881,
+      "step": 18070
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3869514767554516,
+      "learning_rate": 1.2851052729606673e-05,
+      "loss": 0.6097,
+      "step": 18071
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3260531733579919,
+      "learning_rate": 1.2850321349343613e-05,
+      "loss": 0.7471,
+      "step": 18072
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.006680275484889,
+      "learning_rate": 1.2849589952485084e-05,
+      "loss": 0.7935,
+      "step": 18073
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3084988610694217,
+      "learning_rate": 1.2848858539035339e-05,
+      "loss": 0.5642,
+      "step": 18074
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6182679040924814,
+      "learning_rate": 1.2848127108998638e-05,
+      "loss": 0.7673,
+      "step": 18075
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4976739388032025,
+      "learning_rate": 1.2847395662379236e-05,
+      "loss": 0.635,
+      "step": 18076
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5463018280193257,
+      "learning_rate": 1.28466641991814e-05,
+      "loss": 0.8695,
+      "step": 18077
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3430505907285655,
+      "learning_rate": 1.284593271940938e-05,
+      "loss": 0.7122,
+      "step": 18078
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5968124270460787,
+      "learning_rate": 1.2845201223067441e-05,
+      "loss": 0.5218,
+      "step": 18079
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8114271580007686,
+      "learning_rate": 1.2844469710159837e-05,
+      "loss": 0.7441,
+      "step": 18080
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.39551446823405,
+      "learning_rate": 1.2843738180690833e-05,
+      "loss": 0.703,
+      "step": 18081
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5569458401111087,
+      "learning_rate": 1.2843006634664683e-05,
+      "loss": 0.6944,
+      "step": 18082
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.501483806814065,
+      "learning_rate": 1.2842275072085649e-05,
+      "loss": 0.7021,
+      "step": 18083
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.540770896063538,
+      "learning_rate": 1.2841543492957988e-05,
+      "loss": 0.7672,
+      "step": 18084
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5320557382506337,
+      "learning_rate": 1.2840811897285963e-05,
+      "loss": 0.8615,
+      "step": 18085
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.544224996421577,
+      "learning_rate": 1.284008028507383e-05,
+      "loss": 0.7291,
+      "step": 18086
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4882916212111652,
+      "learning_rate": 1.2839348656325851e-05,
+      "loss": 0.7782,
+      "step": 18087
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6941427867001457,
+      "learning_rate": 1.2838617011046288e-05,
+      "loss": 0.6915,
+      "step": 18088
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.274760479389311,
+      "learning_rate": 1.2837885349239396e-05,
+      "loss": 0.7143,
+      "step": 18089
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5894568237577869,
+      "learning_rate": 1.2837153670909437e-05,
+      "loss": 0.5539,
+      "step": 18090
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8313135704294647,
+      "learning_rate": 1.2836421976060672e-05,
+      "loss": 0.7341,
+      "step": 18091
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.537450958889892,
+      "learning_rate": 1.283569026469736e-05,
+      "loss": 0.7739,
+      "step": 18092
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4953840159621894,
+      "learning_rate": 1.2834958536823758e-05,
+      "loss": 0.7631,
+      "step": 18093
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4891885195699028,
+      "learning_rate": 1.2834226792444136e-05,
+      "loss": 0.7566,
+      "step": 18094
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.663670862853625,
+      "learning_rate": 1.2833495031562744e-05,
+      "loss": 0.7238,
+      "step": 18095
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5131395779935408,
+      "learning_rate": 1.283276325418385e-05,
+      "loss": 0.7732,
+      "step": 18096
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6640101956124909,
+      "learning_rate": 1.2832031460311707e-05,
+      "loss": 0.6318,
+      "step": 18097
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5250415737295406,
+      "learning_rate": 1.2831299649950582e-05,
+      "loss": 0.7897,
+      "step": 18098
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6023627037097674,
+      "learning_rate": 1.2830567823104736e-05,
+      "loss": 0.7898,
+      "step": 18099
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4810690916229803,
+      "learning_rate": 1.2829835979778424e-05,
+      "loss": 0.7609,
+      "step": 18100
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7284125964711916,
+      "learning_rate": 1.2829104119975917e-05,
+      "loss": 0.611,
+      "step": 18101
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4229788921648756,
+      "learning_rate": 1.2828372243701463e-05,
+      "loss": 0.6438,
+      "step": 18102
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.670794879624801,
+      "learning_rate": 1.2827640350959338e-05,
+      "loss": 0.7578,
+      "step": 18103
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8814695906625127,
+      "learning_rate": 1.2826908441753789e-05,
+      "loss": 0.7147,
+      "step": 18104
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6862951980455252,
+      "learning_rate": 1.2826176516089088e-05,
+      "loss": 0.7477,
+      "step": 18105
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.460243685963257,
+      "learning_rate": 1.2825444573969494e-05,
+      "loss": 0.7594,
+      "step": 18106
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6261319699423602,
+      "learning_rate": 1.2824712615399264e-05,
+      "loss": 0.7205,
+      "step": 18107
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.928916189140762,
+      "learning_rate": 1.2823980640382663e-05,
+      "loss": 0.7347,
+      "step": 18108
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.270585591870843,
+      "learning_rate": 1.2823248648923954e-05,
+      "loss": 0.7633,
+      "step": 18109
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7969884176375057,
+      "learning_rate": 1.2822516641027396e-05,
+      "loss": 0.8212,
+      "step": 18110
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.369711208388761,
+      "learning_rate": 1.2821784616697253e-05,
+      "loss": 0.7396,
+      "step": 18111
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4999891109078296,
+      "learning_rate": 1.2821052575937786e-05,
+      "loss": 0.6643,
+      "step": 18112
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3874968572177881,
+      "learning_rate": 1.282032051875326e-05,
+      "loss": 0.7158,
+      "step": 18113
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3151871114323737,
+      "learning_rate": 1.2819588445147932e-05,
+      "loss": 0.7719,
+      "step": 18114
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5682646306481678,
+      "learning_rate": 1.2818856355126071e-05,
+      "loss": 0.6826,
+      "step": 18115
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4308703527393454,
+      "learning_rate": 1.2818124248691932e-05,
+      "loss": 0.6497,
+      "step": 18116
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.421999829837559,
+      "learning_rate": 1.2817392125849785e-05,
+      "loss": 0.8101,
+      "step": 18117
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3797249480134344,
+      "learning_rate": 1.2816659986603887e-05,
+      "loss": 0.8531,
+      "step": 18118
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6859624846681454,
+      "learning_rate": 1.2815927830958504e-05,
+      "loss": 0.7176,
+      "step": 18119
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.341233265052055,
+      "learning_rate": 1.2815195658917896e-05,
+      "loss": 0.6807,
+      "step": 18120
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5831663333386001,
+      "learning_rate": 1.2814463470486329e-05,
+      "loss": 0.5168,
+      "step": 18121
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.441615548175108,
+      "learning_rate": 1.2813731265668066e-05,
+      "loss": 0.7257,
+      "step": 18122
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4653794542598206,
+      "learning_rate": 1.2812999044467364e-05,
+      "loss": 0.7358,
+      "step": 18123
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.489009506472994,
+      "learning_rate": 1.2812266806888496e-05,
+      "loss": 0.729,
+      "step": 18124
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.634344444611965,
+      "learning_rate": 1.2811534552935716e-05,
+      "loss": 0.7632,
+      "step": 18125
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3612684229902523,
+      "learning_rate": 1.2810802282613297e-05,
+      "loss": 0.8077,
+      "step": 18126
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5806167491111205,
+      "learning_rate": 1.2810069995925491e-05,
+      "loss": 0.747,
+      "step": 18127
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2865533214613298,
+      "learning_rate": 1.2809337692876573e-05,
+      "loss": 0.6589,
+      "step": 18128
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.535268728090985,
+      "learning_rate": 1.2808605373470799e-05,
+      "loss": 0.706,
+      "step": 18129
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9219418951154366,
+      "learning_rate": 1.2807873037712437e-05,
+      "loss": 0.6773,
+      "step": 18130
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5620213313435223,
+      "learning_rate": 1.2807140685605748e-05,
+      "loss": 0.7329,
+      "step": 18131
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5927871866351972,
+      "learning_rate": 1.2806408317154997e-05,
+      "loss": 0.7488,
+      "step": 18132
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5286489945255253,
+      "learning_rate": 1.2805675932364449e-05,
+      "loss": 0.6554,
+      "step": 18133
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.402827902237445,
+      "learning_rate": 1.2804943531238366e-05,
+      "loss": 0.7249,
+      "step": 18134
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9285611685792554,
+      "learning_rate": 1.2804211113781016e-05,
+      "loss": 0.758,
+      "step": 18135
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4441672124310336,
+      "learning_rate": 1.280347867999666e-05,
+      "loss": 0.8267,
+      "step": 18136
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.370018918709214,
+      "learning_rate": 1.2802746229889563e-05,
+      "loss": 0.7766,
+      "step": 18137
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5659511884525847,
+      "learning_rate": 1.280201376346399e-05,
+      "loss": 0.6631,
+      "step": 18138
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6312332443246509,
+      "learning_rate": 1.2801281280724207e-05,
+      "loss": 0.8043,
+      "step": 18139
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3757223600029773,
+      "learning_rate": 1.2800548781674476e-05,
+      "loss": 0.6913,
+      "step": 18140
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4745506938787005,
+      "learning_rate": 1.2799816266319062e-05,
+      "loss": 0.8705,
+      "step": 18141
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6185288224326536,
+      "learning_rate": 1.2799083734662237e-05,
+      "loss": 0.5739,
+      "step": 18142
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3012540378641246,
+      "learning_rate": 1.2798351186708258e-05,
+      "loss": 0.7017,
+      "step": 18143
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4155118249819363,
+      "learning_rate": 1.2797618622461387e-05,
+      "loss": 0.7541,
+      "step": 18144
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5760122074332863,
+      "learning_rate": 1.2796886041925903e-05,
+      "loss": 0.847,
+      "step": 18145
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6177366183132762,
+      "learning_rate": 1.2796153445106057e-05,
+      "loss": 0.7616,
+      "step": 18146
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.291935596696309,
+      "learning_rate": 1.279542083200612e-05,
+      "loss": 0.7668,
+      "step": 18147
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9486588115753918,
+      "learning_rate": 1.2794688202630361e-05,
+      "loss": 0.7232,
+      "step": 18148
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5815480507324842,
+      "learning_rate": 1.2793955556983042e-05,
+      "loss": 0.6125,
+      "step": 18149
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3746510591418037,
+      "learning_rate": 1.2793222895068431e-05,
+      "loss": 0.6165,
+      "step": 18150
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4607857853426245,
+      "learning_rate": 1.2792490216890786e-05,
+      "loss": 0.777,
+      "step": 18151
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5309703867486486,
+      "learning_rate": 1.2791757522454387e-05,
+      "loss": 0.9436,
+      "step": 18152
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3111350360665688,
+      "learning_rate": 1.2791024811763484e-05,
+      "loss": 0.6246,
+      "step": 18153
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8220915049896578,
+      "learning_rate": 1.2790292084822355e-05,
+      "loss": 0.8235,
+      "step": 18154
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3254790934087424,
+      "learning_rate": 1.278955934163526e-05,
+      "loss": 0.6283,
+      "step": 18155
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4404916367302243,
+      "learning_rate": 1.2788826582206473e-05,
+      "loss": 0.6897,
+      "step": 18156
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6677607969071697,
+      "learning_rate": 1.2788093806540249e-05,
+      "loss": 0.7343,
+      "step": 18157
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.010481330290217,
+      "learning_rate": 1.2787361014640862e-05,
+      "loss": 0.7325,
+      "step": 18158
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6008856673222385,
+      "learning_rate": 1.2786628206512577e-05,
+      "loss": 0.727,
+      "step": 18159
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5557943360549025,
+      "learning_rate": 1.2785895382159658e-05,
+      "loss": 0.7565,
+      "step": 18160
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.450697608046359,
+      "learning_rate": 1.2785162541586378e-05,
+      "loss": 0.816,
+      "step": 18161
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.577730513712235,
+      "learning_rate": 1.2784429684796997e-05,
+      "loss": 0.6328,
+      "step": 18162
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.458779526125477,
+      "learning_rate": 1.2783696811795787e-05,
+      "loss": 0.6168,
+      "step": 18163
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5692793318915297,
+      "learning_rate": 1.278296392258701e-05,
+      "loss": 0.6565,
+      "step": 18164
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.559394202778748,
+      "learning_rate": 1.2782231017174937e-05,
+      "loss": 0.867,
+      "step": 18165
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.259717518348704,
+      "learning_rate": 1.2781498095563835e-05,
+      "loss": 0.7067,
+      "step": 18166
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4372822292930667,
+      "learning_rate": 1.2780765157757971e-05,
+      "loss": 0.6883,
+      "step": 18167
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.481586188589636,
+      "learning_rate": 1.278003220376161e-05,
+      "loss": 0.7035,
+      "step": 18168
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.454602385761706,
+      "learning_rate": 1.2779299233579024e-05,
+      "loss": 0.6944,
+      "step": 18169
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4641849329819665,
+      "learning_rate": 1.2778566247214474e-05,
+      "loss": 0.5816,
+      "step": 18170
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4979699813834269,
+      "learning_rate": 1.2777833244672235e-05,
+      "loss": 0.7223,
+      "step": 18171
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.386346390024407,
+      "learning_rate": 1.277710022595657e-05,
+      "loss": 0.715,
+      "step": 18172
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.342483253605583,
+      "learning_rate": 1.2776367191071752e-05,
+      "loss": 0.6943,
+      "step": 18173
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5101877449940593,
+      "learning_rate": 1.2775634140022039e-05,
+      "loss": 0.7973,
+      "step": 18174
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.417556965736431,
+      "learning_rate": 1.2774901072811709e-05,
+      "loss": 0.7858,
+      "step": 18175
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5661302583425928,
+      "learning_rate": 1.2774167989445025e-05,
+      "loss": 0.5124,
+      "step": 18176
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6150009309613422,
+      "learning_rate": 1.2773434889926259e-05,
+      "loss": 0.5473,
+      "step": 18177
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.353258665197205,
+      "learning_rate": 1.2772701774259674e-05,
+      "loss": 0.7233,
+      "step": 18178
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.585990009223129,
+      "learning_rate": 1.2771968642449544e-05,
+      "loss": 0.8148,
+      "step": 18179
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5745607145849796,
+      "learning_rate": 1.2771235494500133e-05,
+      "loss": 0.7706,
+      "step": 18180
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4354573814014318,
+      "learning_rate": 1.2770502330415712e-05,
+      "loss": 0.8047,
+      "step": 18181
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3545978425713094,
+      "learning_rate": 1.276976915020055e-05,
+      "loss": 0.7178,
+      "step": 18182
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4782409747508878,
+      "learning_rate": 1.2769035953858913e-05,
+      "loss": 0.7464,
+      "step": 18183
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.41289297659326,
+      "learning_rate": 1.2768302741395076e-05,
+      "loss": 0.6865,
+      "step": 18184
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5121826093535102,
+      "learning_rate": 1.27675695128133e-05,
+      "loss": 0.8327,
+      "step": 18185
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.577688334028521,
+      "learning_rate": 1.2766836268117862e-05,
+      "loss": 0.5239,
+      "step": 18186
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4900777679976827,
+      "learning_rate": 1.2766103007313024e-05,
+      "loss": 0.8611,
+      "step": 18187
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6519961490842874,
+      "learning_rate": 1.2765369730403061e-05,
+      "loss": 0.7066,
+      "step": 18188
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4635957794298131,
+      "learning_rate": 1.2764636437392239e-05,
+      "loss": 0.8026,
+      "step": 18189
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6018943188018286,
+      "learning_rate": 1.2763903128284827e-05,
+      "loss": 0.7355,
+      "step": 18190
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3365309918014237,
+      "learning_rate": 1.2763169803085095e-05,
+      "loss": 0.5916,
+      "step": 18191
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5297203474965133,
+      "learning_rate": 1.2762436461797318e-05,
+      "loss": 0.6872,
+      "step": 18192
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4141893736000792,
+      "learning_rate": 1.2761703104425757e-05,
+      "loss": 0.6902,
+      "step": 18193
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3913629556830485,
+      "learning_rate": 1.2760969730974692e-05,
+      "loss": 0.6383,
+      "step": 18194
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3527312933356657,
+      "learning_rate": 1.276023634144838e-05,
+      "loss": 0.7817,
+      "step": 18195
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7262828886750834,
+      "learning_rate": 1.2759502935851102e-05,
+      "loss": 0.5993,
+      "step": 18196
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5795054356060216,
+      "learning_rate": 1.2758769514187123e-05,
+      "loss": 0.6737,
+      "step": 18197
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7948456649603832,
+      "learning_rate": 1.2758036076460716e-05,
+      "loss": 0.7726,
+      "step": 18198
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.540732058264478,
+      "learning_rate": 1.2757302622676149e-05,
+      "loss": 0.8143,
+      "step": 18199
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.593610099904951,
+      "learning_rate": 1.2756569152837693e-05,
+      "loss": 0.6754,
+      "step": 18200
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4136984352604107,
+      "learning_rate": 1.275583566694962e-05,
+      "loss": 0.8319,
+      "step": 18201
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7092021835953333,
+      "learning_rate": 1.2755102165016195e-05,
+      "loss": 0.6108,
+      "step": 18202
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4258909024387014,
+      "learning_rate": 1.2754368647041699e-05,
+      "loss": 0.7115,
+      "step": 18203
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4758174792368024,
+      "learning_rate": 1.275363511303039e-05,
+      "loss": 0.8296,
+      "step": 18204
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.333310110491133,
+      "learning_rate": 1.2752901562986553e-05,
+      "loss": 0.7347,
+      "step": 18205
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6516225764672965,
+      "learning_rate": 1.2752167996914446e-05,
+      "loss": 0.5252,
+      "step": 18206
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9552668728054128,
+      "learning_rate": 1.2751434414818349e-05,
+      "loss": 0.7138,
+      "step": 18207
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.947471744245231,
+      "learning_rate": 1.2750700816702528e-05,
+      "loss": 0.6756,
+      "step": 18208
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4192627918487877,
+      "learning_rate": 1.2749967202571256e-05,
+      "loss": 0.7208,
+      "step": 18209
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3052892395798719,
+      "learning_rate": 1.2749233572428805e-05,
+      "loss": 0.7973,
+      "step": 18210
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5836912504018331,
+      "learning_rate": 1.2748499926279445e-05,
+      "loss": 0.5589,
+      "step": 18211
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7771258481629442,
+      "learning_rate": 1.2747766264127449e-05,
+      "loss": 0.6535,
+      "step": 18212
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7953770648468201,
+      "learning_rate": 1.2747032585977089e-05,
+      "loss": 0.5872,
+      "step": 18213
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.478368155715057,
+      "learning_rate": 1.2746298891832631e-05,
+      "loss": 0.7499,
+      "step": 18214
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5374228923043083,
+      "learning_rate": 1.2745565181698357e-05,
+      "loss": 0.823,
+      "step": 18215
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7101629838623296,
+      "learning_rate": 1.2744831455578528e-05,
+      "loss": 0.8691,
+      "step": 18216
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4264933176259862,
+      "learning_rate": 1.2744097713477426e-05,
+      "loss": 0.7377,
+      "step": 18217
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.066115377365812,
+      "learning_rate": 1.2743363955399315e-05,
+      "loss": 0.7508,
+      "step": 18218
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7113617683120665,
+      "learning_rate": 1.2742630181348473e-05,
+      "loss": 0.6995,
+      "step": 18219
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.512656054790035,
+      "learning_rate": 1.2741896391329167e-05,
+      "loss": 0.6779,
+      "step": 18220
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6378604081558414,
+      "learning_rate": 1.2741162585345674e-05,
+      "loss": 0.7788,
+      "step": 18221
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4410329425124848,
+      "learning_rate": 1.2740428763402266e-05,
+      "loss": 0.7555,
+      "step": 18222
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.315112110386546,
+      "learning_rate": 1.2739694925503209e-05,
+      "loss": 0.7027,
+      "step": 18223
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5872654715629987,
+      "learning_rate": 1.2738961071652787e-05,
+      "loss": 0.556,
+      "step": 18224
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4253705780410197,
+      "learning_rate": 1.2738227201855263e-05,
+      "loss": 0.7581,
+      "step": 18225
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1938218267972482,
+      "learning_rate": 1.2737493316114916e-05,
+      "loss": 0.745,
+      "step": 18226
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.595407107771416,
+      "learning_rate": 1.273675941443601e-05,
+      "loss": 0.7105,
+      "step": 18227
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6111256404650136,
+      "learning_rate": 1.2736025496822833e-05,
+      "loss": 0.7268,
+      "step": 18228
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5065721641979686,
+      "learning_rate": 1.2735291563279644e-05,
+      "loss": 0.7771,
+      "step": 18229
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3437603757056233,
+      "learning_rate": 1.2734557613810723e-05,
+      "loss": 0.7022,
+      "step": 18230
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.869790491568958,
+      "learning_rate": 1.2733823648420341e-05,
+      "loss": 0.689,
+      "step": 18231
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2028495488385293,
+      "learning_rate": 1.273308966711277e-05,
+      "loss": 0.7072,
+      "step": 18232
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9681993712523391,
+      "learning_rate": 1.2732355669892289e-05,
+      "loss": 0.6677,
+      "step": 18233
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7164409135552208,
+      "learning_rate": 1.2731621656763165e-05,
+      "loss": 0.6695,
+      "step": 18234
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4505791319561883,
+      "learning_rate": 1.273088762772968e-05,
+      "loss": 0.8036,
+      "step": 18235
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8022732328920703,
+      "learning_rate": 1.2730153582796097e-05,
+      "loss": 0.7056,
+      "step": 18236
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4825098569825383,
+      "learning_rate": 1.2729419521966702e-05,
+      "loss": 0.7401,
+      "step": 18237
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6148623176083373,
+      "learning_rate": 1.2728685445245757e-05,
+      "loss": 0.4805,
+      "step": 18238
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6217488144743702,
+      "learning_rate": 1.2727951352637544e-05,
+      "loss": 0.5469,
+      "step": 18239
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3331322380698651,
+      "learning_rate": 1.2727217244146334e-05,
+      "loss": 0.6807,
+      "step": 18240
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.938109105529572,
+      "learning_rate": 1.2726483119776401e-05,
+      "loss": 0.7304,
+      "step": 18241
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4248554447971329,
+      "learning_rate": 1.272574897953202e-05,
+      "loss": 0.8175,
+      "step": 18242
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4104666635519307,
+      "learning_rate": 1.2725014823417468e-05,
+      "loss": 0.7225,
+      "step": 18243
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4184991137927139,
+      "learning_rate": 1.2724280651437014e-05,
+      "loss": 0.7163,
+      "step": 18244
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3817236178600303,
+      "learning_rate": 1.2723546463594936e-05,
+      "loss": 0.6516,
+      "step": 18245
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7282664297943766,
+      "learning_rate": 1.272281225989551e-05,
+      "loss": 0.7728,
+      "step": 18246
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.533644197544143,
+      "learning_rate": 1.2722078040343007e-05,
+      "loss": 0.8133,
+      "step": 18247
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4596340474019556,
+      "learning_rate": 1.2721343804941704e-05,
+      "loss": 0.7026,
+      "step": 18248
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5028470816642707,
+      "learning_rate": 1.2720609553695877e-05,
+      "loss": 0.7247,
+      "step": 18249
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5354004365818958,
+      "learning_rate": 1.27198752866098e-05,
+      "loss": 0.7602,
+      "step": 18250
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.508084399170261,
+      "learning_rate": 1.2719141003687746e-05,
+      "loss": 0.7887,
+      "step": 18251
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.321190913165869,
+      "learning_rate": 1.2718406704933997e-05,
+      "loss": 0.7498,
+      "step": 18252
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5942383958040989,
+      "learning_rate": 1.2717672390352818e-05,
+      "loss": 0.6636,
+      "step": 18253
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6297962038643572,
+      "learning_rate": 1.2716938059948493e-05,
+      "loss": 0.6577,
+      "step": 18254
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.95875383775253,
+      "learning_rate": 1.2716203713725293e-05,
+      "loss": 0.7046,
+      "step": 18255
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4995025945194949,
+      "learning_rate": 1.2715469351687498e-05,
+      "loss": 0.75,
+      "step": 18256
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3823929189438933,
+      "learning_rate": 1.2714734973839375e-05,
+      "loss": 0.7419,
+      "step": 18257
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3827972716029064,
+      "learning_rate": 1.2714000580185211e-05,
+      "loss": 0.6498,
+      "step": 18258
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3726282823422076,
+      "learning_rate": 1.2713266170729275e-05,
+      "loss": 0.7896,
+      "step": 18259
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6230877672788933,
+      "learning_rate": 1.2712531745475842e-05,
+      "loss": 0.6528,
+      "step": 18260
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5795091756525376,
+      "learning_rate": 1.2711797304429192e-05,
+      "loss": 0.5442,
+      "step": 18261
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6236338602801978,
+      "learning_rate": 1.27110628475936e-05,
+      "loss": 0.7185,
+      "step": 18262
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4932369326627872,
+      "learning_rate": 1.2710328374973342e-05,
+      "loss": 0.7252,
+      "step": 18263
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4695667021587995,
+      "learning_rate": 1.2709593886572693e-05,
+      "loss": 0.7054,
+      "step": 18264
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3737061066923848,
+      "learning_rate": 1.270885938239593e-05,
+      "loss": 0.7452,
+      "step": 18265
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5824695256664285,
+      "learning_rate": 1.2708124862447331e-05,
+      "loss": 0.7206,
+      "step": 18266
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.343707801187349,
+      "learning_rate": 1.2707390326731173e-05,
+      "loss": 0.7251,
+      "step": 18267
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9118457190327205,
+      "learning_rate": 1.2706655775251728e-05,
+      "loss": 0.5879,
+      "step": 18268
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6191876132336662,
+      "learning_rate": 1.270592120801328e-05,
+      "loss": 0.7402,
+      "step": 18269
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3248683935762378,
+      "learning_rate": 1.2705186625020098e-05,
+      "loss": 0.769,
+      "step": 18270
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.590852376478142,
+      "learning_rate": 1.2704452026276468e-05,
+      "loss": 0.5205,
+      "step": 18271
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7493556859127721,
+      "learning_rate": 1.2703717411786658e-05,
+      "loss": 0.7476,
+      "step": 18272
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3026605367265942,
+      "learning_rate": 1.2702982781554953e-05,
+      "loss": 0.6963,
+      "step": 18273
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3524631840676236,
+      "learning_rate": 1.2702248135585622e-05,
+      "loss": 0.8311,
+      "step": 18274
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6563835212305167,
+      "learning_rate": 1.2701513473882951e-05,
+      "loss": 0.6934,
+      "step": 18275
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5758076448489375,
+      "learning_rate": 1.270077879645121e-05,
+      "loss": 0.5292,
+      "step": 18276
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6032619609276901,
+      "learning_rate": 1.2700044103294683e-05,
+      "loss": 0.664,
+      "step": 18277
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6000421272256466,
+      "learning_rate": 1.2699309394417641e-05,
+      "loss": 0.5459,
+      "step": 18278
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2062279886811331,
+      "learning_rate": 1.2698574669824366e-05,
+      "loss": 0.6727,
+      "step": 18279
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6881302160026952,
+      "learning_rate": 1.2697839929519136e-05,
+      "loss": 0.671,
+      "step": 18280
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7766527434673247,
+      "learning_rate": 1.2697105173506225e-05,
+      "loss": 0.7232,
+      "step": 18281
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5821355118840845,
+      "learning_rate": 1.269637040178992e-05,
+      "loss": 0.7508,
+      "step": 18282
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3401378344774275,
+      "learning_rate": 1.2695635614374484e-05,
+      "loss": 0.7331,
+      "step": 18283
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2428560033176634,
+      "learning_rate": 1.2694900811264212e-05,
+      "loss": 0.7079,
+      "step": 18284
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7032417814375918,
+      "learning_rate": 1.269416599246337e-05,
+      "loss": 0.7466,
+      "step": 18285
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.439668201168845,
+      "learning_rate": 1.2693431157976243e-05,
+      "loss": 0.7003,
+      "step": 18286
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2584600982596654,
+      "learning_rate": 1.2692696307807106e-05,
+      "loss": 0.8283,
+      "step": 18287
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4629753971796153,
+      "learning_rate": 1.2691961441960238e-05,
+      "loss": 0.7627,
+      "step": 18288
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7415862235938908,
+      "learning_rate": 1.269122656043992e-05,
+      "loss": 0.726,
+      "step": 18289
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5771762868969708,
+      "learning_rate": 1.2690491663250428e-05,
+      "loss": 0.7747,
+      "step": 18290
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2858002676676388,
+      "learning_rate": 1.268975675039604e-05,
+      "loss": 0.7369,
+      "step": 18291
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5523527069267904,
+      "learning_rate": 1.268902182188104e-05,
+      "loss": 0.805,
+      "step": 18292
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6848536120491988,
+      "learning_rate": 1.26882868777097e-05,
+      "loss": 0.7353,
+      "step": 18293
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.92173141549714,
+      "learning_rate": 1.2687551917886304e-05,
+      "loss": 0.711,
+      "step": 18294
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4554054222277488,
+      "learning_rate": 1.2686816942415133e-05,
+      "loss": 0.663,
+      "step": 18295
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6605363804369029,
+      "learning_rate": 1.268608195130046e-05,
+      "loss": 0.7245,
+      "step": 18296
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.619597619339043,
+      "learning_rate": 1.2685346944546566e-05,
+      "loss": 0.5147,
+      "step": 18297
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5889269943508157,
+      "learning_rate": 1.2684611922157735e-05,
+      "loss": 0.515,
+      "step": 18298
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9283918868017136,
+      "learning_rate": 1.2683876884138242e-05,
+      "loss": 0.9117,
+      "step": 18299
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6857612143599114,
+      "learning_rate": 1.2683141830492368e-05,
+      "loss": 0.6921,
+      "step": 18300
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4268425756812448,
+      "learning_rate": 1.2682406761224394e-05,
+      "loss": 0.7212,
+      "step": 18301
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4609775689917075,
+      "learning_rate": 1.2681671676338597e-05,
+      "loss": 0.8284,
+      "step": 18302
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3219306120390903,
+      "learning_rate": 1.268093657583926e-05,
+      "loss": 0.6634,
+      "step": 18303
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6228734491496386,
+      "learning_rate": 1.2680201459730661e-05,
+      "loss": 0.7022,
+      "step": 18304
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4705455703630894,
+      "learning_rate": 1.267946632801708e-05,
+      "loss": 0.7185,
+      "step": 18305
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.612228800770022,
+      "learning_rate": 1.2678731180702796e-05,
+      "loss": 0.6619,
+      "step": 18306
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3217665983121936,
+      "learning_rate": 1.2677996017792095e-05,
+      "loss": 0.7857,
+      "step": 18307
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6788220793197015,
+      "learning_rate": 1.2677260839289249e-05,
+      "loss": 0.6744,
+      "step": 18308
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.186930352010777,
+      "learning_rate": 1.267652564519855e-05,
+      "loss": 0.7247,
+      "step": 18309
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4088467591840197,
+      "learning_rate": 1.2675790435524264e-05,
+      "loss": 0.7754,
+      "step": 18310
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3725598716233716,
+      "learning_rate": 1.2675055210270681e-05,
+      "loss": 0.8935,
+      "step": 18311
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6330580405395615,
+      "learning_rate": 1.2674319969442082e-05,
+      "loss": 0.5039,
+      "step": 18312
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5798281312195339,
+      "learning_rate": 1.2673584713042743e-05,
+      "loss": 0.5428,
+      "step": 18313
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5764872648127909,
+      "learning_rate": 1.2672849441076948e-05,
+      "loss": 0.516,
+      "step": 18314
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5650194213177835,
+      "learning_rate": 1.2672114153548977e-05,
+      "loss": 0.6713,
+      "step": 18315
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3075908461640595,
+      "learning_rate": 1.2671378850463112e-05,
+      "loss": 0.699,
+      "step": 18316
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.439690665171221,
+      "learning_rate": 1.2670643531823634e-05,
+      "loss": 0.6529,
+      "step": 18317
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3181848497239996,
+      "learning_rate": 1.2669908197634822e-05,
+      "loss": 0.616,
+      "step": 18318
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5279908875045376,
+      "learning_rate": 1.2669172847900958e-05,
+      "loss": 0.691,
+      "step": 18319
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.265497642496662,
+      "learning_rate": 1.266843748262633e-05,
+      "loss": 0.8749,
+      "step": 18320
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4216720728969825,
+      "learning_rate": 1.266770210181521e-05,
+      "loss": 0.72,
+      "step": 18321
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3190739441757364,
+      "learning_rate": 1.2666966705471886e-05,
+      "loss": 0.6683,
+      "step": 18322
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.399840863701596,
+      "learning_rate": 1.2666231293600634e-05,
+      "loss": 0.6324,
+      "step": 18323
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3362562003723435,
+      "learning_rate": 1.2665495866205745e-05,
+      "loss": 0.7101,
+      "step": 18324
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5401719876601716,
+      "learning_rate": 1.2664760423291492e-05,
+      "loss": 0.7719,
+      "step": 18325
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3707239948295338,
+      "learning_rate": 1.266402496486216e-05,
+      "loss": 0.665,
+      "step": 18326
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4888538036453443,
+      "learning_rate": 1.2663289490922029e-05,
+      "loss": 0.8222,
+      "step": 18327
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6049280315763268,
+      "learning_rate": 1.266255400147539e-05,
+      "loss": 0.5522,
+      "step": 18328
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5423875904547402,
+      "learning_rate": 1.266181849652651e-05,
+      "loss": 0.8381,
+      "step": 18329
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.471986897870071,
+      "learning_rate": 1.2661082976079684e-05,
+      "loss": 0.7544,
+      "step": 18330
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4890405975387444,
+      "learning_rate": 1.2660347440139194e-05,
+      "loss": 0.6295,
+      "step": 18331
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3199193116352892,
+      "learning_rate": 1.2659611888709313e-05,
+      "loss": 0.6961,
+      "step": 18332
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4948340140841718,
+      "learning_rate": 1.2658876321794333e-05,
+      "loss": 0.7514,
+      "step": 18333
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6119319999657616,
+      "learning_rate": 1.2658140739398532e-05,
+      "loss": 0.5431,
+      "step": 18334
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7174012520560302,
+      "learning_rate": 1.2657405141526196e-05,
+      "loss": 0.7687,
+      "step": 18335
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.1717027800538458,
+      "learning_rate": 1.2656669528181603e-05,
+      "loss": 0.8032,
+      "step": 18336
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7061119444941777,
+      "learning_rate": 1.2655933899369042e-05,
+      "loss": 0.8367,
+      "step": 18337
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4861810853944215,
+      "learning_rate": 1.2655198255092791e-05,
+      "loss": 0.8158,
+      "step": 18338
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.313918805210674,
+      "learning_rate": 1.2654462595357136e-05,
+      "loss": 0.7976,
+      "step": 18339
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4691173902819075,
+      "learning_rate": 1.2653726920166358e-05,
+      "loss": 0.782,
+      "step": 18340
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7844619688837051,
+      "learning_rate": 1.2652991229524742e-05,
+      "loss": 0.7579,
+      "step": 18341
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7443763589306067,
+      "learning_rate": 1.2652255523436572e-05,
+      "loss": 0.7719,
+      "step": 18342
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4220053965693753,
+      "learning_rate": 1.2651519801906131e-05,
+      "loss": 0.6805,
+      "step": 18343
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7437199513569437,
+      "learning_rate": 1.2650784064937702e-05,
+      "loss": 0.7805,
+      "step": 18344
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3590977659101477,
+      "learning_rate": 1.265004831253557e-05,
+      "loss": 0.8988,
+      "step": 18345
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5584913762953594,
+      "learning_rate": 1.2649312544704016e-05,
+      "loss": 0.6459,
+      "step": 18346
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8529213937869526,
+      "learning_rate": 1.2648576761447324e-05,
+      "loss": 0.7108,
+      "step": 18347
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5495067196026713,
+      "learning_rate": 1.2647840962769781e-05,
+      "loss": 0.6521,
+      "step": 18348
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5583803617626086,
+      "learning_rate": 1.2647105148675671e-05,
+      "loss": 0.7136,
+      "step": 18349
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4327085510210067,
+      "learning_rate": 1.2646369319169275e-05,
+      "loss": 0.7165,
+      "step": 18350
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7161820371465926,
+      "learning_rate": 1.2645633474254882e-05,
+      "loss": 0.8039,
+      "step": 18351
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.7722812002237593,
+      "learning_rate": 1.2644897613936773e-05,
+      "loss": 0.7054,
+      "step": 18352
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7533773537891901,
+      "learning_rate": 1.2644161738219229e-05,
+      "loss": 0.7372,
+      "step": 18353
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4908437999645001,
+      "learning_rate": 1.2643425847106543e-05,
+      "loss": 0.7272,
+      "step": 18354
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4694518530787937,
+      "learning_rate": 1.264268994060299e-05,
+      "loss": 0.7269,
+      "step": 18355
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.736274411651394,
+      "learning_rate": 1.2641954018712863e-05,
+      "loss": 0.834,
+      "step": 18356
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.455402325019713,
+      "learning_rate": 1.264121808144044e-05,
+      "loss": 0.8052,
+      "step": 18357
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7972303862107437,
+      "learning_rate": 1.2640482128790012e-05,
+      "loss": 0.7402,
+      "step": 18358
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4512618889598452,
+      "learning_rate": 1.2639746160765858e-05,
+      "loss": 0.8024,
+      "step": 18359
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3740878484359405,
+      "learning_rate": 1.2639010177372271e-05,
+      "loss": 0.7249,
+      "step": 18360
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.65818005181303,
+      "learning_rate": 1.2638274178613527e-05,
+      "loss": 0.6331,
+      "step": 18361
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5647082851355647,
+      "learning_rate": 1.2637538164493916e-05,
+      "loss": 0.8042,
+      "step": 18362
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6013610083133859,
+      "learning_rate": 1.2636802135017726e-05,
+      "loss": 0.5026,
+      "step": 18363
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4600425988821408,
+      "learning_rate": 1.2636066090189237e-05,
+      "loss": 0.7976,
+      "step": 18364
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4331566340376012,
+      "learning_rate": 1.2635330030012735e-05,
+      "loss": 0.769,
+      "step": 18365
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5763527283037806,
+      "learning_rate": 1.2634593954492509e-05,
+      "loss": 0.5763,
+      "step": 18366
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.1190279291910725,
+      "learning_rate": 1.2633857863632843e-05,
+      "loss": 0.78,
+      "step": 18367
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6409560885058563,
+      "learning_rate": 1.2633121757438022e-05,
+      "loss": 0.793,
+      "step": 18368
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7585766094394193,
+      "learning_rate": 1.2632385635912332e-05,
+      "loss": 0.7691,
+      "step": 18369
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.696699734686481,
+      "learning_rate": 1.2631649499060061e-05,
+      "loss": 0.6959,
+      "step": 18370
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3508334385279221,
+      "learning_rate": 1.2630913346885491e-05,
+      "loss": 0.6949,
+      "step": 18371
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.511237423267216,
+      "learning_rate": 1.2630177179392913e-05,
+      "loss": 0.6877,
+      "step": 18372
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5299415211679153,
+      "learning_rate": 1.262944099658661e-05,
+      "loss": 0.8329,
+      "step": 18373
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.892278167443152,
+      "learning_rate": 1.262870479847087e-05,
+      "loss": 0.8213,
+      "step": 18374
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4674581676524396,
+      "learning_rate": 1.2627968585049979e-05,
+      "loss": 0.8637,
+      "step": 18375
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5947797587833792,
+      "learning_rate": 1.262723235632822e-05,
+      "loss": 0.5113,
+      "step": 18376
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3030889031402435,
+      "learning_rate": 1.2626496112309885e-05,
+      "loss": 0.7383,
+      "step": 18377
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4479519603246878,
+      "learning_rate": 1.2625759852999256e-05,
+      "loss": 0.8196,
+      "step": 18378
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3200854161365034,
+      "learning_rate": 1.2625023578400622e-05,
+      "loss": 0.582,
+      "step": 18379
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4067658352702213,
+      "learning_rate": 1.2624287288518272e-05,
+      "loss": 0.7585,
+      "step": 18380
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4442551534163581,
+      "learning_rate": 1.2623550983356487e-05,
+      "loss": 0.7945,
+      "step": 18381
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3979426528655388,
+      "learning_rate": 1.2622814662919562e-05,
+      "loss": 0.6064,
+      "step": 18382
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5879677518488273,
+      "learning_rate": 1.2622078327211776e-05,
+      "loss": 0.7327,
+      "step": 18383
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5191374447685493,
+      "learning_rate": 1.2621341976237423e-05,
+      "loss": 0.8099,
+      "step": 18384
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.327605898351801,
+      "learning_rate": 1.2620605610000782e-05,
+      "loss": 0.8284,
+      "step": 18385
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.482499266293183,
+      "learning_rate": 1.2619869228506151e-05,
+      "loss": 0.7396,
+      "step": 18386
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5381053692956383,
+      "learning_rate": 1.2619132831757808e-05,
+      "loss": 0.716,
+      "step": 18387
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3282164279895703,
+      "learning_rate": 1.2618396419760049e-05,
+      "loss": 0.7497,
+      "step": 18388
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5261902953653537,
+      "learning_rate": 1.2617659992517153e-05,
+      "loss": 0.8393,
+      "step": 18389
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4293627045938773,
+      "learning_rate": 1.2616923550033413e-05,
+      "loss": 0.757,
+      "step": 18390
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5954436355350878,
+      "learning_rate": 1.2616187092313114e-05,
+      "loss": 0.6868,
+      "step": 18391
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8689754968487862,
+      "learning_rate": 1.2615450619360547e-05,
+      "loss": 0.7129,
+      "step": 18392
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.4619645207666325,
+      "learning_rate": 1.2614714131179999e-05,
+      "loss": 0.7768,
+      "step": 18393
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5013668181628397,
+      "learning_rate": 1.2613977627775757e-05,
+      "loss": 0.6992,
+      "step": 18394
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4929536479952044,
+      "learning_rate": 1.261324110915211e-05,
+      "loss": 0.6997,
+      "step": 18395
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4791933989519748,
+      "learning_rate": 1.2612504575313344e-05,
+      "loss": 0.6944,
+      "step": 18396
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.610486781948954,
+      "learning_rate": 1.2611768026263752e-05,
+      "loss": 0.5339,
+      "step": 18397
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8892819773914222,
+      "learning_rate": 1.2611031462007619e-05,
+      "loss": 0.7166,
+      "step": 18398
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4718783053993099,
+      "learning_rate": 1.2610294882549232e-05,
+      "loss": 0.569,
+      "step": 18399
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7494700566949846,
+      "learning_rate": 1.2609558287892884e-05,
+      "loss": 0.7713,
+      "step": 18400
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.431751508195031,
+      "learning_rate": 1.2608821678042862e-05,
+      "loss": 0.7703,
+      "step": 18401
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4420414923139406,
+      "learning_rate": 1.2608085053003452e-05,
+      "loss": 0.7546,
+      "step": 18402
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4786594800616626,
+      "learning_rate": 1.2607348412778949e-05,
+      "loss": 0.6549,
+      "step": 18403
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6136497725784125,
+      "learning_rate": 1.2606611757373633e-05,
+      "loss": 0.5353,
+      "step": 18404
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4595985583330593,
+      "learning_rate": 1.2605875086791804e-05,
+      "loss": 0.8457,
+      "step": 18405
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5667212710606399,
+      "learning_rate": 1.2605138401037738e-05,
+      "loss": 0.5037,
+      "step": 18406
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.818586360496744,
+      "learning_rate": 1.2604401700115738e-05,
+      "loss": 0.6782,
+      "step": 18407
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4907910217175218,
+      "learning_rate": 1.2603664984030082e-05,
+      "loss": 0.8252,
+      "step": 18408
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3515666602169214,
+      "learning_rate": 1.2602928252785068e-05,
+      "loss": 0.705,
+      "step": 18409
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7599758787725324,
+      "learning_rate": 1.260219150638498e-05,
+      "loss": 0.7372,
+      "step": 18410
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3722584462872218,
+      "learning_rate": 1.2601454744834108e-05,
+      "loss": 0.793,
+      "step": 18411
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1852158567092939,
+      "learning_rate": 1.2600717968136742e-05,
+      "loss": 0.6208,
+      "step": 18412
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5277373107613965,
+      "learning_rate": 1.2599981176297176e-05,
+      "loss": 0.6966,
+      "step": 18413
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.280394842388331,
+      "learning_rate": 1.2599244369319694e-05,
+      "loss": 0.703,
+      "step": 18414
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6135918463944987,
+      "learning_rate": 1.2598507547208589e-05,
+      "loss": 0.5265,
+      "step": 18415
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.149022459427535,
+      "learning_rate": 1.259777070996815e-05,
+      "loss": 0.6137,
+      "step": 18416
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5226559023617567,
+      "learning_rate": 1.259703385760267e-05,
+      "loss": 0.7273,
+      "step": 18417
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5912706060073074,
+      "learning_rate": 1.2596296990116435e-05,
+      "loss": 0.6251,
+      "step": 18418
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9392161903477987,
+      "learning_rate": 1.2595560107513735e-05,
+      "loss": 0.7302,
+      "step": 18419
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6208038582297257,
+      "learning_rate": 1.2594823209798865e-05,
+      "loss": 0.5369,
+      "step": 18420
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.69306415429756,
+      "learning_rate": 1.2594086296976111e-05,
+      "loss": 0.6625,
+      "step": 18421
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8260081064770197,
+      "learning_rate": 1.2593349369049767e-05,
+      "loss": 0.7442,
+      "step": 18422
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.460005286589313,
+      "learning_rate": 1.259261242602412e-05,
+      "loss": 0.6845,
+      "step": 18423
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2952741110529284,
+      "learning_rate": 1.2591875467903465e-05,
+      "loss": 0.709,
+      "step": 18424
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4099317740487134,
+      "learning_rate": 1.2591138494692087e-05,
+      "loss": 0.6205,
+      "step": 18425
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0179556527260893,
+      "learning_rate": 1.2590401506394285e-05,
+      "loss": 0.7466,
+      "step": 18426
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6223049604926365,
+      "learning_rate": 1.258966450301434e-05,
+      "loss": 0.5205,
+      "step": 18427
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8927724761871143,
+      "learning_rate": 1.2588927484556552e-05,
+      "loss": 0.7592,
+      "step": 18428
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2805394253989875,
+      "learning_rate": 1.2588190451025209e-05,
+      "loss": 0.7008,
+      "step": 18429
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.389921912315184,
+      "learning_rate": 1.2587453402424598e-05,
+      "loss": 0.8236,
+      "step": 18430
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8285390229627059,
+      "learning_rate": 1.2586716338759018e-05,
+      "loss": 0.8699,
+      "step": 18431
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4145290718866865,
+      "learning_rate": 1.2585979260032754e-05,
+      "loss": 0.8787,
+      "step": 18432
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9324381663786923,
+      "learning_rate": 1.2585242166250105e-05,
+      "loss": 0.6477,
+      "step": 18433
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6098449246412445,
+      "learning_rate": 1.2584505057415352e-05,
+      "loss": 0.7393,
+      "step": 18434
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5128257357698338,
+      "learning_rate": 1.2583767933532795e-05,
+      "loss": 0.5087,
+      "step": 18435
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.4965409993399765,
+      "learning_rate": 1.258303079460672e-05,
+      "loss": 0.7758,
+      "step": 18436
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.436017831561995,
+      "learning_rate": 1.2582293640641427e-05,
+      "loss": 0.7586,
+      "step": 18437
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4448331977216995,
+      "learning_rate": 1.2581556471641199e-05,
+      "loss": 0.7522,
+      "step": 18438
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.37139605587389,
+      "learning_rate": 1.2580819287610332e-05,
+      "loss": 0.7211,
+      "step": 18439
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3302797096344643,
+      "learning_rate": 1.2580082088553121e-05,
+      "loss": 0.6291,
+      "step": 18440
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3612050439578376,
+      "learning_rate": 1.2579344874473854e-05,
+      "loss": 0.6806,
+      "step": 18441
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6519504186428267,
+      "learning_rate": 1.2578607645376823e-05,
+      "loss": 0.6504,
+      "step": 18442
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4573721572256297,
+      "learning_rate": 1.2577870401266324e-05,
+      "loss": 0.6939,
+      "step": 18443
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.507237492539536,
+      "learning_rate": 1.2577133142146646e-05,
+      "loss": 0.695,
+      "step": 18444
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5503454312875478,
+      "learning_rate": 1.2576395868022082e-05,
+      "loss": 0.486,
+      "step": 18445
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3692303002353055,
+      "learning_rate": 1.2575658578896928e-05,
+      "loss": 0.7652,
+      "step": 18446
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.353050039044511,
+      "learning_rate": 1.2574921274775472e-05,
+      "loss": 0.7776,
+      "step": 18447
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7921706389082988,
+      "learning_rate": 1.257418395566201e-05,
+      "loss": 0.7789,
+      "step": 18448
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3103334999946026,
+      "learning_rate": 1.2573446621560833e-05,
+      "loss": 0.7789,
+      "step": 18449
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4484886021514247,
+      "learning_rate": 1.2572709272476237e-05,
+      "loss": 0.7549,
+      "step": 18450
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.699010011090285,
+      "learning_rate": 1.2571971908412512e-05,
+      "loss": 0.6676,
+      "step": 18451
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6362670699923829,
+      "learning_rate": 1.2571234529373953e-05,
+      "loss": 0.5288,
+      "step": 18452
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5120109738208964,
+      "learning_rate": 1.2570497135364851e-05,
+      "loss": 0.6899,
+      "step": 18453
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3976682191782068,
+      "learning_rate": 1.2569759726389505e-05,
+      "loss": 0.6465,
+      "step": 18454
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6278253245586988,
+      "learning_rate": 1.2569022302452199e-05,
+      "loss": 0.5045,
+      "step": 18455
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5817407594396546,
+      "learning_rate": 1.2568284863557236e-05,
+      "loss": 0.5177,
+      "step": 18456
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7210674626790496,
+      "learning_rate": 1.25675474097089e-05,
+      "loss": 0.6735,
+      "step": 18457
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6317708009020649,
+      "learning_rate": 1.2566809940911497e-05,
+      "loss": 0.5449,
+      "step": 18458
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.792059320066444,
+      "learning_rate": 1.2566072457169308e-05,
+      "loss": 0.8075,
+      "step": 18459
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5375414631667264,
+      "learning_rate": 1.2565334958486636e-05,
+      "loss": 0.6602,
+      "step": 18460
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6093969389739426,
+      "learning_rate": 1.2564597444867772e-05,
+      "loss": 0.4902,
+      "step": 18461
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5888299264301842,
+      "learning_rate": 1.256385991631701e-05,
+      "loss": 0.7692,
+      "step": 18462
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4201841305465497,
+      "learning_rate": 1.2563122372838642e-05,
+      "loss": 0.7006,
+      "step": 18463
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.709992597442837,
+      "learning_rate": 1.2562384814436961e-05,
+      "loss": 0.6069,
+      "step": 18464
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5241418535382316,
+      "learning_rate": 1.256164724111627e-05,
+      "loss": 0.521,
+      "step": 18465
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7107615259189908,
+      "learning_rate": 1.2560909652880854e-05,
+      "loss": 0.7298,
+      "step": 18466
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4320977301513471,
+      "learning_rate": 1.2560172049735016e-05,
+      "loss": 0.6363,
+      "step": 18467
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5692407379581659,
+      "learning_rate": 1.2559434431683042e-05,
+      "loss": 0.5203,
+      "step": 18468
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5974017564437195,
+      "learning_rate": 1.2558696798729231e-05,
+      "loss": 0.7322,
+      "step": 18469
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3714376420287622,
+      "learning_rate": 1.255795915087788e-05,
+      "loss": 0.709,
+      "step": 18470
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.492748363338846,
+      "learning_rate": 1.2557221488133276e-05,
+      "loss": 0.7042,
+      "step": 18471
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4835365439599557,
+      "learning_rate": 1.255648381049972e-05,
+      "loss": 0.687,
+      "step": 18472
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3830887298468157,
+      "learning_rate": 1.2555746117981506e-05,
+      "loss": 0.5746,
+      "step": 18473
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5512648176167252,
+      "learning_rate": 1.2555008410582932e-05,
+      "loss": 0.7695,
+      "step": 18474
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8167760990479072,
+      "learning_rate": 1.255427068830829e-05,
+      "loss": 0.6948,
+      "step": 18475
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4481606126188755,
+      "learning_rate": 1.255353295116187e-05,
+      "loss": 0.7603,
+      "step": 18476
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4194943977149426,
+      "learning_rate": 1.2552795199147979e-05,
+      "loss": 0.7153,
+      "step": 18477
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4830481441503758,
+      "learning_rate": 1.2552057432270903e-05,
+      "loss": 0.6958,
+      "step": 18478
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4787845818504766,
+      "learning_rate": 1.255131965053494e-05,
+      "loss": 0.7832,
+      "step": 18479
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3354665582606022,
+      "learning_rate": 1.2550581853944388e-05,
+      "loss": 0.606,
+      "step": 18480
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6007479972736979,
+      "learning_rate": 1.254984404250354e-05,
+      "loss": 0.5452,
+      "step": 18481
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.502081908257902,
+      "learning_rate": 1.2549106216216693e-05,
+      "loss": 0.808,
+      "step": 18482
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.465910439497031,
+      "learning_rate": 1.2548368375088143e-05,
+      "loss": 0.8147,
+      "step": 18483
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3975007230756944,
+      "learning_rate": 1.2547630519122187e-05,
+      "loss": 0.7431,
+      "step": 18484
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.641540600850603,
+      "learning_rate": 1.2546892648323114e-05,
+      "loss": 0.8033,
+      "step": 18485
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4659957448984944,
+      "learning_rate": 1.2546154762695233e-05,
+      "loss": 0.7665,
+      "step": 18486
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3269874278684002,
+      "learning_rate": 1.2545416862242827e-05,
+      "loss": 0.8104,
+      "step": 18487
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6649847928625041,
+      "learning_rate": 1.2544678946970201e-05,
+      "loss": 0.655,
+      "step": 18488
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.432043638627126,
+      "learning_rate": 1.2543941016881647e-05,
+      "loss": 0.8186,
+      "step": 18489
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5053069808806152,
+      "learning_rate": 1.2543203071981467e-05,
+      "loss": 0.6736,
+      "step": 18490
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.580627015811951,
+      "learning_rate": 1.254246511227395e-05,
+      "loss": 0.6724,
+      "step": 18491
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6059804670309108,
+      "learning_rate": 1.2541727137763395e-05,
+      "loss": 0.6266,
+      "step": 18492
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3191913374762825,
+      "learning_rate": 1.2540989148454102e-05,
+      "loss": 0.7082,
+      "step": 18493
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3830778640979295,
+      "learning_rate": 1.2540251144350366e-05,
+      "loss": 0.6453,
+      "step": 18494
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.474219044451101,
+      "learning_rate": 1.2539513125456484e-05,
+      "loss": 0.8556,
+      "step": 18495
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7469454780225147,
+      "learning_rate": 1.253877509177675e-05,
+      "loss": 0.6847,
+      "step": 18496
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3531744339199505,
+      "learning_rate": 1.2538037043315465e-05,
+      "loss": 0.67,
+      "step": 18497
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6075466342134166,
+      "learning_rate": 1.2537298980076926e-05,
+      "loss": 0.7766,
+      "step": 18498
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.40468512368462,
+      "learning_rate": 1.2536560902065427e-05,
+      "loss": 0.7457,
+      "step": 18499
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.09689011745446,
+      "learning_rate": 1.253582280928527e-05,
+      "loss": 0.6854,
+      "step": 18500
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.427181121475048,
+      "learning_rate": 1.2535084701740748e-05,
+      "loss": 0.8335,
+      "step": 18501
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6020204484116325,
+      "learning_rate": 1.2534346579436158e-05,
+      "loss": 0.6215,
+      "step": 18502
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4286789561385072,
+      "learning_rate": 1.2533608442375803e-05,
+      "loss": 0.69,
+      "step": 18503
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.167917439137074,
+      "learning_rate": 1.2532870290563976e-05,
+      "loss": 0.7899,
+      "step": 18504
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2550475011433513,
+      "learning_rate": 1.253213212400498e-05,
+      "loss": 0.7805,
+      "step": 18505
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3365799676660997,
+      "learning_rate": 1.2531393942703104e-05,
+      "loss": 0.7362,
+      "step": 18506
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.528124617360454,
+      "learning_rate": 1.2530655746662657e-05,
+      "loss": 0.706,
+      "step": 18507
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.267367596293451,
+      "learning_rate": 1.2529917535887923e-05,
+      "loss": 0.8578,
+      "step": 18508
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3367687051722217,
+      "learning_rate": 1.2529179310383215e-05,
+      "loss": 0.8193,
+      "step": 18509
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4777051670240702,
+      "learning_rate": 1.252844107015282e-05,
+      "loss": 0.7426,
+      "step": 18510
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6629540565810871,
+      "learning_rate": 1.2527702815201043e-05,
+      "loss": 0.5275,
+      "step": 18511
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5528503259308537,
+      "learning_rate": 1.2526964545532182e-05,
+      "loss": 0.8164,
+      "step": 18512
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6270170986879422,
+      "learning_rate": 1.2526226261150527e-05,
+      "loss": 0.5393,
+      "step": 18513
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3515555860588566,
+      "learning_rate": 1.2525487962060389e-05,
+      "loss": 0.747,
+      "step": 18514
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6654674272882952,
+      "learning_rate": 1.2524749648266054e-05,
+      "loss": 0.5839,
+      "step": 18515
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4525939406980286,
+      "learning_rate": 1.2524011319771833e-05,
+      "loss": 0.6431,
+      "step": 18516
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.491353392276304,
+      "learning_rate": 1.2523272976582018e-05,
+      "loss": 0.7388,
+      "step": 18517
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4060641543345975,
+      "learning_rate": 1.252253461870091e-05,
+      "loss": 0.8622,
+      "step": 18518
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.417924029309093,
+      "learning_rate": 1.2521796246132804e-05,
+      "loss": 0.5755,
+      "step": 18519
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2957858154354007,
+      "learning_rate": 1.2521057858882002e-05,
+      "loss": 0.7277,
+      "step": 18520
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5949761242613669,
+      "learning_rate": 1.2520319456952806e-05,
+      "loss": 0.5242,
+      "step": 18521
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.377812512474391,
+      "learning_rate": 1.2519581040349508e-05,
+      "loss": 0.7489,
+      "step": 18522
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.601333115141561,
+      "learning_rate": 1.2518842609076414e-05,
+      "loss": 0.6163,
+      "step": 18523
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3859452366491183,
+      "learning_rate": 1.2518104163137821e-05,
+      "loss": 0.7781,
+      "step": 18524
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3038247371325906,
+      "learning_rate": 1.2517365702538026e-05,
+      "loss": 0.7617,
+      "step": 18525
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.579382296847555,
+      "learning_rate": 1.2516627227281336e-05,
+      "loss": 0.5233,
+      "step": 18526
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.787308165929174,
+      "learning_rate": 1.2515888737372042e-05,
+      "loss": 0.7323,
+      "step": 18527
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4485122942594304,
+      "learning_rate": 1.2515150232814446e-05,
+      "loss": 0.7944,
+      "step": 18528
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9208152404775298,
+      "learning_rate": 1.251441171361285e-05,
+      "loss": 0.8122,
+      "step": 18529
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7692389502227703,
+      "learning_rate": 1.2513673179771555e-05,
+      "loss": 0.6258,
+      "step": 18530
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.557294878219717,
+      "learning_rate": 1.2512934631294858e-05,
+      "loss": 0.7103,
+      "step": 18531
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5074199329314533,
+      "learning_rate": 1.251219606818706e-05,
+      "loss": 0.5818,
+      "step": 18532
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7187333791983068,
+      "learning_rate": 1.251145749045246e-05,
+      "loss": 0.7193,
+      "step": 18533
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3775310487927688,
+      "learning_rate": 1.2510718898095359e-05,
+      "loss": 0.6369,
+      "step": 18534
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3290323500157213,
+      "learning_rate": 1.2509980291120061e-05,
+      "loss": 0.718,
+      "step": 18535
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3578257196697023,
+      "learning_rate": 1.2509241669530858e-05,
+      "loss": 0.7037,
+      "step": 18536
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7921757540942407,
+      "learning_rate": 1.250850303333206e-05,
+      "loss": 0.6897,
+      "step": 18537
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6331236344593701,
+      "learning_rate": 1.250776438252796e-05,
+      "loss": 0.5812,
+      "step": 18538
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4332090619807198,
+      "learning_rate": 1.2507025717122865e-05,
+      "loss": 0.7337,
+      "step": 18539
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.411339458047285,
+      "learning_rate": 1.2506287037121072e-05,
+      "loss": 0.5848,
+      "step": 18540
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.497499527180701,
+      "learning_rate": 1.250554834252688e-05,
+      "loss": 0.6753,
+      "step": 18541
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.437847797897867,
+      "learning_rate": 1.2504809633344594e-05,
+      "loss": 0.7305,
+      "step": 18542
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6035804907114486,
+      "learning_rate": 1.2504070909578515e-05,
+      "loss": 0.6954,
+      "step": 18543
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4855546884705222,
+      "learning_rate": 1.2503332171232938e-05,
+      "loss": 0.8033,
+      "step": 18544
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4991831968296128,
+      "learning_rate": 1.250259341831217e-05,
+      "loss": 0.6217,
+      "step": 18545
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8056741656363238,
+      "learning_rate": 1.2501854650820513e-05,
+      "loss": 0.8008,
+      "step": 18546
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3021909993710925,
+      "learning_rate": 1.2501115868762264e-05,
+      "loss": 0.7417,
+      "step": 18547
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3046868284214455,
+      "learning_rate": 1.2500377072141727e-05,
+      "loss": 0.8913,
+      "step": 18548
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6078583661652921,
+      "learning_rate": 1.2499638260963204e-05,
+      "loss": 0.564,
+      "step": 18549
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6697878950452691,
+      "learning_rate": 1.2498899435230996e-05,
+      "loss": 0.6999,
+      "step": 18550
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.411343159742837,
+      "learning_rate": 1.2498160594949402e-05,
+      "loss": 0.7402,
+      "step": 18551
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.704843355466098,
+      "learning_rate": 1.2497421740122728e-05,
+      "loss": 0.7357,
+      "step": 18552
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7632639582096514,
+      "learning_rate": 1.2496682870755275e-05,
+      "loss": 0.8015,
+      "step": 18553
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.488583902339039,
+      "learning_rate": 1.2495943986851342e-05,
+      "loss": 0.7187,
+      "step": 18554
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2499563179844706,
+      "learning_rate": 1.2495205088415233e-05,
+      "loss": 0.7255,
+      "step": 18555
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4393991742539851,
+      "learning_rate": 1.2494466175451253e-05,
+      "loss": 0.8168,
+      "step": 18556
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5020173904828007,
+      "learning_rate": 1.2493727247963695e-05,
+      "loss": 0.6412,
+      "step": 18557
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6688992216739078,
+      "learning_rate": 1.2492988305956875e-05,
+      "loss": 0.69,
+      "step": 18558
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3676955666037884,
+      "learning_rate": 1.249224934943508e-05,
+      "loss": 0.7603,
+      "step": 18559
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3668893535724544,
+      "learning_rate": 1.2491510378402628e-05,
+      "loss": 0.7296,
+      "step": 18560
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4464834045739257,
+      "learning_rate": 1.2490771392863805e-05,
+      "loss": 0.6903,
+      "step": 18561
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5867161319338854,
+      "learning_rate": 1.2490032392822929e-05,
+      "loss": 0.8354,
+      "step": 18562
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3946433691654738,
+      "learning_rate": 1.2489293378284298e-05,
+      "loss": 0.6641,
+      "step": 18563
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7955429706839448,
+      "learning_rate": 1.2488554349252205e-05,
+      "loss": 0.71,
+      "step": 18564
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4034824239438448,
+      "learning_rate": 1.2487815305730967e-05,
+      "loss": 0.7737,
+      "step": 18565
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6014341638602376,
+      "learning_rate": 1.2487076247724876e-05,
+      "loss": 0.5495,
+      "step": 18566
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9949385396549584,
+      "learning_rate": 1.2486337175238246e-05,
+      "loss": 0.6781,
+      "step": 18567
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3907187895907425,
+      "learning_rate": 1.2485598088275367e-05,
+      "loss": 0.7612,
+      "step": 18568
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3029818910952085,
+      "learning_rate": 1.2484858986840553e-05,
+      "loss": 0.6735,
+      "step": 18569
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.33241030009292,
+      "learning_rate": 1.2484119870938102e-05,
+      "loss": 0.817,
+      "step": 18570
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6228565100373673,
+      "learning_rate": 1.2483380740572319e-05,
+      "loss": 0.6803,
+      "step": 18571
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4358663705300256,
+      "learning_rate": 1.2482641595747508e-05,
+      "loss": 0.6702,
+      "step": 18572
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6034238616522609,
+      "learning_rate": 1.2481902436467972e-05,
+      "loss": 0.5421,
+      "step": 18573
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.378873206511304,
+      "learning_rate": 1.2481163262738011e-05,
+      "loss": 0.6704,
+      "step": 18574
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4822292594906965,
+      "learning_rate": 1.2480424074561934e-05,
+      "loss": 0.7247,
+      "step": 18575
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6523585529779607,
+      "learning_rate": 1.2479684871944044e-05,
+      "loss": 0.7558,
+      "step": 18576
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3409605841948864,
+      "learning_rate": 1.247894565488864e-05,
+      "loss": 0.8276,
+      "step": 18577
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2869597042807177,
+      "learning_rate": 1.2478206423400034e-05,
+      "loss": 0.7993,
+      "step": 18578
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6036474967416579,
+      "learning_rate": 1.2477467177482524e-05,
+      "loss": 0.6577,
+      "step": 18579
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3246486421495467,
+      "learning_rate": 1.2476727917140415e-05,
+      "loss": 0.6471,
+      "step": 18580
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7410574814129864,
+      "learning_rate": 1.2475988642378011e-05,
+      "loss": 0.7224,
+      "step": 18581
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6305188405702722,
+      "learning_rate": 1.247524935319962e-05,
+      "loss": 0.7147,
+      "step": 18582
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7000021807869659,
+      "learning_rate": 1.247451004960954e-05,
+      "loss": 0.7514,
+      "step": 18583
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5367980248762658,
+      "learning_rate": 1.2473770731612084e-05,
+      "loss": 0.7143,
+      "step": 18584
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3738541679740717,
+      "learning_rate": 1.2473031399211548e-05,
+      "loss": 0.6305,
+      "step": 18585
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.429933821296462,
+      "learning_rate": 1.2472292052412243e-05,
+      "loss": 0.7623,
+      "step": 18586
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7192659417877976,
+      "learning_rate": 1.2471552691218465e-05,
+      "loss": 0.7226,
+      "step": 18587
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4490812040780292,
+      "learning_rate": 1.247081331563453e-05,
+      "loss": 0.7839,
+      "step": 18588
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7780016269562773,
+      "learning_rate": 1.2470073925664736e-05,
+      "loss": 0.6749,
+      "step": 18589
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5176789100967387,
+      "learning_rate": 1.2469334521313392e-05,
+      "loss": 0.6265,
+      "step": 18590
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.231161575373524,
+      "learning_rate": 1.2468595102584798e-05,
+      "loss": 0.7672,
+      "step": 18591
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6346861799207514,
+      "learning_rate": 1.2467855669483262e-05,
+      "loss": 0.5422,
+      "step": 18592
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6810475194558985,
+      "learning_rate": 1.2467116222013089e-05,
+      "loss": 0.6674,
+      "step": 18593
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.509519960170771,
+      "learning_rate": 1.2466376760178584e-05,
+      "loss": 0.8266,
+      "step": 18594
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2654862147202293,
+      "learning_rate": 1.2465637283984056e-05,
+      "loss": 0.6217,
+      "step": 18595
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4275420304016997,
+      "learning_rate": 1.2464897793433803e-05,
+      "loss": 0.6655,
+      "step": 18596
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6126205060760779,
+      "learning_rate": 1.2464158288532136e-05,
+      "loss": 0.693,
+      "step": 18597
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6853953848006407,
+      "learning_rate": 1.246341876928336e-05,
+      "loss": 0.756,
+      "step": 18598
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9137446229714261,
+      "learning_rate": 1.2462679235691779e-05,
+      "loss": 0.6613,
+      "step": 18599
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4919645762343352,
+      "learning_rate": 1.24619396877617e-05,
+      "loss": 0.6179,
+      "step": 18600
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.29254187313582,
+      "learning_rate": 1.2461200125497428e-05,
+      "loss": 0.6969,
+      "step": 18601
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5033885455748481,
+      "learning_rate": 1.2460460548903273e-05,
+      "loss": 0.6959,
+      "step": 18602
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7073892500423513,
+      "learning_rate": 1.2459720957983535e-05,
+      "loss": 0.6511,
+      "step": 18603
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3142589007502408,
+      "learning_rate": 1.2458981352742522e-05,
+      "loss": 0.7517,
+      "step": 18604
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.560270364081916,
+      "learning_rate": 1.2458241733184542e-05,
+      "loss": 0.7401,
+      "step": 18605
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5756039510751616,
+      "learning_rate": 1.2457502099313901e-05,
+      "loss": 0.6901,
+      "step": 18606
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4666019112919089,
+      "learning_rate": 1.2456762451134905e-05,
+      "loss": 0.7326,
+      "step": 18607
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8512612596500007,
+      "learning_rate": 1.2456022788651858e-05,
+      "loss": 0.6335,
+      "step": 18608
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3375263901369783,
+      "learning_rate": 1.2455283111869072e-05,
+      "loss": 0.6236,
+      "step": 18609
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.647627962791388,
+      "learning_rate": 1.2454543420790847e-05,
+      "loss": 0.789,
+      "step": 18610
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2846687655172773,
+      "learning_rate": 1.2453803715421496e-05,
+      "loss": 0.59,
+      "step": 18611
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.530295039675972,
+      "learning_rate": 1.245306399576532e-05,
+      "loss": 0.7334,
+      "step": 18612
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5178778496248269,
+      "learning_rate": 1.2452324261826632e-05,
+      "loss": 0.8278,
+      "step": 18613
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6217082522837707,
+      "learning_rate": 1.2451584513609736e-05,
+      "loss": 0.6936,
+      "step": 18614
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6185421937471467,
+      "learning_rate": 1.2450844751118934e-05,
+      "loss": 0.8021,
+      "step": 18615
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6503504956061277,
+      "learning_rate": 1.2450104974358544e-05,
+      "loss": 0.5352,
+      "step": 18616
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4441673525576044,
+      "learning_rate": 1.2449365183332862e-05,
+      "loss": 0.7568,
+      "step": 18617
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6304277265194953,
+      "learning_rate": 1.2448625378046205e-05,
+      "loss": 0.7784,
+      "step": 18618
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7482459719120658,
+      "learning_rate": 1.244788555850287e-05,
+      "loss": 0.8188,
+      "step": 18619
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5070202037464906,
+      "learning_rate": 1.2447145724707178e-05,
+      "loss": 0.5702,
+      "step": 18620
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3363749492375105,
+      "learning_rate": 1.2446405876663424e-05,
+      "loss": 0.7939,
+      "step": 18621
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4179027630030365,
+      "learning_rate": 1.2445666014375918e-05,
+      "loss": 0.7214,
+      "step": 18622
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3802665075140403,
+      "learning_rate": 1.2444926137848974e-05,
+      "loss": 0.6547,
+      "step": 18623
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3534041871101108,
+      "learning_rate": 1.2444186247086895e-05,
+      "loss": 0.5786,
+      "step": 18624
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7318060723705175,
+      "learning_rate": 1.244344634209399e-05,
+      "loss": 0.7024,
+      "step": 18625
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4652010694250046,
+      "learning_rate": 1.2442706422874567e-05,
+      "loss": 0.7482,
+      "step": 18626
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4291917371765839,
+      "learning_rate": 1.2441966489432936e-05,
+      "loss": 0.7983,
+      "step": 18627
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3337743286361543,
+      "learning_rate": 1.24412265417734e-05,
+      "loss": 0.7307,
+      "step": 18628
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.510234510074507,
+      "learning_rate": 1.2440486579900268e-05,
+      "loss": 0.76,
+      "step": 18629
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6049083041476994,
+      "learning_rate": 1.2439746603817855e-05,
+      "loss": 0.5477,
+      "step": 18630
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5856673747672798,
+      "learning_rate": 1.2439006613530462e-05,
+      "loss": 0.6758,
+      "step": 18631
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3752343165212495,
+      "learning_rate": 1.2438266609042403e-05,
+      "loss": 0.7328,
+      "step": 18632
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.518910219449199,
+      "learning_rate": 1.2437526590357982e-05,
+      "loss": 0.7568,
+      "step": 18633
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3093659766587675,
+      "learning_rate": 1.243678655748151e-05,
+      "loss": 0.7404,
+      "step": 18634
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.8745348934274477,
+      "learning_rate": 1.2436046510417297e-05,
+      "loss": 0.8002,
+      "step": 18635
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6370116722687923,
+      "learning_rate": 1.2435306449169646e-05,
+      "loss": 0.5318,
+      "step": 18636
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6629653375747637,
+      "learning_rate": 1.2434566373742873e-05,
+      "loss": 0.7546,
+      "step": 18637
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5960079723892333,
+      "learning_rate": 1.2433826284141282e-05,
+      "loss": 0.762,
+      "step": 18638
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6686493306733594,
+      "learning_rate": 1.2433086180369184e-05,
+      "loss": 0.7441,
+      "step": 18639
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3574100446807302,
+      "learning_rate": 1.2432346062430889e-05,
+      "loss": 0.6255,
+      "step": 18640
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.369167254083286,
+      "learning_rate": 1.2431605930330706e-05,
+      "loss": 0.653,
+      "step": 18641
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.032305634678877,
+      "learning_rate": 1.243086578407294e-05,
+      "loss": 0.7348,
+      "step": 18642
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2899620048896479,
+      "learning_rate": 1.2430125623661909e-05,
+      "loss": 0.8454,
+      "step": 18643
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.381962293948955,
+      "learning_rate": 1.2429385449101912e-05,
+      "loss": 0.7037,
+      "step": 18644
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5660784515060546,
+      "learning_rate": 1.2428645260397267e-05,
+      "loss": 0.741,
+      "step": 18645
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7888793176587665,
+      "learning_rate": 1.2427905057552279e-05,
+      "loss": 0.7492,
+      "step": 18646
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4572324068539182,
+      "learning_rate": 1.242716484057126e-05,
+      "loss": 0.7826,
+      "step": 18647
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6276999689000804,
+      "learning_rate": 1.2426424609458518e-05,
+      "loss": 0.816,
+      "step": 18648
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.493500440223188,
+      "learning_rate": 1.2425684364218365e-05,
+      "loss": 0.7692,
+      "step": 18649
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7823623129995876,
+      "learning_rate": 1.2424944104855107e-05,
+      "loss": 0.7476,
+      "step": 18650
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6328122792988156,
+      "learning_rate": 1.2424203831373058e-05,
+      "loss": 0.8458,
+      "step": 18651
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6004547660745331,
+      "learning_rate": 1.2423463543776529e-05,
+      "loss": 0.5367,
+      "step": 18652
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6541038545577607,
+      "learning_rate": 1.2422723242069826e-05,
+      "loss": 0.6752,
+      "step": 18653
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5615645763802606,
+      "learning_rate": 1.2421982926257262e-05,
+      "loss": 0.6502,
+      "step": 18654
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2881254302023986,
+      "learning_rate": 1.2421242596343144e-05,
+      "loss": 0.7088,
+      "step": 18655
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4418053401823117,
+      "learning_rate": 1.242050225233179e-05,
+      "loss": 0.608,
+      "step": 18656
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4964720412463848,
+      "learning_rate": 1.24197618942275e-05,
+      "loss": 0.6754,
+      "step": 18657
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.17718243104798,
+      "learning_rate": 1.2419021522034596e-05,
+      "loss": 0.6884,
+      "step": 18658
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.288924668483774,
+      "learning_rate": 1.2418281135757378e-05,
+      "loss": 0.7417,
+      "step": 18659
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6503854790214438,
+      "learning_rate": 1.2417540735400161e-05,
+      "loss": 0.759,
+      "step": 18660
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4859294872150626,
+      "learning_rate": 1.241680032096726e-05,
+      "loss": 0.6951,
+      "step": 18661
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5160232800185633,
+      "learning_rate": 1.2416059892462979e-05,
+      "loss": 0.7526,
+      "step": 18662
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4953198382986121,
+      "learning_rate": 1.2415319449891638e-05,
+      "loss": 0.7834,
+      "step": 18663
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.40230250435938,
+      "learning_rate": 1.2414578993257535e-05,
+      "loss": 0.7405,
+      "step": 18664
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4113386996941852,
+      "learning_rate": 1.2413838522564997e-05,
+      "loss": 0.6877,
+      "step": 18665
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4977269043590633,
+      "learning_rate": 1.2413098037818321e-05,
+      "loss": 0.7504,
+      "step": 18666
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5757477814165372,
+      "learning_rate": 1.2412357539021826e-05,
+      "loss": 0.5446,
+      "step": 18667
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3994113692763002,
+      "learning_rate": 1.2411617026179821e-05,
+      "loss": 0.8637,
+      "step": 18668
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6437544857601329,
+      "learning_rate": 1.241087649929662e-05,
+      "loss": 0.5476,
+      "step": 18669
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.380970315537273,
+      "learning_rate": 1.241013595837653e-05,
+      "loss": 0.7323,
+      "step": 18670
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4431934190405942,
+      "learning_rate": 1.240939540342387e-05,
+      "loss": 0.8432,
+      "step": 18671
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.661741704208903,
+      "learning_rate": 1.2408654834442942e-05,
+      "loss": 0.9238,
+      "step": 18672
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3411054483300193,
+      "learning_rate": 1.2407914251438065e-05,
+      "loss": 0.6399,
+      "step": 18673
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8522713423882693,
+      "learning_rate": 1.2407173654413549e-05,
+      "loss": 0.6027,
+      "step": 18674
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1395988817713647,
+      "learning_rate": 1.2406433043373707e-05,
+      "loss": 0.8032,
+      "step": 18675
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4607742033318125,
+      "learning_rate": 1.2405692418322847e-05,
+      "loss": 0.709,
+      "step": 18676
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.399122860964821,
+      "learning_rate": 1.2404951779265287e-05,
+      "loss": 0.6506,
+      "step": 18677
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.538865677510638,
+      "learning_rate": 1.2404211126205335e-05,
+      "loss": 0.6958,
+      "step": 18678
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4347735112493223,
+      "learning_rate": 1.2403470459147305e-05,
+      "loss": 0.7824,
+      "step": 18679
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0498574807965544,
+      "learning_rate": 1.2402729778095507e-05,
+      "loss": 0.6622,
+      "step": 18680
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3765226591258037,
+      "learning_rate": 1.2401989083054258e-05,
+      "loss": 0.7728,
+      "step": 18681
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4750364820795425,
+      "learning_rate": 1.2401248374027868e-05,
+      "loss": 0.7243,
+      "step": 18682
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.398096805238241,
+      "learning_rate": 1.240050765102065e-05,
+      "loss": 0.5881,
+      "step": 18683
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9692272564732858,
+      "learning_rate": 1.2399766914036914e-05,
+      "loss": 0.8124,
+      "step": 18684
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6184359562446897,
+      "learning_rate": 1.2399026163080978e-05,
+      "loss": 0.808,
+      "step": 18685
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.441408442427582,
+      "learning_rate": 1.2398285398157153e-05,
+      "loss": 0.5712,
+      "step": 18686
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.579921165041439,
+      "learning_rate": 1.2397544619269748e-05,
+      "loss": 0.6823,
+      "step": 18687
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3888941766739695,
+      "learning_rate": 1.2396803826423083e-05,
+      "loss": 0.7319,
+      "step": 18688
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.406074539346268,
+      "learning_rate": 1.2396063019621463e-05,
+      "loss": 0.7732,
+      "step": 18689
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.368012166606534,
+      "learning_rate": 1.2395322198869209e-05,
+      "loss": 0.8277,
+      "step": 18690
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7820400476512899,
+      "learning_rate": 1.2394581364170626e-05,
+      "loss": 0.683,
+      "step": 18691
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7239128419818452,
+      "learning_rate": 1.2393840515530038e-05,
+      "loss": 0.5882,
+      "step": 18692
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6630999414438308,
+      "learning_rate": 1.239309965295175e-05,
+      "loss": 0.6651,
+      "step": 18693
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5885447018500999,
+      "learning_rate": 1.2392358776440076e-05,
+      "loss": 0.5514,
+      "step": 18694
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5537627145384474,
+      "learning_rate": 1.2391617885999336e-05,
+      "loss": 0.8051,
+      "step": 18695
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5711512968275088,
+      "learning_rate": 1.2390876981633838e-05,
+      "loss": 0.7753,
+      "step": 18696
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.318329955518924,
+      "learning_rate": 1.2390136063347895e-05,
+      "loss": 0.6696,
+      "step": 18697
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9103896409941905,
+      "learning_rate": 1.2389395131145823e-05,
+      "loss": 0.6783,
+      "step": 18698
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5978685424844987,
+      "learning_rate": 1.2388654185031941e-05,
+      "loss": 0.5546,
+      "step": 18699
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5806050313190527,
+      "learning_rate": 1.2387913225010554e-05,
+      "loss": 0.5436,
+      "step": 18700
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3917106636575847,
+      "learning_rate": 1.2387172251085982e-05,
+      "loss": 0.691,
+      "step": 18701
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2449019616636328,
+      "learning_rate": 1.2386431263262538e-05,
+      "loss": 0.6664,
+      "step": 18702
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5905451584855739,
+      "learning_rate": 1.2385690261544531e-05,
+      "loss": 0.5428,
+      "step": 18703
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5159468138452474,
+      "learning_rate": 1.2384949245936284e-05,
+      "loss": 0.6599,
+      "step": 18704
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4745539573296529,
+      "learning_rate": 1.2384208216442106e-05,
+      "loss": 0.717,
+      "step": 18705
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.490535686112704,
+      "learning_rate": 1.2383467173066315e-05,
+      "loss": 0.6599,
+      "step": 18706
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9442289236209886,
+      "learning_rate": 1.2382726115813224e-05,
+      "loss": 0.6851,
+      "step": 18707
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5809543169684677,
+      "learning_rate": 1.2381985044687144e-05,
+      "loss": 0.5075,
+      "step": 18708
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4689363570796878,
+      "learning_rate": 1.2381243959692394e-05,
+      "loss": 0.7342,
+      "step": 18709
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6029057860444681,
+      "learning_rate": 1.2380502860833287e-05,
+      "loss": 0.5517,
+      "step": 18710
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5152334698081602,
+      "learning_rate": 1.237976174811414e-05,
+      "loss": 0.8219,
+      "step": 18711
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5910422004711133,
+      "learning_rate": 1.2379020621539266e-05,
+      "loss": 0.7271,
+      "step": 18712
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.540486788810985,
+      "learning_rate": 1.2378279481112983e-05,
+      "loss": 0.8036,
+      "step": 18713
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5958379253238477,
+      "learning_rate": 1.2377538326839603e-05,
+      "loss": 0.658,
+      "step": 18714
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6785913092986813,
+      "learning_rate": 1.2376797158723439e-05,
+      "loss": 0.692,
+      "step": 18715
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3644615036233365,
+      "learning_rate": 1.2376055976768815e-05,
+      "loss": 0.5736,
+      "step": 18716
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.287343919953565,
+      "learning_rate": 1.2375314780980034e-05,
+      "loss": 0.6249,
+      "step": 18717
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5928711394063118,
+      "learning_rate": 1.2374573571361426e-05,
+      "loss": 0.5407,
+      "step": 18718
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3276021623814178,
+      "learning_rate": 1.2373832347917292e-05,
+      "loss": 0.7098,
+      "step": 18719
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5387590303892351,
+      "learning_rate": 1.2373091110651958e-05,
+      "loss": 0.7286,
+      "step": 18720
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.395254095821443,
+      "learning_rate": 1.2372349859569737e-05,
+      "loss": 0.6281,
+      "step": 18721
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8735671000594603,
+      "learning_rate": 1.2371608594674943e-05,
+      "loss": 0.6643,
+      "step": 18722
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5900034977175538,
+      "learning_rate": 1.237086731597189e-05,
+      "loss": 0.5533,
+      "step": 18723
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5851061458591764,
+      "learning_rate": 1.2370126023464901e-05,
+      "loss": 0.7763,
+      "step": 18724
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4191044538536965,
+      "learning_rate": 1.2369384717158285e-05,
+      "loss": 0.92,
+      "step": 18725
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.59055841791721,
+      "learning_rate": 1.2368643397056364e-05,
+      "loss": 0.6312,
+      "step": 18726
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.216919831390292,
+      "learning_rate": 1.2367902063163448e-05,
+      "loss": 0.635,
+      "step": 18727
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1816857421079257,
+      "learning_rate": 1.236716071548386e-05,
+      "loss": 0.6562,
+      "step": 18728
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.524453520235544,
+      "learning_rate": 1.236641935402191e-05,
+      "loss": 0.7238,
+      "step": 18729
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6205684829658091,
+      "learning_rate": 1.2365677978781919e-05,
+      "loss": 0.4819,
+      "step": 18730
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.370807400240014,
+      "learning_rate": 1.2364936589768202e-05,
+      "loss": 0.6013,
+      "step": 18731
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2035945072851972,
+      "learning_rate": 1.2364195186985073e-05,
+      "loss": 0.8382,
+      "step": 18732
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3501569015143067,
+      "learning_rate": 1.2363453770436853e-05,
+      "loss": 0.816,
+      "step": 18733
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.37617733804839,
+      "learning_rate": 1.2362712340127856e-05,
+      "loss": 0.7545,
+      "step": 18734
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.669866624623653,
+      "learning_rate": 1.23619708960624e-05,
+      "loss": 0.7953,
+      "step": 18735
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5958828318218268,
+      "learning_rate": 1.2361229438244803e-05,
+      "loss": 0.6838,
+      "step": 18736
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4354685930922075,
+      "learning_rate": 1.2360487966679383e-05,
+      "loss": 0.7842,
+      "step": 18737
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7542055097785272,
+      "learning_rate": 1.2359746481370447e-05,
+      "loss": 0.863,
+      "step": 18738
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5384145961165598,
+      "learning_rate": 1.2359004982322328e-05,
+      "loss": 0.7928,
+      "step": 18739
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3194088038535976,
+      "learning_rate": 1.2358263469539328e-05,
+      "loss": 0.7664,
+      "step": 18740
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4374414389544738,
+      "learning_rate": 1.2357521943025779e-05,
+      "loss": 0.8009,
+      "step": 18741
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.367262547766313,
+      "learning_rate": 1.2356780402785985e-05,
+      "loss": 0.6788,
+      "step": 18742
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8943653670822698,
+      "learning_rate": 1.2356038848824272e-05,
+      "loss": 0.7463,
+      "step": 18743
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3411079646453377,
+      "learning_rate": 1.2355297281144953e-05,
+      "loss": 0.7627,
+      "step": 18744
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6963784445675094,
+      "learning_rate": 1.2354555699752351e-05,
+      "loss": 0.8105,
+      "step": 18745
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5739983687879378,
+      "learning_rate": 1.2353814104650777e-05,
+      "loss": 0.5386,
+      "step": 18746
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3571787343580792,
+      "learning_rate": 1.2353072495844551e-05,
+      "loss": 0.6495,
+      "step": 18747
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.507550547181822,
+      "learning_rate": 1.2352330873337995e-05,
+      "loss": 0.7859,
+      "step": 18748
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5398272936990303,
+      "learning_rate": 1.2351589237135422e-05,
+      "loss": 0.6987,
+      "step": 18749
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6324033638564155,
+      "learning_rate": 1.2350847587241155e-05,
+      "loss": 0.7209,
+      "step": 18750
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.700511672296082,
+      "learning_rate": 1.2350105923659506e-05,
+      "loss": 0.7264,
+      "step": 18751
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4593746286327827,
+      "learning_rate": 1.2349364246394797e-05,
+      "loss": 0.7821,
+      "step": 18752
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3537691145570658,
+      "learning_rate": 1.2348622555451346e-05,
+      "loss": 0.7839,
+      "step": 18753
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3500487719036058,
+      "learning_rate": 1.2347880850833472e-05,
+      "loss": 0.7046,
+      "step": 18754
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4195213970416514,
+      "learning_rate": 1.234713913254549e-05,
+      "loss": 0.7529,
+      "step": 18755
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6259320802559676,
+      "learning_rate": 1.2346397400591721e-05,
+      "loss": 0.5965,
+      "step": 18756
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.139220415071299,
+      "learning_rate": 1.2345655654976485e-05,
+      "loss": 0.6891,
+      "step": 18757
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.439639977700012,
+      "learning_rate": 1.2344913895704099e-05,
+      "loss": 0.7118,
+      "step": 18758
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4496067984492114,
+      "learning_rate": 1.2344172122778882e-05,
+      "loss": 0.7515,
+      "step": 18759
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2939355940386346,
+      "learning_rate": 1.234343033620515e-05,
+      "loss": 0.6541,
+      "step": 18760
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.46336657054244,
+      "learning_rate": 1.2342688535987228e-05,
+      "loss": 0.8068,
+      "step": 18761
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2898244303382673,
+      "learning_rate": 1.234194672212943e-05,
+      "loss": 0.8644,
+      "step": 18762
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.664328569201597,
+      "learning_rate": 1.234120489463608e-05,
+      "loss": 0.7772,
+      "step": 18763
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.410766957436791,
+      "learning_rate": 1.234046305351149e-05,
+      "loss": 0.7334,
+      "step": 18764
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6264587280725173,
+      "learning_rate": 1.2339721198759987e-05,
+      "loss": 0.659,
+      "step": 18765
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4941515485825496,
+      "learning_rate": 1.2338979330385881e-05,
+      "loss": 0.6964,
+      "step": 18766
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.435872691580758,
+      "learning_rate": 1.23382374483935e-05,
+      "loss": 0.7645,
+      "step": 18767
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.479652323629334,
+      "learning_rate": 1.2337495552787157e-05,
+      "loss": 0.8526,
+      "step": 18768
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5792627581153825,
+      "learning_rate": 1.2336753643571181e-05,
+      "loss": 0.5098,
+      "step": 18769
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4369307445409354,
+      "learning_rate": 1.2336011720749881e-05,
+      "loss": 0.8491,
+      "step": 18770
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5969719945966369,
+      "learning_rate": 1.2335269784327583e-05,
+      "loss": 0.563,
+      "step": 18771
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9729825456452124,
+      "learning_rate": 1.2334527834308606e-05,
+      "loss": 0.6483,
+      "step": 18772
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.737134005738476,
+      "learning_rate": 1.2333785870697265e-05,
+      "loss": 0.7788,
+      "step": 18773
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.683765656551236,
+      "learning_rate": 1.2333043893497886e-05,
+      "loss": 0.7112,
+      "step": 18774
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4192979872796192,
+      "learning_rate": 1.2332301902714789e-05,
+      "loss": 0.6555,
+      "step": 18775
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5582639129082467,
+      "learning_rate": 1.233155989835229e-05,
+      "loss": 0.5427,
+      "step": 18776
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.383446048235621,
+      "learning_rate": 1.233081788041471e-05,
+      "loss": 0.8293,
+      "step": 18777
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4270942442771564,
+      "learning_rate": 1.2330075848906371e-05,
+      "loss": 0.5851,
+      "step": 18778
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.82526539468014,
+      "learning_rate": 1.2329333803831593e-05,
+      "loss": 0.7492,
+      "step": 18779
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8841929454129795,
+      "learning_rate": 1.2328591745194696e-05,
+      "loss": 0.7785,
+      "step": 18780
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8749854855531318,
+      "learning_rate": 1.2327849673000001e-05,
+      "loss": 0.7699,
+      "step": 18781
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6166459976465035,
+      "learning_rate": 1.2327107587251827e-05,
+      "loss": 0.5691,
+      "step": 18782
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4750450501682748,
+      "learning_rate": 1.2326365487954497e-05,
+      "loss": 0.7161,
+      "step": 18783
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6830014998991983,
+      "learning_rate": 1.232562337511233e-05,
+      "loss": 0.5228,
+      "step": 18784
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5674179852673795,
+      "learning_rate": 1.232488124872965e-05,
+      "loss": 0.5005,
+      "step": 18785
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6879277227497962,
+      "learning_rate": 1.2324139108810775e-05,
+      "loss": 0.7723,
+      "step": 18786
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5788155359796083,
+      "learning_rate": 1.2323396955360023e-05,
+      "loss": 0.5634,
+      "step": 18787
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7284896359840392,
+      "learning_rate": 1.232265478838172e-05,
+      "loss": 0.745,
+      "step": 18788
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3385733509522717,
+      "learning_rate": 1.2321912607880184e-05,
+      "loss": 0.6871,
+      "step": 18789
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4124086984538127,
+      "learning_rate": 1.2321170413859742e-05,
+      "loss": 0.8002,
+      "step": 18790
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4359567824952255,
+      "learning_rate": 1.2320428206324703e-05,
+      "loss": 0.8513,
+      "step": 18791
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3600599387728467,
+      "learning_rate": 1.2319685985279406e-05,
+      "loss": 0.8575,
+      "step": 18792
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.407494102002925,
+      "learning_rate": 1.2318943750728157e-05,
+      "loss": 0.947,
+      "step": 18793
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5030099539950625,
+      "learning_rate": 1.2318201502675285e-05,
+      "loss": 0.6893,
+      "step": 18794
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4650848284269666,
+      "learning_rate": 1.231745924112511e-05,
+      "loss": 0.8103,
+      "step": 18795
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3897193117935616,
+      "learning_rate": 1.231671696608195e-05,
+      "loss": 0.7271,
+      "step": 18796
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6159270281918696,
+      "learning_rate": 1.2315974677550136e-05,
+      "loss": 0.824,
+      "step": 18797
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0909552451205298,
+      "learning_rate": 1.231523237553398e-05,
+      "loss": 0.6381,
+      "step": 18798
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.468005605185213,
+      "learning_rate": 1.2314490060037812e-05,
+      "loss": 0.745,
+      "step": 18799
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.461288692640808,
+      "learning_rate": 1.2313747731065947e-05,
+      "loss": 0.7519,
+      "step": 18800
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3011480284424874,
+      "learning_rate": 1.2313005388622712e-05,
+      "loss": 0.6627,
+      "step": 18801
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8364315465521186,
+      "learning_rate": 1.2312263032712425e-05,
+      "loss": 0.7977,
+      "step": 18802
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4710838175147791,
+      "learning_rate": 1.2311520663339411e-05,
+      "loss": 0.8111,
+      "step": 18803
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3684267627901994,
+      "learning_rate": 1.2310778280507994e-05,
+      "loss": 0.6261,
+      "step": 18804
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8364263513213492,
+      "learning_rate": 1.2310035884222493e-05,
+      "loss": 0.7812,
+      "step": 18805
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5879550212342586,
+      "learning_rate": 1.2309293474487232e-05,
+      "loss": 0.7008,
+      "step": 18806
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4148367615570094,
+      "learning_rate": 1.2308551051306533e-05,
+      "loss": 0.7716,
+      "step": 18807
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4119803208011628,
+      "learning_rate": 1.230780861468472e-05,
+      "loss": 0.7379,
+      "step": 18808
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.688328559508604,
+      "learning_rate": 1.2307066164626113e-05,
+      "loss": 0.6986,
+      "step": 18809
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5736890635235237,
+      "learning_rate": 1.2306323701135036e-05,
+      "loss": 0.5217,
+      "step": 18810
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.432742063996631,
+      "learning_rate": 1.2305581224215812e-05,
+      "loss": 0.7825,
+      "step": 18811
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5850317885599392,
+      "learning_rate": 1.2304838733872768e-05,
+      "loss": 0.6568,
+      "step": 18812
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5988183346548235,
+      "learning_rate": 1.2304096230110219e-05,
+      "loss": 0.6833,
+      "step": 18813
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.328162406839489,
+      "learning_rate": 1.2303353712932495e-05,
+      "loss": 0.8395,
+      "step": 18814
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8069620938976652,
+      "learning_rate": 1.2302611182343915e-05,
+      "loss": 0.7748,
+      "step": 18815
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6528341343065487,
+      "learning_rate": 1.2301868638348805e-05,
+      "loss": 0.8612,
+      "step": 18816
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.558518954758398,
+      "learning_rate": 1.2301126080951484e-05,
+      "loss": 0.6944,
+      "step": 18817
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6058500884747335,
+      "learning_rate": 1.2300383510156284e-05,
+      "loss": 0.5285,
+      "step": 18818
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7548005366498527,
+      "learning_rate": 1.2299640925967518e-05,
+      "loss": 0.7992,
+      "step": 18819
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5038295353270592,
+      "learning_rate": 1.2298898328389517e-05,
+      "loss": 0.6946,
+      "step": 18820
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5709617394841985,
+      "learning_rate": 1.2298155717426598e-05,
+      "loss": 0.5075,
+      "step": 18821
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4031894721312568,
+      "learning_rate": 1.2297413093083094e-05,
+      "loss": 0.6358,
+      "step": 18822
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7573568485376008,
+      "learning_rate": 1.2296670455363322e-05,
+      "loss": 0.8489,
+      "step": 18823
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3661594231399934,
+      "learning_rate": 1.2295927804271606e-05,
+      "loss": 0.7993,
+      "step": 18824
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.0163109312832077,
+      "learning_rate": 1.2295185139812274e-05,
+      "loss": 0.8145,
+      "step": 18825
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2902967015271316,
+      "learning_rate": 1.2294442461989644e-05,
+      "loss": 0.8178,
+      "step": 18826
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.4149500682514478,
+      "learning_rate": 1.2293699770808047e-05,
+      "loss": 0.7233,
+      "step": 18827
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4755987938424047,
+      "learning_rate": 1.2292957066271802e-05,
+      "loss": 0.6902,
+      "step": 18828
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.285586294591457,
+      "learning_rate": 1.2292214348385235e-05,
+      "loss": 0.6946,
+      "step": 18829
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5128723730496774,
+      "learning_rate": 1.229147161715267e-05,
+      "loss": 0.7419,
+      "step": 18830
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.578458663684803,
+      "learning_rate": 1.2290728872578432e-05,
+      "loss": 0.4987,
+      "step": 18831
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4618642477164197,
+      "learning_rate": 1.2289986114666847e-05,
+      "loss": 0.6431,
+      "step": 18832
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7044104323692895,
+      "learning_rate": 1.2289243343422234e-05,
+      "loss": 0.8055,
+      "step": 18833
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4701915036646385,
+      "learning_rate": 1.2288500558848926e-05,
+      "loss": 0.7923,
+      "step": 18834
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4039292288963512,
+      "learning_rate": 1.228775776095124e-05,
+      "loss": 0.7362,
+      "step": 18835
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.701771060563427,
+      "learning_rate": 1.2287014949733506e-05,
+      "loss": 0.6898,
+      "step": 18836
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4245444041908994,
+      "learning_rate": 1.2286272125200048e-05,
+      "loss": 0.7559,
+      "step": 18837
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2681708128563092,
+      "learning_rate": 1.2285529287355188e-05,
+      "loss": 0.6911,
+      "step": 18838
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3770205610054727,
+      "learning_rate": 1.2284786436203253e-05,
+      "loss": 0.6857,
+      "step": 18839
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7238565309721134,
+      "learning_rate": 1.2284043571748565e-05,
+      "loss": 0.7193,
+      "step": 18840
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8805211500786363,
+      "learning_rate": 1.2283300693995459e-05,
+      "loss": 0.8176,
+      "step": 18841
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7944410078875164,
+      "learning_rate": 1.2282557802948247e-05,
+      "loss": 0.8635,
+      "step": 18842
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4232097709635927,
+      "learning_rate": 1.2281814898611267e-05,
+      "loss": 0.7135,
+      "step": 18843
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.375077774452719,
+      "learning_rate": 1.2281071980988832e-05,
+      "loss": 0.9581,
+      "step": 18844
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5654210291484423,
+      "learning_rate": 1.2280329050085276e-05,
+      "loss": 0.7792,
+      "step": 18845
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5594958465076691,
+      "learning_rate": 1.2279586105904924e-05,
+      "loss": 0.4978,
+      "step": 18846
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.807097567133164,
+      "learning_rate": 1.2278843148452095e-05,
+      "loss": 0.8857,
+      "step": 18847
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6749973805548882,
+      "learning_rate": 1.2278100177731128e-05,
+      "loss": 0.619,
+      "step": 18848
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.495823080763245,
+      "learning_rate": 1.2277357193746333e-05,
+      "loss": 0.5691,
+      "step": 18849
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.406405831434945,
+      "learning_rate": 1.2276614196502047e-05,
+      "loss": 0.7964,
+      "step": 18850
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6491315196685372,
+      "learning_rate": 1.2275871186002592e-05,
+      "loss": 0.7094,
+      "step": 18851
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3757249209352667,
+      "learning_rate": 1.2275128162252297e-05,
+      "loss": 0.6527,
+      "step": 18852
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.325532591546446,
+      "learning_rate": 1.2274385125255481e-05,
+      "loss": 0.7875,
+      "step": 18853
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.870811969514901,
+      "learning_rate": 1.2273642075016476e-05,
+      "loss": 0.7342,
+      "step": 18854
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.056950577212245,
+      "learning_rate": 1.2272899011539607e-05,
+      "loss": 0.7718,
+      "step": 18855
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9426735570459612,
+      "learning_rate": 1.2272155934829203e-05,
+      "loss": 0.8851,
+      "step": 18856
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5337175213160272,
+      "learning_rate": 1.2271412844889585e-05,
+      "loss": 0.7955,
+      "step": 18857
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.397212038566553,
+      "learning_rate": 1.2270669741725083e-05,
+      "loss": 0.804,
+      "step": 18858
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5198174133197744,
+      "learning_rate": 1.2269926625340022e-05,
+      "loss": 0.7838,
+      "step": 18859
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2783761261250324,
+      "learning_rate": 1.226918349573873e-05,
+      "loss": 0.6816,
+      "step": 18860
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0121121378105107,
+      "learning_rate": 1.2268440352925534e-05,
+      "loss": 0.577,
+      "step": 18861
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6042639881673453,
+      "learning_rate": 1.226769719690476e-05,
+      "loss": 0.7242,
+      "step": 18862
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6425670941735458,
+      "learning_rate": 1.2266954027680736e-05,
+      "loss": 0.5033,
+      "step": 18863
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7275791392177877,
+      "learning_rate": 1.2266210845257787e-05,
+      "loss": 0.85,
+      "step": 18864
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.475321419155916,
+      "learning_rate": 1.226546764964024e-05,
+      "loss": 0.7568,
+      "step": 18865
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3063607533134072,
+      "learning_rate": 1.2264724440832423e-05,
+      "loss": 0.6638,
+      "step": 18866
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4588035043558303,
+      "learning_rate": 1.2263981218838668e-05,
+      "loss": 0.843,
+      "step": 18867
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1965353073337262,
+      "learning_rate": 1.226323798366329e-05,
+      "loss": 0.7896,
+      "step": 18868
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3494809671487809,
+      "learning_rate": 1.226249473531063e-05,
+      "loss": 0.7385,
+      "step": 18869
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3812763294868837,
+      "learning_rate": 1.2261751473785004e-05,
+      "loss": 0.7838,
+      "step": 18870
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3878317177053414,
+      "learning_rate": 1.2261008199090751e-05,
+      "loss": 0.8575,
+      "step": 18871
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5862194240105291,
+      "learning_rate": 1.2260264911232187e-05,
+      "loss": 0.5182,
+      "step": 18872
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7694250430954983,
+      "learning_rate": 1.225952161021365e-05,
+      "loss": 0.7097,
+      "step": 18873
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5566107399201967,
+      "learning_rate": 1.225877829603946e-05,
+      "loss": 0.7384,
+      "step": 18874
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6376717015038671,
+      "learning_rate": 1.2258034968713948e-05,
+      "loss": 0.5679,
+      "step": 18875
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6426612325053941,
+      "learning_rate": 1.2257291628241441e-05,
+      "loss": 0.53,
+      "step": 18876
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.57630104072098,
+      "learning_rate": 1.2256548274626267e-05,
+      "loss": 0.7289,
+      "step": 18877
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4547163277304835,
+      "learning_rate": 1.2255804907872755e-05,
+      "loss": 0.7316,
+      "step": 18878
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5588850065584978,
+      "learning_rate": 1.2255061527985235e-05,
+      "loss": 0.7146,
+      "step": 18879
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4753843113491067,
+      "learning_rate": 1.2254318134968029e-05,
+      "loss": 0.6371,
+      "step": 18880
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4329040416961147,
+      "learning_rate": 1.2253574728825472e-05,
+      "loss": 0.7221,
+      "step": 18881
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.498444407799978,
+      "learning_rate": 1.2252831309561888e-05,
+      "loss": 0.7688,
+      "step": 18882
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5520383548486241,
+      "learning_rate": 1.2252087877181606e-05,
+      "loss": 0.4831,
+      "step": 18883
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.468167873381492,
+      "learning_rate": 1.2251344431688959e-05,
+      "loss": 0.7928,
+      "step": 18884
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8037894903385707,
+      "learning_rate": 1.225060097308827e-05,
+      "loss": 0.8159,
+      "step": 18885
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4066642193499967,
+      "learning_rate": 1.2249857501383869e-05,
+      "loss": 0.856,
+      "step": 18886
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5853041921565066,
+      "learning_rate": 1.2249114016580083e-05,
+      "loss": 0.5238,
+      "step": 18887
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4890669290848653,
+      "learning_rate": 1.224837051868125e-05,
+      "loss": 0.7212,
+      "step": 18888
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6005169251980738,
+      "learning_rate": 1.2247627007691684e-05,
+      "loss": 0.6894,
+      "step": 18889
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4856137244041288,
+      "learning_rate": 1.2246883483615731e-05,
+      "loss": 0.851,
+      "step": 18890
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.356563932207103,
+      "learning_rate": 1.2246139946457703e-05,
+      "loss": 0.7829,
+      "step": 18891
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3871589067139185,
+      "learning_rate": 1.2245396396221938e-05,
+      "loss": 0.8133,
+      "step": 18892
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6489939832088634,
+      "learning_rate": 1.2244652832912766e-05,
+      "loss": 0.7129,
+      "step": 18893
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4134945656351532,
+      "learning_rate": 1.2243909256534517e-05,
+      "loss": 0.6707,
+      "step": 18894
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4906546882249818,
+      "learning_rate": 1.2243165667091517e-05,
+      "loss": 0.6808,
+      "step": 18895
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3252752367834104,
+      "learning_rate": 1.224242206458809e-05,
+      "loss": 0.7205,
+      "step": 18896
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5687335584090508,
+      "learning_rate": 1.224167844902858e-05,
+      "loss": 0.5298,
+      "step": 18897
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1104672352375973,
+      "learning_rate": 1.2240934820417303e-05,
+      "loss": 0.6688,
+      "step": 18898
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.569972548422024,
+      "learning_rate": 1.2240191178758598e-05,
+      "loss": 0.7615,
+      "step": 18899
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6143420136198028,
+      "learning_rate": 1.2239447524056786e-05,
+      "loss": 0.6919,
+      "step": 18900
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.436603774332356,
+      "learning_rate": 1.2238703856316208e-05,
+      "loss": 0.6864,
+      "step": 18901
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4682701185605882,
+      "learning_rate": 1.2237960175541184e-05,
+      "loss": 0.6845,
+      "step": 18902
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4160075165253359,
+      "learning_rate": 1.2237216481736047e-05,
+      "loss": 0.8757,
+      "step": 18903
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.646431306924082,
+      "learning_rate": 1.2236472774905128e-05,
+      "loss": 0.7439,
+      "step": 18904
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4300479507411685,
+      "learning_rate": 1.2235729055052757e-05,
+      "loss": 0.7702,
+      "step": 18905
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.448269854803488,
+      "learning_rate": 1.2234985322183262e-05,
+      "loss": 0.7762,
+      "step": 18906
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6191328608382949,
+      "learning_rate": 1.2234241576300974e-05,
+      "loss": 0.7786,
+      "step": 18907
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.495408095436184,
+      "learning_rate": 1.2233497817410226e-05,
+      "loss": 0.7829,
+      "step": 18908
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8138903246208036,
+      "learning_rate": 1.2232754045515348e-05,
+      "loss": 0.6388,
+      "step": 18909
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3290231548098013,
+      "learning_rate": 1.2232010260620665e-05,
+      "loss": 0.6412,
+      "step": 18910
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2584021130976641,
+      "learning_rate": 1.2231266462730514e-05,
+      "loss": 0.7804,
+      "step": 18911
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.557648831508695,
+      "learning_rate": 1.2230522651849224e-05,
+      "loss": 0.7718,
+      "step": 18912
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2928075264462502,
+      "learning_rate": 1.2229778827981122e-05,
+      "loss": 0.6698,
+      "step": 18913
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6782159819905882,
+      "learning_rate": 1.2229034991130544e-05,
+      "loss": 0.6913,
+      "step": 18914
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.445211550561632,
+      "learning_rate": 1.2228291141301818e-05,
+      "loss": 0.7998,
+      "step": 18915
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5220648594427033,
+      "learning_rate": 1.2227547278499276e-05,
+      "loss": 0.7038,
+      "step": 18916
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3965997090284314,
+      "learning_rate": 1.222680340272725e-05,
+      "loss": 0.792,
+      "step": 18917
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3259225680656768,
+      "learning_rate": 1.222605951399007e-05,
+      "loss": 0.6559,
+      "step": 18918
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5483525444589994,
+      "learning_rate": 1.2225315612292061e-05,
+      "loss": 0.8212,
+      "step": 18919
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9250131510014992,
+      "learning_rate": 1.2224571697637566e-05,
+      "loss": 0.757,
+      "step": 18920
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3395160399157273,
+      "learning_rate": 1.2223827770030907e-05,
+      "loss": 0.7827,
+      "step": 18921
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4844236100858175,
+      "learning_rate": 1.2223083829476421e-05,
+      "loss": 0.7552,
+      "step": 18922
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1330431392399456,
+      "learning_rate": 1.2222339875978435e-05,
+      "loss": 0.6873,
+      "step": 18923
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3640735996736326,
+      "learning_rate": 1.2221595909541283e-05,
+      "loss": 0.8215,
+      "step": 18924
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.144869838038427,
+      "learning_rate": 1.22208519301693e-05,
+      "loss": 0.7595,
+      "step": 18925
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5596430094639977,
+      "learning_rate": 1.222010793786681e-05,
+      "loss": 0.5322,
+      "step": 18926
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5810588082777612,
+      "learning_rate": 1.2219363932638148e-05,
+      "loss": 0.6919,
+      "step": 18927
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5694202921813284,
+      "learning_rate": 1.2218619914487648e-05,
+      "loss": 0.5312,
+      "step": 18928
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7576876571565718,
+      "learning_rate": 1.2217875883419642e-05,
+      "loss": 0.7434,
+      "step": 18929
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6181904658285974,
+      "learning_rate": 1.2217131839438457e-05,
+      "loss": 0.5321,
+      "step": 18930
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3161942132325863,
+      "learning_rate": 1.221638778254843e-05,
+      "loss": 0.7535,
+      "step": 18931
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3163173022069867,
+      "learning_rate": 1.2215643712753893e-05,
+      "loss": 0.7401,
+      "step": 18932
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5470599427684297,
+      "learning_rate": 1.2214899630059176e-05,
+      "loss": 0.6299,
+      "step": 18933
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.422881490978293,
+      "learning_rate": 1.2214155534468612e-05,
+      "loss": 0.8197,
+      "step": 18934
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5778804080931648,
+      "learning_rate": 1.2213411425986533e-05,
+      "loss": 0.9093,
+      "step": 18935
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.610143358851623,
+      "learning_rate": 1.2212667304617272e-05,
+      "loss": 0.5419,
+      "step": 18936
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.606130728462785,
+      "learning_rate": 1.2211923170365163e-05,
+      "loss": 0.586,
+      "step": 18937
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6266605802194256,
+      "learning_rate": 1.2211179023234534e-05,
+      "loss": 0.6009,
+      "step": 18938
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5377899196844576,
+      "learning_rate": 1.2210434863229725e-05,
+      "loss": 0.8978,
+      "step": 18939
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4543231850623108,
+      "learning_rate": 1.2209690690355061e-05,
+      "loss": 0.7403,
+      "step": 18940
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.262234063523095,
+      "learning_rate": 1.2208946504614878e-05,
+      "loss": 0.6919,
+      "step": 18941
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9054936773220446,
+      "learning_rate": 1.220820230601351e-05,
+      "loss": 0.8806,
+      "step": 18942
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4110947900971949,
+      "learning_rate": 1.2207458094555289e-05,
+      "loss": 0.8429,
+      "step": 18943
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8758866047324523,
+      "learning_rate": 1.2206713870244549e-05,
+      "loss": 0.6995,
+      "step": 18944
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7272710052340783,
+      "learning_rate": 1.2205969633085621e-05,
+      "loss": 0.7295,
+      "step": 18945
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5263987711425215,
+      "learning_rate": 1.2205225383082844e-05,
+      "loss": 0.6368,
+      "step": 18946
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3165401153694856,
+      "learning_rate": 1.220448112024054e-05,
+      "loss": 0.8302,
+      "step": 18947
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5225951608514436,
+      "learning_rate": 1.2203736844563055e-05,
+      "loss": 0.8491,
+      "step": 18948
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.343170978403408,
+      "learning_rate": 1.2202992556054709e-05,
+      "loss": 0.6737,
+      "step": 18949
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2610708768044658,
+      "learning_rate": 1.2202248254719849e-05,
+      "loss": 0.7559,
+      "step": 18950
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5982203944301623,
+      "learning_rate": 1.2201503940562798e-05,
+      "loss": 0.4954,
+      "step": 18951
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3844957045007527,
+      "learning_rate": 1.22007596135879e-05,
+      "loss": 0.7425,
+      "step": 18952
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6645461671552102,
+      "learning_rate": 1.220001527379948e-05,
+      "loss": 0.828,
+      "step": 18953
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.419699768303785,
+      "learning_rate": 1.2199270921201873e-05,
+      "loss": 0.7419,
+      "step": 18954
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2894984418141842,
+      "learning_rate": 1.2198526555799415e-05,
+      "loss": 0.7529,
+      "step": 18955
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6418804585578761,
+      "learning_rate": 1.219778217759644e-05,
+      "loss": 0.683,
+      "step": 18956
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1918617131038287,
+      "learning_rate": 1.2197037786597278e-05,
+      "loss": 0.6614,
+      "step": 18957
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.428065638765088,
+      "learning_rate": 1.219629338280627e-05,
+      "loss": 0.7503,
+      "step": 18958
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4701352240712207,
+      "learning_rate": 1.2195548966227749e-05,
+      "loss": 0.7094,
+      "step": 18959
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5380631191778722,
+      "learning_rate": 1.2194804536866043e-05,
+      "loss": 0.7604,
+      "step": 18960
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5074231097002888,
+      "learning_rate": 1.219406009472549e-05,
+      "loss": 0.7557,
+      "step": 18961
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4535467978961434,
+      "learning_rate": 1.2193315639810426e-05,
+      "loss": 0.8249,
+      "step": 18962
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.929428157979307,
+      "learning_rate": 1.2192571172125182e-05,
+      "loss": 0.7041,
+      "step": 18963
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7191384977412807,
+      "learning_rate": 1.2191826691674096e-05,
+      "loss": 0.6635,
+      "step": 18964
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5887198597303187,
+      "learning_rate": 1.2191082198461502e-05,
+      "loss": 0.8424,
+      "step": 18965
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.488855707639863,
+      "learning_rate": 1.2190337692491732e-05,
+      "loss": 0.6826,
+      "step": 18966
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3862078767574313,
+      "learning_rate": 1.2189593173769126e-05,
+      "loss": 0.7238,
+      "step": 18967
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6600453839203415,
+      "learning_rate": 1.218884864229801e-05,
+      "loss": 0.7386,
+      "step": 18968
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4615404089530522,
+      "learning_rate": 1.218810409808273e-05,
+      "loss": 0.7251,
+      "step": 18969
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7037985232263118,
+      "learning_rate": 1.2187359541127607e-05,
+      "loss": 0.6725,
+      "step": 18970
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6110567520661474,
+      "learning_rate": 1.2186614971436992e-05,
+      "loss": 0.7447,
+      "step": 18971
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2916434181123821,
+      "learning_rate": 1.2185870389015208e-05,
+      "loss": 0.6589,
+      "step": 18972
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5481323298822323,
+      "learning_rate": 1.21851257938666e-05,
+      "loss": 0.7963,
+      "step": 18973
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4556269782249274,
+      "learning_rate": 1.2184381185995494e-05,
+      "loss": 0.8068,
+      "step": 18974
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5923738006415067,
+      "learning_rate": 1.2183636565406227e-05,
+      "loss": 0.5544,
+      "step": 18975
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4405311549441737,
+      "learning_rate": 1.2182891932103138e-05,
+      "loss": 0.6413,
+      "step": 18976
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6444851799912823,
+      "learning_rate": 1.2182147286090562e-05,
+      "loss": 0.5092,
+      "step": 18977
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6239841026281692,
+      "learning_rate": 1.2181402627372834e-05,
+      "loss": 0.5899,
+      "step": 18978
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7783696575302954,
+      "learning_rate": 1.2180657955954287e-05,
+      "loss": 0.6804,
+      "step": 18979
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4859658507646174,
+      "learning_rate": 1.2179913271839264e-05,
+      "loss": 0.7141,
+      "step": 18980
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3888632736281163,
+      "learning_rate": 1.2179168575032088e-05,
+      "loss": 0.8147,
+      "step": 18981
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.558049970024847,
+      "learning_rate": 1.217842386553711e-05,
+      "loss": 0.466,
+      "step": 18982
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.670812532718855,
+      "learning_rate": 1.2177679143358655e-05,
+      "loss": 0.7106,
+      "step": 18983
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4829847855303413,
+      "learning_rate": 1.2176934408501065e-05,
+      "loss": 0.6746,
+      "step": 18984
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7241730693140382,
+      "learning_rate": 1.2176189660968671e-05,
+      "loss": 0.7514,
+      "step": 18985
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4295778847900118,
+      "learning_rate": 1.2175444900765811e-05,
+      "loss": 0.7095,
+      "step": 18986
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.460668691889576,
+      "learning_rate": 1.2174700127896826e-05,
+      "loss": 0.7082,
+      "step": 18987
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4595842412318516,
+      "learning_rate": 1.2173955342366046e-05,
+      "loss": 0.836,
+      "step": 18988
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.495130909051118,
+      "learning_rate": 1.2173210544177809e-05,
+      "loss": 0.7094,
+      "step": 18989
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3880851209626728,
+      "learning_rate": 1.2172465733336453e-05,
+      "loss": 0.5541,
+      "step": 18990
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6953148446136896,
+      "learning_rate": 1.2171720909846314e-05,
+      "loss": 0.6677,
+      "step": 18991
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.288622819752171,
+      "learning_rate": 1.217097607371173e-05,
+      "loss": 0.6865,
+      "step": 18992
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8336315590791585,
+      "learning_rate": 1.2170231224937032e-05,
+      "loss": 0.7981,
+      "step": 18993
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.439814193980393,
+      "learning_rate": 1.2169486363526564e-05,
+      "loss": 0.8006,
+      "step": 18994
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5469121477537036,
+      "learning_rate": 1.2168741489484656e-05,
+      "loss": 0.6803,
+      "step": 18995
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4474730399115403,
+      "learning_rate": 1.2167996602815652e-05,
+      "loss": 0.7079,
+      "step": 18996
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4790412446154708,
+      "learning_rate": 1.2167251703523886e-05,
+      "loss": 0.7021,
+      "step": 18997
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.592127786193951,
+      "learning_rate": 1.216650679161369e-05,
+      "loss": 0.5542,
+      "step": 18998
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.669913003539619,
+      "learning_rate": 1.2165761867089411e-05,
+      "loss": 0.7666,
+      "step": 18999
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.199909895069516,
+      "learning_rate": 1.2165016929955376e-05,
+      "loss": 0.7499,
+      "step": 19000
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4346527202696362,
+      "learning_rate": 1.2164271980215931e-05,
+      "loss": 0.6279,
+      "step": 19001
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.408349232378826,
+      "learning_rate": 1.2163527017875406e-05,
+      "loss": 0.7513,
+      "step": 19002
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8102242324393765,
+      "learning_rate": 1.2162782042938144e-05,
+      "loss": 0.6229,
+      "step": 19003
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3456136356096824,
+      "learning_rate": 1.216203705540848e-05,
+      "loss": 0.522,
+      "step": 19004
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.354609543864857,
+      "learning_rate": 1.2161292055290751e-05,
+      "loss": 0.7227,
+      "step": 19005
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4232662602922552,
+      "learning_rate": 1.2160547042589294e-05,
+      "loss": 0.811,
+      "step": 19006
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5342999809964888,
+      "learning_rate": 1.2159802017308451e-05,
+      "loss": 0.8077,
+      "step": 19007
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.45126290951542,
+      "learning_rate": 1.2159056979452555e-05,
+      "loss": 0.7805,
+      "step": 19008
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6499959193955396,
+      "learning_rate": 1.2158311929025945e-05,
+      "loss": 0.5875,
+      "step": 19009
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.40390801476779,
+      "learning_rate": 1.215756686603296e-05,
+      "loss": 0.7615,
+      "step": 19010
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.333460092189524,
+      "learning_rate": 1.2156821790477939e-05,
+      "loss": 0.6759,
+      "step": 19011
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.587594931675421,
+      "learning_rate": 1.2156076702365219e-05,
+      "loss": 0.6896,
+      "step": 19012
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5062503156479874,
+      "learning_rate": 1.2155331601699136e-05,
+      "loss": 0.6949,
+      "step": 19013
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5843164457062802,
+      "learning_rate": 1.2154586488484033e-05,
+      "loss": 0.7798,
+      "step": 19014
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.352169184986711,
+      "learning_rate": 1.2153841362724244e-05,
+      "loss": 0.7188,
+      "step": 19015
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5849273615321249,
+      "learning_rate": 1.2153096224424106e-05,
+      "loss": 0.7592,
+      "step": 19016
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4467334598738382,
+      "learning_rate": 1.2152351073587963e-05,
+      "loss": 0.7312,
+      "step": 19017
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.218239804508003,
+      "learning_rate": 1.2151605910220153e-05,
+      "loss": 0.7692,
+      "step": 19018
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.328979188147386,
+      "learning_rate": 1.2150860734325005e-05,
+      "loss": 0.7589,
+      "step": 19019
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3665327428360239,
+      "learning_rate": 1.2150115545906871e-05,
+      "loss": 0.63,
+      "step": 19020
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5610508431016944,
+      "learning_rate": 1.2149370344970079e-05,
+      "loss": 0.4963,
+      "step": 19021
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.683004521910899,
+      "learning_rate": 1.2148625131518976e-05,
+      "loss": 0.7492,
+      "step": 19022
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3469237897714579,
+      "learning_rate": 1.2147879905557896e-05,
+      "loss": 0.6731,
+      "step": 19023
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.205421344874038,
+      "learning_rate": 1.2147134667091181e-05,
+      "loss": 0.7326,
+      "step": 19024
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5432292301697683,
+      "learning_rate": 1.2146389416123166e-05,
+      "loss": 0.686,
+      "step": 19025
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9427468253637683,
+      "learning_rate": 1.2145644152658191e-05,
+      "loss": 0.6625,
+      "step": 19026
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.541082993897666,
+      "learning_rate": 1.2144898876700599e-05,
+      "loss": 0.546,
+      "step": 19027
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5552544030103557,
+      "learning_rate": 1.2144153588254723e-05,
+      "loss": 0.8388,
+      "step": 19028
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.314589631291123,
+      "learning_rate": 1.2143408287324912e-05,
+      "loss": 0.792,
+      "step": 19029
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4054188476751792,
+      "learning_rate": 1.2142662973915494e-05,
+      "loss": 0.7789,
+      "step": 19030
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.230236823151801,
+      "learning_rate": 1.214191764803082e-05,
+      "loss": 0.7597,
+      "step": 19031
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.445023739967234,
+      "learning_rate": 1.2141172309675218e-05,
+      "loss": 0.8838,
+      "step": 19032
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5377241868496248,
+      "learning_rate": 1.2140426958853034e-05,
+      "loss": 0.6786,
+      "step": 19033
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3302740808235132,
+      "learning_rate": 1.2139681595568606e-05,
+      "loss": 0.7031,
+      "step": 19034
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6116318828289034,
+      "learning_rate": 1.2138936219826273e-05,
+      "loss": 0.7947,
+      "step": 19035
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4059948598272791,
+      "learning_rate": 1.2138190831630381e-05,
+      "loss": 0.7822,
+      "step": 19036
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4062485172405472,
+      "learning_rate": 1.2137445430985263e-05,
+      "loss": 0.754,
+      "step": 19037
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3004903035350643,
+      "learning_rate": 1.213670001789526e-05,
+      "loss": 0.7454,
+      "step": 19038
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4386299042707738,
+      "learning_rate": 1.2135954592364712e-05,
+      "loss": 0.7725,
+      "step": 19039
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4341395225218112,
+      "learning_rate": 1.2135209154397962e-05,
+      "loss": 0.7504,
+      "step": 19040
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5717648709411967,
+      "learning_rate": 1.2134463703999346e-05,
+      "loss": 0.5061,
+      "step": 19041
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6360268382080185,
+      "learning_rate": 1.2133718241173209e-05,
+      "loss": 0.6302,
+      "step": 19042
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5182671205659226,
+      "learning_rate": 1.213297276592389e-05,
+      "loss": 0.7491,
+      "step": 19043
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4568201789291813,
+      "learning_rate": 1.2132227278255727e-05,
+      "loss": 0.857,
+      "step": 19044
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.609340722985767,
+      "learning_rate": 1.213148177817306e-05,
+      "loss": 0.6544,
+      "step": 19045
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9151455047292434,
+      "learning_rate": 1.2130736265680231e-05,
+      "loss": 0.7227,
+      "step": 19046
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6567055604538377,
+      "learning_rate": 1.2129990740781583e-05,
+      "loss": 0.7292,
+      "step": 19047
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.731236819893599,
+      "learning_rate": 1.2129245203481457e-05,
+      "loss": 0.7441,
+      "step": 19048
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4208654589529184,
+      "learning_rate": 1.2128499653784184e-05,
+      "loss": 0.7243,
+      "step": 19049
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2154652949114584,
+      "learning_rate": 1.2127754091694119e-05,
+      "loss": 0.6697,
+      "step": 19050
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9285768215638812,
+      "learning_rate": 1.2127008517215593e-05,
+      "loss": 0.6772,
+      "step": 19051
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2004776567305717,
+      "learning_rate": 1.2126262930352953e-05,
+      "loss": 0.6585,
+      "step": 19052
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.331148749543964,
+      "learning_rate": 1.2125517331110532e-05,
+      "loss": 0.8642,
+      "step": 19053
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3207474716528573,
+      "learning_rate": 1.2124771719492681e-05,
+      "loss": 0.7563,
+      "step": 19054
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.30579320064233,
+      "learning_rate": 1.2124026095503735e-05,
+      "loss": 0.7677,
+      "step": 19055
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3025999269933202,
+      "learning_rate": 1.2123280459148034e-05,
+      "loss": 0.7285,
+      "step": 19056
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8026375878422776,
+      "learning_rate": 1.2122534810429922e-05,
+      "loss": 0.7929,
+      "step": 19057
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5739063371562888,
+      "learning_rate": 1.2121789149353745e-05,
+      "loss": 0.5632,
+      "step": 19058
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6010564544270878,
+      "learning_rate": 1.2121043475923834e-05,
+      "loss": 0.5014,
+      "step": 19059
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.623158535422267,
+      "learning_rate": 1.2120297790144539e-05,
+      "loss": 0.7906,
+      "step": 19060
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4314750278952897,
+      "learning_rate": 1.21195520920202e-05,
+      "loss": 0.7907,
+      "step": 19061
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5032381325618467,
+      "learning_rate": 1.2118806381555157e-05,
+      "loss": 0.7356,
+      "step": 19062
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.391765557087529,
+      "learning_rate": 1.2118060658753751e-05,
+      "loss": 0.7876,
+      "step": 19063
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9608265251815153,
+      "learning_rate": 1.2117314923620326e-05,
+      "loss": 0.733,
+      "step": 19064
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.640944003211319,
+      "learning_rate": 1.2116569176159225e-05,
+      "loss": 0.6801,
+      "step": 19065
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5668880791221464,
+      "learning_rate": 1.2115823416374788e-05,
+      "loss": 0.6533,
+      "step": 19066
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6322648784393434,
+      "learning_rate": 1.2115077644271356e-05,
+      "loss": 0.7535,
+      "step": 19067
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3544908085886,
+      "learning_rate": 1.2114331859853272e-05,
+      "loss": 0.8883,
+      "step": 19068
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3744138984995506,
+      "learning_rate": 1.211358606312488e-05,
+      "loss": 0.7631,
+      "step": 19069
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5296151591804379,
+      "learning_rate": 1.2112840254090519e-05,
+      "loss": 0.7622,
+      "step": 19070
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.4701829095403074,
+      "learning_rate": 1.2112094432754536e-05,
+      "loss": 0.7434,
+      "step": 19071
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4528202979908356,
+      "learning_rate": 1.2111348599121266e-05,
+      "loss": 0.7458,
+      "step": 19072
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6415701836181609,
+      "learning_rate": 1.211060275319506e-05,
+      "loss": 0.597,
+      "step": 19073
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.588135608315679,
+      "learning_rate": 1.2109856894980254e-05,
+      "loss": 0.7048,
+      "step": 19074
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6582652253478842,
+      "learning_rate": 1.2109111024481194e-05,
+      "loss": 0.7589,
+      "step": 19075
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.3695831201753625,
+      "learning_rate": 1.2108365141702224e-05,
+      "loss": 0.7097,
+      "step": 19076
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5941656705782281,
+      "learning_rate": 1.2107619246647679e-05,
+      "loss": 0.7954,
+      "step": 19077
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5616631126880052,
+      "learning_rate": 1.2106873339321914e-05,
+      "loss": 0.7644,
+      "step": 19078
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5799070878249125,
+      "learning_rate": 1.210612741972926e-05,
+      "loss": 0.517,
+      "step": 19079
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3960099398674177,
+      "learning_rate": 1.210538148787407e-05,
+      "loss": 0.6907,
+      "step": 19080
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4481624782617284,
+      "learning_rate": 1.210463554376068e-05,
+      "loss": 0.7408,
+      "step": 19081
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8770862602618008,
+      "learning_rate": 1.2103889587393437e-05,
+      "loss": 0.7131,
+      "step": 19082
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.756056543926484,
+      "learning_rate": 1.2103143618776682e-05,
+      "loss": 0.7903,
+      "step": 19083
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3746578440607973,
+      "learning_rate": 1.2102397637914758e-05,
+      "loss": 0.7457,
+      "step": 19084
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5803389332167996,
+      "learning_rate": 1.210165164481201e-05,
+      "loss": 0.6379,
+      "step": 19085
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4664254787122057,
+      "learning_rate": 1.210090563947278e-05,
+      "loss": 0.6505,
+      "step": 19086
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4554488865797932,
+      "learning_rate": 1.2100159621901413e-05,
+      "loss": 0.7639,
+      "step": 19087
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6383546316459859,
+      "learning_rate": 1.2099413592102255e-05,
+      "loss": 0.5413,
+      "step": 19088
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5874862843420067,
+      "learning_rate": 1.2098667550079642e-05,
+      "loss": 0.57,
+      "step": 19089
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2812728793702595,
+      "learning_rate": 1.2097921495837923e-05,
+      "loss": 0.8029,
+      "step": 19090
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8389934415829807,
+      "learning_rate": 1.209717542938144e-05,
+      "loss": 0.8547,
+      "step": 19091
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.354563889495642,
+      "learning_rate": 1.2096429350714538e-05,
+      "loss": 0.6661,
+      "step": 19092
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5172317198435785,
+      "learning_rate": 1.2095683259841562e-05,
+      "loss": 0.7113,
+      "step": 19093
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.555011945281625,
+      "learning_rate": 1.2094937156766855e-05,
+      "loss": 0.4976,
+      "step": 19094
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5233173968831035,
+      "learning_rate": 1.2094191041494759e-05,
+      "loss": 0.7633,
+      "step": 19095
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3299302436399394,
+      "learning_rate": 1.2093444914029622e-05,
+      "loss": 0.8269,
+      "step": 19096
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5749974946929763,
+      "learning_rate": 1.2092698774375786e-05,
+      "loss": 0.718,
+      "step": 19097
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3351237453988656,
+      "learning_rate": 1.209195262253759e-05,
+      "loss": 0.7241,
+      "step": 19098
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4847110441827338,
+      "learning_rate": 1.2091206458519389e-05,
+      "loss": 0.7398,
+      "step": 19099
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.93682809117043,
+      "learning_rate": 1.2090460282325518e-05,
+      "loss": 0.7308,
+      "step": 19100
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4962744542323776,
+      "learning_rate": 1.208971409396033e-05,
+      "loss": 0.7601,
+      "step": 19101
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.197441474739545,
+      "learning_rate": 1.2088967893428161e-05,
+      "loss": 0.5752,
+      "step": 19102
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5417138476741687,
+      "learning_rate": 1.2088221680733362e-05,
+      "loss": 0.7429,
+      "step": 19103
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2908618243750238,
+      "learning_rate": 1.2087475455880272e-05,
+      "loss": 0.7801,
+      "step": 19104
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4417695617598185,
+      "learning_rate": 1.2086729218873242e-05,
+      "loss": 0.8242,
+      "step": 19105
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.438644433202857,
+      "learning_rate": 1.2085982969716615e-05,
+      "loss": 0.7038,
+      "step": 19106
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3781835950267995,
+      "learning_rate": 1.208523670841473e-05,
+      "loss": 0.8325,
+      "step": 19107
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3467561147053566,
+      "learning_rate": 1.2084490434971942e-05,
+      "loss": 0.8032,
+      "step": 19108
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4058060551180218,
+      "learning_rate": 1.2083744149392586e-05,
+      "loss": 0.7648,
+      "step": 19109
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5609198636425963,
+      "learning_rate": 1.2082997851681015e-05,
+      "loss": 0.6847,
+      "step": 19110
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.310293927706387,
+      "learning_rate": 1.2082251541841573e-05,
+      "loss": 0.7042,
+      "step": 19111
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2353114485412098,
+      "learning_rate": 1.2081505219878597e-05,
+      "loss": 0.6687,
+      "step": 19112
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6535752015156644,
+      "learning_rate": 1.2080758885796444e-05,
+      "loss": 0.6944,
+      "step": 19113
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0037148206342645,
+      "learning_rate": 1.208001253959945e-05,
+      "loss": 0.6592,
+      "step": 19114
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1981760250155957,
+      "learning_rate": 1.207926618129197e-05,
+      "loss": 0.7084,
+      "step": 19115
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3974430160001008,
+      "learning_rate": 1.207851981087834e-05,
+      "loss": 0.7083,
+      "step": 19116
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.388829953322943,
+      "learning_rate": 1.2077773428362909e-05,
+      "loss": 0.7098,
+      "step": 19117
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4850255836578596,
+      "learning_rate": 1.2077027033750025e-05,
+      "loss": 0.8111,
+      "step": 19118
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4390709792749146,
+      "learning_rate": 1.207628062704403e-05,
+      "loss": 0.7517,
+      "step": 19119
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4482241378310188,
+      "learning_rate": 1.2075534208249275e-05,
+      "loss": 0.719,
+      "step": 19120
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3045613481879492,
+      "learning_rate": 1.2074787777370101e-05,
+      "loss": 0.73,
+      "step": 19121
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5482095971794279,
+      "learning_rate": 1.2074041334410858e-05,
+      "loss": 0.7804,
+      "step": 19122
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4030740548972984,
+      "learning_rate": 1.2073294879375883e-05,
+      "loss": 0.7769,
+      "step": 19123
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8142605597230963,
+      "learning_rate": 1.2072548412269536e-05,
+      "loss": 0.7935,
+      "step": 19124
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5374604152383016,
+      "learning_rate": 1.2071801933096153e-05,
+      "loss": 0.8233,
+      "step": 19125
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.354465386588621,
+      "learning_rate": 1.2071055441860082e-05,
+      "loss": 0.801,
+      "step": 19126
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3120303714329205,
+      "learning_rate": 1.2070308938565674e-05,
+      "loss": 0.712,
+      "step": 19127
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3211769543148548,
+      "learning_rate": 1.2069562423217268e-05,
+      "loss": 0.7091,
+      "step": 19128
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5624941574114104,
+      "learning_rate": 1.2068815895819219e-05,
+      "loss": 0.7582,
+      "step": 19129
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5102412815364905,
+      "learning_rate": 1.2068069356375864e-05,
+      "loss": 0.7361,
+      "step": 19130
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6192880992004367,
+      "learning_rate": 1.2067322804891556e-05,
+      "loss": 0.5471,
+      "step": 19131
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5548798969850328,
+      "learning_rate": 1.206657624137064e-05,
+      "loss": 0.9012,
+      "step": 19132
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8522348613772903,
+      "learning_rate": 1.2065829665817465e-05,
+      "loss": 0.7386,
+      "step": 19133
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5244661408633302,
+      "learning_rate": 1.2065083078236375e-05,
+      "loss": 0.8394,
+      "step": 19134
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.537187688594843,
+      "learning_rate": 1.2064336478631717e-05,
+      "loss": 0.8607,
+      "step": 19135
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.927759015195561,
+      "learning_rate": 1.2063589867007838e-05,
+      "loss": 0.7375,
+      "step": 19136
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.313033940251771,
+      "learning_rate": 1.2062843243369086e-05,
+      "loss": 0.7362,
+      "step": 19137
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3965221858616184,
+      "learning_rate": 1.2062096607719809e-05,
+      "loss": 0.6696,
+      "step": 19138
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1663365988681575,
+      "learning_rate": 1.206134996006435e-05,
+      "loss": 0.6951,
+      "step": 19139
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4162895365632115,
+      "learning_rate": 1.206060330040706e-05,
+      "loss": 0.6555,
+      "step": 19140
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6098063383547578,
+      "learning_rate": 1.2059856628752285e-05,
+      "loss": 0.8288,
+      "step": 19141
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5674332770277883,
+      "learning_rate": 1.2059109945104373e-05,
+      "loss": 0.6831,
+      "step": 19142
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.235461482323983,
+      "learning_rate": 1.2058363249467671e-05,
+      "loss": 0.6617,
+      "step": 19143
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7584465910736788,
+      "learning_rate": 1.2057616541846526e-05,
+      "loss": 0.7902,
+      "step": 19144
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.50165299336751,
+      "learning_rate": 1.2056869822245285e-05,
+      "loss": 0.6445,
+      "step": 19145
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4613373087300554,
+      "learning_rate": 1.20561230906683e-05,
+      "loss": 0.7334,
+      "step": 19146
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5849781878919214,
+      "learning_rate": 1.2055376347119912e-05,
+      "loss": 0.6574,
+      "step": 19147
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7711928939178165,
+      "learning_rate": 1.2054629591604474e-05,
+      "loss": 0.7187,
+      "step": 19148
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5662092907183152,
+      "learning_rate": 1.2053882824126331e-05,
+      "loss": 0.5418,
+      "step": 19149
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.43698720709732,
+      "learning_rate": 1.2053136044689834e-05,
+      "loss": 0.7833,
+      "step": 19150
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5065346288509691,
+      "learning_rate": 1.2052389253299326e-05,
+      "loss": 0.7076,
+      "step": 19151
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3463292337787869,
+      "learning_rate": 1.2051642449959162e-05,
+      "loss": 0.7469,
+      "step": 19152
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7876421502461095,
+      "learning_rate": 1.2050895634673681e-05,
+      "loss": 0.715,
+      "step": 19153
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2615550793542278,
+      "learning_rate": 1.2050148807447242e-05,
+      "loss": 0.8068,
+      "step": 19154
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5585042057310776,
+      "learning_rate": 1.2049401968284182e-05,
+      "loss": 0.4935,
+      "step": 19155
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.292943276171126,
+      "learning_rate": 1.2048655117188859e-05,
+      "loss": 0.7707,
+      "step": 19156
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6063700072659316,
+      "learning_rate": 1.2047908254165614e-05,
+      "loss": 0.5499,
+      "step": 19157
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4785052953096893,
+      "learning_rate": 1.20471613792188e-05,
+      "loss": 0.5969,
+      "step": 19158
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4168357450976858,
+      "learning_rate": 1.2046414492352766e-05,
+      "loss": 0.7997,
+      "step": 19159
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.623222372141573,
+      "learning_rate": 1.2045667593571858e-05,
+      "loss": 0.848,
+      "step": 19160
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.4307813923813044,
+      "learning_rate": 1.2044920682880425e-05,
+      "loss": 0.7395,
+      "step": 19161
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4240261630924729,
+      "learning_rate": 1.2044173760282818e-05,
+      "loss": 0.6938,
+      "step": 19162
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3735141562373807,
+      "learning_rate": 1.2043426825783383e-05,
+      "loss": 0.707,
+      "step": 19163
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.263724328877641,
+      "learning_rate": 1.204267987938647e-05,
+      "loss": 0.6761,
+      "step": 19164
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.774810518640303,
+      "learning_rate": 1.2041932921096427e-05,
+      "loss": 0.6132,
+      "step": 19165
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7968826043299126,
+      "learning_rate": 1.2041185950917604e-05,
+      "loss": 0.6213,
+      "step": 19166
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.485134485155096,
+      "learning_rate": 1.2040438968854354e-05,
+      "loss": 0.6989,
+      "step": 19167
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4931114964567964,
+      "learning_rate": 1.2039691974911019e-05,
+      "loss": 0.7566,
+      "step": 19168
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5755277725681758,
+      "learning_rate": 1.2038944969091953e-05,
+      "loss": 0.7812,
+      "step": 19169
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2716520034379957,
+      "learning_rate": 1.20381979514015e-05,
+      "loss": 0.7717,
+      "step": 19170
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7220727043055746,
+      "learning_rate": 1.203745092184402e-05,
+      "loss": 0.7319,
+      "step": 19171
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.460869967651582,
+      "learning_rate": 1.203670388042385e-05,
+      "loss": 0.7295,
+      "step": 19172
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5990957581191668,
+      "learning_rate": 1.203595682714535e-05,
+      "loss": 0.5491,
+      "step": 19173
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5792443599256287,
+      "learning_rate": 1.2035209762012858e-05,
+      "loss": 0.7833,
+      "step": 19174
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6582976417511028,
+      "learning_rate": 1.2034462685030734e-05,
+      "loss": 0.5914,
+      "step": 19175
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3918289476158787,
+      "learning_rate": 1.2033715596203326e-05,
+      "loss": 0.8162,
+      "step": 19176
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8178859493185013,
+      "learning_rate": 1.203296849553498e-05,
+      "loss": 0.8087,
+      "step": 19177
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6781730364705898,
+      "learning_rate": 1.203222138303005e-05,
+      "loss": 0.7067,
+      "step": 19178
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1148521923037316,
+      "learning_rate": 1.2031474258692878e-05,
+      "loss": 0.7256,
+      "step": 19179
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.684692546376517,
+      "learning_rate": 1.2030727122527828e-05,
+      "loss": 0.7244,
+      "step": 19180
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4981005506243426,
+      "learning_rate": 1.2029979974539233e-05,
+      "loss": 0.7464,
+      "step": 19181
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4639159172138136,
+      "learning_rate": 1.202923281473146e-05,
+      "loss": 0.7501,
+      "step": 19182
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5471639396140997,
+      "learning_rate": 1.2028485643108848e-05,
+      "loss": 0.525,
+      "step": 19183
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.416213789682133,
+      "learning_rate": 1.202773845967575e-05,
+      "loss": 0.8235,
+      "step": 19184
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.44631727051546,
+      "learning_rate": 1.2026991264436516e-05,
+      "loss": 0.7879,
+      "step": 19185
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.428545539398235,
+      "learning_rate": 1.2026244057395497e-05,
+      "loss": 0.7422,
+      "step": 19186
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3267844054397864,
+      "learning_rate": 1.2025496838557045e-05,
+      "loss": 0.7828,
+      "step": 19187
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6019289998721589,
+      "learning_rate": 1.202474960792551e-05,
+      "loss": 0.6532,
+      "step": 19188
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5394086515728067,
+      "learning_rate": 1.2024002365505238e-05,
+      "loss": 0.6888,
+      "step": 19189
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5561466554283002,
+      "learning_rate": 1.2023255111300587e-05,
+      "loss": 0.5266,
+      "step": 19190
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5907856953171975,
+      "learning_rate": 1.2022507845315903e-05,
+      "loss": 0.7113,
+      "step": 19191
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.235138258334178,
+      "learning_rate": 1.2021760567555539e-05,
+      "loss": 0.6224,
+      "step": 19192
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5956729692820246,
+      "learning_rate": 1.2021013278023843e-05,
+      "loss": 0.5466,
+      "step": 19193
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4859593874658295,
+      "learning_rate": 1.2020265976725167e-05,
+      "loss": 0.7626,
+      "step": 19194
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.033989072635297,
+      "learning_rate": 1.2019518663663865e-05,
+      "loss": 0.7519,
+      "step": 19195
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6058029135178022,
+      "learning_rate": 1.2018771338844287e-05,
+      "loss": 0.7139,
+      "step": 19196
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7535491188782188,
+      "learning_rate": 1.201802400227078e-05,
+      "loss": 0.7939,
+      "step": 19197
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.4066011560785197,
+      "learning_rate": 1.2017276653947701e-05,
+      "loss": 0.7636,
+      "step": 19198
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5638633874640462,
+      "learning_rate": 1.20165292938794e-05,
+      "loss": 0.7934,
+      "step": 19199
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5195973847599327,
+      "learning_rate": 1.2015781922070223e-05,
+      "loss": 0.7668,
+      "step": 19200
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4181803010390737,
+      "learning_rate": 1.2015034538524528e-05,
+      "loss": 0.7757,
+      "step": 19201
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6698009565177774,
+      "learning_rate": 1.2014287143246662e-05,
+      "loss": 0.6642,
+      "step": 19202
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5577241544185024,
+      "learning_rate": 1.201353973624098e-05,
+      "loss": 0.4852,
+      "step": 19203
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7234630263518702,
+      "learning_rate": 1.201279231751183e-05,
+      "loss": 0.8062,
+      "step": 19204
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5624313283056608,
+      "learning_rate": 1.201204488706357e-05,
+      "loss": 0.7344,
+      "step": 19205
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4815314358793605,
+      "learning_rate": 1.2011297444900545e-05,
+      "loss": 0.6614,
+      "step": 19206
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3984744120824686,
+      "learning_rate": 1.2010549991027109e-05,
+      "loss": 0.6949,
+      "step": 19207
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5396083653111716,
+      "learning_rate": 1.2009802525447616e-05,
+      "loss": 0.8062,
+      "step": 19208
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2799223101836317,
+      "learning_rate": 1.2009055048166415e-05,
+      "loss": 0.7674,
+      "step": 19209
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.435515233150538,
+      "learning_rate": 1.2008307559187862e-05,
+      "loss": 0.6772,
+      "step": 19210
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3051742340815293,
+      "learning_rate": 1.2007560058516301e-05,
+      "loss": 0.7491,
+      "step": 19211
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4084538217693823,
+      "learning_rate": 1.2006812546156097e-05,
+      "loss": 0.8997,
+      "step": 19212
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3404907683338385,
+      "learning_rate": 1.2006065022111591e-05,
+      "loss": 0.8022,
+      "step": 19213
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3168271023605227,
+      "learning_rate": 1.200531748638714e-05,
+      "loss": 0.6744,
+      "step": 19214
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7632359673803857,
+      "learning_rate": 1.2004569938987093e-05,
+      "loss": 0.74,
+      "step": 19215
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3859127933613709,
+      "learning_rate": 1.200382237991581e-05,
+      "loss": 0.6193,
+      "step": 19216
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5741106542096786,
+      "learning_rate": 1.2003074809177635e-05,
+      "loss": 0.7509,
+      "step": 19217
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4833619187629759,
+      "learning_rate": 1.2002327226776926e-05,
+      "loss": 0.6167,
+      "step": 19218
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3886169476676786,
+      "learning_rate": 1.2001579632718034e-05,
+      "loss": 0.7281,
+      "step": 19219
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7460051495479747,
+      "learning_rate": 1.2000832027005314e-05,
+      "loss": 0.6848,
+      "step": 19220
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4597763509783126,
+      "learning_rate": 1.200008440964311e-05,
+      "loss": 0.7507,
+      "step": 19221
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6463393481615773,
+      "learning_rate": 1.1999336780635788e-05,
+      "loss": 0.5125,
+      "step": 19222
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2617408877643563,
+      "learning_rate": 1.199858913998769e-05,
+      "loss": 0.6612,
+      "step": 19223
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4048833096236144,
+      "learning_rate": 1.1997841487703175e-05,
+      "loss": 0.6765,
+      "step": 19224
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4625216962553307,
+      "learning_rate": 1.1997093823786594e-05,
+      "loss": 0.8968,
+      "step": 19225
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5828365878654851,
+      "learning_rate": 1.1996346148242299e-05,
+      "loss": 0.5144,
+      "step": 19226
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.291351673369099,
+      "learning_rate": 1.1995598461074647e-05,
+      "loss": 0.7164,
+      "step": 19227
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4206962945330182,
+      "learning_rate": 1.1994850762287988e-05,
+      "loss": 0.6619,
+      "step": 19228
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7721484385950532,
+      "learning_rate": 1.199410305188668e-05,
+      "loss": 0.8329,
+      "step": 19229
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5686935372775946,
+      "learning_rate": 1.1993355329875067e-05,
+      "loss": 0.4976,
+      "step": 19230
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5263389291644573,
+      "learning_rate": 1.1992607596257513e-05,
+      "loss": 0.7761,
+      "step": 19231
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5897517207853848,
+      "learning_rate": 1.1991859851038362e-05,
+      "loss": 0.6137,
+      "step": 19232
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2720516829185897,
+      "learning_rate": 1.1991112094221975e-05,
+      "loss": 0.7027,
+      "step": 19233
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3346109812874596,
+      "learning_rate": 1.19903643258127e-05,
+      "loss": 0.6723,
+      "step": 19234
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6235843358206138,
+      "learning_rate": 1.1989616545814899e-05,
+      "loss": 0.5531,
+      "step": 19235
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.367670245156835,
+      "learning_rate": 1.198886875423292e-05,
+      "loss": 0.7948,
+      "step": 19236
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7312335852767418,
+      "learning_rate": 1.1988120951071113e-05,
+      "loss": 0.7412,
+      "step": 19237
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6061238362287257,
+      "learning_rate": 1.198737313633384e-05,
+      "loss": 0.5673,
+      "step": 19238
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3916195897348485,
+      "learning_rate": 1.1986625310025451e-05,
+      "loss": 0.7299,
+      "step": 19239
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5102245471601556,
+      "learning_rate": 1.1985877472150299e-05,
+      "loss": 0.7497,
+      "step": 19240
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8461902729731887,
+      "learning_rate": 1.1985129622712738e-05,
+      "loss": 0.689,
+      "step": 19241
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6244808793956684,
+      "learning_rate": 1.1984381761717127e-05,
+      "loss": 0.7236,
+      "step": 19242
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7450361898291182,
+      "learning_rate": 1.1983633889167815e-05,
+      "loss": 0.7833,
+      "step": 19243
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5808227753810622,
+      "learning_rate": 1.198288600506916e-05,
+      "loss": 0.6376,
+      "step": 19244
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9262312276855067,
+      "learning_rate": 1.1982138109425516e-05,
+      "loss": 0.7489,
+      "step": 19245
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5319029465706961,
+      "learning_rate": 1.1981390202241235e-05,
+      "loss": 0.6556,
+      "step": 19246
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.627091121084685,
+      "learning_rate": 1.1980642283520672e-05,
+      "loss": 0.7531,
+      "step": 19247
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5638833569589632,
+      "learning_rate": 1.1979894353268185e-05,
+      "loss": 0.534,
+      "step": 19248
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6878260431623004,
+      "learning_rate": 1.1979146411488122e-05,
+      "loss": 0.7001,
+      "step": 19249
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6205083381130978,
+      "learning_rate": 1.1978398458184848e-05,
+      "loss": 0.8223,
+      "step": 19250
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.686308895529327,
+      "learning_rate": 1.1977650493362708e-05,
+      "loss": 0.686,
+      "step": 19251
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6230279137779625,
+      "learning_rate": 1.1976902517026063e-05,
+      "loss": 0.4845,
+      "step": 19252
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3594978982157866,
+      "learning_rate": 1.1976154529179263e-05,
+      "loss": 0.6831,
+      "step": 19253
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5555507158451507,
+      "learning_rate": 1.1975406529826668e-05,
+      "loss": 0.6524,
+      "step": 19254
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3351980266883567,
+      "learning_rate": 1.1974658518972625e-05,
+      "loss": 0.8331,
+      "step": 19255
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6263952751242154,
+      "learning_rate": 1.1973910496621503e-05,
+      "loss": 0.6622,
+      "step": 19256
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4272302523258142,
+      "learning_rate": 1.1973162462777645e-05,
+      "loss": 0.8901,
+      "step": 19257
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.594680376836392,
+      "learning_rate": 1.197241441744541e-05,
+      "loss": 0.7126,
+      "step": 19258
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.385359585991779,
+      "learning_rate": 1.1971666360629157e-05,
+      "loss": 0.8489,
+      "step": 19259
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6050454276264661,
+      "learning_rate": 1.1970918292333234e-05,
+      "loss": 0.5681,
+      "step": 19260
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3980558793475963,
+      "learning_rate": 1.1970170212562004e-05,
+      "loss": 0.7009,
+      "step": 19261
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3508819297235697,
+      "learning_rate": 1.1969422121319815e-05,
+      "loss": 0.7193,
+      "step": 19262
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2950110943315776,
+      "learning_rate": 1.1968674018611033e-05,
+      "loss": 0.8211,
+      "step": 19263
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5311933727917915,
+      "learning_rate": 1.1967925904440005e-05,
+      "loss": 0.8131,
+      "step": 19264
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.646148919241489,
+      "learning_rate": 1.1967177778811087e-05,
+      "loss": 0.8102,
+      "step": 19265
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.185137771738265,
+      "learning_rate": 1.1966429641728641e-05,
+      "loss": 0.704,
+      "step": 19266
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6009604430270356,
+      "learning_rate": 1.1965681493197017e-05,
+      "loss": 0.5291,
+      "step": 19267
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.981360358475841,
+      "learning_rate": 1.1964933333220573e-05,
+      "loss": 0.6417,
+      "step": 19268
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4102654266952577,
+      "learning_rate": 1.1964185161803666e-05,
+      "loss": 0.7146,
+      "step": 19269
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6605813054666116,
+      "learning_rate": 1.196343697895065e-05,
+      "loss": 0.7484,
+      "step": 19270
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4538306525670206,
+      "learning_rate": 1.1962688784665883e-05,
+      "loss": 0.8003,
+      "step": 19271
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4502982922236631,
+      "learning_rate": 1.1961940578953719e-05,
+      "loss": 0.7085,
+      "step": 19272
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4585329536486225,
+      "learning_rate": 1.1961192361818517e-05,
+      "loss": 0.7444,
+      "step": 19273
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3982480243207387,
+      "learning_rate": 1.1960444133264632e-05,
+      "loss": 0.8023,
+      "step": 19274
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.573601218731636,
+      "learning_rate": 1.1959695893296421e-05,
+      "loss": 0.7136,
+      "step": 19275
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1357296836040143,
+      "learning_rate": 1.1958947641918239e-05,
+      "loss": 0.7198,
+      "step": 19276
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.379358575518988,
+      "learning_rate": 1.1958199379134445e-05,
+      "loss": 0.6915,
+      "step": 19277
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6063458794337999,
+      "learning_rate": 1.1957451104949396e-05,
+      "loss": 0.5514,
+      "step": 19278
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6249427079623184,
+      "learning_rate": 1.1956702819367442e-05,
+      "loss": 0.5139,
+      "step": 19279
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6152113173625886,
+      "learning_rate": 1.195595452239295e-05,
+      "loss": 0.5587,
+      "step": 19280
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8647440453796167,
+      "learning_rate": 1.1955206214030266e-05,
+      "loss": 0.7537,
+      "step": 19281
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2169577216027825,
+      "learning_rate": 1.1954457894283758e-05,
+      "loss": 0.679,
+      "step": 19282
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5222798910941902,
+      "learning_rate": 1.1953709563157771e-05,
+      "loss": 0.6043,
+      "step": 19283
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4722023374248903,
+      "learning_rate": 1.1952961220656676e-05,
+      "loss": 0.7144,
+      "step": 19284
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6022121598371721,
+      "learning_rate": 1.1952212866784819e-05,
+      "loss": 0.8069,
+      "step": 19285
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3702794385383976,
+      "learning_rate": 1.1951464501546559e-05,
+      "loss": 0.7264,
+      "step": 19286
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8398289257605078,
+      "learning_rate": 1.1950716124946256e-05,
+      "loss": 0.7422,
+      "step": 19287
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3190017220139407,
+      "learning_rate": 1.1949967736988265e-05,
+      "loss": 0.8557,
+      "step": 19288
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5410776209122874,
+      "learning_rate": 1.1949219337676948e-05,
+      "loss": 0.7599,
+      "step": 19289
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.464994724709054,
+      "learning_rate": 1.1948470927016656e-05,
+      "loss": 0.8522,
+      "step": 19290
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4505466773859184,
+      "learning_rate": 1.1947722505011749e-05,
+      "loss": 0.8021,
+      "step": 19291
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8068208353514688,
+      "learning_rate": 1.1946974071666587e-05,
+      "loss": 0.7052,
+      "step": 19292
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7419991348963477,
+      "learning_rate": 1.1946225626985523e-05,
+      "loss": 0.7993,
+      "step": 19293
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8691515090131217,
+      "learning_rate": 1.194547717097292e-05,
+      "loss": 0.7222,
+      "step": 19294
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5662801372228998,
+      "learning_rate": 1.194472870363313e-05,
+      "loss": 0.7339,
+      "step": 19295
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3866096763461835,
+      "learning_rate": 1.1943980224970516e-05,
+      "loss": 0.7443,
+      "step": 19296
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3969815363659233,
+      "learning_rate": 1.1943231734989434e-05,
+      "loss": 0.6409,
+      "step": 19297
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.308837239887988,
+      "learning_rate": 1.194248323369424e-05,
+      "loss": 0.7726,
+      "step": 19298
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3129439026364458,
+      "learning_rate": 1.1941734721089293e-05,
+      "loss": 0.6415,
+      "step": 19299
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2678231236538295,
+      "learning_rate": 1.1940986197178955e-05,
+      "loss": 0.8566,
+      "step": 19300
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6149626427446864,
+      "learning_rate": 1.1940237661967579e-05,
+      "loss": 0.5507,
+      "step": 19301
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6687533504197823,
+      "learning_rate": 1.1939489115459523e-05,
+      "loss": 0.5579,
+      "step": 19302
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4119946305724795,
+      "learning_rate": 1.1938740557659153e-05,
+      "loss": 0.7934,
+      "step": 19303
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5757063371261415,
+      "learning_rate": 1.1937991988570815e-05,
+      "loss": 0.7045,
+      "step": 19304
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4654993598198067,
+      "learning_rate": 1.193724340819888e-05,
+      "loss": 0.7869,
+      "step": 19305
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8739443184522655,
+      "learning_rate": 1.1936494816547695e-05,
+      "loss": 0.681,
+      "step": 19306
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.53400767203616,
+      "learning_rate": 1.1935746213621629e-05,
+      "loss": 0.8908,
+      "step": 19307
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8767201198186356,
+      "learning_rate": 1.1934997599425034e-05,
+      "loss": 0.8473,
+      "step": 19308
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.355865502406316,
+      "learning_rate": 1.1934248973962266e-05,
+      "loss": 0.7178,
+      "step": 19309
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4829371106698814,
+      "learning_rate": 1.1933500337237693e-05,
+      "loss": 0.691,
+      "step": 19310
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3549423142164696,
+      "learning_rate": 1.1932751689255666e-05,
+      "loss": 0.763,
+      "step": 19311
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4576758896183848,
+      "learning_rate": 1.1932003030020552e-05,
+      "loss": 0.7552,
+      "step": 19312
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9781506281945807,
+      "learning_rate": 1.1931254359536698e-05,
+      "loss": 0.7029,
+      "step": 19313
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5102310984173037,
+      "learning_rate": 1.1930505677808476e-05,
+      "loss": 0.8244,
+      "step": 19314
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5950030783917333,
+      "learning_rate": 1.1929756984840236e-05,
+      "loss": 0.7317,
+      "step": 19315
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4189244285483726,
+      "learning_rate": 1.1929008280636338e-05,
+      "loss": 0.7214,
+      "step": 19316
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6279209920815998,
+      "learning_rate": 1.1928259565201145e-05,
+      "loss": 0.7372,
+      "step": 19317
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4262067484550074,
+      "learning_rate": 1.1927510838539014e-05,
+      "loss": 0.6961,
+      "step": 19318
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3437959503019858,
+      "learning_rate": 1.1926762100654306e-05,
+      "loss": 0.6865,
+      "step": 19319
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3058862443396404,
+      "learning_rate": 1.1926013351551375e-05,
+      "loss": 0.6919,
+      "step": 19320
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6619834726132313,
+      "learning_rate": 1.1925264591234588e-05,
+      "loss": 0.5898,
+      "step": 19321
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1254281254858762,
+      "learning_rate": 1.19245158197083e-05,
+      "loss": 0.7671,
+      "step": 19322
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7166150713089139,
+      "learning_rate": 1.1923767036976872e-05,
+      "loss": 0.8349,
+      "step": 19323
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.390268599688645,
+      "learning_rate": 1.1923018243044664e-05,
+      "loss": 0.8493,
+      "step": 19324
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.9980775109465,
+      "learning_rate": 1.1922269437916035e-05,
+      "loss": 0.6474,
+      "step": 19325
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3977451953794409,
+      "learning_rate": 1.1921520621595342e-05,
+      "loss": 0.9062,
+      "step": 19326
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.422565530199381,
+      "learning_rate": 1.1920771794086949e-05,
+      "loss": 0.7308,
+      "step": 19327
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.561073186794153,
+      "learning_rate": 1.1920022955395217e-05,
+      "loss": 0.982,
+      "step": 19328
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.358761942980462,
+      "learning_rate": 1.1919274105524501e-05,
+      "loss": 0.7127,
+      "step": 19329
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7199588744346976,
+      "learning_rate": 1.1918525244479161e-05,
+      "loss": 0.7052,
+      "step": 19330
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6105921762374182,
+      "learning_rate": 1.1917776372263568e-05,
+      "loss": 0.7653,
+      "step": 19331
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4925733163543884,
+      "learning_rate": 1.1917027488882066e-05,
+      "loss": 0.7191,
+      "step": 19332
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4609597732512611,
+      "learning_rate": 1.1916278594339026e-05,
+      "loss": 0.5905,
+      "step": 19333
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3612856006227496,
+      "learning_rate": 1.1915529688638803e-05,
+      "loss": 0.8703,
+      "step": 19334
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5592850608701014,
+      "learning_rate": 1.1914780771785764e-05,
+      "loss": 0.5757,
+      "step": 19335
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4484405006607606,
+      "learning_rate": 1.1914031843784262e-05,
+      "loss": 0.6813,
+      "step": 19336
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4841942194267739,
+      "learning_rate": 1.1913282904638661e-05,
+      "loss": 0.7884,
+      "step": 19337
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1159479725275894,
+      "learning_rate": 1.1912533954353323e-05,
+      "loss": 0.7585,
+      "step": 19338
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5279513014321036,
+      "learning_rate": 1.1911784992932605e-05,
+      "loss": 0.7229,
+      "step": 19339
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3932471639355875,
+      "learning_rate": 1.1911036020380868e-05,
+      "loss": 0.9077,
+      "step": 19340
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5839313729612299,
+      "learning_rate": 1.1910287036702478e-05,
+      "loss": 0.7441,
+      "step": 19341
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4309671466333744,
+      "learning_rate": 1.1909538041901789e-05,
+      "loss": 0.7847,
+      "step": 19342
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5940099937545545,
+      "learning_rate": 1.1908789035983166e-05,
+      "loss": 0.4878,
+      "step": 19343
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5624873749029198,
+      "learning_rate": 1.1908040018950968e-05,
+      "loss": 0.7911,
+      "step": 19344
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3876179295246571,
+      "learning_rate": 1.190729099080956e-05,
+      "loss": 0.7298,
+      "step": 19345
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4602164475396666,
+      "learning_rate": 1.1906541951563295e-05,
+      "loss": 0.818,
+      "step": 19346
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.240140712109502,
+      "learning_rate": 1.1905792901216544e-05,
+      "loss": 0.6718,
+      "step": 19347
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5637712463303721,
+      "learning_rate": 1.190504383977366e-05,
+      "loss": 0.6939,
+      "step": 19348
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8358973000790964,
+      "learning_rate": 1.1904294767239009e-05,
+      "loss": 0.6897,
+      "step": 19349
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2183682641542375,
+      "learning_rate": 1.1903545683616953e-05,
+      "loss": 0.6523,
+      "step": 19350
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.407037208336658,
+      "learning_rate": 1.1902796588911847e-05,
+      "loss": 0.7701,
+      "step": 19351
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6195975974735082,
+      "learning_rate": 1.1902047483128063e-05,
+      "loss": 0.5284,
+      "step": 19352
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4087048243795945,
+      "learning_rate": 1.190129836626995e-05,
+      "loss": 0.7781,
+      "step": 19353
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5532662742453975,
+      "learning_rate": 1.1900549238341881e-05,
+      "loss": 0.6095,
+      "step": 19354
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5366978482588878,
+      "learning_rate": 1.1899800099348209e-05,
+      "loss": 0.7875,
+      "step": 19355
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2910147646774521,
+      "learning_rate": 1.1899050949293303e-05,
+      "loss": 0.7221,
+      "step": 19356
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8379679234807178,
+      "learning_rate": 1.1898301788181515e-05,
+      "loss": 0.7921,
+      "step": 19357
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9738217743398894,
+      "learning_rate": 1.1897552616017216e-05,
+      "loss": 0.7899,
+      "step": 19358
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.485122017660915,
+      "learning_rate": 1.1896803432804768e-05,
+      "loss": 0.7739,
+      "step": 19359
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.382334176900663,
+      "learning_rate": 1.1896054238548527e-05,
+      "loss": 0.7401,
+      "step": 19360
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4506751109370029,
+      "learning_rate": 1.189530503325286e-05,
+      "loss": 0.7904,
+      "step": 19361
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9677327715077007,
+      "learning_rate": 1.1894555816922122e-05,
+      "loss": 0.6126,
+      "step": 19362
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4919807092398258,
+      "learning_rate": 1.1893806589560689e-05,
+      "loss": 0.7127,
+      "step": 19363
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4254768458964358,
+      "learning_rate": 1.1893057351172905e-05,
+      "loss": 0.741,
+      "step": 19364
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0168799512619393,
+      "learning_rate": 1.1892308101763148e-05,
+      "loss": 0.8056,
+      "step": 19365
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4684184353388239,
+      "learning_rate": 1.1891558841335773e-05,
+      "loss": 0.8128,
+      "step": 19366
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.353724632170608,
+      "learning_rate": 1.189080956989514e-05,
+      "loss": 0.7238,
+      "step": 19367
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.043263257802141,
+      "learning_rate": 1.189006028744562e-05,
+      "loss": 0.8204,
+      "step": 19368
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.718508985368617,
+      "learning_rate": 1.1889310993991567e-05,
+      "loss": 0.6317,
+      "step": 19369
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4047489270562816,
+      "learning_rate": 1.188856168953735e-05,
+      "loss": 0.7367,
+      "step": 19370
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.674549906582209,
+      "learning_rate": 1.1887812374087328e-05,
+      "loss": 0.6863,
+      "step": 19371
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.096755662847002,
+      "learning_rate": 1.1887063047645863e-05,
+      "loss": 0.7009,
+      "step": 19372
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5899582871770689,
+      "learning_rate": 1.1886313710217322e-05,
+      "loss": 0.5216,
+      "step": 19373
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.378503556177498,
+      "learning_rate": 1.1885564361806063e-05,
+      "loss": 0.7004,
+      "step": 19374
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5320240691971785,
+      "learning_rate": 1.1884815002416452e-05,
+      "loss": 0.7678,
+      "step": 19375
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.5363079757804754,
+      "learning_rate": 1.1884065632052852e-05,
+      "loss": 0.8887,
+      "step": 19376
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3821763141859573,
+      "learning_rate": 1.1883316250719625e-05,
+      "loss": 0.7821,
+      "step": 19377
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4546395503932983,
+      "learning_rate": 1.1882566858421137e-05,
+      "loss": 0.8996,
+      "step": 19378
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4774530706950453,
+      "learning_rate": 1.1881817455161747e-05,
+      "loss": 0.7623,
+      "step": 19379
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4346295674022282,
+      "learning_rate": 1.188106804094582e-05,
+      "loss": 0.9037,
+      "step": 19380
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.688299859179843,
+      "learning_rate": 1.1880318615777718e-05,
+      "loss": 0.7207,
+      "step": 19381
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.9533241844169207,
+      "learning_rate": 1.1879569179661808e-05,
+      "loss": 0.6884,
+      "step": 19382
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.302130803027715,
+      "learning_rate": 1.1878819732602448e-05,
+      "loss": 0.6128,
+      "step": 19383
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6401152809167465,
+      "learning_rate": 1.1878070274604008e-05,
+      "loss": 0.5175,
+      "step": 19384
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5547335414259333,
+      "learning_rate": 1.1877320805670845e-05,
+      "loss": 0.6672,
+      "step": 19385
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9080587098960473,
+      "learning_rate": 1.1876571325807328e-05,
+      "loss": 0.6906,
+      "step": 19386
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.388947096139696,
+      "learning_rate": 1.1875821835017819e-05,
+      "loss": 0.6796,
+      "step": 19387
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7490482699715646,
+      "learning_rate": 1.1875072333306681e-05,
+      "loss": 0.7409,
+      "step": 19388
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7281526270517422,
+      "learning_rate": 1.1874322820678277e-05,
+      "loss": 0.6748,
+      "step": 19389
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6013450958681911,
+      "learning_rate": 1.1873573297136973e-05,
+      "loss": 0.5503,
+      "step": 19390
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.769047563862367,
+      "learning_rate": 1.1872823762687133e-05,
+      "loss": 0.7167,
+      "step": 19391
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6210393406027235,
+      "learning_rate": 1.1872074217333118e-05,
+      "loss": 0.8005,
+      "step": 19392
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4567514003011055,
+      "learning_rate": 1.1871324661079295e-05,
+      "loss": 0.7597,
+      "step": 19393
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.648674219852091,
+      "learning_rate": 1.1870575093930027e-05,
+      "loss": 0.6992,
+      "step": 19394
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9154818957251643,
+      "learning_rate": 1.1869825515889682e-05,
+      "loss": 0.7412,
+      "step": 19395
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4776420063583526,
+      "learning_rate": 1.1869075926962616e-05,
+      "loss": 0.766,
+      "step": 19396
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.347350356122319,
+      "learning_rate": 1.1868326327153202e-05,
+      "loss": 0.7401,
+      "step": 19397
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5475111149420717,
+      "learning_rate": 1.1867576716465799e-05,
+      "loss": 0.6539,
+      "step": 19398
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5725102171247982,
+      "learning_rate": 1.1866827094904773e-05,
+      "loss": 0.7863,
+      "step": 19399
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.2561824713143417,
+      "learning_rate": 1.186607746247449e-05,
+      "loss": 0.7864,
+      "step": 19400
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4675659807798271,
+      "learning_rate": 1.1865327819179313e-05,
+      "loss": 0.7252,
+      "step": 19401
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3492314510976835,
+      "learning_rate": 1.1864578165023604e-05,
+      "loss": 0.8483,
+      "step": 19402
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5000005632948314,
+      "learning_rate": 1.1863828500011736e-05,
+      "loss": 0.7242,
+      "step": 19403
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5249411355358893,
+      "learning_rate": 1.1863078824148062e-05,
+      "loss": 0.6774,
+      "step": 19404
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.644016256721175,
+      "learning_rate": 1.1862329137436958e-05,
+      "loss": 0.7516,
+      "step": 19405
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7376689357331114,
+      "learning_rate": 1.1861579439882779e-05,
+      "loss": 0.8431,
+      "step": 19406
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4166301024716672,
+      "learning_rate": 1.18608297314899e-05,
+      "loss": 0.7585,
+      "step": 19407
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4724523270672183,
+      "learning_rate": 1.1860080012262683e-05,
+      "loss": 0.8415,
+      "step": 19408
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.559320016159798,
+      "learning_rate": 1.1859330282205483e-05,
+      "loss": 0.8586,
+      "step": 19409
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5645323546071974,
+      "learning_rate": 1.1858580541322682e-05,
+      "loss": 0.7881,
+      "step": 19410
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6252775725606249,
+      "learning_rate": 1.1857830789618632e-05,
+      "loss": 0.5349,
+      "step": 19411
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3607512472124406,
+      "learning_rate": 1.1857081027097706e-05,
+      "loss": 0.6215,
+      "step": 19412
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3223536580527888,
+      "learning_rate": 1.1856331253764261e-05,
+      "loss": 0.6802,
+      "step": 19413
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6165839323707167,
+      "learning_rate": 1.1855581469622673e-05,
+      "loss": 0.7673,
+      "step": 19414
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5793193568459454,
+      "learning_rate": 1.1854831674677298e-05,
+      "loss": 0.5486,
+      "step": 19415
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5680906583898979,
+      "learning_rate": 1.185408186893251e-05,
+      "loss": 0.5562,
+      "step": 19416
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6130842765282851,
+      "learning_rate": 1.185333205239267e-05,
+      "loss": 0.853,
+      "step": 19417
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.467199189423636,
+      "learning_rate": 1.1852582225062141e-05,
+      "loss": 0.7181,
+      "step": 19418
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.744614828345143,
+      "learning_rate": 1.1851832386945294e-05,
+      "loss": 0.6361,
+      "step": 19419
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.582238760481771,
+      "learning_rate": 1.1851082538046492e-05,
+      "loss": 0.5269,
+      "step": 19420
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3725951911607062,
+      "learning_rate": 1.18503326783701e-05,
+      "loss": 0.5821,
+      "step": 19421
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6092766522564337,
+      "learning_rate": 1.1849582807920488e-05,
+      "loss": 0.5173,
+      "step": 19422
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4879933571426922,
+      "learning_rate": 1.1848832926702017e-05,
+      "loss": 0.7527,
+      "step": 19423
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6650377582414895,
+      "learning_rate": 1.1848083034719055e-05,
+      "loss": 0.8881,
+      "step": 19424
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9422822632597367,
+      "learning_rate": 1.1847333131975971e-05,
+      "loss": 0.7096,
+      "step": 19425
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.450663482700513,
+      "learning_rate": 1.1846583218477129e-05,
+      "loss": 0.6187,
+      "step": 19426
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7463851197219338,
+      "learning_rate": 1.1845833294226892e-05,
+      "loss": 0.7941,
+      "step": 19427
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.644037325557246,
+      "learning_rate": 1.184508335922963e-05,
+      "loss": 0.5773,
+      "step": 19428
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4782582989320427,
+      "learning_rate": 1.1844333413489709e-05,
+      "loss": 0.8296,
+      "step": 19429
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5393051860346414,
+      "learning_rate": 1.1843583457011497e-05,
+      "loss": 0.7782,
+      "step": 19430
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.379752226955083,
+      "learning_rate": 1.1842833489799356e-05,
+      "loss": 0.6812,
+      "step": 19431
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4899458911507368,
+      "learning_rate": 1.1842083511857654e-05,
+      "loss": 0.7874,
+      "step": 19432
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.098800123987136,
+      "learning_rate": 1.184133352319076e-05,
+      "loss": 0.6889,
+      "step": 19433
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5263084470982053,
+      "learning_rate": 1.184058352380304e-05,
+      "loss": 0.6933,
+      "step": 19434
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4153025940888873,
+      "learning_rate": 1.183983351369886e-05,
+      "loss": 0.8104,
+      "step": 19435
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6325060757542291,
+      "learning_rate": 1.1839083492882583e-05,
+      "loss": 0.7794,
+      "step": 19436
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5425258174254564,
+      "learning_rate": 1.1838333461358587e-05,
+      "loss": 0.6462,
+      "step": 19437
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.434577127792099,
+      "learning_rate": 1.1837583419131227e-05,
+      "loss": 0.9156,
+      "step": 19438
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3472507875279383,
+      "learning_rate": 1.1836833366204874e-05,
+      "loss": 0.7237,
+      "step": 19439
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7574912192121281,
+      "learning_rate": 1.1836083302583896e-05,
+      "loss": 0.6523,
+      "step": 19440
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2843209703402683,
+      "learning_rate": 1.183533322827266e-05,
+      "loss": 0.624,
+      "step": 19441
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5792134687056488,
+      "learning_rate": 1.1834583143275534e-05,
+      "loss": 0.5271,
+      "step": 19442
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6149869674213088,
+      "learning_rate": 1.1833833047596884e-05,
+      "loss": 0.5559,
+      "step": 19443
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6188822687710998,
+      "learning_rate": 1.1833082941241076e-05,
+      "loss": 0.6648,
+      "step": 19444
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7065993029327522,
+      "learning_rate": 1.1832332824212478e-05,
+      "loss": 0.8346,
+      "step": 19445
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4095424536079584,
+      "learning_rate": 1.183158269651546e-05,
+      "loss": 0.702,
+      "step": 19446
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5462203450024783,
+      "learning_rate": 1.1830832558154388e-05,
+      "loss": 0.6936,
+      "step": 19447
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3476055331487145,
+      "learning_rate": 1.1830082409133627e-05,
+      "loss": 0.8209,
+      "step": 19448
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.266584070552919,
+      "learning_rate": 1.182933224945755e-05,
+      "loss": 0.6392,
+      "step": 19449
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.029746527556026,
+      "learning_rate": 1.1828582079130519e-05,
+      "loss": 0.7492,
+      "step": 19450
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0508090227380924,
+      "learning_rate": 1.1827831898156905e-05,
+      "loss": 0.6618,
+      "step": 19451
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3244745538896787,
+      "learning_rate": 1.1827081706541076e-05,
+      "loss": 0.7917,
+      "step": 19452
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4499188761049762,
+      "learning_rate": 1.1826331504287396e-05,
+      "loss": 0.7662,
+      "step": 19453
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.516363469906226,
+      "learning_rate": 1.182558129140024e-05,
+      "loss": 0.7829,
+      "step": 19454
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5995628585750399,
+      "learning_rate": 1.1824831067883969e-05,
+      "loss": 0.7369,
+      "step": 19455
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3853829940948745,
+      "learning_rate": 1.1824080833742952e-05,
+      "loss": 0.6505,
+      "step": 19456
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.3364878916594076,
+      "learning_rate": 1.1823330588981562e-05,
+      "loss": 0.6809,
+      "step": 19457
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0561475144092687,
+      "learning_rate": 1.1822580333604163e-05,
+      "loss": 0.8379,
+      "step": 19458
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6840296116572893,
+      "learning_rate": 1.1821830067615128e-05,
+      "loss": 0.6972,
+      "step": 19459
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3749720085611064,
+      "learning_rate": 1.1821079791018815e-05,
+      "loss": 0.6322,
+      "step": 19460
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4430587796959515,
+      "learning_rate": 1.1820329503819604e-05,
+      "loss": 0.6134,
+      "step": 19461
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3510455239140993,
+      "learning_rate": 1.1819579206021854e-05,
+      "loss": 0.6923,
+      "step": 19462
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6737381696829616,
+      "learning_rate": 1.1818828897629941e-05,
+      "loss": 0.7069,
+      "step": 19463
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.699395442131309,
+      "learning_rate": 1.181807857864823e-05,
+      "loss": 0.7006,
+      "step": 19464
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5960122977453304,
+      "learning_rate": 1.1817328249081088e-05,
+      "loss": 0.8388,
+      "step": 19465
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3929351369323202,
+      "learning_rate": 1.1816577908932887e-05,
+      "loss": 0.9071,
+      "step": 19466
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7342898642456117,
+      "learning_rate": 1.1815827558207996e-05,
+      "loss": 0.7492,
+      "step": 19467
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4940634638891863,
+      "learning_rate": 1.181507719691078e-05,
+      "loss": 0.8261,
+      "step": 19468
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4149211117772758,
+      "learning_rate": 1.181432682504561e-05,
+      "loss": 0.7698,
+      "step": 19469
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3963203213354047,
+      "learning_rate": 1.1813576442616854e-05,
+      "loss": 0.7385,
+      "step": 19470
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5266800377799696,
+      "learning_rate": 1.1812826049628883e-05,
+      "loss": 0.7705,
+      "step": 19471
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.370601418388628,
+      "learning_rate": 1.1812075646086066e-05,
+      "loss": 0.7439,
+      "step": 19472
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4278896041272044,
+      "learning_rate": 1.1811325231992768e-05,
+      "loss": 0.6584,
+      "step": 19473
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2942949438521776,
+      "learning_rate": 1.1810574807353362e-05,
+      "loss": 0.7562,
+      "step": 19474
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2875622014974126,
+      "learning_rate": 1.1809824372172217e-05,
+      "loss": 0.6985,
+      "step": 19475
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3681774493675676,
+      "learning_rate": 1.18090739264537e-05,
+      "loss": 0.8229,
+      "step": 19476
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3318676555089513,
+      "learning_rate": 1.1808323470202185e-05,
+      "loss": 0.7399,
+      "step": 19477
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.550999628444814,
+      "learning_rate": 1.1807573003422035e-05,
+      "loss": 0.7132,
+      "step": 19478
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.825820115937883,
+      "learning_rate": 1.1806822526117626e-05,
+      "loss": 0.7076,
+      "step": 19479
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5507163859907276,
+      "learning_rate": 1.1806072038293322e-05,
+      "loss": 0.7443,
+      "step": 19480
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9695521677536185,
+      "learning_rate": 1.1805321539953493e-05,
+      "loss": 0.7447,
+      "step": 19481
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.712637778896956,
+      "learning_rate": 1.1804571031102516e-05,
+      "loss": 0.7355,
+      "step": 19482
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3753505399845882,
+      "learning_rate": 1.1803820511744749e-05,
+      "loss": 0.6829,
+      "step": 19483
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3651224690457757,
+      "learning_rate": 1.1803069981884573e-05,
+      "loss": 0.5557,
+      "step": 19484
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7821166696046833,
+      "learning_rate": 1.1802319441526348e-05,
+      "loss": 0.7889,
+      "step": 19485
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9601891613263092,
+      "learning_rate": 1.1801568890674453e-05,
+      "loss": 0.6299,
+      "step": 19486
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3313926235999125,
+      "learning_rate": 1.180081832933325e-05,
+      "loss": 0.7862,
+      "step": 19487
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.959748322793005,
+      "learning_rate": 1.1800067757507115e-05,
+      "loss": 0.8048,
+      "step": 19488
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.372029415892242,
+      "learning_rate": 1.1799317175200416e-05,
+      "loss": 0.7762,
+      "step": 19489
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3304086569467084,
+      "learning_rate": 1.1798566582417521e-05,
+      "loss": 0.7502,
+      "step": 19490
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4917077564714722,
+      "learning_rate": 1.1797815979162801e-05,
+      "loss": 0.7328,
+      "step": 19491
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4092045390409667,
+      "learning_rate": 1.179706536544063e-05,
+      "loss": 0.74,
+      "step": 19492
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.292263865560351,
+      "learning_rate": 1.1796314741255372e-05,
+      "loss": 0.6385,
+      "step": 19493
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3075340918617915,
+      "learning_rate": 1.1795564106611401e-05,
+      "loss": 0.7263,
+      "step": 19494
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.807068027366235,
+      "learning_rate": 1.1794813461513092e-05,
+      "loss": 0.7684,
+      "step": 19495
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5315445170189166,
+      "learning_rate": 1.1794062805964808e-05,
+      "loss": 0.6205,
+      "step": 19496
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5971721520211666,
+      "learning_rate": 1.1793312139970923e-05,
+      "loss": 0.8204,
+      "step": 19497
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.032721678578642,
+      "learning_rate": 1.1792561463535805e-05,
+      "loss": 0.8986,
+      "step": 19498
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4004976438249517,
+      "learning_rate": 1.1791810776663829e-05,
+      "loss": 0.561,
+      "step": 19499
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4772072539127235,
+      "learning_rate": 1.1791060079359362e-05,
+      "loss": 0.754,
+      "step": 19500
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5633737861296548,
+      "learning_rate": 1.1790309371626775e-05,
+      "loss": 0.7548,
+      "step": 19501
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.349754203712503,
+      "learning_rate": 1.1789558653470443e-05,
+      "loss": 0.7076,
+      "step": 19502
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4531774454128519,
+      "learning_rate": 1.1788807924894734e-05,
+      "loss": 0.6492,
+      "step": 19503
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7442994010841955,
+      "learning_rate": 1.1788057185904015e-05,
+      "loss": 0.7916,
+      "step": 19504
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4177788468706658,
+      "learning_rate": 1.1787306436502664e-05,
+      "loss": 0.6211,
+      "step": 19505
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.404835895791964,
+      "learning_rate": 1.1786555676695049e-05,
+      "loss": 0.6773,
+      "step": 19506
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.478607325053833,
+      "learning_rate": 1.178580490648554e-05,
+      "loss": 0.7683,
+      "step": 19507
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.595712649379673,
+      "learning_rate": 1.178505412587851e-05,
+      "loss": 0.5429,
+      "step": 19508
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5885847647757004,
+      "learning_rate": 1.1784303334878331e-05,
+      "loss": 0.5147,
+      "step": 19509
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8304723454940282,
+      "learning_rate": 1.1783552533489372e-05,
+      "loss": 0.7153,
+      "step": 19510
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4728479687457572,
+      "learning_rate": 1.1782801721716004e-05,
+      "loss": 0.6654,
+      "step": 19511
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.1108279270176036,
+      "learning_rate": 1.1782050899562605e-05,
+      "loss": 0.6645,
+      "step": 19512
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5820170003214024,
+      "learning_rate": 1.1781300067033534e-05,
+      "loss": 0.5439,
+      "step": 19513
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.362103598237454,
+      "learning_rate": 1.1780549224133177e-05,
+      "loss": 0.7331,
+      "step": 19514
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3630860214802858,
+      "learning_rate": 1.1779798370865893e-05,
+      "loss": 0.8636,
+      "step": 19515
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8489667317774832,
+      "learning_rate": 1.1779047507236064e-05,
+      "loss": 0.7821,
+      "step": 19516
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2221745805851076,
+      "learning_rate": 1.1778296633248056e-05,
+      "loss": 0.7115,
+      "step": 19517
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4340258923374132,
+      "learning_rate": 1.177754574890624e-05,
+      "loss": 0.6668,
+      "step": 19518
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4905672735733921,
+      "learning_rate": 1.1776794854214993e-05,
+      "loss": 0.7665,
+      "step": 19519
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4840407927491817,
+      "learning_rate": 1.1776043949178681e-05,
+      "loss": 0.8043,
+      "step": 19520
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7054123440570532,
+      "learning_rate": 1.177529303380168e-05,
+      "loss": 0.7669,
+      "step": 19521
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6678093647752705,
+      "learning_rate": 1.177454210808836e-05,
+      "loss": 0.7954,
+      "step": 19522
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1345086844658354,
+      "learning_rate": 1.1773791172043094e-05,
+      "loss": 0.808,
+      "step": 19523
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6605298546574798,
+      "learning_rate": 1.1773040225670257e-05,
+      "loss": 0.7851,
+      "step": 19524
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6370089258162421,
+      "learning_rate": 1.1772289268974215e-05,
+      "loss": 0.5428,
+      "step": 19525
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4535003512387374,
+      "learning_rate": 1.1771538301959346e-05,
+      "loss": 0.787,
+      "step": 19526
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4102657987826561,
+      "learning_rate": 1.1770787324630019e-05,
+      "loss": 0.6861,
+      "step": 19527
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4957859182526834,
+      "learning_rate": 1.1770036336990609e-05,
+      "loss": 0.8462,
+      "step": 19528
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7216612848141128,
+      "learning_rate": 1.1769285339045484e-05,
+      "loss": 0.6027,
+      "step": 19529
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7690164462610425,
+      "learning_rate": 1.1768534330799021e-05,
+      "loss": 0.8093,
+      "step": 19530
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9850013207852484,
+      "learning_rate": 1.1767783312255595e-05,
+      "loss": 0.7401,
+      "step": 19531
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6316156098064815,
+      "learning_rate": 1.176703228341957e-05,
+      "loss": 0.5634,
+      "step": 19532
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.722018578404178,
+      "learning_rate": 1.1766281244295328e-05,
+      "loss": 0.748,
+      "step": 19533
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5365115155390854,
+      "learning_rate": 1.1765530194887231e-05,
+      "loss": 0.7984,
+      "step": 19534
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4492487675210628,
+      "learning_rate": 1.1764779135199664e-05,
+      "loss": 0.8319,
+      "step": 19535
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6154839890613824,
+      "learning_rate": 1.1764028065236991e-05,
+      "loss": 0.5585,
+      "step": 19536
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.138871067453071,
+      "learning_rate": 1.1763276985003592e-05,
+      "loss": 0.6298,
+      "step": 19537
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8096732192046052,
+      "learning_rate": 1.1762525894503831e-05,
+      "loss": 0.7219,
+      "step": 19538
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7843460253887893,
+      "learning_rate": 1.1761774793742092e-05,
+      "loss": 0.741,
+      "step": 19539
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5664558168601224,
+      "learning_rate": 1.176102368272274e-05,
+      "loss": 0.7401,
+      "step": 19540
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6889537263012477,
+      "learning_rate": 1.1760272561450146e-05,
+      "loss": 0.7565,
+      "step": 19541
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.520040589551499,
+      "learning_rate": 1.1759521429928693e-05,
+      "loss": 0.8129,
+      "step": 19542
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3485285094235346,
+      "learning_rate": 1.1758770288162746e-05,
+      "loss": 0.6483,
+      "step": 19543
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3803462649979181,
+      "learning_rate": 1.1758019136156684e-05,
+      "loss": 0.7387,
+      "step": 19544
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6222447060377252,
+      "learning_rate": 1.1757267973914876e-05,
+      "loss": 0.6538,
+      "step": 19545
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3703081492477924,
+      "learning_rate": 1.17565168014417e-05,
+      "loss": 0.6321,
+      "step": 19546
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5338966326689325,
+      "learning_rate": 1.1755765618741526e-05,
+      "loss": 0.7447,
+      "step": 19547
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5139909270057614,
+      "learning_rate": 1.175501442581873e-05,
+      "loss": 0.769,
+      "step": 19548
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.558461562299602,
+      "learning_rate": 1.1754263222677681e-05,
+      "loss": 0.6954,
+      "step": 19549
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5663658550323447,
+      "learning_rate": 1.1753512009322758e-05,
+      "loss": 0.7214,
+      "step": 19550
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6011050100061823,
+      "learning_rate": 1.1752760785758332e-05,
+      "loss": 0.6164,
+      "step": 19551
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7488807656923324,
+      "learning_rate": 1.1752009551988781e-05,
+      "loss": 0.6177,
+      "step": 19552
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4414252421418017,
+      "learning_rate": 1.1751258308018473e-05,
+      "loss": 0.7741,
+      "step": 19553
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5262859945094578,
+      "learning_rate": 1.1750507053851784e-05,
+      "loss": 0.8148,
+      "step": 19554
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.493813988625946,
+      "learning_rate": 1.1749755789493092e-05,
+      "loss": 0.6971,
+      "step": 19555
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3242730355098438,
+      "learning_rate": 1.1749004514946765e-05,
+      "loss": 0.7543,
+      "step": 19556
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4106079108280318,
+      "learning_rate": 1.174825323021718e-05,
+      "loss": 0.6913,
+      "step": 19557
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.461354251366708,
+      "learning_rate": 1.1747501935308712e-05,
+      "loss": 0.8598,
+      "step": 19558
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5567243858006743,
+      "learning_rate": 1.1746750630225734e-05,
+      "loss": 0.7297,
+      "step": 19559
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.7561234068491074,
+      "learning_rate": 1.1745999314972623e-05,
+      "loss": 0.6803,
+      "step": 19560
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4860204180254346,
+      "learning_rate": 1.174524798955375e-05,
+      "loss": 0.7671,
+      "step": 19561
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.27706181267758,
+      "learning_rate": 1.1744496653973488e-05,
+      "loss": 0.6892,
+      "step": 19562
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.850905667952998,
+      "learning_rate": 1.1743745308236218e-05,
+      "loss": 0.8056,
+      "step": 19563
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6093355804492736,
+      "learning_rate": 1.1742993952346307e-05,
+      "loss": 0.5501,
+      "step": 19564
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4881639280252066,
+      "learning_rate": 1.174224258630814e-05,
+      "loss": 0.7519,
+      "step": 19565
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5822306202540797,
+      "learning_rate": 1.1741491210126077e-05,
+      "loss": 0.7683,
+      "step": 19566
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0601714920608387,
+      "learning_rate": 1.1740739823804506e-05,
+      "loss": 0.5978,
+      "step": 19567
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7349469071933,
+      "learning_rate": 1.1739988427347795e-05,
+      "loss": 0.7098,
+      "step": 19568
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.798721597788486,
+      "learning_rate": 1.1739237020760319e-05,
+      "loss": 0.7838,
+      "step": 19569
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3979595771369813,
+      "learning_rate": 1.1738485604046457e-05,
+      "loss": 0.826,
+      "step": 19570
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8903363601003442,
+      "learning_rate": 1.1737734177210581e-05,
+      "loss": 0.7579,
+      "step": 19571
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.393527925029016,
+      "learning_rate": 1.1736982740257065e-05,
+      "loss": 0.676,
+      "step": 19572
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5288942674770398,
+      "learning_rate": 1.1736231293190286e-05,
+      "loss": 0.679,
+      "step": 19573
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4885932741727508,
+      "learning_rate": 1.1735479836014621e-05,
+      "loss": 0.7372,
+      "step": 19574
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5404630223265665,
+      "learning_rate": 1.1734728368734441e-05,
+      "loss": 0.7263,
+      "step": 19575
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5838923615031277,
+      "learning_rate": 1.1733976891354122e-05,
+      "loss": 0.5268,
+      "step": 19576
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6293279839754866,
+      "learning_rate": 1.1733225403878042e-05,
+      "loss": 0.8006,
+      "step": 19577
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8065688868968275,
+      "learning_rate": 1.1732473906310576e-05,
+      "loss": 0.6832,
+      "step": 19578
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8569123541724175,
+      "learning_rate": 1.1731722398656096e-05,
+      "loss": 0.7149,
+      "step": 19579
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3244049808811296,
+      "learning_rate": 1.1730970880918981e-05,
+      "loss": 0.6851,
+      "step": 19580
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6064578185005209,
+      "learning_rate": 1.1730219353103604e-05,
+      "loss": 0.5239,
+      "step": 19581
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.5116254068935744,
+      "learning_rate": 1.1729467815214348e-05,
+      "loss": 0.763,
+      "step": 19582
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4075552474222652,
+      "learning_rate": 1.1728716267255575e-05,
+      "loss": 0.6637,
+      "step": 19583
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4496830016059388,
+      "learning_rate": 1.1727964709231676e-05,
+      "loss": 0.7848,
+      "step": 19584
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6292639329628786,
+      "learning_rate": 1.1727213141147013e-05,
+      "loss": 0.6619,
+      "step": 19585
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4005774559008521,
+      "learning_rate": 1.1726461563005973e-05,
+      "loss": 0.6785,
+      "step": 19586
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5657155267899032,
+      "learning_rate": 1.1725709974812922e-05,
+      "loss": 0.4912,
+      "step": 19587
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.722997799850947,
+      "learning_rate": 1.1724958376572246e-05,
+      "loss": 0.7877,
+      "step": 19588
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7738274298752967,
+      "learning_rate": 1.1724206768288316e-05,
+      "loss": 0.8111,
+      "step": 19589
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5433209566158173,
+      "learning_rate": 1.1723455149965504e-05,
+      "loss": 0.704,
+      "step": 19590
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7438453715374136,
+      "learning_rate": 1.1722703521608197e-05,
+      "loss": 0.79,
+      "step": 19591
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.284567735608367,
+      "learning_rate": 1.1721951883220756e-05,
+      "loss": 0.7691,
+      "step": 19592
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4702845459092355,
+      "learning_rate": 1.1721200234807574e-05,
+      "loss": 0.6528,
+      "step": 19593
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.378740062341317,
+      "learning_rate": 1.1720448576373013e-05,
+      "loss": 0.7057,
+      "step": 19594
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5623097064977354,
+      "learning_rate": 1.1719696907921463e-05,
+      "loss": 0.6475,
+      "step": 19595
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7623931685582626,
+      "learning_rate": 1.171894522945729e-05,
+      "loss": 0.633,
+      "step": 19596
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.006076092578732,
+      "learning_rate": 1.1718193540984873e-05,
+      "loss": 0.6051,
+      "step": 19597
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5585735751350003,
+      "learning_rate": 1.1717441842508588e-05,
+      "loss": 0.8007,
+      "step": 19598
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5160049070525479,
+      "learning_rate": 1.1716690134032813e-05,
+      "loss": 0.808,
+      "step": 19599
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7500545848795017,
+      "learning_rate": 1.1715938415561928e-05,
+      "loss": 0.7188,
+      "step": 19600
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5277457636254868,
+      "learning_rate": 1.1715186687100302e-05,
+      "loss": 0.8277,
+      "step": 19601
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3183849628270048,
+      "learning_rate": 1.1714434948652317e-05,
+      "loss": 0.6618,
+      "step": 19602
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.077038892637583,
+      "learning_rate": 1.171368320022235e-05,
+      "loss": 0.8683,
+      "step": 19603
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.740706082652377,
+      "learning_rate": 1.1712931441814776e-05,
+      "loss": 0.7385,
+      "step": 19604
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.087981052557559,
+      "learning_rate": 1.1712179673433974e-05,
+      "loss": 0.7598,
+      "step": 19605
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6096263657802186,
+      "learning_rate": 1.171142789508432e-05,
+      "loss": 0.6725,
+      "step": 19606
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6019611511057603,
+      "learning_rate": 1.1710676106770191e-05,
+      "loss": 0.7434,
+      "step": 19607
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4894727423030634,
+      "learning_rate": 1.1709924308495964e-05,
+      "loss": 0.8059,
+      "step": 19608
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7565752596272228,
+      "learning_rate": 1.1709172500266015e-05,
+      "loss": 0.7616,
+      "step": 19609
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4705431933515227,
+      "learning_rate": 1.1708420682084722e-05,
+      "loss": 0.7061,
+      "step": 19610
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.580845603443694,
+      "learning_rate": 1.1707668853956465e-05,
+      "loss": 0.6341,
+      "step": 19611
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4842882303256968,
+      "learning_rate": 1.1706917015885621e-05,
+      "loss": 0.6537,
+      "step": 19612
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4695974303742156,
+      "learning_rate": 1.1706165167876562e-05,
+      "loss": 0.9142,
+      "step": 19613
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.944514912876414,
+      "learning_rate": 1.170541330993367e-05,
+      "loss": 0.8579,
+      "step": 19614
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6498064344076975,
+      "learning_rate": 1.1704661442061323e-05,
+      "loss": 0.8139,
+      "step": 19615
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8514972588397194,
+      "learning_rate": 1.1703909564263898e-05,
+      "loss": 0.5532,
+      "step": 19616
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6192759293767782,
+      "learning_rate": 1.1703157676545768e-05,
+      "loss": 0.6606,
+      "step": 19617
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3230449761485499,
+      "learning_rate": 1.170240577891132e-05,
+      "loss": 0.5269,
+      "step": 19618
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7016309389054578,
+      "learning_rate": 1.1701653871364924e-05,
+      "loss": 0.7397,
+      "step": 19619
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7299218404227108,
+      "learning_rate": 1.1700901953910961e-05,
+      "loss": 0.6507,
+      "step": 19620
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3691831350997008,
+      "learning_rate": 1.1700150026553807e-05,
+      "loss": 0.7872,
+      "step": 19621
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.2247903578484047,
+      "learning_rate": 1.1699398089297842e-05,
+      "loss": 0.7958,
+      "step": 19622
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4155425625289244,
+      "learning_rate": 1.1698646142147446e-05,
+      "loss": 0.6751,
+      "step": 19623
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8646870967901548,
+      "learning_rate": 1.1697894185106991e-05,
+      "loss": 0.8401,
+      "step": 19624
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6279210339675939,
+      "learning_rate": 1.1697142218180859e-05,
+      "loss": 0.5772,
+      "step": 19625
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.5630628237571527,
+      "learning_rate": 1.1696390241373427e-05,
+      "loss": 0.6957,
+      "step": 19626
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6485815900852394,
+      "learning_rate": 1.1695638254689074e-05,
+      "loss": 0.7913,
+      "step": 19627
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3616043055714568,
+      "learning_rate": 1.1694886258132182e-05,
+      "loss": 0.5949,
+      "step": 19628
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7783806103562354,
+      "learning_rate": 1.1694134251707121e-05,
+      "loss": 0.7177,
+      "step": 19629
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.724576262783148,
+      "learning_rate": 1.1693382235418276e-05,
+      "loss": 0.754,
+      "step": 19630
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5688153545014578,
+      "learning_rate": 1.1692630209270024e-05,
+      "loss": 0.5788,
+      "step": 19631
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1097375883804546,
+      "learning_rate": 1.1691878173266743e-05,
+      "loss": 0.7408,
+      "step": 19632
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.1289080870559114,
+      "learning_rate": 1.1691126127412814e-05,
+      "loss": 0.7034,
+      "step": 19633
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0141025553905547,
+      "learning_rate": 1.1690374071712608e-05,
+      "loss": 0.8104,
+      "step": 19634
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4565845808995606,
+      "learning_rate": 1.168962200617051e-05,
+      "loss": 0.791,
+      "step": 19635
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7104429232934513,
+      "learning_rate": 1.1688869930790898e-05,
+      "loss": 0.8556,
+      "step": 19636
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8579031108266455,
+      "learning_rate": 1.1688117845578155e-05,
+      "loss": 0.6709,
+      "step": 19637
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3656401436891032,
+      "learning_rate": 1.168736575053665e-05,
+      "loss": 0.7547,
+      "step": 19638
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6069536596343335,
+      "learning_rate": 1.168661364567077e-05,
+      "loss": 0.7262,
+      "step": 19639
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.561954258834213,
+      "learning_rate": 1.1685861530984893e-05,
+      "loss": 0.6778,
+      "step": 19640
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2899272179507362,
+      "learning_rate": 1.1685109406483392e-05,
+      "loss": 0.7036,
+      "step": 19641
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5640628483938734,
+      "learning_rate": 1.1684357272170655e-05,
+      "loss": 0.7117,
+      "step": 19642
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0355999021297086,
+      "learning_rate": 1.1683605128051053e-05,
+      "loss": 0.6784,
+      "step": 19643
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4409764107875491,
+      "learning_rate": 1.1682852974128974e-05,
+      "loss": 0.6367,
+      "step": 19644
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6296658185062243,
+      "learning_rate": 1.1682100810408785e-05,
+      "loss": 0.7705,
+      "step": 19645
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4082281151932876,
+      "learning_rate": 1.1681348636894879e-05,
+      "loss": 0.7459,
+      "step": 19646
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2728262506918824,
+      "learning_rate": 1.1680596453591626e-05,
+      "loss": 0.7336,
+      "step": 19647
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4046531541534886,
+      "learning_rate": 1.1679844260503407e-05,
+      "loss": 0.7616,
+      "step": 19648
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4196001552264963,
+      "learning_rate": 1.1679092057634605e-05,
+      "loss": 0.7397,
+      "step": 19649
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4538946057196018,
+      "learning_rate": 1.1678339844989597e-05,
+      "loss": 0.8084,
+      "step": 19650
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.82381303909425,
+      "learning_rate": 1.1677587622572763e-05,
+      "loss": 0.6481,
+      "step": 19651
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5725951377859129,
+      "learning_rate": 1.1676835390388484e-05,
+      "loss": 0.7311,
+      "step": 19652
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6359456675387591,
+      "learning_rate": 1.1676083148441135e-05,
+      "loss": 0.6769,
+      "step": 19653
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2757937736176317,
+      "learning_rate": 1.1675330896735103e-05,
+      "loss": 0.5916,
+      "step": 19654
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5045117267893342,
+      "learning_rate": 1.1674578635274763e-05,
+      "loss": 0.8184,
+      "step": 19655
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5228357299150082,
+      "learning_rate": 1.1673826364064496e-05,
+      "loss": 0.7291,
+      "step": 19656
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.670815780064274,
+      "learning_rate": 1.167307408310868e-05,
+      "loss": 0.7936,
+      "step": 19657
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4318952706180086,
+      "learning_rate": 1.1672321792411698e-05,
+      "loss": 0.6755,
+      "step": 19658
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5870543465641457,
+      "learning_rate": 1.167156949197793e-05,
+      "loss": 0.5133,
+      "step": 19659
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5903431411628336,
+      "learning_rate": 1.1670817181811756e-05,
+      "loss": 0.4889,
+      "step": 19660
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.27308919494,
+      "learning_rate": 1.1670064861917554e-05,
+      "loss": 0.6472,
+      "step": 19661
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6015005128621658,
+      "learning_rate": 1.1669312532299705e-05,
+      "loss": 0.5471,
+      "step": 19662
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8563001305238949,
+      "learning_rate": 1.1668560192962595e-05,
+      "loss": 0.7402,
+      "step": 19663
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.561089500123265,
+      "learning_rate": 1.1667807843910592e-05,
+      "loss": 0.6838,
+      "step": 19664
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5744342299111007,
+      "learning_rate": 1.166705548514809e-05,
+      "loss": 0.698,
+      "step": 19665
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.57739812902974,
+      "learning_rate": 1.1666303116679456e-05,
+      "loss": 0.5399,
+      "step": 19666
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4287540597402624,
+      "learning_rate": 1.1665550738509084e-05,
+      "loss": 0.7979,
+      "step": 19667
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6850887432575024,
+      "learning_rate": 1.1664798350641344e-05,
+      "loss": 0.6243,
+      "step": 19668
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8439926778839806,
+      "learning_rate": 1.1664045953080625e-05,
+      "loss": 0.7432,
+      "step": 19669
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5765006439272393,
+      "learning_rate": 1.1663293545831302e-05,
+      "loss": 0.7307,
+      "step": 19670
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5086415831196842,
+      "learning_rate": 1.1662541128897758e-05,
+      "loss": 0.64,
+      "step": 19671
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.464607117861851,
+      "learning_rate": 1.166178870228437e-05,
+      "loss": 0.8272,
+      "step": 19672
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.114395846638678,
+      "learning_rate": 1.1661036265995528e-05,
+      "loss": 0.5849,
+      "step": 19673
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.420199703141536,
+      "learning_rate": 1.1660283820035601e-05,
+      "loss": 0.7953,
+      "step": 19674
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5716283648233602,
+      "learning_rate": 1.1659531364408979e-05,
+      "loss": 0.8088,
+      "step": 19675
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4445993990473995,
+      "learning_rate": 1.1658778899120038e-05,
+      "loss": 0.6305,
+      "step": 19676
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5790933318585652,
+      "learning_rate": 1.1658026424173163e-05,
+      "loss": 0.689,
+      "step": 19677
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4652694984365935,
+      "learning_rate": 1.1657273939572734e-05,
+      "loss": 0.775,
+      "step": 19678
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3337749395598233,
+      "learning_rate": 1.1656521445323129e-05,
+      "loss": 0.7138,
+      "step": 19679
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2441223103906858,
+      "learning_rate": 1.1655768941428734e-05,
+      "loss": 0.6408,
+      "step": 19680
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7458256605712466,
+      "learning_rate": 1.1655016427893926e-05,
+      "loss": 0.7839,
+      "step": 19681
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5324293431820821,
+      "learning_rate": 1.1654263904723089e-05,
+      "loss": 0.6561,
+      "step": 19682
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.472418169566697,
+      "learning_rate": 1.1653511371920603e-05,
+      "loss": 0.6879,
+      "step": 19683
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.441640268267255,
+      "learning_rate": 1.1652758829490854e-05,
+      "loss": 0.6858,
+      "step": 19684
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7548039780390368,
+      "learning_rate": 1.1652006277438214e-05,
+      "loss": 0.7707,
+      "step": 19685
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8763380733701882,
+      "learning_rate": 1.1651253715767077e-05,
+      "loss": 0.5734,
+      "step": 19686
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4119503818506594,
+      "learning_rate": 1.1650501144481812e-05,
+      "loss": 0.7528,
+      "step": 19687
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7708185964956455,
+      "learning_rate": 1.164974856358681e-05,
+      "loss": 0.6858,
+      "step": 19688
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5982520320689003,
+      "learning_rate": 1.1648995973086448e-05,
+      "loss": 0.5253,
+      "step": 19689
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4778272043852747,
+      "learning_rate": 1.1648243372985111e-05,
+      "loss": 0.6382,
+      "step": 19690
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6428068580753865,
+      "learning_rate": 1.1647490763287178e-05,
+      "loss": 0.7001,
+      "step": 19691
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6420569368975628,
+      "learning_rate": 1.1646738143997031e-05,
+      "loss": 0.665,
+      "step": 19692
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5857507722320678,
+      "learning_rate": 1.1645985515119059e-05,
+      "loss": 0.519,
+      "step": 19693
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.449799856549946,
+      "learning_rate": 1.164523287665763e-05,
+      "loss": 0.7866,
+      "step": 19694
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4307177117278007,
+      "learning_rate": 1.164448022861714e-05,
+      "loss": 0.7617,
+      "step": 19695
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5223519859550896,
+      "learning_rate": 1.1643727571001964e-05,
+      "loss": 0.5868,
+      "step": 19696
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.8381354285208635,
+      "learning_rate": 1.1642974903816488e-05,
+      "loss": 0.7827,
+      "step": 19697
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5645378342715869,
+      "learning_rate": 1.164222222706509e-05,
+      "loss": 0.5139,
+      "step": 19698
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6237047846724189,
+      "learning_rate": 1.1641469540752153e-05,
+      "loss": 0.7707,
+      "step": 19699
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.44557969801122,
+      "learning_rate": 1.164071684488206e-05,
+      "loss": 0.8154,
+      "step": 19700
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.754197043310849,
+      "learning_rate": 1.1639964139459196e-05,
+      "loss": 0.6908,
+      "step": 19701
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5834830423066596,
+      "learning_rate": 1.1639211424487941e-05,
+      "loss": 0.5506,
+      "step": 19702
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8171634040065825,
+      "learning_rate": 1.1638458699972678e-05,
+      "loss": 0.7581,
+      "step": 19703
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4112021946432298,
+      "learning_rate": 1.1637705965917791e-05,
+      "loss": 0.7761,
+      "step": 19704
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5859384332589823,
+      "learning_rate": 1.163695322232766e-05,
+      "loss": 0.5001,
+      "step": 19705
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.547267912045534,
+      "learning_rate": 1.1636200469206671e-05,
+      "loss": 0.7245,
+      "step": 19706
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3713000305733065,
+      "learning_rate": 1.1635447706559205e-05,
+      "loss": 0.6399,
+      "step": 19707
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3723623371087936,
+      "learning_rate": 1.1634694934389641e-05,
+      "loss": 0.7104,
+      "step": 19708
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4566552734383313,
+      "learning_rate": 1.1633942152702367e-05,
+      "loss": 0.7582,
+      "step": 19709
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.623523769813778,
+      "learning_rate": 1.1633189361501767e-05,
+      "loss": 0.6298,
+      "step": 19710
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2170339947036932,
+      "learning_rate": 1.1632436560792217e-05,
+      "loss": 0.7546,
+      "step": 19711
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.414959964246927,
+      "learning_rate": 1.1631683750578112e-05,
+      "loss": 0.6927,
+      "step": 19712
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1388864453512988,
+      "learning_rate": 1.163093093086382e-05,
+      "loss": 0.6369,
+      "step": 19713
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6079944643097271,
+      "learning_rate": 1.1630178101653736e-05,
+      "loss": 0.5529,
+      "step": 19714
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5257545414216167,
+      "learning_rate": 1.1629425262952235e-05,
+      "loss": 0.6892,
+      "step": 19715
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6203620758919111,
+      "learning_rate": 1.1628672414763709e-05,
+      "loss": 0.5432,
+      "step": 19716
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6085466236626094,
+      "learning_rate": 1.1627919557092532e-05,
+      "loss": 0.7626,
+      "step": 19717
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5703371646197566,
+      "learning_rate": 1.1627166689943098e-05,
+      "loss": 0.6424,
+      "step": 19718
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4919866129150507,
+      "learning_rate": 1.1626413813319779e-05,
+      "loss": 0.7763,
+      "step": 19719
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3871884538621075,
+      "learning_rate": 1.1625660927226966e-05,
+      "loss": 0.7863,
+      "step": 19720
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4654564049661751,
+      "learning_rate": 1.162490803166904e-05,
+      "loss": 0.7187,
+      "step": 19721
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.518365120133171,
+      "learning_rate": 1.1624155126650385e-05,
+      "loss": 0.6317,
+      "step": 19722
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.385210995882271,
+      "learning_rate": 1.1623402212175384e-05,
+      "loss": 0.6166,
+      "step": 19723
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5557120310771821,
+      "learning_rate": 1.1622649288248422e-05,
+      "loss": 0.679,
+      "step": 19724
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.115247474759605,
+      "learning_rate": 1.1621896354873883e-05,
+      "loss": 0.689,
+      "step": 19725
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7151773362660672,
+      "learning_rate": 1.1621143412056146e-05,
+      "loss": 0.6678,
+      "step": 19726
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5207435725300087,
+      "learning_rate": 1.1620390459799605e-05,
+      "loss": 0.6341,
+      "step": 19727
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8291985509901105,
+      "learning_rate": 1.1619637498108634e-05,
+      "loss": 0.6835,
+      "step": 19728
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.0817098243520524,
+      "learning_rate": 1.1618884526987619e-05,
+      "loss": 0.723,
+      "step": 19729
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3662586080925518,
+      "learning_rate": 1.1618131546440949e-05,
+      "loss": 0.7298,
+      "step": 19730
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5446682136653919,
+      "learning_rate": 1.1617378556473004e-05,
+      "loss": 0.5078,
+      "step": 19731
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3780909010296634,
+      "learning_rate": 1.1616625557088169e-05,
+      "loss": 0.7336,
+      "step": 19732
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4709678882027772,
+      "learning_rate": 1.1615872548290827e-05,
+      "loss": 0.6718,
+      "step": 19733
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.859698586772775,
+      "learning_rate": 1.1615119530085365e-05,
+      "loss": 0.7092,
+      "step": 19734
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6809009562234345,
+      "learning_rate": 1.1614366502476169e-05,
+      "loss": 0.7382,
+      "step": 19735
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3210895948835346,
+      "learning_rate": 1.1613613465467612e-05,
+      "loss": 0.8012,
+      "step": 19736
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4164967225263099,
+      "learning_rate": 1.1612860419064092e-05,
+      "loss": 0.7476,
+      "step": 19737
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5344570957862562,
+      "learning_rate": 1.1612107363269987e-05,
+      "loss": 0.6866,
+      "step": 19738
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4359816564265833,
+      "learning_rate": 1.1611354298089683e-05,
+      "loss": 0.7193,
+      "step": 19739
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3127931118055736,
+      "learning_rate": 1.1610601223527563e-05,
+      "loss": 0.7633,
+      "step": 19740
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5993984435970642,
+      "learning_rate": 1.1609848139588014e-05,
+      "loss": 0.5028,
+      "step": 19741
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.496626655658522,
+      "learning_rate": 1.160909504627542e-05,
+      "loss": 0.8233,
+      "step": 19742
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5332511657239218,
+      "learning_rate": 1.1608341943594162e-05,
+      "loss": 0.7292,
+      "step": 19743
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3000641527750212,
+      "learning_rate": 1.160758883154863e-05,
+      "loss": 0.802,
+      "step": 19744
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4559047680439194,
+      "learning_rate": 1.1606835710143207e-05,
+      "loss": 0.6691,
+      "step": 19745
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.618604441677336,
+      "learning_rate": 1.160608257938228e-05,
+      "loss": 0.9069,
+      "step": 19746
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5482630808672018,
+      "learning_rate": 1.1605329439270227e-05,
+      "loss": 0.7108,
+      "step": 19747
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.690901241285501,
+      "learning_rate": 1.1604576289811442e-05,
+      "loss": 0.7636,
+      "step": 19748
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.040429867745419,
+      "learning_rate": 1.1603823131010303e-05,
+      "loss": 0.5632,
+      "step": 19749
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8397912695052132,
+      "learning_rate": 1.16030699628712e-05,
+      "loss": 0.7226,
+      "step": 19750
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5679397686211585,
+      "learning_rate": 1.1602316785398515e-05,
+      "loss": 0.677,
+      "step": 19751
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5845205625064576,
+      "learning_rate": 1.1601563598596635e-05,
+      "loss": 0.6825,
+      "step": 19752
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6556437424185326,
+      "learning_rate": 1.1600810402469944e-05,
+      "loss": 0.7136,
+      "step": 19753
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0085954284568803,
+      "learning_rate": 1.1600057197022826e-05,
+      "loss": 0.7469,
+      "step": 19754
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4685324717010533,
+      "learning_rate": 1.1599303982259673e-05,
+      "loss": 0.7118,
+      "step": 19755
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.532865821145512,
+      "learning_rate": 1.1598550758184864e-05,
+      "loss": 0.7961,
+      "step": 19756
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4673604355406278,
+      "learning_rate": 1.1597797524802786e-05,
+      "loss": 0.6948,
+      "step": 19757
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5297326064524184,
+      "learning_rate": 1.1597044282117824e-05,
+      "loss": 0.7716,
+      "step": 19758
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6621447062845092,
+      "learning_rate": 1.1596291030134366e-05,
+      "loss": 0.8257,
+      "step": 19759
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3519800042309498,
+      "learning_rate": 1.1595537768856794e-05,
+      "loss": 0.7886,
+      "step": 19760
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1995453672361454,
+      "learning_rate": 1.15947844982895e-05,
+      "loss": 0.7986,
+      "step": 19761
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2999796092048341,
+      "learning_rate": 1.1594031218436864e-05,
+      "loss": 0.8504,
+      "step": 19762
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5523383120839936,
+      "learning_rate": 1.1593277929303275e-05,
+      "loss": 0.4902,
+      "step": 19763
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7575182099658504,
+      "learning_rate": 1.1592524630893111e-05,
+      "loss": 0.734,
+      "step": 19764
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4664773334044512,
+      "learning_rate": 1.1591771323210771e-05,
+      "loss": 0.6602,
+      "step": 19765
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.414505486591085,
+      "learning_rate": 1.1591018006260631e-05,
+      "loss": 0.623,
+      "step": 19766
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.299084118829956,
+      "learning_rate": 1.1590264680047084e-05,
+      "loss": 0.8213,
+      "step": 19767
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5314542797447388,
+      "learning_rate": 1.1589511344574509e-05,
+      "loss": 0.679,
+      "step": 19768
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4428806057828651,
+      "learning_rate": 1.1588757999847297e-05,
+      "loss": 0.7763,
+      "step": 19769
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3010854104355685,
+      "learning_rate": 1.1588004645869835e-05,
+      "loss": 0.6522,
+      "step": 19770
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6105230479139013,
+      "learning_rate": 1.1587251282646505e-05,
+      "loss": 0.5257,
+      "step": 19771
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.447280342285266,
+      "learning_rate": 1.1586497910181695e-05,
+      "loss": 0.7872,
+      "step": 19772
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5346111794261172,
+      "learning_rate": 1.1585744528479794e-05,
+      "loss": 0.7961,
+      "step": 19773
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4962740790932478,
+      "learning_rate": 1.1584991137545185e-05,
+      "loss": 0.762,
+      "step": 19774
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6660275830603317,
+      "learning_rate": 1.1584237737382253e-05,
+      "loss": 0.7106,
+      "step": 19775
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3381260763031322,
+      "learning_rate": 1.158348432799539e-05,
+      "loss": 0.6626,
+      "step": 19776
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4035252580397053,
+      "learning_rate": 1.1582730909388982e-05,
+      "loss": 0.6782,
+      "step": 19777
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.874216674553247,
+      "learning_rate": 1.1581977481567413e-05,
+      "loss": 0.684,
+      "step": 19778
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5499467329808274,
+      "learning_rate": 1.1581224044535067e-05,
+      "loss": 0.7849,
+      "step": 19779
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4751961243745404,
+      "learning_rate": 1.1580470598296337e-05,
+      "loss": 0.6976,
+      "step": 19780
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.619763337392961,
+      "learning_rate": 1.1579717142855605e-05,
+      "loss": 0.7316,
+      "step": 19781
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8424656328967353,
+      "learning_rate": 1.157896367821726e-05,
+      "loss": 0.7732,
+      "step": 19782
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4739629875688671,
+      "learning_rate": 1.1578210204385689e-05,
+      "loss": 0.6868,
+      "step": 19783
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6213922785251781,
+      "learning_rate": 1.157745672136528e-05,
+      "loss": 0.5303,
+      "step": 19784
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5999515226414648,
+      "learning_rate": 1.1576703229160416e-05,
+      "loss": 0.5055,
+      "step": 19785
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3353056694948027,
+      "learning_rate": 1.1575949727775487e-05,
+      "loss": 0.7077,
+      "step": 19786
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4467572387101575,
+      "learning_rate": 1.1575196217214879e-05,
+      "loss": 0.6345,
+      "step": 19787
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4597345834246713,
+      "learning_rate": 1.1574442697482982e-05,
+      "loss": 0.6558,
+      "step": 19788
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.9270337914074736,
+      "learning_rate": 1.1573689168584182e-05,
+      "loss": 0.7302,
+      "step": 19789
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.542385453746203,
+      "learning_rate": 1.1572935630522864e-05,
+      "loss": 0.7316,
+      "step": 19790
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7100565123917078,
+      "learning_rate": 1.1572182083303417e-05,
+      "loss": 0.6903,
+      "step": 19791
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.795197049735433,
+      "learning_rate": 1.1571428526930227e-05,
+      "loss": 0.8582,
+      "step": 19792
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4616740320709243,
+      "learning_rate": 1.1570674961407684e-05,
+      "loss": 0.6945,
+      "step": 19793
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4603325882492844,
+      "learning_rate": 1.156992138674017e-05,
+      "loss": 0.7349,
+      "step": 19794
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.765128013820539,
+      "learning_rate": 1.1569167802932084e-05,
+      "loss": 0.6915,
+      "step": 19795
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5579838664013743,
+      "learning_rate": 1.15684142099878e-05,
+      "loss": 0.7882,
+      "step": 19796
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6588283432531883,
+      "learning_rate": 1.1567660607911718e-05,
+      "loss": 0.5527,
+      "step": 19797
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5788441049688273,
+      "learning_rate": 1.1566906996708212e-05,
+      "loss": 0.6891,
+      "step": 19798
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2797303965005271,
+      "learning_rate": 1.1566153376381683e-05,
+      "loss": 0.7106,
+      "step": 19799
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5933621199567625,
+      "learning_rate": 1.1565399746936513e-05,
+      "loss": 0.5032,
+      "step": 19800
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5737553660781128,
+      "learning_rate": 1.1564646108377088e-05,
+      "loss": 0.5009,
+      "step": 19801
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4892241097568648,
+      "learning_rate": 1.1563892460707799e-05,
+      "loss": 0.6105,
+      "step": 19802
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3180178103995968,
+      "learning_rate": 1.1563138803933031e-05,
+      "loss": 0.8133,
+      "step": 19803
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4074073466245385,
+      "learning_rate": 1.1562385138057177e-05,
+      "loss": 0.663,
+      "step": 19804
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.337563658725447,
+      "learning_rate": 1.1561631463084621e-05,
+      "loss": 0.7733,
+      "step": 19805
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6473331906071396,
+      "learning_rate": 1.1560877779019752e-05,
+      "loss": 0.777,
+      "step": 19806
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5624518903327539,
+      "learning_rate": 1.1560124085866957e-05,
+      "loss": 0.5137,
+      "step": 19807
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8184756674578404,
+      "learning_rate": 1.1559370383630628e-05,
+      "loss": 0.7769,
+      "step": 19808
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4465298320332611,
+      "learning_rate": 1.155861667231515e-05,
+      "loss": 0.8323,
+      "step": 19809
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.345564174088835,
+      "learning_rate": 1.1557862951924912e-05,
+      "loss": 0.7633,
+      "step": 19810
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5405940624693824,
+      "learning_rate": 1.1557109222464302e-05,
+      "loss": 0.6872,
+      "step": 19811
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3605512832389268,
+      "learning_rate": 1.155635548393771e-05,
+      "loss": 0.6719,
+      "step": 19812
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3479257864555632,
+      "learning_rate": 1.1555601736349525e-05,
+      "loss": 0.8125,
+      "step": 19813
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8305455332966651,
+      "learning_rate": 1.1554847979704133e-05,
+      "loss": 0.9053,
+      "step": 19814
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4301093024350124,
+      "learning_rate": 1.1554094214005922e-05,
+      "loss": 0.6766,
+      "step": 19815
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4830034508740537,
+      "learning_rate": 1.1553340439259287e-05,
+      "loss": 0.6785,
+      "step": 19816
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.717960526782985,
+      "learning_rate": 1.1552586655468608e-05,
+      "loss": 0.6816,
+      "step": 19817
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4790868626775917,
+      "learning_rate": 1.1551832862638282e-05,
+      "loss": 0.8384,
+      "step": 19818
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4593978400140315,
+      "learning_rate": 1.1551079060772687e-05,
+      "loss": 0.6371,
+      "step": 19819
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4849878631085294,
+      "learning_rate": 1.1550325249876226e-05,
+      "loss": 0.7311,
+      "step": 19820
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6352661829701525,
+      "learning_rate": 1.1549571429953276e-05,
+      "loss": 0.5408,
+      "step": 19821
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3879198804782078,
+      "learning_rate": 1.1548817601008233e-05,
+      "loss": 0.7781,
+      "step": 19822
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5475460200684508,
+      "learning_rate": 1.1548063763045483e-05,
+      "loss": 0.7252,
+      "step": 19823
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.559230220264272,
+      "learning_rate": 1.154730991606941e-05,
+      "loss": 0.7563,
+      "step": 19824
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6022134821935787,
+      "learning_rate": 1.1546556060084418e-05,
+      "loss": 0.5523,
+      "step": 19825
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4571296226072945,
+      "learning_rate": 1.1545802195094879e-05,
+      "loss": 0.6603,
+      "step": 19826
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5819239622018565,
+      "learning_rate": 1.1545048321105196e-05,
+      "loss": 0.5249,
+      "step": 19827
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2287408117465448,
+      "learning_rate": 1.1544294438119751e-05,
+      "loss": 0.7268,
+      "step": 19828
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.569555231942163,
+      "learning_rate": 1.1543540546142934e-05,
+      "loss": 0.61,
+      "step": 19829
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.329113126232524,
+      "learning_rate": 1.1542786645179139e-05,
+      "loss": 0.687,
+      "step": 19830
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.560759446959174,
+      "learning_rate": 1.1542032735232746e-05,
+      "loss": 0.6692,
+      "step": 19831
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5901135158982097,
+      "learning_rate": 1.1541278816308152e-05,
+      "loss": 0.5594,
+      "step": 19832
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4702330697119868,
+      "learning_rate": 1.1540524888409746e-05,
+      "loss": 0.7606,
+      "step": 19833
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2728888641811518,
+      "learning_rate": 1.1539770951541916e-05,
+      "loss": 0.7315,
+      "step": 19834
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6508442614579841,
+      "learning_rate": 1.1539017005709052e-05,
+      "loss": 0.6646,
+      "step": 19835
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.037982795296407,
+      "learning_rate": 1.1538263050915543e-05,
+      "loss": 0.6436,
+      "step": 19836
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.022989990994929,
+      "learning_rate": 1.1537509087165781e-05,
+      "loss": 0.6592,
+      "step": 19837
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4621489753334822,
+      "learning_rate": 1.1536755114464153e-05,
+      "loss": 0.7227,
+      "step": 19838
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9603387221414506,
+      "learning_rate": 1.1536001132815051e-05,
+      "loss": 0.5921,
+      "step": 19839
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3120398030844,
+      "learning_rate": 1.1535247142222862e-05,
+      "loss": 0.6292,
+      "step": 19840
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9579507630972168,
+      "learning_rate": 1.153449314269198e-05,
+      "loss": 0.8662,
+      "step": 19841
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5382305128074079,
+      "learning_rate": 1.1533739134226797e-05,
+      "loss": 0.9052,
+      "step": 19842
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3940172263309831,
+      "learning_rate": 1.1532985116831692e-05,
+      "loss": 0.6964,
+      "step": 19843
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5055091288110691,
+      "learning_rate": 1.1532231090511069e-05,
+      "loss": 0.7942,
+      "step": 19844
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3996207456635237,
+      "learning_rate": 1.1531477055269305e-05,
+      "loss": 0.6641,
+      "step": 19845
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5662968390757173,
+      "learning_rate": 1.15307230111108e-05,
+      "loss": 0.7173,
+      "step": 19846
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5617302283010412,
+      "learning_rate": 1.152996895803994e-05,
+      "loss": 0.7119,
+      "step": 19847
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2959678588479613,
+      "learning_rate": 1.1529214896061118e-05,
+      "loss": 0.6523,
+      "step": 19848
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3322288332864778,
+      "learning_rate": 1.152846082517872e-05,
+      "loss": 0.731,
+      "step": 19849
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.321442361323902,
+      "learning_rate": 1.1527706745397143e-05,
+      "loss": 0.7298,
+      "step": 19850
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5280315742366612,
+      "learning_rate": 1.1526952656720772e-05,
+      "loss": 0.7908,
+      "step": 19851
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5620159259863498,
+      "learning_rate": 1.1526198559153998e-05,
+      "loss": 0.698,
+      "step": 19852
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.663770491967302,
+      "learning_rate": 1.1525444452701213e-05,
+      "loss": 0.8116,
+      "step": 19853
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2397501231254267,
+      "learning_rate": 1.1524690337366808e-05,
+      "loss": 0.6913,
+      "step": 19854
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0365085034300403,
+      "learning_rate": 1.1523936213155174e-05,
+      "loss": 0.7137,
+      "step": 19855
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.627088605260384,
+      "learning_rate": 1.1523182080070701e-05,
+      "loss": 0.7118,
+      "step": 19856
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3812148137058124,
+      "learning_rate": 1.1522427938117777e-05,
+      "loss": 0.7314,
+      "step": 19857
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8131298316827773,
+      "learning_rate": 1.1521673787300798e-05,
+      "loss": 0.8123,
+      "step": 19858
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3374817231713803,
+      "learning_rate": 1.1520919627624152e-05,
+      "loss": 0.693,
+      "step": 19859
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5602164338263642,
+      "learning_rate": 1.152016545909223e-05,
+      "loss": 0.7615,
+      "step": 19860
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4035361708145306,
+      "learning_rate": 1.1519411281709423e-05,
+      "loss": 0.6899,
+      "step": 19861
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.387989634380038,
+      "learning_rate": 1.1518657095480123e-05,
+      "loss": 0.6723,
+      "step": 19862
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3428159723480821,
+      "learning_rate": 1.1517902900408722e-05,
+      "loss": 0.6861,
+      "step": 19863
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4431002966133988,
+      "learning_rate": 1.1517148696499607e-05,
+      "loss": 0.7596,
+      "step": 19864
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9622631193941842,
+      "learning_rate": 1.1516394483757176e-05,
+      "loss": 0.8144,
+      "step": 19865
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6213046428032648,
+      "learning_rate": 1.1515640262185813e-05,
+      "loss": 0.556,
+      "step": 19866
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.280428476388324,
+      "learning_rate": 1.1514886031789913e-05,
+      "loss": 0.6998,
+      "step": 19867
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7481616556671422,
+      "learning_rate": 1.1514131792573865e-05,
+      "loss": 0.7028,
+      "step": 19868
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4994934123361003,
+      "learning_rate": 1.1513377544542069e-05,
+      "loss": 0.6119,
+      "step": 19869
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5139177602205158,
+      "learning_rate": 1.1512623287698901e-05,
+      "loss": 0.8522,
+      "step": 19870
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6014596113467089,
+      "learning_rate": 1.1511869022048767e-05,
+      "loss": 0.5612,
+      "step": 19871
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5008752279160062,
+      "learning_rate": 1.1511114747596052e-05,
+      "loss": 0.6011,
+      "step": 19872
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3357360046790674,
+      "learning_rate": 1.1510360464345147e-05,
+      "loss": 0.7384,
+      "step": 19873
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3303983047583636,
+      "learning_rate": 1.1509606172300447e-05,
+      "loss": 0.8598,
+      "step": 19874
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6371249572538317,
+      "learning_rate": 1.150885187146634e-05,
+      "loss": 0.5828,
+      "step": 19875
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6754330424607842,
+      "learning_rate": 1.1508097561847223e-05,
+      "loss": 0.7616,
+      "step": 19876
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.577220559713216,
+      "learning_rate": 1.150734324344748e-05,
+      "loss": 0.7757,
+      "step": 19877
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5702399318922622,
+      "learning_rate": 1.1506588916271512e-05,
+      "loss": 0.6773,
+      "step": 19878
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.984219453551847,
+      "learning_rate": 1.1505834580323704e-05,
+      "loss": 0.5665,
+      "step": 19879
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3501345470007557,
+      "learning_rate": 1.150508023560845e-05,
+      "loss": 0.8867,
+      "step": 19880
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3115491232776462,
+      "learning_rate": 1.1504325882130143e-05,
+      "loss": 0.648,
+      "step": 19881
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5075439622484084,
+      "learning_rate": 1.1503571519893175e-05,
+      "loss": 0.685,
+      "step": 19882
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6203245360611258,
+      "learning_rate": 1.1502817148901937e-05,
+      "loss": 0.5126,
+      "step": 19883
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.299675316442844,
+      "learning_rate": 1.1502062769160821e-05,
+      "loss": 0.7733,
+      "step": 19884
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2715016489450492,
+      "learning_rate": 1.1501308380674221e-05,
+      "loss": 0.7946,
+      "step": 19885
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.366773529102826,
+      "learning_rate": 1.1500553983446527e-05,
+      "loss": 0.6005,
+      "step": 19886
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3751814680503276,
+      "learning_rate": 1.1499799577482134e-05,
+      "loss": 0.5775,
+      "step": 19887
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4603214836883653,
+      "learning_rate": 1.1499045162785434e-05,
+      "loss": 0.802,
+      "step": 19888
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3701115565063622,
+      "learning_rate": 1.1498290739360815e-05,
+      "loss": 0.7094,
+      "step": 19889
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3723350858583665,
+      "learning_rate": 1.1497536307212675e-05,
+      "loss": 0.7655,
+      "step": 19890
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5821616652350825,
+      "learning_rate": 1.1496781866345406e-05,
+      "loss": 0.822,
+      "step": 19891
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6571121371516149,
+      "learning_rate": 1.1496027416763399e-05,
+      "loss": 0.6955,
+      "step": 19892
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.351527938349625,
+      "learning_rate": 1.1495272958471046e-05,
+      "loss": 0.679,
+      "step": 19893
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6516455628473027,
+      "learning_rate": 1.1494518491472738e-05,
+      "loss": 0.544,
+      "step": 19894
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.418790794900156,
+      "learning_rate": 1.1493764015772875e-05,
+      "loss": 0.7881,
+      "step": 19895
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8076933489666152,
+      "learning_rate": 1.1493009531375838e-05,
+      "loss": 0.771,
+      "step": 19896
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.38619476491903,
+      "learning_rate": 1.1492255038286034e-05,
+      "loss": 0.6858,
+      "step": 19897
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4356648734595854,
+      "learning_rate": 1.1491500536507845e-05,
+      "loss": 0.7106,
+      "step": 19898
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.778105567148408,
+      "learning_rate": 1.149074602604567e-05,
+      "loss": 0.7098,
+      "step": 19899
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8772327361698034,
+      "learning_rate": 1.1489991506903896e-05,
+      "loss": 0.707,
+      "step": 19900
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4246501920960353,
+      "learning_rate": 1.1489236979086922e-05,
+      "loss": 0.7644,
+      "step": 19901
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4622658078525619,
+      "learning_rate": 1.1488482442599138e-05,
+      "loss": 0.7868,
+      "step": 19902
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6068322316050119,
+      "learning_rate": 1.148772789744494e-05,
+      "loss": 0.6983,
+      "step": 19903
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9601590373650446,
+      "learning_rate": 1.1486973343628719e-05,
+      "loss": 0.7642,
+      "step": 19904
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5719391781696306,
+      "learning_rate": 1.1486218781154866e-05,
+      "loss": 0.7012,
+      "step": 19905
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.835390057372065,
+      "learning_rate": 1.1485464210027778e-05,
+      "loss": 0.6762,
+      "step": 19906
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.398125834622116,
+      "learning_rate": 1.1484709630251847e-05,
+      "loss": 0.8094,
+      "step": 19907
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5813485653436972,
+      "learning_rate": 1.1483955041831465e-05,
+      "loss": 0.6471,
+      "step": 19908
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5653928984525605,
+      "learning_rate": 1.148320044477103e-05,
+      "loss": 0.619,
+      "step": 19909
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4094760739884944,
+      "learning_rate": 1.1482445839074929e-05,
+      "loss": 0.8396,
+      "step": 19910
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5988698778442098,
+      "learning_rate": 1.1481691224747563e-05,
+      "loss": 0.549,
+      "step": 19911
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4131634792234928,
+      "learning_rate": 1.1480936601793318e-05,
+      "loss": 0.9557,
+      "step": 19912
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5993500811167568,
+      "learning_rate": 1.1480181970216593e-05,
+      "loss": 0.5229,
+      "step": 19913
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4713074063215665,
+      "learning_rate": 1.1479427330021778e-05,
+      "loss": 0.6532,
+      "step": 19914
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5553517993478327,
+      "learning_rate": 1.147867268121327e-05,
+      "loss": 0.5171,
+      "step": 19915
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3920520858593084,
+      "learning_rate": 1.1477918023795464e-05,
+      "loss": 0.7228,
+      "step": 19916
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4195000115360732,
+      "learning_rate": 1.1477163357772747e-05,
+      "loss": 0.7475,
+      "step": 19917
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.832137579058893,
+      "learning_rate": 1.1476408683149522e-05,
+      "loss": 0.6951,
+      "step": 19918
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3886906538361237,
+      "learning_rate": 1.1475653999930174e-05,
+      "loss": 0.7894,
+      "step": 19919
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.367119792917865,
+      "learning_rate": 1.1474899308119105e-05,
+      "loss": 0.8968,
+      "step": 19920
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5919458484603226,
+      "learning_rate": 1.1474144607720701e-05,
+      "loss": 0.5679,
+      "step": 19921
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.560724622460951,
+      "learning_rate": 1.1473389898739363e-05,
+      "loss": 0.7057,
+      "step": 19922
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2078661234439667,
+      "learning_rate": 1.1472635181179483e-05,
+      "loss": 0.7026,
+      "step": 19923
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3095373462494893,
+      "learning_rate": 1.1471880455045452e-05,
+      "loss": 0.7413,
+      "step": 19924
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3187245419757754,
+      "learning_rate": 1.1471125720341671e-05,
+      "loss": 0.6186,
+      "step": 19925
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4251198143145791,
+      "learning_rate": 1.1470370977072524e-05,
+      "loss": 0.6972,
+      "step": 19926
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3988685476576923,
+      "learning_rate": 1.1469616225242418e-05,
+      "loss": 0.8563,
+      "step": 19927
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4633634924216055,
+      "learning_rate": 1.1468861464855738e-05,
+      "loss": 0.7089,
+      "step": 19928
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.635230851147919,
+      "learning_rate": 1.1468106695916885e-05,
+      "loss": 0.7255,
+      "step": 19929
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4683397159666236,
+      "learning_rate": 1.1467351918430248e-05,
+      "loss": 0.6276,
+      "step": 19930
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2587875193098308,
+      "learning_rate": 1.1466597132400221e-05,
+      "loss": 0.6268,
+      "step": 19931
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.48166746547722,
+      "learning_rate": 1.1465842337831203e-05,
+      "loss": 0.7681,
+      "step": 19932
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6335914130590528,
+      "learning_rate": 1.1465087534727587e-05,
+      "loss": 0.7482,
+      "step": 19933
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3621135064472956,
+      "learning_rate": 1.1464332723093769e-05,
+      "loss": 0.6757,
+      "step": 19934
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4332676226941286,
+      "learning_rate": 1.146357790293414e-05,
+      "loss": 0.6102,
+      "step": 19935
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5503265996912121,
+      "learning_rate": 1.1462823074253098e-05,
+      "loss": 0.8708,
+      "step": 19936
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3787806672359273,
+      "learning_rate": 1.1462068237055037e-05,
+      "loss": 0.6803,
+      "step": 19937
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7398388685490564,
+      "learning_rate": 1.146131339134435e-05,
+      "loss": 0.7284,
+      "step": 19938
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3447821803073223,
+      "learning_rate": 1.1460558537125437e-05,
+      "loss": 0.6502,
+      "step": 19939
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6366076041936496,
+      "learning_rate": 1.1459803674402687e-05,
+      "loss": 0.5518,
+      "step": 19940
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4150529581296594,
+      "learning_rate": 1.14590488031805e-05,
+      "loss": 0.7717,
+      "step": 19941
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7553466953019787,
+      "learning_rate": 1.1458293923463269e-05,
+      "loss": 0.8042,
+      "step": 19942
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5430788713573778,
+      "learning_rate": 1.1457539035255387e-05,
+      "loss": 0.8561,
+      "step": 19943
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5692899116843878,
+      "learning_rate": 1.1456784138561254e-05,
+      "loss": 0.5129,
+      "step": 19944
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.302196751835498,
+      "learning_rate": 1.1456029233385259e-05,
+      "loss": 0.8012,
+      "step": 19945
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6042863141812231,
+      "learning_rate": 1.1455274319731806e-05,
+      "loss": 0.4878,
+      "step": 19946
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4396760688665995,
+      "learning_rate": 1.1454519397605279e-05,
+      "loss": 0.6752,
+      "step": 19947
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.631318905417039,
+      "learning_rate": 1.1453764467010084e-05,
+      "loss": 0.7839,
+      "step": 19948
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4464313323228803,
+      "learning_rate": 1.145300952795061e-05,
+      "loss": 0.6716,
+      "step": 19949
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3325212312262402,
+      "learning_rate": 1.1452254580431256e-05,
+      "loss": 0.7456,
+      "step": 19950
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.240767346508154,
+      "learning_rate": 1.1451499624456415e-05,
+      "loss": 0.6733,
+      "step": 19951
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.294160900796554,
+      "learning_rate": 1.1450744660030484e-05,
+      "loss": 0.6706,
+      "step": 19952
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3022635944679481,
+      "learning_rate": 1.1449989687157859e-05,
+      "loss": 0.7661,
+      "step": 19953
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3199565235476902,
+      "learning_rate": 1.1449234705842933e-05,
+      "loss": 0.7954,
+      "step": 19954
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3776053405387072,
+      "learning_rate": 1.1448479716090103e-05,
+      "loss": 0.7149,
+      "step": 19955
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6925241221883056,
+      "learning_rate": 1.144772471790377e-05,
+      "loss": 0.7213,
+      "step": 19956
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.837459222253198,
+      "learning_rate": 1.1446969711288322e-05,
+      "loss": 0.6858,
+      "step": 19957
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4353108802051004,
+      "learning_rate": 1.144621469624816e-05,
+      "loss": 0.6787,
+      "step": 19958
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6129537597042766,
+      "learning_rate": 1.1445459672787676e-05,
+      "loss": 0.5749,
+      "step": 19959
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5287984253370213,
+      "learning_rate": 1.144470464091127e-05,
+      "loss": 0.7235,
+      "step": 19960
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4213072001927547,
+      "learning_rate": 1.1443949600623334e-05,
+      "loss": 0.7066,
+      "step": 19961
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4132046141072554,
+      "learning_rate": 1.1443194551928267e-05,
+      "loss": 0.7871,
+      "step": 19962
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4834872461939315,
+      "learning_rate": 1.1442439494830467e-05,
+      "loss": 0.7223,
+      "step": 19963
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7354905666035778,
+      "learning_rate": 1.1441684429334323e-05,
+      "loss": 0.7874,
+      "step": 19964
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5599633733288103,
+      "learning_rate": 1.144092935544424e-05,
+      "loss": 0.8235,
+      "step": 19965
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5924265971135125,
+      "learning_rate": 1.1440174273164608e-05,
+      "loss": 0.8004,
+      "step": 19966
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5921277420438246,
+      "learning_rate": 1.143941918249983e-05,
+      "loss": 0.5054,
+      "step": 19967
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5061430734686139,
+      "learning_rate": 1.143866408345429e-05,
+      "loss": 0.6929,
+      "step": 19968
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.428058766341625,
+      "learning_rate": 1.1437908976032395e-05,
+      "loss": 0.858,
+      "step": 19969
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3390656550507014,
+      "learning_rate": 1.1437153860238541e-05,
+      "loss": 0.6772,
+      "step": 19970
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.375822378917952,
+      "learning_rate": 1.143639873607712e-05,
+      "loss": 0.6327,
+      "step": 19971
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8832968523772293,
+      "learning_rate": 1.1435643603552532e-05,
+      "loss": 0.7787,
+      "step": 19972
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7261902328810792,
+      "learning_rate": 1.143488846266917e-05,
+      "loss": 0.6797,
+      "step": 19973
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2905554883932793,
+      "learning_rate": 1.143413331343144e-05,
+      "loss": 0.6354,
+      "step": 19974
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.593793251375685,
+      "learning_rate": 1.1433378155843725e-05,
+      "loss": 0.6799,
+      "step": 19975
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5337805429468165,
+      "learning_rate": 1.1432622989910434e-05,
+      "loss": 0.7663,
+      "step": 19976
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5927060719050742,
+      "learning_rate": 1.1431867815635951e-05,
+      "loss": 0.5119,
+      "step": 19977
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.197539957550783,
+      "learning_rate": 1.1431112633024689e-05,
+      "loss": 0.6919,
+      "step": 19978
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6821914357496757,
+      "learning_rate": 1.1430357442081027e-05,
+      "loss": 0.8427,
+      "step": 19979
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4635256178369416,
+      "learning_rate": 1.1429602242809379e-05,
+      "loss": 0.7421,
+      "step": 19980
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4393206524438673,
+      "learning_rate": 1.142884703521413e-05,
+      "loss": 0.665,
+      "step": 19981
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3187153630088606,
+      "learning_rate": 1.1428091819299683e-05,
+      "loss": 0.6647,
+      "step": 19982
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5743727797333582,
+      "learning_rate": 1.142733659507043e-05,
+      "loss": 0.7929,
+      "step": 19983
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.3929424980295493,
+      "learning_rate": 1.1426581362530775e-05,
+      "loss": 0.7065,
+      "step": 19984
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4608662186162293,
+      "learning_rate": 1.142582612168511e-05,
+      "loss": 0.6641,
+      "step": 19985
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.137930168556639,
+      "learning_rate": 1.1425070872537836e-05,
+      "loss": 0.7765,
+      "step": 19986
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5597735683742007,
+      "learning_rate": 1.1424315615093343e-05,
+      "loss": 0.7885,
+      "step": 19987
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.315545619907982,
+      "learning_rate": 1.142356034935604e-05,
+      "loss": 0.8114,
+      "step": 19988
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6953355785313304,
+      "learning_rate": 1.1422805075330315e-05,
+      "loss": 0.7927,
+      "step": 19989
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4273972136424478,
+      "learning_rate": 1.1422049793020566e-05,
+      "loss": 0.8097,
+      "step": 19990
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7301553384003718,
+      "learning_rate": 1.1421294502431197e-05,
+      "loss": 0.7894,
+      "step": 19991
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.518509821424483,
+      "learning_rate": 1.1420539203566598e-05,
+      "loss": 0.8191,
+      "step": 19992
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7461644701289265,
+      "learning_rate": 1.1419783896431174e-05,
+      "loss": 0.7087,
+      "step": 19993
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.407221793272387,
+      "learning_rate": 1.1419028581029315e-05,
+      "loss": 0.7499,
+      "step": 19994
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5993640811271994,
+      "learning_rate": 1.1418273257365426e-05,
+      "loss": 0.516,
+      "step": 19995
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.584649139984163,
+      "learning_rate": 1.1417517925443897e-05,
+      "loss": 0.546,
+      "step": 19996
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4482848468464233,
+      "learning_rate": 1.1416762585269134e-05,
+      "loss": 0.7172,
+      "step": 19997
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6096582455703887,
+      "learning_rate": 1.1416007236845526e-05,
+      "loss": 0.6863,
+      "step": 19998
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6070057142045517,
+      "learning_rate": 1.141525188017748e-05,
+      "loss": 0.7628,
+      "step": 19999
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5294404997586677,
+      "learning_rate": 1.1414496515269384e-05,
+      "loss": 0.8566,
+      "step": 20000
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2649347210820419,
+      "learning_rate": 1.141374114212565e-05,
+      "loss": 0.6681,
+      "step": 20001
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7171656124573411,
+      "learning_rate": 1.141298576075066e-05,
+      "loss": 0.6618,
+      "step": 20002
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9294574219321412,
+      "learning_rate": 1.1412230371148822e-05,
+      "loss": 0.668,
+      "step": 20003
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.641102287376841,
+      "learning_rate": 1.1411474973324533e-05,
+      "loss": 0.6971,
+      "step": 20004
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8606386373232888,
+      "learning_rate": 1.1410719567282188e-05,
+      "loss": 0.6833,
+      "step": 20005
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5963276446927892,
+      "learning_rate": 1.1409964153026188e-05,
+      "loss": 0.8234,
+      "step": 20006
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8767392922740977,
+      "learning_rate": 1.1409208730560927e-05,
+      "loss": 0.6178,
+      "step": 20007
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4913888032441411,
+      "learning_rate": 1.1408453299890811e-05,
+      "loss": 0.8218,
+      "step": 20008
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.373837334599395,
+      "learning_rate": 1.1407697861020232e-05,
+      "loss": 0.6599,
+      "step": 20009
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3435732516157124,
+      "learning_rate": 1.1406942413953593e-05,
+      "loss": 0.7523,
+      "step": 20010
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4789966132032832,
+      "learning_rate": 1.1406186958695287e-05,
+      "loss": 0.7156,
+      "step": 20011
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.445203630402162,
+      "learning_rate": 1.1405431495249714e-05,
+      "loss": 0.7167,
+      "step": 20012
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.397391911950544,
+      "learning_rate": 1.1404676023621278e-05,
+      "loss": 0.8846,
+      "step": 20013
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4214273905778885,
+      "learning_rate": 1.140392054381437e-05,
+      "loss": 0.6982,
+      "step": 20014
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5443706187455517,
+      "learning_rate": 1.1403165055833394e-05,
+      "loss": 0.7343,
+      "step": 20015
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2623288014469394,
+      "learning_rate": 1.1402409559682748e-05,
+      "loss": 0.8644,
+      "step": 20016
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7241132362244607,
+      "learning_rate": 1.1401654055366826e-05,
+      "loss": 0.7032,
+      "step": 20017
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3873714971924391,
+      "learning_rate": 1.1400898542890034e-05,
+      "loss": 0.8086,
+      "step": 20018
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5782871461457124,
+      "learning_rate": 1.1400143022256764e-05,
+      "loss": 0.5317,
+      "step": 20019
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2717673344883904,
+      "learning_rate": 1.139938749347142e-05,
+      "loss": 0.7538,
+      "step": 20020
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.582836857199328,
+      "learning_rate": 1.1398631956538398e-05,
+      "loss": 0.6975,
+      "step": 20021
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4164309235469728,
+      "learning_rate": 1.1397876411462096e-05,
+      "loss": 0.8217,
+      "step": 20022
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7007653381011578,
+      "learning_rate": 1.1397120858246921e-05,
+      "loss": 0.7681,
+      "step": 20023
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4326462499851058,
+      "learning_rate": 1.139636529689726e-05,
+      "loss": 0.7733,
+      "step": 20024
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2843907990173773,
+      "learning_rate": 1.1395609727417522e-05,
+      "loss": 0.7873,
+      "step": 20025
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7525679630590247,
+      "learning_rate": 1.13948541498121e-05,
+      "loss": 0.7064,
+      "step": 20026
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3534951018384833,
+      "learning_rate": 1.1394098564085399e-05,
+      "loss": 0.7049,
+      "step": 20027
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.36671179768914,
+      "learning_rate": 1.1393342970241809e-05,
+      "loss": 0.7987,
+      "step": 20028
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.561412236378875,
+      "learning_rate": 1.139258736828574e-05,
+      "loss": 0.6388,
+      "step": 20029
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2454429747795168,
+      "learning_rate": 1.1391831758221585e-05,
+      "loss": 0.6535,
+      "step": 20030
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5715138706132874,
+      "learning_rate": 1.1391076140053745e-05,
+      "loss": 0.5081,
+      "step": 20031
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4890963993579092,
+      "learning_rate": 1.139032051378662e-05,
+      "loss": 0.7046,
+      "step": 20032
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5659765938371545,
+      "learning_rate": 1.1389564879424605e-05,
+      "loss": 0.5072,
+      "step": 20033
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.397376908793972,
+      "learning_rate": 1.1388809236972107e-05,
+      "loss": 0.7985,
+      "step": 20034
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.395312139840723,
+      "learning_rate": 1.138805358643352e-05,
+      "loss": 0.7192,
+      "step": 20035
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.489463207325628,
+      "learning_rate": 1.1387297927813248e-05,
+      "loss": 0.7435,
+      "step": 20036
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9714694042516765,
+      "learning_rate": 1.1386542261115685e-05,
+      "loss": 0.6859,
+      "step": 20037
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6223799664112062,
+      "learning_rate": 1.1385786586345238e-05,
+      "loss": 0.6799,
+      "step": 20038
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6630073304595006,
+      "learning_rate": 1.13850309035063e-05,
+      "loss": 0.8095,
+      "step": 20039
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.7587364454336565,
+      "learning_rate": 1.1384275212603273e-05,
+      "loss": 0.6438,
+      "step": 20040
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4416214826679008,
+      "learning_rate": 1.138351951364056e-05,
+      "loss": 0.6412,
+      "step": 20041
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4774948351620996,
+      "learning_rate": 1.1382763806622556e-05,
+      "loss": 0.8396,
+      "step": 20042
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4778724312628617,
+      "learning_rate": 1.1382008091553664e-05,
+      "loss": 0.7468,
+      "step": 20043
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.790083852797832,
+      "learning_rate": 1.1381252368438283e-05,
+      "loss": 0.6428,
+      "step": 20044
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6016499709892897,
+      "learning_rate": 1.1380496637280816e-05,
+      "loss": 0.628,
+      "step": 20045
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5457859740255842,
+      "learning_rate": 1.1379740898085657e-05,
+      "loss": 0.6475,
+      "step": 20046
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.41339343948476,
+      "learning_rate": 1.1378985150857211e-05,
+      "loss": 0.766,
+      "step": 20047
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.433130515438581,
+      "learning_rate": 1.137822939559988e-05,
+      "loss": 0.8402,
+      "step": 20048
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3896351299558487,
+      "learning_rate": 1.1377473632318057e-05,
+      "loss": 0.7419,
+      "step": 20049
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4119100455551377,
+      "learning_rate": 1.1376717861016149e-05,
+      "loss": 0.802,
+      "step": 20050
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4271122012245403,
+      "learning_rate": 1.137596208169855e-05,
+      "loss": 0.7818,
+      "step": 20051
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4598004214112485,
+      "learning_rate": 1.1375206294369669e-05,
+      "loss": 0.7369,
+      "step": 20052
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7070334199418586,
+      "learning_rate": 1.1374450499033898e-05,
+      "loss": 0.7748,
+      "step": 20053
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3728209643817333,
+      "learning_rate": 1.1373694695695644e-05,
+      "loss": 0.8976,
+      "step": 20054
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.292032326324751,
+      "learning_rate": 1.1372938884359302e-05,
+      "loss": 0.6503,
+      "step": 20055
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.115025206043014,
+      "learning_rate": 1.1372183065029273e-05,
+      "loss": 0.7562,
+      "step": 20056
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2672511463451273,
+      "learning_rate": 1.1371427237709964e-05,
+      "loss": 0.8024,
+      "step": 20057
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6343222714112011,
+      "learning_rate": 1.1370671402405768e-05,
+      "loss": 0.7498,
+      "step": 20058
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4737476228076856,
+      "learning_rate": 1.1369915559121094e-05,
+      "loss": 0.7345,
+      "step": 20059
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4335104853136467,
+      "learning_rate": 1.1369159707860336e-05,
+      "loss": 0.78,
+      "step": 20060
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5655157835398352,
+      "learning_rate": 1.1368403848627896e-05,
+      "loss": 0.7294,
+      "step": 20061
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5056788741570182,
+      "learning_rate": 1.1367647981428173e-05,
+      "loss": 0.7235,
+      "step": 20062
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9315860679897048,
+      "learning_rate": 1.1366892106265575e-05,
+      "loss": 0.8483,
+      "step": 20063
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7966720586957512,
+      "learning_rate": 1.1366136223144495e-05,
+      "loss": 0.6956,
+      "step": 20064
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6038748696504943,
+      "learning_rate": 1.136538033206934e-05,
+      "loss": 0.869,
+      "step": 20065
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2460706621418338,
+      "learning_rate": 1.1364624433044507e-05,
+      "loss": 0.7777,
+      "step": 20066
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6339647401502146,
+      "learning_rate": 1.1363868526074397e-05,
+      "loss": 0.6749,
+      "step": 20067
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7020068857234267,
+      "learning_rate": 1.1363112611163415e-05,
+      "loss": 0.694,
+      "step": 20068
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5935567434516706,
+      "learning_rate": 1.1362356688315959e-05,
+      "loss": 0.5413,
+      "step": 20069
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5500052301267608,
+      "learning_rate": 1.136160075753643e-05,
+      "loss": 0.7876,
+      "step": 20070
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5929065851653679,
+      "learning_rate": 1.1360844818829234e-05,
+      "loss": 0.5236,
+      "step": 20071
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4095686570797459,
+      "learning_rate": 1.1360088872198764e-05,
+      "loss": 0.7949,
+      "step": 20072
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3938161459804275,
+      "learning_rate": 1.1359332917649427e-05,
+      "loss": 0.6412,
+      "step": 20073
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4776178572666545,
+      "learning_rate": 1.135857695518563e-05,
+      "loss": 0.7096,
+      "step": 20074
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4959304982497121,
+      "learning_rate": 1.1357820984811758e-05,
+      "loss": 0.6798,
+      "step": 20075
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.616223267619129,
+      "learning_rate": 1.135706500653223e-05,
+      "loss": 0.6467,
+      "step": 20076
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.425353975738697,
+      "learning_rate": 1.1356309020351435e-05,
+      "loss": 0.8334,
+      "step": 20077
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2572817402094358,
+      "learning_rate": 1.1355553026273783e-05,
+      "loss": 0.7976,
+      "step": 20078
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4138020096162704,
+      "learning_rate": 1.1354797024303668e-05,
+      "loss": 0.8578,
+      "step": 20079
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8235054896766865,
+      "learning_rate": 1.1354041014445499e-05,
+      "loss": 0.7962,
+      "step": 20080
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.623993889877272,
+      "learning_rate": 1.1353284996703672e-05,
+      "loss": 0.7762,
+      "step": 20081
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3632591593725867,
+      "learning_rate": 1.1352528971082594e-05,
+      "loss": 0.8156,
+      "step": 20082
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3462236344545357,
+      "learning_rate": 1.1351772937586663e-05,
+      "loss": 0.8189,
+      "step": 20083
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4047651773229697,
+      "learning_rate": 1.1351016896220282e-05,
+      "loss": 0.8522,
+      "step": 20084
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3941768010587465,
+      "learning_rate": 1.1350260846987852e-05,
+      "loss": 0.6905,
+      "step": 20085
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6910335261679288,
+      "learning_rate": 1.1349504789893776e-05,
+      "loss": 0.6999,
+      "step": 20086
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3127471850974723,
+      "learning_rate": 1.1348748724942458e-05,
+      "loss": 0.5662,
+      "step": 20087
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3264121518933647,
+      "learning_rate": 1.1347992652138298e-05,
+      "loss": 0.8822,
+      "step": 20088
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3274500811210321,
+      "learning_rate": 1.1347236571485696e-05,
+      "loss": 0.5347,
+      "step": 20089
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.312717160283668,
+      "learning_rate": 1.1346480482989055e-05,
+      "loss": 0.5496,
+      "step": 20090
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.775075148206029,
+      "learning_rate": 1.1345724386652781e-05,
+      "loss": 0.5476,
+      "step": 20091
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.436550643754115,
+      "learning_rate": 1.1344968282481273e-05,
+      "loss": 0.8043,
+      "step": 20092
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3370049679324831,
+      "learning_rate": 1.1344212170478932e-05,
+      "loss": 0.7185,
+      "step": 20093
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.144694865470464,
+      "learning_rate": 1.1343456050650163e-05,
+      "loss": 0.7596,
+      "step": 20094
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2776536907498446,
+      "learning_rate": 1.134269992299937e-05,
+      "loss": 0.8449,
+      "step": 20095
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3750977046967103,
+      "learning_rate": 1.1341943787530951e-05,
+      "loss": 0.7267,
+      "step": 20096
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4656135725825818,
+      "learning_rate": 1.1341187644249311e-05,
+      "loss": 0.8052,
+      "step": 20097
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4309018339563866,
+      "learning_rate": 1.134043149315885e-05,
+      "loss": 0.6822,
+      "step": 20098
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1570106296173703,
+      "learning_rate": 1.1339675334263976e-05,
+      "loss": 0.7027,
+      "step": 20099
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4733781394348957,
+      "learning_rate": 1.1338919167569085e-05,
+      "loss": 0.7059,
+      "step": 20100
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4970611126522795,
+      "learning_rate": 1.1338162993078585e-05,
+      "loss": 0.8355,
+      "step": 20101
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4577092639375693,
+      "learning_rate": 1.1337406810796872e-05,
+      "loss": 0.7384,
+      "step": 20102
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3410911570746633,
+      "learning_rate": 1.133665062072836e-05,
+      "loss": 0.6896,
+      "step": 20103
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.530260755237471,
+      "learning_rate": 1.1335894422877441e-05,
+      "loss": 0.8663,
+      "step": 20104
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6193187764841726,
+      "learning_rate": 1.1335138217248518e-05,
+      "loss": 0.7487,
+      "step": 20105
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3417919987867732,
+      "learning_rate": 1.1334382003846006e-05,
+      "loss": 0.6778,
+      "step": 20106
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4594991251883858,
+      "learning_rate": 1.133362578267429e-05,
+      "loss": 0.7602,
+      "step": 20107
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6669882195685963,
+      "learning_rate": 1.133286955373779e-05,
+      "loss": 0.5111,
+      "step": 20108
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6078329927738768,
+      "learning_rate": 1.1332113317040898e-05,
+      "loss": 0.7988,
+      "step": 20109
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3426873622529125,
+      "learning_rate": 1.1331357072588024e-05,
+      "loss": 0.7703,
+      "step": 20110
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4903521841772371,
+      "learning_rate": 1.1330600820383564e-05,
+      "loss": 0.7434,
+      "step": 20111
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5838721397573677,
+      "learning_rate": 1.1329844560431926e-05,
+      "loss": 0.5381,
+      "step": 20112
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3581833903801608,
+      "learning_rate": 1.132908829273751e-05,
+      "loss": 0.8214,
+      "step": 20113
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4040572053588214,
+      "learning_rate": 1.1328332017304724e-05,
+      "loss": 0.7858,
+      "step": 20114
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5296356709989958,
+      "learning_rate": 1.1327575734137968e-05,
+      "loss": 0.7054,
+      "step": 20115
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.511997543086389,
+      "learning_rate": 1.1326819443241645e-05,
+      "loss": 0.7573,
+      "step": 20116
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3444853604229876,
+      "learning_rate": 1.1326063144620158e-05,
+      "loss": 0.7827,
+      "step": 20117
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.597604899682217,
+      "learning_rate": 1.1325306838277914e-05,
+      "loss": 0.6652,
+      "step": 20118
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8568609388118467,
+      "learning_rate": 1.1324550524219313e-05,
+      "loss": 0.7827,
+      "step": 20119
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5542748968388485,
+      "learning_rate": 1.1323794202448759e-05,
+      "loss": 0.7845,
+      "step": 20120
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5992901182254167,
+      "learning_rate": 1.1323037872970657e-05,
+      "loss": 0.5114,
+      "step": 20121
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7438043176962987,
+      "learning_rate": 1.1322281535789408e-05,
+      "loss": 0.68,
+      "step": 20122
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.433032459120619,
+      "learning_rate": 1.1321525190909419e-05,
+      "loss": 0.8085,
+      "step": 20123
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6695885066265304,
+      "learning_rate": 1.1320768838335092e-05,
+      "loss": 0.6867,
+      "step": 20124
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.493233035058888,
+      "learning_rate": 1.1320012478070831e-05,
+      "loss": 0.8354,
+      "step": 20125
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.613473585691146,
+      "learning_rate": 1.1319256110121036e-05,
+      "loss": 0.6306,
+      "step": 20126
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5773636863977325,
+      "learning_rate": 1.131849973449012e-05,
+      "loss": 0.7709,
+      "step": 20127
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4146944158741337,
+      "learning_rate": 1.1317743351182475e-05,
+      "loss": 0.6312,
+      "step": 20128
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5723990796355933,
+      "learning_rate": 1.1316986960202516e-05,
+      "loss": 0.5024,
+      "step": 20129
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0707956703456785,
+      "learning_rate": 1.1316230561554637e-05,
+      "loss": 0.7339,
+      "step": 20130
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3850614015331493,
+      "learning_rate": 1.131547415524325e-05,
+      "loss": 0.7875,
+      "step": 20131
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5305760568181452,
+      "learning_rate": 1.1314717741272757e-05,
+      "loss": 0.8559,
+      "step": 20132
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6079580565024885,
+      "learning_rate": 1.1313961319647561e-05,
+      "loss": 0.8117,
+      "step": 20133
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3964153605151906,
+      "learning_rate": 1.1313204890372065e-05,
+      "loss": 0.6669,
+      "step": 20134
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.317458188182356,
+      "learning_rate": 1.1312448453450672e-05,
+      "loss": 0.6712,
+      "step": 20135
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6228530299491124,
+      "learning_rate": 1.1311692008887793e-05,
+      "loss": 0.5258,
+      "step": 20136
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6106386185697739,
+      "learning_rate": 1.1310935556687825e-05,
+      "loss": 0.5653,
+      "step": 20137
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2715381767588334,
+      "learning_rate": 1.1310179096855176e-05,
+      "loss": 0.7912,
+      "step": 20138
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3057172091347462,
+      "learning_rate": 1.1309422629394251e-05,
+      "loss": 0.6898,
+      "step": 20139
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.50854537786638,
+      "learning_rate": 1.1308666154309451e-05,
+      "loss": 0.7776,
+      "step": 20140
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1754894953312847,
+      "learning_rate": 1.1307909671605182e-05,
+      "loss": 0.6683,
+      "step": 20141
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6901858642354803,
+      "learning_rate": 1.1307153181285848e-05,
+      "loss": 0.8118,
+      "step": 20142
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4565425976246194,
+      "learning_rate": 1.1306396683355858e-05,
+      "loss": 0.721,
+      "step": 20143
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.456403941604184,
+      "learning_rate": 1.130564017781961e-05,
+      "loss": 0.7201,
+      "step": 20144
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0760458623168256,
+      "learning_rate": 1.1304883664681512e-05,
+      "loss": 0.5287,
+      "step": 20145
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3420607435764809,
+      "learning_rate": 1.1304127143945971e-05,
+      "loss": 0.6736,
+      "step": 20146
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3867142880394838,
+      "learning_rate": 1.1303370615617384e-05,
+      "loss": 0.6936,
+      "step": 20147
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.393707157332468,
+      "learning_rate": 1.1302614079700165e-05,
+      "loss": 0.6418,
+      "step": 20148
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2505541898219226,
+      "learning_rate": 1.130185753619871e-05,
+      "loss": 0.6402,
+      "step": 20149
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.577936624589298,
+      "learning_rate": 1.1301100985117433e-05,
+      "loss": 0.586,
+      "step": 20150
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3749716842319621,
+      "learning_rate": 1.130034442646073e-05,
+      "loss": 0.7115,
+      "step": 20151
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.419843868147952,
+      "learning_rate": 1.1299587860233014e-05,
+      "loss": 0.7269,
+      "step": 20152
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4883700343758581,
+      "learning_rate": 1.1298831286438683e-05,
+      "loss": 0.8732,
+      "step": 20153
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3995039588588016,
+      "learning_rate": 1.1298074705082145e-05,
+      "loss": 0.7765,
+      "step": 20154
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4553226651690478,
+      "learning_rate": 1.1297318116167808e-05,
+      "loss": 0.6248,
+      "step": 20155
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.55010866035452,
+      "learning_rate": 1.129656151970007e-05,
+      "loss": 0.6852,
+      "step": 20156
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4753535281429353,
+      "learning_rate": 1.1295804915683343e-05,
+      "loss": 0.8241,
+      "step": 20157
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.658471317929708,
+      "learning_rate": 1.1295048304122029e-05,
+      "loss": 0.7814,
+      "step": 20158
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4968427665969153,
+      "learning_rate": 1.1294291685020535e-05,
+      "loss": 0.7228,
+      "step": 20159
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5004185982333018,
+      "learning_rate": 1.1293535058383262e-05,
+      "loss": 0.8219,
+      "step": 20160
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.61120019967544,
+      "learning_rate": 1.1292778424214623e-05,
+      "loss": 0.8138,
+      "step": 20161
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4076491250352774,
+      "learning_rate": 1.1292021782519015e-05,
+      "loss": 0.7093,
+      "step": 20162
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3602856775864811,
+      "learning_rate": 1.1291265133300849e-05,
+      "loss": 0.8363,
+      "step": 20163
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3308079891651756,
+      "learning_rate": 1.1290508476564528e-05,
+      "loss": 0.7441,
+      "step": 20164
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6008490558241573,
+      "learning_rate": 1.1289751812314458e-05,
+      "loss": 0.664,
+      "step": 20165
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5777259870392022,
+      "learning_rate": 1.1288995140555044e-05,
+      "loss": 0.5231,
+      "step": 20166
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4619981683760437,
+      "learning_rate": 1.1288238461290693e-05,
+      "loss": 0.8084,
+      "step": 20167
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.649542863408364,
+      "learning_rate": 1.128748177452581e-05,
+      "loss": 0.7452,
+      "step": 20168
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6114873147662255,
+      "learning_rate": 1.12867250802648e-05,
+      "loss": 0.56,
+      "step": 20169
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6166216709922535,
+      "learning_rate": 1.1285968378512069e-05,
+      "loss": 0.4946,
+      "step": 20170
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.374769944057181,
+      "learning_rate": 1.1285211669272023e-05,
+      "loss": 0.7159,
+      "step": 20171
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6082715920029866,
+      "learning_rate": 1.1284454952549067e-05,
+      "loss": 0.538,
+      "step": 20172
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.497220123660339,
+      "learning_rate": 1.1283698228347609e-05,
+      "loss": 0.6553,
+      "step": 20173
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4257421401719124,
+      "learning_rate": 1.1282941496672053e-05,
+      "loss": 0.7756,
+      "step": 20174
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3958901107400656,
+      "learning_rate": 1.1282184757526806e-05,
+      "loss": 0.7275,
+      "step": 20175
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.778273355358889,
+      "learning_rate": 1.1281428010916276e-05,
+      "loss": 0.7675,
+      "step": 20176
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.355940584067365,
+      "learning_rate": 1.128067125684486e-05,
+      "loss": 0.7728,
+      "step": 20177
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6841227933951057,
+      "learning_rate": 1.1279914495316975e-05,
+      "loss": 0.8084,
+      "step": 20178
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.408874288560884,
+      "learning_rate": 1.127915772633702e-05,
+      "loss": 0.6832,
+      "step": 20179
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.299698238828893,
+      "learning_rate": 1.1278400949909407e-05,
+      "loss": 0.6954,
+      "step": 20180
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4549769892543905,
+      "learning_rate": 1.1277644166038535e-05,
+      "loss": 0.7512,
+      "step": 20181
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6231841193370844,
+      "learning_rate": 1.127688737472882e-05,
+      "loss": 0.7923,
+      "step": 20182
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.538064544203494,
+      "learning_rate": 1.1276130575984658e-05,
+      "loss": 0.6453,
+      "step": 20183
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5070584536403342,
+      "learning_rate": 1.1275373769810461e-05,
+      "loss": 0.6392,
+      "step": 20184
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.723080160945571,
+      "learning_rate": 1.1274616956210633e-05,
+      "loss": 0.7549,
+      "step": 20185
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5399561384708509,
+      "learning_rate": 1.1273860135189583e-05,
+      "loss": 0.734,
+      "step": 20186
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5234220802385676,
+      "learning_rate": 1.1273103306751713e-05,
+      "loss": 0.5325,
+      "step": 20187
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5909035352026093,
+      "learning_rate": 1.1272346470901436e-05,
+      "loss": 0.4797,
+      "step": 20188
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4551291822030012,
+      "learning_rate": 1.1271589627643154e-05,
+      "loss": 0.7312,
+      "step": 20189
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2073804968519208,
+      "learning_rate": 1.1270832776981272e-05,
+      "loss": 0.79,
+      "step": 20190
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5023906809402514,
+      "learning_rate": 1.1270075918920202e-05,
+      "loss": 0.6585,
+      "step": 20191
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.606783849925919,
+      "learning_rate": 1.126931905346435e-05,
+      "loss": 0.5692,
+      "step": 20192
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3183683451373853,
+      "learning_rate": 1.1268562180618116e-05,
+      "loss": 0.654,
+      "step": 20193
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5939949919514595,
+      "learning_rate": 1.1267805300385914e-05,
+      "loss": 0.753,
+      "step": 20194
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6439645335929233,
+      "learning_rate": 1.1267048412772147e-05,
+      "loss": 0.5565,
+      "step": 20195
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2158098990462882,
+      "learning_rate": 1.1266291517781225e-05,
+      "loss": 0.7306,
+      "step": 20196
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5921647368032389,
+      "learning_rate": 1.1265534615417553e-05,
+      "loss": 0.5156,
+      "step": 20197
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4060512602316464,
+      "learning_rate": 1.1264777705685535e-05,
+      "loss": 0.7757,
+      "step": 20198
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7439107717409146,
+      "learning_rate": 1.1264020788589582e-05,
+      "loss": 0.6867,
+      "step": 20199
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4514277936154356,
+      "learning_rate": 1.12632638641341e-05,
+      "loss": 0.7444,
+      "step": 20200
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4402083469817253,
+      "learning_rate": 1.1262506932323496e-05,
+      "loss": 0.7263,
+      "step": 20201
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8325106153390696,
+      "learning_rate": 1.1261749993162175e-05,
+      "loss": 0.839,
+      "step": 20202
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6344722152788453,
+      "learning_rate": 1.1260993046654546e-05,
+      "loss": 0.7162,
+      "step": 20203
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4300881115823287,
+      "learning_rate": 1.1260236092805022e-05,
+      "loss": 0.7286,
+      "step": 20204
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6227516460521627,
+      "learning_rate": 1.1259479131617999e-05,
+      "loss": 0.5923,
+      "step": 20205
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8543608029486132,
+      "learning_rate": 1.1258722163097893e-05,
+      "loss": 0.8307,
+      "step": 20206
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.31140001246416,
+      "learning_rate": 1.1257965187249102e-05,
+      "loss": 0.6938,
+      "step": 20207
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4703187543326366,
+      "learning_rate": 1.1257208204076046e-05,
+      "loss": 0.9059,
+      "step": 20208
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6308489983549687,
+      "learning_rate": 1.1256451213583121e-05,
+      "loss": 0.5173,
+      "step": 20209
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5209995429702519,
+      "learning_rate": 1.1255694215774743e-05,
+      "loss": 0.6504,
+      "step": 20210
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5647232427829385,
+      "learning_rate": 1.1254937210655316e-05,
+      "loss": 0.5264,
+      "step": 20211
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9254929696458305,
+      "learning_rate": 1.1254180198229245e-05,
+      "loss": 0.7654,
+      "step": 20212
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.597278380943612,
+      "learning_rate": 1.1253423178500943e-05,
+      "loss": 0.8028,
+      "step": 20213
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8635081071969222,
+      "learning_rate": 1.125266615147481e-05,
+      "loss": 0.6836,
+      "step": 20214
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4648729333553403,
+      "learning_rate": 1.125190911715526e-05,
+      "loss": 0.6854,
+      "step": 20215
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5129739771129205,
+      "learning_rate": 1.1251152075546698e-05,
+      "loss": 0.7145,
+      "step": 20216
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5657770522898595,
+      "learning_rate": 1.1250395026653536e-05,
+      "loss": 0.7188,
+      "step": 20217
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.990319196334366,
+      "learning_rate": 1.1249637970480174e-05,
+      "loss": 0.784,
+      "step": 20218
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6004118393563467,
+      "learning_rate": 1.1248880907031025e-05,
+      "loss": 0.8103,
+      "step": 20219
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.607822247512191,
+      "learning_rate": 1.1248123836310499e-05,
+      "loss": 0.8993,
+      "step": 20220
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4266352356677814,
+      "learning_rate": 1.1247366758322997e-05,
+      "loss": 0.6615,
+      "step": 20221
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4402000686515097,
+      "learning_rate": 1.1246609673072932e-05,
+      "loss": 0.6656,
+      "step": 20222
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5173564542390214,
+      "learning_rate": 1.124585258056471e-05,
+      "loss": 0.7412,
+      "step": 20223
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.557158813700727,
+      "learning_rate": 1.1245095480802743e-05,
+      "loss": 0.853,
+      "step": 20224
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5885195054026686,
+      "learning_rate": 1.1244338373791432e-05,
+      "loss": 0.5552,
+      "step": 20225
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4699166664476793,
+      "learning_rate": 1.1243581259535193e-05,
+      "loss": 0.7031,
+      "step": 20226
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4198246632298717,
+      "learning_rate": 1.1242824138038431e-05,
+      "loss": 0.7088,
+      "step": 20227
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3411270723050126,
+      "learning_rate": 1.1242067009305546e-05,
+      "loss": 0.7415,
+      "step": 20228
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.544078517470382,
+      "learning_rate": 1.1241309873340962e-05,
+      "loss": 0.7405,
+      "step": 20229
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5745775724025175,
+      "learning_rate": 1.1240552730149073e-05,
+      "loss": 0.59,
+      "step": 20230
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3382274358831325,
+      "learning_rate": 1.1239795579734296e-05,
+      "loss": 0.6523,
+      "step": 20231
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8314507715507806,
+      "learning_rate": 1.1239038422101033e-05,
+      "loss": 0.6954,
+      "step": 20232
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.440350354816459,
+      "learning_rate": 1.1238281257253703e-05,
+      "loss": 0.7253,
+      "step": 20233
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5922345861790805,
+      "learning_rate": 1.1237524085196703e-05,
+      "loss": 0.7874,
+      "step": 20234
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9164009055018596,
+      "learning_rate": 1.1236766905934447e-05,
+      "loss": 0.714,
+      "step": 20235
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.299153792436446,
+      "learning_rate": 1.123600971947134e-05,
+      "loss": 0.7618,
+      "step": 20236
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4288184083121844,
+      "learning_rate": 1.1235252525811797e-05,
+      "loss": 0.593,
+      "step": 20237
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5833298943134188,
+      "learning_rate": 1.1234495324960221e-05,
+      "loss": 0.5066,
+      "step": 20238
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.612783119779567,
+      "learning_rate": 1.123373811692102e-05,
+      "loss": 0.719,
+      "step": 20239
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.497591869981051,
+      "learning_rate": 1.123298090169861e-05,
+      "loss": 0.8132,
+      "step": 20240
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4974832872472947,
+      "learning_rate": 1.1232223679297391e-05,
+      "loss": 0.7033,
+      "step": 20241
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9681401845132398,
+      "learning_rate": 1.1231466449721777e-05,
+      "loss": 0.6782,
+      "step": 20242
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4925483231202727,
+      "learning_rate": 1.1230709212976176e-05,
+      "loss": 0.7178,
+      "step": 20243
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5908194155356202,
+      "learning_rate": 1.1229951969064996e-05,
+      "loss": 0.5009,
+      "step": 20244
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9806028233621424,
+      "learning_rate": 1.1229194717992644e-05,
+      "loss": 0.8361,
+      "step": 20245
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.673403459463409,
+      "learning_rate": 1.1228437459763535e-05,
+      "loss": 0.7489,
+      "step": 20246
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.334813731669417,
+      "learning_rate": 1.1227680194382072e-05,
+      "loss": 0.7565,
+      "step": 20247
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4088414159743132,
+      "learning_rate": 1.1226922921852668e-05,
+      "loss": 0.7064,
+      "step": 20248
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.445839867428681,
+      "learning_rate": 1.1226165642179727e-05,
+      "loss": 0.8154,
+      "step": 20249
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4044845590387909,
+      "learning_rate": 1.1225408355367665e-05,
+      "loss": 0.8309,
+      "step": 20250
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8249315546809008,
+      "learning_rate": 1.1224651061420884e-05,
+      "loss": 0.7655,
+      "step": 20251
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5827965547622487,
+      "learning_rate": 1.1223893760343798e-05,
+      "loss": 0.702,
+      "step": 20252
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4208335603162,
+      "learning_rate": 1.1223136452140815e-05,
+      "loss": 0.6902,
+      "step": 20253
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5914574478150245,
+      "learning_rate": 1.1222379136816347e-05,
+      "loss": 0.5245,
+      "step": 20254
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.570678656794413,
+      "learning_rate": 1.12216218143748e-05,
+      "loss": 0.6816,
+      "step": 20255
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3737724862091725,
+      "learning_rate": 1.122086448482058e-05,
+      "loss": 0.7179,
+      "step": 20256
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2614170506814713,
+      "learning_rate": 1.1220107148158104e-05,
+      "loss": 0.6322,
+      "step": 20257
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.669225775209259,
+      "learning_rate": 1.1219349804391774e-05,
+      "loss": 0.6628,
+      "step": 20258
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.042651540980266,
+      "learning_rate": 1.121859245352601e-05,
+      "loss": 0.8575,
+      "step": 20259
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3969663427059236,
+      "learning_rate": 1.1217835095565208e-05,
+      "loss": 0.7739,
+      "step": 20260
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.319811206217898,
+      "learning_rate": 1.1217077730513792e-05,
+      "loss": 0.7704,
+      "step": 20261
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0245355929600244,
+      "learning_rate": 1.1216320358376158e-05,
+      "loss": 0.6566,
+      "step": 20262
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6221839757994604,
+      "learning_rate": 1.1215562979156725e-05,
+      "loss": 0.5282,
+      "step": 20263
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5964498542701545,
+      "learning_rate": 1.1214805592859898e-05,
+      "loss": 0.5233,
+      "step": 20264
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6046009387830737,
+      "learning_rate": 1.1214048199490089e-05,
+      "loss": 0.8165,
+      "step": 20265
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3872583596448302,
+      "learning_rate": 1.1213290799051706e-05,
+      "loss": 0.8172,
+      "step": 20266
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3340667651212055,
+      "learning_rate": 1.121253339154916e-05,
+      "loss": 0.6422,
+      "step": 20267
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5940716688733985,
+      "learning_rate": 1.1211775976986861e-05,
+      "loss": 0.5131,
+      "step": 20268
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3652383561730321,
+      "learning_rate": 1.121101855536922e-05,
+      "loss": 0.6895,
+      "step": 20269
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.362097007549273,
+      "learning_rate": 1.1210261126700643e-05,
+      "loss": 0.71,
+      "step": 20270
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.399834838830477,
+      "learning_rate": 1.1209503690985544e-05,
+      "loss": 0.8336,
+      "step": 20271
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1963520586082284,
+      "learning_rate": 1.1208746248228332e-05,
+      "loss": 0.5922,
+      "step": 20272
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6747028779171538,
+      "learning_rate": 1.1207988798433416e-05,
+      "loss": 0.7292,
+      "step": 20273
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4721226264623763,
+      "learning_rate": 1.1207231341605208e-05,
+      "loss": 0.8717,
+      "step": 20274
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.628730961461185,
+      "learning_rate": 1.1206473877748115e-05,
+      "loss": 0.694,
+      "step": 20275
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.648541329389565,
+      "learning_rate": 1.120571640686655e-05,
+      "loss": 0.7445,
+      "step": 20276
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4833080845347162,
+      "learning_rate": 1.1204958928964922e-05,
+      "loss": 0.7814,
+      "step": 20277
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5821582964670342,
+      "learning_rate": 1.1204201444047644e-05,
+      "loss": 0.7252,
+      "step": 20278
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.777774388588723,
+      "learning_rate": 1.120344395211912e-05,
+      "loss": 0.7469,
+      "step": 20279
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2124126153415562,
+      "learning_rate": 1.1202686453183769e-05,
+      "loss": 0.7517,
+      "step": 20280
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5042520798513852,
+      "learning_rate": 1.120192894724599e-05,
+      "loss": 0.6781,
+      "step": 20281
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3784670151840934,
+      "learning_rate": 1.1201171434310207e-05,
+      "loss": 0.6766,
+      "step": 20282
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4511179765692264,
+      "learning_rate": 1.1200413914380819e-05,
+      "loss": 0.7941,
+      "step": 20283
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1342499325487534,
+      "learning_rate": 1.1199656387462244e-05,
+      "loss": 0.6884,
+      "step": 20284
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3478415821321172,
+      "learning_rate": 1.1198898853558889e-05,
+      "loss": 0.7883,
+      "step": 20285
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6159673914354997,
+      "learning_rate": 1.1198141312675164e-05,
+      "loss": 0.7449,
+      "step": 20286
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4762691912397063,
+      "learning_rate": 1.1197383764815482e-05,
+      "loss": 0.6075,
+      "step": 20287
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.201728010756604,
+      "learning_rate": 1.119662620998425e-05,
+      "loss": 0.6979,
+      "step": 20288
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3416352183546538,
+      "learning_rate": 1.1195868648185886e-05,
+      "loss": 0.7357,
+      "step": 20289
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4343033908800187,
+      "learning_rate": 1.1195111079424791e-05,
+      "loss": 0.8134,
+      "step": 20290
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3582454584286432,
+      "learning_rate": 1.1194353503705387e-05,
+      "loss": 0.7338,
+      "step": 20291
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3442416619636093,
+      "learning_rate": 1.1193595921032074e-05,
+      "loss": 0.6803,
+      "step": 20292
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5350378209273756,
+      "learning_rate": 1.119283833140927e-05,
+      "loss": 0.6654,
+      "step": 20293
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4356571832067517,
+      "learning_rate": 1.1192080734841385e-05,
+      "loss": 0.6096,
+      "step": 20294
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5630428317712292,
+      "learning_rate": 1.1191323131332825e-05,
+      "loss": 0.4853,
+      "step": 20295
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5989713829100874,
+      "learning_rate": 1.1190565520888005e-05,
+      "loss": 0.5135,
+      "step": 20296
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8367724846209448,
+      "learning_rate": 1.1189807903511336e-05,
+      "loss": 0.745,
+      "step": 20297
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7239024937666632,
+      "learning_rate": 1.1189050279207229e-05,
+      "loss": 0.8486,
+      "step": 20298
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.58891343563865,
+      "learning_rate": 1.1188292647980099e-05,
+      "loss": 0.4852,
+      "step": 20299
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8169339369734863,
+      "learning_rate": 1.1187535009834347e-05,
+      "loss": 0.703,
+      "step": 20300
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4205695076850495,
+      "learning_rate": 1.1186777364774393e-05,
+      "loss": 0.8732,
+      "step": 20301
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5891226999169972,
+      "learning_rate": 1.1186019712804644e-05,
+      "loss": 0.4954,
+      "step": 20302
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.575496454565868,
+      "learning_rate": 1.1185262053929513e-05,
+      "loss": 0.7547,
+      "step": 20303
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4367790113442265,
+      "learning_rate": 1.1184504388153413e-05,
+      "loss": 0.6819,
+      "step": 20304
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8587740232259173,
+      "learning_rate": 1.118374671548075e-05,
+      "loss": 0.7733,
+      "step": 20305
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2334324546357716,
+      "learning_rate": 1.1182989035915943e-05,
+      "loss": 0.7182,
+      "step": 20306
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3826517767427629,
+      "learning_rate": 1.1182231349463394e-05,
+      "loss": 0.7109,
+      "step": 20307
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6017411346745382,
+      "learning_rate": 1.1181473656127526e-05,
+      "loss": 0.5348,
+      "step": 20308
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4535740006076516,
+      "learning_rate": 1.118071595591274e-05,
+      "loss": 0.7505,
+      "step": 20309
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4251194118372816,
+      "learning_rate": 1.1179958248823455e-05,
+      "loss": 0.7322,
+      "step": 20310
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5198318716454104,
+      "learning_rate": 1.1179200534864077e-05,
+      "loss": 0.783,
+      "step": 20311
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4730084091903142,
+      "learning_rate": 1.1178442814039021e-05,
+      "loss": 0.6456,
+      "step": 20312
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.401414943755266,
+      "learning_rate": 1.1177685086352699e-05,
+      "loss": 0.6698,
+      "step": 20313
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9506931367958207,
+      "learning_rate": 1.1176927351809519e-05,
+      "loss": 0.738,
+      "step": 20314
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4424650873236131,
+      "learning_rate": 1.1176169610413895e-05,
+      "loss": 0.8106,
+      "step": 20315
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7077774577559184,
+      "learning_rate": 1.1175411862170242e-05,
+      "loss": 0.5882,
+      "step": 20316
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.334726064814503,
+      "learning_rate": 1.117465410708297e-05,
+      "loss": 0.7533,
+      "step": 20317
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3934803932481943,
+      "learning_rate": 1.1173896345156486e-05,
+      "loss": 0.718,
+      "step": 20318
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4735265379410558,
+      "learning_rate": 1.1173138576395208e-05,
+      "loss": 0.7005,
+      "step": 20319
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0769122733652647,
+      "learning_rate": 1.1172380800803544e-05,
+      "loss": 0.8066,
+      "step": 20320
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4021371424457238,
+      "learning_rate": 1.117162301838591e-05,
+      "loss": 0.6841,
+      "step": 20321
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.662292777793806,
+      "learning_rate": 1.1170865229146715e-05,
+      "loss": 0.6953,
+      "step": 20322
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6489245574745843,
+      "learning_rate": 1.1170107433090371e-05,
+      "loss": 0.6783,
+      "step": 20323
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6996366449689186,
+      "learning_rate": 1.1169349630221294e-05,
+      "loss": 0.7757,
+      "step": 20324
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4978590505866496,
+      "learning_rate": 1.116859182054389e-05,
+      "loss": 0.8215,
+      "step": 20325
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4600655702128202,
+      "learning_rate": 1.1167834004062576e-05,
+      "loss": 0.809,
+      "step": 20326
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4958289877476543,
+      "learning_rate": 1.1167076180781764e-05,
+      "loss": 0.6853,
+      "step": 20327
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.331177293210226,
+      "learning_rate": 1.116631835070586e-05,
+      "loss": 0.7525,
+      "step": 20328
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.835786948616294,
+      "learning_rate": 1.1165560513839289e-05,
+      "loss": 0.674,
+      "step": 20329
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5919340360967369,
+      "learning_rate": 1.1164802670186448e-05,
+      "loss": 0.4741,
+      "step": 20330
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6941010131333047,
+      "learning_rate": 1.1164044819751763e-05,
+      "loss": 0.7569,
+      "step": 20331
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1983692706819655,
+      "learning_rate": 1.1163286962539637e-05,
+      "loss": 0.663,
+      "step": 20332
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.35322797303753,
+      "learning_rate": 1.1162529098554489e-05,
+      "loss": 0.8708,
+      "step": 20333
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.588436848642337,
+      "learning_rate": 1.1161771227800727e-05,
+      "loss": 0.6939,
+      "step": 20334
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5181103879837827,
+      "learning_rate": 1.1161013350282763e-05,
+      "loss": 0.7801,
+      "step": 20335
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5944810720039955,
+      "learning_rate": 1.1160255466005014e-05,
+      "loss": 0.7437,
+      "step": 20336
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.917733227633337,
+      "learning_rate": 1.1159497574971888e-05,
+      "loss": 0.7066,
+      "step": 20337
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6444331302708148,
+      "learning_rate": 1.1158739677187802e-05,
+      "loss": 0.8709,
+      "step": 20338
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.603799969631851,
+      "learning_rate": 1.1157981772657165e-05,
+      "loss": 0.4968,
+      "step": 20339
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.014231197646824,
+      "learning_rate": 1.1157223861384394e-05,
+      "loss": 0.7658,
+      "step": 20340
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0478049886930827,
+      "learning_rate": 1.1156465943373897e-05,
+      "loss": 0.8255,
+      "step": 20341
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8801393341709538,
+      "learning_rate": 1.1155708018630094e-05,
+      "loss": 0.6657,
+      "step": 20342
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3414132395085703,
+      "learning_rate": 1.1154950087157389e-05,
+      "loss": 0.6648,
+      "step": 20343
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4560913057986782,
+      "learning_rate": 1.1154192148960199e-05,
+      "loss": 0.7164,
+      "step": 20344
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5033789762839986,
+      "learning_rate": 1.1153434204042936e-05,
+      "loss": 0.7131,
+      "step": 20345
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5665181866262086,
+      "learning_rate": 1.1152676252410015e-05,
+      "loss": 0.6257,
+      "step": 20346
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7314660944104299,
+      "learning_rate": 1.115191829406585e-05,
+      "loss": 0.7579,
+      "step": 20347
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6260717719824535,
+      "learning_rate": 1.1151160329014849e-05,
+      "loss": 0.5437,
+      "step": 20348
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5994998597507414,
+      "learning_rate": 1.1150402357261429e-05,
+      "loss": 0.5061,
+      "step": 20349
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0956271506982906,
+      "learning_rate": 1.1149644378810003e-05,
+      "loss": 0.6691,
+      "step": 20350
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6127375069915825,
+      "learning_rate": 1.1148886393664984e-05,
+      "loss": 0.5218,
+      "step": 20351
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6627383950297772,
+      "learning_rate": 1.1148128401830783e-05,
+      "loss": 0.7755,
+      "step": 20352
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6686341994672613,
+      "learning_rate": 1.1147370403311815e-05,
+      "loss": 0.8233,
+      "step": 20353
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.395878566706194,
+      "learning_rate": 1.1146612398112492e-05,
+      "loss": 0.7714,
+      "step": 20354
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.548202410627129,
+      "learning_rate": 1.1145854386237231e-05,
+      "loss": 0.7272,
+      "step": 20355
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4828428618936276,
+      "learning_rate": 1.1145096367690445e-05,
+      "loss": 0.722,
+      "step": 20356
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8194530615600182,
+      "learning_rate": 1.1144338342476544e-05,
+      "loss": 0.6337,
+      "step": 20357
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8504152097022866,
+      "learning_rate": 1.114358031059994e-05,
+      "loss": 0.5997,
+      "step": 20358
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4097769051341016,
+      "learning_rate": 1.1142822272065053e-05,
+      "loss": 0.7727,
+      "step": 20359
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5730267291863615,
+      "learning_rate": 1.1142064226876289e-05,
+      "loss": 0.7284,
+      "step": 20360
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5854486323591392,
+      "learning_rate": 1.1141306175038069e-05,
+      "loss": 0.6727,
+      "step": 20361
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4444671015798534,
+      "learning_rate": 1.1140548116554799e-05,
+      "loss": 0.8307,
+      "step": 20362
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5984135239180376,
+      "learning_rate": 1.1139790051430901e-05,
+      "loss": 0.5006,
+      "step": 20363
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.413155686875955,
+      "learning_rate": 1.1139031979670785e-05,
+      "loss": 0.7401,
+      "step": 20364
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4558760293460766,
+      "learning_rate": 1.1138273901278863e-05,
+      "loss": 0.8115,
+      "step": 20365
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2881937424493333,
+      "learning_rate": 1.1137515816259548e-05,
+      "loss": 0.7398,
+      "step": 20366
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2369259712746605,
+      "learning_rate": 1.1136757724617256e-05,
+      "loss": 0.717,
+      "step": 20367
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6967679106841205,
+      "learning_rate": 1.1135999626356403e-05,
+      "loss": 0.6955,
+      "step": 20368
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.04281674559174,
+      "learning_rate": 1.1135241521481397e-05,
+      "loss": 0.7049,
+      "step": 20369
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2967820421365264,
+      "learning_rate": 1.1134483409996657e-05,
+      "loss": 0.7435,
+      "step": 20370
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4424708164163444,
+      "learning_rate": 1.1133725291906596e-05,
+      "loss": 0.699,
+      "step": 20371
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5633808574365418,
+      "learning_rate": 1.1132967167215627e-05,
+      "loss": 0.5062,
+      "step": 20372
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3703690344540111,
+      "learning_rate": 1.1132209035928166e-05,
+      "loss": 0.7411,
+      "step": 20373
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5148751267690963,
+      "learning_rate": 1.1131450898048624e-05,
+      "loss": 0.7298,
+      "step": 20374
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4870899522045218,
+      "learning_rate": 1.1130692753581416e-05,
+      "loss": 0.6386,
+      "step": 20375
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4458969913994524,
+      "learning_rate": 1.1129934602530959e-05,
+      "loss": 0.7564,
+      "step": 20376
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5669514558782013,
+      "learning_rate": 1.1129176444901663e-05,
+      "loss": 0.4759,
+      "step": 20377
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4323893894468072,
+      "learning_rate": 1.1128418280697948e-05,
+      "loss": 0.7165,
+      "step": 20378
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3052370786722352,
+      "learning_rate": 1.1127660109924218e-05,
+      "loss": 0.6746,
+      "step": 20379
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5756501316702455,
+      "learning_rate": 1.11269019325849e-05,
+      "loss": 0.5212,
+      "step": 20380
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5566081089992125,
+      "learning_rate": 1.1126143748684397e-05,
+      "loss": 0.7082,
+      "step": 20381
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9263276875607565,
+      "learning_rate": 1.1125385558227136e-05,
+      "loss": 0.7455,
+      "step": 20382
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3009622455045595,
+      "learning_rate": 1.1124627361217514e-05,
+      "loss": 0.6697,
+      "step": 20383
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4877606391289275,
+      "learning_rate": 1.1123869157659962e-05,
+      "loss": 0.6577,
+      "step": 20384
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3942849483161808,
+      "learning_rate": 1.1123110947558888e-05,
+      "loss": 0.6822,
+      "step": 20385
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8317049868872528,
+      "learning_rate": 1.1122352730918704e-05,
+      "loss": 0.821,
+      "step": 20386
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5462269633121737,
+      "learning_rate": 1.112159450774383e-05,
+      "loss": 0.5784,
+      "step": 20387
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2822845794884252,
+      "learning_rate": 1.1120836278038674e-05,
+      "loss": 0.8029,
+      "step": 20388
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.575107059383286,
+      "learning_rate": 1.1120078041807658e-05,
+      "loss": 0.7625,
+      "step": 20389
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2847331094361336,
+      "learning_rate": 1.1119319799055188e-05,
+      "loss": 0.5645,
+      "step": 20390
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3303126693410274,
+      "learning_rate": 1.111856154978569e-05,
+      "loss": 0.8086,
+      "step": 20391
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4864559611659545,
+      "learning_rate": 1.1117803294003568e-05,
+      "loss": 0.7293,
+      "step": 20392
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4501606287645827,
+      "learning_rate": 1.1117045031713242e-05,
+      "loss": 0.714,
+      "step": 20393
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5080798272518343,
+      "learning_rate": 1.1116286762919129e-05,
+      "loss": 0.6292,
+      "step": 20394
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4414279165697428,
+      "learning_rate": 1.1115528487625638e-05,
+      "loss": 0.8133,
+      "step": 20395
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4109794677133347,
+      "learning_rate": 1.1114770205837189e-05,
+      "loss": 0.8388,
+      "step": 20396
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5656294534147153,
+      "learning_rate": 1.1114011917558194e-05,
+      "loss": 0.5194,
+      "step": 20397
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3986988239382252,
+      "learning_rate": 1.1113253622793067e-05,
+      "loss": 0.7607,
+      "step": 20398
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5709983255516542,
+      "learning_rate": 1.1112495321546228e-05,
+      "loss": 0.7539,
+      "step": 20399
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9984079266392953,
+      "learning_rate": 1.1111737013822089e-05,
+      "loss": 0.6597,
+      "step": 20400
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3323668325025675,
+      "learning_rate": 1.1110978699625061e-05,
+      "loss": 0.8034,
+      "step": 20401
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4027934680500278,
+      "learning_rate": 1.1110220378959568e-05,
+      "loss": 0.7865,
+      "step": 20402
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8585415571254744,
+      "learning_rate": 1.1109462051830018e-05,
+      "loss": 0.6987,
+      "step": 20403
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4607408914877162,
+      "learning_rate": 1.110870371824083e-05,
+      "loss": 0.7787,
+      "step": 20404
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.467535897754775,
+      "learning_rate": 1.1107945378196417e-05,
+      "loss": 0.7911,
+      "step": 20405
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.1819974953475225,
+      "learning_rate": 1.1107187031701194e-05,
+      "loss": 0.8788,
+      "step": 20406
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4610097755475115,
+      "learning_rate": 1.110642867875958e-05,
+      "loss": 0.6187,
+      "step": 20407
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.414175002245173,
+      "learning_rate": 1.110567031937599e-05,
+      "loss": 0.5903,
+      "step": 20408
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5104799003260174,
+      "learning_rate": 1.1104911953554831e-05,
+      "loss": 0.756,
+      "step": 20409
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3807946925500185,
+      "learning_rate": 1.1104153581300532e-05,
+      "loss": 0.7247,
+      "step": 20410
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6524278415506062,
+      "learning_rate": 1.1103395202617495e-05,
+      "loss": 0.7305,
+      "step": 20411
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4322329679657666,
+      "learning_rate": 1.1102636817510147e-05,
+      "loss": 0.7282,
+      "step": 20412
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3266662165499328,
+      "learning_rate": 1.1101878425982892e-05,
+      "loss": 0.6874,
+      "step": 20413
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2757387822694881,
+      "learning_rate": 1.1101120028040158e-05,
+      "loss": 0.8083,
+      "step": 20414
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2143111777607196,
+      "learning_rate": 1.1100361623686352e-05,
+      "loss": 0.6315,
+      "step": 20415
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4334897026219304,
+      "learning_rate": 1.1099603212925893e-05,
+      "loss": 0.6249,
+      "step": 20416
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.424118057094679,
+      "learning_rate": 1.1098844795763195e-05,
+      "loss": 0.9158,
+      "step": 20417
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3466668455337676,
+      "learning_rate": 1.1098086372202674e-05,
+      "loss": 0.7216,
+      "step": 20418
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4497108477791651,
+      "learning_rate": 1.109732794224875e-05,
+      "loss": 0.7848,
+      "step": 20419
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3961262340192346,
+      "learning_rate": 1.1096569505905832e-05,
+      "loss": 0.6164,
+      "step": 20420
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.421795371696111,
+      "learning_rate": 1.109581106317834e-05,
+      "loss": 0.6959,
+      "step": 20421
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.278746870660772,
+      "learning_rate": 1.109505261407069e-05,
+      "loss": 0.6913,
+      "step": 20422
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2710393413132537,
+      "learning_rate": 1.1094294158587295e-05,
+      "loss": 0.7324,
+      "step": 20423
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3518140498915556,
+      "learning_rate": 1.1093535696732574e-05,
+      "loss": 0.8432,
+      "step": 20424
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3684837040109608,
+      "learning_rate": 1.1092777228510943e-05,
+      "loss": 0.8365,
+      "step": 20425
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.416777334649246,
+      "learning_rate": 1.1092018753926817e-05,
+      "loss": 0.7217,
+      "step": 20426
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4273522427080447,
+      "learning_rate": 1.109126027298461e-05,
+      "loss": 0.6616,
+      "step": 20427
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6167623148408936,
+      "learning_rate": 1.109050178568874e-05,
+      "loss": 0.6812,
+      "step": 20428
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3458644660963828,
+      "learning_rate": 1.1089743292043629e-05,
+      "loss": 0.5804,
+      "step": 20429
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.203123981661278,
+      "learning_rate": 1.1088984792053681e-05,
+      "loss": 0.7184,
+      "step": 20430
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5848908179605784,
+      "learning_rate": 1.1088226285723323e-05,
+      "loss": 0.5524,
+      "step": 20431
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2790575137165576,
+      "learning_rate": 1.1087467773056962e-05,
+      "loss": 0.7742,
+      "step": 20432
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6063438588040146,
+      "learning_rate": 1.1086709254059025e-05,
+      "loss": 0.543,
+      "step": 20433
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4244011724440122,
+      "learning_rate": 1.1085950728733919e-05,
+      "loss": 0.7605,
+      "step": 20434
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2757347618498707,
+      "learning_rate": 1.1085192197086063e-05,
+      "loss": 0.8108,
+      "step": 20435
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4263971292411473,
+      "learning_rate": 1.108443365911988e-05,
+      "loss": 0.7159,
+      "step": 20436
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4650957855721625,
+      "learning_rate": 1.1083675114839775e-05,
+      "loss": 0.7504,
+      "step": 20437
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2562494807727302,
+      "learning_rate": 1.1082916564250175e-05,
+      "loss": 0.8087,
+      "step": 20438
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6361697887624187,
+      "learning_rate": 1.1082158007355487e-05,
+      "loss": 0.7227,
+      "step": 20439
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.689858113285003,
+      "learning_rate": 1.108139944416014e-05,
+      "loss": 0.6037,
+      "step": 20440
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6255152975536131,
+      "learning_rate": 1.1080640874668535e-05,
+      "loss": 0.548,
+      "step": 20441
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.772281959491575,
+      "learning_rate": 1.1079882298885102e-05,
+      "loss": 0.7334,
+      "step": 20442
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.1581750998738336,
+      "learning_rate": 1.107912371681425e-05,
+      "loss": 0.7176,
+      "step": 20443
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5305098441246954,
+      "learning_rate": 1.1078365128460398e-05,
+      "loss": 0.6449,
+      "step": 20444
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7421493211027441,
+      "learning_rate": 1.1077606533827965e-05,
+      "loss": 0.8322,
+      "step": 20445
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9169156044925166,
+      "learning_rate": 1.1076847932921364e-05,
+      "loss": 0.713,
+      "step": 20446
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4662541886061333,
+      "learning_rate": 1.1076089325745013e-05,
+      "loss": 0.802,
+      "step": 20447
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5661890319623448,
+      "learning_rate": 1.1075330712303329e-05,
+      "loss": 0.7135,
+      "step": 20448
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5562982574301976,
+      "learning_rate": 1.107457209260073e-05,
+      "loss": 0.6045,
+      "step": 20449
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8041143771814512,
+      "learning_rate": 1.1073813466641633e-05,
+      "loss": 0.7015,
+      "step": 20450
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4557324087663204,
+      "learning_rate": 1.1073054834430452e-05,
+      "loss": 0.793,
+      "step": 20451
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4996283514497148,
+      "learning_rate": 1.1072296195971606e-05,
+      "loss": 0.758,
+      "step": 20452
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.373711028973864,
+      "learning_rate": 1.1071537551269513e-05,
+      "loss": 0.7253,
+      "step": 20453
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6617968527368483,
+      "learning_rate": 1.107077890032859e-05,
+      "loss": 0.7443,
+      "step": 20454
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2867160985758437,
+      "learning_rate": 1.1070020243153253e-05,
+      "loss": 0.7678,
+      "step": 20455
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5191321132531774,
+      "learning_rate": 1.1069261579747919e-05,
+      "loss": 0.6777,
+      "step": 20456
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5918057656643235,
+      "learning_rate": 1.1068502910117006e-05,
+      "loss": 0.5043,
+      "step": 20457
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5282179359418229,
+      "learning_rate": 1.1067744234264928e-05,
+      "loss": 0.6111,
+      "step": 20458
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.504484061520473,
+      "learning_rate": 1.106698555219611e-05,
+      "loss": 0.7611,
+      "step": 20459
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4859392987531774,
+      "learning_rate": 1.106622686391496e-05,
+      "loss": 0.8558,
+      "step": 20460
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1129684869379384,
+      "learning_rate": 1.1065468169425902e-05,
+      "loss": 0.5597,
+      "step": 20461
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4477749284545576,
+      "learning_rate": 1.1064709468733348e-05,
+      "loss": 0.783,
+      "step": 20462
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6480634301662715,
+      "learning_rate": 1.1063950761841722e-05,
+      "loss": 0.8076,
+      "step": 20463
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.605866420291617,
+      "learning_rate": 1.1063192048755436e-05,
+      "loss": 0.6482,
+      "step": 20464
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.540530408313623,
+      "learning_rate": 1.106243332947891e-05,
+      "loss": 0.6522,
+      "step": 20465
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.474616857081025,
+      "learning_rate": 1.106167460401656e-05,
+      "loss": 0.7961,
+      "step": 20466
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.624074766825429,
+      "learning_rate": 1.1060915872372804e-05,
+      "loss": 0.8129,
+      "step": 20467
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4376892970001822,
+      "learning_rate": 1.1060157134552061e-05,
+      "loss": 0.6419,
+      "step": 20468
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5896068916601632,
+      "learning_rate": 1.105939839055875e-05,
+      "loss": 0.5549,
+      "step": 20469
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7271951237987615,
+      "learning_rate": 1.1058639640397282e-05,
+      "loss": 0.7209,
+      "step": 20470
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.707836503227086,
+      "learning_rate": 1.1057880884072078e-05,
+      "loss": 0.7653,
+      "step": 20471
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6819271506555182,
+      "learning_rate": 1.1057122121587561e-05,
+      "loss": 0.7134,
+      "step": 20472
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5084215222678765,
+      "learning_rate": 1.1056363352948141e-05,
+      "loss": 0.7484,
+      "step": 20473
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.620634773612602,
+      "learning_rate": 1.105560457815824e-05,
+      "loss": 0.5454,
+      "step": 20474
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4952522930953893,
+      "learning_rate": 1.1054845797222276e-05,
+      "loss": 0.6473,
+      "step": 20475
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4058795407462812,
+      "learning_rate": 1.1054087010144662e-05,
+      "loss": 0.6813,
+      "step": 20476
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.3553667050331066,
+      "learning_rate": 1.1053328216929823e-05,
+      "loss": 0.6512,
+      "step": 20477
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3776841850407542,
+      "learning_rate": 1.1052569417582174e-05,
+      "loss": 0.6788,
+      "step": 20478
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4918081838451835,
+      "learning_rate": 1.1051810612106134e-05,
+      "loss": 0.8491,
+      "step": 20479
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4350245833605382,
+      "learning_rate": 1.1051051800506118e-05,
+      "loss": 0.5812,
+      "step": 20480
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.525590704880515,
+      "learning_rate": 1.1050292982786543e-05,
+      "loss": 0.7271,
+      "step": 20481
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4064796932063959,
+      "learning_rate": 1.1049534158951834e-05,
+      "loss": 0.8235,
+      "step": 20482
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6338945790336832,
+      "learning_rate": 1.10487753290064e-05,
+      "loss": 0.797,
+      "step": 20483
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4641421111504365,
+      "learning_rate": 1.1048016492954668e-05,
+      "loss": 0.8133,
+      "step": 20484
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.219574982955543,
+      "learning_rate": 1.104725765080105e-05,
+      "loss": 0.7582,
+      "step": 20485
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.497680331803855,
+      "learning_rate": 1.1046498802549968e-05,
+      "loss": 0.791,
+      "step": 20486
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.391330383652707,
+      "learning_rate": 1.104573994820584e-05,
+      "loss": 0.8496,
+      "step": 20487
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.465681200888665,
+      "learning_rate": 1.104498108777308e-05,
+      "loss": 0.726,
+      "step": 20488
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4879149286454905,
+      "learning_rate": 1.104422222125611e-05,
+      "loss": 0.8213,
+      "step": 20489
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6644326224139492,
+      "learning_rate": 1.1043463348659346e-05,
+      "loss": 0.6906,
+      "step": 20490
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2984631123859571,
+      "learning_rate": 1.1042704469987213e-05,
+      "loss": 0.746,
+      "step": 20491
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4178326950294957,
+      "learning_rate": 1.1041945585244118e-05,
+      "loss": 0.7253,
+      "step": 20492
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4674905956965185,
+      "learning_rate": 1.1041186694434492e-05,
+      "loss": 0.7521,
+      "step": 20493
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5092913943563522,
+      "learning_rate": 1.1040427797562744e-05,
+      "loss": 0.6564,
+      "step": 20494
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.435417833775247,
+      "learning_rate": 1.1039668894633298e-05,
+      "loss": 0.6973,
+      "step": 20495
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4130365598426835,
+      "learning_rate": 1.1038909985650568e-05,
+      "loss": 0.7692,
+      "step": 20496
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.471286523849901,
+      "learning_rate": 1.1038151070618976e-05,
+      "loss": 0.7198,
+      "step": 20497
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5703055339197551,
+      "learning_rate": 1.103739214954294e-05,
+      "loss": 0.509,
+      "step": 20498
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.097501270161648,
+      "learning_rate": 1.103663322242688e-05,
+      "loss": 0.585,
+      "step": 20499
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.603140469473524,
+      "learning_rate": 1.1035874289275212e-05,
+      "loss": 0.6822,
+      "step": 20500
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5155345905183115,
+      "learning_rate": 1.1035115350092356e-05,
+      "loss": 0.641,
+      "step": 20501
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5681198489007272,
+      "learning_rate": 1.1034356404882728e-05,
+      "loss": 0.5044,
+      "step": 20502
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2870285428653354,
+      "learning_rate": 1.1033597453650752e-05,
+      "loss": 0.8085,
+      "step": 20503
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8998354935827448,
+      "learning_rate": 1.1032838496400846e-05,
+      "loss": 0.7385,
+      "step": 20504
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.35854016642718,
+      "learning_rate": 1.1032079533137426e-05,
+      "loss": 0.7132,
+      "step": 20505
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5727110909902466,
+      "learning_rate": 1.1031320563864912e-05,
+      "loss": 0.6517,
+      "step": 20506
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.466829844252207,
+      "learning_rate": 1.1030561588587722e-05,
+      "loss": 0.7427,
+      "step": 20507
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.64965566431979,
+      "learning_rate": 1.102980260731028e-05,
+      "loss": 0.7385,
+      "step": 20508
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6303318483754508,
+      "learning_rate": 1.1029043620036996e-05,
+      "loss": 0.5028,
+      "step": 20509
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2827231698474781,
+      "learning_rate": 1.1028284626772299e-05,
+      "loss": 0.7043,
+      "step": 20510
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3262301926820041,
+      "learning_rate": 1.1027525627520599e-05,
+      "loss": 0.7206,
+      "step": 20511
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5237353060078271,
+      "learning_rate": 1.1026766622286322e-05,
+      "loss": 0.6447,
+      "step": 20512
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.406111566138063,
+      "learning_rate": 1.1026007611073883e-05,
+      "loss": 0.794,
+      "step": 20513
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.295575702706573,
+      "learning_rate": 1.1025248593887707e-05,
+      "loss": 0.8329,
+      "step": 20514
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0835870434590116,
+      "learning_rate": 1.1024489570732207e-05,
+      "loss": 0.7491,
+      "step": 20515
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5073528016382867,
+      "learning_rate": 1.10237305416118e-05,
+      "loss": 0.6959,
+      "step": 20516
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5531339107369104,
+      "learning_rate": 1.1022971506530915e-05,
+      "loss": 0.6687,
+      "step": 20517
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3312642597476763,
+      "learning_rate": 1.1022212465493965e-05,
+      "loss": 0.6474,
+      "step": 20518
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3883681020693133,
+      "learning_rate": 1.102145341850537e-05,
+      "loss": 0.7145,
+      "step": 20519
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3187815883522966,
+      "learning_rate": 1.1020694365569548e-05,
+      "loss": 0.7766,
+      "step": 20520
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9368529992619068,
+      "learning_rate": 1.1019935306690921e-05,
+      "loss": 0.5028,
+      "step": 20521
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.503179290057419,
+      "learning_rate": 1.101917624187391e-05,
+      "loss": 0.7546,
+      "step": 20522
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8172080477064851,
+      "learning_rate": 1.101841717112293e-05,
+      "loss": 0.7657,
+      "step": 20523
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4691016031358373,
+      "learning_rate": 1.1017658094442403e-05,
+      "loss": 0.6764,
+      "step": 20524
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4333833412679013,
+      "learning_rate": 1.1016899011836749e-05,
+      "loss": 0.8045,
+      "step": 20525
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5678579303495677,
+      "learning_rate": 1.1016139923310388e-05,
+      "loss": 0.7756,
+      "step": 20526
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4300266119823037,
+      "learning_rate": 1.1015380828867736e-05,
+      "loss": 0.7205,
+      "step": 20527
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9236831726009826,
+      "learning_rate": 1.1014621728513218e-05,
+      "loss": 0.8216,
+      "step": 20528
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.841706068522066,
+      "learning_rate": 1.101386262225125e-05,
+      "loss": 0.7988,
+      "step": 20529
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2275027750399985,
+      "learning_rate": 1.1013103510086252e-05,
+      "loss": 0.6991,
+      "step": 20530
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4813920593014098,
+      "learning_rate": 1.1012344392022647e-05,
+      "loss": 0.7189,
+      "step": 20531
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6271683754885226,
+      "learning_rate": 1.1011585268064849e-05,
+      "loss": 0.5495,
+      "step": 20532
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8589411700129275,
+      "learning_rate": 1.1010826138217284e-05,
+      "loss": 0.8032,
+      "step": 20533
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5481525656450086,
+      "learning_rate": 1.1010067002484368e-05,
+      "loss": 0.6497,
+      "step": 20534
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4439829020967723,
+      "learning_rate": 1.1009307860870522e-05,
+      "loss": 0.6848,
+      "step": 20535
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4159519452130722,
+      "learning_rate": 1.1008548713380167e-05,
+      "loss": 0.6336,
+      "step": 20536
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6612842433569501,
+      "learning_rate": 1.1007789560017722e-05,
+      "loss": 0.7681,
+      "step": 20537
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.239907344345721,
+      "learning_rate": 1.1007030400787608e-05,
+      "loss": 0.7797,
+      "step": 20538
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.546692976667651,
+      "learning_rate": 1.1006271235694242e-05,
+      "loss": 0.6477,
+      "step": 20539
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.387903405424281,
+      "learning_rate": 1.1005512064742048e-05,
+      "loss": 0.8313,
+      "step": 20540
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4724432122543385,
+      "learning_rate": 1.1004752887935442e-05,
+      "loss": 0.7496,
+      "step": 20541
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5914576754403098,
+      "learning_rate": 1.100399370527885e-05,
+      "loss": 0.6793,
+      "step": 20542
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7926162926120623,
+      "learning_rate": 1.1003234516776685e-05,
+      "loss": 0.8699,
+      "step": 20543
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5544465704771355,
+      "learning_rate": 1.1002475322433374e-05,
+      "loss": 0.8002,
+      "step": 20544
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3206244597459316,
+      "learning_rate": 1.1001716122253335e-05,
+      "loss": 0.7694,
+      "step": 20545
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.467415651752284,
+      "learning_rate": 1.1000956916240985e-05,
+      "loss": 0.6362,
+      "step": 20546
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8820175387081286,
+      "learning_rate": 1.1000197704400748e-05,
+      "loss": 0.7094,
+      "step": 20547
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2733771987723315,
+      "learning_rate": 1.0999438486737043e-05,
+      "loss": 0.6608,
+      "step": 20548
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3645961984110684,
+      "learning_rate": 1.099867926325429e-05,
+      "loss": 0.6961,
+      "step": 20549
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6758282688430144,
+      "learning_rate": 1.0997920033956912e-05,
+      "loss": 0.78,
+      "step": 20550
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2765895740795241,
+      "learning_rate": 1.0997160798849327e-05,
+      "loss": 0.6242,
+      "step": 20551
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.472630515025093,
+      "learning_rate": 1.0996401557935957e-05,
+      "loss": 0.6326,
+      "step": 20552
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6835211708000966,
+      "learning_rate": 1.099564231122122e-05,
+      "loss": 0.8162,
+      "step": 20553
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.330144742342562,
+      "learning_rate": 1.0994883058709538e-05,
+      "loss": 0.6465,
+      "step": 20554
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2729627448987095,
+      "learning_rate": 1.0994123800405332e-05,
+      "loss": 0.8074,
+      "step": 20555
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7718960386041578,
+      "learning_rate": 1.0993364536313022e-05,
+      "loss": 0.7517,
+      "step": 20556
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4517107827665674,
+      "learning_rate": 1.0992605266437032e-05,
+      "loss": 0.7017,
+      "step": 20557
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5240190493616173,
+      "learning_rate": 1.0991845990781777e-05,
+      "loss": 0.6354,
+      "step": 20558
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7005537481946926,
+      "learning_rate": 1.0991086709351686e-05,
+      "loss": 0.6848,
+      "step": 20559
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.674010772408068,
+      "learning_rate": 1.0990327422151168e-05,
+      "loss": 0.5716,
+      "step": 20560
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3668650947135599,
+      "learning_rate": 1.0989568129184654e-05,
+      "loss": 0.5838,
+      "step": 20561
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.198857170396965,
+      "learning_rate": 1.0988808830456557e-05,
+      "loss": 0.6672,
+      "step": 20562
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3125212401473052,
+      "learning_rate": 1.0988049525971307e-05,
+      "loss": 0.6669,
+      "step": 20563
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.362289729675004,
+      "learning_rate": 1.0987290215733315e-05,
+      "loss": 0.681,
+      "step": 20564
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.371922600219552,
+      "learning_rate": 1.098653089974701e-05,
+      "loss": 0.8622,
+      "step": 20565
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.563716513392116,
+      "learning_rate": 1.0985771578016808e-05,
+      "loss": 0.6802,
+      "step": 20566
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4591866696723874,
+      "learning_rate": 1.0985012250547133e-05,
+      "loss": 0.6955,
+      "step": 20567
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5236205985730158,
+      "learning_rate": 1.0984252917342406e-05,
+      "loss": 0.6795,
+      "step": 20568
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3902558221725907,
+      "learning_rate": 1.0983493578407044e-05,
+      "loss": 0.8417,
+      "step": 20569
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2847570626541522,
+      "learning_rate": 1.0982734233745473e-05,
+      "loss": 0.7253,
+      "step": 20570
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4867909201424887,
+      "learning_rate": 1.098197488336211e-05,
+      "loss": 0.6659,
+      "step": 20571
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.34633046535174,
+      "learning_rate": 1.0981215527261383e-05,
+      "loss": 0.6345,
+      "step": 20572
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4395701875027807,
+      "learning_rate": 1.0980456165447704e-05,
+      "loss": 0.6694,
+      "step": 20573
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5827339399664098,
+      "learning_rate": 1.09796967979255e-05,
+      "loss": 0.5273,
+      "step": 20574
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6087990983498219,
+      "learning_rate": 1.0978937424699191e-05,
+      "loss": 0.5152,
+      "step": 20575
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3089703486912732,
+      "learning_rate": 1.0978178045773197e-05,
+      "loss": 0.7143,
+      "step": 20576
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3012059893803813,
+      "learning_rate": 1.0977418661151943e-05,
+      "loss": 0.6285,
+      "step": 20577
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.109821894267246,
+      "learning_rate": 1.097665927083985e-05,
+      "loss": 0.7136,
+      "step": 20578
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6084629035532123,
+      "learning_rate": 1.0975899874841334e-05,
+      "loss": 0.709,
+      "step": 20579
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.068803955553594,
+      "learning_rate": 1.0975140473160824e-05,
+      "loss": 0.8399,
+      "step": 20580
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.383509510541124,
+      "learning_rate": 1.0974381065802731e-05,
+      "loss": 0.7958,
+      "step": 20581
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2060909272733322,
+      "learning_rate": 1.0973621652771487e-05,
+      "loss": 0.6563,
+      "step": 20582
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7276901897840464,
+      "learning_rate": 1.0972862234071508e-05,
+      "loss": 0.7708,
+      "step": 20583
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.1535153196142254,
+      "learning_rate": 1.0972102809707219e-05,
+      "loss": 0.8035,
+      "step": 20584
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3045859905233201,
+      "learning_rate": 1.0971343379683037e-05,
+      "loss": 0.7039,
+      "step": 20585
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3430076261494743,
+      "learning_rate": 1.0970583944003388e-05,
+      "loss": 0.7163,
+      "step": 20586
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3718525314391088,
+      "learning_rate": 1.0969824502672694e-05,
+      "loss": 0.6593,
+      "step": 20587
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5297515600265077,
+      "learning_rate": 1.0969065055695371e-05,
+      "loss": 0.7089,
+      "step": 20588
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.913025750245925,
+      "learning_rate": 1.0968305603075849e-05,
+      "loss": 0.7153,
+      "step": 20589
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6380358169904323,
+      "learning_rate": 1.096754614481854e-05,
+      "loss": 0.7506,
+      "step": 20590
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3497649099111713,
+      "learning_rate": 1.0966786680927875e-05,
+      "loss": 0.7013,
+      "step": 20591
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.880324500708282,
+      "learning_rate": 1.0966027211408268e-05,
+      "loss": 0.7409,
+      "step": 20592
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3933309346789207,
+      "learning_rate": 1.096526773626415e-05,
+      "loss": 0.8039,
+      "step": 20593
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3603280691028614,
+      "learning_rate": 1.0964508255499933e-05,
+      "loss": 0.6697,
+      "step": 20594
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5726270092261845,
+      "learning_rate": 1.0963748769120046e-05,
+      "loss": 0.6437,
+      "step": 20595
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.846968020299983,
+      "learning_rate": 1.0962989277128909e-05,
+      "loss": 0.6109,
+      "step": 20596
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6167084332147347,
+      "learning_rate": 1.0962229779530941e-05,
+      "loss": 0.5248,
+      "step": 20597
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5268328077993991,
+      "learning_rate": 1.0961470276330568e-05,
+      "loss": 0.7811,
+      "step": 20598
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4187105358353376,
+      "learning_rate": 1.0960710767532209e-05,
+      "loss": 0.7525,
+      "step": 20599
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.180803164867487,
+      "learning_rate": 1.095995125314029e-05,
+      "loss": 0.8501,
+      "step": 20600
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5844169611043466,
+      "learning_rate": 1.095919173315923e-05,
+      "loss": 0.8245,
+      "step": 20601
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2948784383193301,
+      "learning_rate": 1.0958432207593453e-05,
+      "loss": 0.6798,
+      "step": 20602
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6630030342277395,
+      "learning_rate": 1.095767267644738e-05,
+      "loss": 0.721,
+      "step": 20603
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.351686522070656,
+      "learning_rate": 1.095691313972543e-05,
+      "loss": 0.8452,
+      "step": 20604
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.534374243163463,
+      "learning_rate": 1.0956153597432033e-05,
+      "loss": 0.5967,
+      "step": 20605
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0396844075287555,
+      "learning_rate": 1.0955394049571606e-05,
+      "loss": 0.5642,
+      "step": 20606
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.593902270219621,
+      "learning_rate": 1.0954634496148573e-05,
+      "loss": 0.4847,
+      "step": 20607
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4711029937527824,
+      "learning_rate": 1.0953874937167355e-05,
+      "loss": 0.7434,
+      "step": 20608
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.849490785279183,
+      "learning_rate": 1.0953115372632375e-05,
+      "loss": 0.7841,
+      "step": 20609
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.591477753314214,
+      "learning_rate": 1.0952355802548058e-05,
+      "loss": 0.5263,
+      "step": 20610
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.646368024522594,
+      "learning_rate": 1.095159622691882e-05,
+      "loss": 0.7506,
+      "step": 20611
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4880544023275653,
+      "learning_rate": 1.095083664574909e-05,
+      "loss": 0.7806,
+      "step": 20612
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3822045618260388,
+      "learning_rate": 1.0950077059043285e-05,
+      "loss": 0.7774,
+      "step": 20613
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3610401420417433,
+      "learning_rate": 1.0949317466805834e-05,
+      "loss": 0.7504,
+      "step": 20614
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4309643608104914,
+      "learning_rate": 1.0948557869041151e-05,
+      "loss": 0.7304,
+      "step": 20615
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4998554018052983,
+      "learning_rate": 1.094779826575367e-05,
+      "loss": 0.7351,
+      "step": 20616
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.492244877115079,
+      "learning_rate": 1.0947038656947806e-05,
+      "loss": 0.6689,
+      "step": 20617
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7511106039188513,
+      "learning_rate": 1.0946279042627981e-05,
+      "loss": 0.8149,
+      "step": 20618
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4076783174030132,
+      "learning_rate": 1.0945519422798622e-05,
+      "loss": 0.6759,
+      "step": 20619
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2563628727974179,
+      "learning_rate": 1.0944759797464146e-05,
+      "loss": 0.6869,
+      "step": 20620
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6147539433215065,
+      "learning_rate": 1.0944000166628986e-05,
+      "loss": 0.5447,
+      "step": 20621
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6375960789825215,
+      "learning_rate": 1.094324053029755e-05,
+      "loss": 0.8078,
+      "step": 20622
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.682967506667788,
+      "learning_rate": 1.0942480888474276e-05,
+      "loss": 0.7071,
+      "step": 20623
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.721092961261505,
+      "learning_rate": 1.0941721241163577e-05,
+      "loss": 0.7928,
+      "step": 20624
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7818629832871282,
+      "learning_rate": 1.0940961588369879e-05,
+      "loss": 0.7294,
+      "step": 20625
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4555716798018832,
+      "learning_rate": 1.0940201930097603e-05,
+      "loss": 0.8383,
+      "step": 20626
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5058514683574014,
+      "learning_rate": 1.0939442266351174e-05,
+      "loss": 0.7202,
+      "step": 20627
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6129264609914217,
+      "learning_rate": 1.0938682597135017e-05,
+      "loss": 0.5151,
+      "step": 20628
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5012663691169361,
+      "learning_rate": 1.0937922922453553e-05,
+      "loss": 0.7161,
+      "step": 20629
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3828698784352649,
+      "learning_rate": 1.0937163242311201e-05,
+      "loss": 0.7266,
+      "step": 20630
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.592424995284404,
+      "learning_rate": 1.0936403556712392e-05,
+      "loss": 0.5733,
+      "step": 20631
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0044036930219837,
+      "learning_rate": 1.0935643865661543e-05,
+      "loss": 0.757,
+      "step": 20632
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3547181153087173,
+      "learning_rate": 1.0934884169163078e-05,
+      "loss": 0.7399,
+      "step": 20633
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4607283153723527,
+      "learning_rate": 1.0934124467221424e-05,
+      "loss": 0.78,
+      "step": 20634
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6041765708384406,
+      "learning_rate": 1.0933364759841001e-05,
+      "loss": 0.5509,
+      "step": 20635
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5634820451192875,
+      "learning_rate": 1.0932605047026232e-05,
+      "loss": 0.497,
+      "step": 20636
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3087890855742288,
+      "learning_rate": 1.0931845328781541e-05,
+      "loss": 0.8906,
+      "step": 20637
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4565466079157325,
+      "learning_rate": 1.0931085605111354e-05,
+      "loss": 0.7137,
+      "step": 20638
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5479943304375547,
+      "learning_rate": 1.0930325876020086e-05,
+      "loss": 0.6935,
+      "step": 20639
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3695235246900752,
+      "learning_rate": 1.0929566141512173e-05,
+      "loss": 0.7446,
+      "step": 20640
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5442438015590056,
+      "learning_rate": 1.0928806401592026e-05,
+      "loss": 0.7692,
+      "step": 20641
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8439986793004377,
+      "learning_rate": 1.092804665626408e-05,
+      "loss": 0.8735,
+      "step": 20642
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4104813907430507,
+      "learning_rate": 1.0927286905532749e-05,
+      "loss": 0.6989,
+      "step": 20643
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1551920775573856,
+      "learning_rate": 1.0926527149402461e-05,
+      "loss": 0.754,
+      "step": 20644
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.847902193205916,
+      "learning_rate": 1.0925767387877638e-05,
+      "loss": 0.7976,
+      "step": 20645
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.560370020384777,
+      "learning_rate": 1.0925007620962703e-05,
+      "loss": 0.6439,
+      "step": 20646
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1571527917078703,
+      "learning_rate": 1.0924247848662084e-05,
+      "loss": 0.6833,
+      "step": 20647
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3819513053932,
+      "learning_rate": 1.0923488070980197e-05,
+      "loss": 0.7658,
+      "step": 20648
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8401428414910088,
+      "learning_rate": 1.0922728287921474e-05,
+      "loss": 0.8044,
+      "step": 20649
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5801683781880023,
+      "learning_rate": 1.0921968499490332e-05,
+      "loss": 0.4903,
+      "step": 20650
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8923122773323795,
+      "learning_rate": 1.09212087056912e-05,
+      "loss": 0.6815,
+      "step": 20651
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3749824878702823,
+      "learning_rate": 1.0920448906528498e-05,
+      "loss": 0.7839,
+      "step": 20652
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.429355103783993,
+      "learning_rate": 1.091968910200665e-05,
+      "loss": 0.8465,
+      "step": 20653
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3358536360071038,
+      "learning_rate": 1.0918929292130082e-05,
+      "loss": 0.6705,
+      "step": 20654
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.246126300715844,
+      "learning_rate": 1.0918169476903217e-05,
+      "loss": 0.7115,
+      "step": 20655
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4832790654099577,
+      "learning_rate": 1.0917409656330477e-05,
+      "loss": 0.7199,
+      "step": 20656
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4461493058562878,
+      "learning_rate": 1.091664983041629e-05,
+      "loss": 0.7538,
+      "step": 20657
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.72199448151399,
+      "learning_rate": 1.0915889999165077e-05,
+      "loss": 0.7329,
+      "step": 20658
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.809960773445074,
+      "learning_rate": 1.091513016258126e-05,
+      "loss": 0.6755,
+      "step": 20659
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4722717492036013,
+      "learning_rate": 1.0914370320669267e-05,
+      "loss": 0.8381,
+      "step": 20660
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4955635851686329,
+      "learning_rate": 1.0913610473433525e-05,
+      "loss": 0.6899,
+      "step": 20661
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.524493271034426,
+      "learning_rate": 1.0912850620878445e-05,
+      "loss": 0.7818,
+      "step": 20662
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6394309844222261,
+      "learning_rate": 1.0912090763008468e-05,
+      "loss": 0.7036,
+      "step": 20663
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3234912955762292,
+      "learning_rate": 1.0911330899828003e-05,
+      "loss": 0.6674,
+      "step": 20664
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5291192254080401,
+      "learning_rate": 1.0910571031341485e-05,
+      "loss": 0.7392,
+      "step": 20665
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.046797977107513,
+      "learning_rate": 1.0909811157553331e-05,
+      "loss": 0.6008,
+      "step": 20666
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7925893544260003,
+      "learning_rate": 1.090905127846797e-05,
+      "loss": 0.7854,
+      "step": 20667
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6676830443158177,
+      "learning_rate": 1.0908291394089828e-05,
+      "loss": 0.7277,
+      "step": 20668
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6775831204069884,
+      "learning_rate": 1.0907531504423321e-05,
+      "loss": 0.7868,
+      "step": 20669
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3446285547487797,
+      "learning_rate": 1.0906771609472882e-05,
+      "loss": 0.6992,
+      "step": 20670
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4773800562398276,
+      "learning_rate": 1.0906011709242927e-05,
+      "loss": 0.7875,
+      "step": 20671
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1934918601682514,
+      "learning_rate": 1.0905251803737891e-05,
+      "loss": 0.6221,
+      "step": 20672
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0055682711187544,
+      "learning_rate": 1.0904491892962186e-05,
+      "loss": 0.7676,
+      "step": 20673
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.594679625073024,
+      "learning_rate": 1.0903731976920249e-05,
+      "loss": 0.4927,
+      "step": 20674
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5555808484010367,
+      "learning_rate": 1.0902972055616494e-05,
+      "loss": 0.4904,
+      "step": 20675
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.465005154638278,
+      "learning_rate": 1.0902212129055353e-05,
+      "loss": 0.6746,
+      "step": 20676
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3633815405191536,
+      "learning_rate": 1.0901452197241245e-05,
+      "loss": 0.7273,
+      "step": 20677
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.645312839515021,
+      "learning_rate": 1.0900692260178596e-05,
+      "loss": 0.5402,
+      "step": 20678
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4655611304172638,
+      "learning_rate": 1.0899932317871832e-05,
+      "loss": 0.7193,
+      "step": 20679
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6228017553771572,
+      "learning_rate": 1.0899172370325379e-05,
+      "loss": 0.7338,
+      "step": 20680
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7411383211313207,
+      "learning_rate": 1.0898412417543658e-05,
+      "loss": 0.8076,
+      "step": 20681
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.785783929623089,
+      "learning_rate": 1.0897652459531098e-05,
+      "loss": 0.7419,
+      "step": 20682
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5846779452262765,
+      "learning_rate": 1.089689249629212e-05,
+      "loss": 0.7526,
+      "step": 20683
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5289765439221394,
+      "learning_rate": 1.0896132527831149e-05,
+      "loss": 0.7035,
+      "step": 20684
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3940945352505247,
+      "learning_rate": 1.0895372554152611e-05,
+      "loss": 0.8269,
+      "step": 20685
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4703359997779344,
+      "learning_rate": 1.089461257526093e-05,
+      "loss": 0.7854,
+      "step": 20686
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.628281708815025,
+      "learning_rate": 1.0893852591160533e-05,
+      "loss": 0.7854,
+      "step": 20687
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3388811404790113,
+      "learning_rate": 1.0893092601855841e-05,
+      "loss": 0.7197,
+      "step": 20688
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5446795703420464,
+      "learning_rate": 1.0892332607351283e-05,
+      "loss": 0.687,
+      "step": 20689
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4416641413963935,
+      "learning_rate": 1.0891572607651281e-05,
+      "loss": 0.6792,
+      "step": 20690
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3581370256041678,
+      "learning_rate": 1.0890812602760263e-05,
+      "loss": 0.7189,
+      "step": 20691
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4728215579097796,
+      "learning_rate": 1.089005259268265e-05,
+      "loss": 0.5554,
+      "step": 20692
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.468083377313463,
+      "learning_rate": 1.0889292577422869e-05,
+      "loss": 0.7482,
+      "step": 20693
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3954766894145723,
+      "learning_rate": 1.0888532556985346e-05,
+      "loss": 0.891,
+      "step": 20694
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6309292310284647,
+      "learning_rate": 1.0887772531374506e-05,
+      "loss": 0.5378,
+      "step": 20695
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.410966808058939,
+      "learning_rate": 1.0887012500594771e-05,
+      "loss": 0.8697,
+      "step": 20696
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5778372557107583,
+      "learning_rate": 1.0886252464650569e-05,
+      "loss": 0.6657,
+      "step": 20697
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4772533655645674,
+      "learning_rate": 1.0885492423546324e-05,
+      "loss": 0.6989,
+      "step": 20698
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4410411026149736,
+      "learning_rate": 1.0884732377286464e-05,
+      "loss": 0.8161,
+      "step": 20699
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6002228533184428,
+      "learning_rate": 1.0883972325875411e-05,
+      "loss": 0.8883,
+      "step": 20700
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4408905097232518,
+      "learning_rate": 1.088321226931759e-05,
+      "loss": 0.7776,
+      "step": 20701
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4782656556592577,
+      "learning_rate": 1.0882452207617428e-05,
+      "loss": 0.6312,
+      "step": 20702
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4763736071156388,
+      "learning_rate": 1.0881692140779352e-05,
+      "loss": 0.8115,
+      "step": 20703
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6278015996326862,
+      "learning_rate": 1.0880932068807783e-05,
+      "loss": 0.7424,
+      "step": 20704
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.614126457050048,
+      "learning_rate": 1.0880171991707149e-05,
+      "loss": 0.5743,
+      "step": 20705
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6204991427993891,
+      "learning_rate": 1.0879411909481874e-05,
+      "loss": 0.5262,
+      "step": 20706
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3916827936473362,
+      "learning_rate": 1.0878651822136386e-05,
+      "loss": 0.8281,
+      "step": 20707
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4760113036566835,
+      "learning_rate": 1.087789172967511e-05,
+      "loss": 0.7605,
+      "step": 20708
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4156497409037634,
+      "learning_rate": 1.0877131632102468e-05,
+      "loss": 0.7334,
+      "step": 20709
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.354531119506233,
+      "learning_rate": 1.087637152942289e-05,
+      "loss": 0.7523,
+      "step": 20710
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5427086936548056,
+      "learning_rate": 1.0875611421640797e-05,
+      "loss": 0.7137,
+      "step": 20711
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.445726884588717,
+      "learning_rate": 1.0874851308760622e-05,
+      "loss": 0.6284,
+      "step": 20712
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2246817419374834,
+      "learning_rate": 1.087409119078678e-05,
+      "loss": 0.7629,
+      "step": 20713
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6007406642247104,
+      "learning_rate": 1.0873331067723704e-05,
+      "loss": 0.8109,
+      "step": 20714
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.367102184539446,
+      "learning_rate": 1.0872570939575818e-05,
+      "loss": 0.8448,
+      "step": 20715
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.2314186834117464,
+      "learning_rate": 1.087181080634755e-05,
+      "loss": 0.6107,
+      "step": 20716
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5147512132688667,
+      "learning_rate": 1.087105066804332e-05,
+      "loss": 0.7605,
+      "step": 20717
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3159257043043178,
+      "learning_rate": 1.087029052466756e-05,
+      "loss": 0.6608,
+      "step": 20718
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.98394128918383,
+      "learning_rate": 1.0869530376224693e-05,
+      "loss": 0.75,
+      "step": 20719
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.303685470396606,
+      "learning_rate": 1.0868770222719143e-05,
+      "loss": 0.6801,
+      "step": 20720
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.369859798443731,
+      "learning_rate": 1.0868010064155338e-05,
+      "loss": 0.7265,
+      "step": 20721
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3703686695116308,
+      "learning_rate": 1.0867249900537703e-05,
+      "loss": 0.7821,
+      "step": 20722
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7456470212959243,
+      "learning_rate": 1.0866489731870668e-05,
+      "loss": 0.6624,
+      "step": 20723
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5719333059062925,
+      "learning_rate": 1.0865729558158652e-05,
+      "loss": 0.8232,
+      "step": 20724
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.54700536167122,
+      "learning_rate": 1.0864969379406088e-05,
+      "loss": 0.8351,
+      "step": 20725
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9396871362524304,
+      "learning_rate": 1.0864209195617395e-05,
+      "loss": 0.6278,
+      "step": 20726
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4818944609674076,
+      "learning_rate": 1.0863449006797005e-05,
+      "loss": 0.6338,
+      "step": 20727
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.511269935156507,
+      "learning_rate": 1.0862688812949338e-05,
+      "loss": 0.6697,
+      "step": 20728
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.423702889428459,
+      "learning_rate": 1.0861928614078827e-05,
+      "loss": 0.7872,
+      "step": 20729
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5762366736294187,
+      "learning_rate": 1.0861168410189894e-05,
+      "loss": 0.7916,
+      "step": 20730
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6092158709160902,
+      "learning_rate": 1.0860408201286964e-05,
+      "loss": 0.5643,
+      "step": 20731
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6501409278352925,
+      "learning_rate": 1.0859647987374467e-05,
+      "loss": 0.6621,
+      "step": 20732
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5182537804979668,
+      "learning_rate": 1.0858887768456824e-05,
+      "loss": 0.7859,
+      "step": 20733
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3358756782920531,
+      "learning_rate": 1.0858127544538468e-05,
+      "loss": 0.6604,
+      "step": 20734
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6161371944906279,
+      "learning_rate": 1.085736731562382e-05,
+      "loss": 0.55,
+      "step": 20735
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.466018541662961,
+      "learning_rate": 1.085660708171731e-05,
+      "loss": 0.7843,
+      "step": 20736
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5890949364961363,
+      "learning_rate": 1.085584684282336e-05,
+      "loss": 0.5271,
+      "step": 20737
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7285672883974814,
+      "learning_rate": 1.0855086598946401e-05,
+      "loss": 0.8169,
+      "step": 20738
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3188708291119693,
+      "learning_rate": 1.0854326350090855e-05,
+      "loss": 0.7588,
+      "step": 20739
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0063434175124875,
+      "learning_rate": 1.0853566096261155e-05,
+      "loss": 0.6637,
+      "step": 20740
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.480485661659802,
+      "learning_rate": 1.0852805837461718e-05,
+      "loss": 0.7356,
+      "step": 20741
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6177444301356889,
+      "learning_rate": 1.0852045573696979e-05,
+      "loss": 0.5398,
+      "step": 20742
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4054775476159636,
+      "learning_rate": 1.0851285304971354e-05,
+      "loss": 0.6645,
+      "step": 20743
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5092083701029348,
+      "learning_rate": 1.0850525031289286e-05,
+      "loss": 0.8184,
+      "step": 20744
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4132064383719571,
+      "learning_rate": 1.0849764752655185e-05,
+      "loss": 0.5169,
+      "step": 20745
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4341274789192895,
+      "learning_rate": 1.0849004469073487e-05,
+      "loss": 0.7647,
+      "step": 20746
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6031606230334073,
+      "learning_rate": 1.0848244180548618e-05,
+      "loss": 0.4907,
+      "step": 20747
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4298890453736737,
+      "learning_rate": 1.0847483887085e-05,
+      "loss": 0.6714,
+      "step": 20748
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.364224457464874,
+      "learning_rate": 1.0846723588687065e-05,
+      "loss": 0.6496,
+      "step": 20749
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.949906333188946,
+      "learning_rate": 1.0845963285359233e-05,
+      "loss": 0.7195,
+      "step": 20750
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5807298612991587,
+      "learning_rate": 1.084520297710594e-05,
+      "loss": 0.7582,
+      "step": 20751
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.2267286348999304,
+      "learning_rate": 1.0844442663931604e-05,
+      "loss": 0.7691,
+      "step": 20752
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3969532412718535,
+      "learning_rate": 1.0843682345840655e-05,
+      "loss": 0.6738,
+      "step": 20753
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.561060762657748,
+      "learning_rate": 1.0842922022837524e-05,
+      "loss": 0.6903,
+      "step": 20754
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6061885959926823,
+      "learning_rate": 1.0842161694926632e-05,
+      "loss": 0.5246,
+      "step": 20755
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2323471216599515,
+      "learning_rate": 1.0841401362112408e-05,
+      "loss": 0.7105,
+      "step": 20756
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.28979797684138,
+      "learning_rate": 1.0840641024399278e-05,
+      "loss": 0.7553,
+      "step": 20757
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6117343842850368,
+      "learning_rate": 1.083988068179167e-05,
+      "loss": 0.5673,
+      "step": 20758
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6611872530923086,
+      "learning_rate": 1.0839120334294013e-05,
+      "loss": 0.7139,
+      "step": 20759
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.37705343338488,
+      "learning_rate": 1.083835998191073e-05,
+      "loss": 0.7062,
+      "step": 20760
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6229732092771718,
+      "learning_rate": 1.0837599624646251e-05,
+      "loss": 0.5295,
+      "step": 20761
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4287694068304542,
+      "learning_rate": 1.0836839262504998e-05,
+      "loss": 0.7122,
+      "step": 20762
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4931819142182925,
+      "learning_rate": 1.0836078895491408e-05,
+      "loss": 0.7815,
+      "step": 20763
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.579982913657766,
+      "learning_rate": 1.0835318523609897e-05,
+      "loss": 0.6921,
+      "step": 20764
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6082661565259146,
+      "learning_rate": 1.08345581468649e-05,
+      "loss": 0.4896,
+      "step": 20765
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.344004824149384,
+      "learning_rate": 1.0833797765260838e-05,
+      "loss": 0.7414,
+      "step": 20766
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.345539847299466,
+      "learning_rate": 1.0833037378802145e-05,
+      "loss": 0.8568,
+      "step": 20767
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5510285733765392,
+      "learning_rate": 1.0832276987493243e-05,
+      "loss": 0.7837,
+      "step": 20768
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4208379521740238,
+      "learning_rate": 1.083151659133856e-05,
+      "loss": 0.6714,
+      "step": 20769
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4647260579970405,
+      "learning_rate": 1.0830756190342528e-05,
+      "loss": 0.6727,
+      "step": 20770
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.70292973136939,
+      "learning_rate": 1.0829995784509565e-05,
+      "loss": 0.7898,
+      "step": 20771
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2405390636741085,
+      "learning_rate": 1.0829235373844109e-05,
+      "loss": 0.6788,
+      "step": 20772
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.377857851828894,
+      "learning_rate": 1.0828474958350577e-05,
+      "loss": 0.7108,
+      "step": 20773
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.729464321203775,
+      "learning_rate": 1.0827714538033407e-05,
+      "loss": 0.7813,
+      "step": 20774
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7256406687994872,
+      "learning_rate": 1.0826954112897014e-05,
+      "loss": 0.7654,
+      "step": 20775
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.242999092211392,
+      "learning_rate": 1.0826193682945839e-05,
+      "loss": 0.6317,
+      "step": 20776
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5958884634378363,
+      "learning_rate": 1.0825433248184301e-05,
+      "loss": 0.5529,
+      "step": 20777
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3848855817811456,
+      "learning_rate": 1.0824672808616827e-05,
+      "loss": 0.7943,
+      "step": 20778
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.483694860383595,
+      "learning_rate": 1.0823912364247848e-05,
+      "loss": 0.7745,
+      "step": 20779
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4946319140422226,
+      "learning_rate": 1.0823151915081791e-05,
+      "loss": 0.7324,
+      "step": 20780
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5749685722322935,
+      "learning_rate": 1.0822391461123083e-05,
+      "loss": 0.5211,
+      "step": 20781
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5379761562640628,
+      "learning_rate": 1.0821631002376152e-05,
+      "loss": 0.7763,
+      "step": 20782
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2843927605191117,
+      "learning_rate": 1.0820870538845423e-05,
+      "loss": 0.7277,
+      "step": 20783
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.94380994513142,
+      "learning_rate": 1.0820110070535326e-05,
+      "loss": 0.5983,
+      "step": 20784
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.353463472511362,
+      "learning_rate": 1.081934959745029e-05,
+      "loss": 0.5965,
+      "step": 20785
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5355249831136248,
+      "learning_rate": 1.081858911959474e-05,
+      "loss": 0.7526,
+      "step": 20786
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4266898527028335,
+      "learning_rate": 1.0817828636973106e-05,
+      "loss": 0.6545,
+      "step": 20787
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.212560338839292,
+      "learning_rate": 1.0817068149589816e-05,
+      "loss": 0.6855,
+      "step": 20788
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5947994340376848,
+      "learning_rate": 1.0816307657449294e-05,
+      "loss": 0.7337,
+      "step": 20789
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3402000880939653,
+      "learning_rate": 1.0815547160555973e-05,
+      "loss": 0.7707,
+      "step": 20790
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7038171715077575,
+      "learning_rate": 1.0814786658914279e-05,
+      "loss": 0.6949,
+      "step": 20791
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.642932729301337,
+      "learning_rate": 1.0814026152528634e-05,
+      "loss": 0.6956,
+      "step": 20792
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6358985922717029,
+      "learning_rate": 1.0813265641403476e-05,
+      "loss": 0.5933,
+      "step": 20793
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4566946760105461,
+      "learning_rate": 1.0812505125543223e-05,
+      "loss": 0.9361,
+      "step": 20794
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5585584332689761,
+      "learning_rate": 1.0811744604952312e-05,
+      "loss": 0.6859,
+      "step": 20795
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.397499098280064,
+      "learning_rate": 1.0810984079635164e-05,
+      "loss": 0.7035,
+      "step": 20796
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6474570111220215,
+      "learning_rate": 1.0810223549596215e-05,
+      "loss": 0.5463,
+      "step": 20797
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7795205170860828,
+      "learning_rate": 1.0809463014839884e-05,
+      "loss": 0.7804,
+      "step": 20798
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8883422571143667,
+      "learning_rate": 1.0808702475370603e-05,
+      "loss": 0.7168,
+      "step": 20799
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4046146460386022,
+      "learning_rate": 1.0807941931192802e-05,
+      "loss": 0.7052,
+      "step": 20800
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.187346619365161,
+      "learning_rate": 1.0807181382310904e-05,
+      "loss": 0.6538,
+      "step": 20801
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4242261948078847,
+      "learning_rate": 1.0806420828729345e-05,
+      "loss": 0.698,
+      "step": 20802
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4834041886503828,
+      "learning_rate": 1.0805660270452542e-05,
+      "loss": 0.7193,
+      "step": 20803
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.3627132836208116,
+      "learning_rate": 1.0804899707484938e-05,
+      "loss": 0.657,
+      "step": 20804
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5763133705833337,
+      "learning_rate": 1.0804139139830947e-05,
+      "loss": 0.794,
+      "step": 20805
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5650419770710194,
+      "learning_rate": 1.0803378567495006e-05,
+      "loss": 0.5169,
+      "step": 20806
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3941610438911958,
+      "learning_rate": 1.0802617990481542e-05,
+      "loss": 0.7794,
+      "step": 20807
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4826220092543017,
+      "learning_rate": 1.0801857408794978e-05,
+      "loss": 0.7238,
+      "step": 20808
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4168768092630326,
+      "learning_rate": 1.080109682243975e-05,
+      "loss": 0.6717,
+      "step": 20809
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.479870431869665,
+      "learning_rate": 1.0800336231420278e-05,
+      "loss": 0.693,
+      "step": 20810
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3685685567032617,
+      "learning_rate": 1.0799575635741e-05,
+      "loss": 0.68,
+      "step": 20811
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3877702769378064,
+      "learning_rate": 1.079881503540634e-05,
+      "loss": 0.8322,
+      "step": 20812
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2956893497385602,
+      "learning_rate": 1.079805443042072e-05,
+      "loss": 0.6443,
+      "step": 20813
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6749454892797282,
+      "learning_rate": 1.0797293820788578e-05,
+      "loss": 0.5965,
+      "step": 20814
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.350395826569398,
+      "learning_rate": 1.079653320651434e-05,
+      "loss": 0.7393,
+      "step": 20815
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.398474449688177,
+      "learning_rate": 1.0795772587602432e-05,
+      "loss": 0.7266,
+      "step": 20816
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5751384843295735,
+      "learning_rate": 1.0795011964057285e-05,
+      "loss": 0.5513,
+      "step": 20817
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2571989743689636,
+      "learning_rate": 1.0794251335883326e-05,
+      "loss": 0.7726,
+      "step": 20818
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5781010733302272,
+      "learning_rate": 1.0793490703084985e-05,
+      "loss": 0.7456,
+      "step": 20819
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9411089915616178,
+      "learning_rate": 1.0792730065666688e-05,
+      "loss": 0.8157,
+      "step": 20820
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.459448230195778,
+      "learning_rate": 1.0791969423632869e-05,
+      "loss": 0.6183,
+      "step": 20821
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4383253664823135,
+      "learning_rate": 1.0791208776987949e-05,
+      "loss": 0.7222,
+      "step": 20822
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.247022989469907,
+      "learning_rate": 1.0790448125736366e-05,
+      "loss": 0.6508,
+      "step": 20823
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5839862609787672,
+      "learning_rate": 1.0789687469882538e-05,
+      "loss": 0.768,
+      "step": 20824
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3991229383463357,
+      "learning_rate": 1.0788926809430904e-05,
+      "loss": 0.6844,
+      "step": 20825
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4649037826674205,
+      "learning_rate": 1.0788166144385888e-05,
+      "loss": 0.6339,
+      "step": 20826
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6252057339984751,
+      "learning_rate": 1.0787405474751917e-05,
+      "loss": 0.6703,
+      "step": 20827
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4503724965413416,
+      "learning_rate": 1.0786644800533423e-05,
+      "loss": 0.7053,
+      "step": 20828
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6218639774581551,
+      "learning_rate": 1.0785884121734836e-05,
+      "loss": 0.5123,
+      "step": 20829
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5950755473480263,
+      "learning_rate": 1.0785123438360578e-05,
+      "loss": 0.5772,
+      "step": 20830
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4123312996812578,
+      "learning_rate": 1.0784362750415088e-05,
+      "loss": 0.839,
+      "step": 20831
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3143634462877958,
+      "learning_rate": 1.0783602057902786e-05,
+      "loss": 0.7954,
+      "step": 20832
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.818372748734988,
+      "learning_rate": 1.0782841360828107e-05,
+      "loss": 0.7738,
+      "step": 20833
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.00395731385693,
+      "learning_rate": 1.0782080659195475e-05,
+      "loss": 0.607,
+      "step": 20834
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6857109175615619,
+      "learning_rate": 1.0781319953009324e-05,
+      "loss": 0.8479,
+      "step": 20835
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4116158812719253,
+      "learning_rate": 1.078055924227408e-05,
+      "loss": 0.7805,
+      "step": 20836
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.599024218671571,
+      "learning_rate": 1.0779798526994173e-05,
+      "loss": 0.7287,
+      "step": 20837
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4132121728308675,
+      "learning_rate": 1.0779037807174032e-05,
+      "loss": 0.7304,
+      "step": 20838
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4947232053785184,
+      "learning_rate": 1.077827708281809e-05,
+      "loss": 0.5439,
+      "step": 20839
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.765740340675002,
+      "learning_rate": 1.0777516353930767e-05,
+      "loss": 0.629,
+      "step": 20840
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.519709193184023,
+      "learning_rate": 1.0776755620516501e-05,
+      "loss": 0.7183,
+      "step": 20841
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5016302956257934,
+      "learning_rate": 1.0775994882579718e-05,
+      "loss": 0.6992,
+      "step": 20842
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5955046740785148,
+      "learning_rate": 1.0775234140124845e-05,
+      "loss": 0.7314,
+      "step": 20843
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5225873460763664,
+      "learning_rate": 1.0774473393156315e-05,
+      "loss": 0.6473,
+      "step": 20844
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.53316461948885,
+      "learning_rate": 1.0773712641678553e-05,
+      "loss": 0.7122,
+      "step": 20845
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.436190719655719,
+      "learning_rate": 1.0772951885695994e-05,
+      "loss": 0.7816,
+      "step": 20846
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5583465877345071,
+      "learning_rate": 1.0772191125213063e-05,
+      "loss": 0.5307,
+      "step": 20847
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5025738423018047,
+      "learning_rate": 1.0771430360234194e-05,
+      "loss": 0.7384,
+      "step": 20848
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.52840512557572,
+      "learning_rate": 1.0770669590763809e-05,
+      "loss": 0.8344,
+      "step": 20849
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4201054535451547,
+      "learning_rate": 1.0769908816806343e-05,
+      "loss": 0.8538,
+      "step": 20850
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3905620628517725,
+      "learning_rate": 1.0769148038366223e-05,
+      "loss": 0.7324,
+      "step": 20851
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5878987148887117,
+      "learning_rate": 1.076838725544788e-05,
+      "loss": 0.5204,
+      "step": 20852
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4834326885762394,
+      "learning_rate": 1.0767626468055747e-05,
+      "loss": 0.797,
+      "step": 20853
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.6560458877425805,
+      "learning_rate": 1.0766865676194245e-05,
+      "loss": 0.5819,
+      "step": 20854
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8758079978395543,
+      "learning_rate": 1.0766104879867811e-05,
+      "loss": 0.4942,
+      "step": 20855
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4197169696075846,
+      "learning_rate": 1.076534407908087e-05,
+      "loss": 0.8018,
+      "step": 20856
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4203036066972798,
+      "learning_rate": 1.0764583273837855e-05,
+      "loss": 0.7841,
+      "step": 20857
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3719987470168504,
+      "learning_rate": 1.0763822464143191e-05,
+      "loss": 0.7443,
+      "step": 20858
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.364701637045551,
+      "learning_rate": 1.0763061650001314e-05,
+      "loss": 0.6793,
+      "step": 20859
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3906528612298275,
+      "learning_rate": 1.076230083141665e-05,
+      "loss": 0.6999,
+      "step": 20860
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7498160917883756,
+      "learning_rate": 1.076154000839363e-05,
+      "loss": 0.6423,
+      "step": 20861
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6041076470565867,
+      "learning_rate": 1.076077918093668e-05,
+      "loss": 0.7313,
+      "step": 20862
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.684069488936098,
+      "learning_rate": 1.0760018349050236e-05,
+      "loss": 0.6764,
+      "step": 20863
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7695054882047427,
+      "learning_rate": 1.0759257512738723e-05,
+      "loss": 0.718,
+      "step": 20864
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.42241597720478,
+      "learning_rate": 1.0758496672006571e-05,
+      "loss": 0.6151,
+      "step": 20865
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4814553829985873,
+      "learning_rate": 1.0757735826858213e-05,
+      "loss": 0.6973,
+      "step": 20866
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4808596409068937,
+      "learning_rate": 1.0756974977298078e-05,
+      "loss": 0.6729,
+      "step": 20867
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6004787083788816,
+      "learning_rate": 1.0756214123330593e-05,
+      "loss": 0.5331,
+      "step": 20868
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5839000376026942,
+      "learning_rate": 1.0755453264960192e-05,
+      "loss": 0.8858,
+      "step": 20869
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3749569075583754,
+      "learning_rate": 1.0754692402191303e-05,
+      "loss": 0.5547,
+      "step": 20870
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4234720411270414,
+      "learning_rate": 1.0753931535028353e-05,
+      "loss": 0.8648,
+      "step": 20871
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3643728553443892,
+      "learning_rate": 1.0753170663475778e-05,
+      "loss": 0.6917,
+      "step": 20872
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8837043664126871,
+      "learning_rate": 1.0752409787538e-05,
+      "loss": 0.7783,
+      "step": 20873
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3405140608657427,
+      "learning_rate": 1.0751648907219461e-05,
+      "loss": 0.7606,
+      "step": 20874
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.734238895885553,
+      "learning_rate": 1.075088802252458e-05,
+      "loss": 0.6701,
+      "step": 20875
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3692544443299006,
+      "learning_rate": 1.0750127133457792e-05,
+      "loss": 0.7189,
+      "step": 20876
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5646868995626537,
+      "learning_rate": 1.0749366240023526e-05,
+      "loss": 0.6862,
+      "step": 20877
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5407288622332957,
+      "learning_rate": 1.0748605342226213e-05,
+      "loss": 0.7606,
+      "step": 20878
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3913024653490023,
+      "learning_rate": 1.0747844440070284e-05,
+      "loss": 0.8429,
+      "step": 20879
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9660399196024394,
+      "learning_rate": 1.0747083533560166e-05,
+      "loss": 0.8003,
+      "step": 20880
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4546357053697527,
+      "learning_rate": 1.0746322622700289e-05,
+      "loss": 0.7527,
+      "step": 20881
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6353975717740052,
+      "learning_rate": 1.0745561707495088e-05,
+      "loss": 0.6485,
+      "step": 20882
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.64430256229802,
+      "learning_rate": 1.074480078794899e-05,
+      "loss": 0.5564,
+      "step": 20883
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3753763840113669,
+      "learning_rate": 1.0744039864066427e-05,
+      "loss": 0.7145,
+      "step": 20884
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.547139323169734,
+      "learning_rate": 1.0743278935851828e-05,
+      "loss": 0.783,
+      "step": 20885
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5878233144084726,
+      "learning_rate": 1.0742518003309624e-05,
+      "loss": 0.5078,
+      "step": 20886
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.6992126294435326,
+      "learning_rate": 1.0741757066444244e-05,
+      "loss": 0.7382,
+      "step": 20887
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7159469322050793,
+      "learning_rate": 1.0740996125260122e-05,
+      "loss": 0.686,
+      "step": 20888
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.658093914277211,
+      "learning_rate": 1.0740235179761682e-05,
+      "loss": 0.7961,
+      "step": 20889
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.555532266379786,
+      "learning_rate": 1.0739474229953362e-05,
+      "loss": 0.6729,
+      "step": 20890
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.491354705189494,
+      "learning_rate": 1.073871327583959e-05,
+      "loss": 0.7002,
+      "step": 20891
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.621885072874121,
+      "learning_rate": 1.0737952317424789e-05,
+      "loss": 0.7237,
+      "step": 20892
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3182228139951095,
+      "learning_rate": 1.0737191354713403e-05,
+      "loss": 0.6676,
+      "step": 20893
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3374551522377043,
+      "learning_rate": 1.0736430387709851e-05,
+      "loss": 0.7277,
+      "step": 20894
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4347151941852734,
+      "learning_rate": 1.0735669416418571e-05,
+      "loss": 0.6659,
+      "step": 20895
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.41364205168628,
+      "learning_rate": 1.0734908440843987e-05,
+      "loss": 0.7362,
+      "step": 20896
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5068740483602245,
+      "learning_rate": 1.0734147460990539e-05,
+      "loss": 0.6976,
+      "step": 20897
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2531306790595766,
+      "learning_rate": 1.0733386476862649e-05,
+      "loss": 0.7189,
+      "step": 20898
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8557211580711506,
+      "learning_rate": 1.0732625488464749e-05,
+      "loss": 0.6256,
+      "step": 20899
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6024645950998144,
+      "learning_rate": 1.0731864495801272e-05,
+      "loss": 0.826,
+      "step": 20900
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.3204155761042298,
+      "learning_rate": 1.0731103498876647e-05,
+      "loss": 0.6514,
+      "step": 20901
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.451653679718752,
+      "learning_rate": 1.0730342497695312e-05,
+      "loss": 0.7641,
+      "step": 20902
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5128751223311216,
+      "learning_rate": 1.0729581492261684e-05,
+      "loss": 0.734,
+      "step": 20903
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4322446931330353,
+      "learning_rate": 1.072882048258021e-05,
+      "loss": 0.689,
+      "step": 20904
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5131873274695857,
+      "learning_rate": 1.0728059468655303e-05,
+      "loss": 0.6548,
+      "step": 20905
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3333085920296746,
+      "learning_rate": 1.0727298450491413e-05,
+      "loss": 0.7778,
+      "step": 20906
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3549472867494057,
+      "learning_rate": 1.0726537428092956e-05,
+      "loss": 0.7397,
+      "step": 20907
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7214730043898088,
+      "learning_rate": 1.0725776401464366e-05,
+      "loss": 0.6062,
+      "step": 20908
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3846810011999002,
+      "learning_rate": 1.0725015370610078e-05,
+      "loss": 0.75,
+      "step": 20909
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2996580984705528,
+      "learning_rate": 1.0724254335534524e-05,
+      "loss": 0.6986,
+      "step": 20910
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4185827958659851,
+      "learning_rate": 1.0723493296242129e-05,
+      "loss": 0.7787,
+      "step": 20911
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4707078888222778,
+      "learning_rate": 1.0722732252737327e-05,
+      "loss": 0.7504,
+      "step": 20912
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.565461721150353,
+      "learning_rate": 1.072197120502455e-05,
+      "loss": 0.8551,
+      "step": 20913
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.427879831691536,
+      "learning_rate": 1.0721210153108226e-05,
+      "loss": 0.7539,
+      "step": 20914
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.414814257571921,
+      "learning_rate": 1.072044909699279e-05,
+      "loss": 0.7723,
+      "step": 20915
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.614804664047182,
+      "learning_rate": 1.0719688036682672e-05,
+      "loss": 0.7585,
+      "step": 20916
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5636242325116876,
+      "learning_rate": 1.07189269721823e-05,
+      "loss": 0.7287,
+      "step": 20917
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.631336111372836,
+      "learning_rate": 1.0718165903496109e-05,
+      "loss": 0.6637,
+      "step": 20918
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.262396110425457,
+      "learning_rate": 1.0717404830628529e-05,
+      "loss": 0.6867,
+      "step": 20919
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5544809993975623,
+      "learning_rate": 1.071664375358399e-05,
+      "loss": 0.7786,
+      "step": 20920
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4006561206833195,
+      "learning_rate": 1.0715882672366926e-05,
+      "loss": 0.8362,
+      "step": 20921
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.681350671449532,
+      "learning_rate": 1.0715121586981763e-05,
+      "loss": 0.889,
+      "step": 20922
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2237352113260462,
+      "learning_rate": 1.071436049743294e-05,
+      "loss": 0.708,
+      "step": 20923
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4513844529129725,
+      "learning_rate": 1.0713599403724879e-05,
+      "loss": 0.6551,
+      "step": 20924
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8551732756371533,
+      "learning_rate": 1.0712838305862023e-05,
+      "loss": 0.735,
+      "step": 20925
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.621916726370615,
+      "learning_rate": 1.071207720384879e-05,
+      "loss": 0.6879,
+      "step": 20926
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2938123724006025,
+      "learning_rate": 1.0711316097689623e-05,
+      "loss": 0.7087,
+      "step": 20927
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5631069723450035,
+      "learning_rate": 1.0710554987388946e-05,
+      "loss": 0.5683,
+      "step": 20928
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6373292802964949,
+      "learning_rate": 1.0709793872951193e-05,
+      "loss": 0.5718,
+      "step": 20929
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.34988720176847,
+      "learning_rate": 1.0709032754380797e-05,
+      "loss": 0.7459,
+      "step": 20930
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.692633768333833,
+      "learning_rate": 1.0708271631682186e-05,
+      "loss": 0.6346,
+      "step": 20931
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.80707193812827,
+      "learning_rate": 1.0707510504859795e-05,
+      "loss": 0.8132,
+      "step": 20932
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.2154976995813365,
+      "learning_rate": 1.070674937391805e-05,
+      "loss": 0.6739,
+      "step": 20933
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5424505674382826,
+      "learning_rate": 1.0705988238861391e-05,
+      "loss": 0.4974,
+      "step": 20934
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.204325489235002,
+      "learning_rate": 1.0705227099694242e-05,
+      "loss": 0.6868,
+      "step": 20935
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3977992740482268,
+      "learning_rate": 1.070446595642104e-05,
+      "loss": 0.7885,
+      "step": 20936
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4103984898989712,
+      "learning_rate": 1.0703704809046211e-05,
+      "loss": 0.7313,
+      "step": 20937
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.358191084255322,
+      "learning_rate": 1.0702943657574193e-05,
+      "loss": 0.8062,
+      "step": 20938
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3268619285042345,
+      "learning_rate": 1.0702182502009412e-05,
+      "loss": 0.5992,
+      "step": 20939
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.62400024622166,
+      "learning_rate": 1.0701421342356304e-05,
+      "loss": 0.5182,
+      "step": 20940
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5870296229055868,
+      "learning_rate": 1.0700660178619297e-05,
+      "loss": 0.4966,
+      "step": 20941
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3520255561740353,
+      "learning_rate": 1.069989901080283e-05,
+      "loss": 0.7261,
+      "step": 20942
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.222827143568881,
+      "learning_rate": 1.0699137838911323e-05,
+      "loss": 0.6789,
+      "step": 20943
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5577013328411549,
+      "learning_rate": 1.0698376662949217e-05,
+      "loss": 0.6785,
+      "step": 20944
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3927648071398773,
+      "learning_rate": 1.0697615482920937e-05,
+      "loss": 0.6202,
+      "step": 20945
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5208478135581567,
+      "learning_rate": 1.0696854298830925e-05,
+      "loss": 0.771,
+      "step": 20946
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.552393536592064,
+      "learning_rate": 1.0696093110683603e-05,
+      "loss": 0.8289,
+      "step": 20947
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9624578993496005,
+      "learning_rate": 1.069533191848341e-05,
+      "loss": 0.8043,
+      "step": 20948
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2936504724237705,
+      "learning_rate": 1.0694570722234769e-05,
+      "loss": 0.7095,
+      "step": 20949
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3092162633210953,
+      "learning_rate": 1.0693809521942118e-05,
+      "loss": 0.7074,
+      "step": 20950
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6462386126116373,
+      "learning_rate": 1.0693048317609893e-05,
+      "loss": 0.7414,
+      "step": 20951
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1955488256620583,
+      "learning_rate": 1.0692287109242516e-05,
+      "loss": 0.7034,
+      "step": 20952
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.604444735968704,
+      "learning_rate": 1.069152589684443e-05,
+      "loss": 0.6481,
+      "step": 20953
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.319136977105975,
+      "learning_rate": 1.0690764680420056e-05,
+      "loss": 0.7017,
+      "step": 20954
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4152918660788922,
+      "learning_rate": 1.0690003459973836e-05,
+      "loss": 0.7739,
+      "step": 20955
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4034286295682354,
+      "learning_rate": 1.0689242235510195e-05,
+      "loss": 0.7455,
+      "step": 20956
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6120903157937894,
+      "learning_rate": 1.0688481007033566e-05,
+      "loss": 0.819,
+      "step": 20957
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5093360591222553,
+      "learning_rate": 1.0687719774548384e-05,
+      "loss": 0.6598,
+      "step": 20958
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2918660517647251,
+      "learning_rate": 1.068695853805908e-05,
+      "loss": 0.8139,
+      "step": 20959
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2662917291876705,
+      "learning_rate": 1.0686197297570087e-05,
+      "loss": 0.774,
+      "step": 20960
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6185264150582293,
+      "learning_rate": 1.0685436053085833e-05,
+      "loss": 0.6297,
+      "step": 20961
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.308408008700374,
+      "learning_rate": 1.0684674804610756e-05,
+      "loss": 0.7413,
+      "step": 20962
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4858195005728976,
+      "learning_rate": 1.0683913552149286e-05,
+      "loss": 0.7199,
+      "step": 20963
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9987747181235036,
+      "learning_rate": 1.0683152295705853e-05,
+      "loss": 0.7049,
+      "step": 20964
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3387441535669538,
+      "learning_rate": 1.0682391035284892e-05,
+      "loss": 0.632,
+      "step": 20965
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.79400815181877,
+      "learning_rate": 1.0681629770890832e-05,
+      "loss": 0.8275,
+      "step": 20966
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.071216297610292,
+      "learning_rate": 1.068086850252811e-05,
+      "loss": 0.7202,
+      "step": 20967
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.3948821783052123,
+      "learning_rate": 1.0680107230201155e-05,
+      "loss": 0.7162,
+      "step": 20968
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6854148068776984,
+      "learning_rate": 1.06793459539144e-05,
+      "loss": 0.7106,
+      "step": 20969
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7476730793518716,
+      "learning_rate": 1.0678584673672279e-05,
+      "loss": 0.6788,
+      "step": 20970
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.533999273949478,
+      "learning_rate": 1.0677823389479223e-05,
+      "loss": 0.6159,
+      "step": 20971
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4578033409964575,
+      "learning_rate": 1.0677062101339666e-05,
+      "loss": 0.7275,
+      "step": 20972
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.525740606754397,
+      "learning_rate": 1.0676300809258035e-05,
+      "loss": 0.6314,
+      "step": 20973
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5758490080988963,
+      "learning_rate": 1.0675539513238774e-05,
+      "loss": 0.6194,
+      "step": 20974
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5723021369468562,
+      "learning_rate": 1.06747782132863e-05,
+      "loss": 0.7665,
+      "step": 20975
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5023920214889186,
+      "learning_rate": 1.0674016909405059e-05,
+      "loss": 0.6024,
+      "step": 20976
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4456216057717342,
+      "learning_rate": 1.0673255601599472e-05,
+      "loss": 0.8788,
+      "step": 20977
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3046692302007876,
+      "learning_rate": 1.0672494289873985e-05,
+      "loss": 0.6621,
+      "step": 20978
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3102061425929297,
+      "learning_rate": 1.067173297423302e-05,
+      "loss": 0.7681,
+      "step": 20979
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5350864561898152,
+      "learning_rate": 1.0670971654681014e-05,
+      "loss": 0.7355,
+      "step": 20980
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4788969057746422,
+      "learning_rate": 1.0670210331222397e-05,
+      "loss": 0.8263,
+      "step": 20981
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.807530008050002,
+      "learning_rate": 1.0669449003861603e-05,
+      "loss": 0.7091,
+      "step": 20982
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6408247406883417,
+      "learning_rate": 1.0668687672603066e-05,
+      "loss": 0.5207,
+      "step": 20983
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4369787492058133,
+      "learning_rate": 1.0667926337451217e-05,
+      "loss": 0.6555,
+      "step": 20984
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6149092786510898,
+      "learning_rate": 1.0667164998410491e-05,
+      "loss": 0.5416,
+      "step": 20985
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6748821094939907,
+      "learning_rate": 1.0666403655485318e-05,
+      "loss": 0.6678,
+      "step": 20986
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.33706251360176,
+      "learning_rate": 1.066564230868013e-05,
+      "loss": 0.7234,
+      "step": 20987
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9025401959017776,
+      "learning_rate": 1.0664880957999365e-05,
+      "loss": 0.7167,
+      "step": 20988
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4606829616897674,
+      "learning_rate": 1.0664119603447451e-05,
+      "loss": 0.7701,
+      "step": 20989
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3853389081118348,
+      "learning_rate": 1.0663358245028822e-05,
+      "loss": 0.7571,
+      "step": 20990
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.641120145363996,
+      "learning_rate": 1.0662596882747913e-05,
+      "loss": 0.5785,
+      "step": 20991
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3969197827315465,
+      "learning_rate": 1.0661835516609153e-05,
+      "loss": 0.7844,
+      "step": 20992
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.47473951672929,
+      "learning_rate": 1.066107414661698e-05,
+      "loss": 0.7223,
+      "step": 20993
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2735108514988236,
+      "learning_rate": 1.066031277277582e-05,
+      "loss": 0.7046,
+      "step": 20994
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.59096719762216,
+      "learning_rate": 1.0659551395090115e-05,
+      "loss": 0.4881,
+      "step": 20995
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8106706029662178,
+      "learning_rate": 1.0658790013564289e-05,
+      "loss": 0.6867,
+      "step": 20996
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5897475682962439,
+      "learning_rate": 1.0658028628202782e-05,
+      "loss": 0.7004,
+      "step": 20997
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.3605008145933515,
+      "learning_rate": 1.0657267239010021e-05,
+      "loss": 0.6727,
+      "step": 20998
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5328449524898575,
+      "learning_rate": 1.0656505845990442e-05,
+      "loss": 0.6983,
+      "step": 20999
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7531418929586606,
+      "learning_rate": 1.0655744449148482e-05,
+      "loss": 0.6994,
+      "step": 21000
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5667421414915859,
+      "learning_rate": 1.0654983048488566e-05,
+      "loss": 0.4677,
+      "step": 21001
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4748599303179768,
+      "learning_rate": 1.0654221644015135e-05,
+      "loss": 0.79,
+      "step": 21002
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4707047465046807,
+      "learning_rate": 1.0653460235732613e-05,
+      "loss": 0.7512,
+      "step": 21003
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.061457667398274,
+      "learning_rate": 1.0652698823645444e-05,
+      "loss": 0.6877,
+      "step": 21004
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.037196654198807,
+      "learning_rate": 1.0651937407758052e-05,
+      "loss": 0.6893,
+      "step": 21005
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.449454308102996,
+      "learning_rate": 1.0651175988074877e-05,
+      "loss": 0.8755,
+      "step": 21006
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4940415987495268,
+      "learning_rate": 1.0650414564600349e-05,
+      "loss": 0.787,
+      "step": 21007
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.524924358432083,
+      "learning_rate": 1.06496531373389e-05,
+      "loss": 0.7016,
+      "step": 21008
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5575877918360315,
+      "learning_rate": 1.0648891706294967e-05,
+      "loss": 0.5451,
+      "step": 21009
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5689361769948333,
+      "learning_rate": 1.0648130271472978e-05,
+      "loss": 0.5468,
+      "step": 21010
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4180253583506246,
+      "learning_rate": 1.0647368832877371e-05,
+      "loss": 0.6576,
+      "step": 21011
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6067427200452875,
+      "learning_rate": 1.0646607390512578e-05,
+      "loss": 0.5202,
+      "step": 21012
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3721007402277094,
+      "learning_rate": 1.064584594438303e-05,
+      "loss": 0.618,
+      "step": 21013
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.030145891803046,
+      "learning_rate": 1.0645084494493166e-05,
+      "loss": 0.7437,
+      "step": 21014
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4565357135198933,
+      "learning_rate": 1.0644323040847414e-05,
+      "loss": 0.8248,
+      "step": 21015
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4643133576020853,
+      "learning_rate": 1.0643561583450208e-05,
+      "loss": 0.6279,
+      "step": 21016
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6888815220091677,
+      "learning_rate": 1.0642800122305983e-05,
+      "loss": 0.7435,
+      "step": 21017
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5216734971238912,
+      "learning_rate": 1.0642038657419174e-05,
+      "loss": 0.6711,
+      "step": 21018
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6471294529091463,
+      "learning_rate": 1.064127718879421e-05,
+      "loss": 0.5812,
+      "step": 21019
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0404635020119755,
+      "learning_rate": 1.0640515716435529e-05,
+      "loss": 0.6878,
+      "step": 21020
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.611277469882775,
+      "learning_rate": 1.0639754240347564e-05,
+      "loss": 0.7964,
+      "step": 21021
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.665078645297906,
+      "learning_rate": 1.0638992760534743e-05,
+      "loss": 0.7407,
+      "step": 21022
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.686393275910466,
+      "learning_rate": 1.063823127700151e-05,
+      "loss": 0.6955,
+      "step": 21023
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4144779109974097,
+      "learning_rate": 1.0637469789752286e-05,
+      "loss": 0.8434,
+      "step": 21024
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3539708296960375,
+      "learning_rate": 1.0636708298791515e-05,
+      "loss": 0.8824,
+      "step": 21025
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4613194625712336,
+      "learning_rate": 1.0635946804123625e-05,
+      "loss": 0.9363,
+      "step": 21026
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3245320670349339,
+      "learning_rate": 1.0635185305753054e-05,
+      "loss": 0.7422,
+      "step": 21027
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1259047508801703,
+      "learning_rate": 1.0634423803684231e-05,
+      "loss": 0.7032,
+      "step": 21028
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.279443343863871,
+      "learning_rate": 1.0633662297921595e-05,
+      "loss": 0.7732,
+      "step": 21029
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5188102679388233,
+      "learning_rate": 1.0632900788469572e-05,
+      "loss": 0.7346,
+      "step": 21030
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.396009825850425,
+      "learning_rate": 1.06321392753326e-05,
+      "loss": 0.7729,
+      "step": 21031
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3800095182798193,
+      "learning_rate": 1.0631377758515116e-05,
+      "loss": 0.6758,
+      "step": 21032
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5743954785153824,
+      "learning_rate": 1.0630616238021549e-05,
+      "loss": 0.4836,
+      "step": 21033
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.101369153947388,
+      "learning_rate": 1.0629854713856336e-05,
+      "loss": 0.7113,
+      "step": 21034
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5043902318356608,
+      "learning_rate": 1.0629093186023906e-05,
+      "loss": 0.7818,
+      "step": 21035
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3926554104761828,
+      "learning_rate": 1.06283316545287e-05,
+      "loss": 0.6778,
+      "step": 21036
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4034289493118617,
+      "learning_rate": 1.0627570119375148e-05,
+      "loss": 0.7495,
+      "step": 21037
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4642406080027932,
+      "learning_rate": 1.0626808580567683e-05,
+      "loss": 0.7079,
+      "step": 21038
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.106835619876897,
+      "learning_rate": 1.0626047038110741e-05,
+      "loss": 0.6851,
+      "step": 21039
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6607871533797203,
+      "learning_rate": 1.0625285492008752e-05,
+      "loss": 0.6328,
+      "step": 21040
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.618334101475909,
+      "learning_rate": 1.0624523942266157e-05,
+      "loss": 0.531,
+      "step": 21041
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5185842372788247,
+      "learning_rate": 1.0623762388887382e-05,
+      "loss": 0.7535,
+      "step": 21042
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3509956636621996,
+      "learning_rate": 1.0623000831876868e-05,
+      "loss": 0.7884,
+      "step": 21043
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.541295057430413,
+      "learning_rate": 1.0622239271239048e-05,
+      "loss": 0.6499,
+      "step": 21044
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4281017723440506,
+      "learning_rate": 1.0621477706978347e-05,
+      "loss": 0.6278,
+      "step": 21045
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4262383432938712,
+      "learning_rate": 1.062071613909921e-05,
+      "loss": 0.6293,
+      "step": 21046
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4674590221094796,
+      "learning_rate": 1.0619954567606066e-05,
+      "loss": 0.8734,
+      "step": 21047
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3683802023310456,
+      "learning_rate": 1.0619192992503351e-05,
+      "loss": 0.6827,
+      "step": 21048
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.422836442639925,
+      "learning_rate": 1.0618431413795498e-05,
+      "loss": 0.7986,
+      "step": 21049
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4865416509487261,
+      "learning_rate": 1.0617669831486944e-05,
+      "loss": 0.778,
+      "step": 21050
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.392912652723282,
+      "learning_rate": 1.0616908245582121e-05,
+      "loss": 0.8347,
+      "step": 21051
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3856480120485033,
+      "learning_rate": 1.0616146656085457e-05,
+      "loss": 0.7564,
+      "step": 21052
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4300404528289419,
+      "learning_rate": 1.0615385063001398e-05,
+      "loss": 0.7455,
+      "step": 21053
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6153892814759024,
+      "learning_rate": 1.0614623466334368e-05,
+      "loss": 0.5409,
+      "step": 21054
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6179535841451227,
+      "learning_rate": 1.0613861866088809e-05,
+      "loss": 0.8168,
+      "step": 21055
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3977487650648763,
+      "learning_rate": 1.0613100262269148e-05,
+      "loss": 0.6261,
+      "step": 21056
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5721795817123716,
+      "learning_rate": 1.0612338654879827e-05,
+      "loss": 0.5173,
+      "step": 21057
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2536453034570814,
+      "learning_rate": 1.0611577043925276e-05,
+      "loss": 0.6659,
+      "step": 21058
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3607706532102417,
+      "learning_rate": 1.0610815429409928e-05,
+      "loss": 0.7242,
+      "step": 21059
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5345809936360335,
+      "learning_rate": 1.0610053811338222e-05,
+      "loss": 0.736,
+      "step": 21060
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9313379184388972,
+      "learning_rate": 1.0609292189714586e-05,
+      "loss": 0.7048,
+      "step": 21061
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3741895382553586,
+      "learning_rate": 1.060853056454346e-05,
+      "loss": 0.8898,
+      "step": 21062
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4936580970095463,
+      "learning_rate": 1.0607768935829277e-05,
+      "loss": 0.6217,
+      "step": 21063
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3308415377365272,
+      "learning_rate": 1.0607007303576469e-05,
+      "loss": 0.7657,
+      "step": 21064
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3473864629446224,
+      "learning_rate": 1.0606245667789473e-05,
+      "loss": 0.8085,
+      "step": 21065
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4774090371350717,
+      "learning_rate": 1.0605484028472723e-05,
+      "loss": 0.7046,
+      "step": 21066
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6639991434201642,
+      "learning_rate": 1.0604722385630652e-05,
+      "loss": 0.8678,
+      "step": 21067
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5948098926294871,
+      "learning_rate": 1.0603960739267698e-05,
+      "loss": 0.5536,
+      "step": 21068
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7115817860450793,
+      "learning_rate": 1.0603199089388293e-05,
+      "loss": 0.7286,
+      "step": 21069
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5157578632990123,
+      "learning_rate": 1.060243743599687e-05,
+      "loss": 0.7053,
+      "step": 21070
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5886008473974139,
+      "learning_rate": 1.0601675779097868e-05,
+      "loss": 0.5072,
+      "step": 21071
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2593310817489896,
+      "learning_rate": 1.060091411869572e-05,
+      "loss": 0.6869,
+      "step": 21072
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5974552722473772,
+      "learning_rate": 1.0600152454794853e-05,
+      "loss": 0.6676,
+      "step": 21073
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.431830932401177,
+      "learning_rate": 1.0599390787399717e-05,
+      "loss": 0.6383,
+      "step": 21074
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5899773823348374,
+      "learning_rate": 1.0598629116514733e-05,
+      "loss": 0.4936,
+      "step": 21075
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3115972711858301,
+      "learning_rate": 1.0597867442144343e-05,
+      "loss": 0.6393,
+      "step": 21076
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4064371453350404,
+      "learning_rate": 1.0597105764292976e-05,
+      "loss": 0.8215,
+      "step": 21077
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2983036947557516,
+      "learning_rate": 1.0596344082965076e-05,
+      "loss": 0.6922,
+      "step": 21078
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3000619297699418,
+      "learning_rate": 1.0595582398165067e-05,
+      "loss": 0.8014,
+      "step": 21079
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6080738797598071,
+      "learning_rate": 1.059482070989739e-05,
+      "loss": 0.5134,
+      "step": 21080
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5793236494388005,
+      "learning_rate": 1.059405901816648e-05,
+      "loss": 0.5094,
+      "step": 21081
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3269110295178428,
+      "learning_rate": 1.059329732297677e-05,
+      "loss": 0.6661,
+      "step": 21082
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.252374258882352,
+      "learning_rate": 1.0592535624332694e-05,
+      "loss": 0.8039,
+      "step": 21083
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5110327412476063,
+      "learning_rate": 1.0591773922238685e-05,
+      "loss": 0.8397,
+      "step": 21084
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.301665456711815,
+      "learning_rate": 1.0591012216699189e-05,
+      "loss": 0.753,
+      "step": 21085
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5761763783911317,
+      "learning_rate": 1.0590250507718624e-05,
+      "loss": 0.5216,
+      "step": 21086
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4018596536396342,
+      "learning_rate": 1.058948879530144e-05,
+      "loss": 0.8646,
+      "step": 21087
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2413243395237017,
+      "learning_rate": 1.0588727079452063e-05,
+      "loss": 0.6432,
+      "step": 21088
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3988797945849032,
+      "learning_rate": 1.058796536017493e-05,
+      "loss": 0.724,
+      "step": 21089
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4410155107623377,
+      "learning_rate": 1.058720363747448e-05,
+      "loss": 0.7745,
+      "step": 21090
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.683494651587013,
+      "learning_rate": 1.0586441911355138e-05,
+      "loss": 0.7254,
+      "step": 21091
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0642152244114307,
+      "learning_rate": 1.0585680181821351e-05,
+      "loss": 0.7971,
+      "step": 21092
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5180955497466315,
+      "learning_rate": 1.0584918448877547e-05,
+      "loss": 0.5366,
+      "step": 21093
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3114933271335327,
+      "learning_rate": 1.0584156712528161e-05,
+      "loss": 0.7832,
+      "step": 21094
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3792713777245993,
+      "learning_rate": 1.0583394972777632e-05,
+      "loss": 0.6805,
+      "step": 21095
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4584694887835818,
+      "learning_rate": 1.0582633229630393e-05,
+      "loss": 0.7138,
+      "step": 21096
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.313229425893257,
+      "learning_rate": 1.0581871483090878e-05,
+      "loss": 0.7769,
+      "step": 21097
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5409201867267996,
+      "learning_rate": 1.0581109733163522e-05,
+      "loss": 0.7747,
+      "step": 21098
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4240570021353665,
+      "learning_rate": 1.0580347979852763e-05,
+      "loss": 0.7537,
+      "step": 21099
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.529291146330149,
+      "learning_rate": 1.0579586223163032e-05,
+      "loss": 0.71,
+      "step": 21100
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6186091480214002,
+      "learning_rate": 1.0578824463098769e-05,
+      "loss": 0.4992,
+      "step": 21101
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8489395580354975,
+      "learning_rate": 1.0578062699664406e-05,
+      "loss": 0.6929,
+      "step": 21102
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.316978229501051,
+      "learning_rate": 1.0577300932864378e-05,
+      "loss": 0.7614,
+      "step": 21103
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.52725140633221,
+      "learning_rate": 1.0576539162703123e-05,
+      "loss": 0.615,
+      "step": 21104
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4721309481757314,
+      "learning_rate": 1.0575777389185071e-05,
+      "loss": 0.7881,
+      "step": 21105
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4928884811903822,
+      "learning_rate": 1.0575015612314665e-05,
+      "loss": 0.6936,
+      "step": 21106
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6098750088414666,
+      "learning_rate": 1.0574253832096332e-05,
+      "loss": 0.6591,
+      "step": 21107
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.091660408726573,
+      "learning_rate": 1.0573492048534515e-05,
+      "loss": 0.6766,
+      "step": 21108
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5442635115950345,
+      "learning_rate": 1.0572730261633644e-05,
+      "loss": 0.8677,
+      "step": 21109
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.466975461360874,
+      "learning_rate": 1.0571968471398157e-05,
+      "loss": 0.6552,
+      "step": 21110
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4523983062413004,
+      "learning_rate": 1.0571206677832488e-05,
+      "loss": 0.7472,
+      "step": 21111
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3869737382311444,
+      "learning_rate": 1.0570444880941074e-05,
+      "loss": 0.6881,
+      "step": 21112
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.39174896009657,
+      "learning_rate": 1.0569683080728345e-05,
+      "loss": 0.7977,
+      "step": 21113
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.41226961465497,
+      "learning_rate": 1.0568921277198744e-05,
+      "loss": 0.6945,
+      "step": 21114
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3752853319919285,
+      "learning_rate": 1.0568159470356702e-05,
+      "loss": 0.531,
+      "step": 21115
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5514644205209471,
+      "learning_rate": 1.0567397660206658e-05,
+      "loss": 0.7814,
+      "step": 21116
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5984335367808645,
+      "learning_rate": 1.0566635846753044e-05,
+      "loss": 0.5466,
+      "step": 21117
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6516786521466331,
+      "learning_rate": 1.0565874030000296e-05,
+      "loss": 0.5614,
+      "step": 21118
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4130715864040342,
+      "learning_rate": 1.056511220995285e-05,
+      "loss": 0.7062,
+      "step": 21119
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5336287610107846,
+      "learning_rate": 1.0564350386615143e-05,
+      "loss": 0.78,
+      "step": 21120
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.521793588525009,
+      "learning_rate": 1.0563588559991608e-05,
+      "loss": 0.6281,
+      "step": 21121
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4092064727177644,
+      "learning_rate": 1.0562826730086684e-05,
+      "loss": 0.8062,
+      "step": 21122
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.427201769177963,
+      "learning_rate": 1.0562064896904804e-05,
+      "loss": 0.716,
+      "step": 21123
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.420515926194133,
+      "learning_rate": 1.0561303060450404e-05,
+      "loss": 0.7494,
+      "step": 21124
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6323156169194866,
+      "learning_rate": 1.056054122072792e-05,
+      "loss": 0.6622,
+      "step": 21125
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8369750229827764,
+      "learning_rate": 1.0559779377741786e-05,
+      "loss": 0.7565,
+      "step": 21126
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9179407084651456,
+      "learning_rate": 1.055901753149644e-05,
+      "loss": 0.7138,
+      "step": 21127
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8845248694123213,
+      "learning_rate": 1.0558255681996316e-05,
+      "loss": 0.7434,
+      "step": 21128
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3520352883017828,
+      "learning_rate": 1.0557493829245854e-05,
+      "loss": 0.7312,
+      "step": 21129
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2270748745841824,
+      "learning_rate": 1.0556731973249486e-05,
+      "loss": 0.7634,
+      "step": 21130
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7987517176489631,
+      "learning_rate": 1.0555970114011644e-05,
+      "loss": 0.737,
+      "step": 21131
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3462649008460632,
+      "learning_rate": 1.055520825153677e-05,
+      "loss": 0.757,
+      "step": 21132
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4638622439046822,
+      "learning_rate": 1.0554446385829297e-05,
+      "loss": 1.0062,
+      "step": 21133
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.355946949624523,
+      "learning_rate": 1.0553684516893665e-05,
+      "loss": 0.7337,
+      "step": 21134
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.571642663326786,
+      "learning_rate": 1.0552922644734302e-05,
+      "loss": 0.5191,
+      "step": 21135
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4906876001264147,
+      "learning_rate": 1.0552160769355653e-05,
+      "loss": 0.6571,
+      "step": 21136
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6012213062633864,
+      "learning_rate": 1.0551398890762148e-05,
+      "loss": 0.5422,
+      "step": 21137
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5583270709804282,
+      "learning_rate": 1.0550637008958222e-05,
+      "loss": 0.8595,
+      "step": 21138
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3975625989164966,
+      "learning_rate": 1.0549875123948314e-05,
+      "loss": 0.7291,
+      "step": 21139
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8334905410411402,
+      "learning_rate": 1.0549113235736858e-05,
+      "loss": 0.7967,
+      "step": 21140
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4184510176107663,
+      "learning_rate": 1.0548351344328293e-05,
+      "loss": 0.7042,
+      "step": 21141
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2958238506848296,
+      "learning_rate": 1.0547589449727051e-05,
+      "loss": 0.6603,
+      "step": 21142
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4191383758034937,
+      "learning_rate": 1.054682755193757e-05,
+      "loss": 0.7007,
+      "step": 21143
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6245743172638727,
+      "learning_rate": 1.0546065650964288e-05,
+      "loss": 0.7798,
+      "step": 21144
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3703146747764499,
+      "learning_rate": 1.0545303746811637e-05,
+      "loss": 0.7979,
+      "step": 21145
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3634365213657507,
+      "learning_rate": 1.0544541839484055e-05,
+      "loss": 0.6739,
+      "step": 21146
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5990737649182323,
+      "learning_rate": 1.0543779928985978e-05,
+      "loss": 0.4887,
+      "step": 21147
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9952036115899345,
+      "learning_rate": 1.0543018015321842e-05,
+      "loss": 0.7573,
+      "step": 21148
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9696118880553593,
+      "learning_rate": 1.0542256098496085e-05,
+      "loss": 0.754,
+      "step": 21149
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5144550473569602,
+      "learning_rate": 1.0541494178513138e-05,
+      "loss": 0.8394,
+      "step": 21150
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6112642321418031,
+      "learning_rate": 1.0540732255377442e-05,
+      "loss": 0.6818,
+      "step": 21151
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7090442728036297,
+      "learning_rate": 1.0539970329093432e-05,
+      "loss": 0.5828,
+      "step": 21152
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5060508690303793,
+      "learning_rate": 1.0539208399665546e-05,
+      "loss": 0.6409,
+      "step": 21153
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6153849069530257,
+      "learning_rate": 1.0538446467098214e-05,
+      "loss": 0.7893,
+      "step": 21154
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.340521795157561,
+      "learning_rate": 1.053768453139588e-05,
+      "loss": 0.711,
+      "step": 21155
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4727218378482714,
+      "learning_rate": 1.0536922592562973e-05,
+      "loss": 0.742,
+      "step": 21156
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5588777132982865,
+      "learning_rate": 1.0536160650603935e-05,
+      "loss": 0.5015,
+      "step": 21157
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5187506170133447,
+      "learning_rate": 1.0535398705523197e-05,
+      "loss": 0.7108,
+      "step": 21158
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5419084095180826,
+      "learning_rate": 1.0534636757325203e-05,
+      "loss": 0.6995,
+      "step": 21159
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3978700505163904,
+      "learning_rate": 1.0533874806014382e-05,
+      "loss": 0.696,
+      "step": 21160
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.9418644154162674,
+      "learning_rate": 1.053311285159517e-05,
+      "loss": 0.8849,
+      "step": 21161
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2300564220541992,
+      "learning_rate": 1.053235089407201e-05,
+      "loss": 0.7624,
+      "step": 21162
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6831899312021914,
+      "learning_rate": 1.0531588933449332e-05,
+      "loss": 0.8279,
+      "step": 21163
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5160627010181082,
+      "learning_rate": 1.0530826969731577e-05,
+      "loss": 0.6895,
+      "step": 21164
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5896478389626469,
+      "learning_rate": 1.053006500292318e-05,
+      "loss": 0.5347,
+      "step": 21165
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3222907501880574,
+      "learning_rate": 1.0529303033028575e-05,
+      "loss": 0.8107,
+      "step": 21166
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.31910383956177,
+      "learning_rate": 1.05285410600522e-05,
+      "loss": 0.7011,
+      "step": 21167
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5326769374288207,
+      "learning_rate": 1.0527779083998492e-05,
+      "loss": 0.763,
+      "step": 21168
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6945649711526725,
+      "learning_rate": 1.0527017104871885e-05,
+      "loss": 0.6087,
+      "step": 21169
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.350995727514182,
+      "learning_rate": 1.0526255122676823e-05,
+      "loss": 0.7178,
+      "step": 21170
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5600048934622,
+      "learning_rate": 1.0525493137417733e-05,
+      "loss": 0.7042,
+      "step": 21171
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4237348910745704,
+      "learning_rate": 1.0524731149099057e-05,
+      "loss": 0.7226,
+      "step": 21172
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.323169047901727,
+      "learning_rate": 1.0523969157725229e-05,
+      "loss": 0.8052,
+      "step": 21173
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5177419684609068,
+      "learning_rate": 1.0523207163300688e-05,
+      "loss": 0.6622,
+      "step": 21174
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3460227986242002,
+      "learning_rate": 1.0522445165829867e-05,
+      "loss": 0.7561,
+      "step": 21175
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3334446603662689,
+      "learning_rate": 1.0521683165317209e-05,
+      "loss": 0.6854,
+      "step": 21176
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.447622838759551,
+      "learning_rate": 1.0520921161767142e-05,
+      "loss": 0.7878,
+      "step": 21177
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3957999327778532,
+      "learning_rate": 1.0520159155184111e-05,
+      "loss": 0.7335,
+      "step": 21178
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3317500224334204,
+      "learning_rate": 1.0519397145572542e-05,
+      "loss": 0.7226,
+      "step": 21179
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5218907640916206,
+      "learning_rate": 1.0518635132936888e-05,
+      "loss": 0.7297,
+      "step": 21180
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5890916205955692,
+      "learning_rate": 1.051787311728157e-05,
+      "loss": 0.8877,
+      "step": 21181
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.401143333847699,
+      "learning_rate": 1.0517111098611028e-05,
+      "loss": 0.6574,
+      "step": 21182
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4377221541364067,
+      "learning_rate": 1.0516349076929708e-05,
+      "loss": 0.6534,
+      "step": 21183
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5770659739327646,
+      "learning_rate": 1.0515587052242035e-05,
+      "loss": 0.8056,
+      "step": 21184
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5205381537560314,
+      "learning_rate": 1.0514825024552456e-05,
+      "loss": 0.7721,
+      "step": 21185
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4959565137182342,
+      "learning_rate": 1.0514062993865396e-05,
+      "loss": 0.7247,
+      "step": 21186
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5952603919110033,
+      "learning_rate": 1.0513300960185307e-05,
+      "loss": 0.5251,
+      "step": 21187
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4635452444302124,
+      "learning_rate": 1.0512538923516613e-05,
+      "loss": 0.6897,
+      "step": 21188
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5182879714120152,
+      "learning_rate": 1.0511776883863753e-05,
+      "loss": 0.7338,
+      "step": 21189
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4574050221956754,
+      "learning_rate": 1.0511014841231168e-05,
+      "loss": 0.717,
+      "step": 21190
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.257741840144957,
+      "learning_rate": 1.0510252795623294e-05,
+      "loss": 0.6323,
+      "step": 21191
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4778241196258146,
+      "learning_rate": 1.0509490747044566e-05,
+      "loss": 0.7683,
+      "step": 21192
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.667827081034669,
+      "learning_rate": 1.050872869549942e-05,
+      "loss": 0.7617,
+      "step": 21193
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6404040396695907,
+      "learning_rate": 1.0507966640992295e-05,
+      "loss": 0.6916,
+      "step": 21194
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5852178215615185,
+      "learning_rate": 1.0507204583527629e-05,
+      "loss": 0.6521,
+      "step": 21195
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2827476068385886,
+      "learning_rate": 1.0506442523109857e-05,
+      "loss": 0.7332,
+      "step": 21196
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5119596923362812,
+      "learning_rate": 1.0505680459743414e-05,
+      "loss": 0.7431,
+      "step": 21197
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5679994163429365,
+      "learning_rate": 1.050491839343274e-05,
+      "loss": 0.6883,
+      "step": 21198
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3330664187995578,
+      "learning_rate": 1.0504156324182274e-05,
+      "loss": 0.7583,
+      "step": 21199
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.665493485774606,
+      "learning_rate": 1.0503394251996446e-05,
+      "loss": 0.8051,
+      "step": 21200
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.612926430749676,
+      "learning_rate": 1.0502632176879701e-05,
+      "loss": 0.72,
+      "step": 21201
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.651721632894199,
+      "learning_rate": 1.0501870098836473e-05,
+      "loss": 0.7723,
+      "step": 21202
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.709303999967277,
+      "learning_rate": 1.0501108017871193e-05,
+      "loss": 0.753,
+      "step": 21203
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4672187154416336,
+      "learning_rate": 1.050034593398831e-05,
+      "loss": 0.8267,
+      "step": 21204
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8042519149066225,
+      "learning_rate": 1.0499583847192249e-05,
+      "loss": 0.6865,
+      "step": 21205
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.624093438039799,
+      "learning_rate": 1.0498821757487456e-05,
+      "loss": 0.5631,
+      "step": 21206
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3762043466335323,
+      "learning_rate": 1.0498059664878361e-05,
+      "loss": 0.691,
+      "step": 21207
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.534126237716195,
+      "learning_rate": 1.0497297569369412e-05,
+      "loss": 0.7029,
+      "step": 21208
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.332991424233178,
+      "learning_rate": 1.0496535470965033e-05,
+      "loss": 0.9585,
+      "step": 21209
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6961414674033972,
+      "learning_rate": 1.049577336966967e-05,
+      "loss": 0.7185,
+      "step": 21210
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3531593365191457,
+      "learning_rate": 1.0495011265487758e-05,
+      "loss": 0.5702,
+      "step": 21211
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5777563043792462,
+      "learning_rate": 1.0494249158423733e-05,
+      "loss": 0.8094,
+      "step": 21212
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1002411252865887,
+      "learning_rate": 1.0493487048482031e-05,
+      "loss": 0.8173,
+      "step": 21213
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5860082194496861,
+      "learning_rate": 1.049272493566709e-05,
+      "loss": 0.4775,
+      "step": 21214
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3764480964344483,
+      "learning_rate": 1.0491962819983351e-05,
+      "loss": 0.6535,
+      "step": 21215
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5756851430451784,
+      "learning_rate": 1.0491200701435252e-05,
+      "loss": 0.6745,
+      "step": 21216
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3425205882168585,
+      "learning_rate": 1.0490438580027223e-05,
+      "loss": 0.753,
+      "step": 21217
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2490037251587,
+      "learning_rate": 1.0489676455763704e-05,
+      "loss": 0.7029,
+      "step": 21218
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4048279855846753,
+      "learning_rate": 1.0488914328649136e-05,
+      "loss": 0.8264,
+      "step": 21219
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3604176162463026,
+      "learning_rate": 1.0488152198687955e-05,
+      "loss": 0.7255,
+      "step": 21220
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.188983491614568,
+      "learning_rate": 1.0487390065884596e-05,
+      "loss": 0.6986,
+      "step": 21221
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4069958162356926,
+      "learning_rate": 1.0486627930243497e-05,
+      "loss": 0.7282,
+      "step": 21222
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.965581403545775,
+      "learning_rate": 1.0485865791769096e-05,
+      "loss": 0.8131,
+      "step": 21223
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1484899550202217,
+      "learning_rate": 1.0485103650465832e-05,
+      "loss": 0.6651,
+      "step": 21224
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.77019494547306,
+      "learning_rate": 1.0484341506338143e-05,
+      "loss": 0.7244,
+      "step": 21225
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5053217475453367,
+      "learning_rate": 1.0483579359390459e-05,
+      "loss": 0.7139,
+      "step": 21226
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3774049666976698,
+      "learning_rate": 1.0482817209627228e-05,
+      "loss": 0.649,
+      "step": 21227
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4505920522812326,
+      "learning_rate": 1.0482055057052878e-05,
+      "loss": 0.7295,
+      "step": 21228
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.418426744667907,
+      "learning_rate": 1.0481292901671853e-05,
+      "loss": 0.682,
+      "step": 21229
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5659656216689337,
+      "learning_rate": 1.0480530743488586e-05,
+      "loss": 0.7474,
+      "step": 21230
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3295695741199702,
+      "learning_rate": 1.0479768582507519e-05,
+      "loss": 0.6219,
+      "step": 21231
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7282893661474403,
+      "learning_rate": 1.0479006418733091e-05,
+      "loss": 0.7125,
+      "step": 21232
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.427702365550587,
+      "learning_rate": 1.047824425216973e-05,
+      "loss": 0.7914,
+      "step": 21233
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.572051084879285,
+      "learning_rate": 1.0477482082821884e-05,
+      "loss": 0.6958,
+      "step": 21234
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5494163020625145,
+      "learning_rate": 1.0476719910693981e-05,
+      "loss": 0.8148,
+      "step": 21235
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.535060222003826,
+      "learning_rate": 1.0475957735790469e-05,
+      "loss": 0.6198,
+      "step": 21236
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6450314346207215,
+      "learning_rate": 1.0475195558115776e-05,
+      "loss": 0.7453,
+      "step": 21237
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3150044344490548,
+      "learning_rate": 1.0474433377674349e-05,
+      "loss": 0.7792,
+      "step": 21238
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8888880566641981,
+      "learning_rate": 1.0473671194470618e-05,
+      "loss": 0.8307,
+      "step": 21239
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.529112295642548,
+      "learning_rate": 1.0472909008509023e-05,
+      "loss": 0.6559,
+      "step": 21240
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4630752414883588,
+      "learning_rate": 1.0472146819794003e-05,
+      "loss": 0.6814,
+      "step": 21241
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6267785271870214,
+      "learning_rate": 1.0471384628329993e-05,
+      "loss": 0.6815,
+      "step": 21242
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.887121635139002,
+      "learning_rate": 1.0470622434121434e-05,
+      "loss": 0.7013,
+      "step": 21243
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6374631755683784,
+      "learning_rate": 1.0469860237172759e-05,
+      "loss": 0.7637,
+      "step": 21244
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6142184112034652,
+      "learning_rate": 1.0469098037488415e-05,
+      "loss": 0.7083,
+      "step": 21245
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5057299080784154,
+      "learning_rate": 1.0468335835072829e-05,
+      "loss": 0.7771,
+      "step": 21246
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3889501267064976,
+      "learning_rate": 1.0467573629930445e-05,
+      "loss": 0.6939,
+      "step": 21247
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.587826260365814,
+      "learning_rate": 1.0466811422065699e-05,
+      "loss": 0.6503,
+      "step": 21248
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.388485892250222,
+      "learning_rate": 1.046604921148303e-05,
+      "loss": 0.6137,
+      "step": 21249
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4594987200476115,
+      "learning_rate": 1.0465286998186872e-05,
+      "loss": 0.6584,
+      "step": 21250
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3116173074396895,
+      "learning_rate": 1.0464524782181669e-05,
+      "loss": 0.7916,
+      "step": 21251
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3102422974474457,
+      "learning_rate": 1.0463762563471854e-05,
+      "loss": 0.8088,
+      "step": 21252
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.9268446950883074,
+      "learning_rate": 1.0463000342061868e-05,
+      "loss": 0.7141,
+      "step": 21253
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4732477070899592,
+      "learning_rate": 1.0462238117956145e-05,
+      "loss": 0.7085,
+      "step": 21254
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4661546578655278,
+      "learning_rate": 1.046147589115913e-05,
+      "loss": 0.7956,
+      "step": 21255
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4800858970866635,
+      "learning_rate": 1.046071366167525e-05,
+      "loss": 0.8128,
+      "step": 21256
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6078431410288965,
+      "learning_rate": 1.0459951429508955e-05,
+      "loss": 0.4992,
+      "step": 21257
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6067529270954226,
+      "learning_rate": 1.0459189194664671e-05,
+      "loss": 0.5929,
+      "step": 21258
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3806738301363495,
+      "learning_rate": 1.045842695714685e-05,
+      "loss": 0.7236,
+      "step": 21259
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2503981945565612,
+      "learning_rate": 1.0457664716959916e-05,
+      "loss": 0.7522,
+      "step": 21260
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2149819782379512,
+      "learning_rate": 1.0456902474108315e-05,
+      "loss": 0.6134,
+      "step": 21261
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.533581248280162,
+      "learning_rate": 1.0456140228596484e-05,
+      "loss": 0.6903,
+      "step": 21262
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5065082500090101,
+      "learning_rate": 1.0455377980428858e-05,
+      "loss": 0.7261,
+      "step": 21263
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7356605745214064,
+      "learning_rate": 1.045461572960988e-05,
+      "loss": 0.6894,
+      "step": 21264
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8296576950943821,
+      "learning_rate": 1.0453853476143982e-05,
+      "loss": 0.7766,
+      "step": 21265
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5951404456028995,
+      "learning_rate": 1.0453091220035607e-05,
+      "loss": 0.6473,
+      "step": 21266
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4482400259658872,
+      "learning_rate": 1.0452328961289192e-05,
+      "loss": 0.6231,
+      "step": 21267
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.429014437395108,
+      "learning_rate": 1.0451566699909174e-05,
+      "loss": 0.6466,
+      "step": 21268
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5258236442954003,
+      "learning_rate": 1.0450804435899993e-05,
+      "loss": 0.7139,
+      "step": 21269
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0309531613142253,
+      "learning_rate": 1.0450042169266083e-05,
+      "loss": 0.722,
+      "step": 21270
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4774184071162846,
+      "learning_rate": 1.0449279900011888e-05,
+      "loss": 0.8495,
+      "step": 21271
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6156797363818738,
+      "learning_rate": 1.044851762814184e-05,
+      "loss": 0.7422,
+      "step": 21272
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.6167232454779272,
+      "learning_rate": 1.0447755353660384e-05,
+      "loss": 0.7147,
+      "step": 21273
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.345802162669652,
+      "learning_rate": 1.0446993076571952e-05,
+      "loss": 0.7716,
+      "step": 21274
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4422112352362113,
+      "learning_rate": 1.0446230796880987e-05,
+      "loss": 0.8128,
+      "step": 21275
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3098190829478737,
+      "learning_rate": 1.0445468514591925e-05,
+      "loss": 0.7253,
+      "step": 21276
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3846507307150164,
+      "learning_rate": 1.0444706229709201e-05,
+      "loss": 0.7616,
+      "step": 21277
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2876394311500827,
+      "learning_rate": 1.044394394223726e-05,
+      "loss": 0.8174,
+      "step": 21278
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.538222685839157,
+      "learning_rate": 1.0443181652180534e-05,
+      "loss": 0.6832,
+      "step": 21279
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6150813961761225,
+      "learning_rate": 1.0442419359543465e-05,
+      "loss": 0.5215,
+      "step": 21280
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4907618168169932,
+      "learning_rate": 1.044165706433049e-05,
+      "loss": 0.7254,
+      "step": 21281
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5688653865188387,
+      "learning_rate": 1.0440894766546051e-05,
+      "loss": 0.8007,
+      "step": 21282
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7343235511152375,
+      "learning_rate": 1.0440132466194582e-05,
+      "loss": 0.6772,
+      "step": 21283
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2160252927693729,
+      "learning_rate": 1.0439370163280518e-05,
+      "loss": 0.705,
+      "step": 21284
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6490842684114644,
+      "learning_rate": 1.0438607857808307e-05,
+      "loss": 0.6184,
+      "step": 21285
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2484031991971838,
+      "learning_rate": 1.0437845549782377e-05,
+      "loss": 0.7159,
+      "step": 21286
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7637499771707557,
+      "learning_rate": 1.0437083239207177e-05,
+      "loss": 0.7264,
+      "step": 21287
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6338480279850167,
+      "learning_rate": 1.0436320926087135e-05,
+      "loss": 0.6403,
+      "step": 21288
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.398484415935265,
+      "learning_rate": 1.0435558610426701e-05,
+      "loss": 0.6722,
+      "step": 21289
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.280835326394142,
+      "learning_rate": 1.0434796292230303e-05,
+      "loss": 0.8235,
+      "step": 21290
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.677422596521454,
+      "learning_rate": 1.0434033971502381e-05,
+      "loss": 0.6876,
+      "step": 21291
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5984601781717831,
+      "learning_rate": 1.043327164824738e-05,
+      "loss": 0.5506,
+      "step": 21292
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7564993344076316,
+      "learning_rate": 1.0432509322469732e-05,
+      "loss": 0.7263,
+      "step": 21293
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6268978113412512,
+      "learning_rate": 1.043174699417388e-05,
+      "loss": 0.5785,
+      "step": 21294
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5904827042354949,
+      "learning_rate": 1.0430984663364256e-05,
+      "loss": 0.5539,
+      "step": 21295
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2605069563323714,
+      "learning_rate": 1.0430222330045306e-05,
+      "loss": 0.807,
+      "step": 21296
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4070057389712496,
+      "learning_rate": 1.0429459994221465e-05,
+      "loss": 0.6678,
+      "step": 21297
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6600629571689247,
+      "learning_rate": 1.0428697655897172e-05,
+      "loss": 0.733,
+      "step": 21298
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4547237602486032,
+      "learning_rate": 1.0427935315076864e-05,
+      "loss": 0.8128,
+      "step": 21299
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.478908534942905,
+      "learning_rate": 1.0427172971764984e-05,
+      "loss": 0.7201,
+      "step": 21300
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4694580717350116,
+      "learning_rate": 1.0426410625965965e-05,
+      "loss": 0.669,
+      "step": 21301
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4208863462879147,
+      "learning_rate": 1.0425648277684251e-05,
+      "loss": 0.7536,
+      "step": 21302
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4957924805347846,
+      "learning_rate": 1.0424885926924275e-05,
+      "loss": 0.6717,
+      "step": 21303
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.890855086667169,
+      "learning_rate": 1.042412357369048e-05,
+      "loss": 0.7355,
+      "step": 21304
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.498511109185583,
+      "learning_rate": 1.0423361217987303e-05,
+      "loss": 0.7723,
+      "step": 21305
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.961977472391381,
+      "learning_rate": 1.0422598859819186e-05,
+      "loss": 0.7253,
+      "step": 21306
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4576252578083917,
+      "learning_rate": 1.0421836499190559e-05,
+      "loss": 0.7674,
+      "step": 21307
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4378579672155194,
+      "learning_rate": 1.0421074136105872e-05,
+      "loss": 0.6311,
+      "step": 21308
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5788175170057503,
+      "learning_rate": 1.0420311770569553e-05,
+      "loss": 0.7575,
+      "step": 21309
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.551566373101865,
+      "learning_rate": 1.0419549402586049e-05,
+      "loss": 0.4911,
+      "step": 21310
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.27637112193634,
+      "learning_rate": 1.0418787032159797e-05,
+      "loss": 0.693,
+      "step": 21311
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5527627088590215,
+      "learning_rate": 1.041802465929523e-05,
+      "loss": 0.628,
+      "step": 21312
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7273196630424623,
+      "learning_rate": 1.0417262283996793e-05,
+      "loss": 0.6638,
+      "step": 21313
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.224619498140443,
+      "learning_rate": 1.0416499906268922e-05,
+      "loss": 0.5949,
+      "step": 21314
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4773163242496623,
+      "learning_rate": 1.0415737526116059e-05,
+      "loss": 0.6597,
+      "step": 21315
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8734783597821476,
+      "learning_rate": 1.0414975143542636e-05,
+      "loss": 0.7107,
+      "step": 21316
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5360747979157503,
+      "learning_rate": 1.0414212758553102e-05,
+      "loss": 0.673,
+      "step": 21317
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.716857401703911,
+      "learning_rate": 1.0413450371151888e-05,
+      "loss": 0.7553,
+      "step": 21318
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.609828396657795,
+      "learning_rate": 1.0412687981343435e-05,
+      "loss": 0.4975,
+      "step": 21319
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4000058916809766,
+      "learning_rate": 1.0411925589132182e-05,
+      "loss": 0.6438,
+      "step": 21320
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7685087347086916,
+      "learning_rate": 1.0411163194522566e-05,
+      "loss": 0.7535,
+      "step": 21321
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5773024503548435,
+      "learning_rate": 1.041040079751903e-05,
+      "loss": 0.5313,
+      "step": 21322
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3867076410873649,
+      "learning_rate": 1.0409638398126009e-05,
+      "loss": 0.6663,
+      "step": 21323
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3401878612922367,
+      "learning_rate": 1.0408875996347945e-05,
+      "loss": 0.6687,
+      "step": 21324
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.437494952222783,
+      "learning_rate": 1.0408113592189275e-05,
+      "loss": 0.7461,
+      "step": 21325
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4850579289415067,
+      "learning_rate": 1.0407351185654437e-05,
+      "loss": 0.7854,
+      "step": 21326
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6216700662850743,
+      "learning_rate": 1.0406588776747875e-05,
+      "loss": 0.5279,
+      "step": 21327
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4451493890790241,
+      "learning_rate": 1.040582636547402e-05,
+      "loss": 0.6793,
+      "step": 21328
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4216944603255466,
+      "learning_rate": 1.0405063951837318e-05,
+      "loss": 0.7546,
+      "step": 21329
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2783751206759728,
+      "learning_rate": 1.0404301535842205e-05,
+      "loss": 0.7793,
+      "step": 21330
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4908387507029426,
+      "learning_rate": 1.040353911749312e-05,
+      "loss": 0.6917,
+      "step": 21331
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8736254835757762,
+      "learning_rate": 1.0402776696794506e-05,
+      "loss": 0.7482,
+      "step": 21332
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.420608561911788,
+      "learning_rate": 1.040201427375079e-05,
+      "loss": 0.7224,
+      "step": 21333
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6969623441836168,
+      "learning_rate": 1.0401251848366429e-05,
+      "loss": 0.707,
+      "step": 21334
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7774600168139563,
+      "learning_rate": 1.0400489420645848e-05,
+      "loss": 0.6772,
+      "step": 21335
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8652021476370024,
+      "learning_rate": 1.0399726990593493e-05,
+      "loss": 0.785,
+      "step": 21336
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6143785379238331,
+      "learning_rate": 1.0398964558213796e-05,
+      "loss": 0.7131,
+      "step": 21337
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6388605850005233,
+      "learning_rate": 1.0398202123511207e-05,
+      "loss": 0.5176,
+      "step": 21338
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5988810782361866,
+      "learning_rate": 1.0397439686490154e-05,
+      "loss": 0.6761,
+      "step": 21339
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4782019234091714,
+      "learning_rate": 1.0396677247155087e-05,
+      "loss": 0.7786,
+      "step": 21340
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.967286287823662,
+      "learning_rate": 1.0395914805510437e-05,
+      "loss": 0.6411,
+      "step": 21341
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5827544883549589,
+      "learning_rate": 1.0395152361560645e-05,
+      "loss": 0.5401,
+      "step": 21342
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4483005272246483,
+      "learning_rate": 1.039438991531015e-05,
+      "loss": 0.6827,
+      "step": 21343
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2414489922590006,
+      "learning_rate": 1.0393627466763395e-05,
+      "loss": 0.7863,
+      "step": 21344
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.714234691191003,
+      "learning_rate": 1.0392865015924813e-05,
+      "loss": 0.6058,
+      "step": 21345
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2997786144998695,
+      "learning_rate": 1.0392102562798847e-05,
+      "loss": 0.6706,
+      "step": 21346
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4049320203844053,
+      "learning_rate": 1.0391340107389938e-05,
+      "loss": 0.761,
+      "step": 21347
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6305859651845347,
+      "learning_rate": 1.039057764970252e-05,
+      "loss": 0.5097,
+      "step": 21348
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3961840925495719,
+      "learning_rate": 1.0389815189741038e-05,
+      "loss": 0.7313,
+      "step": 21349
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4419766364097206,
+      "learning_rate": 1.0389052727509928e-05,
+      "loss": 0.7222,
+      "step": 21350
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.301939275449132,
+      "learning_rate": 1.0388290263013628e-05,
+      "loss": 0.8517,
+      "step": 21351
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6069885244222792,
+      "learning_rate": 1.038752779625658e-05,
+      "loss": 0.7583,
+      "step": 21352
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.319560131811961,
+      "learning_rate": 1.0386765327243222e-05,
+      "loss": 0.7001,
+      "step": 21353
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4564832991821552,
+      "learning_rate": 1.0386002855977995e-05,
+      "loss": 0.7898,
+      "step": 21354
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4037948873306874,
+      "learning_rate": 1.038524038246534e-05,
+      "loss": 0.7248,
+      "step": 21355
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6246911249344745,
+      "learning_rate": 1.0384477906709687e-05,
+      "loss": 0.5128,
+      "step": 21356
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5859542915302536,
+      "learning_rate": 1.0383715428715489e-05,
+      "loss": 0.4859,
+      "step": 21357
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3846478370299964,
+      "learning_rate": 1.038295294848717e-05,
+      "loss": 0.7072,
+      "step": 21358
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5056183535107623,
+      "learning_rate": 1.0382190466029186e-05,
+      "loss": 0.7924,
+      "step": 21359
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.559420524580224,
+      "learning_rate": 1.0381427981345961e-05,
+      "loss": 0.5153,
+      "step": 21360
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4821067713584972,
+      "learning_rate": 1.0380665494441949e-05,
+      "loss": 0.7239,
+      "step": 21361
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.37376614617583,
+      "learning_rate": 1.0379903005321576e-05,
+      "loss": 0.6301,
+      "step": 21362
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2751562155065228,
+      "learning_rate": 1.037914051398929e-05,
+      "loss": 0.7533,
+      "step": 21363
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5040409822962775,
+      "learning_rate": 1.0378378020449527e-05,
+      "loss": 0.7246,
+      "step": 21364
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3992369443875572,
+      "learning_rate": 1.0377615524706726e-05,
+      "loss": 0.7157,
+      "step": 21365
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7452204191661036,
+      "learning_rate": 1.0376853026765332e-05,
+      "loss": 0.8531,
+      "step": 21366
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.83080114662849,
+      "learning_rate": 1.0376090526629773e-05,
+      "loss": 0.692,
+      "step": 21367
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4197957034863942,
+      "learning_rate": 1.0375328024304505e-05,
+      "loss": 0.7205,
+      "step": 21368
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6016378623942613,
+      "learning_rate": 1.0374565519793953e-05,
+      "loss": 0.6872,
+      "step": 21369
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5738076488090889,
+      "learning_rate": 1.0373803013102563e-05,
+      "loss": 0.5207,
+      "step": 21370
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9195464152642454,
+      "learning_rate": 1.0373040504234775e-05,
+      "loss": 0.6513,
+      "step": 21371
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.563827943474739,
+      "learning_rate": 1.0372277993195026e-05,
+      "loss": 0.797,
+      "step": 21372
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6055106642620827,
+      "learning_rate": 1.0371515479987755e-05,
+      "loss": 0.7347,
+      "step": 21373
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4840865606869982,
+      "learning_rate": 1.0370752964617405e-05,
+      "loss": 0.6994,
+      "step": 21374
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3461113545130285,
+      "learning_rate": 1.0369990447088415e-05,
+      "loss": 0.869,
+      "step": 21375
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5887553413406206,
+      "learning_rate": 1.0369227927405223e-05,
+      "loss": 0.536,
+      "step": 21376
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5441384269365541,
+      "learning_rate": 1.0368465405572265e-05,
+      "loss": 0.6723,
+      "step": 21377
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4198479463240854,
+      "learning_rate": 1.0367702881593988e-05,
+      "loss": 0.8115,
+      "step": 21378
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6210569593491303,
+      "learning_rate": 1.0366940355474829e-05,
+      "loss": 0.549,
+      "step": 21379
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6261497815374493,
+      "learning_rate": 1.0366177827219227e-05,
+      "loss": 0.525,
+      "step": 21380
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6211966515893307,
+      "learning_rate": 1.036541529683162e-05,
+      "loss": 0.479,
+      "step": 21381
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.682095752738499,
+      "learning_rate": 1.036465276431645e-05,
+      "loss": 0.7168,
+      "step": 21382
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4378652878976204,
+      "learning_rate": 1.036389022967816e-05,
+      "loss": 0.7649,
+      "step": 21383
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5056654961506866,
+      "learning_rate": 1.036312769292118e-05,
+      "loss": 0.7693,
+      "step": 21384
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3366321826750094,
+      "learning_rate": 1.0362365154049959e-05,
+      "loss": 0.744,
+      "step": 21385
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2240474651657027,
+      "learning_rate": 1.036160261306893e-05,
+      "loss": 0.7373,
+      "step": 21386
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5576807617592654,
+      "learning_rate": 1.036084006998254e-05,
+      "loss": 0.8077,
+      "step": 21387
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5957998656819948,
+      "learning_rate": 1.0360077524795222e-05,
+      "loss": 0.4999,
+      "step": 21388
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.354196516943161,
+      "learning_rate": 1.0359314977511423e-05,
+      "loss": 0.8188,
+      "step": 21389
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5987488807950903,
+      "learning_rate": 1.0358552428135576e-05,
+      "loss": 0.8386,
+      "step": 21390
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5993670474095902,
+      "learning_rate": 1.0357789876672122e-05,
+      "loss": 0.5742,
+      "step": 21391
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6004932862362045,
+      "learning_rate": 1.0357027323125503e-05,
+      "loss": 0.4733,
+      "step": 21392
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5730229961009219,
+      "learning_rate": 1.0356264767500158e-05,
+      "loss": 0.5251,
+      "step": 21393
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.404100908819732,
+      "learning_rate": 1.0355502209800525e-05,
+      "loss": 0.8466,
+      "step": 21394
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3584733800677062,
+      "learning_rate": 1.0354739650031047e-05,
+      "loss": 0.7384,
+      "step": 21395
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.406592973396577,
+      "learning_rate": 1.0353977088196164e-05,
+      "loss": 0.6726,
+      "step": 21396
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.451506838551615,
+      "learning_rate": 1.0353214524300312e-05,
+      "loss": 0.823,
+      "step": 21397
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5181897222544942,
+      "learning_rate": 1.0352451958347936e-05,
+      "loss": 0.7637,
+      "step": 21398
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5978408153335015,
+      "learning_rate": 1.035168939034347e-05,
+      "loss": 0.5114,
+      "step": 21399
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6721684925251683,
+      "learning_rate": 1.0350926820291359e-05,
+      "loss": 0.8449,
+      "step": 21400
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.341418476649921,
+      "learning_rate": 1.035016424819604e-05,
+      "loss": 0.7457,
+      "step": 21401
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.031065441013514,
+      "learning_rate": 1.0349401674061952e-05,
+      "loss": 0.7958,
+      "step": 21402
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4747057155276981,
+      "learning_rate": 1.034863909789354e-05,
+      "loss": 0.6698,
+      "step": 21403
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2958306529837924,
+      "learning_rate": 1.034787651969524e-05,
+      "loss": 0.8083,
+      "step": 21404
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5326297946102319,
+      "learning_rate": 1.0347113939471491e-05,
+      "loss": 0.6746,
+      "step": 21405
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6419173670394547,
+      "learning_rate": 1.0346351357226737e-05,
+      "loss": 0.6193,
+      "step": 21406
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5950848119667962,
+      "learning_rate": 1.0345588772965413e-05,
+      "loss": 0.6447,
+      "step": 21407
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.311148655403095,
+      "learning_rate": 1.0344826186691968e-05,
+      "loss": 0.7457,
+      "step": 21408
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7915124475394952,
+      "learning_rate": 1.0344063598410829e-05,
+      "loss": 0.6931,
+      "step": 21409
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3456060176035938,
+      "learning_rate": 1.0343301008126447e-05,
+      "loss": 0.7713,
+      "step": 21410
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1635666908088216,
+      "learning_rate": 1.0342538415843255e-05,
+      "loss": 0.8639,
+      "step": 21411
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5774745258698566,
+      "learning_rate": 1.0341775821565699e-05,
+      "loss": 0.6698,
+      "step": 21412
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4001059977270212,
+      "learning_rate": 1.0341013225298213e-05,
+      "loss": 0.7276,
+      "step": 21413
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9579235162286426,
+      "learning_rate": 1.0340250627045238e-05,
+      "loss": 0.6737,
+      "step": 21414
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6825984900451925,
+      "learning_rate": 1.0339488026811222e-05,
+      "loss": 0.6773,
+      "step": 21415
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3139324104712424,
+      "learning_rate": 1.0338725424600596e-05,
+      "loss": 0.7283,
+      "step": 21416
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5199063265403874,
+      "learning_rate": 1.0337962820417804e-05,
+      "loss": 0.8559,
+      "step": 21417
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.223853711492162,
+      "learning_rate": 1.0337200214267282e-05,
+      "loss": 0.6849,
+      "step": 21418
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6488402645910198,
+      "learning_rate": 1.033643760615348e-05,
+      "loss": 0.6384,
+      "step": 21419
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6820867757580296,
+      "learning_rate": 1.033567499608083e-05,
+      "loss": 0.5814,
+      "step": 21420
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3485287758414837,
+      "learning_rate": 1.033491238405377e-05,
+      "loss": 0.6811,
+      "step": 21421
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3201420533628303,
+      "learning_rate": 1.0334149770076747e-05,
+      "loss": 0.7416,
+      "step": 21422
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9989220135619867,
+      "learning_rate": 1.0333387154154197e-05,
+      "loss": 0.7309,
+      "step": 21423
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6018648491568592,
+      "learning_rate": 1.0332624536290563e-05,
+      "loss": 0.5421,
+      "step": 21424
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.650412131450101,
+      "learning_rate": 1.0331861916490283e-05,
+      "loss": 0.5708,
+      "step": 21425
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.391657151379336,
+      "learning_rate": 1.0331099294757798e-05,
+      "loss": 0.7513,
+      "step": 21426
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.45558192791856,
+      "learning_rate": 1.0330336671097547e-05,
+      "loss": 0.8062,
+      "step": 21427
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5079458302937723,
+      "learning_rate": 1.0329574045513973e-05,
+      "loss": 0.9028,
+      "step": 21428
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.273643421809284,
+      "learning_rate": 1.0328811418011514e-05,
+      "loss": 0.6809,
+      "step": 21429
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6136842283645172,
+      "learning_rate": 1.0328048788594609e-05,
+      "loss": 0.694,
+      "step": 21430
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.944229792449249,
+      "learning_rate": 1.0327286157267701e-05,
+      "loss": 0.6695,
+      "step": 21431
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.436129830993214,
+      "learning_rate": 1.032652352403523e-05,
+      "loss": 0.7039,
+      "step": 21432
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.559931643423762,
+      "learning_rate": 1.0325760888901638e-05,
+      "loss": 0.5945,
+      "step": 21433
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6157414497543641,
+      "learning_rate": 1.032499825187136e-05,
+      "loss": 0.5356,
+      "step": 21434
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3857256350358702,
+      "learning_rate": 1.0324235612948839e-05,
+      "loss": 0.8055,
+      "step": 21435
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.267434926981166,
+      "learning_rate": 1.0323472972138518e-05,
+      "loss": 0.7053,
+      "step": 21436
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3868285078615739,
+      "learning_rate": 1.0322710329444831e-05,
+      "loss": 0.8338,
+      "step": 21437
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3270219178257205,
+      "learning_rate": 1.0321947684872227e-05,
+      "loss": 0.7905,
+      "step": 21438
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.391870427844645,
+      "learning_rate": 1.0321185038425137e-05,
+      "loss": 0.7417,
+      "step": 21439
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5670982735936283,
+      "learning_rate": 1.032042239010801e-05,
+      "loss": 0.7246,
+      "step": 21440
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1345407964993273,
+      "learning_rate": 1.031965973992528e-05,
+      "loss": 0.7005,
+      "step": 21441
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4123521967183568,
+      "learning_rate": 1.0318897087881392e-05,
+      "loss": 0.7664,
+      "step": 21442
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6600047024810307,
+      "learning_rate": 1.0318134433980781e-05,
+      "loss": 0.6817,
+      "step": 21443
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9583315117679096,
+      "learning_rate": 1.0317371778227893e-05,
+      "loss": 0.763,
+      "step": 21444
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4613553855729087,
+      "learning_rate": 1.0316609120627165e-05,
+      "loss": 0.7576,
+      "step": 21445
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3643484061948101,
+      "learning_rate": 1.0315846461183039e-05,
+      "loss": 0.6659,
+      "step": 21446
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4778925712771684,
+      "learning_rate": 1.0315083799899957e-05,
+      "loss": 0.652,
+      "step": 21447
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5965063601527378,
+      "learning_rate": 1.0314321136782354e-05,
+      "loss": 0.6072,
+      "step": 21448
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4171029087531126,
+      "learning_rate": 1.0313558471834674e-05,
+      "loss": 0.8416,
+      "step": 21449
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6089999778792237,
+      "learning_rate": 1.031279580506136e-05,
+      "loss": 0.7352,
+      "step": 21450
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5715105306490849,
+      "learning_rate": 1.0312033136466848e-05,
+      "loss": 0.8139,
+      "step": 21451
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4201985567884243,
+      "learning_rate": 1.0311270466055578e-05,
+      "loss": 0.6918,
+      "step": 21452
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6062196562165911,
+      "learning_rate": 1.0310507793831997e-05,
+      "loss": 0.5189,
+      "step": 21453
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6214665438651878,
+      "learning_rate": 1.0309745119800539e-05,
+      "loss": 0.7037,
+      "step": 21454
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.273108525063799,
+      "learning_rate": 1.030898244396565e-05,
+      "loss": 0.7279,
+      "step": 21455
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.570829826532549,
+      "learning_rate": 1.0308219766331762e-05,
+      "loss": 0.5348,
+      "step": 21456
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3843408087217173,
+      "learning_rate": 1.0307457086903325e-05,
+      "loss": 0.6668,
+      "step": 21457
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5564029579339944,
+      "learning_rate": 1.0306694405684771e-05,
+      "loss": 0.5239,
+      "step": 21458
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4855124916329765,
+      "learning_rate": 1.030593172268055e-05,
+      "loss": 0.765,
+      "step": 21459
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.533567137495723,
+      "learning_rate": 1.0305169037895095e-05,
+      "loss": 0.7638,
+      "step": 21460
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4462013603797854,
+      "learning_rate": 1.0304406351332852e-05,
+      "loss": 0.8583,
+      "step": 21461
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5967257787055398,
+      "learning_rate": 1.0303643662998255e-05,
+      "loss": 0.6508,
+      "step": 21462
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4838756753941125,
+      "learning_rate": 1.0302880972895751e-05,
+      "loss": 0.6773,
+      "step": 21463
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.57866070741694,
+      "learning_rate": 1.030211828102978e-05,
+      "loss": 0.6949,
+      "step": 21464
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2679747024963337,
+      "learning_rate": 1.0301355587404775e-05,
+      "loss": 0.7227,
+      "step": 21465
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5273669406706976,
+      "learning_rate": 1.0300592892025187e-05,
+      "loss": 0.689,
+      "step": 21466
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5899340764609196,
+      "learning_rate": 1.029983019489545e-05,
+      "loss": 0.8305,
+      "step": 21467
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4881426351923188,
+      "learning_rate": 1.029906749602001e-05,
+      "loss": 0.676,
+      "step": 21468
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3294612911005208,
+      "learning_rate": 1.0298304795403298e-05,
+      "loss": 0.6466,
+      "step": 21469
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5923031877318707,
+      "learning_rate": 1.0297542093049767e-05,
+      "loss": 0.5354,
+      "step": 21470
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8813538580001936,
+      "learning_rate": 1.0296779388963849e-05,
+      "loss": 0.7112,
+      "step": 21471
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5845817094364925,
+      "learning_rate": 1.0296016683149988e-05,
+      "loss": 0.5564,
+      "step": 21472
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6379919975657133,
+      "learning_rate": 1.0295253975612624e-05,
+      "loss": 0.705,
+      "step": 21473
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4666336634914408,
+      "learning_rate": 1.02944912663562e-05,
+      "loss": 0.8019,
+      "step": 21474
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.406932264773895,
+      "learning_rate": 1.0293728555385149e-05,
+      "loss": 0.627,
+      "step": 21475
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6084504970081093,
+      "learning_rate": 1.0292965842703922e-05,
+      "loss": 0.6784,
+      "step": 21476
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5578493854045206,
+      "learning_rate": 1.0292203128316953e-05,
+      "loss": 0.7517,
+      "step": 21477
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.9535066097025218,
+      "learning_rate": 1.0291440412228688e-05,
+      "loss": 0.8166,
+      "step": 21478
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5957275709037543,
+      "learning_rate": 1.029067769444356e-05,
+      "loss": 0.5286,
+      "step": 21479
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4942078970616546,
+      "learning_rate": 1.0289914974966018e-05,
+      "loss": 0.7554,
+      "step": 21480
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5444040054702775,
+      "learning_rate": 1.0289152253800499e-05,
+      "loss": 0.8359,
+      "step": 21481
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.599297124304873,
+      "learning_rate": 1.0288389530951444e-05,
+      "loss": 0.6631,
+      "step": 21482
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8397500738302672,
+      "learning_rate": 1.0287626806423292e-05,
+      "loss": 0.7098,
+      "step": 21483
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.346311199656659,
+      "learning_rate": 1.0286864080220486e-05,
+      "loss": 0.6351,
+      "step": 21484
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.501727521770242,
+      "learning_rate": 1.028610135234747e-05,
+      "loss": 0.6217,
+      "step": 21485
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1649793442724536,
+      "learning_rate": 1.0285338622808676e-05,
+      "loss": 0.6766,
+      "step": 21486
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.448040546965883,
+      "learning_rate": 1.0284575891608556e-05,
+      "loss": 0.8614,
+      "step": 21487
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2656926751888133,
+      "learning_rate": 1.0283813158751539e-05,
+      "loss": 0.7811,
+      "step": 21488
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.579877071887819,
+      "learning_rate": 1.0283050424242076e-05,
+      "loss": 0.494,
+      "step": 21489
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4103144478279281,
+      "learning_rate": 1.02822876880846e-05,
+      "loss": 0.7813,
+      "step": 21490
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.469636125768325,
+      "learning_rate": 1.0281524950283562e-05,
+      "loss": 0.7122,
+      "step": 21491
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.61480033036232,
+      "learning_rate": 1.0280762210843391e-05,
+      "loss": 0.7939,
+      "step": 21492
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.020555966333855,
+      "learning_rate": 1.0279999469768535e-05,
+      "loss": 0.5569,
+      "step": 21493
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.293785437848977,
+      "learning_rate": 1.0279236727063436e-05,
+      "loss": 0.763,
+      "step": 21494
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5886133509414467,
+      "learning_rate": 1.0278473982732529e-05,
+      "loss": 0.5304,
+      "step": 21495
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5193468530113294,
+      "learning_rate": 1.0277711236780261e-05,
+      "loss": 0.7498,
+      "step": 21496
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6348093537137438,
+      "learning_rate": 1.0276948489211068e-05,
+      "loss": 0.5676,
+      "step": 21497
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3022381034010293,
+      "learning_rate": 1.0276185740029393e-05,
+      "loss": 0.6587,
+      "step": 21498
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.605344841156682,
+      "learning_rate": 1.027542298923968e-05,
+      "loss": 0.755,
+      "step": 21499
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3897991412433488,
+      "learning_rate": 1.0274660236846364e-05,
+      "loss": 0.7494,
+      "step": 21500
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5502806321633213,
+      "learning_rate": 1.0273897482853891e-05,
+      "loss": 0.679,
+      "step": 21501
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3836784004537044,
+      "learning_rate": 1.02731347272667e-05,
+      "loss": 0.6179,
+      "step": 21502
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4795506827059326,
+      "learning_rate": 1.0272371970089232e-05,
+      "loss": 0.7255,
+      "step": 21503
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3568567929972035,
+      "learning_rate": 1.0271609211325928e-05,
+      "loss": 0.6874,
+      "step": 21504
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2580183709923691,
+      "learning_rate": 1.0270846450981228e-05,
+      "loss": 0.8053,
+      "step": 21505
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.444648650608758,
+      "learning_rate": 1.0270083689059578e-05,
+      "loss": 0.8127,
+      "step": 21506
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3265737701281528,
+      "learning_rate": 1.0269320925565409e-05,
+      "loss": 0.6321,
+      "step": 21507
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2955034872574243,
+      "learning_rate": 1.0268558160503174e-05,
+      "loss": 0.6297,
+      "step": 21508
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4348111111331987,
+      "learning_rate": 1.0267795393877305e-05,
+      "loss": 0.7035,
+      "step": 21509
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6803865825699424,
+      "learning_rate": 1.026703262569225e-05,
+      "loss": 0.8094,
+      "step": 21510
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.547666229282248,
+      "learning_rate": 1.0266269855952442e-05,
+      "loss": 0.7098,
+      "step": 21511
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6128446416032756,
+      "learning_rate": 1.026550708466233e-05,
+      "loss": 0.7307,
+      "step": 21512
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.030340699173843,
+      "learning_rate": 1.026474431182635e-05,
+      "loss": 0.7361,
+      "step": 21513
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6959560911936427,
+      "learning_rate": 1.0263981537448944e-05,
+      "loss": 0.7837,
+      "step": 21514
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6162924615082278,
+      "learning_rate": 1.0263218761534557e-05,
+      "loss": 0.497,
+      "step": 21515
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2653677526102851,
+      "learning_rate": 1.0262455984087626e-05,
+      "loss": 0.7188,
+      "step": 21516
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5226451327535278,
+      "learning_rate": 1.0261693205112592e-05,
+      "loss": 0.6837,
+      "step": 21517
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.188547128802628,
+      "learning_rate": 1.0260930424613895e-05,
+      "loss": 0.4968,
+      "step": 21518
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3352253131763914,
+      "learning_rate": 1.0260167642595984e-05,
+      "loss": 0.7064,
+      "step": 21519
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.587109440541269,
+      "learning_rate": 1.0259404859063288e-05,
+      "loss": 0.8678,
+      "step": 21520
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5060794879981196,
+      "learning_rate": 1.0258642074020261e-05,
+      "loss": 0.8627,
+      "step": 21521
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5252988938416427,
+      "learning_rate": 1.0257879287471337e-05,
+      "loss": 0.7373,
+      "step": 21522
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8232460397453993,
+      "learning_rate": 1.0257116499420955e-05,
+      "loss": 0.7274,
+      "step": 21523
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4507253442668802,
+      "learning_rate": 1.0256353709873563e-05,
+      "loss": 0.6655,
+      "step": 21524
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6018941752080025,
+      "learning_rate": 1.0255590918833596e-05,
+      "loss": 0.6993,
+      "step": 21525
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.450095750541059,
+      "learning_rate": 1.0254828126305498e-05,
+      "loss": 0.8497,
+      "step": 21526
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.086745289474919,
+      "learning_rate": 1.025406533229371e-05,
+      "loss": 0.6723,
+      "step": 21527
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4687512555213709,
+      "learning_rate": 1.0253302536802674e-05,
+      "loss": 0.7294,
+      "step": 21528
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5868568273273076,
+      "learning_rate": 1.025253973983683e-05,
+      "loss": 0.8516,
+      "step": 21529
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2294810952575512,
+      "learning_rate": 1.025177694140062e-05,
+      "loss": 0.6441,
+      "step": 21530
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.434006051164057,
+      "learning_rate": 1.0251014141498486e-05,
+      "loss": 0.7651,
+      "step": 21531
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.036818756138831,
+      "learning_rate": 1.0250251340134865e-05,
+      "loss": 0.7301,
+      "step": 21532
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.624808468224883,
+      "learning_rate": 1.0249488537314203e-05,
+      "loss": 0.8866,
+      "step": 21533
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4897945075832972,
+      "learning_rate": 1.0248725733040941e-05,
+      "loss": 0.7016,
+      "step": 21534
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3609029925290996,
+      "learning_rate": 1.0247962927319518e-05,
+      "loss": 0.7173,
+      "step": 21535
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6303517225002402,
+      "learning_rate": 1.0247200120154377e-05,
+      "loss": 0.6097,
+      "step": 21536
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5795845357862466,
+      "learning_rate": 1.0246437311549957e-05,
+      "loss": 0.6697,
+      "step": 21537
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3810519285220662,
+      "learning_rate": 1.0245674501510703e-05,
+      "loss": 0.7837,
+      "step": 21538
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2866958154068768,
+      "learning_rate": 1.0244911690041051e-05,
+      "loss": 0.7752,
+      "step": 21539
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.386713747429807,
+      "learning_rate": 1.024414887714545e-05,
+      "loss": 0.8217,
+      "step": 21540
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.481887887594734,
+      "learning_rate": 1.0243386062828333e-05,
+      "loss": 0.5873,
+      "step": 21541
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5461249622968252,
+      "learning_rate": 1.024262324709415e-05,
+      "loss": 0.9114,
+      "step": 21542
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6554321354807112,
+      "learning_rate": 1.0241860429947334e-05,
+      "loss": 0.59,
+      "step": 21543
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5916600700987769,
+      "learning_rate": 1.024109761139233e-05,
+      "loss": 0.6612,
+      "step": 21544
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4646412456629239,
+      "learning_rate": 1.0240334791433581e-05,
+      "loss": 0.7271,
+      "step": 21545
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.708167041169799,
+      "learning_rate": 1.0239571970075526e-05,
+      "loss": 0.7532,
+      "step": 21546
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3599473906154689,
+      "learning_rate": 1.0238809147322606e-05,
+      "loss": 0.7181,
+      "step": 21547
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5045578934591723,
+      "learning_rate": 1.0238046323179265e-05,
+      "loss": 0.8099,
+      "step": 21548
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.985850399481769,
+      "learning_rate": 1.0237283497649942e-05,
+      "loss": 0.7806,
+      "step": 21549
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8056736701613147,
+      "learning_rate": 1.0236520670739081e-05,
+      "loss": 0.6532,
+      "step": 21550
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8359464058951713,
+      "learning_rate": 1.023575784245112e-05,
+      "loss": 0.635,
+      "step": 21551
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.787325405272163,
+      "learning_rate": 1.0234995012790503e-05,
+      "loss": 0.6528,
+      "step": 21552
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3315291577226354,
+      "learning_rate": 1.023423218176167e-05,
+      "loss": 0.6696,
+      "step": 21553
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.314484277323918,
+      "learning_rate": 1.0233469349369062e-05,
+      "loss": 0.6868,
+      "step": 21554
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.399409776865406,
+      "learning_rate": 1.0232706515617123e-05,
+      "loss": 0.6956,
+      "step": 21555
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3451230571028376,
+      "learning_rate": 1.0231943680510295e-05,
+      "loss": 0.7343,
+      "step": 21556
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.592875962321044,
+      "learning_rate": 1.0231180844053017e-05,
+      "loss": 0.5735,
+      "step": 21557
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.490404942858838,
+      "learning_rate": 1.0230418006249728e-05,
+      "loss": 0.835,
+      "step": 21558
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.329778450786939,
+      "learning_rate": 1.0229655167104876e-05,
+      "loss": 0.7858,
+      "step": 21559
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3803051168946645,
+      "learning_rate": 1.0228892326622892e-05,
+      "loss": 0.7048,
+      "step": 21560
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4330522034553577,
+      "learning_rate": 1.022812948480823e-05,
+      "loss": 0.6198,
+      "step": 21561
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.509228631352382,
+      "learning_rate": 1.0227366641665327e-05,
+      "loss": 0.7701,
+      "step": 21562
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.538196789137755,
+      "learning_rate": 1.022660379719862e-05,
+      "loss": 0.6309,
+      "step": 21563
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.615698600490102,
+      "learning_rate": 1.0225840951412557e-05,
+      "loss": 0.7328,
+      "step": 21564
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.738349716180619,
+      "learning_rate": 1.0225078104311572e-05,
+      "loss": 0.68,
+      "step": 21565
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4404659045745876,
+      "learning_rate": 1.0224315255900116e-05,
+      "loss": 0.8394,
+      "step": 21566
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5163537805172196,
+      "learning_rate": 1.0223552406182622e-05,
+      "loss": 0.5124,
+      "step": 21567
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3325589127826196,
+      "learning_rate": 1.0222789555163538e-05,
+      "loss": 0.8736,
+      "step": 21568
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3809922627676403,
+      "learning_rate": 1.02220267028473e-05,
+      "loss": 0.7298,
+      "step": 21569
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4213721237478376,
+      "learning_rate": 1.0221263849238357e-05,
+      "loss": 0.8922,
+      "step": 21570
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5570706840053118,
+      "learning_rate": 1.022050099434114e-05,
+      "loss": 0.5384,
+      "step": 21571
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5607337783785764,
+      "learning_rate": 1.02197381381601e-05,
+      "loss": 0.8312,
+      "step": 21572
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2466221355014144,
+      "learning_rate": 1.0218975280699675e-05,
+      "loss": 0.7316,
+      "step": 21573
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3485355709219837,
+      "learning_rate": 1.0218212421964304e-05,
+      "loss": 0.6475,
+      "step": 21574
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4003454774219581,
+      "learning_rate": 1.0217449561958432e-05,
+      "loss": 0.7749,
+      "step": 21575
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.55783201505835,
+      "learning_rate": 1.02166867006865e-05,
+      "loss": 0.7357,
+      "step": 21576
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6804376900443083,
+      "learning_rate": 1.021592383815295e-05,
+      "loss": 0.7107,
+      "step": 21577
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4561281555969112,
+      "learning_rate": 1.0215160974362224e-05,
+      "loss": 0.8071,
+      "step": 21578
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6013154322788768,
+      "learning_rate": 1.0214398109318761e-05,
+      "loss": 0.506,
+      "step": 21579
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4743411207974977,
+      "learning_rate": 1.0213635243027004e-05,
+      "loss": 0.6156,
+      "step": 21580
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5304196187544583,
+      "learning_rate": 1.0212872375491396e-05,
+      "loss": 0.7221,
+      "step": 21581
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4523932636734362,
+      "learning_rate": 1.021210950671638e-05,
+      "loss": 0.7506,
+      "step": 21582
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5546025900973006,
+      "learning_rate": 1.021134663670639e-05,
+      "loss": 0.7214,
+      "step": 21583
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5854624635700099,
+      "learning_rate": 1.0210583765465875e-05,
+      "loss": 0.4924,
+      "step": 21584
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3507898038737534,
+      "learning_rate": 1.0209820892999275e-05,
+      "loss": 0.7094,
+      "step": 21585
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5991915646776965,
+      "learning_rate": 1.0209058019311034e-05,
+      "loss": 0.5149,
+      "step": 21586
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.459675344112093,
+      "learning_rate": 1.0208295144405592e-05,
+      "loss": 0.7454,
+      "step": 21587
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4650572002384643,
+      "learning_rate": 1.0207532268287384e-05,
+      "loss": 0.8315,
+      "step": 21588
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5276776909515701,
+      "learning_rate": 1.020676939096086e-05,
+      "loss": 0.8779,
+      "step": 21589
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7906312363942567,
+      "learning_rate": 1.0206006512430457e-05,
+      "loss": 0.6551,
+      "step": 21590
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6261737391857407,
+      "learning_rate": 1.0205243632700622e-05,
+      "loss": 0.8141,
+      "step": 21591
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.355406863420643,
+      "learning_rate": 1.0204480751775792e-05,
+      "loss": 0.8217,
+      "step": 21592
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7866891384948322,
+      "learning_rate": 1.0203717869660413e-05,
+      "loss": 0.7168,
+      "step": 21593
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4173594916800005,
+      "learning_rate": 1.0202954986358921e-05,
+      "loss": 0.7356,
+      "step": 21594
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.545136925044818,
+      "learning_rate": 1.0202192101875763e-05,
+      "loss": 0.7386,
+      "step": 21595
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.432531321327721,
+      "learning_rate": 1.0201429216215378e-05,
+      "loss": 0.8137,
+      "step": 21596
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.803958243428134,
+      "learning_rate": 1.0200666329382205e-05,
+      "loss": 0.6376,
+      "step": 21597
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4651425173479669,
+      "learning_rate": 1.0199903441380692e-05,
+      "loss": 0.7671,
+      "step": 21598
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.479882661611428,
+      "learning_rate": 1.0199140552215276e-05,
+      "loss": 0.6623,
+      "step": 21599
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3686330894855478,
+      "learning_rate": 1.0198377661890404e-05,
+      "loss": 0.5999,
+      "step": 21600
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5992591062532542,
+      "learning_rate": 1.0197614770410513e-05,
+      "loss": 0.8507,
+      "step": 21601
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5689739835023765,
+      "learning_rate": 1.0196851877780045e-05,
+      "loss": 0.5339,
+      "step": 21602
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2393518010924498,
+      "learning_rate": 1.0196088984003444e-05,
+      "loss": 0.7506,
+      "step": 21603
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6241153321063088,
+      "learning_rate": 1.0195326089085152e-05,
+      "loss": 0.5171,
+      "step": 21604
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3741971399153705,
+      "learning_rate": 1.0194563193029608e-05,
+      "loss": 0.7605,
+      "step": 21605
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0725749619911724,
+      "learning_rate": 1.0193800295841257e-05,
+      "loss": 0.6928,
+      "step": 21606
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4538416792572157,
+      "learning_rate": 1.0193037397524538e-05,
+      "loss": 0.6691,
+      "step": 21607
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2855789018275505,
+      "learning_rate": 1.0192274498083895e-05,
+      "loss": 0.7258,
+      "step": 21608
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4192083655435435,
+      "learning_rate": 1.0191511597523767e-05,
+      "loss": 0.7198,
+      "step": 21609
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3766659469819096,
+      "learning_rate": 1.01907486958486e-05,
+      "loss": 0.5958,
+      "step": 21610
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3783497158992137,
+      "learning_rate": 1.0189985793062833e-05,
+      "loss": 0.7488,
+      "step": 21611
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6431662580005182,
+      "learning_rate": 1.0189222889170906e-05,
+      "loss": 0.6346,
+      "step": 21612
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4372676687308656,
+      "learning_rate": 1.0188459984177268e-05,
+      "loss": 0.8497,
+      "step": 21613
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3678941335328902,
+      "learning_rate": 1.0187697078086354e-05,
+      "loss": 0.6468,
+      "step": 21614
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5466987789711812,
+      "learning_rate": 1.0186934170902611e-05,
+      "loss": 0.6336,
+      "step": 21615
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5631248811476992,
+      "learning_rate": 1.0186171262630473e-05,
+      "loss": 0.7338,
+      "step": 21616
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9555890600360346,
+      "learning_rate": 1.018540835327439e-05,
+      "loss": 0.6482,
+      "step": 21617
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7058395536847246,
+      "learning_rate": 1.0184645442838802e-05,
+      "loss": 0.6755,
+      "step": 21618
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3345866193466676,
+      "learning_rate": 1.0183882531328148e-05,
+      "loss": 0.6931,
+      "step": 21619
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3419964010074026,
+      "learning_rate": 1.0183119618746871e-05,
+      "loss": 0.6717,
+      "step": 21620
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7769744052238265,
+      "learning_rate": 1.0182356705099416e-05,
+      "loss": 0.7452,
+      "step": 21621
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5411618532799012,
+      "learning_rate": 1.0181593790390221e-05,
+      "loss": 0.721,
+      "step": 21622
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.434134824366844,
+      "learning_rate": 1.0180830874623729e-05,
+      "loss": 0.7415,
+      "step": 21623
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2865255552674832,
+      "learning_rate": 1.0180067957804383e-05,
+      "loss": 0.7253,
+      "step": 21624
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5353448984561282,
+      "learning_rate": 1.0179305039936624e-05,
+      "loss": 0.8559,
+      "step": 21625
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.364923000289633,
+      "learning_rate": 1.0178542121024895e-05,
+      "loss": 0.6837,
+      "step": 21626
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7040710014814364,
+      "learning_rate": 1.0177779201073636e-05,
+      "loss": 0.5828,
+      "step": 21627
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9101123325277511,
+      "learning_rate": 1.017701628008729e-05,
+      "loss": 0.6695,
+      "step": 21628
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9733657240565725,
+      "learning_rate": 1.0176253358070301e-05,
+      "loss": 0.7578,
+      "step": 21629
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6010512369317849,
+      "learning_rate": 1.017549043502711e-05,
+      "loss": 0.7288,
+      "step": 21630
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8075761625276239,
+      "learning_rate": 1.0174727510962154e-05,
+      "loss": 0.6802,
+      "step": 21631
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4963434297330285,
+      "learning_rate": 1.0173964585879884e-05,
+      "loss": 0.6988,
+      "step": 21632
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2834746424368555,
+      "learning_rate": 1.0173201659784732e-05,
+      "loss": 0.7091,
+      "step": 21633
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2552609093345493,
+      "learning_rate": 1.017243873268115e-05,
+      "loss": 0.6879,
+      "step": 21634
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4772774502722321,
+      "learning_rate": 1.017167580457357e-05,
+      "loss": 0.8562,
+      "step": 21635
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9008878876520217,
+      "learning_rate": 1.0170912875466445e-05,
+      "loss": 0.7372,
+      "step": 21636
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5733682876482846,
+      "learning_rate": 1.0170149945364204e-05,
+      "loss": 0.8627,
+      "step": 21637
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.58665696085933,
+      "learning_rate": 1.0169387014271303e-05,
+      "loss": 0.5091,
+      "step": 21638
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7126215370112245,
+      "learning_rate": 1.0168624082192173e-05,
+      "loss": 0.6526,
+      "step": 21639
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9606782279602173,
+      "learning_rate": 1.0167861149131265e-05,
+      "loss": 0.7256,
+      "step": 21640
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0134963319731187,
+      "learning_rate": 1.016709821509301e-05,
+      "loss": 0.6388,
+      "step": 21641
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5248650294212025,
+      "learning_rate": 1.0166335280081861e-05,
+      "loss": 0.8423,
+      "step": 21642
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5530658093884453,
+      "learning_rate": 1.0165572344102252e-05,
+      "loss": 0.8308,
+      "step": 21643
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4529536468721225,
+      "learning_rate": 1.0164809407158632e-05,
+      "loss": 0.7362,
+      "step": 21644
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5150859112581256,
+      "learning_rate": 1.0164046469255437e-05,
+      "loss": 0.7371,
+      "step": 21645
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4847650683539058,
+      "learning_rate": 1.0163283530397112e-05,
+      "loss": 0.6587,
+      "step": 21646
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4128922966798265,
+      "learning_rate": 1.01625205905881e-05,
+      "loss": 0.7307,
+      "step": 21647
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5273838717641848,
+      "learning_rate": 1.0161757649832836e-05,
+      "loss": 0.7691,
+      "step": 21648
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6058296879751811,
+      "learning_rate": 1.0160994708135775e-05,
+      "loss": 0.5189,
+      "step": 21649
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3069098063953772,
+      "learning_rate": 1.0160231765501345e-05,
+      "loss": 0.6717,
+      "step": 21650
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2641003739907628,
+      "learning_rate": 1.0159468821934001e-05,
+      "loss": 0.666,
+      "step": 21651
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.334036752535096,
+      "learning_rate": 1.0158705877438178e-05,
+      "loss": 0.7364,
+      "step": 21652
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5928625449380998,
+      "learning_rate": 1.0157942932018317e-05,
+      "loss": 0.7019,
+      "step": 21653
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.826894639495121,
+      "learning_rate": 1.0157179985678862e-05,
+      "loss": 0.7144,
+      "step": 21654
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.551306017690706,
+      "learning_rate": 1.0156417038424259e-05,
+      "loss": 0.8308,
+      "step": 21655
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4185349788008828,
+      "learning_rate": 1.0155654090258942e-05,
+      "loss": 0.7484,
+      "step": 21656
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0001577451913026,
+      "learning_rate": 1.015489114118736e-05,
+      "loss": 0.6851,
+      "step": 21657
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0553820373970204,
+      "learning_rate": 1.0154128191213952e-05,
+      "loss": 0.641,
+      "step": 21658
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.3110021509717833,
+      "learning_rate": 1.015336524034316e-05,
+      "loss": 0.7458,
+      "step": 21659
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3205834021230696,
+      "learning_rate": 1.015260228857943e-05,
+      "loss": 0.7473,
+      "step": 21660
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.658752491946391,
+      "learning_rate": 1.0151839335927199e-05,
+      "loss": 0.776,
+      "step": 21661
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.689159764642235,
+      "learning_rate": 1.015107638239091e-05,
+      "loss": 0.6605,
+      "step": 21662
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6049560835556633,
+      "learning_rate": 1.015031342797501e-05,
+      "loss": 0.8162,
+      "step": 21663
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4252168253047643,
+      "learning_rate": 1.0149550472683933e-05,
+      "loss": 0.8014,
+      "step": 21664
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4355429620815168,
+      "learning_rate": 1.0148787516522129e-05,
+      "loss": 0.8185,
+      "step": 21665
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.410088059079399,
+      "learning_rate": 1.0148024559494038e-05,
+      "loss": 0.7354,
+      "step": 21666
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3965065865671769,
+      "learning_rate": 1.0147261601604099e-05,
+      "loss": 0.7454,
+      "step": 21667
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5006233647229958,
+      "learning_rate": 1.014649864285676e-05,
+      "loss": 0.7433,
+      "step": 21668
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2978318069529904,
+      "learning_rate": 1.0145735683256453e-05,
+      "loss": 0.7354,
+      "step": 21669
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.47523123398713,
+      "learning_rate": 1.0144972722807633e-05,
+      "loss": 0.6569,
+      "step": 21670
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4266438757324373,
+      "learning_rate": 1.0144209761514729e-05,
+      "loss": 0.7238,
+      "step": 21671
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6626872747396115,
+      "learning_rate": 1.0143446799382197e-05,
+      "loss": 0.6013,
+      "step": 21672
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.385014597354206,
+      "learning_rate": 1.0142683836414472e-05,
+      "loss": 0.7231,
+      "step": 21673
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3663969011132457,
+      "learning_rate": 1.0141920872615992e-05,
+      "loss": 0.7042,
+      "step": 21674
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 10.634271718656292,
+      "learning_rate": 1.0141157907991206e-05,
+      "loss": 0.7121,
+      "step": 21675
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4867778689791966,
+      "learning_rate": 1.0140394942544554e-05,
+      "loss": 0.7129,
+      "step": 21676
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4872529147989184,
+      "learning_rate": 1.0139631976280477e-05,
+      "loss": 0.7155,
+      "step": 21677
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4325576353104046,
+      "learning_rate": 1.013886900920342e-05,
+      "loss": 0.6265,
+      "step": 21678
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5873413009627698,
+      "learning_rate": 1.0138106041317823e-05,
+      "loss": 0.529,
+      "step": 21679
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5519417502066133,
+      "learning_rate": 1.013734307262813e-05,
+      "loss": 0.5793,
+      "step": 21680
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.680483637928396,
+      "learning_rate": 1.013658010313878e-05,
+      "loss": 0.6233,
+      "step": 21681
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8898913251026679,
+      "learning_rate": 1.0135817132854218e-05,
+      "loss": 0.6717,
+      "step": 21682
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.26604776904307,
+      "learning_rate": 1.0135054161778886e-05,
+      "loss": 0.6916,
+      "step": 21683
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6028437973234806,
+      "learning_rate": 1.0134291189917227e-05,
+      "loss": 0.7136,
+      "step": 21684
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3225105870839216,
+      "learning_rate": 1.013352821727368e-05,
+      "loss": 0.7775,
+      "step": 21685
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3920494012542601,
+      "learning_rate": 1.0132765243852691e-05,
+      "loss": 0.7656,
+      "step": 21686
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2951115129797677,
+      "learning_rate": 1.0132002269658703e-05,
+      "loss": 0.7507,
+      "step": 21687
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5477098012854096,
+      "learning_rate": 1.0131239294696151e-05,
+      "loss": 0.7636,
+      "step": 21688
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3928583560884855,
+      "learning_rate": 1.0130476318969487e-05,
+      "loss": 0.7346,
+      "step": 21689
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6180753165838532,
+      "learning_rate": 1.0129713342483145e-05,
+      "loss": 0.5611,
+      "step": 21690
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3630613781034093,
+      "learning_rate": 1.0128950365241575e-05,
+      "loss": 0.7358,
+      "step": 21691
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4884298484203946,
+      "learning_rate": 1.0128187387249208e-05,
+      "loss": 0.6697,
+      "step": 21692
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6960848510086806,
+      "learning_rate": 1.01274244085105e-05,
+      "loss": 0.7422,
+      "step": 21693
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2651853358707212,
+      "learning_rate": 1.0126661429029885e-05,
+      "loss": 0.7163,
+      "step": 21694
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4045533671388006,
+      "learning_rate": 1.0125898448811803e-05,
+      "loss": 0.5741,
+      "step": 21695
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.57657945100866,
+      "learning_rate": 1.0125135467860706e-05,
+      "loss": 0.8115,
+      "step": 21696
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.465221158682219,
+      "learning_rate": 1.0124372486181027e-05,
+      "loss": 0.6574,
+      "step": 21697
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3043262976950558,
+      "learning_rate": 1.0123609503777216e-05,
+      "loss": 0.7606,
+      "step": 21698
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3776247312822691,
+      "learning_rate": 1.0122846520653707e-05,
+      "loss": 0.6932,
+      "step": 21699
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.644160491189509,
+      "learning_rate": 1.0122083536814951e-05,
+      "loss": 0.69,
+      "step": 21700
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.091645542592106,
+      "learning_rate": 1.0121320552265384e-05,
+      "loss": 0.8217,
+      "step": 21701
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7053745348206641,
+      "learning_rate": 1.0120557567009449e-05,
+      "loss": 0.7526,
+      "step": 21702
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.316200373740301,
+      "learning_rate": 1.011979458105159e-05,
+      "loss": 0.7689,
+      "step": 21703
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4134973958697472,
+      "learning_rate": 1.0119031594396249e-05,
+      "loss": 0.7288,
+      "step": 21704
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3579718538537553,
+      "learning_rate": 1.011826860704787e-05,
+      "loss": 0.6406,
+      "step": 21705
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5055651834252464,
+      "learning_rate": 1.0117505619010891e-05,
+      "loss": 0.7122,
+      "step": 21706
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.130204133596837,
+      "learning_rate": 1.011674263028976e-05,
+      "loss": 0.6993,
+      "step": 21707
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.54204528859587,
+      "learning_rate": 1.0115979640888916e-05,
+      "loss": 0.684,
+      "step": 21708
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7375309282738769,
+      "learning_rate": 1.01152166508128e-05,
+      "loss": 0.5631,
+      "step": 21709
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.402370614606216,
+      "learning_rate": 1.0114453660065856e-05,
+      "loss": 0.7339,
+      "step": 21710
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2040848783530602,
+      "learning_rate": 1.0113690668652529e-05,
+      "loss": 0.6413,
+      "step": 21711
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4863785597482309,
+      "learning_rate": 1.0112927676577258e-05,
+      "loss": 0.8657,
+      "step": 21712
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.653072453090286,
+      "learning_rate": 1.0112164683844487e-05,
+      "loss": 0.7428,
+      "step": 21713
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5662753319712093,
+      "learning_rate": 1.0111401690458655e-05,
+      "loss": 0.5604,
+      "step": 21714
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5835519312532336,
+      "learning_rate": 1.0110638696424211e-05,
+      "loss": 0.7184,
+      "step": 21715
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6541689017542187,
+      "learning_rate": 1.010987570174559e-05,
+      "loss": 0.652,
+      "step": 21716
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.528962331575243,
+      "learning_rate": 1.0109112706427242e-05,
+      "loss": 0.5912,
+      "step": 21717
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4093838069767082,
+      "learning_rate": 1.01083497104736e-05,
+      "loss": 0.8221,
+      "step": 21718
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3897969616203378,
+      "learning_rate": 1.0107586713889117e-05,
+      "loss": 0.8173,
+      "step": 21719
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6628677005490635,
+      "learning_rate": 1.0106823716678225e-05,
+      "loss": 0.7569,
+      "step": 21720
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.466379194625394,
+      "learning_rate": 1.0106060718845377e-05,
+      "loss": 0.7,
+      "step": 21721
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1801621269645377,
+      "learning_rate": 1.0105297720395003e-05,
+      "loss": 0.6765,
+      "step": 21722
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3427157663510152,
+      "learning_rate": 1.0104534721331558e-05,
+      "loss": 0.6932,
+      "step": 21723
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.854162725151865,
+      "learning_rate": 1.0103771721659476e-05,
+      "loss": 0.6262,
+      "step": 21724
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0102315865682043,
+      "learning_rate": 1.0103008721383202e-05,
+      "loss": 0.6325,
+      "step": 21725
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4110021752326694,
+      "learning_rate": 1.010224572050718e-05,
+      "loss": 0.7481,
+      "step": 21726
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.540887814300462,
+      "learning_rate": 1.0101482719035851e-05,
+      "loss": 0.7639,
+      "step": 21727
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4489413605585544,
+      "learning_rate": 1.0100719716973657e-05,
+      "loss": 0.6807,
+      "step": 21728
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4096641586309697,
+      "learning_rate": 1.009995671432504e-05,
+      "loss": 0.679,
+      "step": 21729
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0032196342030257,
+      "learning_rate": 1.0099193711094446e-05,
+      "loss": 0.7008,
+      "step": 21730
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.492747284141309,
+      "learning_rate": 1.0098430707286313e-05,
+      "loss": 0.691,
+      "step": 21731
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5904059255385234,
+      "learning_rate": 1.0097667702905085e-05,
+      "loss": 0.5175,
+      "step": 21732
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6037576018495963,
+      "learning_rate": 1.0096904697955205e-05,
+      "loss": 0.5405,
+      "step": 21733
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5458231386019217,
+      "learning_rate": 1.0096141692441117e-05,
+      "loss": 0.7012,
+      "step": 21734
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.653091114333632,
+      "learning_rate": 1.0095378686367258e-05,
+      "loss": 0.7886,
+      "step": 21735
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4372026048022708,
+      "learning_rate": 1.0094615679738078e-05,
+      "loss": 0.6489,
+      "step": 21736
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3952898833632095,
+      "learning_rate": 1.0093852672558013e-05,
+      "loss": 0.6862,
+      "step": 21737
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.500121475672505,
+      "learning_rate": 1.009308966483151e-05,
+      "loss": 0.6704,
+      "step": 21738
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4555705381996695,
+      "learning_rate": 1.0092326656563006e-05,
+      "loss": 0.6756,
+      "step": 21739
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4489901493624309,
+      "learning_rate": 1.0091563647756953e-05,
+      "loss": 0.7389,
+      "step": 21740
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3449928734898386,
+      "learning_rate": 1.0090800638417779e-05,
+      "loss": 0.6883,
+      "step": 21741
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3481075197440162,
+      "learning_rate": 1.0090037628549943e-05,
+      "loss": 0.7293,
+      "step": 21742
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7571853666375379,
+      "learning_rate": 1.0089274618157874e-05,
+      "loss": 0.7179,
+      "step": 21743
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6311837600329877,
+      "learning_rate": 1.0088511607246021e-05,
+      "loss": 0.7253,
+      "step": 21744
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5305214433610588,
+      "learning_rate": 1.008774859581883e-05,
+      "loss": 0.696,
+      "step": 21745
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.174188794341609,
+      "learning_rate": 1.0086985583880731e-05,
+      "loss": 0.7555,
+      "step": 21746
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.337411083908045,
+      "learning_rate": 1.0086222571436182e-05,
+      "loss": 0.6056,
+      "step": 21747
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.56770913277779,
+      "learning_rate": 1.0085459558489612e-05,
+      "loss": 0.5353,
+      "step": 21748
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4024157203881853,
+      "learning_rate": 1.0084696545045474e-05,
+      "loss": 0.8434,
+      "step": 21749
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1850636284530267,
+      "learning_rate": 1.0083933531108202e-05,
+      "loss": 0.587,
+      "step": 21750
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5971724931908315,
+      "learning_rate": 1.0083170516682244e-05,
+      "loss": 0.516,
+      "step": 21751
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.460400532212978,
+      "learning_rate": 1.0082407501772043e-05,
+      "loss": 0.742,
+      "step": 21752
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4992703542468564,
+      "learning_rate": 1.0081644486382036e-05,
+      "loss": 0.8355,
+      "step": 21753
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8539389994781006,
+      "learning_rate": 1.008088147051667e-05,
+      "loss": 0.8636,
+      "step": 21754
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.624836169208028,
+      "learning_rate": 1.0080118454180387e-05,
+      "loss": 0.8096,
+      "step": 21755
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4232964290173695,
+      "learning_rate": 1.007935543737763e-05,
+      "loss": 0.7395,
+      "step": 21756
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5907781675635384,
+      "learning_rate": 1.0078592420112838e-05,
+      "loss": 0.5186,
+      "step": 21757
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5932868238301007,
+      "learning_rate": 1.0077829402390458e-05,
+      "loss": 0.5251,
+      "step": 21758
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.959514455096124,
+      "learning_rate": 1.007706638421493e-05,
+      "loss": 0.6338,
+      "step": 21759
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6004667859540642,
+      "learning_rate": 1.0076303365590696e-05,
+      "loss": 0.5146,
+      "step": 21760
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4997036687156142,
+      "learning_rate": 1.0075540346522202e-05,
+      "loss": 0.8031,
+      "step": 21761
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8794283125098694,
+      "learning_rate": 1.0074777327013887e-05,
+      "loss": 0.6365,
+      "step": 21762
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8239060777369644,
+      "learning_rate": 1.0074014307070194e-05,
+      "loss": 0.6039,
+      "step": 21763
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4907419318243214,
+      "learning_rate": 1.0073251286695567e-05,
+      "loss": 0.7781,
+      "step": 21764
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4746011033323534,
+      "learning_rate": 1.007248826589445e-05,
+      "loss": 0.8669,
+      "step": 21765
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5149463772346385,
+      "learning_rate": 1.0071725244671281e-05,
+      "loss": 0.7547,
+      "step": 21766
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4744565448170601,
+      "learning_rate": 1.0070962223030502e-05,
+      "loss": 0.7252,
+      "step": 21767
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.619343179686104,
+      "learning_rate": 1.0070199200976564e-05,
+      "loss": 0.812,
+      "step": 21768
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.91668692395265,
+      "learning_rate": 1.00694361785139e-05,
+      "loss": 0.6896,
+      "step": 21769
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6013279060041248,
+      "learning_rate": 1.006867315564696e-05,
+      "loss": 0.7202,
+      "step": 21770
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5061996931848098,
+      "learning_rate": 1.0067910132380179e-05,
+      "loss": 0.7219,
+      "step": 21771
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.63609491470218,
+      "learning_rate": 1.006714710871801e-05,
+      "loss": 0.8547,
+      "step": 21772
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4667008874282879,
+      "learning_rate": 1.0066384084664882e-05,
+      "loss": 0.6623,
+      "step": 21773
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5212250728330692,
+      "learning_rate": 1.006562106022525e-05,
+      "loss": 0.7181,
+      "step": 21774
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2066197440488717,
+      "learning_rate": 1.0064858035403549e-05,
+      "loss": 0.5764,
+      "step": 21775
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5942279420279912,
+      "learning_rate": 1.0064095010204225e-05,
+      "loss": 0.7456,
+      "step": 21776
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5801493371979375,
+      "learning_rate": 1.0063331984631719e-05,
+      "loss": 0.5367,
+      "step": 21777
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4713992685112058,
+      "learning_rate": 1.0062568958690475e-05,
+      "loss": 0.6887,
+      "step": 21778
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2994304496258304,
+      "learning_rate": 1.0061805932384933e-05,
+      "loss": 0.6474,
+      "step": 21779
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5770064655885353,
+      "learning_rate": 1.0061042905719538e-05,
+      "loss": 0.4992,
+      "step": 21780
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.128768134018374,
+      "learning_rate": 1.0060279878698734e-05,
+      "loss": 0.721,
+      "step": 21781
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4272069692081975,
+      "learning_rate": 1.0059516851326958e-05,
+      "loss": 0.7575,
+      "step": 21782
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3311854083997048,
+      "learning_rate": 1.0058753823608658e-05,
+      "loss": 0.7344,
+      "step": 21783
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.13342091273971,
+      "learning_rate": 1.0057990795548274e-05,
+      "loss": 0.8284,
+      "step": 21784
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.479147099487117,
+      "learning_rate": 1.0057227767150249e-05,
+      "loss": 0.6391,
+      "step": 21785
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6219311396419978,
+      "learning_rate": 1.0056464738419028e-05,
+      "loss": 0.7422,
+      "step": 21786
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3976016396504547,
+      "learning_rate": 1.005570170935905e-05,
+      "loss": 0.6742,
+      "step": 21787
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6287866857698527,
+      "learning_rate": 1.0054938679974757e-05,
+      "loss": 0.6556,
+      "step": 21788
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4119864269679954,
+      "learning_rate": 1.0054175650270596e-05,
+      "loss": 0.7636,
+      "step": 21789
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2919758933548025,
+      "learning_rate": 1.0053412620251006e-05,
+      "loss": 0.8206,
+      "step": 21790
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8717049872547356,
+      "learning_rate": 1.0052649589920433e-05,
+      "loss": 0.6778,
+      "step": 21791
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6171824275773858,
+      "learning_rate": 1.0051886559283315e-05,
+      "loss": 0.8201,
+      "step": 21792
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4912470758910454,
+      "learning_rate": 1.0051123528344096e-05,
+      "loss": 0.6771,
+      "step": 21793
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8674919418006621,
+      "learning_rate": 1.0050360497107221e-05,
+      "loss": 0.7298,
+      "step": 21794
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.40293551176038,
+      "learning_rate": 1.0049597465577133e-05,
+      "loss": 0.8989,
+      "step": 21795
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3790166289730543,
+      "learning_rate": 1.0048834433758275e-05,
+      "loss": 0.6839,
+      "step": 21796
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.387863468925304,
+      "learning_rate": 1.004807140165508e-05,
+      "loss": 0.6449,
+      "step": 21797
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5349018376694472,
+      "learning_rate": 1.0047308369272004e-05,
+      "loss": 0.7241,
+      "step": 21798
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9914226695364432,
+      "learning_rate": 1.0046545336613481e-05,
+      "loss": 0.8297,
+      "step": 21799
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3153603000998764,
+      "learning_rate": 1.0045782303683958e-05,
+      "loss": 0.6613,
+      "step": 21800
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.373472924253787,
+      "learning_rate": 1.0045019270487873e-05,
+      "loss": 0.765,
+      "step": 21801
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4011533905554954,
+      "learning_rate": 1.0044256237029674e-05,
+      "loss": 0.6086,
+      "step": 21802
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8872305594238497,
+      "learning_rate": 1.0043493203313803e-05,
+      "loss": 0.6478,
+      "step": 21803
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4376370198408714,
+      "learning_rate": 1.0042730169344698e-05,
+      "loss": 0.6429,
+      "step": 21804
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4833005442845326,
+      "learning_rate": 1.0041967135126805e-05,
+      "loss": 0.7346,
+      "step": 21805
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4560601637413961,
+      "learning_rate": 1.0041204100664566e-05,
+      "loss": 0.7017,
+      "step": 21806
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8068278761268066,
+      "learning_rate": 1.0040441065962423e-05,
+      "loss": 0.7201,
+      "step": 21807
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4572161266356085,
+      "learning_rate": 1.0039678031024819e-05,
+      "loss": 0.7222,
+      "step": 21808
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4760678089101573,
+      "learning_rate": 1.0038914995856199e-05,
+      "loss": 0.7112,
+      "step": 21809
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5825796349377173,
+      "learning_rate": 1.0038151960461002e-05,
+      "loss": 0.7294,
+      "step": 21810
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5560320090179238,
+      "learning_rate": 1.0037388924843674e-05,
+      "loss": 0.7236,
+      "step": 21811
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6039302449451035,
+      "learning_rate": 1.0036625889008656e-05,
+      "loss": 0.5383,
+      "step": 21812
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5169224763776066,
+      "learning_rate": 1.0035862852960388e-05,
+      "loss": 0.7043,
+      "step": 21813
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5934373428354154,
+      "learning_rate": 1.0035099816703315e-05,
+      "loss": 0.5164,
+      "step": 21814
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5536501575185775,
+      "learning_rate": 1.0034336780241882e-05,
+      "loss": 0.7198,
+      "step": 21815
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4156452481763706,
+      "learning_rate": 1.0033573743580527e-05,
+      "loss": 0.7093,
+      "step": 21816
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6148007659644004,
+      "learning_rate": 1.0032810706723699e-05,
+      "loss": 0.7641,
+      "step": 21817
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7358780151759512,
+      "learning_rate": 1.0032047669675832e-05,
+      "loss": 0.7722,
+      "step": 21818
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.287966999254964,
+      "learning_rate": 1.0031284632441378e-05,
+      "loss": 0.7187,
+      "step": 21819
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.291236825504768,
+      "learning_rate": 1.003052159502477e-05,
+      "loss": 0.8386,
+      "step": 21820
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6077950096138245,
+      "learning_rate": 1.002975855743046e-05,
+      "loss": 0.5162,
+      "step": 21821
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4356555634220285,
+      "learning_rate": 1.0028995519662883e-05,
+      "loss": 0.5862,
+      "step": 21822
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.507640641908206,
+      "learning_rate": 1.0028232481726488e-05,
+      "loss": 0.6881,
+      "step": 21823
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3478340669233762,
+      "learning_rate": 1.0027469443625712e-05,
+      "loss": 0.7772,
+      "step": 21824
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5488613677997232,
+      "learning_rate": 1.0026706405365001e-05,
+      "loss": 0.8488,
+      "step": 21825
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5482285724944667,
+      "learning_rate": 1.0025943366948795e-05,
+      "loss": 0.7742,
+      "step": 21826
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3428730204458954,
+      "learning_rate": 1.002518032838154e-05,
+      "loss": 0.7487,
+      "step": 21827
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.368449315228545,
+      "learning_rate": 1.0024417289667678e-05,
+      "loss": 0.7244,
+      "step": 21828
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.149618283597841,
+      "learning_rate": 1.0023654250811652e-05,
+      "loss": 0.552,
+      "step": 21829
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.273133473517169,
+      "learning_rate": 1.00228912118179e-05,
+      "loss": 0.7076,
+      "step": 21830
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3802613804946307,
+      "learning_rate": 1.0022128172690868e-05,
+      "loss": 0.5848,
+      "step": 21831
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3594448832152584,
+      "learning_rate": 1.0021365133435004e-05,
+      "loss": 0.7086,
+      "step": 21832
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6829098227340804,
+      "learning_rate": 1.002060209405474e-05,
+      "loss": 0.6613,
+      "step": 21833
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5911044625396558,
+      "learning_rate": 1.0019839054554528e-05,
+      "loss": 0.5578,
+      "step": 21834
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.464452039370312,
+      "learning_rate": 1.0019076014938805e-05,
+      "loss": 0.7356,
+      "step": 21835
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3493700806326412,
+      "learning_rate": 1.0018312975212015e-05,
+      "loss": 0.7339,
+      "step": 21836
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.738894789087943,
+      "learning_rate": 1.0017549935378601e-05,
+      "loss": 0.7475,
+      "step": 21837
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8342567249385437,
+      "learning_rate": 1.0016786895443005e-05,
+      "loss": 0.7907,
+      "step": 21838
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8463657765665618,
+      "learning_rate": 1.0016023855409671e-05,
+      "loss": 0.6,
+      "step": 21839
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3692495179305313,
+      "learning_rate": 1.0015260815283043e-05,
+      "loss": 0.7809,
+      "step": 21840
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7397140170633105,
+      "learning_rate": 1.001449777506756e-05,
+      "loss": 0.7047,
+      "step": 21841
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5507756584136339,
+      "learning_rate": 1.0013734734767666e-05,
+      "loss": 0.7226,
+      "step": 21842
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3013605510913735,
+      "learning_rate": 1.0012971694387803e-05,
+      "loss": 0.8097,
+      "step": 21843
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5713152860783124,
+      "learning_rate": 1.0012208653932416e-05,
+      "loss": 0.7486,
+      "step": 21844
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4277184636104152,
+      "learning_rate": 1.0011445613405948e-05,
+      "loss": 0.7252,
+      "step": 21845
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6231956956136162,
+      "learning_rate": 1.0010682572812837e-05,
+      "loss": 0.6734,
+      "step": 21846
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5414428491919046,
+      "learning_rate": 1.0009919532157532e-05,
+      "loss": 0.6907,
+      "step": 21847
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5339465208638547,
+      "learning_rate": 1.000915649144447e-05,
+      "loss": 0.7398,
+      "step": 21848
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4693906623484299,
+      "learning_rate": 1.0008393450678098e-05,
+      "loss": 0.7541,
+      "step": 21849
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5976042125875052,
+      "learning_rate": 1.0007630409862854e-05,
+      "loss": 0.5097,
+      "step": 21850
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.30588916474564,
+      "learning_rate": 1.0006867369003185e-05,
+      "loss": 0.7364,
+      "step": 21851
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7807308813755602,
+      "learning_rate": 1.000610432810353e-05,
+      "loss": 0.6595,
+      "step": 21852
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5741023277542857,
+      "learning_rate": 1.0005341287168338e-05,
+      "loss": 0.4948,
+      "step": 21853
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6124251776042441,
+      "learning_rate": 1.0004578246202046e-05,
+      "loss": 0.5359,
+      "step": 21854
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4082371831244493,
+      "learning_rate": 1.0003815205209095e-05,
+      "loss": 0.6337,
+      "step": 21855
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3799812157536238,
+      "learning_rate": 1.0003052164193931e-05,
+      "loss": 0.6101,
+      "step": 21856
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3601509567756411,
+      "learning_rate": 1.0002289123161e-05,
+      "loss": 0.8373,
+      "step": 21857
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4687615554880031,
+      "learning_rate": 1.0001526082114738e-05,
+      "loss": 0.7589,
+      "step": 21858
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4633132948023229,
+      "learning_rate": 1.0000763041059591e-05,
+      "loss": 0.7613,
+      "step": 21859
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5454877499214871,
+      "learning_rate": 1e-05,
+      "loss": 0.4885,
+      "step": 21860
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5301899961587246,
+      "learning_rate": 9.999236958940412e-06,
+      "loss": 0.8201,
+      "step": 21861
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4068955170359414,
+      "learning_rate": 9.998473917885268e-06,
+      "loss": 0.7956,
+      "step": 21862
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.481995961382815,
+      "learning_rate": 9.997710876839005e-06,
+      "loss": 0.8184,
+      "step": 21863
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4367452404599879,
+      "learning_rate": 9.99694783580607e-06,
+      "loss": 0.6726,
+      "step": 21864
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.444527657820193,
+      "learning_rate": 9.996184794790907e-06,
+      "loss": 0.6795,
+      "step": 21865
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9171642746453437,
+      "learning_rate": 9.99542175379796e-06,
+      "loss": 0.7212,
+      "step": 21866
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.388687582809905,
+      "learning_rate": 9.994658712831666e-06,
+      "loss": 0.6852,
+      "step": 21867
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.40470861209452,
+      "learning_rate": 9.99389567189647e-06,
+      "loss": 0.7392,
+      "step": 21868
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2061993934788835,
+      "learning_rate": 9.99313263099682e-06,
+      "loss": 0.7593,
+      "step": 21869
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.291230229805198,
+      "learning_rate": 9.99236959013715e-06,
+      "loss": 0.8339,
+      "step": 21870
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3605902044383484,
+      "learning_rate": 9.991606549321904e-06,
+      "loss": 0.7517,
+      "step": 21871
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3717988782452186,
+      "learning_rate": 9.99084350855553e-06,
+      "loss": 0.6333,
+      "step": 21872
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5989532567354694,
+      "learning_rate": 9.990080467842472e-06,
+      "loss": 0.6861,
+      "step": 21873
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7274400450767002,
+      "learning_rate": 9.989317427187165e-06,
+      "loss": 0.8753,
+      "step": 21874
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9220951297587432,
+      "learning_rate": 9.988554386594052e-06,
+      "loss": 0.7771,
+      "step": 21875
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1610445851859343,
+      "learning_rate": 9.987791346067586e-06,
+      "loss": 0.6406,
+      "step": 21876
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4256658167377936,
+      "learning_rate": 9.987028305612199e-06,
+      "loss": 0.8039,
+      "step": 21877
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5951228688601904,
+      "learning_rate": 9.986265265232337e-06,
+      "loss": 0.8231,
+      "step": 21878
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3767348058183848,
+      "learning_rate": 9.985502224932442e-06,
+      "loss": 0.6855,
+      "step": 21879
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4297114468204724,
+      "learning_rate": 9.98473918471696e-06,
+      "loss": 0.7947,
+      "step": 21880
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4347701213309176,
+      "learning_rate": 9.983976144590332e-06,
+      "loss": 0.7295,
+      "step": 21881
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.305427568949692,
+      "learning_rate": 9.983213104556995e-06,
+      "loss": 0.8113,
+      "step": 21882
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.315158041262251,
+      "learning_rate": 9.982450064621405e-06,
+      "loss": 0.5753,
+      "step": 21883
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.396942541923444,
+      "learning_rate": 9.98168702478799e-06,
+      "loss": 0.6529,
+      "step": 21884
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2576765286250258,
+      "learning_rate": 9.980923985061199e-06,
+      "loss": 0.681,
+      "step": 21885
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.431597817460848,
+      "learning_rate": 9.980160945445474e-06,
+      "loss": 0.6452,
+      "step": 21886
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.576589767371008,
+      "learning_rate": 9.979397905945264e-06,
+      "loss": 0.6803,
+      "step": 21887
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4894715663261058,
+      "learning_rate": 9.978634866565e-06,
+      "loss": 0.6698,
+      "step": 21888
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.47743547846498,
+      "learning_rate": 9.977871827309132e-06,
+      "loss": 0.6262,
+      "step": 21889
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5046462779616507,
+      "learning_rate": 9.977108788182104e-06,
+      "loss": 0.7684,
+      "step": 21890
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.348709501492505,
+      "learning_rate": 9.976345749188353e-06,
+      "loss": 0.7298,
+      "step": 21891
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3465250435224636,
+      "learning_rate": 9.975582710332325e-06,
+      "loss": 0.657,
+      "step": 21892
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3712215658528668,
+      "learning_rate": 9.974819671618462e-06,
+      "loss": 0.6268,
+      "step": 21893
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3972682606167595,
+      "learning_rate": 9.97405663305121e-06,
+      "loss": 0.7026,
+      "step": 21894
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4616010069271304,
+      "learning_rate": 9.973293594635002e-06,
+      "loss": 0.6635,
+      "step": 21895
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9393036284932252,
+      "learning_rate": 9.972530556374291e-06,
+      "loss": 0.6538,
+      "step": 21896
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6509851322982696,
+      "learning_rate": 9.971767518273517e-06,
+      "loss": 0.7246,
+      "step": 21897
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.407526378190059,
+      "learning_rate": 9.97100448033712e-06,
+      "loss": 0.8036,
+      "step": 21898
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3320206495186426,
+      "learning_rate": 9.970241442569543e-06,
+      "loss": 0.6754,
+      "step": 21899
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3086189591302397,
+      "learning_rate": 9.96947840497523e-06,
+      "loss": 0.7731,
+      "step": 21900
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.364179333133299,
+      "learning_rate": 9.968715367558627e-06,
+      "loss": 0.8776,
+      "step": 21901
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.158786622320702,
+      "learning_rate": 9.967952330324171e-06,
+      "loss": 0.8026,
+      "step": 21902
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.812954168884262,
+      "learning_rate": 9.967189293276303e-06,
+      "loss": 0.7661,
+      "step": 21903
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3398883059141677,
+      "learning_rate": 9.966426256419475e-06,
+      "loss": 0.7203,
+      "step": 21904
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4404825341604641,
+      "learning_rate": 9.965663219758121e-06,
+      "loss": 0.618,
+      "step": 21905
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.419565249947515,
+      "learning_rate": 9.964900183296688e-06,
+      "loss": 0.7374,
+      "step": 21906
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.066932824774433,
+      "learning_rate": 9.964137147039613e-06,
+      "loss": 0.7597,
+      "step": 21907
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5853251016015704,
+      "learning_rate": 9.96337411099135e-06,
+      "loss": 0.7031,
+      "step": 21908
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9413561222063718,
+      "learning_rate": 9.962611075156329e-06,
+      "loss": 0.729,
+      "step": 21909
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5189387500554377,
+      "learning_rate": 9.961848039539e-06,
+      "loss": 0.7771,
+      "step": 21910
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.866482794539026,
+      "learning_rate": 9.961085004143806e-06,
+      "loss": 0.5565,
+      "step": 21911
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4145715403440888,
+      "learning_rate": 9.960321968975182e-06,
+      "loss": 0.6422,
+      "step": 21912
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6177607776683265,
+      "learning_rate": 9.959558934037578e-06,
+      "loss": 0.5134,
+      "step": 21913
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6015827198344079,
+      "learning_rate": 9.958795899335438e-06,
+      "loss": 0.5465,
+      "step": 21914
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6396819034040938,
+      "learning_rate": 9.9580328648732e-06,
+      "loss": 0.5612,
+      "step": 21915
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3603345600404584,
+      "learning_rate": 9.957269830655305e-06,
+      "loss": 0.6909,
+      "step": 21916
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.484667289223177,
+      "learning_rate": 9.9565067966862e-06,
+      "loss": 0.7767,
+      "step": 21917
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4750585124863653,
+      "learning_rate": 9.955743762970329e-06,
+      "loss": 0.6311,
+      "step": 21918
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5894659243585512,
+      "learning_rate": 9.95498072951213e-06,
+      "loss": 0.7222,
+      "step": 21919
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.435066612022664,
+      "learning_rate": 9.954217696316043e-06,
+      "loss": 0.6734,
+      "step": 21920
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.502183323691935,
+      "learning_rate": 9.95345466338652e-06,
+      "loss": 0.6747,
+      "step": 21921
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6755542249050657,
+      "learning_rate": 9.952691630728001e-06,
+      "loss": 0.6958,
+      "step": 21922
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3731357030887068,
+      "learning_rate": 9.951928598344923e-06,
+      "loss": 0.7903,
+      "step": 21923
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.30790898080876,
+      "learning_rate": 9.95116556624173e-06,
+      "loss": 0.6544,
+      "step": 21924
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3209865297220473,
+      "learning_rate": 9.950402534422869e-06,
+      "loss": 0.6567,
+      "step": 21925
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3798750361178225,
+      "learning_rate": 9.94963950289278e-06,
+      "loss": 0.6874,
+      "step": 21926
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6401690239807403,
+      "learning_rate": 9.948876471655906e-06,
+      "loss": 0.7724,
+      "step": 21927
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2880875706052262,
+      "learning_rate": 9.948113440716687e-06,
+      "loss": 0.7066,
+      "step": 21928
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3256126041443448,
+      "learning_rate": 9.947350410079572e-06,
+      "loss": 0.7205,
+      "step": 21929
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5299530562978594,
+      "learning_rate": 9.946587379748997e-06,
+      "loss": 0.6894,
+      "step": 21930
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6330443502026409,
+      "learning_rate": 9.945824349729405e-06,
+      "loss": 0.5339,
+      "step": 21931
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6820556789399135,
+      "learning_rate": 9.945061320025243e-06,
+      "loss": 0.9,
+      "step": 21932
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3789725376047353,
+      "learning_rate": 9.944298290640955e-06,
+      "loss": 0.7053,
+      "step": 21933
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.477210033448212,
+      "learning_rate": 9.943535261580977e-06,
+      "loss": 0.7181,
+      "step": 21934
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5863296033035219,
+      "learning_rate": 9.942772232849751e-06,
+      "loss": 0.554,
+      "step": 21935
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.492554909532948,
+      "learning_rate": 9.942009204451731e-06,
+      "loss": 0.6591,
+      "step": 21936
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5194626131746756,
+      "learning_rate": 9.941246176391345e-06,
+      "loss": 0.7685,
+      "step": 21937
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5900899935885913,
+      "learning_rate": 9.940483148673045e-06,
+      "loss": 0.5506,
+      "step": 21938
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.597584689633039,
+      "learning_rate": 9.939720121301268e-06,
+      "loss": 0.867,
+      "step": 21939
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0930124324293924,
+      "learning_rate": 9.938957094280464e-06,
+      "loss": 0.644,
+      "step": 21940
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6059217094494146,
+      "learning_rate": 9.93819406761507e-06,
+      "loss": 0.5315,
+      "step": 21941
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4497060297249937,
+      "learning_rate": 9.937431041309528e-06,
+      "loss": 0.6538,
+      "step": 21942
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6736705743803633,
+      "learning_rate": 9.936668015368286e-06,
+      "loss": 0.7346,
+      "step": 21943
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5488112569158152,
+      "learning_rate": 9.935904989795778e-06,
+      "loss": 0.6901,
+      "step": 21944
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3837413182507259,
+      "learning_rate": 9.935141964596453e-06,
+      "loss": 0.6993,
+      "step": 21945
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5616816167221498,
+      "learning_rate": 9.934378939774753e-06,
+      "loss": 0.5119,
+      "step": 21946
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.387825177310298,
+      "learning_rate": 9.933615915335121e-06,
+      "loss": 0.6359,
+      "step": 21947
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7821433369836392,
+      "learning_rate": 9.932852891281994e-06,
+      "loss": 0.7296,
+      "step": 21948
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3016332424569652,
+      "learning_rate": 9.932089867619821e-06,
+      "loss": 0.6908,
+      "step": 21949
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6452246106186978,
+      "learning_rate": 9.931326844353044e-06,
+      "loss": 0.6323,
+      "step": 21950
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7913333961007356,
+      "learning_rate": 9.930563821486101e-06,
+      "loss": 0.679,
+      "step": 21951
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6429886173093577,
+      "learning_rate": 9.929800799023438e-06,
+      "loss": 0.7036,
+      "step": 21952
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.787750432072134,
+      "learning_rate": 9.929037776969498e-06,
+      "loss": 0.6536,
+      "step": 21953
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5629212019145735,
+      "learning_rate": 9.928274755328724e-06,
+      "loss": 0.7381,
+      "step": 21954
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5235514376785217,
+      "learning_rate": 9.927511734105555e-06,
+      "loss": 0.7356,
+      "step": 21955
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3595059385684354,
+      "learning_rate": 9.926748713304433e-06,
+      "loss": 0.692,
+      "step": 21956
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0220138424385596,
+      "learning_rate": 9.925985692929809e-06,
+      "loss": 0.6798,
+      "step": 21957
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2606379591607404,
+      "learning_rate": 9.925222672986116e-06,
+      "loss": 0.6031,
+      "step": 21958
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3028718345567414,
+      "learning_rate": 9.924459653477802e-06,
+      "loss": 0.7451,
+      "step": 21959
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4450515206150425,
+      "learning_rate": 9.923696634409304e-06,
+      "loss": 0.7113,
+      "step": 21960
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5007415034059066,
+      "learning_rate": 9.922933615785073e-06,
+      "loss": 0.5723,
+      "step": 21961
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6665093581275747,
+      "learning_rate": 9.922170597609545e-06,
+      "loss": 0.7971,
+      "step": 21962
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4680669059932097,
+      "learning_rate": 9.921407579887164e-06,
+      "loss": 0.8601,
+      "step": 21963
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6285629426775867,
+      "learning_rate": 9.920644562622375e-06,
+      "loss": 0.7291,
+      "step": 21964
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3434360943615855,
+      "learning_rate": 9.919881545819614e-06,
+      "loss": 0.741,
+      "step": 21965
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5746128885741703,
+      "learning_rate": 9.919118529483333e-06,
+      "loss": 0.5118,
+      "step": 21966
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5814707720586124,
+      "learning_rate": 9.918355513617964e-06,
+      "loss": 0.5583,
+      "step": 21967
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3940085184054383,
+      "learning_rate": 9.917592498227964e-06,
+      "loss": 0.7171,
+      "step": 21968
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6363262329526598,
+      "learning_rate": 9.916829483317758e-06,
+      "loss": 0.5144,
+      "step": 21969
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5717997674264325,
+      "learning_rate": 9.9160664688918e-06,
+      "loss": 0.5071,
+      "step": 21970
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5992462626460834,
+      "learning_rate": 9.915303454954533e-06,
+      "loss": 0.5265,
+      "step": 21971
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6035437954941254,
+      "learning_rate": 9.914540441510392e-06,
+      "loss": 0.5079,
+      "step": 21972
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4323416487098881,
+      "learning_rate": 9.913777428563823e-06,
+      "loss": 0.6039,
+      "step": 21973
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6307082344620362,
+      "learning_rate": 9.91301441611927e-06,
+      "loss": 0.6012,
+      "step": 21974
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4623151615669094,
+      "learning_rate": 9.912251404181177e-06,
+      "loss": 0.8378,
+      "step": 21975
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5566840730012783,
+      "learning_rate": 9.911488392753982e-06,
+      "loss": 0.7414,
+      "step": 21976
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3080181972077325,
+      "learning_rate": 9.910725381842127e-06,
+      "loss": 0.7865,
+      "step": 21977
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5137260056630406,
+      "learning_rate": 9.909962371450062e-06,
+      "loss": 0.6663,
+      "step": 21978
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2513021831964357,
+      "learning_rate": 9.909199361582223e-06,
+      "loss": 0.7207,
+      "step": 21979
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.5784498261492947,
+      "learning_rate": 9.908436352243052e-06,
+      "loss": 0.582,
+      "step": 21980
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3894593007664573,
+      "learning_rate": 9.907673343436994e-06,
+      "loss": 0.7085,
+      "step": 21981
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2825548054996587,
+      "learning_rate": 9.906910335168495e-06,
+      "loss": 0.6519,
+      "step": 21982
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.286400244020165,
+      "learning_rate": 9.90614732744199e-06,
+      "loss": 0.757,
+      "step": 21983
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4175289692853215,
+      "learning_rate": 9.905384320261924e-06,
+      "loss": 0.5329,
+      "step": 21984
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.515056882957619,
+      "learning_rate": 9.904621313632747e-06,
+      "loss": 0.6919,
+      "step": 21985
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5310188430954306,
+      "learning_rate": 9.903858307558887e-06,
+      "loss": 0.7357,
+      "step": 21986
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8078672110173999,
+      "learning_rate": 9.903095302044798e-06,
+      "loss": 0.8832,
+      "step": 21987
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7369795405230584,
+      "learning_rate": 9.902332297094915e-06,
+      "loss": 0.7951,
+      "step": 21988
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8153589997603916,
+      "learning_rate": 9.901569292713689e-06,
+      "loss": 0.6922,
+      "step": 21989
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5842250220761702,
+      "learning_rate": 9.900806288905557e-06,
+      "loss": 0.6379,
+      "step": 21990
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5522761140074666,
+      "learning_rate": 9.900043285674961e-06,
+      "loss": 0.7152,
+      "step": 21991
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4341686218558198,
+      "learning_rate": 9.899280283026346e-06,
+      "loss": 0.8421,
+      "step": 21992
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0133396551556055,
+      "learning_rate": 9.898517280964152e-06,
+      "loss": 0.7152,
+      "step": 21993
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.387931059979297,
+      "learning_rate": 9.897754279492822e-06,
+      "loss": 0.7214,
+      "step": 21994
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4331539626036003,
+      "learning_rate": 9.8969912786168e-06,
+      "loss": 0.7527,
+      "step": 21995
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.920330853337897,
+      "learning_rate": 9.896228278340529e-06,
+      "loss": 0.7063,
+      "step": 21996
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4865399428268924,
+      "learning_rate": 9.895465278668445e-06,
+      "loss": 0.6635,
+      "step": 21997
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7277701431311496,
+      "learning_rate": 9.894702279604997e-06,
+      "loss": 0.7133,
+      "step": 21998
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.59350931045006,
+      "learning_rate": 9.89393928115463e-06,
+      "loss": 0.6689,
+      "step": 21999
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5009183188871589,
+      "learning_rate": 9.893176283321779e-06,
+      "loss": 0.7319,
+      "step": 22000
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5223069760887034,
+      "learning_rate": 9.892413286110886e-06,
+      "loss": 0.885,
+      "step": 22001
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7357252530141303,
+      "learning_rate": 9.8916502895264e-06,
+      "loss": 0.6644,
+      "step": 22002
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5972547399682206,
+      "learning_rate": 9.890887293572763e-06,
+      "loss": 0.8053,
+      "step": 22003
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5888236693118244,
+      "learning_rate": 9.890124298254412e-06,
+      "loss": 0.4892,
+      "step": 22004
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6420076932727423,
+      "learning_rate": 9.88936130357579e-06,
+      "loss": 0.7488,
+      "step": 22005
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3136432040341575,
+      "learning_rate": 9.888598309541347e-06,
+      "loss": 0.7483,
+      "step": 22006
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1257978490653597,
+      "learning_rate": 9.887835316155516e-06,
+      "loss": 0.7283,
+      "step": 22007
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5812650106551384,
+      "learning_rate": 9.887072323422743e-06,
+      "loss": 0.5302,
+      "step": 22008
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5963677663697847,
+      "learning_rate": 9.886309331347471e-06,
+      "loss": 0.5319,
+      "step": 22009
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4579576186764243,
+      "learning_rate": 9.885546339934145e-06,
+      "loss": 0.7383,
+      "step": 22010
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3791913310117936,
+      "learning_rate": 9.884783349187202e-06,
+      "loss": 0.6517,
+      "step": 22011
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2705648695517926,
+      "learning_rate": 9.884020359111087e-06,
+      "loss": 0.5851,
+      "step": 22012
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6039471335547385,
+      "learning_rate": 9.883257369710244e-06,
+      "loss": 0.4954,
+      "step": 22013
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5770540906630792,
+      "learning_rate": 9.88249438098911e-06,
+      "loss": 0.5298,
+      "step": 22014
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5699632836393221,
+      "learning_rate": 9.881731392952134e-06,
+      "loss": 0.502,
+      "step": 22015
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3048639550466248,
+      "learning_rate": 9.880968405603751e-06,
+      "loss": 0.6576,
+      "step": 22016
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7473725293283766,
+      "learning_rate": 9.880205418948414e-06,
+      "loss": 0.6992,
+      "step": 22017
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8231997798823585,
+      "learning_rate": 9.879442432990553e-06,
+      "loss": 0.6692,
+      "step": 22018
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4965118612377568,
+      "learning_rate": 9.87867944773462e-06,
+      "loss": 0.6587,
+      "step": 22019
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5051995424792248,
+      "learning_rate": 9.877916463185054e-06,
+      "loss": 0.633,
+      "step": 22020
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0983216510300524,
+      "learning_rate": 9.877153479346296e-06,
+      "loss": 0.8199,
+      "step": 22021
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6219343630998786,
+      "learning_rate": 9.876390496222787e-06,
+      "loss": 0.554,
+      "step": 22022
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6046396968460326,
+      "learning_rate": 9.875627513818973e-06,
+      "loss": 0.7483,
+      "step": 22023
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5535301350595343,
+      "learning_rate": 9.874864532139299e-06,
+      "loss": 0.6599,
+      "step": 22024
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.57736250663546,
+      "learning_rate": 9.874101551188198e-06,
+      "loss": 0.7281,
+      "step": 22025
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.944457253932777,
+      "learning_rate": 9.873338570970119e-06,
+      "loss": 0.8344,
+      "step": 22026
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3868609911939485,
+      "learning_rate": 9.872575591489505e-06,
+      "loss": 0.7332,
+      "step": 22027
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3382437515802217,
+      "learning_rate": 9.871812612750794e-06,
+      "loss": 0.7541,
+      "step": 22028
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0655159085606236,
+      "learning_rate": 9.87104963475843e-06,
+      "loss": 0.8031,
+      "step": 22029
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3672207095910975,
+      "learning_rate": 9.870286657516856e-06,
+      "loss": 0.6931,
+      "step": 22030
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3675043846053476,
+      "learning_rate": 9.869523681030518e-06,
+      "loss": 0.8222,
+      "step": 22031
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4554482800669342,
+      "learning_rate": 9.868760705303852e-06,
+      "loss": 0.7775,
+      "step": 22032
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8805845780550001,
+      "learning_rate": 9.8679977303413e-06,
+      "loss": 0.797,
+      "step": 22033
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5815435882050788,
+      "learning_rate": 9.867234756147314e-06,
+      "loss": 0.4893,
+      "step": 22034
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.109500034807664,
+      "learning_rate": 9.866471782726323e-06,
+      "loss": 0.6869,
+      "step": 22035
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3380693017245706,
+      "learning_rate": 9.865708810082776e-06,
+      "loss": 0.8144,
+      "step": 22036
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.335236023982448,
+      "learning_rate": 9.864945838221114e-06,
+      "loss": 0.7412,
+      "step": 22037
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8195166418742974,
+      "learning_rate": 9.864182867145784e-06,
+      "loss": 0.7001,
+      "step": 22038
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6654859801641304,
+      "learning_rate": 9.863419896861222e-06,
+      "loss": 0.6999,
+      "step": 22039
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.512951575196908,
+      "learning_rate": 9.862656927371873e-06,
+      "loss": 0.8275,
+      "step": 22040
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8002046212955132,
+      "learning_rate": 9.861893958682183e-06,
+      "loss": 0.7006,
+      "step": 22041
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.733872277913778,
+      "learning_rate": 9.861130990796583e-06,
+      "loss": 0.8449,
+      "step": 22042
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4017460582864572,
+      "learning_rate": 9.860368023719525e-06,
+      "loss": 0.7967,
+      "step": 22043
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6034728546441955,
+      "learning_rate": 9.85960505745545e-06,
+      "loss": 0.5428,
+      "step": 22044
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3450646857630808,
+      "learning_rate": 9.858842092008799e-06,
+      "loss": 0.7178,
+      "step": 22045
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4696319479721998,
+      "learning_rate": 9.85807912738401e-06,
+      "loss": 0.742,
+      "step": 22046
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3414493341749574,
+      "learning_rate": 9.857316163585533e-06,
+      "loss": 0.6126,
+      "step": 22047
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3371048277708084,
+      "learning_rate": 9.856553200617805e-06,
+      "loss": 0.697,
+      "step": 22048
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.462632075230537,
+      "learning_rate": 9.855790238485273e-06,
+      "loss": 0.6755,
+      "step": 22049
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2816023964150536,
+      "learning_rate": 9.85502727719237e-06,
+      "loss": 0.6685,
+      "step": 22050
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.358225881264408,
+      "learning_rate": 9.854264316743549e-06,
+      "loss": 0.783,
+      "step": 22051
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4354857416845093,
+      "learning_rate": 9.853501357143247e-06,
+      "loss": 0.6933,
+      "step": 22052
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.303750062160771,
+      "learning_rate": 9.852738398395904e-06,
+      "loss": 0.7441,
+      "step": 22053
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.530288376264346,
+      "learning_rate": 9.851975440505964e-06,
+      "loss": 0.7218,
+      "step": 22054
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6012021331879671,
+      "learning_rate": 9.851212483477871e-06,
+      "loss": 0.7385,
+      "step": 22055
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.072497595400695,
+      "learning_rate": 9.850449527316069e-06,
+      "loss": 0.8089,
+      "step": 22056
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2244006339353533,
+      "learning_rate": 9.849686572024995e-06,
+      "loss": 0.7071,
+      "step": 22057
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3006670372509452,
+      "learning_rate": 9.84892361760909e-06,
+      "loss": 0.7378,
+      "step": 22058
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4056916088342846,
+      "learning_rate": 9.848160664072806e-06,
+      "loss": 0.7298,
+      "step": 22059
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3875986208549251,
+      "learning_rate": 9.847397711420574e-06,
+      "loss": 0.7373,
+      "step": 22060
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5820427833969934,
+      "learning_rate": 9.846634759656843e-06,
+      "loss": 0.525,
+      "step": 22061
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.535160843569918,
+      "learning_rate": 9.84587180878605e-06,
+      "loss": 0.7178,
+      "step": 22062
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6783514414508065,
+      "learning_rate": 9.845108858812645e-06,
+      "loss": 0.8039,
+      "step": 22063
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9817591468461864,
+      "learning_rate": 9.844345909741061e-06,
+      "loss": 0.6694,
+      "step": 22064
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6711444300389808,
+      "learning_rate": 9.843582961575743e-06,
+      "loss": 0.7204,
+      "step": 22065
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3646191137526493,
+      "learning_rate": 9.842820014321143e-06,
+      "loss": 0.6629,
+      "step": 22066
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4475684399759858,
+      "learning_rate": 9.842057067981686e-06,
+      "loss": 0.7035,
+      "step": 22067
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.069309543894283,
+      "learning_rate": 9.841294122561826e-06,
+      "loss": 0.5998,
+      "step": 22068
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.316148466518658,
+      "learning_rate": 9.840531178066e-06,
+      "loss": 0.7423,
+      "step": 22069
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3551770318783147,
+      "learning_rate": 9.839768234498658e-06,
+      "loss": 0.6592,
+      "step": 22070
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7968510394290234,
+      "learning_rate": 9.839005291864229e-06,
+      "loss": 0.7013,
+      "step": 22071
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3510815598380095,
+      "learning_rate": 9.838242350167164e-06,
+      "loss": 0.6856,
+      "step": 22072
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7611615582109754,
+      "learning_rate": 9.837479409411907e-06,
+      "loss": 0.6813,
+      "step": 22073
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3669996229471615,
+      "learning_rate": 9.836716469602891e-06,
+      "loss": 0.7846,
+      "step": 22074
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6940210036833576,
+      "learning_rate": 9.835953530744567e-06,
+      "loss": 0.7335,
+      "step": 22075
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4577680644042792,
+      "learning_rate": 9.835190592841371e-06,
+      "loss": 0.8372,
+      "step": 22076
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5648146569019766,
+      "learning_rate": 9.834427655897751e-06,
+      "loss": 0.4563,
+      "step": 22077
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.631983725393446,
+      "learning_rate": 9.83366471991814e-06,
+      "loss": 0.6958,
+      "step": 22078
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5877857825664974,
+      "learning_rate": 9.83290178490699e-06,
+      "loss": 0.5345,
+      "step": 22079
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3416844597483728,
+      "learning_rate": 9.83213885086874e-06,
+      "loss": 0.7024,
+      "step": 22080
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5438569984827353,
+      "learning_rate": 9.831375917807829e-06,
+      "loss": 0.5958,
+      "step": 22081
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4262663284813157,
+      "learning_rate": 9.830612985728699e-06,
+      "loss": 0.7004,
+      "step": 22082
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.321883927376179,
+      "learning_rate": 9.829850054635794e-06,
+      "loss": 0.662,
+      "step": 22083
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8966318688459802,
+      "learning_rate": 9.829087124533559e-06,
+      "loss": 0.6784,
+      "step": 22084
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6425214863930557,
+      "learning_rate": 9.828324195426431e-06,
+      "loss": 0.7565,
+      "step": 22085
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5824053256866669,
+      "learning_rate": 9.827561267318852e-06,
+      "loss": 0.5054,
+      "step": 22086
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5921487799987314,
+      "learning_rate": 9.82679834021527e-06,
+      "loss": 0.5639,
+      "step": 22087
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.439901860187816,
+      "learning_rate": 9.826035414120122e-06,
+      "loss": 0.7527,
+      "step": 22088
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3694437905170294,
+      "learning_rate": 9.825272489037847e-06,
+      "loss": 0.6562,
+      "step": 22089
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.418607633502527,
+      "learning_rate": 9.824509564972892e-06,
+      "loss": 0.5743,
+      "step": 22090
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5945758053176877,
+      "learning_rate": 9.823746641929702e-06,
+      "loss": 0.7094,
+      "step": 22091
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.58665382431976,
+      "learning_rate": 9.822983719912712e-06,
+      "loss": 0.5652,
+      "step": 22092
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7125466130013134,
+      "learning_rate": 9.822220798926367e-06,
+      "loss": 0.7257,
+      "step": 22093
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2660856511729068,
+      "learning_rate": 9.82145787897511e-06,
+      "loss": 0.6416,
+      "step": 22094
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.008933578471413,
+      "learning_rate": 9.82069496006338e-06,
+      "loss": 0.7703,
+      "step": 22095
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3905167248998267,
+      "learning_rate": 9.81993204219562e-06,
+      "loss": 0.7075,
+      "step": 22096
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6810630684060088,
+      "learning_rate": 9.819169125376275e-06,
+      "loss": 0.6726,
+      "step": 22097
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.307395470150421,
+      "learning_rate": 9.818406209609784e-06,
+      "loss": 0.7697,
+      "step": 22098
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0331033597004344,
+      "learning_rate": 9.817643294900587e-06,
+      "loss": 0.6486,
+      "step": 22099
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.423015791115181,
+      "learning_rate": 9.81688038125313e-06,
+      "loss": 0.722,
+      "step": 22100
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.616008904278332,
+      "learning_rate": 9.816117468671858e-06,
+      "loss": 0.762,
+      "step": 22101
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4465621083526372,
+      "learning_rate": 9.815354557161203e-06,
+      "loss": 0.8546,
+      "step": 22102
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.715486772392455,
+      "learning_rate": 9.814591646725611e-06,
+      "loss": 0.6063,
+      "step": 22103
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5830779119342273,
+      "learning_rate": 9.813828737369527e-06,
+      "loss": 0.512,
+      "step": 22104
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.784468098259085,
+      "learning_rate": 9.813065829097394e-06,
+      "loss": 0.6758,
+      "step": 22105
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8204757031678367,
+      "learning_rate": 9.81230292191365e-06,
+      "loss": 0.6257,
+      "step": 22106
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4925127616971885,
+      "learning_rate": 9.811540015822733e-06,
+      "loss": 0.7767,
+      "step": 22107
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4567410317386393,
+      "learning_rate": 9.810777110829096e-06,
+      "loss": 0.7651,
+      "step": 22108
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5167979314132911,
+      "learning_rate": 9.81001420693717e-06,
+      "loss": 0.8178,
+      "step": 22109
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3663867309100495,
+      "learning_rate": 9.809251304151404e-06,
+      "loss": 0.6911,
+      "step": 22110
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4033316435585406,
+      "learning_rate": 9.808488402476234e-06,
+      "loss": 0.686,
+      "step": 22111
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.183530752113148,
+      "learning_rate": 9.80772550191611e-06,
+      "loss": 0.7897,
+      "step": 22112
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7252961290169502,
+      "learning_rate": 9.806962602475466e-06,
+      "loss": 0.7138,
+      "step": 22113
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9480407680637006,
+      "learning_rate": 9.806199704158745e-06,
+      "loss": 0.7768,
+      "step": 22114
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4919195136201338,
+      "learning_rate": 9.805436806970397e-06,
+      "loss": 0.7378,
+      "step": 22115
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4564955341397252,
+      "learning_rate": 9.804673910914851e-06,
+      "loss": 0.7558,
+      "step": 22116
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6465548776188323,
+      "learning_rate": 9.803911015996557e-06,
+      "loss": 0.8875,
+      "step": 22117
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2738547589652314,
+      "learning_rate": 9.803148122219955e-06,
+      "loss": 0.6781,
+      "step": 22118
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.619582101390276,
+      "learning_rate": 9.802385229589493e-06,
+      "loss": 0.6482,
+      "step": 22119
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3406107454230538,
+      "learning_rate": 9.801622338109599e-06,
+      "loss": 0.5846,
+      "step": 22120
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4800788767683073,
+      "learning_rate": 9.800859447784726e-06,
+      "loss": 0.6689,
+      "step": 22121
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.279646626978623,
+      "learning_rate": 9.800096558619313e-06,
+      "loss": 0.6625,
+      "step": 22122
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6690338217339313,
+      "learning_rate": 9.799333670617797e-06,
+      "loss": 0.758,
+      "step": 22123
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0585529530863558,
+      "learning_rate": 9.798570783784627e-06,
+      "loss": 0.6563,
+      "step": 22124
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5414692600142315,
+      "learning_rate": 9.79780789812424e-06,
+      "loss": 0.7861,
+      "step": 22125
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4344161876677293,
+      "learning_rate": 9.797045013641084e-06,
+      "loss": 0.7302,
+      "step": 22126
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5053576423638424,
+      "learning_rate": 9.79628213033959e-06,
+      "loss": 0.7396,
+      "step": 22127
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2696509344281905,
+      "learning_rate": 9.79551924822421e-06,
+      "loss": 0.7457,
+      "step": 22128
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4738150830443628,
+      "learning_rate": 9.794756367299383e-06,
+      "loss": 0.7589,
+      "step": 22129
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4666733556911726,
+      "learning_rate": 9.793993487569544e-06,
+      "loss": 0.7452,
+      "step": 22130
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4147899021778796,
+      "learning_rate": 9.793230609039142e-06,
+      "loss": 0.77,
+      "step": 22131
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.369048893280283,
+      "learning_rate": 9.792467731712618e-06,
+      "loss": 0.6848,
+      "step": 22132
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3315313242639888,
+      "learning_rate": 9.791704855594415e-06,
+      "loss": 0.6913,
+      "step": 22133
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7302442211347904,
+      "learning_rate": 9.79094198068897e-06,
+      "loss": 0.7767,
+      "step": 22134
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6037275866106863,
+      "learning_rate": 9.790179107000723e-06,
+      "loss": 0.4981,
+      "step": 22135
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3227708516605592,
+      "learning_rate": 9.789416234534126e-06,
+      "loss": 0.5048,
+      "step": 22136
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3982974360976286,
+      "learning_rate": 9.788653363293612e-06,
+      "loss": 0.662,
+      "step": 22137
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.595814509963699,
+      "learning_rate": 9.787890493283626e-06,
+      "loss": 0.5593,
+      "step": 22138
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5755213904142673,
+      "learning_rate": 9.787127624508604e-06,
+      "loss": 0.5352,
+      "step": 22139
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5981523782108342,
+      "learning_rate": 9.786364756972997e-06,
+      "loss": 0.5734,
+      "step": 22140
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8035250567106613,
+      "learning_rate": 9.785601890681242e-06,
+      "loss": 0.6835,
+      "step": 22141
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3799309303473468,
+      "learning_rate": 9.78483902563778e-06,
+      "loss": 0.6709,
+      "step": 22142
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6265734549267258,
+      "learning_rate": 9.784076161847055e-06,
+      "loss": 0.5191,
+      "step": 22143
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.463726322242534,
+      "learning_rate": 9.783313299313502e-06,
+      "loss": 0.7359,
+      "step": 22144
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3375757005994688,
+      "learning_rate": 9.78255043804157e-06,
+      "loss": 0.6608,
+      "step": 22145
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5829090443017761,
+      "learning_rate": 9.781787578035698e-06,
+      "loss": 0.6923,
+      "step": 22146
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6288473276167557,
+      "learning_rate": 9.781024719300332e-06,
+      "loss": 0.6713,
+      "step": 22147
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.712718591766087,
+      "learning_rate": 9.780261861839903e-06,
+      "loss": 0.89,
+      "step": 22148
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.499595328401775,
+      "learning_rate": 9.779499005658862e-06,
+      "loss": 0.7366,
+      "step": 22149
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.612522313224308,
+      "learning_rate": 9.77873615076165e-06,
+      "loss": 0.5311,
+      "step": 22150
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2996353581807831,
+      "learning_rate": 9.777973297152703e-06,
+      "loss": 0.7174,
+      "step": 22151
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.577800781332203,
+      "learning_rate": 9.777210444836463e-06,
+      "loss": 0.503,
+      "step": 22152
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.202975235435661,
+      "learning_rate": 9.77644759381738e-06,
+      "loss": 0.7605,
+      "step": 22153
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3505949164595137,
+      "learning_rate": 9.775684744099889e-06,
+      "loss": 0.7535,
+      "step": 22154
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1213324398627527,
+      "learning_rate": 9.77492189568843e-06,
+      "loss": 0.7338,
+      "step": 22155
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5835556340591894,
+      "learning_rate": 9.774159048587445e-06,
+      "loss": 0.7373,
+      "step": 22156
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3832360226223743,
+      "learning_rate": 9.773396202801383e-06,
+      "loss": 0.7254,
+      "step": 22157
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4505173029287162,
+      "learning_rate": 9.772633358334678e-06,
+      "loss": 0.7266,
+      "step": 22158
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3975910844293769,
+      "learning_rate": 9.771870515191773e-06,
+      "loss": 0.6985,
+      "step": 22159
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3684753508686716,
+      "learning_rate": 9.771107673377106e-06,
+      "loss": 0.7224,
+      "step": 22160
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2979785610360315,
+      "learning_rate": 9.77034483289513e-06,
+      "loss": 0.7056,
+      "step": 22161
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4360633951978858,
+      "learning_rate": 9.769581993750277e-06,
+      "loss": 0.6522,
+      "step": 22162
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5940851244318837,
+      "learning_rate": 9.768819155946986e-06,
+      "loss": 0.6339,
+      "step": 22163
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9673002874943923,
+      "learning_rate": 9.768056319489711e-06,
+      "loss": 0.7712,
+      "step": 22164
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5330762671442133,
+      "learning_rate": 9.767293484382879e-06,
+      "loss": 0.678,
+      "step": 22165
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6062584921497317,
+      "learning_rate": 9.76653065063094e-06,
+      "loss": 0.5247,
+      "step": 22166
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6744186588155243,
+      "learning_rate": 9.76576781823833e-06,
+      "loss": 0.7284,
+      "step": 22167
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6629262466623398,
+      "learning_rate": 9.765004987209504e-06,
+      "loss": 0.655,
+      "step": 22168
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4853911202161725,
+      "learning_rate": 9.764242157548884e-06,
+      "loss": 0.6217,
+      "step": 22169
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0165889553494,
+      "learning_rate": 9.763479329260924e-06,
+      "loss": 0.7207,
+      "step": 22170
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3330969605901088,
+      "learning_rate": 9.762716502350058e-06,
+      "loss": 0.7682,
+      "step": 22171
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6479106961155975,
+      "learning_rate": 9.761953676820738e-06,
+      "loss": 0.5594,
+      "step": 22172
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3561453491361257,
+      "learning_rate": 9.761190852677397e-06,
+      "loss": 0.7007,
+      "step": 22173
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.512876081826534,
+      "learning_rate": 9.760428029924479e-06,
+      "loss": 0.8765,
+      "step": 22174
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.275783290891928,
+      "learning_rate": 9.759665208566425e-06,
+      "loss": 0.7652,
+      "step": 22175
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.630125103334511,
+      "learning_rate": 9.758902388607672e-06,
+      "loss": 0.5738,
+      "step": 22176
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.324263643250872,
+      "learning_rate": 9.75813957005267e-06,
+      "loss": 0.812,
+      "step": 22177
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5095051795101109,
+      "learning_rate": 9.757376752905854e-06,
+      "loss": 0.7151,
+      "step": 22178
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4722306885315675,
+      "learning_rate": 9.756613937171669e-06,
+      "loss": 0.7653,
+      "step": 22179
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5430008183237234,
+      "learning_rate": 9.755851122854553e-06,
+      "loss": 0.6081,
+      "step": 22180
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.871813916009758,
+      "learning_rate": 9.755088309958949e-06,
+      "loss": 0.668,
+      "step": 22181
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5925186888942202,
+      "learning_rate": 9.7543254984893e-06,
+      "loss": 0.6384,
+      "step": 22182
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6241301334950735,
+      "learning_rate": 9.753562688450047e-06,
+      "loss": 0.549,
+      "step": 22183
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.355628585504077,
+      "learning_rate": 9.752799879845624e-06,
+      "loss": 0.7374,
+      "step": 22184
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3755088794199855,
+      "learning_rate": 9.752037072680484e-06,
+      "loss": 0.7493,
+      "step": 22185
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3856027305332368,
+      "learning_rate": 9.751274266959064e-06,
+      "loss": 0.7594,
+      "step": 22186
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3498830666138757,
+      "learning_rate": 9.7505114626858e-06,
+      "loss": 0.5836,
+      "step": 22187
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4146346819282554,
+      "learning_rate": 9.749748659865136e-06,
+      "loss": 0.6601,
+      "step": 22188
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8585808080293,
+      "learning_rate": 9.74898585850152e-06,
+      "loss": 0.8171,
+      "step": 22189
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3739455590180134,
+      "learning_rate": 9.748223058599383e-06,
+      "loss": 0.706,
+      "step": 22190
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4099633482296359,
+      "learning_rate": 9.747460260163173e-06,
+      "loss": 0.9076,
+      "step": 22191
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4585868192447504,
+      "learning_rate": 9.746697463197327e-06,
+      "loss": 0.6875,
+      "step": 22192
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9483622686497726,
+      "learning_rate": 9.745934667706291e-06,
+      "loss": 0.7739,
+      "step": 22193
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6817782331164102,
+      "learning_rate": 9.745171873694504e-06,
+      "loss": 0.7307,
+      "step": 22194
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.991990082388415,
+      "learning_rate": 9.744409081166406e-06,
+      "loss": 0.6808,
+      "step": 22195
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.175474098373127,
+      "learning_rate": 9.743646290126444e-06,
+      "loss": 0.6763,
+      "step": 22196
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4500782210509182,
+      "learning_rate": 9.742883500579046e-06,
+      "loss": 0.5724,
+      "step": 22197
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2788429221280935,
+      "learning_rate": 9.742120712528665e-06,
+      "loss": 0.6807,
+      "step": 22198
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5662585681366246,
+      "learning_rate": 9.741357925979737e-06,
+      "loss": 0.8571,
+      "step": 22199
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3582528550916606,
+      "learning_rate": 9.740595140936714e-06,
+      "loss": 0.6253,
+      "step": 22200
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3873155110600202,
+      "learning_rate": 9.73983235740402e-06,
+      "loss": 0.7345,
+      "step": 22201
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4301165536210767,
+      "learning_rate": 9.739069575386105e-06,
+      "loss": 0.7611,
+      "step": 22202
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8665822538873957,
+      "learning_rate": 9.738306794887413e-06,
+      "loss": 0.7032,
+      "step": 22203
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4638041706339804,
+      "learning_rate": 9.73754401591238e-06,
+      "loss": 0.7879,
+      "step": 22204
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5958860285711841,
+      "learning_rate": 9.736781238465444e-06,
+      "loss": 0.5163,
+      "step": 22205
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4642619283364555,
+      "learning_rate": 9.736018462551057e-06,
+      "loss": 0.695,
+      "step": 22206
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4180029889042218,
+      "learning_rate": 9.735255688173654e-06,
+      "loss": 0.7445,
+      "step": 22207
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4703197268625188,
+      "learning_rate": 9.734492915337675e-06,
+      "loss": 0.7901,
+      "step": 22208
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5550009645316814,
+      "learning_rate": 9.733730144047558e-06,
+      "loss": 0.6754,
+      "step": 22209
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.82693754900246,
+      "learning_rate": 9.732967374307755e-06,
+      "loss": 0.5658,
+      "step": 22210
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2944875831285005,
+      "learning_rate": 9.732204606122698e-06,
+      "loss": 0.6347,
+      "step": 22211
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0202295579507887,
+      "learning_rate": 9.731441839496828e-06,
+      "loss": 0.6815,
+      "step": 22212
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.403111542663189,
+      "learning_rate": 9.73067907443459e-06,
+      "loss": 0.7302,
+      "step": 22213
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4259075562824286,
+      "learning_rate": 9.729916310940427e-06,
+      "loss": 0.6959,
+      "step": 22214
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.396622291904864,
+      "learning_rate": 9.729153549018775e-06,
+      "loss": 0.8498,
+      "step": 22215
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2735886368459781,
+      "learning_rate": 9.728390788674074e-06,
+      "loss": 0.6808,
+      "step": 22216
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.185118639604415,
+      "learning_rate": 9.727628029910774e-06,
+      "loss": 0.6421,
+      "step": 22217
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0317281214904206,
+      "learning_rate": 9.726865272733304e-06,
+      "loss": 0.684,
+      "step": 22218
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1474729998568352,
+      "learning_rate": 9.726102517146112e-06,
+      "loss": 0.7253,
+      "step": 22219
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3549282178276558,
+      "learning_rate": 9.725339763153636e-06,
+      "loss": 0.6138,
+      "step": 22220
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5139005472419254,
+      "learning_rate": 9.724577010760324e-06,
+      "loss": 0.7586,
+      "step": 22221
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2057427937632579,
+      "learning_rate": 9.723814259970608e-06,
+      "loss": 0.6279,
+      "step": 22222
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4612314173837253,
+      "learning_rate": 9.723051510788934e-06,
+      "loss": 0.901,
+      "step": 22223
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7468703295449954,
+      "learning_rate": 9.722288763219744e-06,
+      "loss": 0.7073,
+      "step": 22224
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4466562002439123,
+      "learning_rate": 9.721526017267473e-06,
+      "loss": 0.7513,
+      "step": 22225
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2507168078038966,
+      "learning_rate": 9.720763272936568e-06,
+      "loss": 0.6608,
+      "step": 22226
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4963785446903044,
+      "learning_rate": 9.720000530231466e-06,
+      "loss": 0.7663,
+      "step": 22227
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2728947518800453,
+      "learning_rate": 9.719237789156612e-06,
+      "loss": 0.7234,
+      "step": 22228
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9676535660611145,
+      "learning_rate": 9.718475049716441e-06,
+      "loss": 0.754,
+      "step": 22229
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.409347121663833,
+      "learning_rate": 9.7177123119154e-06,
+      "loss": 0.6821,
+      "step": 22230
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.669571199958091,
+      "learning_rate": 9.71694957575793e-06,
+      "loss": 0.6701,
+      "step": 22231
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4612709957820333,
+      "learning_rate": 9.716186841248465e-06,
+      "loss": 0.6456,
+      "step": 22232
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5187144033923383,
+      "learning_rate": 9.715424108391448e-06,
+      "loss": 0.6857,
+      "step": 22233
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5313127872072796,
+      "learning_rate": 9.714661377191324e-06,
+      "loss": 0.6675,
+      "step": 22234
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5884312449495083,
+      "learning_rate": 9.713898647652535e-06,
+      "loss": 0.5067,
+      "step": 22235
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5883950018746342,
+      "learning_rate": 9.713135919779516e-06,
+      "loss": 0.5519,
+      "step": 22236
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4757554418750465,
+      "learning_rate": 9.712373193576708e-06,
+      "loss": 0.7056,
+      "step": 22237
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.210695544990854,
+      "learning_rate": 9.71161046904856e-06,
+      "loss": 0.6293,
+      "step": 22238
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6997335910331537,
+      "learning_rate": 9.710847746199505e-06,
+      "loss": 0.6125,
+      "step": 22239
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5700724013080334,
+      "learning_rate": 9.710085025033984e-06,
+      "loss": 0.6371,
+      "step": 22240
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4968633951593282,
+      "learning_rate": 9.709322305556438e-06,
+      "loss": 0.7723,
+      "step": 22241
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4651507290509438,
+      "learning_rate": 9.708559587771316e-06,
+      "loss": 0.7129,
+      "step": 22242
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4324197001041823,
+      "learning_rate": 9.707796871683049e-06,
+      "loss": 0.7227,
+      "step": 22243
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2293808373240551,
+      "learning_rate": 9.70703415729608e-06,
+      "loss": 0.7194,
+      "step": 22244
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5223539031699165,
+      "learning_rate": 9.706271444614854e-06,
+      "loss": 0.7627,
+      "step": 22245
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5932047686596619,
+      "learning_rate": 9.705508733643805e-06,
+      "loss": 0.5416,
+      "step": 22246
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.565465582656481,
+      "learning_rate": 9.704746024387377e-06,
+      "loss": 0.7354,
+      "step": 22247
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4592367004157332,
+      "learning_rate": 9.703983316850012e-06,
+      "loss": 0.7477,
+      "step": 22248
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4018202863399496,
+      "learning_rate": 9.703220611036156e-06,
+      "loss": 0.7786,
+      "step": 22249
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5262433598720377,
+      "learning_rate": 9.702457906950235e-06,
+      "loss": 0.6833,
+      "step": 22250
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4077966715610402,
+      "learning_rate": 9.701695204596702e-06,
+      "loss": 0.7773,
+      "step": 22251
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3734382628231174,
+      "learning_rate": 9.700932503979996e-06,
+      "loss": 0.6349,
+      "step": 22252
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4933926180319728,
+      "learning_rate": 9.700169805104553e-06,
+      "loss": 0.7376,
+      "step": 22253
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5772168522653769,
+      "learning_rate": 9.699407107974815e-06,
+      "loss": 0.5008,
+      "step": 22254
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3740571711225644,
+      "learning_rate": 9.698644412595225e-06,
+      "loss": 0.6866,
+      "step": 22255
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.844425884760357,
+      "learning_rate": 9.697881718970226e-06,
+      "loss": 0.7832,
+      "step": 22256
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.383622804789294,
+      "learning_rate": 9.697119027104252e-06,
+      "loss": 0.7421,
+      "step": 22257
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6943731665042485,
+      "learning_rate": 9.696356337001745e-06,
+      "loss": 0.6524,
+      "step": 22258
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2274225922252393,
+      "learning_rate": 9.695593648667152e-06,
+      "loss": 0.7866,
+      "step": 22259
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5097109086795273,
+      "learning_rate": 9.694830962104907e-06,
+      "loss": 0.7219,
+      "step": 22260
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7207095591943593,
+      "learning_rate": 9.694068277319451e-06,
+      "loss": 0.6322,
+      "step": 22261
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3418836062059443,
+      "learning_rate": 9.693305594315229e-06,
+      "loss": 0.7795,
+      "step": 22262
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0145393384643464,
+      "learning_rate": 9.69254291309668e-06,
+      "loss": 0.684,
+      "step": 22263
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.363887651571428,
+      "learning_rate": 9.691780233668241e-06,
+      "loss": 0.6287,
+      "step": 22264
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6021499291914638,
+      "learning_rate": 9.691017556034354e-06,
+      "loss": 0.7136,
+      "step": 22265
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3657262112777846,
+      "learning_rate": 9.690254880199466e-06,
+      "loss": 0.8063,
+      "step": 22266
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4032590066845567,
+      "learning_rate": 9.689492206168006e-06,
+      "loss": 0.7231,
+      "step": 22267
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4455373936325966,
+      "learning_rate": 9.688729533944423e-06,
+      "loss": 0.7293,
+      "step": 22268
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5612157497697488,
+      "learning_rate": 9.687966863533154e-06,
+      "loss": 0.6564,
+      "step": 22269
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6736704682017658,
+      "learning_rate": 9.687204194938645e-06,
+      "loss": 0.6412,
+      "step": 22270
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4512469092511822,
+      "learning_rate": 9.686441528165327e-06,
+      "loss": 0.7395,
+      "step": 22271
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.56100538616808,
+      "learning_rate": 9.685678863217649e-06,
+      "loss": 0.7123,
+      "step": 22272
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5923859288694733,
+      "learning_rate": 9.68491620010005e-06,
+      "loss": 0.8002,
+      "step": 22273
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1884914046113093,
+      "learning_rate": 9.684153538816963e-06,
+      "loss": 0.7366,
+      "step": 22274
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3866666086394335,
+      "learning_rate": 9.683390879372837e-06,
+      "loss": 0.6719,
+      "step": 22275
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.434390908596401,
+      "learning_rate": 9.682628221772109e-06,
+      "loss": 0.7213,
+      "step": 22276
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4710700221994313,
+      "learning_rate": 9.681865566019224e-06,
+      "loss": 0.7905,
+      "step": 22277
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5489228203938628,
+      "learning_rate": 9.68110291211861e-06,
+      "loss": 0.7338,
+      "step": 22278
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2049832857277547,
+      "learning_rate": 9.680340260074721e-06,
+      "loss": 0.7574,
+      "step": 22279
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5240462982030796,
+      "learning_rate": 9.679577609891994e-06,
+      "loss": 0.8427,
+      "step": 22280
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.417896714520269,
+      "learning_rate": 9.678814961574866e-06,
+      "loss": 0.7905,
+      "step": 22281
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5122858544234326,
+      "learning_rate": 9.678052315127776e-06,
+      "loss": 0.7318,
+      "step": 22282
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4734708761301007,
+      "learning_rate": 9.67728967055517e-06,
+      "loss": 0.7549,
+      "step": 22283
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.599056020861403,
+      "learning_rate": 9.676527027861487e-06,
+      "loss": 0.72,
+      "step": 22284
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.620089111582816,
+      "learning_rate": 9.675764387051164e-06,
+      "loss": 0.6797,
+      "step": 22285
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.194802133953243,
+      "learning_rate": 9.675001748128641e-06,
+      "loss": 0.6632,
+      "step": 22286
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.506136316044918,
+      "learning_rate": 9.674239111098367e-06,
+      "loss": 0.6265,
+      "step": 22287
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.351566078664164,
+      "learning_rate": 9.673476475964773e-06,
+      "loss": 0.7199,
+      "step": 22288
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.299359799610735,
+      "learning_rate": 9.6727138427323e-06,
+      "loss": 0.6552,
+      "step": 22289
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3240571207021845,
+      "learning_rate": 9.671951211405391e-06,
+      "loss": 0.6389,
+      "step": 22290
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6676796155505766,
+      "learning_rate": 9.67118858198849e-06,
+      "loss": 0.6472,
+      "step": 22291
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.321658817252861,
+      "learning_rate": 9.67042595448603e-06,
+      "loss": 0.7699,
+      "step": 22292
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5926770232335987,
+      "learning_rate": 9.669663328902455e-06,
+      "loss": 0.5319,
+      "step": 22293
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5178970462705592,
+      "learning_rate": 9.668900705242208e-06,
+      "loss": 0.8299,
+      "step": 22294
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4845434102914854,
+      "learning_rate": 9.66813808350972e-06,
+      "loss": 0.6392,
+      "step": 22295
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4051403680343892,
+      "learning_rate": 9.66737546370944e-06,
+      "loss": 0.6879,
+      "step": 22296
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4559887216532879,
+      "learning_rate": 9.666612845845803e-06,
+      "loss": 0.7719,
+      "step": 22297
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6358634497403757,
+      "learning_rate": 9.665850229923258e-06,
+      "loss": 0.7802,
+      "step": 22298
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8332994842006545,
+      "learning_rate": 9.665087615946231e-06,
+      "loss": 0.6272,
+      "step": 22299
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0932160312425294,
+      "learning_rate": 9.664325003919176e-06,
+      "loss": 0.7725,
+      "step": 22300
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3141875613631617,
+      "learning_rate": 9.66356239384652e-06,
+      "loss": 0.7189,
+      "step": 22301
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8480595659702084,
+      "learning_rate": 9.66279978573272e-06,
+      "loss": 0.7361,
+      "step": 22302
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.613877385388613,
+      "learning_rate": 9.662037179582199e-06,
+      "loss": 0.7484,
+      "step": 22303
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.447054695899046,
+      "learning_rate": 9.661274575399406e-06,
+      "loss": 0.7749,
+      "step": 22304
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5294282228859963,
+      "learning_rate": 9.660511973188783e-06,
+      "loss": 0.638,
+      "step": 22305
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6578991311558549,
+      "learning_rate": 9.659749372954764e-06,
+      "loss": 0.8014,
+      "step": 22306
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6303484119047378,
+      "learning_rate": 9.65898677470179e-06,
+      "loss": 0.5876,
+      "step": 22307
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.468502981929465,
+      "learning_rate": 9.658224178434305e-06,
+      "loss": 0.7316,
+      "step": 22308
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.517595001267095,
+      "learning_rate": 9.657461584156748e-06,
+      "loss": 0.6528,
+      "step": 22309
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4292267654115525,
+      "learning_rate": 9.656698991873555e-06,
+      "loss": 0.7059,
+      "step": 22310
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.703439209847221,
+      "learning_rate": 9.655936401589173e-06,
+      "loss": 0.7036,
+      "step": 22311
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4148256087081414,
+      "learning_rate": 9.655173813308037e-06,
+      "loss": 0.6572,
+      "step": 22312
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4045444534608795,
+      "learning_rate": 9.654411227034588e-06,
+      "loss": 0.7483,
+      "step": 22313
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.461919579632079,
+      "learning_rate": 9.653648642773264e-06,
+      "loss": 0.742,
+      "step": 22314
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4066728190653928,
+      "learning_rate": 9.652886060528509e-06,
+      "loss": 0.6905,
+      "step": 22315
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.937909384742153,
+      "learning_rate": 9.652123480304764e-06,
+      "loss": 0.6749,
+      "step": 22316
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4767652730426326,
+      "learning_rate": 9.651360902106463e-06,
+      "loss": 0.695,
+      "step": 22317
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.433924593630992,
+      "learning_rate": 9.650598325938048e-06,
+      "loss": 0.8557,
+      "step": 22318
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.296590326388873,
+      "learning_rate": 9.649835751803965e-06,
+      "loss": 0.7618,
+      "step": 22319
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4013004332353893,
+      "learning_rate": 9.649073179708646e-06,
+      "loss": 0.7475,
+      "step": 22320
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4502025510646646,
+      "learning_rate": 9.648310609656534e-06,
+      "loss": 0.8554,
+      "step": 22321
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.009447021384184,
+      "learning_rate": 9.647548041652066e-06,
+      "loss": 0.7415,
+      "step": 22322
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3396165286663009,
+      "learning_rate": 9.646785475699691e-06,
+      "loss": 0.6708,
+      "step": 22323
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6569276022417723,
+      "learning_rate": 9.64602291180384e-06,
+      "loss": 0.7659,
+      "step": 22324
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7072956342726395,
+      "learning_rate": 9.645260349968955e-06,
+      "loss": 0.8063,
+      "step": 22325
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3939105347594607,
+      "learning_rate": 9.644497790199478e-06,
+      "loss": 0.7456,
+      "step": 22326
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4336601497621722,
+      "learning_rate": 9.643735232499846e-06,
+      "loss": 0.6743,
+      "step": 22327
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4049026425543973,
+      "learning_rate": 9.6429726768745e-06,
+      "loss": 0.6629,
+      "step": 22328
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9940227870178144,
+      "learning_rate": 9.642210123327882e-06,
+      "loss": 0.8204,
+      "step": 22329
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2207014253127664,
+      "learning_rate": 9.641447571864429e-06,
+      "loss": 0.5577,
+      "step": 22330
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4922115909263194,
+      "learning_rate": 9.64068502248858e-06,
+      "loss": 0.7712,
+      "step": 22331
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.432846914605378,
+      "learning_rate": 9.639922475204778e-06,
+      "loss": 0.6471,
+      "step": 22332
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4541776753491522,
+      "learning_rate": 9.639159930017465e-06,
+      "loss": 0.8039,
+      "step": 22333
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3530911834431687,
+      "learning_rate": 9.638397386931072e-06,
+      "loss": 0.6981,
+      "step": 22334
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.294496045668733,
+      "learning_rate": 9.637634845950043e-06,
+      "loss": 0.6988,
+      "step": 22335
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6091777458211379,
+      "learning_rate": 9.636872307078821e-06,
+      "loss": 0.5315,
+      "step": 22336
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4976096711541105,
+      "learning_rate": 9.636109770321846e-06,
+      "loss": 0.7288,
+      "step": 22337
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9942434078172795,
+      "learning_rate": 9.635347235683552e-06,
+      "loss": 0.8773,
+      "step": 22338
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6467746950126085,
+      "learning_rate": 9.634584703168381e-06,
+      "loss": 0.6248,
+      "step": 22339
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4780703397335997,
+      "learning_rate": 9.633822172780778e-06,
+      "loss": 0.6299,
+      "step": 22340
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4535828018290584,
+      "learning_rate": 9.633059644525175e-06,
+      "loss": 0.6936,
+      "step": 22341
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5721993492263777,
+      "learning_rate": 9.632297118406015e-06,
+      "loss": 0.7636,
+      "step": 22342
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5454220154101594,
+      "learning_rate": 9.631534594427735e-06,
+      "loss": 0.7739,
+      "step": 22343
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5695912426684833,
+      "learning_rate": 9.630772072594784e-06,
+      "loss": 0.719,
+      "step": 22344
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7876851899668356,
+      "learning_rate": 9.63000955291159e-06,
+      "loss": 0.8023,
+      "step": 22345
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4413128798736916,
+      "learning_rate": 9.629247035382597e-06,
+      "loss": 0.6607,
+      "step": 22346
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3362604903100563,
+      "learning_rate": 9.62848452001225e-06,
+      "loss": 0.7634,
+      "step": 22347
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2532709224027205,
+      "learning_rate": 9.627722006804977e-06,
+      "loss": 0.6517,
+      "step": 22348
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4694513301828833,
+      "learning_rate": 9.626959495765229e-06,
+      "loss": 0.7137,
+      "step": 22349
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.461673906808929,
+      "learning_rate": 9.626196986897437e-06,
+      "loss": 0.6702,
+      "step": 22350
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.808007263599861,
+      "learning_rate": 9.62543448020605e-06,
+      "loss": 0.7157,
+      "step": 22351
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4887290450509298,
+      "learning_rate": 9.624671975695497e-06,
+      "loss": 0.7105,
+      "step": 22352
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.688524610658029,
+      "learning_rate": 9.623909473370225e-06,
+      "loss": 0.7332,
+      "step": 22353
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2556025857258655,
+      "learning_rate": 9.623146973234673e-06,
+      "loss": 0.6828,
+      "step": 22354
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3900260329251417,
+      "learning_rate": 9.622384475293278e-06,
+      "loss": 0.8817,
+      "step": 22355
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.383541777560759,
+      "learning_rate": 9.621621979550475e-06,
+      "loss": 0.6698,
+      "step": 22356
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4131830314899776,
+      "learning_rate": 9.620859486010713e-06,
+      "loss": 0.8266,
+      "step": 22357
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7315938659390102,
+      "learning_rate": 9.620096994678429e-06,
+      "loss": 0.7467,
+      "step": 22358
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2922153707175896,
+      "learning_rate": 9.619334505558055e-06,
+      "loss": 0.7795,
+      "step": 22359
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.512477872567027,
+      "learning_rate": 9.618572018654039e-06,
+      "loss": 0.7373,
+      "step": 22360
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2646589896758793,
+      "learning_rate": 9.61780953397082e-06,
+      "loss": 0.6329,
+      "step": 22361
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3354154058818644,
+      "learning_rate": 9.617047051512832e-06,
+      "loss": 0.6799,
+      "step": 22362
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.380378634265154,
+      "learning_rate": 9.616284571284515e-06,
+      "loss": 0.8429,
+      "step": 22363
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4974204006125522,
+      "learning_rate": 9.615522093290313e-06,
+      "loss": 0.7264,
+      "step": 22364
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5789667940848485,
+      "learning_rate": 9.614759617534665e-06,
+      "loss": 0.5176,
+      "step": 22365
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4806615920037418,
+      "learning_rate": 9.613997144022006e-06,
+      "loss": 0.6543,
+      "step": 22366
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6392694051977448,
+      "learning_rate": 9.613234672756777e-06,
+      "loss": 0.838,
+      "step": 22367
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2918290857414982,
+      "learning_rate": 9.612472203743423e-06,
+      "loss": 0.7151,
+      "step": 22368
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.534141137887145,
+      "learning_rate": 9.611709736986375e-06,
+      "loss": 0.7721,
+      "step": 22369
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3559508627776817,
+      "learning_rate": 9.610947272490077e-06,
+      "loss": 0.8636,
+      "step": 22370
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5834207920641198,
+      "learning_rate": 9.610184810258964e-06,
+      "loss": 0.5086,
+      "step": 22371
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.603549528159211,
+      "learning_rate": 9.609422350297482e-06,
+      "loss": 0.7686,
+      "step": 22372
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4029604152012787,
+      "learning_rate": 9.608659892610067e-06,
+      "loss": 0.752,
+      "step": 22373
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.329432920085414,
+      "learning_rate": 9.607897437201155e-06,
+      "loss": 0.6003,
+      "step": 22374
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5848790942249755,
+      "learning_rate": 9.607134984075193e-06,
+      "loss": 0.7161,
+      "step": 22375
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4186127250155678,
+      "learning_rate": 9.60637253323661e-06,
+      "loss": 0.7083,
+      "step": 22376
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8695063501021763,
+      "learning_rate": 9.605610084689853e-06,
+      "loss": 0.6909,
+      "step": 22377
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4680291931042413,
+      "learning_rate": 9.604847638439358e-06,
+      "loss": 0.7094,
+      "step": 22378
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6035526962747109,
+      "learning_rate": 9.604085194489568e-06,
+      "loss": 0.5182,
+      "step": 22379
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4187259404196975,
+      "learning_rate": 9.603322752844917e-06,
+      "loss": 0.6623,
+      "step": 22380
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5738743718987416,
+      "learning_rate": 9.602560313509846e-06,
+      "loss": 0.719,
+      "step": 22381
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.785263369994858,
+      "learning_rate": 9.601797876488798e-06,
+      "loss": 0.7107,
+      "step": 22382
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6647817668183165,
+      "learning_rate": 9.601035441786206e-06,
+      "loss": 0.6325,
+      "step": 22383
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4913275296101336,
+      "learning_rate": 9.60027300940651e-06,
+      "loss": 0.9009,
+      "step": 22384
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.358664925612962,
+      "learning_rate": 9.599510579354153e-06,
+      "loss": 0.6875,
+      "step": 22385
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5472588190322039,
+      "learning_rate": 9.598748151633576e-06,
+      "loss": 0.6411,
+      "step": 22386
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5844576567503438,
+      "learning_rate": 9.597985726249211e-06,
+      "loss": 0.5458,
+      "step": 22387
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5165710056004402,
+      "learning_rate": 9.5972233032055e-06,
+      "loss": 0.7288,
+      "step": 22388
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2792955587376673,
+      "learning_rate": 9.596460882506885e-06,
+      "loss": 0.6489,
+      "step": 22389
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3965164743269747,
+      "learning_rate": 9.595698464157799e-06,
+      "loss": 0.8428,
+      "step": 22390
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9948160098782532,
+      "learning_rate": 9.594936048162685e-06,
+      "loss": 0.6337,
+      "step": 22391
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7014497347477235,
+      "learning_rate": 9.59417363452598e-06,
+      "loss": 0.6789,
+      "step": 22392
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4312701073116096,
+      "learning_rate": 9.59341122325213e-06,
+      "loss": 0.7338,
+      "step": 22393
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3930323894242091,
+      "learning_rate": 9.592648814345565e-06,
+      "loss": 0.6369,
+      "step": 22394
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7350748570283938,
+      "learning_rate": 9.591886407810727e-06,
+      "loss": 0.696,
+      "step": 22395
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.49951972377217,
+      "learning_rate": 9.591124003652062e-06,
+      "loss": 0.7238,
+      "step": 22396
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2953555217455908,
+      "learning_rate": 9.590361601873994e-06,
+      "loss": 0.6975,
+      "step": 22397
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.415612285252854,
+      "learning_rate": 9.589599202480974e-06,
+      "loss": 0.6555,
+      "step": 22398
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.557235328304591,
+      "learning_rate": 9.588836805477435e-06,
+      "loss": 0.783,
+      "step": 22399
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.391817796078095,
+      "learning_rate": 9.588074410867825e-06,
+      "loss": 0.6864,
+      "step": 22400
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2566540855548944,
+      "learning_rate": 9.587312018656568e-06,
+      "loss": 0.603,
+      "step": 22401
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4884830914221259,
+      "learning_rate": 9.586549628848115e-06,
+      "loss": 0.8331,
+      "step": 22402
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.768573241298432,
+      "learning_rate": 9.585787241446903e-06,
+      "loss": 0.7312,
+      "step": 22403
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3058179487902921,
+      "learning_rate": 9.585024856457367e-06,
+      "loss": 0.7826,
+      "step": 22404
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7179614698582861,
+      "learning_rate": 9.584262473883945e-06,
+      "loss": 0.7887,
+      "step": 22405
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8091681886857023,
+      "learning_rate": 9.58350009373108e-06,
+      "loss": 0.7589,
+      "step": 22406
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6457455305771511,
+      "learning_rate": 9.582737716003212e-06,
+      "loss": 0.6422,
+      "step": 22407
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6037678001354535,
+      "learning_rate": 9.581975340704772e-06,
+      "loss": 0.5459,
+      "step": 22408
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3890453433222598,
+      "learning_rate": 9.581212967840208e-06,
+      "loss": 0.7225,
+      "step": 22409
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7076735577128257,
+      "learning_rate": 9.580450597413955e-06,
+      "loss": 0.8002,
+      "step": 22410
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.497896832958898,
+      "learning_rate": 9.57968822943045e-06,
+      "loss": 0.7371,
+      "step": 22411
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3315138266448923,
+      "learning_rate": 9.578925863894132e-06,
+      "loss": 0.7105,
+      "step": 22412
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.402027543836479,
+      "learning_rate": 9.578163500809443e-06,
+      "loss": 0.7557,
+      "step": 22413
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6609787125406847,
+      "learning_rate": 9.577401140180819e-06,
+      "loss": 0.7499,
+      "step": 22414
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3603805549942156,
+      "learning_rate": 9.5766387820127e-06,
+      "loss": 0.6775,
+      "step": 22415
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4938311777940259,
+      "learning_rate": 9.575876426309521e-06,
+      "loss": 0.7024,
+      "step": 22416
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5787556199058972,
+      "learning_rate": 9.575114073075729e-06,
+      "loss": 0.7081,
+      "step": 22417
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4641107776391777,
+      "learning_rate": 9.574351722315754e-06,
+      "loss": 0.6818,
+      "step": 22418
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.434714929257389,
+      "learning_rate": 9.573589374034038e-06,
+      "loss": 0.7819,
+      "step": 22419
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1995775629513996,
+      "learning_rate": 9.572827028235018e-06,
+      "loss": 0.7436,
+      "step": 22420
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1072561029153607,
+      "learning_rate": 9.572064684923138e-06,
+      "loss": 0.7166,
+      "step": 22421
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6831641717391714,
+      "learning_rate": 9.57130234410283e-06,
+      "loss": 0.7367,
+      "step": 22422
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5550033539996344,
+      "learning_rate": 9.570540005778539e-06,
+      "loss": 0.6601,
+      "step": 22423
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.871886703451714,
+      "learning_rate": 9.569777669954694e-06,
+      "loss": 0.7595,
+      "step": 22424
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.589576657462555,
+      "learning_rate": 9.569015336635746e-06,
+      "loss": 0.53,
+      "step": 22425
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.285235880654272,
+      "learning_rate": 9.568253005826124e-06,
+      "loss": 0.5964,
+      "step": 22426
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6410698442380627,
+      "learning_rate": 9.56749067753027e-06,
+      "loss": 0.6245,
+      "step": 22427
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5608581751796988,
+      "learning_rate": 9.566728351752626e-06,
+      "loss": 0.7323,
+      "step": 22428
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4976508142700107,
+      "learning_rate": 9.56596602849762e-06,
+      "loss": 0.8735,
+      "step": 22429
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7970031985308683,
+      "learning_rate": 9.5652037077697e-06,
+      "loss": 0.7582,
+      "step": 22430
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4347365610666205,
+      "learning_rate": 9.5644413895733e-06,
+      "loss": 0.6331,
+      "step": 22431
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2518305236335419,
+      "learning_rate": 9.563679073912866e-06,
+      "loss": 0.6877,
+      "step": 22432
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4929244570943954,
+      "learning_rate": 9.562916760792825e-06,
+      "loss": 0.7284,
+      "step": 22433
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5030956725836535,
+      "learning_rate": 9.562154450217623e-06,
+      "loss": 0.6632,
+      "step": 22434
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4761441162289715,
+      "learning_rate": 9.561392142191698e-06,
+      "loss": 0.6531,
+      "step": 22435
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.553933143666935,
+      "learning_rate": 9.560629836719485e-06,
+      "loss": 0.8155,
+      "step": 22436
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5773202464746443,
+      "learning_rate": 9.559867533805421e-06,
+      "loss": 0.605,
+      "step": 22437
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3885517865470745,
+      "learning_rate": 9.55910523345395e-06,
+      "loss": 0.8012,
+      "step": 22438
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4422977930736909,
+      "learning_rate": 9.558342935669512e-06,
+      "loss": 0.8185,
+      "step": 22439
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.322808078199481,
+      "learning_rate": 9.557580640456538e-06,
+      "loss": 0.6852,
+      "step": 22440
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.668837087280597,
+      "learning_rate": 9.556818347819467e-06,
+      "loss": 0.709,
+      "step": 22441
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5866216917487936,
+      "learning_rate": 9.556056057762746e-06,
+      "loss": 0.8047,
+      "step": 22442
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1930130894281556,
+      "learning_rate": 9.555293770290802e-06,
+      "loss": 0.7291,
+      "step": 22443
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.677542889797731,
+      "learning_rate": 9.554531485408078e-06,
+      "loss": 0.8196,
+      "step": 22444
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3936814859036408,
+      "learning_rate": 9.553769203119014e-06,
+      "loss": 0.7588,
+      "step": 22445
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.422990847268132,
+      "learning_rate": 9.55300692342805e-06,
+      "loss": 0.7051,
+      "step": 22446
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5971350484848326,
+      "learning_rate": 9.552244646339619e-06,
+      "loss": 0.5337,
+      "step": 22447
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2217267784545998,
+      "learning_rate": 9.55148237185816e-06,
+      "loss": 0.6785,
+      "step": 22448
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2673511224488867,
+      "learning_rate": 9.550720099988117e-06,
+      "loss": 0.6738,
+      "step": 22449
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.771524935134995,
+      "learning_rate": 9.54995783073392e-06,
+      "loss": 0.8153,
+      "step": 22450
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.613320275010697,
+      "learning_rate": 9.54919556410001e-06,
+      "loss": 0.68,
+      "step": 22451
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4329919872791008,
+      "learning_rate": 9.548433300090826e-06,
+      "loss": 0.7072,
+      "step": 22452
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.557776676104403,
+      "learning_rate": 9.547671038710813e-06,
+      "loss": 0.7232,
+      "step": 22453
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5280442097474911,
+      "learning_rate": 9.546908779964394e-06,
+      "loss": 0.6826,
+      "step": 22454
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6605779932397606,
+      "learning_rate": 9.546146523856021e-06,
+      "loss": 0.7457,
+      "step": 22455
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1897314564519883,
+      "learning_rate": 9.545384270390125e-06,
+      "loss": 0.7146,
+      "step": 22456
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.344690015439224,
+      "learning_rate": 9.544622019571144e-06,
+      "loss": 0.817,
+      "step": 22457
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6004039979095885,
+      "learning_rate": 9.543859771403518e-06,
+      "loss": 0.6843,
+      "step": 22458
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.539324268751374,
+      "learning_rate": 9.543097525891686e-06,
+      "loss": 0.8226,
+      "step": 22459
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9702934527683598,
+      "learning_rate": 9.542335283040089e-06,
+      "loss": 0.7074,
+      "step": 22460
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9669862193804313,
+      "learning_rate": 9.541573042853153e-06,
+      "loss": 0.7623,
+      "step": 22461
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.528298806154955,
+      "learning_rate": 9.540810805335327e-06,
+      "loss": 0.8695,
+      "step": 22462
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5531877196309012,
+      "learning_rate": 9.540048570491049e-06,
+      "loss": 0.5246,
+      "step": 22463
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6012011388399116,
+      "learning_rate": 9.539286338324753e-06,
+      "loss": 0.5038,
+      "step": 22464
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5730801123085115,
+      "learning_rate": 9.538524108840873e-06,
+      "loss": 0.5483,
+      "step": 22465
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2855987773060948,
+      "learning_rate": 9.537761882043855e-06,
+      "loss": 0.7197,
+      "step": 22466
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.242700175612685,
+      "learning_rate": 9.536999657938135e-06,
+      "loss": 0.8148,
+      "step": 22467
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.562778386569422,
+      "learning_rate": 9.536237436528148e-06,
+      "loss": 0.4978,
+      "step": 22468
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.385842424406107,
+      "learning_rate": 9.535475217818331e-06,
+      "loss": 0.7103,
+      "step": 22469
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8506515652492728,
+      "learning_rate": 9.534713001813131e-06,
+      "loss": 0.7817,
+      "step": 22470
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5018349141644391,
+      "learning_rate": 9.533950788516974e-06,
+      "loss": 0.6982,
+      "step": 22471
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3544026151087236,
+      "learning_rate": 9.533188577934304e-06,
+      "loss": 0.5421,
+      "step": 22472
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3806803509190804,
+      "learning_rate": 9.532426370069557e-06,
+      "loss": 0.7297,
+      "step": 22473
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4252317087036794,
+      "learning_rate": 9.531664164927175e-06,
+      "loss": 0.6935,
+      "step": 22474
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.7037504264283063,
+      "learning_rate": 9.53090196251159e-06,
+      "loss": 0.7182,
+      "step": 22475
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.312773827146625,
+      "learning_rate": 9.530139762827243e-06,
+      "loss": 0.7114,
+      "step": 22476
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3585657754604596,
+      "learning_rate": 9.529377565878571e-06,
+      "loss": 0.7302,
+      "step": 22477
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.418856632662237,
+      "learning_rate": 9.52861537167001e-06,
+      "loss": 0.8475,
+      "step": 22478
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.6330791763819628,
+      "learning_rate": 9.527853180206e-06,
+      "loss": 0.7034,
+      "step": 22479
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6531250910273603,
+      "learning_rate": 9.527090991490977e-06,
+      "loss": 0.8675,
+      "step": 22480
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1487465115098345,
+      "learning_rate": 9.526328805529387e-06,
+      "loss": 0.6413,
+      "step": 22481
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6978610975084045,
+      "learning_rate": 9.525566622325653e-06,
+      "loss": 0.666,
+      "step": 22482
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4134683350287625,
+      "learning_rate": 9.524804441884224e-06,
+      "loss": 0.6157,
+      "step": 22483
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3345100380672286,
+      "learning_rate": 9.524042264209535e-06,
+      "loss": 0.6985,
+      "step": 22484
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7183109069780726,
+      "learning_rate": 9.523280089306022e-06,
+      "loss": 0.6565,
+      "step": 22485
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3573513725594188,
+      "learning_rate": 9.52251791717812e-06,
+      "loss": 0.6982,
+      "step": 22486
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.630670300952991,
+      "learning_rate": 9.52175574783027e-06,
+      "loss": 0.715,
+      "step": 22487
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.328900745412052,
+      "learning_rate": 9.520993581266916e-06,
+      "loss": 0.7183,
+      "step": 22488
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2557526479893217,
+      "learning_rate": 9.520231417492484e-06,
+      "loss": 0.6112,
+      "step": 22489
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3356489720139633,
+      "learning_rate": 9.519469256511415e-06,
+      "loss": 0.6946,
+      "step": 22490
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8469084415689925,
+      "learning_rate": 9.518707098328152e-06,
+      "loss": 0.7016,
+      "step": 22491
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5785753918517222,
+      "learning_rate": 9.517944942947126e-06,
+      "loss": 0.5319,
+      "step": 22492
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.68240821053195,
+      "learning_rate": 9.517182790372776e-06,
+      "loss": 0.7982,
+      "step": 22493
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3912480106381977,
+      "learning_rate": 9.516420640609541e-06,
+      "loss": 0.838,
+      "step": 22494
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5109540390800473,
+      "learning_rate": 9.515658493661863e-06,
+      "loss": 0.7075,
+      "step": 22495
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7177361289160915,
+      "learning_rate": 9.51489634953417e-06,
+      "loss": 0.8854,
+      "step": 22496
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6128561913055486,
+      "learning_rate": 9.514134208230904e-06,
+      "loss": 0.538,
+      "step": 22497
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4396699580594954,
+      "learning_rate": 9.513372069756508e-06,
+      "loss": 0.8317,
+      "step": 22498
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2772327215440569,
+      "learning_rate": 9.512609934115407e-06,
+      "loss": 0.7202,
+      "step": 22499
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3552705029825236,
+      "learning_rate": 9.511847801312048e-06,
+      "loss": 0.7651,
+      "step": 22500
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5139621653629558,
+      "learning_rate": 9.511085671350864e-06,
+      "loss": 0.744,
+      "step": 22501
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3794655181122737,
+      "learning_rate": 9.510323544236298e-06,
+      "loss": 0.7889,
+      "step": 22502
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3296909605748088,
+      "learning_rate": 9.50956141997278e-06,
+      "loss": 0.6801,
+      "step": 22503
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.525203085935309,
+      "learning_rate": 9.508799298564753e-06,
+      "loss": 0.7695,
+      "step": 22504
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3498739909857842,
+      "learning_rate": 9.508037180016652e-06,
+      "loss": 0.8086,
+      "step": 22505
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5566754558409022,
+      "learning_rate": 9.50727506433291e-06,
+      "loss": 0.7074,
+      "step": 22506
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7047697747771076,
+      "learning_rate": 9.506512951517972e-06,
+      "loss": 0.6758,
+      "step": 22507
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6308725401654128,
+      "learning_rate": 9.50575084157627e-06,
+      "loss": 0.5648,
+      "step": 22508
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3878716829289925,
+      "learning_rate": 9.504988734512249e-06,
+      "loss": 0.6819,
+      "step": 22509
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5160508087090265,
+      "learning_rate": 9.504226630330333e-06,
+      "loss": 0.6472,
+      "step": 22510
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4865467504773615,
+      "learning_rate": 9.503464529034967e-06,
+      "loss": 0.6916,
+      "step": 22511
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4092370178647802,
+      "learning_rate": 9.502702430630593e-06,
+      "loss": 0.6303,
+      "step": 22512
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3916031008845744,
+      "learning_rate": 9.50194033512164e-06,
+      "loss": 0.6982,
+      "step": 22513
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5836306808689795,
+      "learning_rate": 9.501178242512546e-06,
+      "loss": 0.8113,
+      "step": 22514
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.787036257789341,
+      "learning_rate": 9.500416152807752e-06,
+      "loss": 0.6307,
+      "step": 22515
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3281699241977813,
+      "learning_rate": 9.499654066011696e-06,
+      "loss": 0.8514,
+      "step": 22516
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.624160321040473,
+      "learning_rate": 9.498891982128809e-06,
+      "loss": 0.576,
+      "step": 22517
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.574585431822903,
+      "learning_rate": 9.49812990116353e-06,
+      "loss": 0.7343,
+      "step": 22518
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.575959960582517,
+      "learning_rate": 9.497367823120302e-06,
+      "loss": 0.7045,
+      "step": 22519
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5678589584439453,
+      "learning_rate": 9.496605748003556e-06,
+      "loss": 0.5264,
+      "step": 22520
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5520878721220555,
+      "learning_rate": 9.49584367581773e-06,
+      "loss": 0.6678,
+      "step": 22521
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5759522591049036,
+      "learning_rate": 9.495081606567261e-06,
+      "loss": 0.4518,
+      "step": 22522
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5873880228215347,
+      "learning_rate": 9.49431954025659e-06,
+      "loss": 0.6274,
+      "step": 22523
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.228392249959317,
+      "learning_rate": 9.493557476890146e-06,
+      "loss": 0.7807,
+      "step": 22524
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4899905035188068,
+      "learning_rate": 9.492795416472374e-06,
+      "loss": 0.7572,
+      "step": 22525
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5254090277944465,
+      "learning_rate": 9.49203335900771e-06,
+      "loss": 0.7636,
+      "step": 22526
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3938728007359993,
+      "learning_rate": 9.491271304500581e-06,
+      "loss": 0.7854,
+      "step": 22527
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4024875731880893,
+      "learning_rate": 9.490509252955437e-06,
+      "loss": 0.6817,
+      "step": 22528
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4409348395126511,
+      "learning_rate": 9.489747204376707e-06,
+      "loss": 0.7982,
+      "step": 22529
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5803769613001903,
+      "learning_rate": 9.488985158768836e-06,
+      "loss": 0.7256,
+      "step": 22530
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.407754551595511,
+      "learning_rate": 9.488223116136248e-06,
+      "loss": 0.5862,
+      "step": 22531
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7270835157702413,
+      "learning_rate": 9.48746107648339e-06,
+      "loss": 0.722,
+      "step": 22532
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3091411395149681,
+      "learning_rate": 9.486699039814698e-06,
+      "loss": 0.6158,
+      "step": 22533
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6408747931866908,
+      "learning_rate": 9.485937006134605e-06,
+      "loss": 0.5304,
+      "step": 22534
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6099650017510636,
+      "learning_rate": 9.485174975447546e-06,
+      "loss": 0.6577,
+      "step": 22535
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5907754595091371,
+      "learning_rate": 9.484412947757965e-06,
+      "loss": 0.4918,
+      "step": 22536
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5494318154415063,
+      "learning_rate": 9.483650923070297e-06,
+      "loss": 0.831,
+      "step": 22537
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5797477306409541,
+      "learning_rate": 9.482888901388973e-06,
+      "loss": 0.5148,
+      "step": 22538
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.29937616965322,
+      "learning_rate": 9.482126882718433e-06,
+      "loss": 0.7561,
+      "step": 22539
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.204583079654821,
+      "learning_rate": 9.481364867063118e-06,
+      "loss": 0.6577,
+      "step": 22540
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3510078720245144,
+      "learning_rate": 9.48060285442746e-06,
+      "loss": 0.784,
+      "step": 22541
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.808808467846831,
+      "learning_rate": 9.479840844815892e-06,
+      "loss": 0.6788,
+      "step": 22542
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4437380205608972,
+      "learning_rate": 9.47907883823286e-06,
+      "loss": 0.639,
+      "step": 22543
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.46654082097847,
+      "learning_rate": 9.478316834682798e-06,
+      "loss": 0.7814,
+      "step": 22544
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1288957509918807,
+      "learning_rate": 9.477554834170137e-06,
+      "loss": 0.651,
+      "step": 22545
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.781649245579577,
+      "learning_rate": 9.476792836699315e-06,
+      "loss": 0.6958,
+      "step": 22546
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.219923707040806,
+      "learning_rate": 9.476030842274773e-06,
+      "loss": 0.6514,
+      "step": 22547
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3778970182124524,
+      "learning_rate": 9.475268850900947e-06,
+      "loss": 0.6355,
+      "step": 22548
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5696478620032023,
+      "learning_rate": 9.47450686258227e-06,
+      "loss": 0.5125,
+      "step": 22549
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5564202218172751,
+      "learning_rate": 9.473744877323178e-06,
+      "loss": 0.4623,
+      "step": 22550
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4746375052745613,
+      "learning_rate": 9.472982895128116e-06,
+      "loss": 0.6228,
+      "step": 22551
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5324562303329299,
+      "learning_rate": 9.472220916001511e-06,
+      "loss": 0.5756,
+      "step": 22552
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4951851474264315,
+      "learning_rate": 9.471458939947804e-06,
+      "loss": 0.6688,
+      "step": 22553
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.485970708671698,
+      "learning_rate": 9.470696966971427e-06,
+      "loss": 0.7393,
+      "step": 22554
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3915958392188685,
+      "learning_rate": 9.469934997076825e-06,
+      "loss": 0.7216,
+      "step": 22555
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9265639601431435,
+      "learning_rate": 9.469173030268424e-06,
+      "loss": 0.668,
+      "step": 22556
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5510008811292993,
+      "learning_rate": 9.46841106655067e-06,
+      "loss": 0.4972,
+      "step": 22557
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2929319227074887,
+      "learning_rate": 9.467649105927995e-06,
+      "loss": 0.5122,
+      "step": 22558
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5815174827744445,
+      "learning_rate": 9.466887148404831e-06,
+      "loss": 0.7052,
+      "step": 22559
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3716655707160788,
+      "learning_rate": 9.466125193985623e-06,
+      "loss": 0.7027,
+      "step": 22560
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2057055801077592,
+      "learning_rate": 9.465363242674802e-06,
+      "loss": 0.7884,
+      "step": 22561
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1912519393254661,
+      "learning_rate": 9.464601294476807e-06,
+      "loss": 0.6342,
+      "step": 22562
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5921342259596731,
+      "learning_rate": 9.463839349396069e-06,
+      "loss": 0.7676,
+      "step": 22563
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.8357679570116088,
+      "learning_rate": 9.463077407437028e-06,
+      "loss": 0.6171,
+      "step": 22564
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.264765211254782,
+      "learning_rate": 9.462315468604126e-06,
+      "loss": 0.6683,
+      "step": 22565
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.477521946117544,
+      "learning_rate": 9.461553532901789e-06,
+      "loss": 0.7218,
+      "step": 22566
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2652539932286138,
+      "learning_rate": 9.460791600334457e-06,
+      "loss": 0.6555,
+      "step": 22567
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.335422610622803,
+      "learning_rate": 9.460029670906568e-06,
+      "loss": 0.682,
+      "step": 22568
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4200974198119323,
+      "learning_rate": 9.45926774462256e-06,
+      "loss": 0.6871,
+      "step": 22569
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3479870442375708,
+      "learning_rate": 9.458505821486864e-06,
+      "loss": 0.8038,
+      "step": 22570
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.624522693420388,
+      "learning_rate": 9.457743901503917e-06,
+      "loss": 0.8031,
+      "step": 22571
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6208114238141835,
+      "learning_rate": 9.45698198467816e-06,
+      "loss": 0.7569,
+      "step": 22572
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5875114391989054,
+      "learning_rate": 9.456220071014025e-06,
+      "loss": 0.7421,
+      "step": 22573
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5056405361843002,
+      "learning_rate": 9.455458160515948e-06,
+      "loss": 0.6503,
+      "step": 22574
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8032407922963167,
+      "learning_rate": 9.454696253188365e-06,
+      "loss": 0.7074,
+      "step": 22575
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.382879003451336,
+      "learning_rate": 9.453934349035715e-06,
+      "loss": 0.7494,
+      "step": 22576
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3687946902185733,
+      "learning_rate": 9.453172448062433e-06,
+      "loss": 0.6435,
+      "step": 22577
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5629897479034343,
+      "learning_rate": 9.452410550272949e-06,
+      "loss": 0.7599,
+      "step": 22578
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3963665756645818,
+      "learning_rate": 9.451648655671712e-06,
+      "loss": 0.8104,
+      "step": 22579
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5815734852021518,
+      "learning_rate": 9.450886764263144e-06,
+      "loss": 0.521,
+      "step": 22580
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6054533290845996,
+      "learning_rate": 9.450124876051688e-06,
+      "loss": 0.5686,
+      "step": 22581
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3796765085028018,
+      "learning_rate": 9.449362991041778e-06,
+      "loss": 0.7065,
+      "step": 22582
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.515868327931914,
+      "learning_rate": 9.448601109237859e-06,
+      "loss": 0.6615,
+      "step": 22583
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.688305941961322,
+      "learning_rate": 9.447839230644349e-06,
+      "loss": 0.6629,
+      "step": 22584
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4760921887410596,
+      "learning_rate": 9.447077355265698e-06,
+      "loss": 0.6845,
+      "step": 22585
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4012351884900578,
+      "learning_rate": 9.44631548310634e-06,
+      "loss": 0.678,
+      "step": 22586
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4821800849619753,
+      "learning_rate": 9.445553614170705e-06,
+      "loss": 0.7862,
+      "step": 22587
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4483089589181353,
+      "learning_rate": 9.444791748463232e-06,
+      "loss": 0.8717,
+      "step": 22588
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5236070893112241,
+      "learning_rate": 9.444029885988359e-06,
+      "loss": 0.7012,
+      "step": 22589
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.588305303463317,
+      "learning_rate": 9.443268026750521e-06,
+      "loss": 0.5143,
+      "step": 22590
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3647059225497904,
+      "learning_rate": 9.442506170754149e-06,
+      "loss": 0.6536,
+      "step": 22591
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5789237478429987,
+      "learning_rate": 9.441744318003685e-06,
+      "loss": 0.6767,
+      "step": 22592
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3709313240636176,
+      "learning_rate": 9.440982468503563e-06,
+      "loss": 0.7424,
+      "step": 22593
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.345275945106589,
+      "learning_rate": 9.440220622258217e-06,
+      "loss": 0.664,
+      "step": 22594
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.380081104410358,
+      "learning_rate": 9.439458779272082e-06,
+      "loss": 0.6231,
+      "step": 22595
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.364453263494619,
+      "learning_rate": 9.438696939549598e-06,
+      "loss": 0.7574,
+      "step": 22596
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5272152754930097,
+      "learning_rate": 9.437935103095201e-06,
+      "loss": 0.6803,
+      "step": 22597
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5960366710501894,
+      "learning_rate": 9.43717326991332e-06,
+      "loss": 0.5015,
+      "step": 22598
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.541927021193443,
+      "learning_rate": 9.436411440008392e-06,
+      "loss": 0.7704,
+      "step": 22599
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6252926343014025,
+      "learning_rate": 9.435649613384859e-06,
+      "loss": 0.7035,
+      "step": 22600
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.73643439524557,
+      "learning_rate": 9.434887790047152e-06,
+      "loss": 0.7855,
+      "step": 22601
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.321718643893687,
+      "learning_rate": 9.434125969999707e-06,
+      "loss": 0.6298,
+      "step": 22602
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3700747204654695,
+      "learning_rate": 9.433364153246958e-06,
+      "loss": 0.6794,
+      "step": 22603
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.318597460613499,
+      "learning_rate": 9.432602339793346e-06,
+      "loss": 0.7664,
+      "step": 22604
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4340894950025325,
+      "learning_rate": 9.4318405296433e-06,
+      "loss": 0.7095,
+      "step": 22605
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.710235887832972,
+      "learning_rate": 9.43107872280126e-06,
+      "loss": 0.5993,
+      "step": 22606
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4440969184214723,
+      "learning_rate": 9.43031691927166e-06,
+      "loss": 0.7456,
+      "step": 22607
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3070319847472829,
+      "learning_rate": 9.429555119058931e-06,
+      "loss": 0.7292,
+      "step": 22608
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3786463220439527,
+      "learning_rate": 9.428793322167516e-06,
+      "loss": 0.719,
+      "step": 22609
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4988143045877278,
+      "learning_rate": 9.428031528601846e-06,
+      "loss": 0.8853,
+      "step": 22610
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4321244680239646,
+      "learning_rate": 9.427269738366361e-06,
+      "loss": 0.7754,
+      "step": 22611
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3985886732143937,
+      "learning_rate": 9.426507951465487e-06,
+      "loss": 0.6607,
+      "step": 22612
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3718896770909046,
+      "learning_rate": 9.42574616790367e-06,
+      "loss": 0.7166,
+      "step": 22613
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2660064996463423,
+      "learning_rate": 9.42498438768534e-06,
+      "loss": 0.7264,
+      "step": 22614
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.488923360500206,
+      "learning_rate": 9.424222610814932e-06,
+      "loss": 0.6144,
+      "step": 22615
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4702184353263223,
+      "learning_rate": 9.423460837296879e-06,
+      "loss": 0.8008,
+      "step": 22616
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.778390209025696,
+      "learning_rate": 9.422699067135624e-06,
+      "loss": 0.7711,
+      "step": 22617
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5834103402216007,
+      "learning_rate": 9.4219373003356e-06,
+      "loss": 0.6391,
+      "step": 22618
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2798605185214416,
+      "learning_rate": 9.421175536901234e-06,
+      "loss": 0.7583,
+      "step": 22619
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5527207213902119,
+      "learning_rate": 9.420413776836968e-06,
+      "loss": 0.7381,
+      "step": 22620
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.548174102735827,
+      "learning_rate": 9.41965202014724e-06,
+      "loss": 0.8207,
+      "step": 22621
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4058471037544957,
+      "learning_rate": 9.41889026683648e-06,
+      "loss": 0.8061,
+      "step": 22622
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6053481185027315,
+      "learning_rate": 9.418128516909126e-06,
+      "loss": 0.5736,
+      "step": 22623
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6042739217328513,
+      "learning_rate": 9.417366770369609e-06,
+      "loss": 0.5141,
+      "step": 22624
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4786675034951704,
+      "learning_rate": 9.41660502722237e-06,
+      "loss": 0.6761,
+      "step": 22625
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.536854223089941,
+      "learning_rate": 9.41584328747184e-06,
+      "loss": 0.7962,
+      "step": 22626
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2871170844284399,
+      "learning_rate": 9.415081551122453e-06,
+      "loss": 0.7209,
+      "step": 22627
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.627204220421388,
+      "learning_rate": 9.414319818178654e-06,
+      "loss": 0.8338,
+      "step": 22628
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6098048443281199,
+      "learning_rate": 9.413558088644864e-06,
+      "loss": 0.4642,
+      "step": 22629
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7740244172713389,
+      "learning_rate": 9.412796362525526e-06,
+      "loss": 0.6719,
+      "step": 22630
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6165508851228992,
+      "learning_rate": 9.41203463982507e-06,
+      "loss": 0.5185,
+      "step": 22631
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3740673861040453,
+      "learning_rate": 9.411272920547942e-06,
+      "loss": 0.7565,
+      "step": 22632
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3834608653185239,
+      "learning_rate": 9.410511204698562e-06,
+      "loss": 0.7515,
+      "step": 22633
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4692408902830336,
+      "learning_rate": 9.409749492281378e-06,
+      "loss": 0.7826,
+      "step": 22634
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4831675651824134,
+      "learning_rate": 9.408987783300818e-06,
+      "loss": 0.6521,
+      "step": 22635
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5994665855639199,
+      "learning_rate": 9.408226077761316e-06,
+      "loss": 0.749,
+      "step": 22636
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9874183593278731,
+      "learning_rate": 9.407464375667309e-06,
+      "loss": 0.7363,
+      "step": 22637
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3443689946682462,
+      "learning_rate": 9.406702677023233e-06,
+      "loss": 0.7295,
+      "step": 22638
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.580377480209962,
+      "learning_rate": 9.405940981833525e-06,
+      "loss": 0.8421,
+      "step": 22639
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.578413205717961,
+      "learning_rate": 9.405179290102611e-06,
+      "loss": 0.708,
+      "step": 22640
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.577555137307436,
+      "learning_rate": 9.404417601834935e-06,
+      "loss": 0.8078,
+      "step": 22641
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5996325636587254,
+      "learning_rate": 9.40365591703493e-06,
+      "loss": 0.7956,
+      "step": 22642
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4240056949796636,
+      "learning_rate": 9.402894235707026e-06,
+      "loss": 0.6443,
+      "step": 22643
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4102443609917774,
+      "learning_rate": 9.402132557855659e-06,
+      "loss": 0.8094,
+      "step": 22644
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.492763746827171,
+      "learning_rate": 9.401370883485268e-06,
+      "loss": 0.7188,
+      "step": 22645
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3975731131148992,
+      "learning_rate": 9.400609212600288e-06,
+      "loss": 0.6856,
+      "step": 22646
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.414479531021845,
+      "learning_rate": 9.399847545205148e-06,
+      "loss": 0.8157,
+      "step": 22647
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.732116706477415,
+      "learning_rate": 9.399085881304283e-06,
+      "loss": 0.7607,
+      "step": 22648
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1535719870644239,
+      "learning_rate": 9.398324220902135e-06,
+      "loss": 0.6816,
+      "step": 22649
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6201880898477684,
+      "learning_rate": 9.397562564003131e-06,
+      "loss": 0.5105,
+      "step": 22650
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3111833513937212,
+      "learning_rate": 9.39680091061171e-06,
+      "loss": 0.6791,
+      "step": 22651
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4005827610251549,
+      "learning_rate": 9.396039260732302e-06,
+      "loss": 0.683,
+      "step": 22652
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3496647440561493,
+      "learning_rate": 9.39527761436935e-06,
+      "loss": 0.6293,
+      "step": 22653
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.434490316190449,
+      "learning_rate": 9.39451597152728e-06,
+      "loss": 0.7883,
+      "step": 22654
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.376701990467236,
+      "learning_rate": 9.393754332210529e-06,
+      "loss": 0.6981,
+      "step": 22655
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5212257991826617,
+      "learning_rate": 9.392992696423536e-06,
+      "loss": 0.6083,
+      "step": 22656
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3628463362966226,
+      "learning_rate": 9.392231064170728e-06,
+      "loss": 0.7683,
+      "step": 22657
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.270998766602191,
+      "learning_rate": 9.391469435456543e-06,
+      "loss": 0.642,
+      "step": 22658
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5928779375337795,
+      "learning_rate": 9.390707810285416e-06,
+      "loss": 0.5121,
+      "step": 22659
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6360688851728198,
+      "learning_rate": 9.389946188661785e-06,
+      "loss": 0.7738,
+      "step": 22660
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6131483352410647,
+      "learning_rate": 9.389184570590075e-06,
+      "loss": 0.7692,
+      "step": 22661
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3363295241655984,
+      "learning_rate": 9.388422956074727e-06,
+      "loss": 0.6481,
+      "step": 22662
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8902317292223032,
+      "learning_rate": 9.387661345120178e-06,
+      "loss": 0.7216,
+      "step": 22663
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6367237478916226,
+      "learning_rate": 9.386899737730854e-06,
+      "loss": 0.6421,
+      "step": 22664
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.42266969919263,
+      "learning_rate": 9.386138133911193e-06,
+      "loss": 0.7291,
+      "step": 22665
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3300171413109143,
+      "learning_rate": 9.385376533665633e-06,
+      "loss": 0.6304,
+      "step": 22666
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6974253834017976,
+      "learning_rate": 9.384614936998607e-06,
+      "loss": 0.6929,
+      "step": 22667
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.12645033831828,
+      "learning_rate": 9.383853343914546e-06,
+      "loss": 0.7646,
+      "step": 22668
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2496362790284796,
+      "learning_rate": 9.383091754417884e-06,
+      "loss": 0.793,
+      "step": 22669
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4628791675736483,
+      "learning_rate": 9.382330168513058e-06,
+      "loss": 0.7549,
+      "step": 22670
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2981875457571443,
+      "learning_rate": 9.381568586204504e-06,
+      "loss": 0.7753,
+      "step": 22671
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5863963116145839,
+      "learning_rate": 9.380807007496652e-06,
+      "loss": 0.4928,
+      "step": 22672
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4594896084882394,
+      "learning_rate": 9.380045432393934e-06,
+      "loss": 0.7167,
+      "step": 22673
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.467522569596371,
+      "learning_rate": 9.379283860900793e-06,
+      "loss": 0.5739,
+      "step": 22674
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4607535190975738,
+      "learning_rate": 9.378522293021656e-06,
+      "loss": 0.7899,
+      "step": 22675
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7078482565253619,
+      "learning_rate": 9.377760728760957e-06,
+      "loss": 0.7878,
+      "step": 22676
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6284450201156826,
+      "learning_rate": 9.376999168123134e-06,
+      "loss": 0.7771,
+      "step": 22677
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9033568113860855,
+      "learning_rate": 9.37623761111262e-06,
+      "loss": 0.7606,
+      "step": 22678
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5731662861140472,
+      "learning_rate": 9.375476057733847e-06,
+      "loss": 0.5503,
+      "step": 22679
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5199578082957732,
+      "learning_rate": 9.374714507991248e-06,
+      "loss": 0.7307,
+      "step": 22680
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2683044653508984,
+      "learning_rate": 9.373952961889265e-06,
+      "loss": 0.6476,
+      "step": 22681
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9048890582162135,
+      "learning_rate": 9.373191419432319e-06,
+      "loss": 0.7724,
+      "step": 22682
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3879287897134835,
+      "learning_rate": 9.372429880624856e-06,
+      "loss": 0.7249,
+      "step": 22683
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2232515641835535,
+      "learning_rate": 9.3716683454713e-06,
+      "loss": 0.6509,
+      "step": 22684
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3546510993307856,
+      "learning_rate": 9.370906813976097e-06,
+      "loss": 0.6793,
+      "step": 22685
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9658386403098538,
+      "learning_rate": 9.370145286143667e-06,
+      "loss": 0.7207,
+      "step": 22686
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5279550540079505,
+      "learning_rate": 9.369383761978453e-06,
+      "loss": 0.6851,
+      "step": 22687
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1500284660703457,
+      "learning_rate": 9.36862224148489e-06,
+      "loss": 0.6254,
+      "step": 22688
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5998949629829142,
+      "learning_rate": 9.367860724667401e-06,
+      "loss": 0.8018,
+      "step": 22689
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2343118100321289,
+      "learning_rate": 9.367099211530432e-06,
+      "loss": 0.7827,
+      "step": 22690
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5766814450800815,
+      "learning_rate": 9.36633770207841e-06,
+      "loss": 0.7032,
+      "step": 22691
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4005473941165116,
+      "learning_rate": 9.365576196315772e-06,
+      "loss": 0.6654,
+      "step": 22692
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.043941717080874,
+      "learning_rate": 9.364814694246947e-06,
+      "loss": 0.7286,
+      "step": 22693
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4196341410927968,
+      "learning_rate": 9.364053195876375e-06,
+      "loss": 0.7509,
+      "step": 22694
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3623775915046128,
+      "learning_rate": 9.363291701208488e-06,
+      "loss": 0.7484,
+      "step": 22695
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4408594089959772,
+      "learning_rate": 9.362530210247715e-06,
+      "loss": 0.6613,
+      "step": 22696
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3170423737532444,
+      "learning_rate": 9.361768722998494e-06,
+      "loss": 0.7754,
+      "step": 22697
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3558830457818474,
+      "learning_rate": 9.361007239465257e-06,
+      "loss": 0.7727,
+      "step": 22698
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6089109518714526,
+      "learning_rate": 9.36024575965244e-06,
+      "loss": 0.8063,
+      "step": 22699
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7713783764272728,
+      "learning_rate": 9.359484283564474e-06,
+      "loss": 0.7313,
+      "step": 22700
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2800038120602937,
+      "learning_rate": 9.35872281120579e-06,
+      "loss": 0.6118,
+      "step": 22701
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3940057273748856,
+      "learning_rate": 9.35796134258083e-06,
+      "loss": 0.7492,
+      "step": 22702
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.448368522565023,
+      "learning_rate": 9.35719987769402e-06,
+      "loss": 0.8239,
+      "step": 22703
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4879391545572007,
+      "learning_rate": 9.356438416549795e-06,
+      "loss": 0.8383,
+      "step": 22704
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3796578806591466,
+      "learning_rate": 9.355676959152588e-06,
+      "loss": 0.8049,
+      "step": 22705
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3395658277260432,
+      "learning_rate": 9.354915505506839e-06,
+      "loss": 0.7697,
+      "step": 22706
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6080270934328864,
+      "learning_rate": 9.354154055616971e-06,
+      "loss": 0.7387,
+      "step": 22707
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3531932558404371,
+      "learning_rate": 9.353392609487427e-06,
+      "loss": 0.7128,
+      "step": 22708
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6758918515390278,
+      "learning_rate": 9.352631167122635e-06,
+      "loss": 0.8279,
+      "step": 22709
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.471749230031062,
+      "learning_rate": 9.351869728527024e-06,
+      "loss": 0.6618,
+      "step": 22710
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.323717257889503,
+      "learning_rate": 9.351108293705037e-06,
+      "loss": 0.6566,
+      "step": 22711
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4416308823217645,
+      "learning_rate": 9.3503468626611e-06,
+      "loss": 0.78,
+      "step": 22712
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.342818840342979,
+      "learning_rate": 9.349585435399656e-06,
+      "loss": 0.7812,
+      "step": 22713
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2181440388531288,
+      "learning_rate": 9.348824011925124e-06,
+      "loss": 0.7532,
+      "step": 22714
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.44974097490483,
+      "learning_rate": 9.34806259224195e-06,
+      "loss": 0.7006,
+      "step": 22715
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4544196807935894,
+      "learning_rate": 9.34730117635456e-06,
+      "loss": 0.7499,
+      "step": 22716
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4392769789690263,
+      "learning_rate": 9.346539764267389e-06,
+      "loss": 0.618,
+      "step": 22717
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3461433686109112,
+      "learning_rate": 9.345778355984869e-06,
+      "loss": 0.7094,
+      "step": 22718
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.569175135152292,
+      "learning_rate": 9.345016951511436e-06,
+      "loss": 0.728,
+      "step": 22719
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.610185278915055,
+      "learning_rate": 9.344255550851523e-06,
+      "loss": 0.523,
+      "step": 22720
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5695359663224315,
+      "learning_rate": 9.34349415400956e-06,
+      "loss": 0.7043,
+      "step": 22721
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4846124362880433,
+      "learning_rate": 9.34273276098998e-06,
+      "loss": 0.8273,
+      "step": 22722
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3370106566532332,
+      "learning_rate": 9.341971371797223e-06,
+      "loss": 0.8,
+      "step": 22723
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2568313982530723,
+      "learning_rate": 9.341209986435715e-06,
+      "loss": 0.7496,
+      "step": 22724
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5617222527925019,
+      "learning_rate": 9.340448604909888e-06,
+      "loss": 0.4852,
+      "step": 22725
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4757989499735253,
+      "learning_rate": 9.339687227224178e-06,
+      "loss": 0.7878,
+      "step": 22726
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3740480111166722,
+      "learning_rate": 9.338925853383024e-06,
+      "loss": 0.6866,
+      "step": 22727
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6443352388022854,
+      "learning_rate": 9.338164483390849e-06,
+      "loss": 0.5076,
+      "step": 22728
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.334641014575364,
+      "learning_rate": 9.337403117252089e-06,
+      "loss": 0.5613,
+      "step": 22729
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.687841687416059,
+      "learning_rate": 9.336641754971183e-06,
+      "loss": 0.7106,
+      "step": 22730
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.8724910199910982,
+      "learning_rate": 9.335880396552552e-06,
+      "loss": 0.6812,
+      "step": 22731
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0046913788196092,
+      "learning_rate": 9.335119042000637e-06,
+      "loss": 0.6596,
+      "step": 22732
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4476171138420775,
+      "learning_rate": 9.33435769131987e-06,
+      "loss": 0.6335,
+      "step": 22733
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5700970076850977,
+      "learning_rate": 9.333596344514689e-06,
+      "loss": 0.5723,
+      "step": 22734
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3078233192250657,
+      "learning_rate": 9.332835001589514e-06,
+      "loss": 0.7713,
+      "step": 22735
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6184486070478503,
+      "learning_rate": 9.332073662548785e-06,
+      "loss": 0.5508,
+      "step": 22736
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3227077294132497,
+      "learning_rate": 9.33131232739694e-06,
+      "loss": 0.796,
+      "step": 22737
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6606922820189559,
+      "learning_rate": 9.3305509961384e-06,
+      "loss": 0.5587,
+      "step": 22738
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.322423570186456,
+      "learning_rate": 9.329789668777606e-06,
+      "loss": 0.7799,
+      "step": 22739
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3275046009143883,
+      "learning_rate": 9.32902834531899e-06,
+      "loss": 0.7016,
+      "step": 22740
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.163783496893843,
+      "learning_rate": 9.328267025766984e-06,
+      "loss": 0.7797,
+      "step": 22741
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2818328899083014,
+      "learning_rate": 9.327505710126018e-06,
+      "loss": 0.7088,
+      "step": 22742
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3998716775104425,
+      "learning_rate": 9.326744398400527e-06,
+      "loss": 0.7724,
+      "step": 22743
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1793396912499203,
+      "learning_rate": 9.325983090594946e-06,
+      "loss": 0.6083,
+      "step": 22744
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3292338736491087,
+      "learning_rate": 9.325221786713704e-06,
+      "loss": 0.7298,
+      "step": 22745
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.466268906096316,
+      "learning_rate": 9.324460486761231e-06,
+      "loss": 0.5878,
+      "step": 22746
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4104666188376505,
+      "learning_rate": 9.323699190741964e-06,
+      "loss": 0.6806,
+      "step": 22747
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.250917774874531,
+      "learning_rate": 9.322937898660339e-06,
+      "loss": 0.6456,
+      "step": 22748
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5513219565609513,
+      "learning_rate": 9.32217661052078e-06,
+      "loss": 0.7032,
+      "step": 22749
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7450355513175462,
+      "learning_rate": 9.321415326327721e-06,
+      "loss": 0.577,
+      "step": 22750
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6170585712712972,
+      "learning_rate": 9.320654046085602e-06,
+      "loss": 0.5151,
+      "step": 22751
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5293301658109673,
+      "learning_rate": 9.319892769798848e-06,
+      "loss": 0.7243,
+      "step": 22752
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5834339423279171,
+      "learning_rate": 9.319131497471894e-06,
+      "loss": 0.6849,
+      "step": 22753
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5487266938621123,
+      "learning_rate": 9.318370229109168e-06,
+      "loss": 0.6889,
+      "step": 22754
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5368320187951552,
+      "learning_rate": 9.317608964715113e-06,
+      "loss": 0.7165,
+      "step": 22755
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2635834465614724,
+      "learning_rate": 9.31684770429415e-06,
+      "loss": 0.7418,
+      "step": 22756
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5474490275945396,
+      "learning_rate": 9.316086447850718e-06,
+      "loss": 0.699,
+      "step": 22757
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3619315222421329,
+      "learning_rate": 9.31532519538925e-06,
+      "loss": 0.6332,
+      "step": 22758
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5046731776652815,
+      "learning_rate": 9.314563946914169e-06,
+      "loss": 0.8079,
+      "step": 22759
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.60633625592276,
+      "learning_rate": 9.313802702429918e-06,
+      "loss": 0.7396,
+      "step": 22760
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.437779138221576,
+      "learning_rate": 9.31304146194092e-06,
+      "loss": 0.6663,
+      "step": 22761
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5638656688213625,
+      "learning_rate": 9.31228022545162e-06,
+      "loss": 0.508,
+      "step": 22762
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5437922042330612,
+      "learning_rate": 9.311518992966436e-06,
+      "loss": 0.6816,
+      "step": 22763
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5813736721924294,
+      "learning_rate": 9.310757764489808e-06,
+      "loss": 0.7386,
+      "step": 22764
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3009496992709784,
+      "learning_rate": 9.309996540026169e-06,
+      "loss": 0.7245,
+      "step": 22765
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.376721798172426,
+      "learning_rate": 9.309235319579947e-06,
+      "loss": 0.6956,
+      "step": 22766
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.768377777529567,
+      "learning_rate": 9.308474103155573e-06,
+      "loss": 0.6848,
+      "step": 22767
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3141424409750806,
+      "learning_rate": 9.307712890757483e-06,
+      "loss": 0.7254,
+      "step": 22768
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5604664385563238,
+      "learning_rate": 9.306951682390112e-06,
+      "loss": 0.5328,
+      "step": 22769
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5280316998586112,
+      "learning_rate": 9.306190478057884e-06,
+      "loss": 0.735,
+      "step": 22770
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5135170291811106,
+      "learning_rate": 9.305429277765233e-06,
+      "loss": 0.7573,
+      "step": 22771
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.556536196347006,
+      "learning_rate": 9.304668081516598e-06,
+      "loss": 0.7129,
+      "step": 22772
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4785311736598863,
+      "learning_rate": 9.3039068893164e-06,
+      "loss": 0.637,
+      "step": 22773
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3660997621977669,
+      "learning_rate": 9.303145701169077e-06,
+      "loss": 0.8199,
+      "step": 22774
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2703332593940897,
+      "learning_rate": 9.302384517079061e-06,
+      "loss": 0.6794,
+      "step": 22775
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.32787342806101,
+      "learning_rate": 9.301623337050788e-06,
+      "loss": 0.5622,
+      "step": 22776
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4749835361782475,
+      "learning_rate": 9.30086216108868e-06,
+      "loss": 0.7502,
+      "step": 22777
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.20896730514196,
+      "learning_rate": 9.300100989197174e-06,
+      "loss": 0.6881,
+      "step": 22778
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4280509790733698,
+      "learning_rate": 9.299339821380708e-06,
+      "loss": 0.7988,
+      "step": 22779
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2620943452645033,
+      "learning_rate": 9.2985786576437e-06,
+      "loss": 0.551,
+      "step": 22780
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4163764163068682,
+      "learning_rate": 9.29781749799059e-06,
+      "loss": 0.6968,
+      "step": 22781
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3515495863809892,
+      "learning_rate": 9.297056342425807e-06,
+      "loss": 0.7819,
+      "step": 22782
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4766334733923037,
+      "learning_rate": 9.296295190953792e-06,
+      "loss": 0.7477,
+      "step": 22783
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.508559429685745,
+      "learning_rate": 9.295534043578963e-06,
+      "loss": 0.6214,
+      "step": 22784
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4472804814820055,
+      "learning_rate": 9.29477290030576e-06,
+      "loss": 0.7567,
+      "step": 22785
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2738754975028463,
+      "learning_rate": 9.294011761138614e-06,
+      "loss": 0.7987,
+      "step": 22786
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4217168234608577,
+      "learning_rate": 9.293250626081951e-06,
+      "loss": 0.6942,
+      "step": 22787
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.290286089249724,
+      "learning_rate": 9.292489495140209e-06,
+      "loss": 0.7585,
+      "step": 22788
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4247307848319677,
+      "learning_rate": 9.291728368317816e-06,
+      "loss": 0.7651,
+      "step": 22789
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4258851569097957,
+      "learning_rate": 9.290967245619208e-06,
+      "loss": 0.8443,
+      "step": 22790
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6035247272743292,
+      "learning_rate": 9.290206127048809e-06,
+      "loss": 0.5303,
+      "step": 22791
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.364304743107969,
+      "learning_rate": 9.289445012611056e-06,
+      "loss": 0.6421,
+      "step": 22792
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4244198070333725,
+      "learning_rate": 9.28868390231038e-06,
+      "loss": 0.6839,
+      "step": 22793
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.452862177160038,
+      "learning_rate": 9.287922796151214e-06,
+      "loss": 0.7183,
+      "step": 22794
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7491943550440634,
+      "learning_rate": 9.28716169413798e-06,
+      "loss": 0.6611,
+      "step": 22795
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4354456251754264,
+      "learning_rate": 9.28640059627512e-06,
+      "loss": 0.7384,
+      "step": 22796
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.346003268478148,
+      "learning_rate": 9.285639502567065e-06,
+      "loss": 0.7287,
+      "step": 22797
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6397284748042011,
+      "learning_rate": 9.284878413018238e-06,
+      "loss": 0.7872,
+      "step": 22798
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3460053442059945,
+      "learning_rate": 9.284117327633077e-06,
+      "loss": 0.8929,
+      "step": 22799
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5672009045463016,
+      "learning_rate": 9.28335624641601e-06,
+      "loss": 0.6699,
+      "step": 22800
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6209476495056554,
+      "learning_rate": 9.282595169371474e-06,
+      "loss": 0.7872,
+      "step": 22801
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.248689148558527,
+      "learning_rate": 9.281834096503895e-06,
+      "loss": 0.7136,
+      "step": 22802
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4149554376638214,
+      "learning_rate": 9.281073027817702e-06,
+      "loss": 0.6533,
+      "step": 22803
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5157567678520578,
+      "learning_rate": 9.280311963317332e-06,
+      "loss": 0.6543,
+      "step": 22804
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4446999664676954,
+      "learning_rate": 9.279550903007212e-06,
+      "loss": 0.604,
+      "step": 22805
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.441241912594797,
+      "learning_rate": 9.278789846891777e-06,
+      "loss": 0.7463,
+      "step": 22806
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2521150121046234,
+      "learning_rate": 9.278028794975452e-06,
+      "loss": 0.6292,
+      "step": 22807
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4970486340015405,
+      "learning_rate": 9.277267747262676e-06,
+      "loss": 0.7648,
+      "step": 22808
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4609274053846402,
+      "learning_rate": 9.276506703757875e-06,
+      "loss": 0.7751,
+      "step": 22809
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.439991030506164,
+      "learning_rate": 9.275745664465478e-06,
+      "loss": 0.782,
+      "step": 22810
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5538197734272008,
+      "learning_rate": 9.274984629389927e-06,
+      "loss": 0.8086,
+      "step": 22811
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7246082268170524,
+      "learning_rate": 9.274223598535635e-06,
+      "loss": 0.6912,
+      "step": 22812
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3193086535998382,
+      "learning_rate": 9.273462571907049e-06,
+      "loss": 0.6786,
+      "step": 22813
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3444391650266434,
+      "learning_rate": 9.272701549508589e-06,
+      "loss": 0.7135,
+      "step": 22814
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.424044396537965,
+      "learning_rate": 9.271940531344699e-06,
+      "loss": 0.799,
+      "step": 22815
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7786415284833779,
+      "learning_rate": 9.271179517419794e-06,
+      "loss": 0.694,
+      "step": 22816
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2597285739769273,
+      "learning_rate": 9.270418507738316e-06,
+      "loss": 0.7986,
+      "step": 22817
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3333450480841584,
+      "learning_rate": 9.269657502304694e-06,
+      "loss": 0.7056,
+      "step": 22818
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.554644592487687,
+      "learning_rate": 9.268896501123354e-06,
+      "loss": 0.6615,
+      "step": 22819
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.568734432408795,
+      "learning_rate": 9.26813550419873e-06,
+      "loss": 0.6581,
+      "step": 22820
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4135791359805197,
+      "learning_rate": 9.267374511535254e-06,
+      "loss": 0.7586,
+      "step": 22821
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3256571037048024,
+      "learning_rate": 9.266613523137358e-06,
+      "loss": 0.8126,
+      "step": 22822
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4426730046795568,
+      "learning_rate": 9.265852539009466e-06,
+      "loss": 0.7591,
+      "step": 22823
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.264585863087887,
+      "learning_rate": 9.265091559156014e-06,
+      "loss": 0.6621,
+      "step": 22824
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6696640505906808,
+      "learning_rate": 9.264330583581434e-06,
+      "loss": 0.5292,
+      "step": 22825
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6860935311620102,
+      "learning_rate": 9.263569612290154e-06,
+      "loss": 0.7025,
+      "step": 22826
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3471810069599095,
+      "learning_rate": 9.2628086452866e-06,
+      "loss": 0.7079,
+      "step": 22827
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4583571230205346,
+      "learning_rate": 9.262047682575211e-06,
+      "loss": 0.6873,
+      "step": 22828
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3492612587314146,
+      "learning_rate": 9.261286724160417e-06,
+      "loss": 0.7523,
+      "step": 22829
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6621077856727671,
+      "learning_rate": 9.260525770046642e-06,
+      "loss": 0.484,
+      "step": 22830
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2507735607070412,
+      "learning_rate": 9.259764820238318e-06,
+      "loss": 0.8016,
+      "step": 22831
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5706402768531056,
+      "learning_rate": 9.259003874739885e-06,
+      "loss": 0.7925,
+      "step": 22832
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6117260791996446,
+      "learning_rate": 9.258242933555759e-06,
+      "loss": 0.6545,
+      "step": 22833
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4109130554201195,
+      "learning_rate": 9.25748199669038e-06,
+      "loss": 0.798,
+      "step": 22834
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6005660339435992,
+      "learning_rate": 9.256721064148173e-06,
+      "loss": 0.5006,
+      "step": 22835
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4139148607361123,
+      "learning_rate": 9.255960135933577e-06,
+      "loss": 0.6783,
+      "step": 22836
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2767944175806467,
+      "learning_rate": 9.255199212051012e-06,
+      "loss": 0.6851,
+      "step": 22837
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.304759680142779,
+      "learning_rate": 9.254438292504915e-06,
+      "loss": 0.7662,
+      "step": 22838
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6010099647042954,
+      "learning_rate": 9.253677377299716e-06,
+      "loss": 0.4927,
+      "step": 22839
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.572511854867742,
+      "learning_rate": 9.252916466439839e-06,
+      "loss": 0.6811,
+      "step": 22840
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3591357808017999,
+      "learning_rate": 9.252155559929722e-06,
+      "loss": 0.6846,
+      "step": 22841
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3801597072652785,
+      "learning_rate": 9.25139465777379e-06,
+      "loss": 0.6917,
+      "step": 22842
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.628223460466356,
+      "learning_rate": 9.25063375997648e-06,
+      "loss": 0.7722,
+      "step": 22843
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4716069078655134,
+      "learning_rate": 9.249872866542211e-06,
+      "loss": 0.7661,
+      "step": 22844
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.667804181605466,
+      "learning_rate": 9.249111977475422e-06,
+      "loss": 0.6715,
+      "step": 22845
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3426570968295848,
+      "learning_rate": 9.248351092780544e-06,
+      "loss": 0.7678,
+      "step": 22846
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5813062942034,
+      "learning_rate": 9.247590212462001e-06,
+      "loss": 0.807,
+      "step": 22847
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4002297612483117,
+      "learning_rate": 9.246829336524225e-06,
+      "loss": 0.7663,
+      "step": 22848
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.660114294293126,
+      "learning_rate": 9.246068464971649e-06,
+      "loss": 0.5424,
+      "step": 22849
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4440231392070435,
+      "learning_rate": 9.245307597808702e-06,
+      "loss": 0.622,
+      "step": 22850
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2701375167044084,
+      "learning_rate": 9.244546735039811e-06,
+      "loss": 0.6364,
+      "step": 22851
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.288728785272235,
+      "learning_rate": 9.243785876669407e-06,
+      "loss": 0.7133,
+      "step": 22852
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6020992073390958,
+      "learning_rate": 9.243025022701926e-06,
+      "loss": 0.5735,
+      "step": 22853
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5738173388346021,
+      "learning_rate": 9.242264173141789e-06,
+      "loss": 0.5091,
+      "step": 22854
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7035493507803472,
+      "learning_rate": 9.24150332799343e-06,
+      "loss": 0.8047,
+      "step": 22855
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3350665698762088,
+      "learning_rate": 9.240742487261279e-06,
+      "loss": 0.7541,
+      "step": 22856
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3899790813512536,
+      "learning_rate": 9.239981650949768e-06,
+      "loss": 0.7198,
+      "step": 22857
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3531709387120738,
+      "learning_rate": 9.239220819063322e-06,
+      "loss": 0.5723,
+      "step": 22858
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.513703647249309,
+      "learning_rate": 9.238459991606371e-06,
+      "loss": 0.7383,
+      "step": 22859
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6167882988436497,
+      "learning_rate": 9.237699168583355e-06,
+      "loss": 0.5831,
+      "step": 22860
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.543233706470909,
+      "learning_rate": 9.236938349998688e-06,
+      "loss": 0.6477,
+      "step": 22861
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.948730259961982,
+      "learning_rate": 9.23617753585681e-06,
+      "loss": 0.7566,
+      "step": 22862
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5217333451905428,
+      "learning_rate": 9.235416726162147e-06,
+      "loss": 0.7475,
+      "step": 22863
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5140280669705188,
+      "learning_rate": 9.234655920919135e-06,
+      "loss": 0.7608,
+      "step": 22864
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5915768399328036,
+      "learning_rate": 9.233895120132192e-06,
+      "loss": 0.6379,
+      "step": 22865
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4799037211832007,
+      "learning_rate": 9.233134323805757e-06,
+      "loss": 0.7908,
+      "step": 22866
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6266173522169671,
+      "learning_rate": 9.23237353194426e-06,
+      "loss": 0.7173,
+      "step": 22867
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3136543141242527,
+      "learning_rate": 9.231612744552123e-06,
+      "loss": 0.6151,
+      "step": 22868
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.524502501600528,
+      "learning_rate": 9.230851961633779e-06,
+      "loss": 0.6547,
+      "step": 22869
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.239844820681537,
+      "learning_rate": 9.230091183193659e-06,
+      "loss": 0.5727,
+      "step": 22870
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7303566226303524,
+      "learning_rate": 9.229330409236196e-06,
+      "loss": 0.7317,
+      "step": 22871
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6572414823324408,
+      "learning_rate": 9.22856963976581e-06,
+      "loss": 0.6982,
+      "step": 22872
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5467264627972972,
+      "learning_rate": 9.227808874786937e-06,
+      "loss": 0.8263,
+      "step": 22873
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3044503722224567,
+      "learning_rate": 9.22704811430401e-06,
+      "loss": 0.7151,
+      "step": 22874
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7477887928687494,
+      "learning_rate": 9.22628735832145e-06,
+      "loss": 0.7575,
+      "step": 22875
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7603156926961734,
+      "learning_rate": 9.225526606843687e-06,
+      "loss": 0.8137,
+      "step": 22876
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5146086522989977,
+      "learning_rate": 9.224765859875157e-06,
+      "loss": 0.7887,
+      "step": 22877
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4513081825658394,
+      "learning_rate": 9.224005117420287e-06,
+      "loss": 0.606,
+      "step": 22878
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5969861801185041,
+      "learning_rate": 9.223244379483502e-06,
+      "loss": 0.5199,
+      "step": 22879
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3943006200574992,
+      "learning_rate": 9.222483646069233e-06,
+      "loss": 0.7386,
+      "step": 22880
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.279278147441846,
+      "learning_rate": 9.221722917181917e-06,
+      "loss": 0.6634,
+      "step": 22881
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.490071223019938,
+      "learning_rate": 9.22096219282597e-06,
+      "loss": 0.6341,
+      "step": 22882
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.7218257297717354,
+      "learning_rate": 9.220201473005828e-06,
+      "loss": 0.7393,
+      "step": 22883
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3205474467335672,
+      "learning_rate": 9.21944075772592e-06,
+      "loss": 0.7389,
+      "step": 22884
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4334868907547782,
+      "learning_rate": 9.218680046990678e-06,
+      "loss": 0.7599,
+      "step": 22885
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8760073360155443,
+      "learning_rate": 9.217919340804527e-06,
+      "loss": 0.7934,
+      "step": 22886
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5827191176600579,
+      "learning_rate": 9.217158639171898e-06,
+      "loss": 0.4935,
+      "step": 22887
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3286526845055757,
+      "learning_rate": 9.216397942097219e-06,
+      "loss": 0.6474,
+      "step": 22888
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.579255267943053,
+      "learning_rate": 9.215637249584915e-06,
+      "loss": 0.6892,
+      "step": 22889
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3916320008893373,
+      "learning_rate": 9.214876561639424e-06,
+      "loss": 0.8103,
+      "step": 22890
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3241971780198278,
+      "learning_rate": 9.214115878265168e-06,
+      "loss": 0.7062,
+      "step": 22891
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5326962167106104,
+      "learning_rate": 9.213355199466582e-06,
+      "loss": 0.6439,
+      "step": 22892
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.531436807751043,
+      "learning_rate": 9.212594525248086e-06,
+      "loss": 0.7398,
+      "step": 22893
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.484855396716464,
+      "learning_rate": 9.211833855614115e-06,
+      "loss": 0.6766,
+      "step": 22894
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5496033404281522,
+      "learning_rate": 9.2110731905691e-06,
+      "loss": 0.7192,
+      "step": 22895
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3282298489339335,
+      "learning_rate": 9.210312530117464e-06,
+      "loss": 0.7279,
+      "step": 22896
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7977466163459477,
+      "learning_rate": 9.209551874263637e-06,
+      "loss": 0.6991,
+      "step": 22897
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4329585181287763,
+      "learning_rate": 9.208791223012051e-06,
+      "loss": 0.6728,
+      "step": 22898
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.351105472607449,
+      "learning_rate": 9.208030576367136e-06,
+      "loss": 0.7464,
+      "step": 22899
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.317739575247488,
+      "learning_rate": 9.207269934333315e-06,
+      "loss": 0.7116,
+      "step": 22900
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4023229828159527,
+      "learning_rate": 9.206509296915016e-06,
+      "loss": 0.6882,
+      "step": 22901
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3636775992120749,
+      "learning_rate": 9.205748664116677e-06,
+      "loss": 0.6643,
+      "step": 22902
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6289575234650024,
+      "learning_rate": 9.20498803594272e-06,
+      "loss": 0.5639,
+      "step": 22903
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4489079820624728,
+      "learning_rate": 9.20422741239757e-06,
+      "loss": 0.6531,
+      "step": 22904
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.550900235092325,
+      "learning_rate": 9.203466793485662e-06,
+      "loss": 0.6892,
+      "step": 22905
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5652843165489134,
+      "learning_rate": 9.202706179211425e-06,
+      "loss": 0.5628,
+      "step": 22906
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5017556091641238,
+      "learning_rate": 9.201945569579283e-06,
+      "loss": 0.6916,
+      "step": 22907
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.480087614141506,
+      "learning_rate": 9.201184964593664e-06,
+      "loss": 0.6813,
+      "step": 22908
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.469509946545059,
+      "learning_rate": 9.200424364259006e-06,
+      "loss": 0.7294,
+      "step": 22909
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4058450262773703,
+      "learning_rate": 9.199663768579724e-06,
+      "loss": 0.6815,
+      "step": 22910
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2092025070276287,
+      "learning_rate": 9.198903177560255e-06,
+      "loss": 0.7171,
+      "step": 22911
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7409415754436648,
+      "learning_rate": 9.198142591205022e-06,
+      "loss": 0.7027,
+      "step": 22912
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2551880131839053,
+      "learning_rate": 9.197382009518465e-06,
+      "loss": 0.7247,
+      "step": 22913
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3691405905530214,
+      "learning_rate": 9.196621432504996e-06,
+      "loss": 0.8267,
+      "step": 22914
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5753935995589728,
+      "learning_rate": 9.195860860169054e-06,
+      "loss": 0.5231,
+      "step": 22915
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4223987880180518,
+      "learning_rate": 9.195100292515064e-06,
+      "loss": 0.6797,
+      "step": 22916
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2287199936429596,
+      "learning_rate": 9.194339729547461e-06,
+      "loss": 0.6305,
+      "step": 22917
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4800557527914546,
+      "learning_rate": 9.193579171270659e-06,
+      "loss": 0.8883,
+      "step": 22918
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7683953066232327,
+      "learning_rate": 9.192818617689097e-06,
+      "loss": 0.7494,
+      "step": 22919
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8987156874094764,
+      "learning_rate": 9.192058068807203e-06,
+      "loss": 0.7226,
+      "step": 22920
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5221725382963123,
+      "learning_rate": 9.191297524629398e-06,
+      "loss": 0.6833,
+      "step": 22921
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4852055935454798,
+      "learning_rate": 9.190536985160119e-06,
+      "loss": 0.7172,
+      "step": 22922
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6615693058861076,
+      "learning_rate": 9.189776450403789e-06,
+      "loss": 0.7095,
+      "step": 22923
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4145570844322852,
+      "learning_rate": 9.189015920364839e-06,
+      "loss": 0.5476,
+      "step": 22924
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.193389961944519,
+      "learning_rate": 9.18825539504769e-06,
+      "loss": 0.6352,
+      "step": 22925
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3786970482419758,
+      "learning_rate": 9.187494874456777e-06,
+      "loss": 0.6613,
+      "step": 22926
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6238604890505812,
+      "learning_rate": 9.186734358596529e-06,
+      "loss": 0.4569,
+      "step": 22927
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5632192922096186,
+      "learning_rate": 9.185973847471369e-06,
+      "loss": 0.8744,
+      "step": 22928
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.65353403494845,
+      "learning_rate": 9.185213341085725e-06,
+      "loss": 0.823,
+      "step": 22929
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3755277319872325,
+      "learning_rate": 9.184452839444029e-06,
+      "loss": 0.7488,
+      "step": 22930
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3259117670384915,
+      "learning_rate": 9.183692342550708e-06,
+      "loss": 0.6999,
+      "step": 22931
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.56277827050883,
+      "learning_rate": 9.182931850410187e-06,
+      "loss": 0.6199,
+      "step": 22932
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3902760852190328,
+      "learning_rate": 9.182171363026894e-06,
+      "loss": 0.7737,
+      "step": 22933
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2978272526143817,
+      "learning_rate": 9.181410880405262e-06,
+      "loss": 0.6647,
+      "step": 22934
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3612610320549239,
+      "learning_rate": 9.180650402549713e-06,
+      "loss": 0.7457,
+      "step": 22935
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3402948207588665,
+      "learning_rate": 9.179889929464676e-06,
+      "loss": 0.7991,
+      "step": 22936
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.629188753024008,
+      "learning_rate": 9.179129461154578e-06,
+      "loss": 0.5595,
+      "step": 22937
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.1961057079995854,
+      "learning_rate": 9.178368997623853e-06,
+      "loss": 0.6836,
+      "step": 22938
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.300390806127205,
+      "learning_rate": 9.17760853887692e-06,
+      "loss": 0.6777,
+      "step": 22939
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2397050756268893,
+      "learning_rate": 9.176848084918212e-06,
+      "loss": 0.7145,
+      "step": 22940
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5937764143281012,
+      "learning_rate": 9.176087635752157e-06,
+      "loss": 0.7688,
+      "step": 22941
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6018083904963427,
+      "learning_rate": 9.175327191383175e-06,
+      "loss": 0.4992,
+      "step": 22942
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3092256189236449,
+      "learning_rate": 9.174566751815704e-06,
+      "loss": 0.6432,
+      "step": 22943
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3772093225848896,
+      "learning_rate": 9.173806317054161e-06,
+      "loss": 0.8075,
+      "step": 22944
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3011850296517653,
+      "learning_rate": 9.173045887102988e-06,
+      "loss": 0.6811,
+      "step": 22945
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3589303825629364,
+      "learning_rate": 9.172285461966597e-06,
+      "loss": 0.7019,
+      "step": 22946
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3741090427139246,
+      "learning_rate": 9.171525041649423e-06,
+      "loss": 0.7236,
+      "step": 22947
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3778445548952907,
+      "learning_rate": 9.170764626155896e-06,
+      "loss": 0.6384,
+      "step": 22948
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4411480796173228,
+      "learning_rate": 9.170004215490438e-06,
+      "loss": 0.7686,
+      "step": 22949
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.420752225201321,
+      "learning_rate": 9.169243809657476e-06,
+      "loss": 0.7127,
+      "step": 22950
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4458508972087545,
+      "learning_rate": 9.16848340866144e-06,
+      "loss": 0.5974,
+      "step": 22951
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.395129056733623,
+      "learning_rate": 9.16772301250676e-06,
+      "loss": 0.8372,
+      "step": 22952
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4608441052419936,
+      "learning_rate": 9.16696262119786e-06,
+      "loss": 0.7065,
+      "step": 22953
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4029710454638697,
+      "learning_rate": 9.166202234739161e-06,
+      "loss": 0.7533,
+      "step": 22954
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4283482720691547,
+      "learning_rate": 9.165441853135104e-06,
+      "loss": 0.7594,
+      "step": 22955
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4269650489189236,
+      "learning_rate": 9.164681476390107e-06,
+      "loss": 0.7811,
+      "step": 22956
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3310741651758389,
+      "learning_rate": 9.163921104508593e-06,
+      "loss": 0.6367,
+      "step": 22957
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.396918793123303,
+      "learning_rate": 9.163160737495e-06,
+      "loss": 0.7396,
+      "step": 22958
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3571721102539165,
+      "learning_rate": 9.162400375353752e-06,
+      "loss": 0.7475,
+      "step": 22959
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3638661510859682,
+      "learning_rate": 9.161640018089273e-06,
+      "loss": 0.6912,
+      "step": 22960
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3549147868131524,
+      "learning_rate": 9.160879665705988e-06,
+      "loss": 0.8384,
+      "step": 22961
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3036658202894016,
+      "learning_rate": 9.160119318208334e-06,
+      "loss": 0.7237,
+      "step": 22962
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4470066950666407,
+      "learning_rate": 9.159358975600723e-06,
+      "loss": 0.6602,
+      "step": 22963
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3771925448994573,
+      "learning_rate": 9.158598637887594e-06,
+      "loss": 0.6745,
+      "step": 22964
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.376362332735835,
+      "learning_rate": 9.15783830507337e-06,
+      "loss": 0.776,
+      "step": 22965
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3080459257142887,
+      "learning_rate": 9.157077977162481e-06,
+      "loss": 0.6712,
+      "step": 22966
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4522140193564692,
+      "learning_rate": 9.156317654159346e-06,
+      "loss": 0.725,
+      "step": 22967
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3240584238330786,
+      "learning_rate": 9.155557336068399e-06,
+      "loss": 0.71,
+      "step": 22968
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.798592199505768,
+      "learning_rate": 9.154797022894067e-06,
+      "loss": 0.7392,
+      "step": 22969
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5805288055677443,
+      "learning_rate": 9.154036714640768e-06,
+      "loss": 0.5535,
+      "step": 22970
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.578016095144511,
+      "learning_rate": 9.15327641131294e-06,
+      "loss": 0.6746,
+      "step": 22971
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.36828175236129,
+      "learning_rate": 9.152516112915002e-06,
+      "loss": 0.74,
+      "step": 22972
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4028958706249985,
+      "learning_rate": 9.151755819451387e-06,
+      "loss": 0.6852,
+      "step": 22973
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5204343603757329,
+      "learning_rate": 9.150995530926515e-06,
+      "loss": 0.659,
+      "step": 22974
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8498927185776262,
+      "learning_rate": 9.150235247344817e-06,
+      "loss": 0.6819,
+      "step": 22975
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5089928881108259,
+      "learning_rate": 9.14947496871072e-06,
+      "loss": 0.6253,
+      "step": 22976
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4613326074177575,
+      "learning_rate": 9.148714695028647e-06,
+      "loss": 0.6736,
+      "step": 22977
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3245859345071387,
+      "learning_rate": 9.147954426303024e-06,
+      "loss": 0.6458,
+      "step": 22978
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5919702603395325,
+      "learning_rate": 9.147194162538284e-06,
+      "loss": 0.5261,
+      "step": 22979
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7767168661001649,
+      "learning_rate": 9.14643390373885e-06,
+      "loss": 0.6675,
+      "step": 22980
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2546337664460596,
+      "learning_rate": 9.145673649909147e-06,
+      "loss": 0.7758,
+      "step": 22981
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5781712837591525,
+      "learning_rate": 9.144913401053599e-06,
+      "loss": 0.5111,
+      "step": 22982
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3516528201338212,
+      "learning_rate": 9.144153157176641e-06,
+      "loss": 0.8601,
+      "step": 22983
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4525582615216976,
+      "learning_rate": 9.143392918282693e-06,
+      "loss": 0.6107,
+      "step": 22984
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6038754257112923,
+      "learning_rate": 9.14263268437618e-06,
+      "loss": 0.5316,
+      "step": 22985
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.443588574609737,
+      "learning_rate": 9.141872455461532e-06,
+      "loss": 0.6178,
+      "step": 22986
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5531903964983944,
+      "learning_rate": 9.141112231543177e-06,
+      "loss": 0.8727,
+      "step": 22987
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7536631075346931,
+      "learning_rate": 9.140352012625538e-06,
+      "loss": 0.7846,
+      "step": 22988
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.535979402125163,
+      "learning_rate": 9.139591798713038e-06,
+      "loss": 0.6984,
+      "step": 22989
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3947140472075055,
+      "learning_rate": 9.138831589810111e-06,
+      "loss": 0.7168,
+      "step": 22990
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5525636878578045,
+      "learning_rate": 9.138071385921176e-06,
+      "loss": 0.5078,
+      "step": 22991
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.485153826594586,
+      "learning_rate": 9.137311187050663e-06,
+      "loss": 0.7761,
+      "step": 22992
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.443615983606844,
+      "learning_rate": 9.136550993202997e-06,
+      "loss": 0.7238,
+      "step": 22993
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.620074202519543,
+      "learning_rate": 9.13579080438261e-06,
+      "loss": 0.4875,
+      "step": 22994
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5573877158369658,
+      "learning_rate": 9.135030620593915e-06,
+      "loss": 0.6698,
+      "step": 22995
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4029345507397568,
+      "learning_rate": 9.13427044184135e-06,
+      "loss": 0.7975,
+      "step": 22996
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3303705130528358,
+      "learning_rate": 9.133510268129337e-06,
+      "loss": 0.6788,
+      "step": 22997
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5609664630087017,
+      "learning_rate": 9.1327500994623e-06,
+      "loss": 0.4766,
+      "step": 22998
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5516998231303718,
+      "learning_rate": 9.131989935844663e-06,
+      "loss": 0.7805,
+      "step": 22999
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5203004303338963,
+      "learning_rate": 9.131229777280859e-06,
+      "loss": 0.5255,
+      "step": 23000
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3564966273737282,
+      "learning_rate": 9.130469623775312e-06,
+      "loss": 0.7269,
+      "step": 23001
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7005844992962003,
+      "learning_rate": 9.129709475332445e-06,
+      "loss": 0.6425,
+      "step": 23002
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6136532398574587,
+      "learning_rate": 9.12894933195668e-06,
+      "loss": 0.4986,
+      "step": 23003
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3693672319388255,
+      "learning_rate": 9.128189193652455e-06,
+      "loss": 0.7433,
+      "step": 23004
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2932709640832862,
+      "learning_rate": 9.127429060424184e-06,
+      "loss": 0.662,
+      "step": 23005
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4504378320302542,
+      "learning_rate": 9.126668932276298e-06,
+      "loss": 0.7895,
+      "step": 23006
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5641455527174429,
+      "learning_rate": 9.125908809213221e-06,
+      "loss": 0.5114,
+      "step": 23007
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5624750464625582,
+      "learning_rate": 9.125148691239383e-06,
+      "loss": 0.5095,
+      "step": 23008
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7601399652167817,
+      "learning_rate": 9.124388578359206e-06,
+      "loss": 0.7784,
+      "step": 23009
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9357742852227604,
+      "learning_rate": 9.123628470577111e-06,
+      "loss": 0.7381,
+      "step": 23010
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.381830817283034,
+      "learning_rate": 9.122868367897537e-06,
+      "loss": 0.6914,
+      "step": 23011
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.308874423943274,
+      "learning_rate": 9.122108270324894e-06,
+      "loss": 0.5874,
+      "step": 23012
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6183598339477872,
+      "learning_rate": 9.121348177863615e-06,
+      "loss": 0.6527,
+      "step": 23013
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9049752809180656,
+      "learning_rate": 9.120588090518126e-06,
+      "loss": 0.714,
+      "step": 23014
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5091934662490811,
+      "learning_rate": 9.119828008292856e-06,
+      "loss": 0.6505,
+      "step": 23015
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4845112349473784,
+      "learning_rate": 9.11906793119222e-06,
+      "loss": 0.7502,
+      "step": 23016
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6395683778099943,
+      "learning_rate": 9.118307859220651e-06,
+      "loss": 0.8223,
+      "step": 23017
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5654813386264076,
+      "learning_rate": 9.117547792382577e-06,
+      "loss": 0.5051,
+      "step": 23018
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3653010745282537,
+      "learning_rate": 9.116787730682413e-06,
+      "loss": 0.779,
+      "step": 23019
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.181886637698245,
+      "learning_rate": 9.116027674124592e-06,
+      "loss": 0.7597,
+      "step": 23020
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6055983670593938,
+      "learning_rate": 9.115267622713539e-06,
+      "loss": 0.5685,
+      "step": 23021
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4999473921091833,
+      "learning_rate": 9.11450757645368e-06,
+      "loss": 0.8113,
+      "step": 23022
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3857809988422747,
+      "learning_rate": 9.113747535349433e-06,
+      "loss": 0.742,
+      "step": 23023
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3460645900005082,
+      "learning_rate": 9.112987499405232e-06,
+      "loss": 0.6357,
+      "step": 23024
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3345659310969455,
+      "learning_rate": 9.112227468625499e-06,
+      "loss": 0.6091,
+      "step": 23025
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.848872199095552,
+      "learning_rate": 9.11146744301466e-06,
+      "loss": 0.6593,
+      "step": 23026
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4005661770460713,
+      "learning_rate": 9.110707422577133e-06,
+      "loss": 0.6718,
+      "step": 23027
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5643863777405222,
+      "learning_rate": 9.109947407317352e-06,
+      "loss": 0.7256,
+      "step": 23028
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.456558683071159,
+      "learning_rate": 9.109187397239742e-06,
+      "loss": 0.7072,
+      "step": 23029
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6040092260013504,
+      "learning_rate": 9.108427392348722e-06,
+      "loss": 0.5565,
+      "step": 23030
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5002215923724733,
+      "learning_rate": 9.107667392648718e-06,
+      "loss": 0.8878,
+      "step": 23031
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.578325140709709,
+      "learning_rate": 9.10690739814416e-06,
+      "loss": 0.5348,
+      "step": 23032
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2546201166924082,
+      "learning_rate": 9.10614740883947e-06,
+      "loss": 0.7223,
+      "step": 23033
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3545809861953848,
+      "learning_rate": 9.105387424739073e-06,
+      "loss": 0.7607,
+      "step": 23034
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4668501823799343,
+      "learning_rate": 9.10462744584739e-06,
+      "loss": 0.6773,
+      "step": 23035
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.275439711287178,
+      "learning_rate": 9.103867472168854e-06,
+      "loss": 0.6317,
+      "step": 23036
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4216506837692657,
+      "learning_rate": 9.103107503707885e-06,
+      "loss": 0.7608,
+      "step": 23037
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4369245888823332,
+      "learning_rate": 9.102347540468906e-06,
+      "loss": 0.8019,
+      "step": 23038
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5865395212673533,
+      "learning_rate": 9.101587582456342e-06,
+      "loss": 0.5548,
+      "step": 23039
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3962221169988869,
+      "learning_rate": 9.100827629674623e-06,
+      "loss": 0.8251,
+      "step": 23040
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5408141107681608,
+      "learning_rate": 9.10006768212817e-06,
+      "loss": 0.7838,
+      "step": 23041
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7498830693175773,
+      "learning_rate": 9.099307739821404e-06,
+      "loss": 0.5866,
+      "step": 23042
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5929123136894312,
+      "learning_rate": 9.098547802758762e-06,
+      "loss": 0.7124,
+      "step": 23043
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.428357316305938,
+      "learning_rate": 9.09778787094465e-06,
+      "loss": 0.6894,
+      "step": 23044
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5140988094974446,
+      "learning_rate": 9.097027944383508e-06,
+      "loss": 0.7231,
+      "step": 23045
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5421899450970669,
+      "learning_rate": 9.096268023079753e-06,
+      "loss": 0.7487,
+      "step": 23046
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.266871229180114,
+      "learning_rate": 9.095508107037817e-06,
+      "loss": 0.7823,
+      "step": 23047
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3831304259874286,
+      "learning_rate": 9.094748196262114e-06,
+      "loss": 0.7896,
+      "step": 23048
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2931074053997071,
+      "learning_rate": 9.093988290757075e-06,
+      "loss": 0.6804,
+      "step": 23049
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3551423458755598,
+      "learning_rate": 9.093228390527123e-06,
+      "loss": 0.6986,
+      "step": 23050
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.490275198056559,
+      "learning_rate": 9.092468495576682e-06,
+      "loss": 0.6653,
+      "step": 23051
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3980270275024864,
+      "learning_rate": 9.091708605910175e-06,
+      "loss": 0.7135,
+      "step": 23052
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3222474795092019,
+      "learning_rate": 9.09094872153203e-06,
+      "loss": 0.7731,
+      "step": 23053
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3976846104585015,
+      "learning_rate": 9.090188842446672e-06,
+      "loss": 0.8278,
+      "step": 23054
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7111108126665402,
+      "learning_rate": 9.089428968658518e-06,
+      "loss": 0.719,
+      "step": 23055
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.013093816787791,
+      "learning_rate": 9.088669100171999e-06,
+      "loss": 0.6952,
+      "step": 23056
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3632507348022713,
+      "learning_rate": 9.08790923699154e-06,
+      "loss": 0.7538,
+      "step": 23057
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4851838783014888,
+      "learning_rate": 9.087149379121556e-06,
+      "loss": 0.7723,
+      "step": 23058
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4501420153551854,
+      "learning_rate": 9.08638952656648e-06,
+      "loss": 0.618,
+      "step": 23059
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.471811588447836,
+      "learning_rate": 9.085629679330733e-06,
+      "loss": 0.6694,
+      "step": 23060
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.635468363866237,
+      "learning_rate": 9.084869837418742e-06,
+      "loss": 0.7043,
+      "step": 23061
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6029084728673172,
+      "learning_rate": 9.084110000834928e-06,
+      "loss": 0.5524,
+      "step": 23062
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5203889806185382,
+      "learning_rate": 9.083350169583711e-06,
+      "loss": 0.6196,
+      "step": 23063
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5639159668425735,
+      "learning_rate": 9.082590343669526e-06,
+      "loss": 0.7314,
+      "step": 23064
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3564207161993358,
+      "learning_rate": 9.081830523096786e-06,
+      "loss": 0.5793,
+      "step": 23065
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2308780225272493,
+      "learning_rate": 9.081070707869921e-06,
+      "loss": 0.7962,
+      "step": 23066
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5632668628708553,
+      "learning_rate": 9.08031089799335e-06,
+      "loss": 0.8717,
+      "step": 23067
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4410147114555256,
+      "learning_rate": 9.079551093471507e-06,
+      "loss": 0.6886,
+      "step": 23068
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.349855785580122,
+      "learning_rate": 9.078791294308803e-06,
+      "loss": 0.6456,
+      "step": 23069
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.35830948926742,
+      "learning_rate": 9.07803150050967e-06,
+      "loss": 0.5914,
+      "step": 23070
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3636375823349072,
+      "learning_rate": 9.077271712078531e-06,
+      "loss": 0.6301,
+      "step": 23071
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6057764076114809,
+      "learning_rate": 9.076511929019804e-06,
+      "loss": 0.5502,
+      "step": 23072
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3689864509039464,
+      "learning_rate": 9.07575215133792e-06,
+      "loss": 0.7432,
+      "step": 23073
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.513814935509072,
+      "learning_rate": 9.074992379037298e-06,
+      "loss": 0.7334,
+      "step": 23074
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4355177079513122,
+      "learning_rate": 9.074232612122367e-06,
+      "loss": 0.6345,
+      "step": 23075
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2640574609767248,
+      "learning_rate": 9.073472850597542e-06,
+      "loss": 0.7249,
+      "step": 23076
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3436491108030009,
+      "learning_rate": 9.072713094467253e-06,
+      "loss": 0.7345,
+      "step": 23077
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5298067020929453,
+      "learning_rate": 9.071953343735925e-06,
+      "loss": 0.7987,
+      "step": 23078
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4544601464041556,
+      "learning_rate": 9.071193598407975e-06,
+      "loss": 0.6908,
+      "step": 23079
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7292472228109261,
+      "learning_rate": 9.070433858487829e-06,
+      "loss": 0.7157,
+      "step": 23080
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5342170879248396,
+      "learning_rate": 9.069674123979912e-06,
+      "loss": 0.7317,
+      "step": 23081
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6035377278035248,
+      "learning_rate": 9.068914394888651e-06,
+      "loss": 0.4916,
+      "step": 23082
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.585781599958291,
+      "learning_rate": 9.068154671218462e-06,
+      "loss": 0.6257,
+      "step": 23083
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5056201586945126,
+      "learning_rate": 9.06739495297377e-06,
+      "loss": 0.705,
+      "step": 23084
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.255391492537868,
+      "learning_rate": 9.066635240159002e-06,
+      "loss": 0.698,
+      "step": 23085
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.366130404655325,
+      "learning_rate": 9.065875532778578e-06,
+      "loss": 0.8071,
+      "step": 23086
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4956097235776338,
+      "learning_rate": 9.065115830836923e-06,
+      "loss": 0.8423,
+      "step": 23087
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3305782241526132,
+      "learning_rate": 9.064356134338459e-06,
+      "loss": 0.6542,
+      "step": 23088
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4288508582456525,
+      "learning_rate": 9.063596443287612e-06,
+      "loss": 0.6471,
+      "step": 23089
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5671653223252254,
+      "learning_rate": 9.0628367576888e-06,
+      "loss": 0.5024,
+      "step": 23090
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.426353606245247,
+      "learning_rate": 9.062077077546449e-06,
+      "loss": 0.805,
+      "step": 23091
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6410520506201207,
+      "learning_rate": 9.061317402864988e-06,
+      "loss": 0.6463,
+      "step": 23092
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5861061337702174,
+      "learning_rate": 9.060557733648827e-06,
+      "loss": 0.6607,
+      "step": 23093
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4603473619201928,
+      "learning_rate": 9.059798069902399e-06,
+      "loss": 0.699,
+      "step": 23094
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3647983691198098,
+      "learning_rate": 9.059038411630123e-06,
+      "loss": 0.7196,
+      "step": 23095
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2820844785360763,
+      "learning_rate": 9.058278758836428e-06,
+      "loss": 0.6707,
+      "step": 23096
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5319569614753012,
+      "learning_rate": 9.057519111525728e-06,
+      "loss": 0.7089,
+      "step": 23097
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3673730831232769,
+      "learning_rate": 9.05675946970245e-06,
+      "loss": 0.7062,
+      "step": 23098
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1117416336625765,
+      "learning_rate": 9.05599983337102e-06,
+      "loss": 0.7882,
+      "step": 23099
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2412780420146317,
+      "learning_rate": 9.055240202535856e-06,
+      "loss": 0.6272,
+      "step": 23100
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6034498646665837,
+      "learning_rate": 9.054480577201381e-06,
+      "loss": 0.5572,
+      "step": 23101
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1285473381082394,
+      "learning_rate": 9.05372095737202e-06,
+      "loss": 0.5866,
+      "step": 23102
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5354233571682445,
+      "learning_rate": 9.052961343052199e-06,
+      "loss": 0.8164,
+      "step": 23103
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.40324817058828,
+      "learning_rate": 9.052201734246332e-06,
+      "loss": 0.7683,
+      "step": 23104
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6003919556162212,
+      "learning_rate": 9.051442130958848e-06,
+      "loss": 0.7663,
+      "step": 23105
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4687258299393315,
+      "learning_rate": 9.050682533194171e-06,
+      "loss": 0.6647,
+      "step": 23106
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4049936842848283,
+      "learning_rate": 9.049922940956719e-06,
+      "loss": 0.7572,
+      "step": 23107
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3899012706756735,
+      "learning_rate": 9.049163354250914e-06,
+      "loss": 0.7965,
+      "step": 23108
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4629110045037692,
+      "learning_rate": 9.048403773081182e-06,
+      "loss": 0.7339,
+      "step": 23109
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3938171798974088,
+      "learning_rate": 9.047644197451949e-06,
+      "loss": 0.6693,
+      "step": 23110
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.55278111652132,
+      "learning_rate": 9.046884627367628e-06,
+      "loss": 0.7104,
+      "step": 23111
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9773122266195164,
+      "learning_rate": 9.046125062832646e-06,
+      "loss": 0.7054,
+      "step": 23112
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5823790508568503,
+      "learning_rate": 9.045365503851432e-06,
+      "loss": 0.5398,
+      "step": 23113
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.520945477456668,
+      "learning_rate": 9.044605950428397e-06,
+      "loss": 0.6559,
+      "step": 23114
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1882863329307385,
+      "learning_rate": 9.043846402567969e-06,
+      "loss": 0.6583,
+      "step": 23115
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.603690242844687,
+      "learning_rate": 9.043086860274567e-06,
+      "loss": 0.6718,
+      "step": 23116
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.30564530900319,
+      "learning_rate": 9.042327323552624e-06,
+      "loss": 0.7091,
+      "step": 23117
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6150221275698837,
+      "learning_rate": 9.04156779240655e-06,
+      "loss": 0.5672,
+      "step": 23118
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3751916551995078,
+      "learning_rate": 9.040808266840773e-06,
+      "loss": 0.7464,
+      "step": 23119
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0530749618545667,
+      "learning_rate": 9.040048746859715e-06,
+      "loss": 0.6888,
+      "step": 23120
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5927789946001754,
+      "learning_rate": 9.039289232467792e-06,
+      "loss": 0.6331,
+      "step": 23121
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2550822442539156,
+      "learning_rate": 9.038529723669435e-06,
+      "loss": 0.6865,
+      "step": 23122
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3321327693485359,
+      "learning_rate": 9.037770220469062e-06,
+      "loss": 0.6083,
+      "step": 23123
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4575296214627436,
+      "learning_rate": 9.037010722871098e-06,
+      "loss": 0.7118,
+      "step": 23124
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3898570796035594,
+      "learning_rate": 9.036251230879956e-06,
+      "loss": 0.6794,
+      "step": 23125
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4187730157776457,
+      "learning_rate": 9.035491744500069e-06,
+      "loss": 0.7479,
+      "step": 23126
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3150630645749979,
+      "learning_rate": 9.034732263735856e-06,
+      "loss": 0.6639,
+      "step": 23127
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9279953000070622,
+      "learning_rate": 9.033972788591735e-06,
+      "loss": 0.7929,
+      "step": 23128
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3568008146365342,
+      "learning_rate": 9.033213319072126e-06,
+      "loss": 0.6759,
+      "step": 23129
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8830770828515997,
+      "learning_rate": 9.032453855181462e-06,
+      "loss": 0.7139,
+      "step": 23130
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.331444543257667,
+      "learning_rate": 9.031694396924156e-06,
+      "loss": 0.7277,
+      "step": 23131
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4457888665826761,
+      "learning_rate": 9.030934944304632e-06,
+      "loss": 0.7279,
+      "step": 23132
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5427841735359258,
+      "learning_rate": 9.030175497327308e-06,
+      "loss": 0.8269,
+      "step": 23133
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3296530177550312,
+      "learning_rate": 9.029416055996614e-06,
+      "loss": 0.6598,
+      "step": 23134
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.097491607455712,
+      "learning_rate": 9.028656620316964e-06,
+      "loss": 0.6606,
+      "step": 23135
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2730858816635906,
+      "learning_rate": 9.027897190292784e-06,
+      "loss": 0.7117,
+      "step": 23136
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.641163064266935,
+      "learning_rate": 9.027137765928492e-06,
+      "loss": 0.586,
+      "step": 23137
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4921412649983787,
+      "learning_rate": 9.026378347228516e-06,
+      "loss": 0.6648,
+      "step": 23138
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.413862830173498,
+      "learning_rate": 9.025618934197272e-06,
+      "loss": 0.7173,
+      "step": 23139
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5463608593211937,
+      "learning_rate": 9.02485952683918e-06,
+      "loss": 0.5502,
+      "step": 23140
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5130286527466217,
+      "learning_rate": 9.024100125158671e-06,
+      "loss": 0.6836,
+      "step": 23141
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3466893453736672,
+      "learning_rate": 9.023340729160154e-06,
+      "loss": 0.7145,
+      "step": 23142
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2745902422458468,
+      "learning_rate": 9.022581338848058e-06,
+      "loss": 0.757,
+      "step": 23143
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.501577751897927,
+      "learning_rate": 9.021821954226801e-06,
+      "loss": 0.8348,
+      "step": 23144
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7181206727208111,
+      "learning_rate": 9.021062575300814e-06,
+      "loss": 0.8693,
+      "step": 23145
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4795211547295326,
+      "learning_rate": 9.020303202074503e-06,
+      "loss": 0.6499,
+      "step": 23146
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3935930799539307,
+      "learning_rate": 9.0195438345523e-06,
+      "loss": 0.6346,
+      "step": 23147
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6891053764453965,
+      "learning_rate": 9.018784472738624e-06,
+      "loss": 0.6793,
+      "step": 23148
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5603169086980433,
+      "learning_rate": 9.018025116637895e-06,
+      "loss": 0.7297,
+      "step": 23149
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2126714158764096,
+      "learning_rate": 9.01726576625453e-06,
+      "loss": 0.6407,
+      "step": 23150
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3675914616351255,
+      "learning_rate": 9.016506421592959e-06,
+      "loss": 0.6304,
+      "step": 23151
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1824721542489904,
+      "learning_rate": 9.0157470826576e-06,
+      "loss": 0.7442,
+      "step": 23152
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6012001766040416,
+      "learning_rate": 9.01498774945287e-06,
+      "loss": 0.5007,
+      "step": 23153
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.22305578735061,
+      "learning_rate": 9.014228421983194e-06,
+      "loss": 0.6123,
+      "step": 23154
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5878247993585164,
+      "learning_rate": 9.013469100252995e-06,
+      "loss": 0.4355,
+      "step": 23155
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6959807960862918,
+      "learning_rate": 9.012709784266688e-06,
+      "loss": 0.6866,
+      "step": 23156
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4493834164699906,
+      "learning_rate": 9.011950474028697e-06,
+      "loss": 0.7086,
+      "step": 23157
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6006395955292188,
+      "learning_rate": 9.011191169543444e-06,
+      "loss": 0.5224,
+      "step": 23158
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.440265065834088,
+      "learning_rate": 9.010431870815351e-06,
+      "loss": 0.7444,
+      "step": 23159
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5773972828403202,
+      "learning_rate": 9.009672577848836e-06,
+      "loss": 0.5128,
+      "step": 23160
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4716963662148967,
+      "learning_rate": 9.008913290648317e-06,
+      "loss": 0.6822,
+      "step": 23161
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.125460494541385,
+      "learning_rate": 9.008154009218228e-06,
+      "loss": 0.7582,
+      "step": 23162
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4162229498305736,
+      "learning_rate": 9.00739473356297e-06,
+      "loss": 0.6687,
+      "step": 23163
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3660642075044993,
+      "learning_rate": 9.00663546368698e-06,
+      "loss": 0.6751,
+      "step": 23164
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5680961579793953,
+      "learning_rate": 9.005876199594668e-06,
+      "loss": 0.6867,
+      "step": 23165
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6627061360353177,
+      "learning_rate": 9.005116941290465e-06,
+      "loss": 0.7416,
+      "step": 23166
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4219725691744514,
+      "learning_rate": 9.004357688778784e-06,
+      "loss": 0.7107,
+      "step": 23167
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5787409278198274,
+      "learning_rate": 9.003598442064048e-06,
+      "loss": 0.5209,
+      "step": 23168
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5938461135335467,
+      "learning_rate": 9.002839201150673e-06,
+      "loss": 0.5363,
+      "step": 23169
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2819410271827558,
+      "learning_rate": 9.00207996604309e-06,
+      "loss": 0.6341,
+      "step": 23170
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4307948951312124,
+      "learning_rate": 9.001320736745711e-06,
+      "loss": 0.6563,
+      "step": 23171
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6742501995033499,
+      "learning_rate": 9.00056151326296e-06,
+      "loss": 0.7209,
+      "step": 23172
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2004240766404157,
+      "learning_rate": 8.999802295599257e-06,
+      "loss": 0.6492,
+      "step": 23173
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5526653735274631,
+      "learning_rate": 8.999043083759016e-06,
+      "loss": 0.6748,
+      "step": 23174
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3798174034996298,
+      "learning_rate": 8.998283877746669e-06,
+      "loss": 0.711,
+      "step": 23175
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5167038771380827,
+      "learning_rate": 8.997524677566627e-06,
+      "loss": 0.7093,
+      "step": 23176
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.431212079423237,
+      "learning_rate": 8.996765483223317e-06,
+      "loss": 0.8379,
+      "step": 23177
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.152233173178862,
+      "learning_rate": 8.996006294721153e-06,
+      "loss": 0.6617,
+      "step": 23178
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1953876108020953,
+      "learning_rate": 8.99524711206456e-06,
+      "loss": 0.5722,
+      "step": 23179
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9582018930772478,
+      "learning_rate": 8.994487935257957e-06,
+      "loss": 0.7582,
+      "step": 23180
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4682727302797103,
+      "learning_rate": 8.993728764305762e-06,
+      "loss": 0.7457,
+      "step": 23181
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9074894622768055,
+      "learning_rate": 8.992969599212395e-06,
+      "loss": 0.6215,
+      "step": 23182
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.404109165086267,
+      "learning_rate": 8.99221043998228e-06,
+      "loss": 0.6917,
+      "step": 23183
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.399721408290975,
+      "learning_rate": 8.991451286619837e-06,
+      "loss": 0.6864,
+      "step": 23184
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2774515308946974,
+      "learning_rate": 8.990692139129481e-06,
+      "loss": 0.7271,
+      "step": 23185
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3633613104461304,
+      "learning_rate": 8.989932997515634e-06,
+      "loss": 0.7449,
+      "step": 23186
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5845325624955529,
+      "learning_rate": 8.98917386178272e-06,
+      "loss": 0.5121,
+      "step": 23187
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3427682227228515,
+      "learning_rate": 8.988414731935153e-06,
+      "loss": 0.7397,
+      "step": 23188
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4767659110438893,
+      "learning_rate": 8.987655607977357e-06,
+      "loss": 0.7467,
+      "step": 23189
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5805959207147663,
+      "learning_rate": 8.98689648991375e-06,
+      "loss": 0.4725,
+      "step": 23190
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2379014599320253,
+      "learning_rate": 8.986137377748754e-06,
+      "loss": 0.7187,
+      "step": 23191
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3023912204255461,
+      "learning_rate": 8.985378271486786e-06,
+      "loss": 0.6437,
+      "step": 23192
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.79743713970002,
+      "learning_rate": 8.984619171132264e-06,
+      "loss": 0.7082,
+      "step": 23193
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5309493748767897,
+      "learning_rate": 8.983860076689617e-06,
+      "loss": 0.6033,
+      "step": 23194
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5396875891172583,
+      "learning_rate": 8.983100988163253e-06,
+      "loss": 0.6467,
+      "step": 23195
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3052639474338985,
+      "learning_rate": 8.982341905557599e-06,
+      "loss": 0.5985,
+      "step": 23196
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5212596005287762,
+      "learning_rate": 8.981582828877072e-06,
+      "loss": 0.7628,
+      "step": 23197
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3314272212254963,
+      "learning_rate": 8.980823758126096e-06,
+      "loss": 0.5677,
+      "step": 23198
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4969635795024372,
+      "learning_rate": 8.98006469330908e-06,
+      "loss": 0.8342,
+      "step": 23199
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6064283431038826,
+      "learning_rate": 8.979305634430453e-06,
+      "loss": 0.7298,
+      "step": 23200
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3535027366598418,
+      "learning_rate": 8.978546581494637e-06,
+      "loss": 0.8097,
+      "step": 23201
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5843489876418074,
+      "learning_rate": 8.977787534506038e-06,
+      "loss": 0.4964,
+      "step": 23202
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.048221999888769,
+      "learning_rate": 8.977028493469088e-06,
+      "loss": 0.7251,
+      "step": 23203
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6161165409541169,
+      "learning_rate": 8.976269458388201e-06,
+      "loss": 0.5242,
+      "step": 23204
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7190474763125914,
+      "learning_rate": 8.9755104292678e-06,
+      "loss": 0.6982,
+      "step": 23205
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4723095137839963,
+      "learning_rate": 8.974751406112296e-06,
+      "loss": 0.7026,
+      "step": 23206
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5897957549693724,
+      "learning_rate": 8.973992388926117e-06,
+      "loss": 0.7295,
+      "step": 23207
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5930236079298927,
+      "learning_rate": 8.973233377713681e-06,
+      "loss": 0.4713,
+      "step": 23208
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4108189206077935,
+      "learning_rate": 8.972474372479404e-06,
+      "loss": 0.7003,
+      "step": 23209
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4518867917763159,
+      "learning_rate": 8.971715373227704e-06,
+      "loss": 0.7929,
+      "step": 23210
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7033955490368509,
+      "learning_rate": 8.970956379963006e-06,
+      "loss": 0.6522,
+      "step": 23211
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5960523493211143,
+      "learning_rate": 8.970197392689725e-06,
+      "loss": 0.7192,
+      "step": 23212
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6745405605423196,
+      "learning_rate": 8.96943841141228e-06,
+      "loss": 0.732,
+      "step": 23213
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3686675836669495,
+      "learning_rate": 8.96867943613509e-06,
+      "loss": 0.8553,
+      "step": 23214
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.499727153249136,
+      "learning_rate": 8.967920466862577e-06,
+      "loss": 0.7754,
+      "step": 23215
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1070732919769886,
+      "learning_rate": 8.967161503599157e-06,
+      "loss": 0.7502,
+      "step": 23216
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.558126402329838,
+      "learning_rate": 8.96640254634925e-06,
+      "loss": 0.6641,
+      "step": 23217
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.661854126852827,
+      "learning_rate": 8.96564359511727e-06,
+      "loss": 0.6944,
+      "step": 23218
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2957800708570482,
+      "learning_rate": 8.964884649907648e-06,
+      "loss": 0.5923,
+      "step": 23219
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4706609351903355,
+      "learning_rate": 8.964125710724791e-06,
+      "loss": 0.6501,
+      "step": 23220
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.211772052630598,
+      "learning_rate": 8.963366777573123e-06,
+      "loss": 0.8298,
+      "step": 23221
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.601011160865644,
+      "learning_rate": 8.962607850457063e-06,
+      "loss": 0.6413,
+      "step": 23222
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6944845946763163,
+      "learning_rate": 8.961848929381027e-06,
+      "loss": 0.7471,
+      "step": 23223
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3997836761012428,
+      "learning_rate": 8.961090014349433e-06,
+      "loss": 0.7614,
+      "step": 23224
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5706985532433462,
+      "learning_rate": 8.960331105366705e-06,
+      "loss": 0.6033,
+      "step": 23225
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4248623081243164,
+      "learning_rate": 8.95957220243726e-06,
+      "loss": 0.5676,
+      "step": 23226
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.579581230475881,
+      "learning_rate": 8.95881330556551e-06,
+      "loss": 0.4751,
+      "step": 23227
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5231287881962632,
+      "learning_rate": 8.958054414755882e-06,
+      "loss": 0.7244,
+      "step": 23228
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.451946639108578,
+      "learning_rate": 8.957295530012792e-06,
+      "loss": 0.7851,
+      "step": 23229
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.559521835180739,
+      "learning_rate": 8.956536651340655e-06,
+      "loss": 0.6516,
+      "step": 23230
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.655851490051642,
+      "learning_rate": 8.955777778743891e-06,
+      "loss": 0.6768,
+      "step": 23231
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8241826755753348,
+      "learning_rate": 8.95501891222692e-06,
+      "loss": 0.6323,
+      "step": 23232
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.223669612790238,
+      "learning_rate": 8.954260051794165e-06,
+      "loss": 0.6114,
+      "step": 23233
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3016802907069502,
+      "learning_rate": 8.953501197450035e-06,
+      "loss": 0.7366,
+      "step": 23234
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5981456113903296,
+      "learning_rate": 8.952742349198951e-06,
+      "loss": 0.723,
+      "step": 23235
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3250962873117136,
+      "learning_rate": 8.951983507045335e-06,
+      "loss": 0.7406,
+      "step": 23236
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3918348289516451,
+      "learning_rate": 8.951224670993603e-06,
+      "loss": 0.6297,
+      "step": 23237
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.516553731369105,
+      "learning_rate": 8.950465841048168e-06,
+      "loss": 0.5204,
+      "step": 23238
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8496318992453038,
+      "learning_rate": 8.949707017213457e-06,
+      "loss": 0.5568,
+      "step": 23239
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3652832311855048,
+      "learning_rate": 8.948948199493887e-06,
+      "loss": 0.8175,
+      "step": 23240
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4368408536902093,
+      "learning_rate": 8.948189387893871e-06,
+      "loss": 0.6739,
+      "step": 23241
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2995778462381713,
+      "learning_rate": 8.947430582417827e-06,
+      "loss": 0.7423,
+      "step": 23242
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.587159963280999,
+      "learning_rate": 8.946671783070182e-06,
+      "loss": 0.8066,
+      "step": 23243
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.1149511214204586,
+      "learning_rate": 8.94591298985534e-06,
+      "loss": 0.7292,
+      "step": 23244
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3457639712833935,
+      "learning_rate": 8.945154202777729e-06,
+      "loss": 0.6568,
+      "step": 23245
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9436821010196752,
+      "learning_rate": 8.94439542184176e-06,
+      "loss": 0.6032,
+      "step": 23246
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5593241279731006,
+      "learning_rate": 8.943636647051864e-06,
+      "loss": 0.527,
+      "step": 23247
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3800056424195781,
+      "learning_rate": 8.942877878412444e-06,
+      "loss": 0.6698,
+      "step": 23248
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4978363357935474,
+      "learning_rate": 8.942119115927925e-06,
+      "loss": 0.7443,
+      "step": 23249
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4595881939246749,
+      "learning_rate": 8.941360359602724e-06,
+      "loss": 0.6886,
+      "step": 23250
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6071459831307708,
+      "learning_rate": 8.940601609441256e-06,
+      "loss": 0.4874,
+      "step": 23251
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.290852936506971,
+      "learning_rate": 8.93984286544794e-06,
+      "loss": 0.7011,
+      "step": 23252
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2228812694863405,
+      "learning_rate": 8.939084127627198e-06,
+      "loss": 0.6752,
+      "step": 23253
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3890319457288447,
+      "learning_rate": 8.938325395983445e-06,
+      "loss": 0.694,
+      "step": 23254
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.324385666013213,
+      "learning_rate": 8.937566670521092e-06,
+      "loss": 0.5964,
+      "step": 23255
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1634534547734614,
+      "learning_rate": 8.936807951244566e-06,
+      "loss": 0.6583,
+      "step": 23256
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.413960535334134,
+      "learning_rate": 8.936049238158283e-06,
+      "loss": 0.7498,
+      "step": 23257
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7031493413257102,
+      "learning_rate": 8.935290531266656e-06,
+      "loss": 0.7118,
+      "step": 23258
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3745880460847641,
+      "learning_rate": 8.934531830574101e-06,
+      "loss": 0.7227,
+      "step": 23259
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0042865873591498,
+      "learning_rate": 8.933773136085042e-06,
+      "loss": 0.8176,
+      "step": 23260
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2905815174279396,
+      "learning_rate": 8.933014447803896e-06,
+      "loss": 0.7487,
+      "step": 23261
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3641932893728688,
+      "learning_rate": 8.932255765735075e-06,
+      "loss": 0.6571,
+      "step": 23262
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.320426929428979,
+      "learning_rate": 8.931497089882997e-06,
+      "loss": 0.879,
+      "step": 23263
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5864991209512571,
+      "learning_rate": 8.930738420252085e-06,
+      "loss": 0.6484,
+      "step": 23264
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1654696394692245,
+      "learning_rate": 8.92997975684675e-06,
+      "loss": 0.6884,
+      "step": 23265
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6379931741284458,
+      "learning_rate": 8.929221099671414e-06,
+      "loss": 0.5407,
+      "step": 23266
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4645756056158377,
+      "learning_rate": 8.928462448730487e-06,
+      "loss": 0.728,
+      "step": 23267
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3520837585436316,
+      "learning_rate": 8.927703804028396e-06,
+      "loss": 0.8039,
+      "step": 23268
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5495622487195562,
+      "learning_rate": 8.926945165569551e-06,
+      "loss": 0.7506,
+      "step": 23269
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5145662059718856,
+      "learning_rate": 8.92618653335837e-06,
+      "loss": 0.7469,
+      "step": 23270
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3012448279879099,
+      "learning_rate": 8.925427907399276e-06,
+      "loss": 0.6922,
+      "step": 23271
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3535052918008574,
+      "learning_rate": 8.924669287696673e-06,
+      "loss": 0.5739,
+      "step": 23272
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.454712323707455,
+      "learning_rate": 8.92391067425499e-06,
+      "loss": 0.7685,
+      "step": 23273
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5620525554935827,
+      "learning_rate": 8.92315206707864e-06,
+      "loss": 0.6491,
+      "step": 23274
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2177953135883213,
+      "learning_rate": 8.922393466172042e-06,
+      "loss": 0.6062,
+      "step": 23275
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3710933889439676,
+      "learning_rate": 8.921634871539603e-06,
+      "loss": 0.7515,
+      "step": 23276
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3648660832878732,
+      "learning_rate": 8.920876283185753e-06,
+      "loss": 0.7566,
+      "step": 23277
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5084203347461953,
+      "learning_rate": 8.920117701114903e-06,
+      "loss": 0.7254,
+      "step": 23278
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5914624196912471,
+      "learning_rate": 8.919359125331468e-06,
+      "loss": 0.5587,
+      "step": 23279
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.429241170915429,
+      "learning_rate": 8.918600555839865e-06,
+      "loss": 0.6069,
+      "step": 23280
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2768720444319988,
+      "learning_rate": 8.917841992644513e-06,
+      "loss": 0.6358,
+      "step": 23281
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.304184261290343,
+      "learning_rate": 8.91708343574983e-06,
+      "loss": 0.6445,
+      "step": 23282
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5843804357211952,
+      "learning_rate": 8.916324885160228e-06,
+      "loss": 0.6348,
+      "step": 23283
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4245300932136262,
+      "learning_rate": 8.915566340880122e-06,
+      "loss": 0.6776,
+      "step": 23284
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9733972350300153,
+      "learning_rate": 8.914807802913939e-06,
+      "loss": 0.7411,
+      "step": 23285
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3399354475763554,
+      "learning_rate": 8.914049271266085e-06,
+      "loss": 0.6734,
+      "step": 23286
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1820556337570394,
+      "learning_rate": 8.913290745940979e-06,
+      "loss": 0.668,
+      "step": 23287
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7499993783587018,
+      "learning_rate": 8.912532226943038e-06,
+      "loss": 0.8749,
+      "step": 23288
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3610959711958268,
+      "learning_rate": 8.911773714276682e-06,
+      "loss": 0.7775,
+      "step": 23289
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1889493255121462,
+      "learning_rate": 8.911015207946324e-06,
+      "loss": 0.6107,
+      "step": 23290
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3975851377098947,
+      "learning_rate": 8.910256707956375e-06,
+      "loss": 0.6477,
+      "step": 23291
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6470414070763633,
+      "learning_rate": 8.90949821431126e-06,
+      "loss": 0.7912,
+      "step": 23292
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.716713748707629,
+      "learning_rate": 8.908739727015394e-06,
+      "loss": 0.7206,
+      "step": 23293
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5318920171990116,
+      "learning_rate": 8.907981246073188e-06,
+      "loss": 0.7872,
+      "step": 23294
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4549464044539735,
+      "learning_rate": 8.907222771489059e-06,
+      "loss": 0.7124,
+      "step": 23295
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.397414842564463,
+      "learning_rate": 8.90646430326743e-06,
+      "loss": 0.7854,
+      "step": 23296
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6810917907977025,
+      "learning_rate": 8.905705841412708e-06,
+      "loss": 0.7015,
+      "step": 23297
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2117881753026327,
+      "learning_rate": 8.904947385929314e-06,
+      "loss": 0.6761,
+      "step": 23298
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6712191672587182,
+      "learning_rate": 8.90418893682166e-06,
+      "loss": 0.7587,
+      "step": 23299
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3731605074328486,
+      "learning_rate": 8.903430494094171e-06,
+      "loss": 0.8145,
+      "step": 23300
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5974670118794931,
+      "learning_rate": 8.902672057751254e-06,
+      "loss": 0.6769,
+      "step": 23301
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.272175249254194,
+      "learning_rate": 8.901913627797328e-06,
+      "loss": 0.7245,
+      "step": 23302
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6523600948593211,
+      "learning_rate": 8.90115520423681e-06,
+      "loss": 0.7051,
+      "step": 23303
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4701719814865108,
+      "learning_rate": 8.90039678707411e-06,
+      "loss": 0.6435,
+      "step": 23304
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4490510753964292,
+      "learning_rate": 8.899638376313651e-06,
+      "loss": 0.6263,
+      "step": 23305
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7593479006278843,
+      "learning_rate": 8.898879971959844e-06,
+      "loss": 0.6884,
+      "step": 23306
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3467785131942651,
+      "learning_rate": 8.89812157401711e-06,
+      "loss": 0.7988,
+      "step": 23307
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6538950355951278,
+      "learning_rate": 8.897363182489856e-06,
+      "loss": 0.7084,
+      "step": 23308
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5132910091724927,
+      "learning_rate": 8.896604797382507e-06,
+      "loss": 0.719,
+      "step": 23309
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5394324784323137,
+      "learning_rate": 8.895846418699475e-06,
+      "loss": 0.7469,
+      "step": 23310
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.367096679115785,
+      "learning_rate": 8.89508804644517e-06,
+      "loss": 0.6663,
+      "step": 23311
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2933063952746462,
+      "learning_rate": 8.894329680624012e-06,
+      "loss": 0.8563,
+      "step": 23312
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5598171056683763,
+      "learning_rate": 8.89357132124042e-06,
+      "loss": 0.5368,
+      "step": 23313
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2445520800333347,
+      "learning_rate": 8.892812968298807e-06,
+      "loss": 0.8695,
+      "step": 23314
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6177059379785965,
+      "learning_rate": 8.892054621803586e-06,
+      "loss": 0.5754,
+      "step": 23315
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8302458983892342,
+      "learning_rate": 8.89129628175917e-06,
+      "loss": 0.7421,
+      "step": 23316
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4657903661713092,
+      "learning_rate": 8.890537948169984e-06,
+      "loss": 0.7512,
+      "step": 23317
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.216524423471647,
+      "learning_rate": 8.889779621040435e-06,
+      "loss": 0.6201,
+      "step": 23318
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5570915026058558,
+      "learning_rate": 8.88902130037494e-06,
+      "loss": 0.7484,
+      "step": 23319
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2891043031301044,
+      "learning_rate": 8.888262986177913e-06,
+      "loss": 0.7679,
+      "step": 23320
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4180565835084393,
+      "learning_rate": 8.887504678453775e-06,
+      "loss": 0.6902,
+      "step": 23321
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2849673541731341,
+      "learning_rate": 8.886746377206934e-06,
+      "loss": 0.6767,
+      "step": 23322
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7096048256576815,
+      "learning_rate": 8.88598808244181e-06,
+      "loss": 0.6705,
+      "step": 23323
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4509986075412502,
+      "learning_rate": 8.885229794162818e-06,
+      "loss": 0.7006,
+      "step": 23324
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.296962453784119,
+      "learning_rate": 8.884471512374363e-06,
+      "loss": 0.6366,
+      "step": 23325
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3877981806624584,
+      "learning_rate": 8.883713237080875e-06,
+      "loss": 0.7098,
+      "step": 23326
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4553549544508935,
+      "learning_rate": 8.882954968286756e-06,
+      "loss": 0.8558,
+      "step": 23327
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4944279809724437,
+      "learning_rate": 8.882196705996436e-06,
+      "loss": 0.6523,
+      "step": 23328
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3032641385131596,
+      "learning_rate": 8.881438450214313e-06,
+      "loss": 0.6235,
+      "step": 23329
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.467917578327067,
+      "learning_rate": 8.880680200944812e-06,
+      "loss": 0.6889,
+      "step": 23330
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3774757865237863,
+      "learning_rate": 8.879921958192347e-06,
+      "loss": 0.694,
+      "step": 23331
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3088907648862966,
+      "learning_rate": 8.87916372196133e-06,
+      "loss": 0.7039,
+      "step": 23332
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3226221662629105,
+      "learning_rate": 8.878405492256174e-06,
+      "loss": 0.7534,
+      "step": 23333
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5551502005600664,
+      "learning_rate": 8.877647269081298e-06,
+      "loss": 0.6889,
+      "step": 23334
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.22214037452415,
+      "learning_rate": 8.876889052441117e-06,
+      "loss": 0.7251,
+      "step": 23335
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4050576180964354,
+      "learning_rate": 8.876130842340039e-06,
+      "loss": 0.8434,
+      "step": 23336
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5630851652714752,
+      "learning_rate": 8.875372638782485e-06,
+      "loss": 0.8038,
+      "step": 23337
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3393254909505166,
+      "learning_rate": 8.874614441772871e-06,
+      "loss": 0.632,
+      "step": 23338
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4720959011747727,
+      "learning_rate": 8.873856251315605e-06,
+      "loss": 0.7322,
+      "step": 23339
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.57287672804806,
+      "learning_rate": 8.873098067415103e-06,
+      "loss": 0.6518,
+      "step": 23340
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4218537186491316,
+      "learning_rate": 8.872339890075782e-06,
+      "loss": 0.8073,
+      "step": 23341
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3609184486499477,
+      "learning_rate": 8.871581719302057e-06,
+      "loss": 0.7212,
+      "step": 23342
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3038520142620207,
+      "learning_rate": 8.870823555098339e-06,
+      "loss": 0.6997,
+      "step": 23343
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.506618499577174,
+      "learning_rate": 8.870065397469043e-06,
+      "loss": 0.7793,
+      "step": 23344
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8948215824482133,
+      "learning_rate": 8.869307246418589e-06,
+      "loss": 0.7551,
+      "step": 23345
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6475007821225622,
+      "learning_rate": 8.86854910195138e-06,
+      "loss": 0.6849,
+      "step": 23346
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3664308828232112,
+      "learning_rate": 8.867790964071838e-06,
+      "loss": 0.597,
+      "step": 23347
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6129041437684206,
+      "learning_rate": 8.867032832784372e-06,
+      "loss": 0.5109,
+      "step": 23348
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2819342648671577,
+      "learning_rate": 8.866274708093407e-06,
+      "loss": 0.7763,
+      "step": 23349
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4779689707959507,
+      "learning_rate": 8.865516590003344e-06,
+      "loss": 0.8234,
+      "step": 23350
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3195906996630347,
+      "learning_rate": 8.864758478518606e-06,
+      "loss": 0.6014,
+      "step": 23351
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7302741595197282,
+      "learning_rate": 8.864000373643604e-06,
+      "loss": 0.7873,
+      "step": 23352
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6029643526404684,
+      "learning_rate": 8.863242275382747e-06,
+      "loss": 0.7382,
+      "step": 23353
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5937816841455954,
+      "learning_rate": 8.862484183740457e-06,
+      "loss": 0.5442,
+      "step": 23354
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6729771547061238,
+      "learning_rate": 8.861726098721142e-06,
+      "loss": 0.698,
+      "step": 23355
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6306301503128722,
+      "learning_rate": 8.860968020329222e-06,
+      "loss": 0.5134,
+      "step": 23356
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1637316917947327,
+      "learning_rate": 8.8602099485691e-06,
+      "loss": 0.6691,
+      "step": 23357
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.288988293930851,
+      "learning_rate": 8.859451883445201e-06,
+      "loss": 0.7685,
+      "step": 23358
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4549893361322659,
+      "learning_rate": 8.858693824961936e-06,
+      "loss": 0.676,
+      "step": 23359
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2083796905680837,
+      "learning_rate": 8.857935773123715e-06,
+      "loss": 0.6441,
+      "step": 23360
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.2497996654173797,
+      "learning_rate": 8.85717772793495e-06,
+      "loss": 0.6872,
+      "step": 23361
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3866776872695374,
+      "learning_rate": 8.856419689400062e-06,
+      "loss": 0.78,
+      "step": 23362
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3463608811734442,
+      "learning_rate": 8.855661657523461e-06,
+      "loss": 0.7002,
+      "step": 23363
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3694528081092545,
+      "learning_rate": 8.85490363230956e-06,
+      "loss": 0.6705,
+      "step": 23364
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5799692759306612,
+      "learning_rate": 8.854145613762767e-06,
+      "loss": 0.6737,
+      "step": 23365
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3703795551742397,
+      "learning_rate": 8.85338760188751e-06,
+      "loss": 0.7711,
+      "step": 23366
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5738626181688659,
+      "learning_rate": 8.852629596688189e-06,
+      "loss": 0.4843,
+      "step": 23367
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2538221491805415,
+      "learning_rate": 8.85187159816922e-06,
+      "loss": 0.7532,
+      "step": 23368
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3259710334956887,
+      "learning_rate": 8.851113606335018e-06,
+      "loss": 0.8328,
+      "step": 23369
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3274483807565147,
+      "learning_rate": 8.85035562119e-06,
+      "loss": 0.7477,
+      "step": 23370
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6273941724528214,
+      "learning_rate": 8.849597642738573e-06,
+      "loss": 0.535,
+      "step": 23371
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3946472192668398,
+      "learning_rate": 8.848839670985151e-06,
+      "loss": 0.7635,
+      "step": 23372
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6549220809563905,
+      "learning_rate": 8.848081705934155e-06,
+      "loss": 0.713,
+      "step": 23373
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.612229976623959,
+      "learning_rate": 8.847323747589987e-06,
+      "loss": 0.5209,
+      "step": 23374
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5164555996671834,
+      "learning_rate": 8.846565795957066e-06,
+      "loss": 0.7506,
+      "step": 23375
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.856761635619312,
+      "learning_rate": 8.845807851039801e-06,
+      "loss": 0.6318,
+      "step": 23376
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4286853762186933,
+      "learning_rate": 8.845049912842616e-06,
+      "loss": 0.6028,
+      "step": 23377
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5595513537782724,
+      "learning_rate": 8.844291981369911e-06,
+      "loss": 0.7059,
+      "step": 23378
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5182960290539829,
+      "learning_rate": 8.843534056626103e-06,
+      "loss": 0.6974,
+      "step": 23379
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4666434292705537,
+      "learning_rate": 8.842776138615609e-06,
+      "loss": 0.5928,
+      "step": 23380
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.450959946040377,
+      "learning_rate": 8.842018227342839e-06,
+      "loss": 0.6484,
+      "step": 23381
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.305854499338979,
+      "learning_rate": 8.8412603228122e-06,
+      "loss": 0.6701,
+      "step": 23382
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3148970328865757,
+      "learning_rate": 8.840502425028114e-06,
+      "loss": 0.683,
+      "step": 23383
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.424848947622364,
+      "learning_rate": 8.839744533994991e-06,
+      "loss": 0.7153,
+      "step": 23384
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4732983424989987,
+      "learning_rate": 8.83898664971724e-06,
+      "loss": 0.8338,
+      "step": 23385
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2204277616342711,
+      "learning_rate": 8.838228772199278e-06,
+      "loss": 0.6356,
+      "step": 23386
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4206281210886742,
+      "learning_rate": 8.837470901445518e-06,
+      "loss": 0.6825,
+      "step": 23387
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.429234689854588,
+      "learning_rate": 8.836713037460368e-06,
+      "loss": 0.6799,
+      "step": 23388
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.514361862856576,
+      "learning_rate": 8.83595518024824e-06,
+      "loss": 0.7482,
+      "step": 23389
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3654421657073388,
+      "learning_rate": 8.835197329813552e-06,
+      "loss": 0.6808,
+      "step": 23390
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.28664361959012,
+      "learning_rate": 8.834439486160716e-06,
+      "loss": 0.7419,
+      "step": 23391
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5872163509832788,
+      "learning_rate": 8.833681649294142e-06,
+      "loss": 0.5441,
+      "step": 23392
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5481470394037504,
+      "learning_rate": 8.832923819218238e-06,
+      "loss": 0.665,
+      "step": 23393
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.999795779571238,
+      "learning_rate": 8.83216599593743e-06,
+      "loss": 0.6886,
+      "step": 23394
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6134143348435616,
+      "learning_rate": 8.831408179456112e-06,
+      "loss": 0.5232,
+      "step": 23395
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6999307122157348,
+      "learning_rate": 8.830650369778711e-06,
+      "loss": 0.7592,
+      "step": 23396
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3415974595881754,
+      "learning_rate": 8.829892566909629e-06,
+      "loss": 0.765,
+      "step": 23397
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5718657967163676,
+      "learning_rate": 8.829134770853289e-06,
+      "loss": 0.4686,
+      "step": 23398
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.615730127402054,
+      "learning_rate": 8.828376981614093e-06,
+      "loss": 0.7908,
+      "step": 23399
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4455135903103178,
+      "learning_rate": 8.827619199196457e-06,
+      "loss": 0.7829,
+      "step": 23400
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3029671270563599,
+      "learning_rate": 8.826861423604797e-06,
+      "loss": 0.6407,
+      "step": 23401
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4844272928022437,
+      "learning_rate": 8.826103654843516e-06,
+      "loss": 0.6876,
+      "step": 23402
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5603491904538027,
+      "learning_rate": 8.825345892917035e-06,
+      "loss": 0.6361,
+      "step": 23403
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.625867221503607,
+      "learning_rate": 8.824588137829761e-06,
+      "loss": 0.5652,
+      "step": 23404
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3974132323623785,
+      "learning_rate": 8.823830389586108e-06,
+      "loss": 0.6508,
+      "step": 23405
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5164811789744992,
+      "learning_rate": 8.823072648190484e-06,
+      "loss": 0.8187,
+      "step": 23406
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.386750271180363,
+      "learning_rate": 8.822314913647305e-06,
+      "loss": 0.7865,
+      "step": 23407
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3503393609707417,
+      "learning_rate": 8.821557185960984e-06,
+      "loss": 0.6717,
+      "step": 23408
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.387369471974605,
+      "learning_rate": 8.820799465135928e-06,
+      "loss": 0.6112,
+      "step": 23409
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3390994225034056,
+      "learning_rate": 8.820041751176549e-06,
+      "loss": 0.7274,
+      "step": 23410
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.280748961565471,
+      "learning_rate": 8.81928404408726e-06,
+      "loss": 0.6875,
+      "step": 23411
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5654546970726613,
+      "learning_rate": 8.81852634387248e-06,
+      "loss": 0.6808,
+      "step": 23412
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3518325201391979,
+      "learning_rate": 8.817768650536607e-06,
+      "loss": 0.7234,
+      "step": 23413
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4398888129422267,
+      "learning_rate": 8.817010964084059e-06,
+      "loss": 0.7489,
+      "step": 23414
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9327239812544645,
+      "learning_rate": 8.81625328451925e-06,
+      "loss": 0.7075,
+      "step": 23415
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4422028309176669,
+      "learning_rate": 8.815495611846592e-06,
+      "loss": 0.6019,
+      "step": 23416
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6331166787917928,
+      "learning_rate": 8.814737946070489e-06,
+      "loss": 0.5361,
+      "step": 23417
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.311503937271648,
+      "learning_rate": 8.813980287195358e-06,
+      "loss": 0.6835,
+      "step": 23418
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7707411888205096,
+      "learning_rate": 8.81322263522561e-06,
+      "loss": 0.7446,
+      "step": 23419
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2686442920036414,
+      "learning_rate": 8.812464990165656e-06,
+      "loss": 0.7302,
+      "step": 23420
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6005640602093167,
+      "learning_rate": 8.811707352019905e-06,
+      "loss": 0.7107,
+      "step": 23421
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5399927335638455,
+      "learning_rate": 8.810949720792771e-06,
+      "loss": 0.7594,
+      "step": 23422
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4696367091474378,
+      "learning_rate": 8.810192096488666e-06,
+      "loss": 0.805,
+      "step": 23423
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.414531853089237,
+      "learning_rate": 8.809434479111998e-06,
+      "loss": 0.734,
+      "step": 23424
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7676664673643527,
+      "learning_rate": 8.808676868667177e-06,
+      "loss": 0.7086,
+      "step": 23425
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3829373612107596,
+      "learning_rate": 8.807919265158622e-06,
+      "loss": 0.8043,
+      "step": 23426
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4640148279470142,
+      "learning_rate": 8.807161668590733e-06,
+      "loss": 0.6553,
+      "step": 23427
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5476520649783976,
+      "learning_rate": 8.806404078967928e-06,
+      "loss": 0.8155,
+      "step": 23428
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.244627060410402,
+      "learning_rate": 8.805646496294613e-06,
+      "loss": 0.5971,
+      "step": 23429
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4210892080598279,
+      "learning_rate": 8.80488892057521e-06,
+      "loss": 0.6836,
+      "step": 23430
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2600862946831544,
+      "learning_rate": 8.804131351814117e-06,
+      "loss": 0.6706,
+      "step": 23431
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3250895643316194,
+      "learning_rate": 8.80337379001575e-06,
+      "loss": 0.7361,
+      "step": 23432
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7372266505215634,
+      "learning_rate": 8.802616235184524e-06,
+      "loss": 0.7102,
+      "step": 23433
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.385993520235653,
+      "learning_rate": 8.801858687324839e-06,
+      "loss": 0.7195,
+      "step": 23434
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.652494471031016,
+      "learning_rate": 8.801101146441114e-06,
+      "loss": 0.8097,
+      "step": 23435
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5257533773286445,
+      "learning_rate": 8.800343612537758e-06,
+      "loss": 0.6864,
+      "step": 23436
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5236243350822254,
+      "learning_rate": 8.799586085619184e-06,
+      "loss": 0.6479,
+      "step": 23437
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3581519239480528,
+      "learning_rate": 8.798828565689796e-06,
+      "loss": 0.8113,
+      "step": 23438
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4347822906986145,
+      "learning_rate": 8.79807105275401e-06,
+      "loss": 0.6199,
+      "step": 23439
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3376493342877145,
+      "learning_rate": 8.797313546816238e-06,
+      "loss": 0.7947,
+      "step": 23440
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3483512211619582,
+      "learning_rate": 8.796556047880883e-06,
+      "loss": 0.8031,
+      "step": 23441
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5064127391650137,
+      "learning_rate": 8.79579855595236e-06,
+      "loss": 0.6599,
+      "step": 23442
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6895215743281762,
+      "learning_rate": 8.795041071035078e-06,
+      "loss": 0.7302,
+      "step": 23443
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3756507355205245,
+      "learning_rate": 8.794283593133454e-06,
+      "loss": 0.5957,
+      "step": 23444
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5917730407151539,
+      "learning_rate": 8.793526122251888e-06,
+      "loss": 0.5158,
+      "step": 23445
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4475057600546284,
+      "learning_rate": 8.792768658394794e-06,
+      "loss": 0.6801,
+      "step": 23446
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3418281217638726,
+      "learning_rate": 8.792011201566587e-06,
+      "loss": 0.7077,
+      "step": 23447
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4797427865755832,
+      "learning_rate": 8.791253751771671e-06,
+      "loss": 0.967,
+      "step": 23448
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4106526088425564,
+      "learning_rate": 8.790496309014457e-06,
+      "loss": 0.7446,
+      "step": 23449
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5869179921193253,
+      "learning_rate": 8.789738873299356e-06,
+      "loss": 0.8283,
+      "step": 23450
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.046111414727263,
+      "learning_rate": 8.788981444630784e-06,
+      "loss": 0.7718,
+      "step": 23451
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5278537821139606,
+      "learning_rate": 8.78822402301314e-06,
+      "loss": 0.7564,
+      "step": 23452
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.264926763299226,
+      "learning_rate": 8.787466608450843e-06,
+      "loss": 0.6079,
+      "step": 23453
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.61651649784055,
+      "learning_rate": 8.786709200948299e-06,
+      "loss": 0.4959,
+      "step": 23454
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5791243771732616,
+      "learning_rate": 8.785951800509915e-06,
+      "loss": 0.5129,
+      "step": 23455
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4057150172796014,
+      "learning_rate": 8.785194407140105e-06,
+      "loss": 0.6868,
+      "step": 23456
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6022369052117345,
+      "learning_rate": 8.784437020843278e-06,
+      "loss": 0.6725,
+      "step": 23457
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2797703359616115,
+      "learning_rate": 8.783679641623845e-06,
+      "loss": 0.7038,
+      "step": 23458
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.439755696135006,
+      "learning_rate": 8.782922269486211e-06,
+      "loss": 0.7101,
+      "step": 23459
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3368526529580604,
+      "learning_rate": 8.782164904434791e-06,
+      "loss": 0.6621,
+      "step": 23460
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.228929007380731,
+      "learning_rate": 8.781407546473995e-06,
+      "loss": 0.6564,
+      "step": 23461
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4402490142078783,
+      "learning_rate": 8.780650195608228e-06,
+      "loss": 0.6838,
+      "step": 23462
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3257196207775692,
+      "learning_rate": 8.779892851841897e-06,
+      "loss": 0.7631,
+      "step": 23463
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6137673993963197,
+      "learning_rate": 8.77913551517942e-06,
+      "loss": 0.5456,
+      "step": 23464
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6523666799633765,
+      "learning_rate": 8.778378185625206e-06,
+      "loss": 0.5658,
+      "step": 23465
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4948513207024157,
+      "learning_rate": 8.777620863183658e-06,
+      "loss": 0.6254,
+      "step": 23466
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.344014679568182,
+      "learning_rate": 8.776863547859185e-06,
+      "loss": 0.7469,
+      "step": 23467
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3208825301372404,
+      "learning_rate": 8.776106239656206e-06,
+      "loss": 0.6567,
+      "step": 23468
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5501943940285797,
+      "learning_rate": 8.77534893857912e-06,
+      "loss": 0.51,
+      "step": 23469
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.603960775882922,
+      "learning_rate": 8.774591644632338e-06,
+      "loss": 0.4787,
+      "step": 23470
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.257903171882718,
+      "learning_rate": 8.773834357820275e-06,
+      "loss": 0.7689,
+      "step": 23471
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5884204279155191,
+      "learning_rate": 8.773077078147337e-06,
+      "loss": 0.5303,
+      "step": 23472
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.661002780123285,
+      "learning_rate": 8.772319805617931e-06,
+      "loss": 0.7458,
+      "step": 23473
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5726383311694137,
+      "learning_rate": 8.771562540236467e-06,
+      "loss": 0.6756,
+      "step": 23474
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.350237814104251,
+      "learning_rate": 8.77080528200736e-06,
+      "loss": 0.7331,
+      "step": 23475
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8132765139531304,
+      "learning_rate": 8.770048030935007e-06,
+      "loss": 0.6884,
+      "step": 23476
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9659757346759044,
+      "learning_rate": 8.769290787023828e-06,
+      "loss": 0.7281,
+      "step": 23477
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7331782945430336,
+      "learning_rate": 8.768533550278223e-06,
+      "loss": 0.7775,
+      "step": 23478
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3884051221150198,
+      "learning_rate": 8.767776320702614e-06,
+      "loss": 0.6444,
+      "step": 23479
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.488746285042012,
+      "learning_rate": 8.767019098301394e-06,
+      "loss": 0.6769,
+      "step": 23480
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.525356863384088,
+      "learning_rate": 8.766261883078981e-06,
+      "loss": 0.7262,
+      "step": 23481
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3758110281176468,
+      "learning_rate": 8.765504675039784e-06,
+      "loss": 0.6548,
+      "step": 23482
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7638388317704374,
+      "learning_rate": 8.764747474188205e-06,
+      "loss": 0.8227,
+      "step": 23483
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3631536221341818,
+      "learning_rate": 8.763990280528661e-06,
+      "loss": 0.7855,
+      "step": 23484
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4109362857351229,
+      "learning_rate": 8.763233094065556e-06,
+      "loss": 0.6049,
+      "step": 23485
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.559447141719736,
+      "learning_rate": 8.7624759148033e-06,
+      "loss": 0.6534,
+      "step": 23486
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4728753559970906,
+      "learning_rate": 8.7617187427463e-06,
+      "loss": 0.5972,
+      "step": 23487
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.258211994813461,
+      "learning_rate": 8.760961577898965e-06,
+      "loss": 0.6939,
+      "step": 23488
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4072821441709675,
+      "learning_rate": 8.760204420265707e-06,
+      "loss": 0.812,
+      "step": 23489
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0972335274569494,
+      "learning_rate": 8.75944726985093e-06,
+      "loss": 0.7562,
+      "step": 23490
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6424371088968313,
+      "learning_rate": 8.758690126659041e-06,
+      "loss": 0.5549,
+      "step": 23491
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5920742431513917,
+      "learning_rate": 8.757932990694454e-06,
+      "loss": 0.5331,
+      "step": 23492
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5421911409053446,
+      "learning_rate": 8.757175861961576e-06,
+      "loss": 0.7764,
+      "step": 23493
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.382696969796548,
+      "learning_rate": 8.75641874046481e-06,
+      "loss": 0.7665,
+      "step": 23494
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5343004018714375,
+      "learning_rate": 8.755661626208568e-06,
+      "loss": 0.6588,
+      "step": 23495
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4153998925697722,
+      "learning_rate": 8.75490451919726e-06,
+      "loss": 0.7151,
+      "step": 23496
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3711624460921152,
+      "learning_rate": 8.754147419435291e-06,
+      "loss": 0.6414,
+      "step": 23497
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2299439344111545,
+      "learning_rate": 8.75339032692707e-06,
+      "loss": 0.7196,
+      "step": 23498
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4952697844802816,
+      "learning_rate": 8.752633241677003e-06,
+      "loss": 0.676,
+      "step": 23499
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.626512998078805,
+      "learning_rate": 8.751876163689506e-06,
+      "loss": 0.7601,
+      "step": 23500
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6626423930513705,
+      "learning_rate": 8.751119092968978e-06,
+      "loss": 0.7823,
+      "step": 23501
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4258978897735413,
+      "learning_rate": 8.75036202951983e-06,
+      "loss": 0.6629,
+      "step": 23502
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7127528373493144,
+      "learning_rate": 8.74960497334647e-06,
+      "loss": 0.6794,
+      "step": 23503
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5198540739101274,
+      "learning_rate": 8.748847924453304e-06,
+      "loss": 0.6522,
+      "step": 23504
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4253475926954629,
+      "learning_rate": 8.748090882844742e-06,
+      "loss": 0.7547,
+      "step": 23505
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6015076105804441,
+      "learning_rate": 8.747333848525193e-06,
+      "loss": 0.5047,
+      "step": 23506
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.85657896064063,
+      "learning_rate": 8.746576821499062e-06,
+      "loss": 0.7363,
+      "step": 23507
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3652054263308757,
+      "learning_rate": 8.745819801770756e-06,
+      "loss": 0.8756,
+      "step": 23508
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4687792621015827,
+      "learning_rate": 8.745062789344686e-06,
+      "loss": 0.596,
+      "step": 23509
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4923095794591925,
+      "learning_rate": 8.74430578422526e-06,
+      "loss": 0.6907,
+      "step": 23510
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5331235533159826,
+      "learning_rate": 8.74354878641688e-06,
+      "loss": 0.7027,
+      "step": 23511
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3910234051912713,
+      "learning_rate": 8.742791795923956e-06,
+      "loss": 0.8134,
+      "step": 23512
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4015493388580342,
+      "learning_rate": 8.742034812750896e-06,
+      "loss": 0.7671,
+      "step": 23513
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3941188787621484,
+      "learning_rate": 8.741277836902114e-06,
+      "loss": 0.6151,
+      "step": 23514
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6286055570114404,
+      "learning_rate": 8.740520868382005e-06,
+      "loss": 0.5266,
+      "step": 23515
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.319119528408429,
+      "learning_rate": 8.739763907194981e-06,
+      "loss": 0.6381,
+      "step": 23516
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.457811130714406,
+      "learning_rate": 8.739006953345455e-06,
+      "loss": 0.6817,
+      "step": 23517
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2097489921330316,
+      "learning_rate": 8.738250006837828e-06,
+      "loss": 0.6222,
+      "step": 23518
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4512753852329165,
+      "learning_rate": 8.737493067676508e-06,
+      "loss": 0.6885,
+      "step": 23519
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.655477625297395,
+      "learning_rate": 8.736736135865901e-06,
+      "loss": 0.7105,
+      "step": 23520
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4906308995821842,
+      "learning_rate": 8.735979211410423e-06,
+      "loss": 0.6746,
+      "step": 23521
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3055482829674905,
+      "learning_rate": 8.73522229431447e-06,
+      "loss": 0.6946,
+      "step": 23522
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4500085462070444,
+      "learning_rate": 8.73446538458245e-06,
+      "loss": 0.6849,
+      "step": 23523
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5914307538923709,
+      "learning_rate": 8.733708482218782e-06,
+      "loss": 0.7113,
+      "step": 23524
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.326885376290209,
+      "learning_rate": 8.732951587227856e-06,
+      "loss": 0.829,
+      "step": 23525
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4335609025921041,
+      "learning_rate": 8.732194699614089e-06,
+      "loss": 0.6864,
+      "step": 23526
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3704421565419254,
+      "learning_rate": 8.731437819381884e-06,
+      "loss": 0.6841,
+      "step": 23527
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.047640031155172,
+      "learning_rate": 8.730680946535657e-06,
+      "loss": 0.6736,
+      "step": 23528
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3932094185933999,
+      "learning_rate": 8.7299240810798e-06,
+      "loss": 0.6802,
+      "step": 23529
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3853873591489445,
+      "learning_rate": 8.72916722301873e-06,
+      "loss": 0.8242,
+      "step": 23530
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5249947780951556,
+      "learning_rate": 8.72841037235685e-06,
+      "loss": 0.7932,
+      "step": 23531
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3406034155341973,
+      "learning_rate": 8.727653529098567e-06,
+      "loss": 0.5838,
+      "step": 23532
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.473581102662497,
+      "learning_rate": 8.726896693248288e-06,
+      "loss": 0.738,
+      "step": 23533
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7274799449599207,
+      "learning_rate": 8.726139864810422e-06,
+      "loss": 0.7138,
+      "step": 23534
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.577857752858987,
+      "learning_rate": 8.725383043789372e-06,
+      "loss": 0.5051,
+      "step": 23535
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.720637315477805,
+      "learning_rate": 8.724626230189544e-06,
+      "loss": 0.7436,
+      "step": 23536
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.89533150781055,
+      "learning_rate": 8.723869424015345e-06,
+      "loss": 0.6276,
+      "step": 23537
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5173465379244,
+      "learning_rate": 8.723112625271184e-06,
+      "loss": 0.7087,
+      "step": 23538
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4936939269390639,
+      "learning_rate": 8.722355833961467e-06,
+      "loss": 0.8907,
+      "step": 23539
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6881381176939347,
+      "learning_rate": 8.721599050090596e-06,
+      "loss": 0.7433,
+      "step": 23540
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2517449580164637,
+      "learning_rate": 8.72084227366298e-06,
+      "loss": 0.7698,
+      "step": 23541
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3053297947573106,
+      "learning_rate": 8.720085504683028e-06,
+      "loss": 0.7217,
+      "step": 23542
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4658808089137483,
+      "learning_rate": 8.719328743155144e-06,
+      "loss": 0.7551,
+      "step": 23543
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5117491885304062,
+      "learning_rate": 8.718571989083729e-06,
+      "loss": 0.75,
+      "step": 23544
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6111649642307349,
+      "learning_rate": 8.717815242473196e-06,
+      "loss": 0.5405,
+      "step": 23545
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.38624869897002,
+      "learning_rate": 8.71705850332795e-06,
+      "loss": 0.7363,
+      "step": 23546
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8687021052407369,
+      "learning_rate": 8.716301771652393e-06,
+      "loss": 0.6707,
+      "step": 23547
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5890259329218683,
+      "learning_rate": 8.715545047450933e-06,
+      "loss": 0.4985,
+      "step": 23548
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.365820838565652,
+      "learning_rate": 8.71478833072798e-06,
+      "loss": 0.478,
+      "step": 23549
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5999388376489692,
+      "learning_rate": 8.714031621487934e-06,
+      "loss": 0.8101,
+      "step": 23550
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6309063451023449,
+      "learning_rate": 8.713274919735204e-06,
+      "loss": 0.5214,
+      "step": 23551
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.663388406506874,
+      "learning_rate": 8.712518225474191e-06,
+      "loss": 0.7276,
+      "step": 23552
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4124291092302121,
+      "learning_rate": 8.71176153870931e-06,
+      "loss": 0.7545,
+      "step": 23553
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.482321549316531,
+      "learning_rate": 8.711004859444957e-06,
+      "loss": 0.7686,
+      "step": 23554
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4312317769770095,
+      "learning_rate": 8.710248187685545e-06,
+      "loss": 0.7569,
+      "step": 23555
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.307389156336245,
+      "learning_rate": 8.709491523435477e-06,
+      "loss": 0.5439,
+      "step": 23556
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5402598188237773,
+      "learning_rate": 8.708734866699153e-06,
+      "loss": 0.7077,
+      "step": 23557
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3303327748810612,
+      "learning_rate": 8.707978217480986e-06,
+      "loss": 0.6739,
+      "step": 23558
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6838801918998112,
+      "learning_rate": 8.707221575785379e-06,
+      "loss": 0.7293,
+      "step": 23559
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4783716647823988,
+      "learning_rate": 8.706464941616741e-06,
+      "loss": 0.675,
+      "step": 23560
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4443104007240997,
+      "learning_rate": 8.705708314979468e-06,
+      "loss": 0.7104,
+      "step": 23561
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4158961592215435,
+      "learning_rate": 8.704951695877973e-06,
+      "loss": 0.6587,
+      "step": 23562
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6528323779666046,
+      "learning_rate": 8.70419508431666e-06,
+      "loss": 0.5444,
+      "step": 23563
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6715953982483382,
+      "learning_rate": 8.703438480299932e-06,
+      "loss": 0.7986,
+      "step": 23564
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5159472645864103,
+      "learning_rate": 8.702681883832196e-06,
+      "loss": 0.7637,
+      "step": 23565
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5932704783727981,
+      "learning_rate": 8.701925294917857e-06,
+      "loss": 0.4828,
+      "step": 23566
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2951807312035029,
+      "learning_rate": 8.70116871356132e-06,
+      "loss": 0.6395,
+      "step": 23567
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6680262107668449,
+      "learning_rate": 8.700412139766989e-06,
+      "loss": 0.6751,
+      "step": 23568
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.191888664878389,
+      "learning_rate": 8.69965557353927e-06,
+      "loss": 0.7072,
+      "step": 23569
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2737437539139724,
+      "learning_rate": 8.698899014882572e-06,
+      "loss": 0.5853,
+      "step": 23570
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.446562524460132,
+      "learning_rate": 8.698142463801292e-06,
+      "loss": 0.8203,
+      "step": 23571
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8222666132837122,
+      "learning_rate": 8.697385920299838e-06,
+      "loss": 0.7317,
+      "step": 23572
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3770713525643579,
+      "learning_rate": 8.696629384382616e-06,
+      "loss": 0.6002,
+      "step": 23573
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0994305107673172,
+      "learning_rate": 8.695872856054034e-06,
+      "loss": 0.7877,
+      "step": 23574
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.50454346098921,
+      "learning_rate": 8.695116335318492e-06,
+      "loss": 0.6173,
+      "step": 23575
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.025856652638905,
+      "learning_rate": 8.694359822180391e-06,
+      "loss": 0.7026,
+      "step": 23576
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4280999891878383,
+      "learning_rate": 8.693603316644147e-06,
+      "loss": 0.6591,
+      "step": 23577
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3692070071816973,
+      "learning_rate": 8.692846818714154e-06,
+      "loss": 0.8525,
+      "step": 23578
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6187586784008188,
+      "learning_rate": 8.692090328394821e-06,
+      "loss": 0.5879,
+      "step": 23579
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3320604906296922,
+      "learning_rate": 8.69133384569055e-06,
+      "loss": 0.6544,
+      "step": 23580
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.600726568717636,
+      "learning_rate": 8.690577370605752e-06,
+      "loss": 0.7583,
+      "step": 23581
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.260710310437177,
+      "learning_rate": 8.689820903144827e-06,
+      "loss": 0.805,
+      "step": 23582
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3995775525667922,
+      "learning_rate": 8.689064443312177e-06,
+      "loss": 0.6557,
+      "step": 23583
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3890375288626038,
+      "learning_rate": 8.688307991112212e-06,
+      "loss": 0.7968,
+      "step": 23584
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.052139881026927,
+      "learning_rate": 8.68755154654933e-06,
+      "loss": 0.6625,
+      "step": 23585
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3765484568431845,
+      "learning_rate": 8.68679510962794e-06,
+      "loss": 0.8618,
+      "step": 23586
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4472873274635756,
+      "learning_rate": 8.686038680352442e-06,
+      "loss": 0.7052,
+      "step": 23587
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2193594321385788,
+      "learning_rate": 8.685282258727248e-06,
+      "loss": 0.6844,
+      "step": 23588
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5026054487739515,
+      "learning_rate": 8.684525844756751e-06,
+      "loss": 0.7239,
+      "step": 23589
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.7015582168323418,
+      "learning_rate": 8.683769438445365e-06,
+      "loss": 0.6983,
+      "step": 23590
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4203964010131906,
+      "learning_rate": 8.68301303979749e-06,
+      "loss": 0.8409,
+      "step": 23591
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3516275752065972,
+      "learning_rate": 8.682256648817528e-06,
+      "loss": 0.6627,
+      "step": 23592
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5960104254122331,
+      "learning_rate": 8.681500265509884e-06,
+      "loss": 0.7519,
+      "step": 23593
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8910129371770532,
+      "learning_rate": 8.680743889878964e-06,
+      "loss": 0.7988,
+      "step": 23594
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3904038685555955,
+      "learning_rate": 8.679987521929174e-06,
+      "loss": 0.7274,
+      "step": 23595
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4628632249398048,
+      "learning_rate": 8.679231161664911e-06,
+      "loss": 0.6402,
+      "step": 23596
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4217988902822127,
+      "learning_rate": 8.678474809090583e-06,
+      "loss": 0.7293,
+      "step": 23597
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3081866392531705,
+      "learning_rate": 8.677718464210594e-06,
+      "loss": 0.6421,
+      "step": 23598
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.231730693319538,
+      "learning_rate": 8.676962127029348e-06,
+      "loss": 0.7183,
+      "step": 23599
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5987827559690222,
+      "learning_rate": 8.676205797551243e-06,
+      "loss": 0.521,
+      "step": 23600
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5904910898980305,
+      "learning_rate": 8.675449475780689e-06,
+      "loss": 0.5051,
+      "step": 23601
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7781304364151942,
+      "learning_rate": 8.674693161722089e-06,
+      "loss": 0.7119,
+      "step": 23602
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6206622656492653,
+      "learning_rate": 8.673936855379844e-06,
+      "loss": 0.5114,
+      "step": 23603
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4706681152450558,
+      "learning_rate": 8.673180556758358e-06,
+      "loss": 0.6686,
+      "step": 23604
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8084267226864705,
+      "learning_rate": 8.672424265862037e-06,
+      "loss": 0.7383,
+      "step": 23605
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4217493349604517,
+      "learning_rate": 8.671667982695279e-06,
+      "loss": 0.692,
+      "step": 23606
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4760061193752307,
+      "learning_rate": 8.670911707262491e-06,
+      "loss": 0.6881,
+      "step": 23607
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4355020485459253,
+      "learning_rate": 8.670155439568076e-06,
+      "loss": 0.7424,
+      "step": 23608
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5707579587496052,
+      "learning_rate": 8.669399179616441e-06,
+      "loss": 0.745,
+      "step": 23609
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5164494598909766,
+      "learning_rate": 8.66864292741198e-06,
+      "loss": 0.6984,
+      "step": 23610
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.569272077334662,
+      "learning_rate": 8.667886682959103e-06,
+      "loss": 0.7693,
+      "step": 23611
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3287060328943214,
+      "learning_rate": 8.667130446262214e-06,
+      "loss": 0.7152,
+      "step": 23612
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5486843780361295,
+      "learning_rate": 8.666374217325713e-06,
+      "loss": 0.7106,
+      "step": 23613
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3142216533182765,
+      "learning_rate": 8.665617996154e-06,
+      "loss": 0.6075,
+      "step": 23614
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4345549277242808,
+      "learning_rate": 8.664861782751482e-06,
+      "loss": 0.8095,
+      "step": 23615
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3691378244291115,
+      "learning_rate": 8.664105577122566e-06,
+      "loss": 0.7896,
+      "step": 23616
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1635972364791685,
+      "learning_rate": 8.663349379271645e-06,
+      "loss": 0.7187,
+      "step": 23617
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2462343460737428,
+      "learning_rate": 8.662593189203128e-06,
+      "loss": 0.6229,
+      "step": 23618
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3549831015114508,
+      "learning_rate": 8.661837006921422e-06,
+      "loss": 0.7356,
+      "step": 23619
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5552986451681323,
+      "learning_rate": 8.661080832430919e-06,
+      "loss": 0.6682,
+      "step": 23620
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4627358265514836,
+      "learning_rate": 8.660324665736026e-06,
+      "loss": 0.8246,
+      "step": 23621
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.518241866864074,
+      "learning_rate": 8.659568506841151e-06,
+      "loss": 0.8254,
+      "step": 23622
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.55922900958136,
+      "learning_rate": 8.658812355750692e-06,
+      "loss": 0.7066,
+      "step": 23623
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2265976045928157,
+      "learning_rate": 8.658056212469052e-06,
+      "loss": 0.6002,
+      "step": 23624
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.359113013576872,
+      "learning_rate": 8.657300077000632e-06,
+      "loss": 0.7313,
+      "step": 23625
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.512047652973963,
+      "learning_rate": 8.65654394934984e-06,
+      "loss": 0.7683,
+      "step": 23626
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3926862159913964,
+      "learning_rate": 8.65578782952107e-06,
+      "loss": 0.8724,
+      "step": 23627
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.857763604265939,
+      "learning_rate": 8.655031717518732e-06,
+      "loss": 0.7252,
+      "step": 23628
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6176443474813358,
+      "learning_rate": 8.65427561334722e-06,
+      "loss": 0.5233,
+      "step": 23629
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5181320790253852,
+      "learning_rate": 8.653519517010947e-06,
+      "loss": 0.7246,
+      "step": 23630
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3939707027275012,
+      "learning_rate": 8.652763428514309e-06,
+      "loss": 0.7723,
+      "step": 23631
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8331212104253225,
+      "learning_rate": 8.652007347861707e-06,
+      "loss": 0.6426,
+      "step": 23632
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.7792699734673634,
+      "learning_rate": 8.651251275057547e-06,
+      "loss": 0.7786,
+      "step": 23633
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4200764040949783,
+      "learning_rate": 8.650495210106225e-06,
+      "loss": 0.7006,
+      "step": 23634
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.343862504980939,
+      "learning_rate": 8.64973915301215e-06,
+      "loss": 0.6572,
+      "step": 23635
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.533695776292753,
+      "learning_rate": 8.648983103779722e-06,
+      "loss": 0.7293,
+      "step": 23636
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3492505980079508,
+      "learning_rate": 8.648227062413342e-06,
+      "loss": 0.6191,
+      "step": 23637
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4552740035608256,
+      "learning_rate": 8.647471028917409e-06,
+      "loss": 0.753,
+      "step": 23638
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4509285713270734,
+      "learning_rate": 8.64671500329633e-06,
+      "loss": 0.7205,
+      "step": 23639
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.590325358773272,
+      "learning_rate": 8.645958985554506e-06,
+      "loss": 0.6381,
+      "step": 23640
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4857039478184646,
+      "learning_rate": 8.645202975696336e-06,
+      "loss": 0.6628,
+      "step": 23641
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3109149560201017,
+      "learning_rate": 8.644446973726222e-06,
+      "loss": 0.6843,
+      "step": 23642
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9121086521261166,
+      "learning_rate": 8.643690979648567e-06,
+      "loss": 0.8265,
+      "step": 23643
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.438104364220211,
+      "learning_rate": 8.642934993467776e-06,
+      "loss": 0.6871,
+      "step": 23644
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3209751334352233,
+      "learning_rate": 8.642179015188245e-06,
+      "loss": 0.7682,
+      "step": 23645
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5272753848864318,
+      "learning_rate": 8.641423044814375e-06,
+      "loss": 0.7059,
+      "step": 23646
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5851629243611397,
+      "learning_rate": 8.640667082350574e-06,
+      "loss": 0.5182,
+      "step": 23647
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4546100106186677,
+      "learning_rate": 8.639911127801238e-06,
+      "loss": 0.7906,
+      "step": 23648
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6893599108934119,
+      "learning_rate": 8.639155181170771e-06,
+      "loss": 0.6856,
+      "step": 23649
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0008608382048005,
+      "learning_rate": 8.63839924246357e-06,
+      "loss": 0.7127,
+      "step": 23650
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6352167955926347,
+      "learning_rate": 8.637643311684045e-06,
+      "loss": 0.6789,
+      "step": 23651
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4654298781041704,
+      "learning_rate": 8.636887388836588e-06,
+      "loss": 0.7506,
+      "step": 23652
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3611757179638841,
+      "learning_rate": 8.636131473925605e-06,
+      "loss": 0.7295,
+      "step": 23653
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3017920389888749,
+      "learning_rate": 8.6353755669555e-06,
+      "loss": 0.6007,
+      "step": 23654
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8784499193490076,
+      "learning_rate": 8.634619667930664e-06,
+      "loss": 0.7617,
+      "step": 23655
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7102095902714913,
+      "learning_rate": 8.633863776855508e-06,
+      "loss": 0.7217,
+      "step": 23656
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5833128718784875,
+      "learning_rate": 8.633107893734427e-06,
+      "loss": 0.4963,
+      "step": 23657
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.190741907747485,
+      "learning_rate": 8.63235201857183e-06,
+      "loss": 0.7599,
+      "step": 23658
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.166921682913948,
+      "learning_rate": 8.631596151372107e-06,
+      "loss": 0.6924,
+      "step": 23659
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.331460340487849,
+      "learning_rate": 8.630840292139668e-06,
+      "loss": 0.7098,
+      "step": 23660
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2672632094786442,
+      "learning_rate": 8.630084440878906e-06,
+      "loss": 0.7085,
+      "step": 23661
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.345359084363138,
+      "learning_rate": 8.629328597594233e-06,
+      "loss": 0.7351,
+      "step": 23662
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5834440079084586,
+      "learning_rate": 8.628572762290038e-06,
+      "loss": 0.7762,
+      "step": 23663
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6004145435043529,
+      "learning_rate": 8.627816934970728e-06,
+      "loss": 0.5941,
+      "step": 23664
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2351185205304587,
+      "learning_rate": 8.627061115640703e-06,
+      "loss": 0.721,
+      "step": 23665
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5596786612003153,
+      "learning_rate": 8.62630530430436e-06,
+      "loss": 0.8413,
+      "step": 23666
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7087676585849725,
+      "learning_rate": 8.625549500966103e-06,
+      "loss": 0.6692,
+      "step": 23667
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7746516851003586,
+      "learning_rate": 8.624793705630334e-06,
+      "loss": 0.8431,
+      "step": 23668
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8019417666104498,
+      "learning_rate": 8.624037918301453e-06,
+      "loss": 0.6553,
+      "step": 23669
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.598355621694506,
+      "learning_rate": 8.623282138983854e-06,
+      "loss": 0.4966,
+      "step": 23670
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2695881574168522,
+      "learning_rate": 8.622526367681944e-06,
+      "loss": 0.6107,
+      "step": 23671
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.37188086819244,
+      "learning_rate": 8.621770604400126e-06,
+      "loss": 0.7976,
+      "step": 23672
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4221430215396054,
+      "learning_rate": 8.621014849142792e-06,
+      "loss": 0.8233,
+      "step": 23673
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6043014652493346,
+      "learning_rate": 8.620259101914344e-06,
+      "loss": 0.764,
+      "step": 23674
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7315891536954886,
+      "learning_rate": 8.619503362719186e-06,
+      "loss": 0.7769,
+      "step": 23675
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.44092887634861,
+      "learning_rate": 8.618747631561719e-06,
+      "loss": 0.8182,
+      "step": 23676
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.331852562265808,
+      "learning_rate": 8.617991908446339e-06,
+      "loss": 0.7095,
+      "step": 23677
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2651540782593982,
+      "learning_rate": 8.617236193377445e-06,
+      "loss": 0.737,
+      "step": 23678
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3564549898524325,
+      "learning_rate": 8.616480486359444e-06,
+      "loss": 0.8432,
+      "step": 23679
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6511388492483743,
+      "learning_rate": 8.61572478739673e-06,
+      "loss": 0.6848,
+      "step": 23680
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0796353542602763,
+      "learning_rate": 8.614969096493704e-06,
+      "loss": 0.8553,
+      "step": 23681
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.292052350500543,
+      "learning_rate": 8.614213413654764e-06,
+      "loss": 0.6779,
+      "step": 23682
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5905537605098912,
+      "learning_rate": 8.613457738884316e-06,
+      "loss": 0.5459,
+      "step": 23683
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3814748414184161,
+      "learning_rate": 8.612702072186757e-06,
+      "loss": 0.5775,
+      "step": 23684
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6183825071652621,
+      "learning_rate": 8.611946413566481e-06,
+      "loss": 0.5093,
+      "step": 23685
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7745836056349302,
+      "learning_rate": 8.611190763027898e-06,
+      "loss": 0.8193,
+      "step": 23686
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5263025621220254,
+      "learning_rate": 8.610435120575396e-06,
+      "loss": 0.6715,
+      "step": 23687
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2521054574981558,
+      "learning_rate": 8.609679486213384e-06,
+      "loss": 0.6098,
+      "step": 23688
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.518763006738659,
+      "learning_rate": 8.608923859946259e-06,
+      "loss": 0.8168,
+      "step": 23689
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3674095977323752,
+      "learning_rate": 8.60816824177842e-06,
+      "loss": 0.6883,
+      "step": 23690
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3341326152743676,
+      "learning_rate": 8.607412631714263e-06,
+      "loss": 0.7986,
+      "step": 23691
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6829535297735163,
+      "learning_rate": 8.606657029758193e-06,
+      "loss": 0.7208,
+      "step": 23692
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6961029733001987,
+      "learning_rate": 8.605901435914608e-06,
+      "loss": 0.7563,
+      "step": 23693
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4147652071715595,
+      "learning_rate": 8.605145850187904e-06,
+      "loss": 0.7118,
+      "step": 23694
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4517359053114887,
+      "learning_rate": 8.60439027258248e-06,
+      "loss": 0.7182,
+      "step": 23695
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.463089263698371,
+      "learning_rate": 8.60363470310274e-06,
+      "loss": 0.7044,
+      "step": 23696
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5643443219091846,
+      "learning_rate": 8.602879141753084e-06,
+      "loss": 0.5224,
+      "step": 23697
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2688483698457997,
+      "learning_rate": 8.602123588537906e-06,
+      "loss": 0.6395,
+      "step": 23698
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5468886322258573,
+      "learning_rate": 8.601368043461604e-06,
+      "loss": 0.7356,
+      "step": 23699
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2440938453403116,
+      "learning_rate": 8.600612506528584e-06,
+      "loss": 0.7062,
+      "step": 23700
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5045557554094662,
+      "learning_rate": 8.59985697774324e-06,
+      "loss": 0.5992,
+      "step": 23701
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.386627642178539,
+      "learning_rate": 8.599101457109971e-06,
+      "loss": 0.8577,
+      "step": 23702
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7842994429797807,
+      "learning_rate": 8.598345944633174e-06,
+      "loss": 0.7645,
+      "step": 23703
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3730054253849284,
+      "learning_rate": 8.597590440317257e-06,
+      "loss": 0.6802,
+      "step": 23704
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3522726463306367,
+      "learning_rate": 8.596834944166608e-06,
+      "loss": 0.6784,
+      "step": 23705
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8207589906176938,
+      "learning_rate": 8.59607945618563e-06,
+      "loss": 0.786,
+      "step": 23706
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.477909819590539,
+      "learning_rate": 8.595323976378727e-06,
+      "loss": 0.7646,
+      "step": 23707
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3004358613109783,
+      "learning_rate": 8.594568504750287e-06,
+      "loss": 0.5646,
+      "step": 23708
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3464187317431118,
+      "learning_rate": 8.593813041304716e-06,
+      "loss": 0.6756,
+      "step": 23709
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.730150064965771,
+      "learning_rate": 8.593057586046409e-06,
+      "loss": 0.707,
+      "step": 23710
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7267974219705802,
+      "learning_rate": 8.592302138979773e-06,
+      "loss": 0.6844,
+      "step": 23711
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5394793483978277,
+      "learning_rate": 8.59154670010919e-06,
+      "loss": 0.7607,
+      "step": 23712
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3605311062422936,
+      "learning_rate": 8.590791269439074e-06,
+      "loss": 0.6168,
+      "step": 23713
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3825161267318182,
+      "learning_rate": 8.590035846973817e-06,
+      "loss": 0.6881,
+      "step": 23714
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2730785822519988,
+      "learning_rate": 8.589280432717814e-06,
+      "loss": 0.7539,
+      "step": 23715
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4411415750297383,
+      "learning_rate": 8.58852502667547e-06,
+      "loss": 0.7326,
+      "step": 23716
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6011938381750234,
+      "learning_rate": 8.58776962885118e-06,
+      "loss": 0.5429,
+      "step": 23717
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5997027682215907,
+      "learning_rate": 8.587014239249344e-06,
+      "loss": 0.7511,
+      "step": 23718
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.404047647480911,
+      "learning_rate": 8.586258857874355e-06,
+      "loss": 0.7678,
+      "step": 23719
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1364855545212942,
+      "learning_rate": 8.585503484730616e-06,
+      "loss": 0.6236,
+      "step": 23720
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5137618661574856,
+      "learning_rate": 8.584748119822525e-06,
+      "loss": 0.6559,
+      "step": 23721
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.2729823726411307,
+      "learning_rate": 8.583992763154476e-06,
+      "loss": 0.6881,
+      "step": 23722
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2587861560187954,
+      "learning_rate": 8.583237414730869e-06,
+      "loss": 0.7622,
+      "step": 23723
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3901673934175556,
+      "learning_rate": 8.582482074556103e-06,
+      "loss": 0.6733,
+      "step": 23724
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.609718817771261,
+      "learning_rate": 8.581726742634579e-06,
+      "loss": 0.5277,
+      "step": 23725
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.350362483477858,
+      "learning_rate": 8.580971418970687e-06,
+      "loss": 0.6275,
+      "step": 23726
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5443130368974476,
+      "learning_rate": 8.580216103568828e-06,
+      "loss": 0.5231,
+      "step": 23727
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4440627453494523,
+      "learning_rate": 8.579460796433403e-06,
+      "loss": 0.7126,
+      "step": 23728
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4056788451166438,
+      "learning_rate": 8.578705497568807e-06,
+      "loss": 0.662,
+      "step": 23729
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3099690308191305,
+      "learning_rate": 8.577950206979436e-06,
+      "loss": 0.6263,
+      "step": 23730
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6118325829163622,
+      "learning_rate": 8.577194924669687e-06,
+      "loss": 0.8025,
+      "step": 23731
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5217960649025892,
+      "learning_rate": 8.576439650643964e-06,
+      "loss": 0.7946,
+      "step": 23732
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6189744856218536,
+      "learning_rate": 8.575684384906658e-06,
+      "loss": 0.5033,
+      "step": 23733
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3625398975342626,
+      "learning_rate": 8.574929127462168e-06,
+      "loss": 0.7881,
+      "step": 23734
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5046590306838818,
+      "learning_rate": 8.574173878314895e-06,
+      "loss": 0.7775,
+      "step": 23735
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3843878700121324,
+      "learning_rate": 8.573418637469228e-06,
+      "loss": 0.7519,
+      "step": 23736
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2654594996653676,
+      "learning_rate": 8.572663404929573e-06,
+      "loss": 0.8588,
+      "step": 23737
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2398123767212885,
+      "learning_rate": 8.57190818070032e-06,
+      "loss": 0.7469,
+      "step": 23738
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5876247737002169,
+      "learning_rate": 8.571152964785876e-06,
+      "loss": 0.5368,
+      "step": 23739
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.443648279662277,
+      "learning_rate": 8.570397757190624e-06,
+      "loss": 0.83,
+      "step": 23740
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6704758001549878,
+      "learning_rate": 8.569642557918974e-06,
+      "loss": 0.7194,
+      "step": 23741
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.128483425211595,
+      "learning_rate": 8.568887366975318e-06,
+      "loss": 0.6928,
+      "step": 23742
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.642679186633704,
+      "learning_rate": 8.56813218436405e-06,
+      "loss": 0.7356,
+      "step": 23743
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3868135654391858,
+      "learning_rate": 8.567377010089571e-06,
+      "loss": 0.7961,
+      "step": 23744
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2138040661344325,
+      "learning_rate": 8.566621844156275e-06,
+      "loss": 0.6911,
+      "step": 23745
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.402245616961743,
+      "learning_rate": 8.565866686568565e-06,
+      "loss": 0.7042,
+      "step": 23746
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4669551166175574,
+      "learning_rate": 8.565111537330831e-06,
+      "loss": 0.7642,
+      "step": 23747
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4082245625959489,
+      "learning_rate": 8.56435639644747e-06,
+      "loss": 0.7408,
+      "step": 23748
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3337169292571698,
+      "learning_rate": 8.563601263922883e-06,
+      "loss": 0.843,
+      "step": 23749
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4825726840526665,
+      "learning_rate": 8.562846139761462e-06,
+      "loss": 0.7809,
+      "step": 23750
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.424678505845095,
+      "learning_rate": 8.562091023967606e-06,
+      "loss": 0.7083,
+      "step": 23751
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.345524705029482,
+      "learning_rate": 8.56133591654571e-06,
+      "loss": 0.5716,
+      "step": 23752
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8887094747396407,
+      "learning_rate": 8.560580817500177e-06,
+      "loss": 0.6684,
+      "step": 23753
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.336265256976062,
+      "learning_rate": 8.559825726835395e-06,
+      "loss": 0.6845,
+      "step": 23754
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4723076803205932,
+      "learning_rate": 8.559070644555761e-06,
+      "loss": 0.7876,
+      "step": 23755
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.579962289662744,
+      "learning_rate": 8.55831557066568e-06,
+      "loss": 0.6948,
+      "step": 23756
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.690414655553235,
+      "learning_rate": 8.557560505169538e-06,
+      "loss": 0.659,
+      "step": 23757
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4409438606356975,
+      "learning_rate": 8.556805448071736e-06,
+      "loss": 0.8056,
+      "step": 23758
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2775011484111087,
+      "learning_rate": 8.556050399376667e-06,
+      "loss": 0.5901,
+      "step": 23759
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8658071143539974,
+      "learning_rate": 8.555295359088736e-06,
+      "loss": 0.6471,
+      "step": 23760
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5988908994042859,
+      "learning_rate": 8.554540327212328e-06,
+      "loss": 0.7668,
+      "step": 23761
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4388768490706045,
+      "learning_rate": 8.553785303751844e-06,
+      "loss": 0.7547,
+      "step": 23762
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4992160667624306,
+      "learning_rate": 8.553030288711683e-06,
+      "loss": 0.6914,
+      "step": 23763
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8145976277477902,
+      "learning_rate": 8.552275282096233e-06,
+      "loss": 0.8356,
+      "step": 23764
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5656442248307159,
+      "learning_rate": 8.551520283909898e-06,
+      "loss": 0.7848,
+      "step": 23765
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3840631386847333,
+      "learning_rate": 8.550765294157069e-06,
+      "loss": 0.7767,
+      "step": 23766
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7011784926680902,
+      "learning_rate": 8.550010312842148e-06,
+      "loss": 0.7046,
+      "step": 23767
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5627623553666514,
+      "learning_rate": 8.54925533996952e-06,
+      "loss": 0.6601,
+      "step": 23768
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6370810321924092,
+      "learning_rate": 8.548500375543587e-06,
+      "loss": 0.5433,
+      "step": 23769
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3766164685368392,
+      "learning_rate": 8.547745419568749e-06,
+      "loss": 0.7536,
+      "step": 23770
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4634824213782578,
+      "learning_rate": 8.546990472049394e-06,
+      "loss": 0.7049,
+      "step": 23771
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4978574798569755,
+      "learning_rate": 8.546235532989919e-06,
+      "loss": 0.748,
+      "step": 23772
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.613474102867472,
+      "learning_rate": 8.545480602394721e-06,
+      "loss": 0.7113,
+      "step": 23773
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7467287897160368,
+      "learning_rate": 8.544725680268199e-06,
+      "loss": 0.8269,
+      "step": 23774
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8382048421247836,
+      "learning_rate": 8.543970766614743e-06,
+      "loss": 0.8905,
+      "step": 23775
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6522328750796846,
+      "learning_rate": 8.543215861438749e-06,
+      "loss": 0.7383,
+      "step": 23776
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.287444335928611,
+      "learning_rate": 8.542460964744615e-06,
+      "loss": 0.6405,
+      "step": 23777
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.7756737066459043,
+      "learning_rate": 8.541706076536735e-06,
+      "loss": 0.6854,
+      "step": 23778
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.658135146946668,
+      "learning_rate": 8.540951196819502e-06,
+      "loss": 0.5703,
+      "step": 23779
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.287578750103298,
+      "learning_rate": 8.540196325597312e-06,
+      "loss": 0.7387,
+      "step": 23780
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6429911678431105,
+      "learning_rate": 8.539441462874566e-06,
+      "loss": 0.524,
+      "step": 23781
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5608993981115424,
+      "learning_rate": 8.538686608655652e-06,
+      "loss": 0.7438,
+      "step": 23782
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4365194831254968,
+      "learning_rate": 8.537931762944966e-06,
+      "loss": 0.5961,
+      "step": 23783
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.75266853227751,
+      "learning_rate": 8.537176925746902e-06,
+      "loss": 0.6494,
+      "step": 23784
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.2725477642799743,
+      "learning_rate": 8.536422097065863e-06,
+      "loss": 0.6399,
+      "step": 23785
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.46714489148579,
+      "learning_rate": 8.535667276906234e-06,
+      "loss": 0.7059,
+      "step": 23786
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.745697069540704,
+      "learning_rate": 8.534912465272414e-06,
+      "loss": 0.7536,
+      "step": 23787
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3851580660192089,
+      "learning_rate": 8.5341576621688e-06,
+      "loss": 0.7826,
+      "step": 23788
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5956104327625775,
+      "learning_rate": 8.53340286759978e-06,
+      "loss": 0.7249,
+      "step": 23789
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.784310068263616,
+      "learning_rate": 8.532648081569756e-06,
+      "loss": 0.6812,
+      "step": 23790
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3686823141385314,
+      "learning_rate": 8.531893304083117e-06,
+      "loss": 0.5865,
+      "step": 23791
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3253742441273704,
+      "learning_rate": 8.531138535144265e-06,
+      "loss": 0.662,
+      "step": 23792
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5403901596489464,
+      "learning_rate": 8.530383774757583e-06,
+      "loss": 0.7036,
+      "step": 23793
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3068510393218178,
+      "learning_rate": 8.529629022927474e-06,
+      "loss": 0.7412,
+      "step": 23794
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.40940378400801,
+      "learning_rate": 8.528874279658335e-06,
+      "loss": 0.7347,
+      "step": 23795
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7229044641250917,
+      "learning_rate": 8.528119544954551e-06,
+      "loss": 0.6908,
+      "step": 23796
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4322537434386295,
+      "learning_rate": 8.52736481882052e-06,
+      "loss": 0.7134,
+      "step": 23797
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3551911518687747,
+      "learning_rate": 8.526610101260639e-06,
+      "loss": 0.7532,
+      "step": 23798
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5747867738693293,
+      "learning_rate": 8.525855392279302e-06,
+      "loss": 0.7088,
+      "step": 23799
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6457336540820349,
+      "learning_rate": 8.525100691880898e-06,
+      "loss": 0.5601,
+      "step": 23800
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2193959331486062,
+      "learning_rate": 8.524346000069828e-06,
+      "loss": 0.7478,
+      "step": 23801
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3223402370644626,
+      "learning_rate": 8.523591316850483e-06,
+      "loss": 0.7246,
+      "step": 23802
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3720913182569043,
+      "learning_rate": 8.522836642227254e-06,
+      "loss": 0.815,
+      "step": 23803
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4703650682064997,
+      "learning_rate": 8.522081976204539e-06,
+      "loss": 0.5778,
+      "step": 23804
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5555989403410918,
+      "learning_rate": 8.52132731878673e-06,
+      "loss": 0.486,
+      "step": 23805
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3375025640735108,
+      "learning_rate": 8.520572669978224e-06,
+      "loss": 0.6738,
+      "step": 23806
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.389535654966582,
+      "learning_rate": 8.519818029783412e-06,
+      "loss": 0.65,
+      "step": 23807
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.537803880520017,
+      "learning_rate": 8.519063398206684e-06,
+      "loss": 0.6236,
+      "step": 23808
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4064553286197166,
+      "learning_rate": 8.518308775252444e-06,
+      "loss": 0.5993,
+      "step": 23809
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.592914040468563,
+      "learning_rate": 8.517554160925073e-06,
+      "loss": 0.5116,
+      "step": 23810
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5769708863150111,
+      "learning_rate": 8.516799555228974e-06,
+      "loss": 0.5358,
+      "step": 23811
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5777103865018621,
+      "learning_rate": 8.516044958168535e-06,
+      "loss": 0.5398,
+      "step": 23812
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4068537007441506,
+      "learning_rate": 8.515290369748158e-06,
+      "loss": 0.7868,
+      "step": 23813
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5844352465143736,
+      "learning_rate": 8.514535789972226e-06,
+      "loss": 0.6636,
+      "step": 23814
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.23329923637442,
+      "learning_rate": 8.513781218845137e-06,
+      "loss": 0.6521,
+      "step": 23815
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4688922594205938,
+      "learning_rate": 8.513026656371288e-06,
+      "loss": 0.7637,
+      "step": 23816
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3580826307115275,
+      "learning_rate": 8.512272102555063e-06,
+      "loss": 0.7149,
+      "step": 23817
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3979016264415796,
+      "learning_rate": 8.511517557400864e-06,
+      "loss": 0.7348,
+      "step": 23818
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.410199775052573,
+      "learning_rate": 8.51076302091308e-06,
+      "loss": 0.6563,
+      "step": 23819
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6694629281896888,
+      "learning_rate": 8.510008493096109e-06,
+      "loss": 0.489,
+      "step": 23820
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.192918745833377,
+      "learning_rate": 8.509253973954334e-06,
+      "loss": 0.6447,
+      "step": 23821
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3828962045501787,
+      "learning_rate": 8.508499463492157e-06,
+      "loss": 0.8122,
+      "step": 23822
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3173884947125274,
+      "learning_rate": 8.507744961713972e-06,
+      "loss": 0.6459,
+      "step": 23823
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3746043676070434,
+      "learning_rate": 8.506990468624164e-06,
+      "loss": 0.6754,
+      "step": 23824
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6130228942620082,
+      "learning_rate": 8.506235984227129e-06,
+      "loss": 0.5207,
+      "step": 23825
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4687773444066483,
+      "learning_rate": 8.505481508527263e-06,
+      "loss": 0.7614,
+      "step": 23826
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.831366715706316,
+      "learning_rate": 8.504727041528959e-06,
+      "loss": 0.7074,
+      "step": 23827
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9937963977969333,
+      "learning_rate": 8.503972583236605e-06,
+      "loss": 0.7099,
+      "step": 23828
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5722041512049652,
+      "learning_rate": 8.503218133654595e-06,
+      "loss": 0.847,
+      "step": 23829
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3856538984157143,
+      "learning_rate": 8.502463692787327e-06,
+      "loss": 0.7724,
+      "step": 23830
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4622812251212023,
+      "learning_rate": 8.501709260639187e-06,
+      "loss": 0.8072,
+      "step": 23831
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3818568988939794,
+      "learning_rate": 8.50095483721457e-06,
+      "loss": 0.7885,
+      "step": 23832
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7339002898570604,
+      "learning_rate": 8.500200422517866e-06,
+      "loss": 0.578,
+      "step": 23833
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3198058691021113,
+      "learning_rate": 8.499446016553475e-06,
+      "loss": 0.8493,
+      "step": 23834
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4031938566269606,
+      "learning_rate": 8.498691619325782e-06,
+      "loss": 0.7169,
+      "step": 23835
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6877065569301677,
+      "learning_rate": 8.497937230839182e-06,
+      "loss": 0.7977,
+      "step": 23836
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3329159250941751,
+      "learning_rate": 8.497182851098068e-06,
+      "loss": 0.8076,
+      "step": 23837
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2666022356600652,
+      "learning_rate": 8.496428480106828e-06,
+      "loss": 0.6036,
+      "step": 23838
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.445882778590655,
+      "learning_rate": 8.495674117869859e-06,
+      "loss": 0.7344,
+      "step": 23839
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3830396290812,
+      "learning_rate": 8.494919764391548e-06,
+      "loss": 0.6289,
+      "step": 23840
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5002057776539588,
+      "learning_rate": 8.4941654196763e-06,
+      "loss": 0.7446,
+      "step": 23841
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.422820115015986,
+      "learning_rate": 8.493411083728491e-06,
+      "loss": 0.6384,
+      "step": 23842
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8681423284886085,
+      "learning_rate": 8.49265675655252e-06,
+      "loss": 0.6801,
+      "step": 23843
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5841163784386304,
+      "learning_rate": 8.49190243815278e-06,
+      "loss": 0.5074,
+      "step": 23844
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4376280717695311,
+      "learning_rate": 8.491148128533662e-06,
+      "loss": 0.7353,
+      "step": 23845
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5165261032917587,
+      "learning_rate": 8.490393827699554e-06,
+      "loss": 0.793,
+      "step": 23846
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8109111398799584,
+      "learning_rate": 8.489639535654855e-06,
+      "loss": 0.6681,
+      "step": 23847
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4728182032781734,
+      "learning_rate": 8.488885252403953e-06,
+      "loss": 0.6445,
+      "step": 23848
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4916015567861647,
+      "learning_rate": 8.488130977951235e-06,
+      "loss": 0.6694,
+      "step": 23849
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.415254667501784,
+      "learning_rate": 8.487376712301099e-06,
+      "loss": 0.7345,
+      "step": 23850
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7113329949136344,
+      "learning_rate": 8.486622455457938e-06,
+      "loss": 0.6147,
+      "step": 23851
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5808397152503664,
+      "learning_rate": 8.485868207426136e-06,
+      "loss": 0.8774,
+      "step": 23852
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.755445435316615,
+      "learning_rate": 8.485113968210089e-06,
+      "loss": 0.8183,
+      "step": 23853
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4006572088744453,
+      "learning_rate": 8.484359737814189e-06,
+      "loss": 0.6525,
+      "step": 23854
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3360555894242567,
+      "learning_rate": 8.483605516242829e-06,
+      "loss": 0.661,
+      "step": 23855
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6083641429578326,
+      "learning_rate": 8.482851303500395e-06,
+      "loss": 0.8184,
+      "step": 23856
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6789977154113267,
+      "learning_rate": 8.48209709959128e-06,
+      "loss": 0.8144,
+      "step": 23857
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4512497252217593,
+      "learning_rate": 8.481342904519882e-06,
+      "loss": 0.6566,
+      "step": 23858
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4060642920333093,
+      "learning_rate": 8.480588718290579e-06,
+      "loss": 0.8334,
+      "step": 23859
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3096531734247019,
+      "learning_rate": 8.479834540907774e-06,
+      "loss": 0.669,
+      "step": 23860
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6032400173191942,
+      "learning_rate": 8.479080372375848e-06,
+      "loss": 0.5345,
+      "step": 23861
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.578133913451106,
+      "learning_rate": 8.478326212699205e-06,
+      "loss": 0.7657,
+      "step": 23862
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3507683805008417,
+      "learning_rate": 8.477572061882225e-06,
+      "loss": 0.7404,
+      "step": 23863
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6078442431531007,
+      "learning_rate": 8.476817919929304e-06,
+      "loss": 0.513,
+      "step": 23864
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3733914875788342,
+      "learning_rate": 8.47606378684483e-06,
+      "loss": 0.8152,
+      "step": 23865
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6384420642990533,
+      "learning_rate": 8.475309662633195e-06,
+      "loss": 0.5165,
+      "step": 23866
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3496962619954176,
+      "learning_rate": 8.474555547298789e-06,
+      "loss": 0.7412,
+      "step": 23867
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3459697385075529,
+      "learning_rate": 8.473801440846005e-06,
+      "loss": 0.7927,
+      "step": 23868
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3220862318709483,
+      "learning_rate": 8.473047343279233e-06,
+      "loss": 0.6715,
+      "step": 23869
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.358573427822603,
+      "learning_rate": 8.472293254602858e-06,
+      "loss": 0.675,
+      "step": 23870
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.390582563691716,
+      "learning_rate": 8.47153917482128e-06,
+      "loss": 0.6994,
+      "step": 23871
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2803556445783912,
+      "learning_rate": 8.470785103938887e-06,
+      "loss": 0.7109,
+      "step": 23872
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.380011385571657,
+      "learning_rate": 8.470031041960062e-06,
+      "loss": 0.6306,
+      "step": 23873
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.522216031239523,
+      "learning_rate": 8.469276988889202e-06,
+      "loss": 0.6943,
+      "step": 23874
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6920834031137677,
+      "learning_rate": 8.468522944730697e-06,
+      "loss": 0.7185,
+      "step": 23875
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4266569767770538,
+      "learning_rate": 8.467768909488938e-06,
+      "loss": 0.7568,
+      "step": 23876
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6282474645440952,
+      "learning_rate": 8.46701488316831e-06,
+      "loss": 0.8303,
+      "step": 23877
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.512862401968997,
+      "learning_rate": 8.466260865773207e-06,
+      "loss": 0.6914,
+      "step": 23878
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1637542865880286,
+      "learning_rate": 8.465506857308021e-06,
+      "loss": 0.683,
+      "step": 23879
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4049513388589434,
+      "learning_rate": 8.46475285777714e-06,
+      "loss": 0.7474,
+      "step": 23880
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4756403435457486,
+      "learning_rate": 8.463998867184952e-06,
+      "loss": 0.6704,
+      "step": 23881
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.858706144401248,
+      "learning_rate": 8.463244885535849e-06,
+      "loss": 0.7809,
+      "step": 23882
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5557804529962613,
+      "learning_rate": 8.462490912834222e-06,
+      "loss": 0.7822,
+      "step": 23883
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.486061607250021,
+      "learning_rate": 8.461736949084458e-06,
+      "loss": 0.7899,
+      "step": 23884
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4525473490539793,
+      "learning_rate": 8.460982994290951e-06,
+      "loss": 0.7222,
+      "step": 23885
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9418686194715273,
+      "learning_rate": 8.46022904845809e-06,
+      "loss": 0.7863,
+      "step": 23886
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3581193904121336,
+      "learning_rate": 8.459475111590256e-06,
+      "loss": 0.621,
+      "step": 23887
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.62610396635378,
+      "learning_rate": 8.45872118369185e-06,
+      "loss": 0.4963,
+      "step": 23888
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3629623412885215,
+      "learning_rate": 8.457967264767253e-06,
+      "loss": 0.7251,
+      "step": 23889
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4463737934034977,
+      "learning_rate": 8.457213354820868e-06,
+      "loss": 0.6955,
+      "step": 23890
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2877071458358555,
+      "learning_rate": 8.456459453857067e-06,
+      "loss": 0.7004,
+      "step": 23891
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4042819509000033,
+      "learning_rate": 8.45570556188025e-06,
+      "loss": 0.7082,
+      "step": 23892
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4119965225747277,
+      "learning_rate": 8.454951678894807e-06,
+      "loss": 0.6827,
+      "step": 23893
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3536073677811948,
+      "learning_rate": 8.454197804905123e-06,
+      "loss": 0.681,
+      "step": 23894
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2318027282153619,
+      "learning_rate": 8.453443939915586e-06,
+      "loss": 0.6165,
+      "step": 23895
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4437252280892774,
+      "learning_rate": 8.45269008393059e-06,
+      "loss": 0.6914,
+      "step": 23896
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.629837110988919,
+      "learning_rate": 8.451936236954524e-06,
+      "loss": 0.5328,
+      "step": 23897
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7205052725316805,
+      "learning_rate": 8.45118239899177e-06,
+      "loss": 0.6378,
+      "step": 23898
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8092690567428662,
+      "learning_rate": 8.450428570046725e-06,
+      "loss": 0.7637,
+      "step": 23899
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6005034475827322,
+      "learning_rate": 8.44967475012378e-06,
+      "loss": 0.7327,
+      "step": 23900
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5021948925677973,
+      "learning_rate": 8.448920939227316e-06,
+      "loss": 0.763,
+      "step": 23901
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.300455182308234,
+      "learning_rate": 8.448167137361723e-06,
+      "loss": 0.6895,
+      "step": 23902
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5891661677987674,
+      "learning_rate": 8.447413344531394e-06,
+      "loss": 0.5035,
+      "step": 23903
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5731693649682896,
+      "learning_rate": 8.446659560740718e-06,
+      "loss": 0.73,
+      "step": 23904
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4239467427661112,
+      "learning_rate": 8.44590578599408e-06,
+      "loss": 0.5959,
+      "step": 23905
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2389558870835846,
+      "learning_rate": 8.445152020295869e-06,
+      "loss": 0.7834,
+      "step": 23906
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.24787879876723,
+      "learning_rate": 8.444398263650477e-06,
+      "loss": 0.7114,
+      "step": 23907
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4611015198367494,
+      "learning_rate": 8.443644516062293e-06,
+      "loss": 0.6723,
+      "step": 23908
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6781366299753944,
+      "learning_rate": 8.4428907775357e-06,
+      "loss": 0.7456,
+      "step": 23909
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6152924715415052,
+      "learning_rate": 8.44213704807509e-06,
+      "loss": 0.5265,
+      "step": 23910
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4443313190595606,
+      "learning_rate": 8.441383327684854e-06,
+      "loss": 0.7549,
+      "step": 23911
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3079072146406485,
+      "learning_rate": 8.440629616369375e-06,
+      "loss": 0.7715,
+      "step": 23912
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.560289618248308,
+      "learning_rate": 8.439875914133044e-06,
+      "loss": 0.844,
+      "step": 23913
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.541287306189884,
+      "learning_rate": 8.439122220980249e-06,
+      "loss": 0.7323,
+      "step": 23914
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3459180030914641,
+      "learning_rate": 8.438368536915382e-06,
+      "loss": 0.7063,
+      "step": 23915
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4336628114223187,
+      "learning_rate": 8.437614861942826e-06,
+      "loss": 0.7192,
+      "step": 23916
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5153855234912934,
+      "learning_rate": 8.43686119606697e-06,
+      "loss": 0.7677,
+      "step": 23917
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4564372593634805,
+      "learning_rate": 8.436107539292206e-06,
+      "loss": 0.768,
+      "step": 23918
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6003095310126189,
+      "learning_rate": 8.435353891622915e-06,
+      "loss": 0.5449,
+      "step": 23919
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.333980001327482,
+      "learning_rate": 8.434600253063489e-06,
+      "loss": 0.604,
+      "step": 23920
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.360104351082175,
+      "learning_rate": 8.433846623618319e-06,
+      "loss": 0.6996,
+      "step": 23921
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3932035101379678,
+      "learning_rate": 8.43309300329179e-06,
+      "loss": 0.6644,
+      "step": 23922
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4551904991779825,
+      "learning_rate": 8.432339392088286e-06,
+      "loss": 0.795,
+      "step": 23923
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3133748627243274,
+      "learning_rate": 8.4315857900122e-06,
+      "loss": 0.7036,
+      "step": 23924
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8016336845236547,
+      "learning_rate": 8.43083219706792e-06,
+      "loss": 0.7467,
+      "step": 23925
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7000373793802876,
+      "learning_rate": 8.430078613259832e-06,
+      "loss": 0.6691,
+      "step": 23926
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5002612321813558,
+      "learning_rate": 8.42932503859232e-06,
+      "loss": 0.8241,
+      "step": 23927
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3755164992900362,
+      "learning_rate": 8.428571473069775e-06,
+      "loss": 0.772,
+      "step": 23928
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3221205172227988,
+      "learning_rate": 8.427817916696588e-06,
+      "loss": 0.564,
+      "step": 23929
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4610321988901074,
+      "learning_rate": 8.42706436947714e-06,
+      "loss": 0.7359,
+      "step": 23930
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5542831103427047,
+      "learning_rate": 8.42631083141582e-06,
+      "loss": 0.537,
+      "step": 23931
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.459980931342491,
+      "learning_rate": 8.42555730251702e-06,
+      "loss": 0.7201,
+      "step": 23932
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3571252580273738,
+      "learning_rate": 8.424803782785123e-06,
+      "loss": 0.7753,
+      "step": 23933
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4299645779738575,
+      "learning_rate": 8.424050272224515e-06,
+      "loss": 0.6772,
+      "step": 23934
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6899650231639725,
+      "learning_rate": 8.423296770839584e-06,
+      "loss": 0.5739,
+      "step": 23935
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3992347694249359,
+      "learning_rate": 8.422543278634726e-06,
+      "loss": 0.8064,
+      "step": 23936
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.182549214571428,
+      "learning_rate": 8.421789795614315e-06,
+      "loss": 0.6597,
+      "step": 23937
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7007665941016836,
+      "learning_rate": 8.421036321782741e-06,
+      "loss": 0.6612,
+      "step": 23938
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2777885279786305,
+      "learning_rate": 8.4202828571444e-06,
+      "loss": 0.7246,
+      "step": 23939
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.432140525288048,
+      "learning_rate": 8.419529401703667e-06,
+      "loss": 0.8149,
+      "step": 23940
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4273936542330046,
+      "learning_rate": 8.418775955464935e-06,
+      "loss": 0.7671,
+      "step": 23941
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.416169271673204,
+      "learning_rate": 8.418022518432588e-06,
+      "loss": 0.8167,
+      "step": 23942
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5890514493137812,
+      "learning_rate": 8.417269090611023e-06,
+      "loss": 0.5226,
+      "step": 23943
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5879937506911994,
+      "learning_rate": 8.416515672004611e-06,
+      "loss": 0.57,
+      "step": 23944
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3893898269368712,
+      "learning_rate": 8.415762262617748e-06,
+      "loss": 0.7567,
+      "step": 23945
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4510300547418318,
+      "learning_rate": 8.41500886245482e-06,
+      "loss": 0.6603,
+      "step": 23946
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6066234369934159,
+      "learning_rate": 8.41425547152021e-06,
+      "loss": 0.5068,
+      "step": 23947
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6909874168383614,
+      "learning_rate": 8.413502089818307e-06,
+      "loss": 0.5863,
+      "step": 23948
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.720801639456546,
+      "learning_rate": 8.412748717353499e-06,
+      "loss": 0.8341,
+      "step": 23949
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.532333408799886,
+      "learning_rate": 8.41199535413017e-06,
+      "loss": 0.7028,
+      "step": 23950
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2371321472262806,
+      "learning_rate": 8.411242000152705e-06,
+      "loss": 0.6991,
+      "step": 23951
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6270798321796742,
+      "learning_rate": 8.410488655425493e-06,
+      "loss": 0.6496,
+      "step": 23952
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5757948533691677,
+      "learning_rate": 8.40973531995292e-06,
+      "loss": 0.5243,
+      "step": 23953
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3952540277162793,
+      "learning_rate": 8.40898199373937e-06,
+      "loss": 0.5369,
+      "step": 23954
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.308202740225176,
+      "learning_rate": 8.40822867678923e-06,
+      "loss": 0.6969,
+      "step": 23955
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.343968286869781,
+      "learning_rate": 8.407475369106889e-06,
+      "loss": 0.6843,
+      "step": 23956
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6547442226474851,
+      "learning_rate": 8.406722070696731e-06,
+      "loss": 0.5616,
+      "step": 23957
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.309898149837266,
+      "learning_rate": 8.405968781563141e-06,
+      "loss": 0.6996,
+      "step": 23958
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4354144875978596,
+      "learning_rate": 8.405215501710501e-06,
+      "loss": 0.6911,
+      "step": 23959
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4727704475775762,
+      "learning_rate": 8.404462231143208e-06,
+      "loss": 0.725,
+      "step": 23960
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4481825228555314,
+      "learning_rate": 8.403708969865638e-06,
+      "loss": 0.5882,
+      "step": 23961
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5686246273646158,
+      "learning_rate": 8.402955717882179e-06,
+      "loss": 0.7532,
+      "step": 23962
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.475061226461315,
+      "learning_rate": 8.402202475197216e-06,
+      "loss": 0.6544,
+      "step": 23963
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6067066997781438,
+      "learning_rate": 8.40144924181514e-06,
+      "loss": 0.659,
+      "step": 23964
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.716856849363187,
+      "learning_rate": 8.400696017740329e-06,
+      "loss": 0.7766,
+      "step": 23965
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5615695446070336,
+      "learning_rate": 8.399942802977175e-06,
+      "loss": 0.4653,
+      "step": 23966
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.556600589963544,
+      "learning_rate": 8.39918959753006e-06,
+      "loss": 0.7619,
+      "step": 23967
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4252389269568144,
+      "learning_rate": 8.398436401403368e-06,
+      "loss": 0.8195,
+      "step": 23968
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3953422422368067,
+      "learning_rate": 8.397683214601487e-06,
+      "loss": 0.6927,
+      "step": 23969
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4289330858989315,
+      "learning_rate": 8.396930037128803e-06,
+      "loss": 0.6884,
+      "step": 23970
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5693242351055087,
+      "learning_rate": 8.3961768689897e-06,
+      "loss": 0.523,
+      "step": 23971
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4831438639160053,
+      "learning_rate": 8.395423710188561e-06,
+      "loss": 0.7289,
+      "step": 23972
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7096006012849334,
+      "learning_rate": 8.394670560729774e-06,
+      "loss": 0.697,
+      "step": 23973
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5654453356452486,
+      "learning_rate": 8.393917420617726e-06,
+      "loss": 0.5221,
+      "step": 23974
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3249496654482955,
+      "learning_rate": 8.393164289856797e-06,
+      "loss": 0.7248,
+      "step": 23975
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5315382186184416,
+      "learning_rate": 8.392411168451371e-06,
+      "loss": 0.7491,
+      "step": 23976
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5904925829254591,
+      "learning_rate": 8.39165805640584e-06,
+      "loss": 0.4861,
+      "step": 23977
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5806864048682501,
+      "learning_rate": 8.390904953724586e-06,
+      "loss": 0.5238,
+      "step": 23978
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5498585580078752,
+      "learning_rate": 8.390151860411991e-06,
+      "loss": 0.7048,
+      "step": 23979
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5207803282882344,
+      "learning_rate": 8.389398776472439e-06,
+      "loss": 0.765,
+      "step": 23980
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6165751693264659,
+      "learning_rate": 8.38864570191032e-06,
+      "loss": 0.5628,
+      "step": 23981
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3976575366101438,
+      "learning_rate": 8.387892636730016e-06,
+      "loss": 0.8334,
+      "step": 23982
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3068231921087752,
+      "learning_rate": 8.387139580935912e-06,
+      "loss": 0.6482,
+      "step": 23983
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.253746879879544,
+      "learning_rate": 8.386386534532388e-06,
+      "loss": 0.7281,
+      "step": 23984
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5362058897413011,
+      "learning_rate": 8.385633497523838e-06,
+      "loss": 0.6802,
+      "step": 23985
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.418808927769851,
+      "learning_rate": 8.384880469914638e-06,
+      "loss": 0.6336,
+      "step": 23986
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.480333822840489,
+      "learning_rate": 8.384127451709173e-06,
+      "loss": 0.6816,
+      "step": 23987
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.749281284876867,
+      "learning_rate": 8.383374442911836e-06,
+      "loss": 0.6299,
+      "step": 23988
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3958433702382442,
+      "learning_rate": 8.382621443527e-06,
+      "loss": 0.7623,
+      "step": 23989
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5950274269708598,
+      "learning_rate": 8.381868453559053e-06,
+      "loss": 0.7265,
+      "step": 23990
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4979477192170028,
+      "learning_rate": 8.381115473012381e-06,
+      "loss": 0.5926,
+      "step": 23991
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4631477676645672,
+      "learning_rate": 8.380362501891373e-06,
+      "loss": 0.9359,
+      "step": 23992
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.63608797572107,
+      "learning_rate": 8.3796095402004e-06,
+      "loss": 0.8208,
+      "step": 23993
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4097914166108922,
+      "learning_rate": 8.378856587943856e-06,
+      "loss": 0.8227,
+      "step": 23994
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.776199239476284,
+      "learning_rate": 8.378103645126124e-06,
+      "loss": 0.7323,
+      "step": 23995
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6213706397571273,
+      "learning_rate": 8.377350711751583e-06,
+      "loss": 0.7782,
+      "step": 23996
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3150355478963143,
+      "learning_rate": 8.376597787824618e-06,
+      "loss": 0.7218,
+      "step": 23997
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3506447097518544,
+      "learning_rate": 8.37584487334962e-06,
+      "loss": 0.7299,
+      "step": 23998
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3165416664792018,
+      "learning_rate": 8.375091968330966e-06,
+      "loss": 0.8002,
+      "step": 23999
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.520896668900491,
+      "learning_rate": 8.374339072773036e-06,
+      "loss": 0.7912,
+      "step": 24000
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.342737420184254,
+      "learning_rate": 8.373586186680224e-06,
+      "loss": 0.7136,
+      "step": 24001
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5128572533937261,
+      "learning_rate": 8.37283331005691e-06,
+      "loss": 0.6284,
+      "step": 24002
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.595014366694462,
+      "learning_rate": 8.372080442907471e-06,
+      "loss": 0.5643,
+      "step": 24003
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5772648227945727,
+      "learning_rate": 8.371327585236294e-06,
+      "loss": 0.5072,
+      "step": 24004
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3973698721518704,
+      "learning_rate": 8.370574737047765e-06,
+      "loss": 0.6256,
+      "step": 24005
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1727170169254528,
+      "learning_rate": 8.369821898346269e-06,
+      "loss": 0.778,
+      "step": 24006
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3980399315570826,
+      "learning_rate": 8.369069069136184e-06,
+      "loss": 0.6799,
+      "step": 24007
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.394219288452312,
+      "learning_rate": 8.368316249421894e-06,
+      "loss": 0.7575,
+      "step": 24008
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.466889245441374,
+      "learning_rate": 8.367563439207784e-06,
+      "loss": 0.7031,
+      "step": 24009
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4023497257368502,
+      "learning_rate": 8.366810638498237e-06,
+      "loss": 0.661,
+      "step": 24010
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3569895930081608,
+      "learning_rate": 8.366057847297635e-06,
+      "loss": 0.7601,
+      "step": 24011
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5326042092165169,
+      "learning_rate": 8.36530506561036e-06,
+      "loss": 0.7068,
+      "step": 24012
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5962945575797038,
+      "learning_rate": 8.3645522934408e-06,
+      "loss": 0.7269,
+      "step": 24013
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3445886634599737,
+      "learning_rate": 8.363799530793332e-06,
+      "loss": 0.6313,
+      "step": 24014
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4485782288810432,
+      "learning_rate": 8.363046777672342e-06,
+      "loss": 0.7895,
+      "step": 24015
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3933759589775891,
+      "learning_rate": 8.362294034082214e-06,
+      "loss": 0.697,
+      "step": 24016
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2924851507572979,
+      "learning_rate": 8.361541300027323e-06,
+      "loss": 0.7518,
+      "step": 24017
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4261896295875398,
+      "learning_rate": 8.36078857551206e-06,
+      "loss": 0.7536,
+      "step": 24018
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.049250592944413,
+      "learning_rate": 8.360035860540806e-06,
+      "loss": 0.7046,
+      "step": 24019
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4406963993312252,
+      "learning_rate": 8.359283155117943e-06,
+      "loss": 0.7176,
+      "step": 24020
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4246467673540595,
+      "learning_rate": 8.35853045924785e-06,
+      "loss": 0.682,
+      "step": 24021
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8115386042078894,
+      "learning_rate": 8.357777772934914e-06,
+      "loss": 0.7408,
+      "step": 24022
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.598368013483563,
+      "learning_rate": 8.357025096183518e-06,
+      "loss": 0.7488,
+      "step": 24023
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6453628886554745,
+      "learning_rate": 8.356272428998039e-06,
+      "loss": 0.7074,
+      "step": 24024
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6388497872757084,
+      "learning_rate": 8.355519771382861e-06,
+      "loss": 0.6375,
+      "step": 24025
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5680688468505405,
+      "learning_rate": 8.35476712334237e-06,
+      "loss": 0.5338,
+      "step": 24026
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5537041895165096,
+      "learning_rate": 8.354014484880946e-06,
+      "loss": 0.6064,
+      "step": 24027
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5629707428712534,
+      "learning_rate": 8.35326185600297e-06,
+      "loss": 0.7303,
+      "step": 24028
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6190431119025255,
+      "learning_rate": 8.352509236712823e-06,
+      "loss": 0.6144,
+      "step": 24029
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3093162994503367,
+      "learning_rate": 8.351756627014894e-06,
+      "loss": 0.5999,
+      "step": 24030
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.405749815606497,
+      "learning_rate": 8.351004026913555e-06,
+      "loss": 0.7899,
+      "step": 24031
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4626603134428289,
+      "learning_rate": 8.350251436413194e-06,
+      "loss": 0.732,
+      "step": 24032
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3289694058739019,
+      "learning_rate": 8.34949885551819e-06,
+      "loss": 0.6881,
+      "step": 24033
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.372794737810375,
+      "learning_rate": 8.34874628423293e-06,
+      "loss": 0.7331,
+      "step": 24034
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6337618906718355,
+      "learning_rate": 8.347993722561789e-06,
+      "loss": 0.516,
+      "step": 24035
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6217431955266065,
+      "learning_rate": 8.34724117050915e-06,
+      "loss": 0.5485,
+      "step": 24036
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4335293268111884,
+      "learning_rate": 8.346488628079397e-06,
+      "loss": 0.7515,
+      "step": 24037
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2765564998277008,
+      "learning_rate": 8.345736095276914e-06,
+      "loss": 0.8027,
+      "step": 24038
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.367638680068648,
+      "learning_rate": 8.344983572106077e-06,
+      "loss": 0.7864,
+      "step": 24039
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4259062271445315,
+      "learning_rate": 8.344231058571268e-06,
+      "loss": 0.7611,
+      "step": 24040
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3782844758153783,
+      "learning_rate": 8.343478554676876e-06,
+      "loss": 0.6521,
+      "step": 24041
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4634963500932707,
+      "learning_rate": 8.342726060427271e-06,
+      "loss": 0.6544,
+      "step": 24042
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.39759439722507,
+      "learning_rate": 8.34197357582684e-06,
+      "loss": 0.7907,
+      "step": 24043
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3787076722407148,
+      "learning_rate": 8.341221100879962e-06,
+      "loss": 0.68,
+      "step": 24044
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6116818884554822,
+      "learning_rate": 8.340468635591023e-06,
+      "loss": 0.5463,
+      "step": 24045
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3526887175135007,
+      "learning_rate": 8.3397161799644e-06,
+      "loss": 0.6707,
+      "step": 24046
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.525090283973851,
+      "learning_rate": 8.338963734004477e-06,
+      "loss": 0.6969,
+      "step": 24047
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6331694513459458,
+      "learning_rate": 8.338211297715633e-06,
+      "loss": 0.7038,
+      "step": 24048
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4298100252577408,
+      "learning_rate": 8.337458871102245e-06,
+      "loss": 0.6826,
+      "step": 24049
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3768695377810478,
+      "learning_rate": 8.336706454168701e-06,
+      "loss": 0.7479,
+      "step": 24050
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4959626625451516,
+      "learning_rate": 8.335954046919376e-06,
+      "loss": 0.7511,
+      "step": 24051
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3254728011803547,
+      "learning_rate": 8.335201649358658e-06,
+      "loss": 0.7121,
+      "step": 24052
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3053777646638123,
+      "learning_rate": 8.334449261490917e-06,
+      "loss": 0.7372,
+      "step": 24053
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2112047912634858,
+      "learning_rate": 8.333696883320544e-06,
+      "loss": 0.5824,
+      "step": 24054
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4098121341438294,
+      "learning_rate": 8.332944514851916e-06,
+      "loss": 0.7952,
+      "step": 24055
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5749280369739841,
+      "learning_rate": 8.332192156089411e-06,
+      "loss": 0.5055,
+      "step": 24056
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3103428575929637,
+      "learning_rate": 8.33143980703741e-06,
+      "loss": 0.6751,
+      "step": 24057
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5611884955478894,
+      "learning_rate": 8.330687467700295e-06,
+      "loss": 0.6726,
+      "step": 24058
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3336694828215636,
+      "learning_rate": 8.32993513808245e-06,
+      "loss": 0.7461,
+      "step": 24059
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3359502944319241,
+      "learning_rate": 8.329182818188247e-06,
+      "loss": 0.7842,
+      "step": 24060
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2155551433085117,
+      "learning_rate": 8.32843050802207e-06,
+      "loss": 0.7307,
+      "step": 24061
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3666426506632203,
+      "learning_rate": 8.327678207588304e-06,
+      "loss": 0.7421,
+      "step": 24062
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.37369311289937,
+      "learning_rate": 8.326925916891322e-06,
+      "loss": 0.6397,
+      "step": 24063
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6006713592250517,
+      "learning_rate": 8.326173635935509e-06,
+      "loss": 0.5112,
+      "step": 24064
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3803888377405142,
+      "learning_rate": 8.325421364725237e-06,
+      "loss": 0.7247,
+      "step": 24065
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6128925811973867,
+      "learning_rate": 8.324669103264899e-06,
+      "loss": 0.517,
+      "step": 24066
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5710799461109052,
+      "learning_rate": 8.323916851558867e-06,
+      "loss": 0.5346,
+      "step": 24067
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6336257458494314,
+      "learning_rate": 8.32316460961152e-06,
+      "loss": 0.6531,
+      "step": 24068
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4495329331848954,
+      "learning_rate": 8.322412377427242e-06,
+      "loss": 0.6665,
+      "step": 24069
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.691714091453182,
+      "learning_rate": 8.321660155010407e-06,
+      "loss": 0.6871,
+      "step": 24070
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.633823372646785,
+      "learning_rate": 8.320907942365397e-06,
+      "loss": 0.7536,
+      "step": 24071
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3553258003143316,
+      "learning_rate": 8.320155739496593e-06,
+      "loss": 0.5782,
+      "step": 24072
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2598295959782937,
+      "learning_rate": 8.31940354640838e-06,
+      "loss": 0.6459,
+      "step": 24073
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.2772063358196,
+      "learning_rate": 8.318651363105125e-06,
+      "loss": 0.6878,
+      "step": 24074
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5608680661009742,
+      "learning_rate": 8.317899189591216e-06,
+      "loss": 0.8855,
+      "step": 24075
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2754327168568265,
+      "learning_rate": 8.317147025871033e-06,
+      "loss": 0.8384,
+      "step": 24076
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6243714467431924,
+      "learning_rate": 8.31639487194895e-06,
+      "loss": 0.5746,
+      "step": 24077
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3608554597672253,
+      "learning_rate": 8.315642727829347e-06,
+      "loss": 0.6955,
+      "step": 24078
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2827641049652572,
+      "learning_rate": 8.31489059351661e-06,
+      "loss": 0.7792,
+      "step": 24079
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.337843509461385,
+      "learning_rate": 8.314138469015112e-06,
+      "loss": 0.5921,
+      "step": 24080
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4412261916473161,
+      "learning_rate": 8.313386354329234e-06,
+      "loss": 0.7563,
+      "step": 24081
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5125523632929967,
+      "learning_rate": 8.31263424946335e-06,
+      "loss": 0.6593,
+      "step": 24082
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2981796009969109,
+      "learning_rate": 8.31188215442185e-06,
+      "loss": 0.7618,
+      "step": 24083
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3498155790694615,
+      "learning_rate": 8.311130069209103e-06,
+      "loss": 0.6075,
+      "step": 24084
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.534493921927694,
+      "learning_rate": 8.310377993829491e-06,
+      "loss": 0.6969,
+      "step": 24085
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.179043998514147,
+      "learning_rate": 8.309625928287394e-06,
+      "loss": 0.6987,
+      "step": 24086
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4229799652793182,
+      "learning_rate": 8.308873872587193e-06,
+      "loss": 0.7634,
+      "step": 24087
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5047766037765151,
+      "learning_rate": 8.30812182673326e-06,
+      "loss": 0.7167,
+      "step": 24088
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6538513875626786,
+      "learning_rate": 8.307369790729977e-06,
+      "loss": 0.7851,
+      "step": 24089
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4255883714285624,
+      "learning_rate": 8.306617764581728e-06,
+      "loss": 0.825,
+      "step": 24090
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.384155604487934,
+      "learning_rate": 8.305865748292882e-06,
+      "loss": 0.608,
+      "step": 24091
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2543765951120132,
+      "learning_rate": 8.305113741867822e-06,
+      "loss": 0.6469,
+      "step": 24092
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5336171266932976,
+      "learning_rate": 8.304361745310926e-06,
+      "loss": 0.5366,
+      "step": 24093
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3760832186044314,
+      "learning_rate": 8.303609758626574e-06,
+      "loss": 0.7768,
+      "step": 24094
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4174157093906563,
+      "learning_rate": 8.302857781819144e-06,
+      "loss": 0.6208,
+      "step": 24095
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4729419944894877,
+      "learning_rate": 8.302105814893014e-06,
+      "loss": 0.7119,
+      "step": 24096
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5780934237470117,
+      "learning_rate": 8.301353857852561e-06,
+      "loss": 0.5157,
+      "step": 24097
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.016015410638946,
+      "learning_rate": 8.30060191070216e-06,
+      "loss": 0.6861,
+      "step": 24098
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4150234662406915,
+      "learning_rate": 8.299849973446196e-06,
+      "loss": 0.6388,
+      "step": 24099
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1575120894158317,
+      "learning_rate": 8.299098046089042e-06,
+      "loss": 0.7363,
+      "step": 24100
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.552694974789649,
+      "learning_rate": 8.298346128635081e-06,
+      "loss": 0.6547,
+      "step": 24101
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.361259067646113,
+      "learning_rate": 8.297594221088683e-06,
+      "loss": 0.7301,
+      "step": 24102
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6555347003474656,
+      "learning_rate": 8.296842323454234e-06,
+      "loss": 0.7874,
+      "step": 24103
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.2693531621379766,
+      "learning_rate": 8.296090435736107e-06,
+      "loss": 0.5052,
+      "step": 24104
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.361885673313601,
+      "learning_rate": 8.295338557938682e-06,
+      "loss": 0.7383,
+      "step": 24105
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3561391795889097,
+      "learning_rate": 8.294586690066331e-06,
+      "loss": 0.6078,
+      "step": 24106
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.362847152427338,
+      "learning_rate": 8.29383483212344e-06,
+      "loss": 0.6701,
+      "step": 24107
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7388869203760235,
+      "learning_rate": 8.293082984114384e-06,
+      "loss": 0.6655,
+      "step": 24108
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.113216898507092,
+      "learning_rate": 8.292331146043537e-06,
+      "loss": 0.6791,
+      "step": 24109
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3330370249075616,
+      "learning_rate": 8.291579317915278e-06,
+      "loss": 0.479,
+      "step": 24110
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2630187564528406,
+      "learning_rate": 8.290827499733988e-06,
+      "loss": 0.7226,
+      "step": 24111
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6535272600772002,
+      "learning_rate": 8.290075691504041e-06,
+      "loss": 0.7387,
+      "step": 24112
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3885772120841535,
+      "learning_rate": 8.289323893229812e-06,
+      "loss": 0.7005,
+      "step": 24113
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.654405874362035,
+      "learning_rate": 8.28857210491568e-06,
+      "loss": 0.8398,
+      "step": 24114
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5160527882585961,
+      "learning_rate": 8.287820326566028e-06,
+      "loss": 0.7004,
+      "step": 24115
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2751254387746966,
+      "learning_rate": 8.287068558185225e-06,
+      "loss": 0.7849,
+      "step": 24116
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3560211854074569,
+      "learning_rate": 8.286316799777651e-06,
+      "loss": 0.7302,
+      "step": 24117
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6072840007805511,
+      "learning_rate": 8.285565051347688e-06,
+      "loss": 0.5335,
+      "step": 24118
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.00122285436651,
+      "learning_rate": 8.284813312899701e-06,
+      "loss": 0.7652,
+      "step": 24119
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5648552489580771,
+      "learning_rate": 8.284061584438077e-06,
+      "loss": 0.5026,
+      "step": 24120
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.516539307394722,
+      "learning_rate": 8.283309865967187e-06,
+      "loss": 0.5954,
+      "step": 24121
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5469503332343855,
+      "learning_rate": 8.282558157491417e-06,
+      "loss": 0.6767,
+      "step": 24122
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4313420889640525,
+      "learning_rate": 8.281806459015132e-06,
+      "loss": 0.7387,
+      "step": 24123
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5952115526745965,
+      "learning_rate": 8.281054770542713e-06,
+      "loss": 0.5123,
+      "step": 24124
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3728848931827702,
+      "learning_rate": 8.280303092078542e-06,
+      "loss": 0.7691,
+      "step": 24125
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4488413817371182,
+      "learning_rate": 8.279551423626988e-06,
+      "loss": 0.7822,
+      "step": 24126
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3623763918165142,
+      "learning_rate": 8.278799765192427e-06,
+      "loss": 0.8201,
+      "step": 24127
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2658144417838735,
+      "learning_rate": 8.278048116779242e-06,
+      "loss": 0.7093,
+      "step": 24128
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6304054428541104,
+      "learning_rate": 8.27729647839181e-06,
+      "loss": 0.7194,
+      "step": 24129
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7055774413411056,
+      "learning_rate": 8.276544850034498e-06,
+      "loss": 0.6631,
+      "step": 24130
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3681978374209542,
+      "learning_rate": 8.275793231711687e-06,
+      "loss": 0.7792,
+      "step": 24131
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5865376257972621,
+      "learning_rate": 8.275041623427757e-06,
+      "loss": 0.5589,
+      "step": 24132
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4050555905629314,
+      "learning_rate": 8.27429002518708e-06,
+      "loss": 0.7719,
+      "step": 24133
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7067460059399144,
+      "learning_rate": 8.27353843699403e-06,
+      "loss": 0.8556,
+      "step": 24134
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2147608203028437,
+      "learning_rate": 8.272786858852989e-06,
+      "loss": 0.7122,
+      "step": 24135
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.314117596148936,
+      "learning_rate": 8.27203529076833e-06,
+      "loss": 0.8004,
+      "step": 24136
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.351304362279702,
+      "learning_rate": 8.271283732744427e-06,
+      "loss": 0.7274,
+      "step": 24137
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2976024944455442,
+      "learning_rate": 8.270532184785656e-06,
+      "loss": 0.7336,
+      "step": 24138
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4137332466089976,
+      "learning_rate": 8.2697806468964e-06,
+      "loss": 0.5835,
+      "step": 24139
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.434553962105168,
+      "learning_rate": 8.269029119081022e-06,
+      "loss": 0.8225,
+      "step": 24140
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6032196502415708,
+      "learning_rate": 8.268277601343908e-06,
+      "loss": 0.8163,
+      "step": 24141
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3579867239519032,
+      "learning_rate": 8.267526093689426e-06,
+      "loss": 0.6343,
+      "step": 24142
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3981310131501328,
+      "learning_rate": 8.266774596121962e-06,
+      "loss": 0.7345,
+      "step": 24143
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3181882742279367,
+      "learning_rate": 8.26602310864588e-06,
+      "loss": 0.6064,
+      "step": 24144
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6074106003222417,
+      "learning_rate": 8.265271631265564e-06,
+      "loss": 0.5172,
+      "step": 24145
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5956591140648986,
+      "learning_rate": 8.264520163985385e-06,
+      "loss": 0.5341,
+      "step": 24146
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2844934669627017,
+      "learning_rate": 8.263768706809715e-06,
+      "loss": 0.6291,
+      "step": 24147
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4250571423206837,
+      "learning_rate": 8.263017259742938e-06,
+      "loss": 0.7165,
+      "step": 24148
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1882792024757554,
+      "learning_rate": 8.262265822789422e-06,
+      "loss": 0.626,
+      "step": 24149
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3435364179919886,
+      "learning_rate": 8.261514395953548e-06,
+      "loss": 0.7456,
+      "step": 24150
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5242686777099141,
+      "learning_rate": 8.260762979239683e-06,
+      "loss": 0.7932,
+      "step": 24151
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3216832390552071,
+      "learning_rate": 8.260011572652208e-06,
+      "loss": 0.6856,
+      "step": 24152
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5896571616625789,
+      "learning_rate": 8.259260176195499e-06,
+      "loss": 0.5109,
+      "step": 24153
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4110743372550874,
+      "learning_rate": 8.258508789873926e-06,
+      "loss": 0.7358,
+      "step": 24154
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4468335779698185,
+      "learning_rate": 8.257757413691865e-06,
+      "loss": 0.6189,
+      "step": 24155
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.505151699348851,
+      "learning_rate": 8.257006047653693e-06,
+      "loss": 0.7254,
+      "step": 24156
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.318351885997288,
+      "learning_rate": 8.256254691763787e-06,
+      "loss": 0.7756,
+      "step": 24157
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4354720107291161,
+      "learning_rate": 8.255503346026514e-06,
+      "loss": 0.822,
+      "step": 24158
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5736892966682945,
+      "learning_rate": 8.254752010446252e-06,
+      "loss": 0.6501,
+      "step": 24159
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3542026046145426,
+      "learning_rate": 8.254000685027379e-06,
+      "loss": 0.7798,
+      "step": 24160
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3515086847846427,
+      "learning_rate": 8.253249369774268e-06,
+      "loss": 0.7174,
+      "step": 24161
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4427593520145217,
+      "learning_rate": 8.252498064691291e-06,
+      "loss": 0.5948,
+      "step": 24162
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2573692819357614,
+      "learning_rate": 8.251746769782821e-06,
+      "loss": 0.6721,
+      "step": 24163
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3371871108802345,
+      "learning_rate": 8.250995485053238e-06,
+      "loss": 0.6788,
+      "step": 24164
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9272455364457357,
+      "learning_rate": 8.250244210506913e-06,
+      "loss": 0.8387,
+      "step": 24165
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9991915160781777,
+      "learning_rate": 8.249492946148217e-06,
+      "loss": 0.7423,
+      "step": 24166
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5752109432139647,
+      "learning_rate": 8.248741691981528e-06,
+      "loss": 0.7546,
+      "step": 24167
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1338878037098166,
+      "learning_rate": 8.247990448011222e-06,
+      "loss": 0.6332,
+      "step": 24168
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5106759856132894,
+      "learning_rate": 8.24723921424167e-06,
+      "loss": 0.7716,
+      "step": 24169
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2688536213473063,
+      "learning_rate": 8.246487990677242e-06,
+      "loss": 0.7153,
+      "step": 24170
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6170080962110105,
+      "learning_rate": 8.245736777322324e-06,
+      "loss": 0.6582,
+      "step": 24171
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.285211603790978,
+      "learning_rate": 8.244985574181274e-06,
+      "loss": 0.8572,
+      "step": 24172
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6007051344596785,
+      "learning_rate": 8.244234381258477e-06,
+      "loss": 0.6433,
+      "step": 24173
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2985536926532277,
+      "learning_rate": 8.243483198558299e-06,
+      "loss": 0.6854,
+      "step": 24174
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5844096064054205,
+      "learning_rate": 8.242732026085127e-06,
+      "loss": 0.6545,
+      "step": 24175
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.331285376306354,
+      "learning_rate": 8.24198086384332e-06,
+      "loss": 0.6154,
+      "step": 24176
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2888832213712642,
+      "learning_rate": 8.241229711837256e-06,
+      "loss": 0.6622,
+      "step": 24177
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.567349950239424,
+      "learning_rate": 8.240478570071312e-06,
+      "loss": 0.6255,
+      "step": 24178
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3711758821355728,
+      "learning_rate": 8.239727438549859e-06,
+      "loss": 0.6223,
+      "step": 24179
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2140042960688433,
+      "learning_rate": 8.238976317277266e-06,
+      "loss": 0.5954,
+      "step": 24180
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.684561519280234,
+      "learning_rate": 8.238225206257913e-06,
+      "loss": 0.6836,
+      "step": 24181
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3786622976774625,
+      "learning_rate": 8.237474105496172e-06,
+      "loss": 0.8843,
+      "step": 24182
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.529608931664293,
+      "learning_rate": 8.23672301499641e-06,
+      "loss": 0.6242,
+      "step": 24183
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3604335935719776,
+      "learning_rate": 8.23597193476301e-06,
+      "loss": 0.6949,
+      "step": 24184
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4996869899875351,
+      "learning_rate": 8.23522086480034e-06,
+      "loss": 0.7137,
+      "step": 24185
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4571826712274456,
+      "learning_rate": 8.23446980511277e-06,
+      "loss": 0.8439,
+      "step": 24186
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8945116552711105,
+      "learning_rate": 8.233718755704676e-06,
+      "loss": 0.7488,
+      "step": 24187
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4955365836343768,
+      "learning_rate": 8.232967716580431e-06,
+      "loss": 0.7571,
+      "step": 24188
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.520761405115797,
+      "learning_rate": 8.23221668774441e-06,
+      "loss": 0.6919,
+      "step": 24189
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3872887581607967,
+      "learning_rate": 8.23146566920098e-06,
+      "loss": 0.7645,
+      "step": 24190
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.399286078175421,
+      "learning_rate": 8.230714660954516e-06,
+      "loss": 0.6671,
+      "step": 24191
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.315722644186618,
+      "learning_rate": 8.229963663009394e-06,
+      "loss": 0.7025,
+      "step": 24192
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2075588626130673,
+      "learning_rate": 8.229212675369984e-06,
+      "loss": 0.6744,
+      "step": 24193
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9857554776213608,
+      "learning_rate": 8.228461698040657e-06,
+      "loss": 0.6382,
+      "step": 24194
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2935272641636353,
+      "learning_rate": 8.227710731025785e-06,
+      "loss": 0.7755,
+      "step": 24195
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4327021393148638,
+      "learning_rate": 8.226959774329747e-06,
+      "loss": 0.7112,
+      "step": 24196
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.35828759883741,
+      "learning_rate": 8.226208827956908e-06,
+      "loss": 0.6337,
+      "step": 24197
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6044801671265059,
+      "learning_rate": 8.225457891911643e-06,
+      "loss": 0.5439,
+      "step": 24198
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1818571204187198,
+      "learning_rate": 8.224706966198325e-06,
+      "loss": 0.627,
+      "step": 24199
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.399928129984948,
+      "learning_rate": 8.223956050821322e-06,
+      "loss": 0.7022,
+      "step": 24200
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.32455599188997,
+      "learning_rate": 8.223205145785012e-06,
+      "loss": 0.7386,
+      "step": 24201
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.282745458310989,
+      "learning_rate": 8.222454251093763e-06,
+      "loss": 0.7351,
+      "step": 24202
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.389097330611913,
+      "learning_rate": 8.22170336675195e-06,
+      "loss": 0.7534,
+      "step": 24203
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4581845701987295,
+      "learning_rate": 8.220952492763938e-06,
+      "loss": 0.6766,
+      "step": 24204
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.553284522540558,
+      "learning_rate": 8.220201629134109e-06,
+      "loss": 0.6933,
+      "step": 24205
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4323543456356098,
+      "learning_rate": 8.219450775866828e-06,
+      "loss": 0.698,
+      "step": 24206
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3027257293712042,
+      "learning_rate": 8.218699932966468e-06,
+      "loss": 0.673,
+      "step": 24207
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2520563280717114,
+      "learning_rate": 8.2179491004374e-06,
+      "loss": 0.6766,
+      "step": 24208
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.861871992746311,
+      "learning_rate": 8.217198278283997e-06,
+      "loss": 0.5897,
+      "step": 24209
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.733047617272432,
+      "learning_rate": 8.216447466510633e-06,
+      "loss": 0.7402,
+      "step": 24210
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3440504582931074,
+      "learning_rate": 8.215696665121674e-06,
+      "loss": 0.5728,
+      "step": 24211
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.368281324922924,
+      "learning_rate": 8.214945874121491e-06,
+      "loss": 0.694,
+      "step": 24212
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3461940597782853,
+      "learning_rate": 8.214195093514463e-06,
+      "loss": 0.6424,
+      "step": 24213
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3733798926627692,
+      "learning_rate": 8.213444323304955e-06,
+      "loss": 0.6642,
+      "step": 24214
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3216626148904331,
+      "learning_rate": 8.21269356349734e-06,
+      "loss": 0.7026,
+      "step": 24215
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4395549030656085,
+      "learning_rate": 8.211942814095986e-06,
+      "loss": 0.6233,
+      "step": 24216
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3528909962776843,
+      "learning_rate": 8.211192075105271e-06,
+      "loss": 0.7502,
+      "step": 24217
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4220454078600906,
+      "learning_rate": 8.21044134652956e-06,
+      "loss": 0.6445,
+      "step": 24218
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2170328815904838,
+      "learning_rate": 8.209690628373225e-06,
+      "loss": 0.5903,
+      "step": 24219
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.570624299343568,
+      "learning_rate": 8.208939920640644e-06,
+      "loss": 0.7314,
+      "step": 24220
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3699905055657697,
+      "learning_rate": 8.208189223336175e-06,
+      "loss": 0.7079,
+      "step": 24221
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3445084029390288,
+      "learning_rate": 8.207438536464197e-06,
+      "loss": 0.6654,
+      "step": 24222
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1730454457936772,
+      "learning_rate": 8.206687860029079e-06,
+      "loss": 0.6169,
+      "step": 24223
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.726975880002813,
+      "learning_rate": 8.205937194035197e-06,
+      "loss": 0.6714,
+      "step": 24224
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2424159352879642,
+      "learning_rate": 8.205186538486911e-06,
+      "loss": 0.6497,
+      "step": 24225
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.588883488798752,
+      "learning_rate": 8.204435893388599e-06,
+      "loss": 0.4876,
+      "step": 24226
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7384775173332008,
+      "learning_rate": 8.203685258744633e-06,
+      "loss": 0.7975,
+      "step": 24227
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3785326435909522,
+      "learning_rate": 8.202934634559375e-06,
+      "loss": 0.5706,
+      "step": 24228
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3986765735759663,
+      "learning_rate": 8.202184020837202e-06,
+      "loss": 0.6472,
+      "step": 24229
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3871472193954726,
+      "learning_rate": 8.201433417582482e-06,
+      "loss": 0.8246,
+      "step": 24230
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.341820960786525,
+      "learning_rate": 8.200682824799591e-06,
+      "loss": 0.6244,
+      "step": 24231
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5129272366819688,
+      "learning_rate": 8.199932242492887e-06,
+      "loss": 0.6003,
+      "step": 24232
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6793400288069926,
+      "learning_rate": 8.199181670666751e-06,
+      "loss": 0.6096,
+      "step": 24233
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3511070202935733,
+      "learning_rate": 8.198431109325552e-06,
+      "loss": 0.7201,
+      "step": 24234
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9345709881821669,
+      "learning_rate": 8.197680558473655e-06,
+      "loss": 0.6882,
+      "step": 24235
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3176414147647686,
+      "learning_rate": 8.196930018115429e-06,
+      "loss": 0.7214,
+      "step": 24236
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4835174806016695,
+      "learning_rate": 8.196179488255251e-06,
+      "loss": 0.7942,
+      "step": 24237
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.30982446575604,
+      "learning_rate": 8.19542896889749e-06,
+      "loss": 0.5332,
+      "step": 24238
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.5126011356336635,
+      "learning_rate": 8.194678460046509e-06,
+      "loss": 0.7588,
+      "step": 24239
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.413871077316081,
+      "learning_rate": 8.19392796170668e-06,
+      "loss": 0.7262,
+      "step": 24240
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4049952285080578,
+      "learning_rate": 8.19317747388238e-06,
+      "loss": 0.7464,
+      "step": 24241
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4633666253445687,
+      "learning_rate": 8.192426996577967e-06,
+      "loss": 0.6726,
+      "step": 24242
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.404057571971256,
+      "learning_rate": 8.191676529797818e-06,
+      "loss": 0.7388,
+      "step": 24243
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3350676139754305,
+      "learning_rate": 8.1909260735463e-06,
+      "loss": 0.6795,
+      "step": 24244
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5679391766843078,
+      "learning_rate": 8.190175627827785e-06,
+      "loss": 0.6606,
+      "step": 24245
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4146060280394026,
+      "learning_rate": 8.18942519264664e-06,
+      "loss": 0.8442,
+      "step": 24246
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4360141980723804,
+      "learning_rate": 8.188674768007234e-06,
+      "loss": 0.6633,
+      "step": 24247
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.471372782873973,
+      "learning_rate": 8.18792435391394e-06,
+      "loss": 0.4911,
+      "step": 24248
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5805646113242462,
+      "learning_rate": 8.18717395037112e-06,
+      "loss": 0.7417,
+      "step": 24249
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.757862726038713,
+      "learning_rate": 8.18642355738315e-06,
+      "loss": 0.7658,
+      "step": 24250
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.642663728602688,
+      "learning_rate": 8.185673174954394e-06,
+      "loss": 0.7286,
+      "step": 24251
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3059210337055618,
+      "learning_rate": 8.184922803089226e-06,
+      "loss": 0.8295,
+      "step": 24252
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6165582931975006,
+      "learning_rate": 8.184172441792008e-06,
+      "loss": 0.5239,
+      "step": 24253
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3460667110927613,
+      "learning_rate": 8.183422091067115e-06,
+      "loss": 0.759,
+      "step": 24254
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3801758980034873,
+      "learning_rate": 8.182671750918915e-06,
+      "loss": 0.72,
+      "step": 24255
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.47538332231301,
+      "learning_rate": 8.181921421351775e-06,
+      "loss": 0.6477,
+      "step": 24256
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4779215646111172,
+      "learning_rate": 8.181171102370062e-06,
+      "loss": 0.6961,
+      "step": 24257
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3359932602014883,
+      "learning_rate": 8.180420793978148e-06,
+      "loss": 0.7477,
+      "step": 24258
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1779638010795932,
+      "learning_rate": 8.179670496180401e-06,
+      "loss": 0.6375,
+      "step": 24259
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7219782056072337,
+      "learning_rate": 8.178920208981189e-06,
+      "loss": 0.7734,
+      "step": 24260
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3511030537024817,
+      "learning_rate": 8.178169932384877e-06,
+      "loss": 0.658,
+      "step": 24261
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5747073657937398,
+      "learning_rate": 8.177419666395839e-06,
+      "loss": 0.7601,
+      "step": 24262
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.392410721960174,
+      "learning_rate": 8.176669411018441e-06,
+      "loss": 0.6233,
+      "step": 24263
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5961792101167134,
+      "learning_rate": 8.17591916625705e-06,
+      "loss": 0.4816,
+      "step": 24264
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.436089482788888,
+      "learning_rate": 8.175168932116033e-06,
+      "loss": 0.7921,
+      "step": 24265
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4139993072707693,
+      "learning_rate": 8.174418708599764e-06,
+      "loss": 0.6134,
+      "step": 24266
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3254853409574727,
+      "learning_rate": 8.173668495712605e-06,
+      "loss": 0.8061,
+      "step": 24267
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5024180131334843,
+      "learning_rate": 8.172918293458925e-06,
+      "loss": 0.6377,
+      "step": 24268
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3740959245265298,
+      "learning_rate": 8.1721681018431e-06,
+      "loss": 0.655,
+      "step": 24269
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4771280747280164,
+      "learning_rate": 8.171417920869483e-06,
+      "loss": 0.705,
+      "step": 24270
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3987091063712374,
+      "learning_rate": 8.170667750542453e-06,
+      "loss": 0.5447,
+      "step": 24271
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4798022683826006,
+      "learning_rate": 8.169917590866372e-06,
+      "loss": 0.7158,
+      "step": 24272
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4923763321011974,
+      "learning_rate": 8.169167441845617e-06,
+      "loss": 0.6982,
+      "step": 24273
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.359430068266756,
+      "learning_rate": 8.168417303484542e-06,
+      "loss": 0.6926,
+      "step": 24274
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3666417043068213,
+      "learning_rate": 8.167667175787524e-06,
+      "loss": 0.7373,
+      "step": 24275
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2333612414463222,
+      "learning_rate": 8.16691705875893e-06,
+      "loss": 0.6871,
+      "step": 24276
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3416921226345013,
+      "learning_rate": 8.16616695240312e-06,
+      "loss": 0.6795,
+      "step": 24277
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5958633806285414,
+      "learning_rate": 8.165416856724468e-06,
+      "loss": 0.7128,
+      "step": 24278
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.501294274052343,
+      "learning_rate": 8.164666771727343e-06,
+      "loss": 0.7177,
+      "step": 24279
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4241540141093822,
+      "learning_rate": 8.163916697416109e-06,
+      "loss": 0.7619,
+      "step": 24280
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3843511636164767,
+      "learning_rate": 8.163166633795129e-06,
+      "loss": 0.7376,
+      "step": 24281
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4137408026194416,
+      "learning_rate": 8.162416580868777e-06,
+      "loss": 0.6515,
+      "step": 24282
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4607761715097856,
+      "learning_rate": 8.161666538641418e-06,
+      "loss": 0.7049,
+      "step": 24283
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1886415274084199,
+      "learning_rate": 8.160916507117418e-06,
+      "loss": 0.7127,
+      "step": 24284
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4072116452107457,
+      "learning_rate": 8.160166486301141e-06,
+      "loss": 0.7147,
+      "step": 24285
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8394699859134975,
+      "learning_rate": 8.159416476196962e-06,
+      "loss": 0.6434,
+      "step": 24286
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2548995274328567,
+      "learning_rate": 8.158666476809243e-06,
+      "loss": 0.6393,
+      "step": 24287
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2587669946232156,
+      "learning_rate": 8.157916488142349e-06,
+      "loss": 0.6879,
+      "step": 24288
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.584380973213623,
+      "learning_rate": 8.157166510200647e-06,
+      "loss": 0.7709,
+      "step": 24289
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3276493478706959,
+      "learning_rate": 8.156416542988505e-06,
+      "loss": 0.697,
+      "step": 24290
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2974992866711685,
+      "learning_rate": 8.155666586510293e-06,
+      "loss": 0.6434,
+      "step": 24291
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6069277938689133,
+      "learning_rate": 8.154916640770373e-06,
+      "loss": 0.7428,
+      "step": 24292
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4533712079324859,
+      "learning_rate": 8.15416670577311e-06,
+      "loss": 0.7484,
+      "step": 24293
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3058217455811285,
+      "learning_rate": 8.153416781522876e-06,
+      "loss": 0.6642,
+      "step": 24294
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.383785536546361,
+      "learning_rate": 8.152666868024032e-06,
+      "loss": 0.6865,
+      "step": 24295
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6630085584410206,
+      "learning_rate": 8.151916965280947e-06,
+      "loss": 0.7479,
+      "step": 24296
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2605784989295192,
+      "learning_rate": 8.151167073297983e-06,
+      "loss": 0.7105,
+      "step": 24297
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3460732341848114,
+      "learning_rate": 8.150417192079516e-06,
+      "loss": 0.7435,
+      "step": 24298
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.467686830774089,
+      "learning_rate": 8.149667321629902e-06,
+      "loss": 0.7943,
+      "step": 24299
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5425624014413541,
+      "learning_rate": 8.148917461953511e-06,
+      "loss": 0.7424,
+      "step": 24300
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4897261623880493,
+      "learning_rate": 8.148167613054711e-06,
+      "loss": 0.6565,
+      "step": 24301
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4587207996598053,
+      "learning_rate": 8.14741777493786e-06,
+      "loss": 0.7213,
+      "step": 24302
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3700204920991397,
+      "learning_rate": 8.146667947607335e-06,
+      "loss": 0.8623,
+      "step": 24303
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3627903429503756,
+      "learning_rate": 8.14591813106749e-06,
+      "loss": 0.7501,
+      "step": 24304
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3275916583426102,
+      "learning_rate": 8.145168325322704e-06,
+      "loss": 0.785,
+      "step": 24305
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3440227919854224,
+      "learning_rate": 8.144418530377329e-06,
+      "loss": 0.664,
+      "step": 24306
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5343307763645524,
+      "learning_rate": 8.14366874623574e-06,
+      "loss": 0.7167,
+      "step": 24307
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2670636888011473,
+      "learning_rate": 8.142918972902301e-06,
+      "loss": 0.7101,
+      "step": 24308
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6923640983867443,
+      "learning_rate": 8.142169210381373e-06,
+      "loss": 0.6561,
+      "step": 24309
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3658201522770288,
+      "learning_rate": 8.141419458677321e-06,
+      "loss": 0.7008,
+      "step": 24310
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.324102818792324,
+      "learning_rate": 8.140669717794515e-06,
+      "loss": 0.7274,
+      "step": 24311
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5867063887863933,
+      "learning_rate": 8.139919987737324e-06,
+      "loss": 0.7862,
+      "step": 24312
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4345815248039853,
+      "learning_rate": 8.139170268510104e-06,
+      "loss": 0.7273,
+      "step": 24313
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.410192852142937,
+      "learning_rate": 8.13842056011722e-06,
+      "loss": 0.8807,
+      "step": 24314
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3182457518236284,
+      "learning_rate": 8.137670862563047e-06,
+      "loss": 0.725,
+      "step": 24315
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4072316845502661,
+      "learning_rate": 8.136921175851942e-06,
+      "loss": 0.5986,
+      "step": 24316
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2517984127125323,
+      "learning_rate": 8.136171499988267e-06,
+      "loss": 0.6656,
+      "step": 24317
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5254875916155952,
+      "learning_rate": 8.135421834976396e-06,
+      "loss": 0.6013,
+      "step": 24318
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1042342213804304,
+      "learning_rate": 8.134672180820692e-06,
+      "loss": 0.7002,
+      "step": 24319
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.307587393607504,
+      "learning_rate": 8.133922537525515e-06,
+      "loss": 0.7598,
+      "step": 24320
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4146781626766323,
+      "learning_rate": 8.133172905095227e-06,
+      "loss": 0.6205,
+      "step": 24321
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.39560645933038,
+      "learning_rate": 8.132423283534206e-06,
+      "loss": 0.7731,
+      "step": 24322
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3157096511238004,
+      "learning_rate": 8.1316736728468e-06,
+      "loss": 0.8472,
+      "step": 24323
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3427092885386578,
+      "learning_rate": 8.130924073037385e-06,
+      "loss": 0.6336,
+      "step": 24324
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.312664057585049,
+      "learning_rate": 8.13017448411032e-06,
+      "loss": 0.7976,
+      "step": 24325
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2535482616614118,
+      "learning_rate": 8.129424906069974e-06,
+      "loss": 0.7295,
+      "step": 24326
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2924662042349635,
+      "learning_rate": 8.128675338920707e-06,
+      "loss": 0.7104,
+      "step": 24327
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4140967850672288,
+      "learning_rate": 8.127925782666884e-06,
+      "loss": 0.7849,
+      "step": 24328
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.258572263934021,
+      "learning_rate": 8.127176237312872e-06,
+      "loss": 0.7447,
+      "step": 24329
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.708904712370025,
+      "learning_rate": 8.126426702863029e-06,
+      "loss": 0.617,
+      "step": 24330
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3854393060768477,
+      "learning_rate": 8.125677179321725e-06,
+      "loss": 0.7377,
+      "step": 24331
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.373033270935511,
+      "learning_rate": 8.124927666693324e-06,
+      "loss": 0.7013,
+      "step": 24332
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.388308970069781,
+      "learning_rate": 8.124178164982186e-06,
+      "loss": 0.6522,
+      "step": 24333
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5916826861400188,
+      "learning_rate": 8.123428674192675e-06,
+      "loss": 0.6324,
+      "step": 24334
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2382849003266714,
+      "learning_rate": 8.122679194329157e-06,
+      "loss": 0.7189,
+      "step": 24335
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3204742537745495,
+      "learning_rate": 8.121929725396e-06,
+      "loss": 0.8097,
+      "step": 24336
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.736830976927194,
+      "learning_rate": 8.121180267397556e-06,
+      "loss": 0.6461,
+      "step": 24337
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.828830417946878,
+      "learning_rate": 8.120430820338195e-06,
+      "loss": 0.7091,
+      "step": 24338
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3420931302059143,
+      "learning_rate": 8.119681384222284e-06,
+      "loss": 0.7489,
+      "step": 24339
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6788850509403306,
+      "learning_rate": 8.118931959054185e-06,
+      "loss": 0.5991,
+      "step": 24340
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0746926067741187,
+      "learning_rate": 8.118182544838258e-06,
+      "loss": 0.8006,
+      "step": 24341
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6556288560723549,
+      "learning_rate": 8.117433141578865e-06,
+      "loss": 0.554,
+      "step": 24342
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5880825840580983,
+      "learning_rate": 8.116683749280376e-06,
+      "loss": 0.5094,
+      "step": 24343
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3269464053146536,
+      "learning_rate": 8.115934367947151e-06,
+      "loss": 0.6794,
+      "step": 24344
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5918239170635898,
+      "learning_rate": 8.11518499758355e-06,
+      "loss": 0.5271,
+      "step": 24345
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5574334259947201,
+      "learning_rate": 8.114435638193937e-06,
+      "loss": 0.4951,
+      "step": 24346
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4753264105505972,
+      "learning_rate": 8.113686289782682e-06,
+      "loss": 0.7531,
+      "step": 24347
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3764273830742193,
+      "learning_rate": 8.112936952354139e-06,
+      "loss": 0.72,
+      "step": 24348
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4580370367576874,
+      "learning_rate": 8.112187625912675e-06,
+      "loss": 0.5655,
+      "step": 24349
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3340219004449614,
+      "learning_rate": 8.111438310462656e-06,
+      "loss": 0.6836,
+      "step": 24350
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.400360443682074,
+      "learning_rate": 8.110689006008435e-06,
+      "loss": 0.8091,
+      "step": 24351
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5072214710127478,
+      "learning_rate": 8.109939712554383e-06,
+      "loss": 0.6797,
+      "step": 24352
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4263169212260765,
+      "learning_rate": 8.109190430104858e-06,
+      "loss": 0.7074,
+      "step": 24353
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5766361061259924,
+      "learning_rate": 8.108441158664234e-06,
+      "loss": 0.5148,
+      "step": 24354
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2956944327991715,
+      "learning_rate": 8.107691898236855e-06,
+      "loss": 0.7799,
+      "step": 24355
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3808258224043537,
+      "learning_rate": 8.106942648827097e-06,
+      "loss": 0.701,
+      "step": 24356
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3207481009478088,
+      "learning_rate": 8.106193410439318e-06,
+      "loss": 0.6489,
+      "step": 24357
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5071025064434167,
+      "learning_rate": 8.10544418307788e-06,
+      "loss": 0.7151,
+      "step": 24358
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6825098776078182,
+      "learning_rate": 8.104694966747143e-06,
+      "loss": 0.6263,
+      "step": 24359
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5654686058653102,
+      "learning_rate": 8.103945761451474e-06,
+      "loss": 0.5928,
+      "step": 24360
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3508202097565096,
+      "learning_rate": 8.103196567195237e-06,
+      "loss": 0.8128,
+      "step": 24361
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.035519896860238,
+      "learning_rate": 8.102447383982785e-06,
+      "loss": 0.7817,
+      "step": 24362
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3356921901494687,
+      "learning_rate": 8.101698211818485e-06,
+      "loss": 0.6268,
+      "step": 24363
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.477262131824795,
+      "learning_rate": 8.100949050706704e-06,
+      "loss": 0.6781,
+      "step": 24364
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3852653361865985,
+      "learning_rate": 8.100199900651794e-06,
+      "loss": 0.683,
+      "step": 24365
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5502380085762304,
+      "learning_rate": 8.099450761658122e-06,
+      "loss": 0.6407,
+      "step": 24366
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.799991281438821,
+      "learning_rate": 8.098701633730051e-06,
+      "loss": 0.7546,
+      "step": 24367
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3862963350100403,
+      "learning_rate": 8.097952516871944e-06,
+      "loss": 0.7456,
+      "step": 24368
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.209516371027269,
+      "learning_rate": 8.097203411088155e-06,
+      "loss": 0.6912,
+      "step": 24369
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2890849271136466,
+      "learning_rate": 8.096454316383049e-06,
+      "loss": 0.7628,
+      "step": 24370
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5955850687367559,
+      "learning_rate": 8.095705232760994e-06,
+      "loss": 0.7895,
+      "step": 24371
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3902906350475845,
+      "learning_rate": 8.094956160226342e-06,
+      "loss": 0.6928,
+      "step": 24372
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.46052197745089,
+      "learning_rate": 8.09420709878346e-06,
+      "loss": 0.6393,
+      "step": 24373
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.235762703665289,
+      "learning_rate": 8.093458048436705e-06,
+      "loss": 0.658,
+      "step": 24374
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5858317253328824,
+      "learning_rate": 8.092709009190446e-06,
+      "loss": 0.7914,
+      "step": 24375
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4082910077667212,
+      "learning_rate": 8.091959981049034e-06,
+      "loss": 0.5757,
+      "step": 24376
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7921898257332756,
+      "learning_rate": 8.091210964016837e-06,
+      "loss": 0.7354,
+      "step": 24377
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2770111926971628,
+      "learning_rate": 8.090461958098216e-06,
+      "loss": 0.6209,
+      "step": 24378
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6588800776897845,
+      "learning_rate": 8.089712963297527e-06,
+      "loss": 0.7275,
+      "step": 24379
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6215912110232541,
+      "learning_rate": 8.088963979619135e-06,
+      "loss": 0.5385,
+      "step": 24380
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.466540412464486,
+      "learning_rate": 8.0882150070674e-06,
+      "loss": 0.6485,
+      "step": 24381
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.331236889905196,
+      "learning_rate": 8.087466045646684e-06,
+      "loss": 0.6712,
+      "step": 24382
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5890757217833654,
+      "learning_rate": 8.086717095361342e-06,
+      "loss": 0.4894,
+      "step": 24383
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.281664204648067,
+      "learning_rate": 8.085968156215741e-06,
+      "loss": 0.7368,
+      "step": 24384
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4380754662776818,
+      "learning_rate": 8.085219228214241e-06,
+      "loss": 0.6742,
+      "step": 24385
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5997752358274173,
+      "learning_rate": 8.0844703113612e-06,
+      "loss": 0.4999,
+      "step": 24386
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3282332402428614,
+      "learning_rate": 8.083721405660978e-06,
+      "loss": 0.7447,
+      "step": 24387
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.561878684647522,
+      "learning_rate": 8.082972511117936e-06,
+      "loss": 0.7136,
+      "step": 24388
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3522400326956114,
+      "learning_rate": 8.082223627736439e-06,
+      "loss": 0.5906,
+      "step": 24389
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6090950648979762,
+      "learning_rate": 8.08147475552084e-06,
+      "loss": 0.5245,
+      "step": 24390
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.403432781973298,
+      "learning_rate": 8.0807258944755e-06,
+      "loss": 0.7879,
+      "step": 24391
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4212411198121968,
+      "learning_rate": 8.079977044604788e-06,
+      "loss": 0.7114,
+      "step": 24392
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4361600393214244,
+      "learning_rate": 8.079228205913053e-06,
+      "loss": 0.5679,
+      "step": 24393
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.314238649879781,
+      "learning_rate": 8.07847937840466e-06,
+      "loss": 0.6815,
+      "step": 24394
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5493223804347573,
+      "learning_rate": 8.077730562083967e-06,
+      "loss": 0.6978,
+      "step": 24395
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5113481937856983,
+      "learning_rate": 8.07698175695534e-06,
+      "loss": 0.7365,
+      "step": 24396
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7603033889255062,
+      "learning_rate": 8.07623296302313e-06,
+      "loss": 0.6381,
+      "step": 24397
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4805010634641351,
+      "learning_rate": 8.075484180291702e-06,
+      "loss": 0.745,
+      "step": 24398
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3299186440487052,
+      "learning_rate": 8.074735408765417e-06,
+      "loss": 0.6477,
+      "step": 24399
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3522925161867838,
+      "learning_rate": 8.073986648448627e-06,
+      "loss": 0.6961,
+      "step": 24400
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6101678420126028,
+      "learning_rate": 8.073237899345699e-06,
+      "loss": 0.5186,
+      "step": 24401
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4206369397202276,
+      "learning_rate": 8.072489161460986e-06,
+      "loss": 0.7215,
+      "step": 24402
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3830417737578429,
+      "learning_rate": 8.07174043479886e-06,
+      "loss": 0.8579,
+      "step": 24403
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5525025990229666,
+      "learning_rate": 8.070991719363664e-06,
+      "loss": 0.7827,
+      "step": 24404
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4583633844660158,
+      "learning_rate": 8.070243015159767e-06,
+      "loss": 0.7448,
+      "step": 24405
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1897435790125552,
+      "learning_rate": 8.069494322191526e-06,
+      "loss": 0.6412,
+      "step": 24406
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2801900409807503,
+      "learning_rate": 8.068745640463305e-06,
+      "loss": 0.728,
+      "step": 24407
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5620947628052955,
+      "learning_rate": 8.067996969979451e-06,
+      "loss": 0.6681,
+      "step": 24408
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7312621471177063,
+      "learning_rate": 8.067248310744334e-06,
+      "loss": 0.79,
+      "step": 24409
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5849091096060375,
+      "learning_rate": 8.066499662762312e-06,
+      "loss": 0.7372,
+      "step": 24410
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.617431138452416,
+      "learning_rate": 8.065751026037738e-06,
+      "loss": 0.8292,
+      "step": 24411
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5703278469370169,
+      "learning_rate": 8.065002400574971e-06,
+      "loss": 0.5369,
+      "step": 24412
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.206779849681071,
+      "learning_rate": 8.064253786378375e-06,
+      "loss": 0.7607,
+      "step": 24413
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2005319072908416,
+      "learning_rate": 8.063505183452309e-06,
+      "loss": 0.734,
+      "step": 24414
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.597515360640098,
+      "learning_rate": 8.062756591801124e-06,
+      "loss": 0.4976,
+      "step": 24415
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6564184951131699,
+      "learning_rate": 8.062008011429187e-06,
+      "loss": 0.5349,
+      "step": 24416
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.596094907893313,
+      "learning_rate": 8.061259442340852e-06,
+      "loss": 0.5122,
+      "step": 24417
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2964223772378363,
+      "learning_rate": 8.06051088454048e-06,
+      "loss": 0.7364,
+      "step": 24418
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.27800751119969,
+      "learning_rate": 8.059762338032423e-06,
+      "loss": 0.7625,
+      "step": 24419
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4943764491131049,
+      "learning_rate": 8.059013802821047e-06,
+      "loss": 0.6991,
+      "step": 24420
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.777205162471382,
+      "learning_rate": 8.05826527891071e-06,
+      "loss": 0.8175,
+      "step": 24421
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5028042516491744,
+      "learning_rate": 8.057516766305765e-06,
+      "loss": 0.6419,
+      "step": 24422
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7240261476549603,
+      "learning_rate": 8.056768265010568e-06,
+      "loss": 0.7419,
+      "step": 24423
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4680596187517179,
+      "learning_rate": 8.056019775029488e-06,
+      "loss": 0.6292,
+      "step": 24424
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3405803626470343,
+      "learning_rate": 8.055271296366874e-06,
+      "loss": 0.7197,
+      "step": 24425
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.426454741372481,
+      "learning_rate": 8.054522829027084e-06,
+      "loss": 0.7506,
+      "step": 24426
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4046778701736173,
+      "learning_rate": 8.053774373014478e-06,
+      "loss": 0.7875,
+      "step": 24427
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.381124236296779,
+      "learning_rate": 8.053025928333418e-06,
+      "loss": 0.6413,
+      "step": 24428
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.368890226816033,
+      "learning_rate": 8.052277494988254e-06,
+      "loss": 0.6214,
+      "step": 24429
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9047722676225416,
+      "learning_rate": 8.051529072983349e-06,
+      "loss": 0.6951,
+      "step": 24430
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1788973152545779,
+      "learning_rate": 8.050780662323059e-06,
+      "loss": 0.7078,
+      "step": 24431
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5319154157556,
+      "learning_rate": 8.050032263011738e-06,
+      "loss": 0.671,
+      "step": 24432
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4171530911877523,
+      "learning_rate": 8.049283875053747e-06,
+      "loss": 0.8094,
+      "step": 24433
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4130728206070224,
+      "learning_rate": 8.048535498453445e-06,
+      "loss": 0.6388,
+      "step": 24434
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3276542689026973,
+      "learning_rate": 8.047787133215188e-06,
+      "loss": 0.6861,
+      "step": 24435
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4656346513244303,
+      "learning_rate": 8.047038779343327e-06,
+      "loss": 0.6422,
+      "step": 24436
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.51929894584587,
+      "learning_rate": 8.046290436842229e-06,
+      "loss": 0.6586,
+      "step": 24437
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4078958429228263,
+      "learning_rate": 8.045542105716247e-06,
+      "loss": 0.7435,
+      "step": 24438
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4174633977475894,
+      "learning_rate": 8.044793785969736e-06,
+      "loss": 0.8069,
+      "step": 24439
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2872698873660018,
+      "learning_rate": 8.044045477607053e-06,
+      "loss": 0.7686,
+      "step": 24440
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4897772452712559,
+      "learning_rate": 8.04329718063256e-06,
+      "loss": 0.6339,
+      "step": 24441
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.264097340280342,
+      "learning_rate": 8.042548895050609e-06,
+      "loss": 0.71,
+      "step": 24442
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7805290433884462,
+      "learning_rate": 8.041800620865558e-06,
+      "loss": 0.7224,
+      "step": 24443
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3834691644640802,
+      "learning_rate": 8.041052358081761e-06,
+      "loss": 0.7405,
+      "step": 24444
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4531204081518547,
+      "learning_rate": 8.040304106703582e-06,
+      "loss": 0.6469,
+      "step": 24445
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0820989826227074,
+      "learning_rate": 8.039555866735371e-06,
+      "loss": 0.6439,
+      "step": 24446
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4049249613415895,
+      "learning_rate": 8.038807638181486e-06,
+      "loss": 0.7349,
+      "step": 24447
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2340917241959606,
+      "learning_rate": 8.038059421046281e-06,
+      "loss": 0.7798,
+      "step": 24448
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5301060703440834,
+      "learning_rate": 8.037311215334122e-06,
+      "loss": 0.6531,
+      "step": 24449
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4582235590439072,
+      "learning_rate": 8.036563021049354e-06,
+      "loss": 0.7144,
+      "step": 24450
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.451749671125813,
+      "learning_rate": 8.035814838196335e-06,
+      "loss": 0.6712,
+      "step": 24451
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3834898619695892,
+      "learning_rate": 8.035066666779432e-06,
+      "loss": 0.8419,
+      "step": 24452
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5151997844052418,
+      "learning_rate": 8.034318506802987e-06,
+      "loss": 0.752,
+      "step": 24453
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5241801396407355,
+      "learning_rate": 8.033570358271362e-06,
+      "loss": 0.7583,
+      "step": 24454
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.283288314129357,
+      "learning_rate": 8.032822221188913e-06,
+      "loss": 0.7792,
+      "step": 24455
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4412409054599542,
+      "learning_rate": 8.03207409556e-06,
+      "loss": 0.6607,
+      "step": 24456
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6128090422001435,
+      "learning_rate": 8.03132598138897e-06,
+      "loss": 0.6507,
+      "step": 24457
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.368488381018696,
+      "learning_rate": 8.030577878680185e-06,
+      "loss": 0.55,
+      "step": 24458
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4902295550622149,
+      "learning_rate": 8.029829787438001e-06,
+      "loss": 0.647,
+      "step": 24459
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6483205805572094,
+      "learning_rate": 8.02908170766677e-06,
+      "loss": 0.8543,
+      "step": 24460
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.482165626718497,
+      "learning_rate": 8.028333639370846e-06,
+      "loss": 0.8016,
+      "step": 24461
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3348885142787816,
+      "learning_rate": 8.027585582554592e-06,
+      "loss": 0.6871,
+      "step": 24462
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2864219130538654,
+      "learning_rate": 8.02683753722236e-06,
+      "loss": 0.6969,
+      "step": 24463
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2350670941973139,
+      "learning_rate": 8.0260895033785e-06,
+      "loss": 0.7316,
+      "step": 24464
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7937441068737685,
+      "learning_rate": 8.025341481027374e-06,
+      "loss": 0.7615,
+      "step": 24465
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4292496732421236,
+      "learning_rate": 8.024593470173337e-06,
+      "loss": 0.6811,
+      "step": 24466
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3595754376771945,
+      "learning_rate": 8.023845470820742e-06,
+      "loss": 0.803,
+      "step": 24467
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3197607018020894,
+      "learning_rate": 8.02309748297394e-06,
+      "loss": 0.6463,
+      "step": 24468
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2709206475899688,
+      "learning_rate": 8.022349506637294e-06,
+      "loss": 0.7357,
+      "step": 24469
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6314704471073493,
+      "learning_rate": 8.021601541815156e-06,
+      "loss": 0.6344,
+      "step": 24470
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6677912759816687,
+      "learning_rate": 8.02085358851188e-06,
+      "loss": 0.8284,
+      "step": 24471
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.475188172931721,
+      "learning_rate": 8.020105646731815e-06,
+      "loss": 0.7344,
+      "step": 24472
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2644734173609347,
+      "learning_rate": 8.01935771647933e-06,
+      "loss": 0.5856,
+      "step": 24473
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3868983278387899,
+      "learning_rate": 8.018609797758768e-06,
+      "loss": 0.6242,
+      "step": 24474
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4392324475041312,
+      "learning_rate": 8.017861890574488e-06,
+      "loss": 0.7235,
+      "step": 24475
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4215787501212767,
+      "learning_rate": 8.01711399493084e-06,
+      "loss": 0.7663,
+      "step": 24476
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.680952698057034,
+      "learning_rate": 8.016366110832186e-06,
+      "loss": 0.9416,
+      "step": 24477
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5003681205738602,
+      "learning_rate": 8.015618238282876e-06,
+      "loss": 0.8339,
+      "step": 24478
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4317601436480374,
+      "learning_rate": 8.014870377287263e-06,
+      "loss": 0.702,
+      "step": 24479
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5913838585519496,
+      "learning_rate": 8.014122527849706e-06,
+      "loss": 0.5164,
+      "step": 24480
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.764020308138768,
+      "learning_rate": 8.013374689974554e-06,
+      "loss": 0.618,
+      "step": 24481
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.405218961640575,
+      "learning_rate": 8.012626863666163e-06,
+      "loss": 0.7965,
+      "step": 24482
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4387709097802404,
+      "learning_rate": 8.011879048928888e-06,
+      "loss": 0.736,
+      "step": 24483
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3885693375712238,
+      "learning_rate": 8.011131245767086e-06,
+      "loss": 0.6776,
+      "step": 24484
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5741373058092958,
+      "learning_rate": 8.010383454185103e-06,
+      "loss": 0.6585,
+      "step": 24485
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8802139265499955,
+      "learning_rate": 8.0096356741873e-06,
+      "loss": 0.6425,
+      "step": 24486
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5848355784226191,
+      "learning_rate": 8.00888790577803e-06,
+      "loss": 0.4897,
+      "step": 24487
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1819737073372534,
+      "learning_rate": 8.008140148961642e-06,
+      "loss": 0.6396,
+      "step": 24488
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9301187168149012,
+      "learning_rate": 8.007392403742492e-06,
+      "loss": 0.753,
+      "step": 24489
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3213577374248864,
+      "learning_rate": 8.006644670124935e-06,
+      "loss": 0.7899,
+      "step": 24490
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4352718055474527,
+      "learning_rate": 8.005896948113326e-06,
+      "loss": 0.8375,
+      "step": 24491
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.463005457793768,
+      "learning_rate": 8.005149237712014e-06,
+      "loss": 0.6907,
+      "step": 24492
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2419415001937775,
+      "learning_rate": 8.004401538925353e-06,
+      "loss": 0.6979,
+      "step": 24493
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8688346228243742,
+      "learning_rate": 8.003653851757703e-06,
+      "loss": 0.6333,
+      "step": 24494
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.565630677159985,
+      "learning_rate": 8.00290617621341e-06,
+      "loss": 0.631,
+      "step": 24495
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3225062401766632,
+      "learning_rate": 8.002158512296829e-06,
+      "loss": 0.5966,
+      "step": 24496
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9699742676331797,
+      "learning_rate": 8.001410860012311e-06,
+      "loss": 0.6474,
+      "step": 24497
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4279882391893528,
+      "learning_rate": 8.000663219364217e-06,
+      "loss": 0.9075,
+      "step": 24498
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3646337402129871,
+      "learning_rate": 7.99991559035689e-06,
+      "loss": 0.7595,
+      "step": 24499
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3254714725422552,
+      "learning_rate": 7.99916797299469e-06,
+      "loss": 0.5709,
+      "step": 24500
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3165347999100272,
+      "learning_rate": 7.99842036728197e-06,
+      "loss": 0.6743,
+      "step": 24501
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3208799617632014,
+      "learning_rate": 7.997672773223077e-06,
+      "loss": 0.6309,
+      "step": 24502
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3541223707480357,
+      "learning_rate": 7.996925190822367e-06,
+      "loss": 0.7338,
+      "step": 24503
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4689267815328442,
+      "learning_rate": 7.996177620084191e-06,
+      "loss": 0.6753,
+      "step": 24504
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4895663426573846,
+      "learning_rate": 7.99543006101291e-06,
+      "loss": 0.7777,
+      "step": 24505
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7936070937265676,
+      "learning_rate": 7.994682513612863e-06,
+      "loss": 0.7789,
+      "step": 24506
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4229540432279182,
+      "learning_rate": 7.993934977888412e-06,
+      "loss": 0.6144,
+      "step": 24507
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4397476088674956,
+      "learning_rate": 7.993187453843908e-06,
+      "loss": 0.744,
+      "step": 24508
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5537550455850146,
+      "learning_rate": 7.992439941483702e-06,
+      "loss": 0.4886,
+      "step": 24509
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3948223325878466,
+      "learning_rate": 7.991692440812143e-06,
+      "loss": 0.5853,
+      "step": 24510
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6118411809446352,
+      "learning_rate": 7.990944951833586e-06,
+      "loss": 0.5594,
+      "step": 24511
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.612008377388218,
+      "learning_rate": 7.99019747455239e-06,
+      "loss": 0.7658,
+      "step": 24512
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.439123688318636,
+      "learning_rate": 7.989450008972893e-06,
+      "loss": 0.7404,
+      "step": 24513
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5529797353223975,
+      "learning_rate": 7.988702555099458e-06,
+      "loss": 0.8008,
+      "step": 24514
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5588471111515563,
+      "learning_rate": 7.987955112936435e-06,
+      "loss": 0.7065,
+      "step": 24515
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3022675207507006,
+      "learning_rate": 7.987207682488172e-06,
+      "loss": 0.77,
+      "step": 24516
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7150862007030971,
+      "learning_rate": 7.986460263759022e-06,
+      "loss": 0.5921,
+      "step": 24517
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.324346221035003,
+      "learning_rate": 7.98571285675334e-06,
+      "loss": 0.6273,
+      "step": 24518
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3618642194381596,
+      "learning_rate": 7.984965461475477e-06,
+      "loss": 0.7258,
+      "step": 24519
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6993497200277885,
+      "learning_rate": 7.984218077929782e-06,
+      "loss": 0.6889,
+      "step": 24520
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4487576012855798,
+      "learning_rate": 7.983470706120604e-06,
+      "loss": 0.7902,
+      "step": 24521
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.637957050772335,
+      "learning_rate": 7.982723346052302e-06,
+      "loss": 0.6929,
+      "step": 24522
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5717021801203008,
+      "learning_rate": 7.981975997729222e-06,
+      "loss": 0.5151,
+      "step": 24523
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5884195427741459,
+      "learning_rate": 7.981228661155718e-06,
+      "loss": 0.5352,
+      "step": 24524
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4879633659525184,
+      "learning_rate": 7.980481336336135e-06,
+      "loss": 0.6172,
+      "step": 24525
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5140637504429582,
+      "learning_rate": 7.979734023274835e-06,
+      "loss": 0.6341,
+      "step": 24526
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7293075662259647,
+      "learning_rate": 7.97898672197616e-06,
+      "loss": 0.7565,
+      "step": 24527
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.437218795611204,
+      "learning_rate": 7.978239432444466e-06,
+      "loss": 0.7326,
+      "step": 24528
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.435665867234323,
+      "learning_rate": 7.977492154684097e-06,
+      "loss": 0.7737,
+      "step": 24529
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3828462498974154,
+      "learning_rate": 7.976744888699416e-06,
+      "loss": 0.6636,
+      "step": 24530
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2505320235913027,
+      "learning_rate": 7.975997634494764e-06,
+      "loss": 0.7554,
+      "step": 24531
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6204707562608808,
+      "learning_rate": 7.975250392074493e-06,
+      "loss": 0.5651,
+      "step": 24532
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5558263973358697,
+      "learning_rate": 7.97450316144296e-06,
+      "loss": 0.6786,
+      "step": 24533
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.409114732574518,
+      "learning_rate": 7.973755942604505e-06,
+      "loss": 0.6653,
+      "step": 24534
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2875191991273356,
+      "learning_rate": 7.973008735563486e-06,
+      "loss": 0.7998,
+      "step": 24535
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6428277605136481,
+      "learning_rate": 7.972261540324252e-06,
+      "loss": 0.5286,
+      "step": 24536
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4631021560109907,
+      "learning_rate": 7.971514356891157e-06,
+      "loss": 0.6065,
+      "step": 24537
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.456231091871352,
+      "learning_rate": 7.970767185268541e-06,
+      "loss": 0.7316,
+      "step": 24538
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7729835762021608,
+      "learning_rate": 7.970020025460765e-06,
+      "loss": 0.6502,
+      "step": 24539
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7130307932564155,
+      "learning_rate": 7.969272877472178e-06,
+      "loss": 0.7381,
+      "step": 24540
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.7880681245892833,
+      "learning_rate": 7.968525741307123e-06,
+      "loss": 0.6459,
+      "step": 24541
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6062543592095805,
+      "learning_rate": 7.967778616969952e-06,
+      "loss": 0.496,
+      "step": 24542
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3922887607510446,
+      "learning_rate": 7.96703150446502e-06,
+      "loss": 0.7825,
+      "step": 24543
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4815661905030413,
+      "learning_rate": 7.966284403796677e-06,
+      "loss": 0.7305,
+      "step": 24544
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6134006189884658,
+      "learning_rate": 7.965537314969267e-06,
+      "loss": 0.551,
+      "step": 24545
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5893266113066897,
+      "learning_rate": 7.964790237987142e-06,
+      "loss": 0.8224,
+      "step": 24546
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8704003726188074,
+      "learning_rate": 7.964043172854657e-06,
+      "loss": 0.6924,
+      "step": 24547
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2875748950986556,
+      "learning_rate": 7.963296119576154e-06,
+      "loss": 0.7647,
+      "step": 24548
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3052312219137296,
+      "learning_rate": 7.962549078155984e-06,
+      "loss": 0.8231,
+      "step": 24549
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3542436940311653,
+      "learning_rate": 7.9618020485985e-06,
+      "loss": 0.6249,
+      "step": 24550
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4936312816699298,
+      "learning_rate": 7.961055030908052e-06,
+      "loss": 0.6709,
+      "step": 24551
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.259575438674793,
+      "learning_rate": 7.960308025088985e-06,
+      "loss": 0.6028,
+      "step": 24552
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.56791061042112,
+      "learning_rate": 7.959561031145648e-06,
+      "loss": 0.4719,
+      "step": 24553
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4714087788649683,
+      "learning_rate": 7.958814049082399e-06,
+      "loss": 0.8746,
+      "step": 24554
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3708980154321337,
+      "learning_rate": 7.958067078903574e-06,
+      "loss": 0.6845,
+      "step": 24555
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7621637771280536,
+      "learning_rate": 7.957320120613535e-06,
+      "loss": 0.731,
+      "step": 24556
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4585878298863402,
+      "learning_rate": 7.956573174216619e-06,
+      "loss": 0.7565,
+      "step": 24557
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4064083372509504,
+      "learning_rate": 7.955826239717189e-06,
+      "loss": 0.7186,
+      "step": 24558
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4304952839151304,
+      "learning_rate": 7.955079317119578e-06,
+      "loss": 0.7795,
+      "step": 24559
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.320263385674799,
+      "learning_rate": 7.954332406428145e-06,
+      "loss": 0.7665,
+      "step": 24560
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4465760005985724,
+      "learning_rate": 7.953585507647239e-06,
+      "loss": 0.6556,
+      "step": 24561
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2397814033533066,
+      "learning_rate": 7.952838620781201e-06,
+      "loss": 0.641,
+      "step": 24562
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5019829526236543,
+      "learning_rate": 7.952091745834387e-06,
+      "loss": 0.8654,
+      "step": 24563
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6771405045635719,
+      "learning_rate": 7.951344882811145e-06,
+      "loss": 0.67,
+      "step": 24564
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4484274470028062,
+      "learning_rate": 7.950598031715822e-06,
+      "loss": 0.6741,
+      "step": 24565
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5082040940433121,
+      "learning_rate": 7.949851192552762e-06,
+      "loss": 0.7406,
+      "step": 24566
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3392016211732989,
+      "learning_rate": 7.94910436532632e-06,
+      "loss": 0.7485,
+      "step": 24567
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4540089519163413,
+      "learning_rate": 7.948357550040843e-06,
+      "loss": 0.6555,
+      "step": 24568
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2647829936957293,
+      "learning_rate": 7.947610746700677e-06,
+      "loss": 0.6916,
+      "step": 24569
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3665852373841278,
+      "learning_rate": 7.946863955310169e-06,
+      "loss": 0.6664,
+      "step": 24570
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4713863800739722,
+      "learning_rate": 7.94611717587367e-06,
+      "loss": 0.8313,
+      "step": 24571
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6022035131154502,
+      "learning_rate": 7.94537040839553e-06,
+      "loss": 0.7212,
+      "step": 24572
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3718896855395437,
+      "learning_rate": 7.944623652880092e-06,
+      "loss": 0.7916,
+      "step": 24573
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2671464811834627,
+      "learning_rate": 7.943876909331702e-06,
+      "loss": 0.7894,
+      "step": 24574
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5751568274392098,
+      "learning_rate": 7.943130177754718e-06,
+      "loss": 0.7819,
+      "step": 24575
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.357261840366737,
+      "learning_rate": 7.942383458153478e-06,
+      "loss": 0.7202,
+      "step": 24576
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2729638125675375,
+      "learning_rate": 7.941636750532334e-06,
+      "loss": 0.6842,
+      "step": 24577
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4043827429999778,
+      "learning_rate": 7.940890054895627e-06,
+      "loss": 0.6535,
+      "step": 24578
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3590742640200713,
+      "learning_rate": 7.940143371247718e-06,
+      "loss": 0.6778,
+      "step": 24579
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2737302671870732,
+      "learning_rate": 7.939396699592942e-06,
+      "loss": 0.754,
+      "step": 24580
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3278071958317912,
+      "learning_rate": 7.938650039935654e-06,
+      "loss": 0.6678,
+      "step": 24581
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.480220034625424,
+      "learning_rate": 7.937903392280198e-06,
+      "loss": 0.6695,
+      "step": 24582
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9302125964850358,
+      "learning_rate": 7.937156756630915e-06,
+      "loss": 0.6818,
+      "step": 24583
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6248569033441032,
+      "learning_rate": 7.936410132992163e-06,
+      "loss": 0.5163,
+      "step": 24584
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.428077960937648,
+      "learning_rate": 7.935663521368283e-06,
+      "loss": 0.6461,
+      "step": 24585
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.448015840108849,
+      "learning_rate": 7.93491692176363e-06,
+      "loss": 0.6804,
+      "step": 24586
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.050668827605633,
+      "learning_rate": 7.934170334182537e-06,
+      "loss": 0.5901,
+      "step": 24587
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7661681063728378,
+      "learning_rate": 7.93342375862936e-06,
+      "loss": 0.6067,
+      "step": 24588
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5799940892956559,
+      "learning_rate": 7.932677195108447e-06,
+      "loss": 0.5265,
+      "step": 24589
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8905687686852977,
+      "learning_rate": 7.93193064362414e-06,
+      "loss": 0.7202,
+      "step": 24590
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.429322587912736,
+      "learning_rate": 7.931184104180784e-06,
+      "loss": 0.6874,
+      "step": 24591
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.329113385188559,
+      "learning_rate": 7.930437576782732e-06,
+      "loss": 0.7103,
+      "step": 24592
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2905853652670858,
+      "learning_rate": 7.92969106143433e-06,
+      "loss": 0.6473,
+      "step": 24593
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5459215814846212,
+      "learning_rate": 7.928944558139921e-06,
+      "loss": 0.6577,
+      "step": 24594
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3798565320746439,
+      "learning_rate": 7.928198066903849e-06,
+      "loss": 0.7632,
+      "step": 24595
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4216243510815463,
+      "learning_rate": 7.927451587730467e-06,
+      "loss": 0.8113,
+      "step": 24596
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5125140703336903,
+      "learning_rate": 7.926705120624119e-06,
+      "loss": 0.7734,
+      "step": 24597
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5065954610639085,
+      "learning_rate": 7.925958665589146e-06,
+      "loss": 0.6672,
+      "step": 24598
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5886828357580702,
+      "learning_rate": 7.9252122226299e-06,
+      "loss": 0.6291,
+      "step": 24599
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4503670601379406,
+      "learning_rate": 7.924465791750728e-06,
+      "loss": 0.7915,
+      "step": 24600
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5608682588159306,
+      "learning_rate": 7.923719372955973e-06,
+      "loss": 0.6432,
+      "step": 24601
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3213080789981928,
+      "learning_rate": 7.922972966249977e-06,
+      "loss": 0.6469,
+      "step": 24602
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4815427407794048,
+      "learning_rate": 7.922226571637096e-06,
+      "loss": 0.8847,
+      "step": 24603
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4223201398260423,
+      "learning_rate": 7.921480189121665e-06,
+      "loss": 0.627,
+      "step": 24604
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.376967538986871,
+      "learning_rate": 7.920733818708036e-06,
+      "loss": 0.7196,
+      "step": 24605
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0605600612001513,
+      "learning_rate": 7.919987460400549e-06,
+      "loss": 0.6929,
+      "step": 24606
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2980167895463548,
+      "learning_rate": 7.919241114203563e-06,
+      "loss": 0.7039,
+      "step": 24607
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2501959091315502,
+      "learning_rate": 7.918494780121404e-06,
+      "loss": 0.7115,
+      "step": 24608
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.608615168755437,
+      "learning_rate": 7.917748458158432e-06,
+      "loss": 0.5583,
+      "step": 24609
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.429965767820341,
+      "learning_rate": 7.917002148318989e-06,
+      "loss": 0.786,
+      "step": 24610
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.364185002982182,
+      "learning_rate": 7.916255850607415e-06,
+      "loss": 0.6236,
+      "step": 24611
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.712283314848165,
+      "learning_rate": 7.915509565028061e-06,
+      "loss": 0.7184,
+      "step": 24612
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.290356953515272,
+      "learning_rate": 7.914763291585271e-06,
+      "loss": 0.7603,
+      "step": 24613
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3822138719576147,
+      "learning_rate": 7.91401703028339e-06,
+      "loss": 0.8193,
+      "step": 24614
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3362222335921723,
+      "learning_rate": 7.91327078112676e-06,
+      "loss": 0.7478,
+      "step": 24615
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3467582791500372,
+      "learning_rate": 7.912524544119728e-06,
+      "loss": 0.7257,
+      "step": 24616
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.576702535828449,
+      "learning_rate": 7.911778319266643e-06,
+      "loss": 0.4978,
+      "step": 24617
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4058968784651684,
+      "learning_rate": 7.911032106571842e-06,
+      "loss": 0.7683,
+      "step": 24618
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4493524555061201,
+      "learning_rate": 7.910285906039673e-06,
+      "loss": 0.6797,
+      "step": 24619
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4385228322487782,
+      "learning_rate": 7.909539717674483e-06,
+      "loss": 0.7015,
+      "step": 24620
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5016632238033796,
+      "learning_rate": 7.908793541480615e-06,
+      "loss": 0.7781,
+      "step": 24621
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4765479968020354,
+      "learning_rate": 7.908047377462413e-06,
+      "loss": 0.8428,
+      "step": 24622
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2942125457440588,
+      "learning_rate": 7.907301225624218e-06,
+      "loss": 0.7189,
+      "step": 24623
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6018471954074126,
+      "learning_rate": 7.906555085970382e-06,
+      "loss": 0.5492,
+      "step": 24624
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4831291837587317,
+      "learning_rate": 7.905808958505243e-06,
+      "loss": 0.7675,
+      "step": 24625
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.1942324310636097,
+      "learning_rate": 7.905062843233148e-06,
+      "loss": 0.6654,
+      "step": 24626
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4497478233802823,
+      "learning_rate": 7.904316740158437e-06,
+      "loss": 0.8054,
+      "step": 24627
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5044505345074177,
+      "learning_rate": 7.903570649285463e-06,
+      "loss": 0.7504,
+      "step": 24628
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5033714834548917,
+      "learning_rate": 7.902824570618562e-06,
+      "loss": 0.7901,
+      "step": 24629
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.287965608973019,
+      "learning_rate": 7.90207850416208e-06,
+      "loss": 0.7211,
+      "step": 24630
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4012515292493266,
+      "learning_rate": 7.901332449920365e-06,
+      "loss": 0.722,
+      "step": 24631
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6800401022623985,
+      "learning_rate": 7.90058640789775e-06,
+      "loss": 0.8465,
+      "step": 24632
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4645802151729204,
+      "learning_rate": 7.899840378098589e-06,
+      "loss": 0.7528,
+      "step": 24633
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4951985234560554,
+      "learning_rate": 7.89909436052722e-06,
+      "loss": 0.7294,
+      "step": 24634
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6017697439649261,
+      "learning_rate": 7.898348355187994e-06,
+      "loss": 0.4628,
+      "step": 24635
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1746173876629564,
+      "learning_rate": 7.897602362085246e-06,
+      "loss": 0.7237,
+      "step": 24636
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.478404467469745,
+      "learning_rate": 7.896856381223321e-06,
+      "loss": 0.8258,
+      "step": 24637
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3040630909050384,
+      "learning_rate": 7.896110412606569e-06,
+      "loss": 0.7249,
+      "step": 24638
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4243212772815894,
+      "learning_rate": 7.895364456239324e-06,
+      "loss": 0.6219,
+      "step": 24639
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4689332788217424,
+      "learning_rate": 7.894618512125932e-06,
+      "loss": 0.6827,
+      "step": 24640
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4300088373434516,
+      "learning_rate": 7.893872580270741e-06,
+      "loss": 0.7884,
+      "step": 24641
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6176348695286116,
+      "learning_rate": 7.893126660678091e-06,
+      "loss": 0.5353,
+      "step": 24642
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3947258483633886,
+      "learning_rate": 7.892380753352324e-06,
+      "loss": 0.6555,
+      "step": 24643
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3282855757134064,
+      "learning_rate": 7.89163485829778e-06,
+      "loss": 0.6473,
+      "step": 24644
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2022696964081527,
+      "learning_rate": 7.89088897551881e-06,
+      "loss": 0.6244,
+      "step": 24645
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.318483679836369,
+      "learning_rate": 7.89014310501975e-06,
+      "loss": 0.6435,
+      "step": 24646
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4270200304415168,
+      "learning_rate": 7.889397246804943e-06,
+      "loss": 0.6521,
+      "step": 24647
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6537580773961578,
+      "learning_rate": 7.888651400878736e-06,
+      "loss": 0.5051,
+      "step": 24648
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.367380339037819,
+      "learning_rate": 7.88790556724547e-06,
+      "loss": 0.7082,
+      "step": 24649
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6123007168148663,
+      "learning_rate": 7.887159745909484e-06,
+      "loss": 0.5528,
+      "step": 24650
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5809491172189716,
+      "learning_rate": 7.886413936875122e-06,
+      "loss": 0.709,
+      "step": 24651
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6285586871273343,
+      "learning_rate": 7.88566814014673e-06,
+      "loss": 0.7659,
+      "step": 24652
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.261339803989675,
+      "learning_rate": 7.884922355728649e-06,
+      "loss": 0.6485,
+      "step": 24653
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.434276362335467,
+      "learning_rate": 7.884176583625215e-06,
+      "loss": 0.82,
+      "step": 24654
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7934547784510344,
+      "learning_rate": 7.883430823840775e-06,
+      "loss": 0.6603,
+      "step": 24655
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3721754833997541,
+      "learning_rate": 7.882685076379675e-06,
+      "loss": 0.6456,
+      "step": 24656
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.657778456170376,
+      "learning_rate": 7.88193934124625e-06,
+      "loss": 0.7009,
+      "step": 24657
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7386599169286763,
+      "learning_rate": 7.881193618444846e-06,
+      "loss": 0.6583,
+      "step": 24658
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9413716552443905,
+      "learning_rate": 7.8804479079798e-06,
+      "loss": 0.8484,
+      "step": 24659
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6080283306229992,
+      "learning_rate": 7.879702209855463e-06,
+      "loss": 0.8401,
+      "step": 24660
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3247624006987226,
+      "learning_rate": 7.878956524076167e-06,
+      "loss": 0.6195,
+      "step": 24661
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.345050388268681,
+      "learning_rate": 7.87821085064626e-06,
+      "loss": 0.6239,
+      "step": 24662
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5912726194254756,
+      "learning_rate": 7.877465189570081e-06,
+      "loss": 0.6943,
+      "step": 24663
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3613273943404989,
+      "learning_rate": 7.876719540851969e-06,
+      "loss": 0.7374,
+      "step": 24664
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3898103934643011,
+      "learning_rate": 7.87597390449627e-06,
+      "loss": 0.6907,
+      "step": 24665
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6945012396722323,
+      "learning_rate": 7.875228280507322e-06,
+      "loss": 0.7445,
+      "step": 24666
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6151162269894577,
+      "learning_rate": 7.874482668889472e-06,
+      "loss": 0.553,
+      "step": 24667
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4268392922930402,
+      "learning_rate": 7.87373706964705e-06,
+      "loss": 0.749,
+      "step": 24668
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4483743795249402,
+      "learning_rate": 7.872991482784408e-06,
+      "loss": 0.6617,
+      "step": 24669
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3765508368121473,
+      "learning_rate": 7.872245908305885e-06,
+      "loss": 0.7849,
+      "step": 24670
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8480418859178795,
+      "learning_rate": 7.871500346215817e-06,
+      "loss": 0.7633,
+      "step": 24671
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3796238161468382,
+      "learning_rate": 7.870754796518547e-06,
+      "loss": 0.7349,
+      "step": 24672
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.488616339362912,
+      "learning_rate": 7.870009259218417e-06,
+      "loss": 0.7059,
+      "step": 24673
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3457917259448808,
+      "learning_rate": 7.86926373431977e-06,
+      "loss": 0.5996,
+      "step": 24674
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3144857514181851,
+      "learning_rate": 7.868518221826943e-06,
+      "loss": 0.664,
+      "step": 24675
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2427922312199424,
+      "learning_rate": 7.867772721744275e-06,
+      "loss": 0.7882,
+      "step": 24676
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6545123848222267,
+      "learning_rate": 7.867027234076115e-06,
+      "loss": 0.7139,
+      "step": 24677
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3635157962017979,
+      "learning_rate": 7.866281758826793e-06,
+      "loss": 0.8634,
+      "step": 24678
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1989470670390223,
+      "learning_rate": 7.865536296000655e-06,
+      "loss": 0.7553,
+      "step": 24679
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6750963128070535,
+      "learning_rate": 7.86479084560204e-06,
+      "loss": 0.7552,
+      "step": 24680
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.59230768300905,
+      "learning_rate": 7.86404540763529e-06,
+      "loss": 0.5884,
+      "step": 24681
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4781460173764975,
+      "learning_rate": 7.863299982104745e-06,
+      "loss": 0.6457,
+      "step": 24682
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3755907897597563,
+      "learning_rate": 7.862554569014739e-06,
+      "loss": 0.5182,
+      "step": 24683
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3282957153573298,
+      "learning_rate": 7.861809168369626e-06,
+      "loss": 0.7785,
+      "step": 24684
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4773165640044,
+      "learning_rate": 7.861063780173728e-06,
+      "loss": 0.7195,
+      "step": 24685
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.486473544368858,
+      "learning_rate": 7.860318404431397e-06,
+      "loss": 0.7936,
+      "step": 24686
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.432084308539362,
+      "learning_rate": 7.859573041146968e-06,
+      "loss": 0.6228,
+      "step": 24687
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4100791831542339,
+      "learning_rate": 7.858827690324788e-06,
+      "loss": 0.8551,
+      "step": 24688
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5852604016253962,
+      "learning_rate": 7.858082351969185e-06,
+      "loss": 0.5505,
+      "step": 24689
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2302590802134161,
+      "learning_rate": 7.857337026084508e-06,
+      "loss": 0.6259,
+      "step": 24690
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.435429694378824,
+      "learning_rate": 7.856591712675093e-06,
+      "loss": 0.7171,
+      "step": 24691
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5839844915554758,
+      "learning_rate": 7.855846411745278e-06,
+      "loss": 0.4848,
+      "step": 24692
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3249943569384472,
+      "learning_rate": 7.855101123299403e-06,
+      "loss": 0.6143,
+      "step": 24693
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.69037938303106,
+      "learning_rate": 7.85435584734181e-06,
+      "loss": 0.6513,
+      "step": 24694
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.394847844307916,
+      "learning_rate": 7.853610583876839e-06,
+      "loss": 0.6507,
+      "step": 24695
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.1476112301443147,
+      "learning_rate": 7.852865332908822e-06,
+      "loss": 0.6882,
+      "step": 24696
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5164583983585957,
+      "learning_rate": 7.852120094442105e-06,
+      "loss": 0.7791,
+      "step": 24697
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5011363101141624,
+      "learning_rate": 7.851374868481027e-06,
+      "loss": 0.8238,
+      "step": 24698
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3242932942684176,
+      "learning_rate": 7.850629655029923e-06,
+      "loss": 0.7889,
+      "step": 24699
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7159248824739055,
+      "learning_rate": 7.849884454093132e-06,
+      "loss": 0.8447,
+      "step": 24700
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2871310362219597,
+      "learning_rate": 7.849139265674995e-06,
+      "loss": 0.7857,
+      "step": 24701
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5811037310351885,
+      "learning_rate": 7.848394089779854e-06,
+      "loss": 0.9138,
+      "step": 24702
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.597385932156323,
+      "learning_rate": 7.84764892641204e-06,
+      "loss": 0.5382,
+      "step": 24703
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3964476958206244,
+      "learning_rate": 7.846903775575893e-06,
+      "loss": 0.5713,
+      "step": 24704
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.41341408860662,
+      "learning_rate": 7.846158637275761e-06,
+      "loss": 0.708,
+      "step": 24705
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5365243052401905,
+      "learning_rate": 7.84541351151597e-06,
+      "loss": 0.7404,
+      "step": 24706
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5643984682323617,
+      "learning_rate": 7.844668398300866e-06,
+      "loss": 0.7825,
+      "step": 24707
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.205786048049071,
+      "learning_rate": 7.843923297634781e-06,
+      "loss": 0.7469,
+      "step": 24708
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4872620370842957,
+      "learning_rate": 7.843178209522063e-06,
+      "loss": 0.6823,
+      "step": 24709
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5837041133412815,
+      "learning_rate": 7.842433133967041e-06,
+      "loss": 0.567,
+      "step": 24710
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.833969262875125,
+      "learning_rate": 7.841688070974056e-06,
+      "loss": 0.7615,
+      "step": 24711
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9444724280268508,
+      "learning_rate": 7.84094302054745e-06,
+      "loss": 0.6614,
+      "step": 24712
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4333928192992509,
+      "learning_rate": 7.840197982691552e-06,
+      "loss": 0.6984,
+      "step": 24713
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.314136295548336,
+      "learning_rate": 7.839452957410707e-06,
+      "loss": 0.6784,
+      "step": 24714
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4731023101976322,
+      "learning_rate": 7.838707944709252e-06,
+      "loss": 0.7348,
+      "step": 24715
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5489664747872445,
+      "learning_rate": 7.837962944591525e-06,
+      "loss": 0.6699,
+      "step": 24716
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3166086554223488,
+      "learning_rate": 7.837217957061858e-06,
+      "loss": 0.7179,
+      "step": 24717
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3919592792661957,
+      "learning_rate": 7.836472982124596e-06,
+      "loss": 0.8064,
+      "step": 24718
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5100005986630063,
+      "learning_rate": 7.835728019784074e-06,
+      "loss": 0.849,
+      "step": 24719
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9694546543748512,
+      "learning_rate": 7.834983070044626e-06,
+      "loss": 0.6236,
+      "step": 24720
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6090626604855613,
+      "learning_rate": 7.83423813291059e-06,
+      "loss": 0.5402,
+      "step": 24721
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5725508710455947,
+      "learning_rate": 7.833493208386308e-06,
+      "loss": 0.7304,
+      "step": 24722
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5882350407410141,
+      "learning_rate": 7.832748296476117e-06,
+      "loss": 0.526,
+      "step": 24723
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4647235778450876,
+      "learning_rate": 7.832003397184351e-06,
+      "loss": 0.7117,
+      "step": 24724
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2502742205194013,
+      "learning_rate": 7.831258510515342e-06,
+      "loss": 0.6694,
+      "step": 24725
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1525818817894709,
+      "learning_rate": 7.83051363647344e-06,
+      "loss": 0.6595,
+      "step": 24726
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3703226454282365,
+      "learning_rate": 7.82976877506297e-06,
+      "loss": 0.6609,
+      "step": 24727
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.862924825204757,
+      "learning_rate": 7.829023926288275e-06,
+      "loss": 0.7066,
+      "step": 24728
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.400701181795859,
+      "learning_rate": 7.828279090153686e-06,
+      "loss": 0.7224,
+      "step": 24729
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.437382663569177,
+      "learning_rate": 7.827534266663548e-06,
+      "loss": 0.7099,
+      "step": 24730
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5858756932658746,
+      "learning_rate": 7.826789455822194e-06,
+      "loss": 0.7372,
+      "step": 24731
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.576479075948658,
+      "learning_rate": 7.826044657633956e-06,
+      "loss": 0.7928,
+      "step": 24732
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2942893430453863,
+      "learning_rate": 7.825299872103181e-06,
+      "loss": 0.7194,
+      "step": 24733
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5528288466595157,
+      "learning_rate": 7.82455509923419e-06,
+      "loss": 0.7729,
+      "step": 24734
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6203298591983142,
+      "learning_rate": 7.823810339031332e-06,
+      "loss": 0.7617,
+      "step": 24735
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4603024476213646,
+      "learning_rate": 7.823065591498937e-06,
+      "loss": 0.6059,
+      "step": 24736
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.45227757015164,
+      "learning_rate": 7.82232085664135e-06,
+      "loss": 0.6408,
+      "step": 24737
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3097127048665265,
+      "learning_rate": 7.821576134462893e-06,
+      "loss": 0.6988,
+      "step": 24738
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.744993734317497,
+      "learning_rate": 7.820831424967912e-06,
+      "loss": 0.6446,
+      "step": 24739
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5930643038273102,
+      "learning_rate": 7.820086728160743e-06,
+      "loss": 0.6902,
+      "step": 24740
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5947507812262954,
+      "learning_rate": 7.819342044045716e-06,
+      "loss": 0.492,
+      "step": 24741
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.642703055830963,
+      "learning_rate": 7.81859737262717e-06,
+      "loss": 0.5481,
+      "step": 24742
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7917741944272736,
+      "learning_rate": 7.817852713909442e-06,
+      "loss": 0.7693,
+      "step": 24743
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5759585730911003,
+      "learning_rate": 7.817108067896865e-06,
+      "loss": 0.5026,
+      "step": 24744
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3448410400888018,
+      "learning_rate": 7.816363434593776e-06,
+      "loss": 0.7783,
+      "step": 24745
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4379096559189988,
+      "learning_rate": 7.815618814004511e-06,
+      "loss": 0.9126,
+      "step": 24746
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3231138239910687,
+      "learning_rate": 7.814874206133407e-06,
+      "loss": 0.7727,
+      "step": 24747
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3120990227635927,
+      "learning_rate": 7.814129610984795e-06,
+      "loss": 0.8398,
+      "step": 24748
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.60145946100719,
+      "learning_rate": 7.81338502856301e-06,
+      "loss": 0.7225,
+      "step": 24749
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3367162568191648,
+      "learning_rate": 7.812640458872393e-06,
+      "loss": 0.8366,
+      "step": 24750
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6860947396285078,
+      "learning_rate": 7.811895901917277e-06,
+      "loss": 0.6517,
+      "step": 24751
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3057916622195462,
+      "learning_rate": 7.811151357701993e-06,
+      "loss": 0.8134,
+      "step": 24752
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4024706725652436,
+      "learning_rate": 7.810406826230877e-06,
+      "loss": 0.7763,
+      "step": 24753
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.262457484315001,
+      "learning_rate": 7.809662307508271e-06,
+      "loss": 0.7256,
+      "step": 24754
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.33034909304523,
+      "learning_rate": 7.808917801538502e-06,
+      "loss": 0.6533,
+      "step": 24755
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.216606858446803,
+      "learning_rate": 7.808173308325905e-06,
+      "loss": 0.6587,
+      "step": 24756
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6005016023973369,
+      "learning_rate": 7.807428827874818e-06,
+      "loss": 0.8032,
+      "step": 24757
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6064989837471768,
+      "learning_rate": 7.806684360189579e-06,
+      "loss": 0.7366,
+      "step": 24758
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3252143661824536,
+      "learning_rate": 7.805939905274514e-06,
+      "loss": 0.799,
+      "step": 24759
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3708300015039963,
+      "learning_rate": 7.80519546313396e-06,
+      "loss": 0.8268,
+      "step": 24760
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4572530761287827,
+      "learning_rate": 7.804451033772258e-06,
+      "loss": 0.5951,
+      "step": 24761
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.52921508117621,
+      "learning_rate": 7.803706617193731e-06,
+      "loss": 0.6259,
+      "step": 24762
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6023054152574097,
+      "learning_rate": 7.802962213402723e-06,
+      "loss": 0.5512,
+      "step": 24763
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3184100920361421,
+      "learning_rate": 7.802217822403563e-06,
+      "loss": 0.8517,
+      "step": 24764
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6549688325427614,
+      "learning_rate": 7.80147344420059e-06,
+      "loss": 0.7408,
+      "step": 24765
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7080125372095012,
+      "learning_rate": 7.80072907879813e-06,
+      "loss": 0.7292,
+      "step": 24766
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5090994862451446,
+      "learning_rate": 7.799984726200524e-06,
+      "loss": 0.8666,
+      "step": 24767
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.389367406247383,
+      "learning_rate": 7.799240386412106e-06,
+      "loss": 0.6473,
+      "step": 24768
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4952657137489274,
+      "learning_rate": 7.798496059437204e-06,
+      "loss": 0.6791,
+      "step": 24769
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.7381858675524184,
+      "learning_rate": 7.797751745280153e-06,
+      "loss": 0.6812,
+      "step": 24770
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.395940512656977,
+      "learning_rate": 7.797007443945291e-06,
+      "loss": 0.6755,
+      "step": 24771
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.209950289285724,
+      "learning_rate": 7.796263155436952e-06,
+      "loss": 0.7149,
+      "step": 24772
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.307042779064855,
+      "learning_rate": 7.795518879759462e-06,
+      "loss": 0.7404,
+      "step": 24773
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2623693790655623,
+      "learning_rate": 7.79477461691716e-06,
+      "loss": 0.7755,
+      "step": 24774
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4496907281938602,
+      "learning_rate": 7.794030366914379e-06,
+      "loss": 0.634,
+      "step": 24775
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5756289128262768,
+      "learning_rate": 7.793286129755453e-06,
+      "loss": 0.7357,
+      "step": 24776
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7446574289329544,
+      "learning_rate": 7.792541905444713e-06,
+      "loss": 0.8459,
+      "step": 24777
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4752365720459457,
+      "learning_rate": 7.79179769398649e-06,
+      "loss": 0.7217,
+      "step": 24778
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.37609735625839,
+      "learning_rate": 7.791053495385124e-06,
+      "loss": 0.7633,
+      "step": 24779
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5804233722551575,
+      "learning_rate": 7.790309309644942e-06,
+      "loss": 0.6649,
+      "step": 24780
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5375317316642716,
+      "learning_rate": 7.789565136770278e-06,
+      "loss": 0.7126,
+      "step": 24781
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4762965971911337,
+      "learning_rate": 7.788820976765466e-06,
+      "loss": 0.6874,
+      "step": 24782
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3534440432398873,
+      "learning_rate": 7.788076829634842e-06,
+      "loss": 0.7349,
+      "step": 24783
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5964482355769354,
+      "learning_rate": 7.78733269538273e-06,
+      "loss": 0.6776,
+      "step": 24784
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5814480801606124,
+      "learning_rate": 7.786588574013467e-06,
+      "loss": 0.5062,
+      "step": 24785
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5864006762592675,
+      "learning_rate": 7.785844465531393e-06,
+      "loss": 0.6729,
+      "step": 24786
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5670408911236644,
+      "learning_rate": 7.785100369940826e-06,
+      "loss": 0.7355,
+      "step": 24787
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.460659745056894,
+      "learning_rate": 7.78435628724611e-06,
+      "loss": 0.7983,
+      "step": 24788
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3449997935989595,
+      "learning_rate": 7.78361221745157e-06,
+      "loss": 0.709,
+      "step": 24789
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5026272202884976,
+      "learning_rate": 7.782868160561548e-06,
+      "loss": 0.7587,
+      "step": 24790
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7931608398152956,
+      "learning_rate": 7.782124116580362e-06,
+      "loss": 0.5992,
+      "step": 24791
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5147134063991805,
+      "learning_rate": 7.781380085512355e-06,
+      "loss": 0.813,
+      "step": 24792
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.479259546936098,
+      "learning_rate": 7.780636067361857e-06,
+      "loss": 0.688,
+      "step": 24793
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3788135478998402,
+      "learning_rate": 7.779892062133193e-06,
+      "loss": 0.6545,
+      "step": 24794
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5309647808158948,
+      "learning_rate": 7.779148069830705e-06,
+      "loss": 0.6033,
+      "step": 24795
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.591408810829791,
+      "learning_rate": 7.778404090458718e-06,
+      "loss": 0.7024,
+      "step": 24796
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3547729676970865,
+      "learning_rate": 7.777660124021569e-06,
+      "loss": 0.6636,
+      "step": 24797
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4933394229649253,
+      "learning_rate": 7.77691617052358e-06,
+      "loss": 0.7732,
+      "step": 24798
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2553822225207598,
+      "learning_rate": 7.776172229969095e-06,
+      "loss": 0.7555,
+      "step": 24799
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2810630202831172,
+      "learning_rate": 7.775428302362437e-06,
+      "loss": 0.6647,
+      "step": 24800
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2824147630649954,
+      "learning_rate": 7.77468438770794e-06,
+      "loss": 0.6805,
+      "step": 24801
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1872331045904647,
+      "learning_rate": 7.773940486009934e-06,
+      "loss": 0.7157,
+      "step": 24802
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2863226637431473,
+      "learning_rate": 7.773196597272751e-06,
+      "loss": 0.7902,
+      "step": 24803
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.355379676037116,
+      "learning_rate": 7.772452721500726e-06,
+      "loss": 0.6369,
+      "step": 24804
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.223556603650021,
+      "learning_rate": 7.771708858698184e-06,
+      "loss": 0.6422,
+      "step": 24805
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.52219662361383,
+      "learning_rate": 7.770965008869455e-06,
+      "loss": 0.5649,
+      "step": 24806
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4831300922074906,
+      "learning_rate": 7.77022117201888e-06,
+      "loss": 0.7549,
+      "step": 24807
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3303859862334468,
+      "learning_rate": 7.76947734815078e-06,
+      "loss": 0.6466,
+      "step": 24808
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.597443269237025,
+      "learning_rate": 7.768733537269488e-06,
+      "loss": 0.8117,
+      "step": 24809
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4128723130214909,
+      "learning_rate": 7.767989739379336e-06,
+      "loss": 0.7402,
+      "step": 24810
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4906219841100898,
+      "learning_rate": 7.767245954484657e-06,
+      "loss": 0.7956,
+      "step": 24811
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3905576441365464,
+      "learning_rate": 7.766502182589776e-06,
+      "loss": 0.8416,
+      "step": 24812
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5843595356427201,
+      "learning_rate": 7.765758423699027e-06,
+      "loss": 0.4925,
+      "step": 24813
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3280133055998593,
+      "learning_rate": 7.765014677816745e-06,
+      "loss": 0.7423,
+      "step": 24814
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.554714683823414,
+      "learning_rate": 7.764270944947248e-06,
+      "loss": 0.6827,
+      "step": 24815
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.436803986560859,
+      "learning_rate": 7.763527225094876e-06,
+      "loss": 0.7829,
+      "step": 24816
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4516836619924018,
+      "learning_rate": 7.762783518263953e-06,
+      "loss": 0.6629,
+      "step": 24817
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5991608627184882,
+      "learning_rate": 7.76203982445882e-06,
+      "loss": 0.5143,
+      "step": 24818
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.492266159401338,
+      "learning_rate": 7.761296143683793e-06,
+      "loss": 0.7603,
+      "step": 24819
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4106171339740923,
+      "learning_rate": 7.760552475943212e-06,
+      "loss": 0.6678,
+      "step": 24820
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3990437594462002,
+      "learning_rate": 7.759808821241406e-06,
+      "loss": 0.6429,
+      "step": 24821
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.387296218509638,
+      "learning_rate": 7.759065179582698e-06,
+      "loss": 0.624,
+      "step": 24822
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4611081637086991,
+      "learning_rate": 7.758321550971421e-06,
+      "loss": 0.6737,
+      "step": 24823
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.359922257900328,
+      "learning_rate": 7.757577935411908e-06,
+      "loss": 0.7715,
+      "step": 24824
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3693967962952223,
+      "learning_rate": 7.75683433290849e-06,
+      "loss": 0.8526,
+      "step": 24825
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4677464329568144,
+      "learning_rate": 7.756090743465488e-06,
+      "loss": 0.5112,
+      "step": 24826
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3171324498314212,
+      "learning_rate": 7.755347167087234e-06,
+      "loss": 0.7271,
+      "step": 24827
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4300691112803525,
+      "learning_rate": 7.754603603778063e-06,
+      "loss": 0.6557,
+      "step": 24828
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2479178131591087,
+      "learning_rate": 7.753860053542302e-06,
+      "loss": 0.725,
+      "step": 24829
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.304368977205972,
+      "learning_rate": 7.753116516384274e-06,
+      "loss": 0.6383,
+      "step": 24830
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5642125630683859,
+      "learning_rate": 7.752372992308316e-06,
+      "loss": 0.6625,
+      "step": 24831
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.583251626857321,
+      "learning_rate": 7.751629481318756e-06,
+      "loss": 0.7394,
+      "step": 24832
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.330383984119327,
+      "learning_rate": 7.750885983419919e-06,
+      "loss": 0.6677,
+      "step": 24833
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7924995109527464,
+      "learning_rate": 7.750142498616133e-06,
+      "loss": 0.5956,
+      "step": 24834
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7426255310042582,
+      "learning_rate": 7.749399026911736e-06,
+      "loss": 0.6091,
+      "step": 24835
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5655258148265243,
+      "learning_rate": 7.748655568311045e-06,
+      "loss": 0.5103,
+      "step": 24836
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2787528001502058,
+      "learning_rate": 7.747912122818395e-06,
+      "loss": 0.6114,
+      "step": 24837
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5832826919946433,
+      "learning_rate": 7.747168690438112e-06,
+      "loss": 0.5005,
+      "step": 24838
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6344507404509331,
+      "learning_rate": 7.746425271174533e-06,
+      "loss": 0.5108,
+      "step": 24839
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5266310279859452,
+      "learning_rate": 7.745681865031973e-06,
+      "loss": 0.7545,
+      "step": 24840
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3408736090089723,
+      "learning_rate": 7.74493847201477e-06,
+      "loss": 0.832,
+      "step": 24841
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.288921674279797,
+      "learning_rate": 7.74419509212725e-06,
+      "loss": 0.7295,
+      "step": 24842
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4358563171144965,
+      "learning_rate": 7.743451725373734e-06,
+      "loss": 0.6867,
+      "step": 24843
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4388070033388558,
+      "learning_rate": 7.742708371758562e-06,
+      "loss": 0.5725,
+      "step": 24844
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3408407693699882,
+      "learning_rate": 7.741965031286055e-06,
+      "loss": 0.7652,
+      "step": 24845
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6019789436626752,
+      "learning_rate": 7.741221703960547e-06,
+      "loss": 0.4763,
+      "step": 24846
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.219760123211397,
+      "learning_rate": 7.740478389786354e-06,
+      "loss": 0.7049,
+      "step": 24847
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3440405558389532,
+      "learning_rate": 7.739735088767815e-06,
+      "loss": 0.6281,
+      "step": 24848
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6028982647763401,
+      "learning_rate": 7.738991800909254e-06,
+      "loss": 0.5043,
+      "step": 24849
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7331331524873645,
+      "learning_rate": 7.738248526214998e-06,
+      "loss": 0.6801,
+      "step": 24850
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7601587975834803,
+      "learning_rate": 7.737505264689374e-06,
+      "loss": 0.7588,
+      "step": 24851
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4233434413611714,
+      "learning_rate": 7.73676201633671e-06,
+      "loss": 0.728,
+      "step": 24852
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2540183987949574,
+      "learning_rate": 7.736018781161339e-06,
+      "loss": 0.6817,
+      "step": 24853
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3065489871181186,
+      "learning_rate": 7.73527555916758e-06,
+      "loss": 0.6326,
+      "step": 24854
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6036110512433519,
+      "learning_rate": 7.734532350359762e-06,
+      "loss": 0.5073,
+      "step": 24855
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2147100894608116,
+      "learning_rate": 7.733789154742217e-06,
+      "loss": 0.7354,
+      "step": 24856
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3611197893496685,
+      "learning_rate": 7.733045972319268e-06,
+      "loss": 0.7113,
+      "step": 24857
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3294633710052064,
+      "learning_rate": 7.732302803095242e-06,
+      "loss": 0.7061,
+      "step": 24858
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2358627915901195,
+      "learning_rate": 7.731559647074466e-06,
+      "loss": 0.6944,
+      "step": 24859
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9736197454581716,
+      "learning_rate": 7.730816504261271e-06,
+      "loss": 0.7451,
+      "step": 24860
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7179397667045953,
+      "learning_rate": 7.73007337465998e-06,
+      "loss": 0.706,
+      "step": 24861
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3065384963125692,
+      "learning_rate": 7.729330258274919e-06,
+      "loss": 0.646,
+      "step": 24862
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5872804209933574,
+      "learning_rate": 7.72858715511042e-06,
+      "loss": 0.5372,
+      "step": 24863
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4142153175783352,
+      "learning_rate": 7.7278440651708e-06,
+      "loss": 0.69,
+      "step": 24864
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8716228391648118,
+      "learning_rate": 7.727100988460394e-06,
+      "loss": 0.7423,
+      "step": 24865
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3949407358568264,
+      "learning_rate": 7.726357924983524e-06,
+      "loss": 0.7313,
+      "step": 24866
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5757098222042476,
+      "learning_rate": 7.725614874744524e-06,
+      "loss": 0.5203,
+      "step": 24867
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2989878423399475,
+      "learning_rate": 7.724871837747708e-06,
+      "loss": 0.6625,
+      "step": 24868
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4782816336477884,
+      "learning_rate": 7.72412881399741e-06,
+      "loss": 0.6462,
+      "step": 24869
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3691779698403925,
+      "learning_rate": 7.723385803497956e-06,
+      "loss": 0.7177,
+      "step": 24870
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5812024005139391,
+      "learning_rate": 7.72264280625367e-06,
+      "loss": 0.6786,
+      "step": 24871
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3217824503080757,
+      "learning_rate": 7.721899822268875e-06,
+      "loss": 0.6954,
+      "step": 24872
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.7551124017045465,
+      "learning_rate": 7.721156851547903e-06,
+      "loss": 0.6779,
+      "step": 24873
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6264609673222917,
+      "learning_rate": 7.72041389409508e-06,
+      "loss": 0.5418,
+      "step": 24874
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2352801160509341,
+      "learning_rate": 7.719670949914729e-06,
+      "loss": 0.719,
+      "step": 24875
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.861396930840782,
+      "learning_rate": 7.71892801901117e-06,
+      "loss": 0.651,
+      "step": 24876
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4633947673438585,
+      "learning_rate": 7.71818510138874e-06,
+      "loss": 0.8887,
+      "step": 24877
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4351614039726626,
+      "learning_rate": 7.717442197051756e-06,
+      "loss": 0.7989,
+      "step": 24878
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.283438499449137,
+      "learning_rate": 7.716699306004544e-06,
+      "loss": 0.6993,
+      "step": 24879
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5527020344575133,
+      "learning_rate": 7.715956428251433e-06,
+      "loss": 0.7199,
+      "step": 24880
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5959890444648421,
+      "learning_rate": 7.715213563796752e-06,
+      "loss": 0.5111,
+      "step": 24881
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3870829666799485,
+      "learning_rate": 7.714470712644817e-06,
+      "loss": 0.658,
+      "step": 24882
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3699859498636786,
+      "learning_rate": 7.713727874799954e-06,
+      "loss": 0.6298,
+      "step": 24883
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7469748989172182,
+      "learning_rate": 7.712985050266499e-06,
+      "loss": 0.7674,
+      "step": 24884
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3746868320549992,
+      "learning_rate": 7.712242239048761e-06,
+      "loss": 0.5854,
+      "step": 24885
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.51993746755024,
+      "learning_rate": 7.711499441151077e-06,
+      "loss": 0.7088,
+      "step": 24886
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3322851216129665,
+      "learning_rate": 7.710756656577764e-06,
+      "loss": 0.7078,
+      "step": 24887
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6361178940350872,
+      "learning_rate": 7.710013885333158e-06,
+      "loss": 0.5336,
+      "step": 24888
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4859342152668928,
+      "learning_rate": 7.70927112742157e-06,
+      "loss": 0.671,
+      "step": 24889
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.377518270813091,
+      "learning_rate": 7.708528382847333e-06,
+      "loss": 0.681,
+      "step": 24890
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8066140829372583,
+      "learning_rate": 7.70778565161477e-06,
+      "loss": 0.8339,
+      "step": 24891
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8130630315908154,
+      "learning_rate": 7.707042933728202e-06,
+      "loss": 0.8203,
+      "step": 24892
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4997784463671024,
+      "learning_rate": 7.706300229191956e-06,
+      "loss": 0.7234,
+      "step": 24893
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4878925849446463,
+      "learning_rate": 7.705557538010358e-06,
+      "loss": 0.7448,
+      "step": 24894
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8327840777096827,
+      "learning_rate": 7.704814860187731e-06,
+      "loss": 0.7432,
+      "step": 24895
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.407884676992239,
+      "learning_rate": 7.704072195728396e-06,
+      "loss": 0.6094,
+      "step": 24896
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4463996707672335,
+      "learning_rate": 7.70332954463668e-06,
+      "loss": 0.7747,
+      "step": 24897
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.588066819751138,
+      "learning_rate": 7.702586906916911e-06,
+      "loss": 0.5975,
+      "step": 24898
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4640154125423033,
+      "learning_rate": 7.701844282573405e-06,
+      "loss": 0.7536,
+      "step": 24899
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.400431292026857,
+      "learning_rate": 7.701101671610485e-06,
+      "loss": 0.7183,
+      "step": 24900
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3708077495031576,
+      "learning_rate": 7.700359074032483e-06,
+      "loss": 0.6134,
+      "step": 24901
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5406907205146294,
+      "learning_rate": 7.699616489843723e-06,
+      "loss": 0.6645,
+      "step": 24902
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3257746930922092,
+      "learning_rate": 7.698873919048518e-06,
+      "loss": 0.6557,
+      "step": 24903
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2925647199503543,
+      "learning_rate": 7.698131361651196e-06,
+      "loss": 0.6144,
+      "step": 24904
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2640735247565404,
+      "learning_rate": 7.697388817656087e-06,
+      "loss": 0.7081,
+      "step": 24905
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4259688530649233,
+      "learning_rate": 7.696646287067509e-06,
+      "loss": 0.6663,
+      "step": 24906
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.476716734701903,
+      "learning_rate": 7.695903769889783e-06,
+      "loss": 0.7596,
+      "step": 24907
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7670685647719313,
+      "learning_rate": 7.695161266127234e-06,
+      "loss": 0.6789,
+      "step": 24908
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5492822357517025,
+      "learning_rate": 7.69441877578419e-06,
+      "loss": 0.4953,
+      "step": 24909
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4668739827025945,
+      "learning_rate": 7.693676298864967e-06,
+      "loss": 0.7657,
+      "step": 24910
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4106469779679989,
+      "learning_rate": 7.69293383537389e-06,
+      "loss": 0.812,
+      "step": 24911
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5021640540980896,
+      "learning_rate": 7.692191385315282e-06,
+      "loss": 0.6741,
+      "step": 24912
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4259643481926108,
+      "learning_rate": 7.691448948693469e-06,
+      "loss": 0.6694,
+      "step": 24913
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.288221139519189,
+      "learning_rate": 7.690706525512771e-06,
+      "loss": 0.5809,
+      "step": 24914
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.252908726257258,
+      "learning_rate": 7.689964115777507e-06,
+      "loss": 0.6941,
+      "step": 24915
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3390394898543103,
+      "learning_rate": 7.68922171949201e-06,
+      "loss": 0.6865,
+      "step": 24916
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4932807489680633,
+      "learning_rate": 7.68847933666059e-06,
+      "loss": 0.8425,
+      "step": 24917
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3251653154531724,
+      "learning_rate": 7.687736967287578e-06,
+      "loss": 0.7661,
+      "step": 24918
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2381127748948397,
+      "learning_rate": 7.68699461137729e-06,
+      "loss": 0.711,
+      "step": 24919
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3602350772965353,
+      "learning_rate": 7.686252268934058e-06,
+      "loss": 0.5825,
+      "step": 24920
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3680958678432404,
+      "learning_rate": 7.685509939962191e-06,
+      "loss": 0.7032,
+      "step": 24921
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3896226829498701,
+      "learning_rate": 7.68476762446602e-06,
+      "loss": 0.6938,
+      "step": 24922
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3031249413906187,
+      "learning_rate": 7.684025322449869e-06,
+      "loss": 0.8311,
+      "step": 24923
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4522995301197483,
+      "learning_rate": 7.683283033918052e-06,
+      "loss": 0.6847,
+      "step": 24924
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.821796071869256,
+      "learning_rate": 7.682540758874894e-06,
+      "loss": 0.8096,
+      "step": 24925
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7559608180561908,
+      "learning_rate": 7.681798497324717e-06,
+      "loss": 0.6267,
+      "step": 24926
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2957478443524724,
+      "learning_rate": 7.681056249271848e-06,
+      "loss": 0.683,
+      "step": 24927
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4739009817266377,
+      "learning_rate": 7.680314014720597e-06,
+      "loss": 0.6683,
+      "step": 24928
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4417603068779725,
+      "learning_rate": 7.679571793675295e-06,
+      "loss": 0.8133,
+      "step": 24929
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.548202519366852,
+      "learning_rate": 7.678829586140263e-06,
+      "loss": 0.6983,
+      "step": 24930
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5409201689632885,
+      "learning_rate": 7.678087392119818e-06,
+      "loss": 0.6359,
+      "step": 24931
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4056829005670985,
+      "learning_rate": 7.677345211618281e-06,
+      "loss": 0.649,
+      "step": 24932
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2877381916089172,
+      "learning_rate": 7.676603044639978e-06,
+      "loss": 0.6964,
+      "step": 24933
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5957813323022203,
+      "learning_rate": 7.67586089118923e-06,
+      "loss": 0.7856,
+      "step": 24934
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5188033297397234,
+      "learning_rate": 7.675118751270355e-06,
+      "loss": 0.5475,
+      "step": 24935
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6351595083976767,
+      "learning_rate": 7.674376624887669e-06,
+      "loss": 0.5597,
+      "step": 24936
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1911108632324505,
+      "learning_rate": 7.673634512045506e-06,
+      "loss": 0.7569,
+      "step": 24937
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8039017108343818,
+      "learning_rate": 7.672892412748175e-06,
+      "loss": 0.7437,
+      "step": 24938
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5110993609230183,
+      "learning_rate": 7.672150327000002e-06,
+      "loss": 0.8563,
+      "step": 24939
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.573568250102758,
+      "learning_rate": 7.671408254805304e-06,
+      "loss": 0.6332,
+      "step": 24940
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3450156620954503,
+      "learning_rate": 7.67066619616841e-06,
+      "loss": 0.6862,
+      "step": 24941
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5862065509324423,
+      "learning_rate": 7.669924151093632e-06,
+      "loss": 0.6461,
+      "step": 24942
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3162095217880954,
+      "learning_rate": 7.669182119585292e-06,
+      "loss": 0.6098,
+      "step": 24943
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4996862197780665,
+      "learning_rate": 7.668440101647717e-06,
+      "loss": 0.7115,
+      "step": 24944
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2929351180356907,
+      "learning_rate": 7.667698097285214e-06,
+      "loss": 0.742,
+      "step": 24945
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5285355672283611,
+      "learning_rate": 7.666956106502115e-06,
+      "loss": 0.802,
+      "step": 24946
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.246344810489047,
+      "learning_rate": 7.666214129302737e-06,
+      "loss": 0.6626,
+      "step": 24947
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9768577325112056,
+      "learning_rate": 7.6654721656914e-06,
+      "loss": 0.8239,
+      "step": 24948
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.398844176314326,
+      "learning_rate": 7.664730215672418e-06,
+      "loss": 0.8141,
+      "step": 24949
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.250802734961667,
+      "learning_rate": 7.66398827925012e-06,
+      "loss": 0.6931,
+      "step": 24950
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.419668587702727,
+      "learning_rate": 7.663246356428824e-06,
+      "loss": 0.7176,
+      "step": 24951
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6121440568402314,
+      "learning_rate": 7.662504447212845e-06,
+      "loss": 0.7167,
+      "step": 24952
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3632118905993695,
+      "learning_rate": 7.661762551606501e-06,
+      "loss": 0.7427,
+      "step": 24953
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8295972322135878,
+      "learning_rate": 7.66102066961412e-06,
+      "loss": 0.7876,
+      "step": 24954
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5874501902598456,
+      "learning_rate": 7.66027880124002e-06,
+      "loss": 0.4813,
+      "step": 24955
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3449163200632857,
+      "learning_rate": 7.659536946488514e-06,
+      "loss": 0.7096,
+      "step": 24956
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3368771958331533,
+      "learning_rate": 7.658795105363922e-06,
+      "loss": 0.7558,
+      "step": 24957
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.301532850749822,
+      "learning_rate": 7.658053277870571e-06,
+      "loss": 0.7063,
+      "step": 24958
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.783794815157342,
+      "learning_rate": 7.657311464012774e-06,
+      "loss": 0.6457,
+      "step": 24959
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4551008329923751,
+      "learning_rate": 7.656569663794851e-06,
+      "loss": 0.7619,
+      "step": 24960
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.949436617773049,
+      "learning_rate": 7.65582787722112e-06,
+      "loss": 0.7844,
+      "step": 24961
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.472933844149762,
+      "learning_rate": 7.655086104295904e-06,
+      "loss": 0.7204,
+      "step": 24962
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3552148549447498,
+      "learning_rate": 7.654344345023517e-06,
+      "loss": 0.6934,
+      "step": 24963
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.294765017135556,
+      "learning_rate": 7.653602599408279e-06,
+      "loss": 0.7497,
+      "step": 24964
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4681373956607031,
+      "learning_rate": 7.652860867454515e-06,
+      "loss": 0.6499,
+      "step": 24965
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.435200957564193,
+      "learning_rate": 7.652119149166533e-06,
+      "loss": 0.7099,
+      "step": 24966
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7789057278205667,
+      "learning_rate": 7.651377444548657e-06,
+      "loss": 0.696,
+      "step": 24967
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2988229675746286,
+      "learning_rate": 7.650635753605203e-06,
+      "loss": 0.633,
+      "step": 24968
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1680630451889695,
+      "learning_rate": 7.649894076340499e-06,
+      "loss": 0.734,
+      "step": 24969
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3581712219171829,
+      "learning_rate": 7.649152412758848e-06,
+      "loss": 0.871,
+      "step": 24970
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3664260347636337,
+      "learning_rate": 7.64841076286458e-06,
+      "loss": 0.7023,
+      "step": 24971
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2961907177043523,
+      "learning_rate": 7.64766912666201e-06,
+      "loss": 0.8121,
+      "step": 24972
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3420206456345611,
+      "learning_rate": 7.646927504155452e-06,
+      "loss": 0.6931,
+      "step": 24973
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.314483759061313,
+      "learning_rate": 7.646185895349227e-06,
+      "loss": 0.73,
+      "step": 24974
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.183306041857088,
+      "learning_rate": 7.645444300247654e-06,
+      "loss": 0.6088,
+      "step": 24975
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2251980464683179,
+      "learning_rate": 7.64470271885505e-06,
+      "loss": 0.6619,
+      "step": 24976
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6248859966274913,
+      "learning_rate": 7.64396115117573e-06,
+      "loss": 0.5046,
+      "step": 24977
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4131733680433658,
+      "learning_rate": 7.643219597214017e-06,
+      "loss": 0.7627,
+      "step": 24978
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3629617615018694,
+      "learning_rate": 7.642478056974228e-06,
+      "loss": 0.8253,
+      "step": 24979
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4029926178736767,
+      "learning_rate": 7.641736530460673e-06,
+      "loss": 0.7284,
+      "step": 24980
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5363624922010297,
+      "learning_rate": 7.640995017677674e-06,
+      "loss": 0.628,
+      "step": 24981
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.505837917832128,
+      "learning_rate": 7.640253518629553e-06,
+      "loss": 0.6672,
+      "step": 24982
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4250261918125808,
+      "learning_rate": 7.639512033320623e-06,
+      "loss": 0.7322,
+      "step": 24983
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5923167686436561,
+      "learning_rate": 7.6387705617552e-06,
+      "loss": 0.4934,
+      "step": 24984
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5845501125488837,
+      "learning_rate": 7.6380291039376e-06,
+      "loss": 0.7666,
+      "step": 24985
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.231012662594561,
+      "learning_rate": 7.637287659872147e-06,
+      "loss": 0.7105,
+      "step": 24986
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2922907872335154,
+      "learning_rate": 7.63654622956315e-06,
+      "loss": 0.7196,
+      "step": 24987
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8088614523026663,
+      "learning_rate": 7.63580481301493e-06,
+      "loss": 0.7132,
+      "step": 24988
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3643211108229762,
+      "learning_rate": 7.6350634102318e-06,
+      "loss": 0.8023,
+      "step": 24989
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3866599037300558,
+      "learning_rate": 7.634322021218085e-06,
+      "loss": 0.6055,
+      "step": 24990
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.857783794460783,
+      "learning_rate": 7.633580645978093e-06,
+      "loss": 0.8087,
+      "step": 24991
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5780775070668401,
+      "learning_rate": 7.632839284516144e-06,
+      "loss": 0.5379,
+      "step": 24992
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4817985144444885,
+      "learning_rate": 7.632097936836555e-06,
+      "loss": 0.6577,
+      "step": 24993
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6789254736011605,
+      "learning_rate": 7.63135660294364e-06,
+      "loss": 0.6445,
+      "step": 24994
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5911292091578397,
+      "learning_rate": 7.630615282841717e-06,
+      "loss": 0.7413,
+      "step": 24995
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7014760335699641,
+      "learning_rate": 7.629873976535102e-06,
+      "loss": 0.6273,
+      "step": 24996
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5646361512965604,
+      "learning_rate": 7.629132684028113e-06,
+      "loss": 0.496,
+      "step": 24997
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3310650892788838,
+      "learning_rate": 7.6283914053250605e-06,
+      "loss": 0.7341,
+      "step": 24998
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6019304266583023,
+      "learning_rate": 7.627650140430267e-06,
+      "loss": 0.6466,
+      "step": 24999
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3774691716244474,
+      "learning_rate": 7.626908889348046e-06,
+      "loss": 0.6621,
+      "step": 25000
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5606364555896934,
+      "learning_rate": 7.626167652082712e-06,
+      "loss": 0.7296,
+      "step": 25001
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8737217965218975,
+      "learning_rate": 7.625426428638579e-06,
+      "loss": 0.7045,
+      "step": 25002
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.330914108403429,
+      "learning_rate": 7.624685219019966e-06,
+      "loss": 0.7108,
+      "step": 25003
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4013268525083278,
+      "learning_rate": 7.623944023231191e-06,
+      "loss": 0.6985,
+      "step": 25004
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.327219222038022,
+      "learning_rate": 7.623202841276564e-06,
+      "loss": 0.5717,
+      "step": 25005
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6575189421252743,
+      "learning_rate": 7.6224616731604e-06,
+      "loss": 0.6885,
+      "step": 25006
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4577527086719024,
+      "learning_rate": 7.621720518887022e-06,
+      "loss": 0.7245,
+      "step": 25007
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.221807355014669,
+      "learning_rate": 7.620979378460736e-06,
+      "loss": 0.6475,
+      "step": 25008
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5319291947529565,
+      "learning_rate": 7.620238251885863e-06,
+      "loss": 0.7348,
+      "step": 25009
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6089620210610404,
+      "learning_rate": 7.6194971391667126e-06,
+      "loss": 0.5483,
+      "step": 25010
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3052761105635693,
+      "learning_rate": 7.618756040307609e-06,
+      "loss": 0.7301,
+      "step": 25011
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5857290699284929,
+      "learning_rate": 7.61801495531286e-06,
+      "loss": 0.456,
+      "step": 25012
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3266534003633121,
+      "learning_rate": 7.617273884186779e-06,
+      "loss": 0.7284,
+      "step": 25013
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4667104856625182,
+      "learning_rate": 7.6165328269336905e-06,
+      "loss": 0.7242,
+      "step": 25014
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5063509168620501,
+      "learning_rate": 7.615791783557895e-06,
+      "loss": 0.6172,
+      "step": 25015
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4320672365952154,
+      "learning_rate": 7.615050754063717e-06,
+      "loss": 0.6773,
+      "step": 25016
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5700545902358328,
+      "learning_rate": 7.614309738455468e-06,
+      "loss": 0.5144,
+      "step": 25017
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7105957779954963,
+      "learning_rate": 7.613568736737469e-06,
+      "loss": 0.6449,
+      "step": 25018
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0312958129745033,
+      "learning_rate": 7.612827748914021e-06,
+      "loss": 0.7022,
+      "step": 25019
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4587865780507605,
+      "learning_rate": 7.612086774989449e-06,
+      "loss": 0.7132,
+      "step": 25020
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.589808585270553,
+      "learning_rate": 7.611345814968065e-06,
+      "loss": 0.4821,
+      "step": 25021
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4645475555219254,
+      "learning_rate": 7.610604868854179e-06,
+      "loss": 0.7754,
+      "step": 25022
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3549737998974418,
+      "learning_rate": 7.609863936652106e-06,
+      "loss": 0.7351,
+      "step": 25023
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4097339539088578,
+      "learning_rate": 7.6091230183661655e-06,
+      "loss": 0.7892,
+      "step": 25024
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.324563722123362,
+      "learning_rate": 7.60838211400067e-06,
+      "loss": 0.7248,
+      "step": 25025
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3958517977701088,
+      "learning_rate": 7.607641223559925e-06,
+      "loss": 0.7238,
+      "step": 25026
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.252208251651825,
+      "learning_rate": 7.606900347048254e-06,
+      "loss": 0.6646,
+      "step": 25027
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4803079573064388,
+      "learning_rate": 7.6061594844699645e-06,
+      "loss": 0.741,
+      "step": 25028
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3630626810117141,
+      "learning_rate": 7.6054186358293765e-06,
+      "loss": 0.7603,
+      "step": 25029
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6932914331159379,
+      "learning_rate": 7.604677801130794e-06,
+      "loss": 0.7734,
+      "step": 25030
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7310735419083694,
+      "learning_rate": 7.603936980378539e-06,
+      "loss": 0.6988,
+      "step": 25031
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.438461173220696,
+      "learning_rate": 7.603196173576923e-06,
+      "loss": 0.7066,
+      "step": 25032
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2025755832553782,
+      "learning_rate": 7.602455380730256e-06,
+      "loss": 0.5849,
+      "step": 25033
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.42720818005326,
+      "learning_rate": 7.601714601842849e-06,
+      "loss": 0.6462,
+      "step": 25034
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5725055681717327,
+      "learning_rate": 7.600973836919022e-06,
+      "loss": 0.5088,
+      "step": 25035
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3242905871320005,
+      "learning_rate": 7.600233085963088e-06,
+      "loss": 0.7796,
+      "step": 25036
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5740484108333752,
+      "learning_rate": 7.5994923489793535e-06,
+      "loss": 0.6813,
+      "step": 25037
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.3141769892615365,
+      "learning_rate": 7.598751625972133e-06,
+      "loss": 0.6321,
+      "step": 25038
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.54033200218343,
+      "learning_rate": 7.598010916945744e-06,
+      "loss": 0.8197,
+      "step": 25039
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4585320338004997,
+      "learning_rate": 7.597270221904494e-06,
+      "loss": 0.6243,
+      "step": 25040
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6059210769113533,
+      "learning_rate": 7.596529540852699e-06,
+      "loss": 0.5189,
+      "step": 25041
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3995015744394035,
+      "learning_rate": 7.595788873794666e-06,
+      "loss": 0.7773,
+      "step": 25042
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6693665168584955,
+      "learning_rate": 7.5950482207347165e-06,
+      "loss": 0.6062,
+      "step": 25043
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5711951584940553,
+      "learning_rate": 7.594307581677155e-06,
+      "loss": 0.6846,
+      "step": 25044
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0191847091192594,
+      "learning_rate": 7.5935669566262974e-06,
+      "loss": 0.6619,
+      "step": 25045
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4596147785000195,
+      "learning_rate": 7.592826345586456e-06,
+      "loss": 0.799,
+      "step": 25046
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.290586827803316,
+      "learning_rate": 7.592085748561937e-06,
+      "loss": 0.841,
+      "step": 25047
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.330342386869853,
+      "learning_rate": 7.5913451655570604e-06,
+      "loss": 0.6796,
+      "step": 25048
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4371939957020516,
+      "learning_rate": 7.590604596576134e-06,
+      "loss": 0.7064,
+      "step": 25049
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3350098232052525,
+      "learning_rate": 7.589864041623473e-06,
+      "loss": 0.6552,
+      "step": 25050
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1413145253336263,
+      "learning_rate": 7.589123500703383e-06,
+      "loss": 0.6726,
+      "step": 25051
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.475105300560413,
+      "learning_rate": 7.588382973820181e-06,
+      "loss": 0.7419,
+      "step": 25052
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5825752123649781,
+      "learning_rate": 7.587642460978178e-06,
+      "loss": 0.7265,
+      "step": 25053
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5776606445942944,
+      "learning_rate": 7.586901962181683e-06,
+      "loss": 0.6736,
+      "step": 25054
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4401033192987274,
+      "learning_rate": 7.586161477435006e-06,
+      "loss": 0.8037,
+      "step": 25055
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.870249118568345,
+      "learning_rate": 7.585421006742464e-06,
+      "loss": 0.6259,
+      "step": 25056
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2987809281091727,
+      "learning_rate": 7.584680550108367e-06,
+      "loss": 0.5786,
+      "step": 25057
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3546194481387137,
+      "learning_rate": 7.5839401075370225e-06,
+      "loss": 0.7007,
+      "step": 25058
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3779811004999365,
+      "learning_rate": 7.583199679032741e-06,
+      "loss": 0.7153,
+      "step": 25059
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6269362666279437,
+      "learning_rate": 7.582459264599842e-06,
+      "loss": 0.6933,
+      "step": 25060
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2596546667718638,
+      "learning_rate": 7.581718864242626e-06,
+      "loss": 0.6014,
+      "step": 25061
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3743859193490235,
+      "learning_rate": 7.580978477965408e-06,
+      "loss": 0.7127,
+      "step": 25062
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5871822450522718,
+      "learning_rate": 7.5802381057725e-06,
+      "loss": 0.5319,
+      "step": 25063
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5286473475150089,
+      "learning_rate": 7.579497747668216e-06,
+      "loss": 0.7692,
+      "step": 25064
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.654655789881964,
+      "learning_rate": 7.578757403656859e-06,
+      "loss": 0.6257,
+      "step": 25065
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3725272341180141,
+      "learning_rate": 7.578017073742741e-06,
+      "loss": 0.6899,
+      "step": 25066
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5069190331614586,
+      "learning_rate": 7.577276757930181e-06,
+      "loss": 0.6248,
+      "step": 25067
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.330660564444203,
+      "learning_rate": 7.5765364562234756e-06,
+      "loss": 0.7907,
+      "step": 25068
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3721714242679808,
+      "learning_rate": 7.575796168626944e-06,
+      "loss": 0.6088,
+      "step": 25069
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6536288046710418,
+      "learning_rate": 7.575055895144894e-06,
+      "loss": 0.5272,
+      "step": 25070
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3311492802194669,
+      "learning_rate": 7.574315635781642e-06,
+      "loss": 0.6547,
+      "step": 25071
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5138228446810302,
+      "learning_rate": 7.573575390541486e-06,
+      "loss": 0.7839,
+      "step": 25072
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4921532827577797,
+      "learning_rate": 7.572835159428744e-06,
+      "loss": 0.6475,
+      "step": 25073
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.426656213714112,
+      "learning_rate": 7.572094942447727e-06,
+      "loss": 0.7706,
+      "step": 25074
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.394858629861785,
+      "learning_rate": 7.571354739602736e-06,
+      "loss": 0.6717,
+      "step": 25075
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.271631985128478,
+      "learning_rate": 7.57061455089809e-06,
+      "loss": 0.8486,
+      "step": 25076
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4789904954097772,
+      "learning_rate": 7.569874376338095e-06,
+      "loss": 0.6001,
+      "step": 25077
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3407092751718757,
+      "learning_rate": 7.569134215927063e-06,
+      "loss": 0.8263,
+      "step": 25078
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6013945317185374,
+      "learning_rate": 7.568394069669297e-06,
+      "loss": 0.5184,
+      "step": 25079
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3694355684981057,
+      "learning_rate": 7.567653937569113e-06,
+      "loss": 0.7888,
+      "step": 25080
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.109637021176165,
+      "learning_rate": 7.566913819630819e-06,
+      "loss": 0.6926,
+      "step": 25081
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6942505718618248,
+      "learning_rate": 7.566173715858722e-06,
+      "loss": 0.6735,
+      "step": 25082
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3364820964460906,
+      "learning_rate": 7.565433626257129e-06,
+      "loss": 0.6835,
+      "step": 25083
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9178312978412355,
+      "learning_rate": 7.564693550830355e-06,
+      "loss": 0.571,
+      "step": 25084
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7341005217719907,
+      "learning_rate": 7.563953489582708e-06,
+      "loss": 0.7533,
+      "step": 25085
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8301985991492697,
+      "learning_rate": 7.5632134425184936e-06,
+      "loss": 0.6366,
+      "step": 25086
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.492801076477851,
+      "learning_rate": 7.562473409642019e-06,
+      "loss": 0.8122,
+      "step": 25087
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4514023594298033,
+      "learning_rate": 7.5617333909576006e-06,
+      "loss": 0.6783,
+      "step": 25088
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2798830851064027,
+      "learning_rate": 7.56099338646954e-06,
+      "loss": 0.5428,
+      "step": 25089
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6665177626743481,
+      "learning_rate": 7.560253396182148e-06,
+      "loss": 0.6783,
+      "step": 25090
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4637860630920119,
+      "learning_rate": 7.559513420099731e-06,
+      "loss": 0.6916,
+      "step": 25091
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3122416060139361,
+      "learning_rate": 7.558773458226605e-06,
+      "loss": 0.6099,
+      "step": 25092
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3690480824728826,
+      "learning_rate": 7.558033510567069e-06,
+      "loss": 0.6225,
+      "step": 25093
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.545964709968134,
+      "learning_rate": 7.557293577125435e-06,
+      "loss": 0.7943,
+      "step": 25094
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2602278894200192,
+      "learning_rate": 7.556553657906014e-06,
+      "loss": 0.7171,
+      "step": 25095
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4511008958345446,
+      "learning_rate": 7.555813752913107e-06,
+      "loss": 0.6964,
+      "step": 25096
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3483883241002534,
+      "learning_rate": 7.555073862151028e-06,
+      "loss": 0.6225,
+      "step": 25097
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3706906850932286,
+      "learning_rate": 7.554333985624082e-06,
+      "loss": 0.8119,
+      "step": 25098
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3417056274890933,
+      "learning_rate": 7.553594123336583e-06,
+      "loss": 0.6753,
+      "step": 25099
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4533221302631008,
+      "learning_rate": 7.5528542752928265e-06,
+      "loss": 0.7673,
+      "step": 25100
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6056958195356137,
+      "learning_rate": 7.552114441497131e-06,
+      "loss": 0.8051,
+      "step": 25101
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5881071170798069,
+      "learning_rate": 7.551374621953801e-06,
+      "loss": 0.5195,
+      "step": 25102
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.179807306890279,
+      "learning_rate": 7.550634816667142e-06,
+      "loss": 0.6708,
+      "step": 25103
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6181608241515888,
+      "learning_rate": 7.549895025641459e-06,
+      "loss": 0.739,
+      "step": 25104
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5807294320670896,
+      "learning_rate": 7.549155248881066e-06,
+      "loss": 0.5386,
+      "step": 25105
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1329422691187494,
+      "learning_rate": 7.5484154863902705e-06,
+      "loss": 0.6954,
+      "step": 25106
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6677251198667595,
+      "learning_rate": 7.547675738173372e-06,
+      "loss": 0.7567,
+      "step": 25107
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5643993094303159,
+      "learning_rate": 7.546936004234679e-06,
+      "loss": 0.5051,
+      "step": 25108
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1663603250622454,
+      "learning_rate": 7.546196284578509e-06,
+      "loss": 0.6899,
+      "step": 25109
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.071847088909678,
+      "learning_rate": 7.545456579209156e-06,
+      "loss": 0.648,
+      "step": 25110
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4372671963017956,
+      "learning_rate": 7.5447168881309295e-06,
+      "loss": 0.6678,
+      "step": 25111
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6152295960175593,
+      "learning_rate": 7.543977211348142e-06,
+      "loss": 0.8211,
+      "step": 25112
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4285600598291148,
+      "learning_rate": 7.543237548865099e-06,
+      "loss": 0.5966,
+      "step": 25113
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3816956128693125,
+      "learning_rate": 7.542497900686102e-06,
+      "loss": 0.6506,
+      "step": 25114
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3185729482133242,
+      "learning_rate": 7.5417582668154585e-06,
+      "loss": 0.6436,
+      "step": 25115
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4363602693109592,
+      "learning_rate": 7.5410186472574835e-06,
+      "loss": 0.6643,
+      "step": 25116
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4480338223950433,
+      "learning_rate": 7.540279042016469e-06,
+      "loss": 0.7491,
+      "step": 25117
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7131898675298245,
+      "learning_rate": 7.539539451096732e-06,
+      "loss": 0.588,
+      "step": 25118
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4263978775887762,
+      "learning_rate": 7.538799874502572e-06,
+      "loss": 0.7373,
+      "step": 25119
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2615676156749864,
+      "learning_rate": 7.538060312238304e-06,
+      "loss": 0.6918,
+      "step": 25120
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5246966800003594,
+      "learning_rate": 7.5373207643082245e-06,
+      "loss": 0.7708,
+      "step": 25121
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.380024445524584,
+      "learning_rate": 7.536581230716643e-06,
+      "loss": 0.7618,
+      "step": 25122
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2726496078824538,
+      "learning_rate": 7.535841711467868e-06,
+      "loss": 0.7106,
+      "step": 25123
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.48799404575534,
+      "learning_rate": 7.535102206566199e-06,
+      "loss": 0.6857,
+      "step": 25124
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4588809683579198,
+      "learning_rate": 7.534362716015948e-06,
+      "loss": 0.7542,
+      "step": 25125
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.457665442826108,
+      "learning_rate": 7.533623239821417e-06,
+      "loss": 0.6897,
+      "step": 25126
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2860745862466951,
+      "learning_rate": 7.532883777986915e-06,
+      "loss": 0.6368,
+      "step": 25127
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2359558982820456,
+      "learning_rate": 7.53214433051674e-06,
+      "loss": 0.6799,
+      "step": 25128
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3392604669524555,
+      "learning_rate": 7.531404897415204e-06,
+      "loss": 0.7042,
+      "step": 25129
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.388110575636613,
+      "learning_rate": 7.530665478686613e-06,
+      "loss": 0.8547,
+      "step": 25130
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6131621140374325,
+      "learning_rate": 7.5299260743352675e-06,
+      "loss": 0.5271,
+      "step": 25131
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4359642815788876,
+      "learning_rate": 7.529186684365471e-06,
+      "loss": 0.6176,
+      "step": 25132
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3490110831221724,
+      "learning_rate": 7.528447308781534e-06,
+      "loss": 0.6285,
+      "step": 25133
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.341343184166323,
+      "learning_rate": 7.527707947587764e-06,
+      "loss": 0.6942,
+      "step": 25134
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.500636317347622,
+      "learning_rate": 7.526968600788456e-06,
+      "loss": 0.6437,
+      "step": 25135
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3976645219949146,
+      "learning_rate": 7.526229268387919e-06,
+      "loss": 0.6751,
+      "step": 25136
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5148489662243403,
+      "learning_rate": 7.525489950390461e-06,
+      "loss": 0.6934,
+      "step": 25137
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3005860058596066,
+      "learning_rate": 7.524750646800384e-06,
+      "loss": 0.7385,
+      "step": 25138
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5841345601196917,
+      "learning_rate": 7.524011357621991e-06,
+      "loss": 0.6645,
+      "step": 25139
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4452918579888463,
+      "learning_rate": 7.523272082859587e-06,
+      "loss": 0.7646,
+      "step": 25140
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2740102141713154,
+      "learning_rate": 7.52253282251748e-06,
+      "loss": 0.6954,
+      "step": 25141
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5863683501565939,
+      "learning_rate": 7.52179357659997e-06,
+      "loss": 0.5246,
+      "step": 25142
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4178747109283192,
+      "learning_rate": 7.521054345111361e-06,
+      "loss": 0.709,
+      "step": 25143
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8678205892813602,
+      "learning_rate": 7.520315128055961e-06,
+      "loss": 0.5534,
+      "step": 25144
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5115102286948674,
+      "learning_rate": 7.5195759254380674e-06,
+      "loss": 0.7445,
+      "step": 25145
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.146812279939807,
+      "learning_rate": 7.518836737261991e-06,
+      "loss": 0.603,
+      "step": 25146
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4709245158054718,
+      "learning_rate": 7.51809756353203e-06,
+      "loss": 0.7822,
+      "step": 25147
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2107651703252078,
+      "learning_rate": 7.517358404252497e-06,
+      "loss": 0.6303,
+      "step": 25148
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.980865156529855,
+      "learning_rate": 7.516619259427682e-06,
+      "loss": 0.7239,
+      "step": 25149
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4032025762966434,
+      "learning_rate": 7.5158801290619e-06,
+      "loss": 0.7371,
+      "step": 25150
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3180352371929216,
+      "learning_rate": 7.515141013159447e-06,
+      "loss": 0.8022,
+      "step": 25151
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.040138906696076,
+      "learning_rate": 7.514401911724636e-06,
+      "loss": 0.7973,
+      "step": 25152
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.367925913240349,
+      "learning_rate": 7.513662824761758e-06,
+      "loss": 0.6983,
+      "step": 25153
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.452708299600246,
+      "learning_rate": 7.512923752275124e-06,
+      "loss": 0.5803,
+      "step": 25154
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5248678282429702,
+      "learning_rate": 7.5121846942690376e-06,
+      "loss": 0.6039,
+      "step": 25155
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9118774175149182,
+      "learning_rate": 7.511445650747797e-06,
+      "loss": 0.7048,
+      "step": 25156
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5519169840356943,
+      "learning_rate": 7.5107066217157065e-06,
+      "loss": 0.6108,
+      "step": 25157
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4568177612830118,
+      "learning_rate": 7.509967607177072e-06,
+      "loss": 0.7291,
+      "step": 25158
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.59579880568443,
+      "learning_rate": 7.509228607136196e-06,
+      "loss": 0.5224,
+      "step": 25159
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.348007755172124,
+      "learning_rate": 7.508489621597376e-06,
+      "loss": 0.7975,
+      "step": 25160
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5507699959603924,
+      "learning_rate": 7.50775065056492e-06,
+      "loss": 0.6647,
+      "step": 25161
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5542242670115585,
+      "learning_rate": 7.5070116940431315e-06,
+      "loss": 0.8317,
+      "step": 25162
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5043430223208611,
+      "learning_rate": 7.506272752036308e-06,
+      "loss": 0.6753,
+      "step": 25163
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3116845973817013,
+      "learning_rate": 7.5055338245487506e-06,
+      "loss": 0.5842,
+      "step": 25164
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4259781395432458,
+      "learning_rate": 7.504794911584767e-06,
+      "loss": 0.8673,
+      "step": 25165
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3824760512123007,
+      "learning_rate": 7.504056013148661e-06,
+      "loss": 0.675,
+      "step": 25166
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.381210761753471,
+      "learning_rate": 7.503317129244729e-06,
+      "loss": 0.6416,
+      "step": 25167
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3202367562059332,
+      "learning_rate": 7.502578259877272e-06,
+      "loss": 0.5664,
+      "step": 25168
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.252135985507146,
+      "learning_rate": 7.501839405050601e-06,
+      "loss": 0.5416,
+      "step": 25169
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.476741510570843,
+      "learning_rate": 7.501100564769007e-06,
+      "loss": 0.6977,
+      "step": 25170
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2601267125605624,
+      "learning_rate": 7.500361739036799e-06,
+      "loss": 0.6664,
+      "step": 25171
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5877605040316775,
+      "learning_rate": 7.499622927858272e-06,
+      "loss": 0.559,
+      "step": 25172
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6860376241532224,
+      "learning_rate": 7.498884131237738e-06,
+      "loss": 0.6622,
+      "step": 25173
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4195000470060455,
+      "learning_rate": 7.49814534917949e-06,
+      "loss": 0.6821,
+      "step": 25174
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3380070088425984,
+      "learning_rate": 7.497406581687832e-06,
+      "loss": 0.7339,
+      "step": 25175
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6716014532471186,
+      "learning_rate": 7.496667828767066e-06,
+      "loss": 0.7927,
+      "step": 25176
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.411117840868155,
+      "learning_rate": 7.4959290904214895e-06,
+      "loss": 0.6654,
+      "step": 25177
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7975731591019102,
+      "learning_rate": 7.495190366655409e-06,
+      "loss": 0.6273,
+      "step": 25178
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8977763386065882,
+      "learning_rate": 7.4944516574731215e-06,
+      "loss": 0.7214,
+      "step": 25179
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.841102649971348,
+      "learning_rate": 7.493712962878934e-06,
+      "loss": 0.7293,
+      "step": 25180
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5848895433715329,
+      "learning_rate": 7.492974282877137e-06,
+      "loss": 0.5129,
+      "step": 25181
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4043675532837434,
+      "learning_rate": 7.492235617472041e-06,
+      "loss": 0.6853,
+      "step": 25182
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.489566153378554,
+      "learning_rate": 7.4914969666679435e-06,
+      "loss": 0.7118,
+      "step": 25183
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3707806062231116,
+      "learning_rate": 7.490758330469144e-06,
+      "loss": 0.7236,
+      "step": 25184
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.561769489155198,
+      "learning_rate": 7.490019708879942e-06,
+      "loss": 0.5012,
+      "step": 25185
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.502204093639762,
+      "learning_rate": 7.489281101904642e-06,
+      "loss": 0.5679,
+      "step": 25186
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3297812462185115,
+      "learning_rate": 7.488542509547544e-06,
+      "loss": 0.7073,
+      "step": 25187
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5153170758484156,
+      "learning_rate": 7.487803931812945e-06,
+      "loss": 0.6741,
+      "step": 25188
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.478219753981741,
+      "learning_rate": 7.487065368705144e-06,
+      "loss": 0.6174,
+      "step": 25189
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.604139147842225,
+      "learning_rate": 7.486326820228449e-06,
+      "loss": 0.7075,
+      "step": 25190
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4343672814666677,
+      "learning_rate": 7.485588286387152e-06,
+      "loss": 0.778,
+      "step": 25191
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2510066049508042,
+      "learning_rate": 7.484849767185556e-06,
+      "loss": 0.7754,
+      "step": 25192
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5192421418563657,
+      "learning_rate": 7.48411126262796e-06,
+      "loss": 0.7597,
+      "step": 25193
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7976042257744496,
+      "learning_rate": 7.483372772718668e-06,
+      "loss": 0.7483,
+      "step": 25194
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6582748346733884,
+      "learning_rate": 7.482634297461975e-06,
+      "loss": 0.6709,
+      "step": 25195
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2919201082911056,
+      "learning_rate": 7.48189583686218e-06,
+      "loss": 0.6478,
+      "step": 25196
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4426983357486625,
+      "learning_rate": 7.48115739092359e-06,
+      "loss": 0.6537,
+      "step": 25197
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3607932841122832,
+      "learning_rate": 7.480418959650493e-06,
+      "loss": 0.6609,
+      "step": 25198
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.694093052033766,
+      "learning_rate": 7.4796805430471975e-06,
+      "loss": 0.772,
+      "step": 25199
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.419454526984732,
+      "learning_rate": 7.478942141117998e-06,
+      "loss": 0.7954,
+      "step": 25200
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3630392787987144,
+      "learning_rate": 7.478203753867202e-06,
+      "loss": 0.7458,
+      "step": 25201
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.583150040787388,
+      "learning_rate": 7.477465381299094e-06,
+      "loss": 0.4883,
+      "step": 25202
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4037589340289351,
+      "learning_rate": 7.476727023417985e-06,
+      "loss": 0.6545,
+      "step": 25203
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.204029179643973,
+      "learning_rate": 7.47598868022817e-06,
+      "loss": 0.6499,
+      "step": 25204
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4456246202429712,
+      "learning_rate": 7.475250351733947e-06,
+      "loss": 0.7454,
+      "step": 25205
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3709856413370731,
+      "learning_rate": 7.474512037939614e-06,
+      "loss": 0.7893,
+      "step": 25206
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.372804954055265,
+      "learning_rate": 7.473773738849475e-06,
+      "loss": 0.698,
+      "step": 25207
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6853331490478207,
+      "learning_rate": 7.473035454467825e-06,
+      "loss": 0.6152,
+      "step": 25208
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.33054065666306,
+      "learning_rate": 7.472297184798959e-06,
+      "loss": 0.6225,
+      "step": 25209
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.446640224169685,
+      "learning_rate": 7.471558929847181e-06,
+      "loss": 0.8867,
+      "step": 25210
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.080334093511769,
+      "learning_rate": 7.4708206896167904e-06,
+      "loss": 0.7345,
+      "step": 25211
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3743382107258573,
+      "learning_rate": 7.470082464112079e-06,
+      "loss": 0.8461,
+      "step": 25212
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6813202138242953,
+      "learning_rate": 7.469344253337347e-06,
+      "loss": 0.899,
+      "step": 25213
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7003472612898989,
+      "learning_rate": 7.468606057296896e-06,
+      "loss": 0.6863,
+      "step": 25214
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3680914522243388,
+      "learning_rate": 7.467867875995025e-06,
+      "loss": 0.6058,
+      "step": 25215
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6067992408217497,
+      "learning_rate": 7.467129709436025e-06,
+      "loss": 0.8462,
+      "step": 25216
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.725121085752235,
+      "learning_rate": 7.466391557624197e-06,
+      "loss": 0.7614,
+      "step": 25217
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1477466718596252,
+      "learning_rate": 7.465653420563846e-06,
+      "loss": 0.6684,
+      "step": 25218
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2845640248054062,
+      "learning_rate": 7.464915298259256e-06,
+      "loss": 0.7063,
+      "step": 25219
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4203118206827012,
+      "learning_rate": 7.464177190714734e-06,
+      "loss": 0.6205,
+      "step": 25220
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4824419906850923,
+      "learning_rate": 7.463439097934572e-06,
+      "loss": 0.6865,
+      "step": 25221
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6006679209160262,
+      "learning_rate": 7.462701019923078e-06,
+      "loss": 0.5323,
+      "step": 25222
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3087024128514153,
+      "learning_rate": 7.461962956684538e-06,
+      "loss": 0.7202,
+      "step": 25223
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2815770527616168,
+      "learning_rate": 7.4612249082232524e-06,
+      "loss": 0.7426,
+      "step": 25224
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.344011762804057,
+      "learning_rate": 7.460486874543522e-06,
+      "loss": 0.7213,
+      "step": 25225
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.632529592560049,
+      "learning_rate": 7.459748855649637e-06,
+      "loss": 0.5896,
+      "step": 25226
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3890723157105171,
+      "learning_rate": 7.459010851545901e-06,
+      "loss": 0.7682,
+      "step": 25227
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2538132996915206,
+      "learning_rate": 7.458272862236606e-06,
+      "loss": 0.6141,
+      "step": 25228
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3009686225558945,
+      "learning_rate": 7.457534887726056e-06,
+      "loss": 0.8271,
+      "step": 25229
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4209386642487123,
+      "learning_rate": 7.456796928018537e-06,
+      "loss": 0.7421,
+      "step": 25230
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5621234671080366,
+      "learning_rate": 7.456058983118354e-06,
+      "loss": 0.6022,
+      "step": 25231
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2248460542460315,
+      "learning_rate": 7.4553210530298025e-06,
+      "loss": 0.601,
+      "step": 25232
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5413850039161578,
+      "learning_rate": 7.454583137757176e-06,
+      "loss": 0.6714,
+      "step": 25233
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.391749440479956,
+      "learning_rate": 7.4538452373047694e-06,
+      "loss": 0.7282,
+      "step": 25234
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9016233056900045,
+      "learning_rate": 7.453107351676885e-06,
+      "loss": 0.6403,
+      "step": 25235
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5682143909951398,
+      "learning_rate": 7.452369480877818e-06,
+      "loss": 0.6944,
+      "step": 25236
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6288141474818394,
+      "learning_rate": 7.45163162491186e-06,
+      "loss": 0.6942,
+      "step": 25237
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3138120387235028,
+      "learning_rate": 7.450893783783308e-06,
+      "loss": 0.7788,
+      "step": 25238
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4148582667819882,
+      "learning_rate": 7.4501559574964634e-06,
+      "loss": 0.6336,
+      "step": 25239
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.48159495534932,
+      "learning_rate": 7.449418146055616e-06,
+      "loss": 0.7313,
+      "step": 25240
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4072460686983292,
+      "learning_rate": 7.448680349465063e-06,
+      "loss": 0.7841,
+      "step": 25241
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.607732377450871,
+      "learning_rate": 7.447942567729098e-06,
+      "loss": 0.5235,
+      "step": 25242
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2551403441846674,
+      "learning_rate": 7.447204800852025e-06,
+      "loss": 0.7856,
+      "step": 25243
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6263628295624207,
+      "learning_rate": 7.446467048838131e-06,
+      "loss": 0.5563,
+      "step": 25244
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5377403627094155,
+      "learning_rate": 7.445729311691713e-06,
+      "loss": 0.7947,
+      "step": 25245
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4976588364483596,
+      "learning_rate": 7.444991589417073e-06,
+      "loss": 0.7105,
+      "step": 25246
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5096614339797207,
+      "learning_rate": 7.444253882018495e-06,
+      "loss": 0.7279,
+      "step": 25247
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.715182350389645,
+      "learning_rate": 7.4435161895002815e-06,
+      "loss": 0.8003,
+      "step": 25248
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4354331588429021,
+      "learning_rate": 7.442778511866725e-06,
+      "loss": 0.8192,
+      "step": 25249
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6590603473156837,
+      "learning_rate": 7.442040849122127e-06,
+      "loss": 0.6423,
+      "step": 25250
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4073209071615072,
+      "learning_rate": 7.441303201270772e-06,
+      "loss": 0.7249,
+      "step": 25251
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.7562447280857207,
+      "learning_rate": 7.4405655683169606e-06,
+      "loss": 0.6619,
+      "step": 25252
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2784894151171649,
+      "learning_rate": 7.43982795026499e-06,
+      "loss": 0.6842,
+      "step": 25253
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3323019311719155,
+      "learning_rate": 7.439090347119149e-06,
+      "loss": 0.8273,
+      "step": 25254
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2942201913059275,
+      "learning_rate": 7.438352758883732e-06,
+      "loss": 0.6721,
+      "step": 25255
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2830259784167042,
+      "learning_rate": 7.437615185563038e-06,
+      "loss": 0.6043,
+      "step": 25256
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2472308119782627,
+      "learning_rate": 7.436877627161364e-06,
+      "loss": 0.6152,
+      "step": 25257
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.301015547824134,
+      "learning_rate": 7.4361400836829945e-06,
+      "loss": 0.7759,
+      "step": 25258
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4080538735806667,
+      "learning_rate": 7.43540255513223e-06,
+      "loss": 0.6764,
+      "step": 25259
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4184637810239549,
+      "learning_rate": 7.434665041513367e-06,
+      "loss": 0.6688,
+      "step": 25260
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.482710115251383,
+      "learning_rate": 7.433927542830695e-06,
+      "loss": 0.7778,
+      "step": 25261
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.225831350685286,
+      "learning_rate": 7.433190059088506e-06,
+      "loss": 0.6818,
+      "step": 25262
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.323606469687399,
+      "learning_rate": 7.4324525902910996e-06,
+      "loss": 0.6952,
+      "step": 25263
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5790544356258198,
+      "learning_rate": 7.43171513644277e-06,
+      "loss": 0.8168,
+      "step": 25264
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.467302714215415,
+      "learning_rate": 7.4309776975478046e-06,
+      "loss": 0.6956,
+      "step": 25265
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4215564984606879,
+      "learning_rate": 7.4302402736104985e-06,
+      "loss": 0.6603,
+      "step": 25266
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.398999572298572,
+      "learning_rate": 7.429502864635154e-06,
+      "loss": 0.6808,
+      "step": 25267
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4656160988274785,
+      "learning_rate": 7.428765470626051e-06,
+      "loss": 0.7215,
+      "step": 25268
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.383269839413294,
+      "learning_rate": 7.428028091587491e-06,
+      "loss": 0.678,
+      "step": 25269
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4031964906031962,
+      "learning_rate": 7.427290727523764e-06,
+      "loss": 0.7735,
+      "step": 25270
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.53790969190806,
+      "learning_rate": 7.42655337843917e-06,
+      "loss": 0.7303,
+      "step": 25271
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3893701086608117,
+      "learning_rate": 7.4258160443379935e-06,
+      "loss": 0.7443,
+      "step": 25272
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1721558865732271,
+      "learning_rate": 7.4250787252245316e-06,
+      "loss": 0.7379,
+      "step": 25273
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6241796884674984,
+      "learning_rate": 7.424341421103074e-06,
+      "loss": 0.4914,
+      "step": 25274
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5346885293226697,
+      "learning_rate": 7.423604131977921e-06,
+      "loss": 0.6522,
+      "step": 25275
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6719728481183882,
+      "learning_rate": 7.422866857853358e-06,
+      "loss": 0.71,
+      "step": 25276
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3255708285576946,
+      "learning_rate": 7.422129598733681e-06,
+      "loss": 0.6743,
+      "step": 25277
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.8814097858713295,
+      "learning_rate": 7.421392354623181e-06,
+      "loss": 0.8012,
+      "step": 25278
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3835273852506647,
+      "learning_rate": 7.4206551255261495e-06,
+      "loss": 0.6819,
+      "step": 25279
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.446815621736763,
+      "learning_rate": 7.419917911446881e-06,
+      "loss": 0.6258,
+      "step": 25280
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3992871221108854,
+      "learning_rate": 7.4191807123896686e-06,
+      "loss": 0.6967,
+      "step": 25281
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.411126924745031,
+      "learning_rate": 7.418443528358805e-06,
+      "loss": 0.7127,
+      "step": 25282
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6015843638702778,
+      "learning_rate": 7.417706359358576e-06,
+      "loss": 0.7578,
+      "step": 25283
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6786622592038956,
+      "learning_rate": 7.416969205393279e-06,
+      "loss": 0.6609,
+      "step": 25284
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.263257572421828,
+      "learning_rate": 7.4162320664672084e-06,
+      "loss": 0.7838,
+      "step": 25285
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3461496076007184,
+      "learning_rate": 7.415494942584651e-06,
+      "loss": 0.6876,
+      "step": 25286
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6274630933797893,
+      "learning_rate": 7.414757833749897e-06,
+      "loss": 0.7987,
+      "step": 25287
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3436546316640459,
+      "learning_rate": 7.414020739967245e-06,
+      "loss": 0.7644,
+      "step": 25288
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2038319056914222,
+      "learning_rate": 7.413283661240984e-06,
+      "loss": 0.7212,
+      "step": 25289
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8167155782167361,
+      "learning_rate": 7.412546597575403e-06,
+      "loss": 0.7956,
+      "step": 25290
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2920911754894364,
+      "learning_rate": 7.411809548974792e-06,
+      "loss": 0.6214,
+      "step": 25291
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4643069055026192,
+      "learning_rate": 7.411072515443449e-06,
+      "loss": 0.7205,
+      "step": 25292
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8166918960346914,
+      "learning_rate": 7.410335496985661e-06,
+      "loss": 0.704,
+      "step": 25293
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2731479929744325,
+      "learning_rate": 7.409598493605716e-06,
+      "loss": 0.6603,
+      "step": 25294
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4854875409811146,
+      "learning_rate": 7.408861505307912e-06,
+      "loss": 0.7169,
+      "step": 25295
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.479424593590397,
+      "learning_rate": 7.408124532096539e-06,
+      "loss": 0.6993,
+      "step": 25296
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6810728831850479,
+      "learning_rate": 7.407387573975881e-06,
+      "loss": 0.6612,
+      "step": 25297
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4969114029724353,
+      "learning_rate": 7.406650630950234e-06,
+      "loss": 0.6948,
+      "step": 25298
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4798006533064785,
+      "learning_rate": 7.405913703023894e-06,
+      "loss": 0.6558,
+      "step": 25299
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5800829596666193,
+      "learning_rate": 7.405176790201139e-06,
+      "loss": 0.4904,
+      "step": 25300
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.40823964819215,
+      "learning_rate": 7.404439892486267e-06,
+      "loss": 0.6652,
+      "step": 25301
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5972413306977975,
+      "learning_rate": 7.403703009883567e-06,
+      "loss": 0.8942,
+      "step": 25302
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2844523978112639,
+      "learning_rate": 7.402966142397335e-06,
+      "loss": 0.6575,
+      "step": 25303
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1460512146038715,
+      "learning_rate": 7.402229290031851e-06,
+      "loss": 0.7375,
+      "step": 25304
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4941249643623935,
+      "learning_rate": 7.401492452791414e-06,
+      "loss": 0.603,
+      "step": 25305
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5212726267774557,
+      "learning_rate": 7.4007556306803095e-06,
+      "loss": 0.65,
+      "step": 25306
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4283970093574085,
+      "learning_rate": 7.400018823702827e-06,
+      "loss": 0.6975,
+      "step": 25307
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3264435983667278,
+      "learning_rate": 7.39928203186326e-06,
+      "loss": 0.7041,
+      "step": 25308
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.409075312948229,
+      "learning_rate": 7.3985452551658956e-06,
+      "loss": 0.6033,
+      "step": 25309
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6664620852867271,
+      "learning_rate": 7.397808493615026e-06,
+      "loss": 0.7376,
+      "step": 25310
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5176745787525665,
+      "learning_rate": 7.3970717472149354e-06,
+      "loss": 0.773,
+      "step": 25311
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3696091597369466,
+      "learning_rate": 7.39633501596992e-06,
+      "loss": 0.7778,
+      "step": 25312
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2471312962223384,
+      "learning_rate": 7.3955982998842675e-06,
+      "loss": 0.7379,
+      "step": 25313
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.583766968452764,
+      "learning_rate": 7.394861598962265e-06,
+      "loss": 0.499,
+      "step": 25314
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3764389494694869,
+      "learning_rate": 7.394124913208201e-06,
+      "loss": 0.7871,
+      "step": 25315
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3206034405385332,
+      "learning_rate": 7.393388242626368e-06,
+      "loss": 0.7338,
+      "step": 25316
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.614832693425981,
+      "learning_rate": 7.3926515872210555e-06,
+      "loss": 0.6832,
+      "step": 25317
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3777639276051172,
+      "learning_rate": 7.3919149469965505e-06,
+      "loss": 0.7279,
+      "step": 25318
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5410546173043829,
+      "learning_rate": 7.39117832195714e-06,
+      "loss": 0.6679,
+      "step": 25319
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6158249523247985,
+      "learning_rate": 7.390441712107119e-06,
+      "loss": 0.6672,
+      "step": 25320
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3237798288963096,
+      "learning_rate": 7.389705117450769e-06,
+      "loss": 0.8004,
+      "step": 25321
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.371959226665958,
+      "learning_rate": 7.388968537992385e-06,
+      "loss": 0.7403,
+      "step": 25322
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4469548980186269,
+      "learning_rate": 7.388231973736248e-06,
+      "loss": 0.7772,
+      "step": 25323
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4036897642490658,
+      "learning_rate": 7.387495424686657e-06,
+      "loss": 0.7457,
+      "step": 25324
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.609785437602835,
+      "learning_rate": 7.3867588908478934e-06,
+      "loss": 0.5211,
+      "step": 25325
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.341339692530581,
+      "learning_rate": 7.386022372224246e-06,
+      "loss": 0.7653,
+      "step": 25326
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6010014013100315,
+      "learning_rate": 7.385285868820006e-06,
+      "loss": 0.5314,
+      "step": 25327
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5332447058512257,
+      "learning_rate": 7.384549380639454e-06,
+      "loss": 0.6603,
+      "step": 25328
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5550309803887234,
+      "learning_rate": 7.383812907686888e-06,
+      "loss": 0.7711,
+      "step": 25329
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7397770439183813,
+      "learning_rate": 7.383076449966591e-06,
+      "loss": 0.7611,
+      "step": 25330
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5822763298784108,
+      "learning_rate": 7.382340007482852e-06,
+      "loss": 0.8272,
+      "step": 25331
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5910678286930711,
+      "learning_rate": 7.381603580239956e-06,
+      "loss": 0.517,
+      "step": 25332
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4658971466710538,
+      "learning_rate": 7.380867168242193e-06,
+      "loss": 0.7919,
+      "step": 25333
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5666385036563413,
+      "learning_rate": 7.380130771493854e-06,
+      "loss": 0.6419,
+      "step": 25334
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5910504387372513,
+      "learning_rate": 7.37939438999922e-06,
+      "loss": 0.7308,
+      "step": 25335
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.705547055184291,
+      "learning_rate": 7.3786580237625804e-06,
+      "loss": 0.7292,
+      "step": 25336
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.388581469604457,
+      "learning_rate": 7.377921672788225e-06,
+      "loss": 0.7715,
+      "step": 25337
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6728377083207062,
+      "learning_rate": 7.377185337080443e-06,
+      "loss": 0.6432,
+      "step": 25338
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.218763957329124,
+      "learning_rate": 7.376449016643516e-06,
+      "loss": 0.6427,
+      "step": 25339
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.406771969180829,
+      "learning_rate": 7.375712711481731e-06,
+      "loss": 0.7248,
+      "step": 25340
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.514606838728038,
+      "learning_rate": 7.3749764215993804e-06,
+      "loss": 0.6111,
+      "step": 25341
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4374263947308066,
+      "learning_rate": 7.374240147000748e-06,
+      "loss": 0.6333,
+      "step": 25342
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8448360518551732,
+      "learning_rate": 7.373503887690117e-06,
+      "loss": 0.705,
+      "step": 25343
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3646256302826572,
+      "learning_rate": 7.372767643671781e-06,
+      "loss": 0.7209,
+      "step": 25344
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4072441580141086,
+      "learning_rate": 7.372031414950025e-06,
+      "loss": 0.657,
+      "step": 25345
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2004743499553794,
+      "learning_rate": 7.371295201529133e-06,
+      "loss": 0.6934,
+      "step": 25346
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5340971910213586,
+      "learning_rate": 7.370559003413389e-06,
+      "loss": 0.7011,
+      "step": 25347
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3868371327889424,
+      "learning_rate": 7.369822820607091e-06,
+      "loss": 0.5818,
+      "step": 25348
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3729525730696783,
+      "learning_rate": 7.36908665311451e-06,
+      "loss": 0.7741,
+      "step": 25349
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.337618622530583,
+      "learning_rate": 7.3683505009399415e-06,
+      "loss": 0.6861,
+      "step": 25350
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3551516896614004,
+      "learning_rate": 7.3676143640876675e-06,
+      "loss": 0.6091,
+      "step": 25351
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.51910705786231,
+      "learning_rate": 7.366878242561984e-06,
+      "loss": 0.77,
+      "step": 25352
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5879743996201785,
+      "learning_rate": 7.36614213636716e-06,
+      "loss": 0.7269,
+      "step": 25353
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.9446879554052297,
+      "learning_rate": 7.365406045507493e-06,
+      "loss": 0.5788,
+      "step": 25354
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5309082994977212,
+      "learning_rate": 7.364669969987269e-06,
+      "loss": 0.7047,
+      "step": 25355
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4933349215818876,
+      "learning_rate": 7.363933909810766e-06,
+      "loss": 0.6087,
+      "step": 25356
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4690463896900778,
+      "learning_rate": 7.363197864982277e-06,
+      "loss": 0.7278,
+      "step": 25357
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4967667704047791,
+      "learning_rate": 7.362461835506085e-06,
+      "loss": 0.593,
+      "step": 25358
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7212089307464424,
+      "learning_rate": 7.361725821386476e-06,
+      "loss": 0.7948,
+      "step": 25359
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3268459797469914,
+      "learning_rate": 7.360989822627732e-06,
+      "loss": 0.7117,
+      "step": 25360
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.663946742616798,
+      "learning_rate": 7.3602538392341415e-06,
+      "loss": 0.7246,
+      "step": 25361
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.565644231206758,
+      "learning_rate": 7.359517871209993e-06,
+      "loss": 0.6122,
+      "step": 25362
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5769747489964561,
+      "learning_rate": 7.358781918559563e-06,
+      "loss": 0.5246,
+      "step": 25363
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3824295285874157,
+      "learning_rate": 7.358045981287141e-06,
+      "loss": 0.7069,
+      "step": 25364
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.472757713827412,
+      "learning_rate": 7.357310059397011e-06,
+      "loss": 0.7091,
+      "step": 25365
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.22380673309366,
+      "learning_rate": 7.356574152893463e-06,
+      "loss": 0.7088,
+      "step": 25366
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2576342191166245,
+      "learning_rate": 7.355838261780774e-06,
+      "loss": 0.7014,
+      "step": 25367
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3309316843304746,
+      "learning_rate": 7.355102386063229e-06,
+      "loss": 0.6162,
+      "step": 25368
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2921308782172591,
+      "learning_rate": 7.354366525745121e-06,
+      "loss": 0.646,
+      "step": 25369
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5473082228544361,
+      "learning_rate": 7.3536306808307256e-06,
+      "loss": 0.6761,
+      "step": 25370
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8564065279091124,
+      "learning_rate": 7.35289485132433e-06,
+      "loss": 0.538,
+      "step": 25371
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4218863752205335,
+      "learning_rate": 7.352159037230218e-06,
+      "loss": 0.6201,
+      "step": 25372
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.508038665852874,
+      "learning_rate": 7.351423238552677e-06,
+      "loss": 0.6867,
+      "step": 25373
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6019150551904607,
+      "learning_rate": 7.3506874552959886e-06,
+      "loss": 0.512,
+      "step": 25374
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3510375129054941,
+      "learning_rate": 7.349951687464434e-06,
+      "loss": 0.7365,
+      "step": 25375
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4598772620718978,
+      "learning_rate": 7.349215935062302e-06,
+      "loss": 0.705,
+      "step": 25376
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3565190274318588,
+      "learning_rate": 7.348480198093871e-06,
+      "loss": 0.5024,
+      "step": 25377
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0559050256237756,
+      "learning_rate": 7.34774447656343e-06,
+      "loss": 0.7625,
+      "step": 25378
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5940689726597355,
+      "learning_rate": 7.347008770475259e-06,
+      "loss": 0.6884,
+      "step": 25379
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.579050221313779,
+      "learning_rate": 7.346273079833645e-06,
+      "loss": 0.5453,
+      "step": 25380
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5904839555283077,
+      "learning_rate": 7.3455374046428675e-06,
+      "loss": 0.7485,
+      "step": 25381
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.291660845558241,
+      "learning_rate": 7.344801744907212e-06,
+      "loss": 0.7065,
+      "step": 25382
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.32882205583216,
+      "learning_rate": 7.344066100630963e-06,
+      "loss": 0.6618,
+      "step": 25383
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3230944902525208,
+      "learning_rate": 7.3433304718184004e-06,
+      "loss": 0.7099,
+      "step": 25384
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4892937319559036,
+      "learning_rate": 7.342594858473807e-06,
+      "loss": 0.6219,
+      "step": 25385
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5973006675005738,
+      "learning_rate": 7.34185926060147e-06,
+      "loss": 0.5079,
+      "step": 25386
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4516738889508503,
+      "learning_rate": 7.34112367820567e-06,
+      "loss": 0.6384,
+      "step": 25387
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5244882280039556,
+      "learning_rate": 7.3403881112906905e-06,
+      "loss": 0.77,
+      "step": 25388
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2555807529390417,
+      "learning_rate": 7.3396525598608105e-06,
+      "loss": 0.6847,
+      "step": 25389
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3838032961589861,
+      "learning_rate": 7.338917023920319e-06,
+      "loss": 0.6882,
+      "step": 25390
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.495870405599385,
+      "learning_rate": 7.338181503473493e-06,
+      "loss": 0.6867,
+      "step": 25391
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2306907283520048,
+      "learning_rate": 7.3374459985246146e-06,
+      "loss": 0.6618,
+      "step": 25392
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.436134889673555,
+      "learning_rate": 7.336710509077971e-06,
+      "loss": 0.7675,
+      "step": 25393
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.413378358462905,
+      "learning_rate": 7.335975035137845e-06,
+      "loss": 0.7505,
+      "step": 25394
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4127030711764912,
+      "learning_rate": 7.335239576708513e-06,
+      "loss": 0.7196,
+      "step": 25395
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6372779134772437,
+      "learning_rate": 7.334504133794258e-06,
+      "loss": 0.7444,
+      "step": 25396
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.485487233943382,
+      "learning_rate": 7.333768706399365e-06,
+      "loss": 0.6725,
+      "step": 25397
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5019771742546306,
+      "learning_rate": 7.333033294528118e-06,
+      "loss": 0.8409,
+      "step": 25398
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.467444780360034,
+      "learning_rate": 7.332297898184793e-06,
+      "loss": 0.8212,
+      "step": 25399
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2406934827291303,
+      "learning_rate": 7.331562517373672e-06,
+      "loss": 0.6939,
+      "step": 25400
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4824601968555553,
+      "learning_rate": 7.330827152099045e-06,
+      "loss": 0.6984,
+      "step": 25401
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3471550798806309,
+      "learning_rate": 7.3300918023651805e-06,
+      "loss": 0.8121,
+      "step": 25402
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6627344817382121,
+      "learning_rate": 7.32935646817637e-06,
+      "loss": 0.7474,
+      "step": 25403
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3751613489637506,
+      "learning_rate": 7.32862114953689e-06,
+      "loss": 0.6153,
+      "step": 25404
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6166465165417118,
+      "learning_rate": 7.327885846451026e-06,
+      "loss": 0.5349,
+      "step": 25405
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7621615239777504,
+      "learning_rate": 7.327150558923055e-06,
+      "loss": 0.6224,
+      "step": 25406
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5878900504505222,
+      "learning_rate": 7.32641528695726e-06,
+      "loss": 0.7568,
+      "step": 25407
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2200986086675805,
+      "learning_rate": 7.3256800305579234e-06,
+      "loss": 0.7866,
+      "step": 25408
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5676931217681496,
+      "learning_rate": 7.3249447897293205e-06,
+      "loss": 0.4527,
+      "step": 25409
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3651912887849973,
+      "learning_rate": 7.324209564475737e-06,
+      "loss": 0.7166,
+      "step": 25410
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4052248017017097,
+      "learning_rate": 7.323474354801454e-06,
+      "loss": 0.7081,
+      "step": 25411
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.22914882343976,
+      "learning_rate": 7.322739160710752e-06,
+      "loss": 0.6441,
+      "step": 25412
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4832820516476908,
+      "learning_rate": 7.3220039822079066e-06,
+      "loss": 0.5876,
+      "step": 25413
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3015702370637297,
+      "learning_rate": 7.321268819297204e-06,
+      "loss": 0.7058,
+      "step": 25414
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4818012237937226,
+      "learning_rate": 7.320533671982924e-06,
+      "loss": 0.5925,
+      "step": 25415
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5952107235585575,
+      "learning_rate": 7.319798540269343e-06,
+      "loss": 0.4823,
+      "step": 25416
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2640791925463053,
+      "learning_rate": 7.319063424160742e-06,
+      "loss": 0.6729,
+      "step": 25417
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2788463205383065,
+      "learning_rate": 7.318328323661403e-06,
+      "loss": 0.6682,
+      "step": 25418
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3767442867419226,
+      "learning_rate": 7.31759323877561e-06,
+      "loss": 0.6951,
+      "step": 25419
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5763458243701888,
+      "learning_rate": 7.316858169507635e-06,
+      "loss": 0.5293,
+      "step": 25420
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3163855913566196,
+      "learning_rate": 7.31612311586176e-06,
+      "loss": 0.6035,
+      "step": 25421
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6037911109377048,
+      "learning_rate": 7.315388077842269e-06,
+      "loss": 0.6761,
+      "step": 25422
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4479124274420219,
+      "learning_rate": 7.3146530554534355e-06,
+      "loss": 0.6749,
+      "step": 25423
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4486572071943302,
+      "learning_rate": 7.313918048699544e-06,
+      "loss": 0.7114,
+      "step": 25424
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.41598009535594,
+      "learning_rate": 7.31318305758487e-06,
+      "loss": 0.6352,
+      "step": 25425
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.61718408601979,
+      "learning_rate": 7.3124480821136974e-06,
+      "loss": 0.6601,
+      "step": 25426
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3365798278497703,
+      "learning_rate": 7.3117131222903025e-06,
+      "loss": 0.6118,
+      "step": 25427
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8051058272579161,
+      "learning_rate": 7.3109781781189635e-06,
+      "loss": 0.6576,
+      "step": 25428
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4657021437026958,
+      "learning_rate": 7.310243249603963e-06,
+      "loss": 0.7356,
+      "step": 25429
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5792138604678325,
+      "learning_rate": 7.309508336749576e-06,
+      "loss": 0.7705,
+      "step": 25430
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6039410931800553,
+      "learning_rate": 7.308773439560083e-06,
+      "loss": 0.5173,
+      "step": 25431
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3910649252366385,
+      "learning_rate": 7.308038558039762e-06,
+      "loss": 0.7742,
+      "step": 25432
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4463985352060538,
+      "learning_rate": 7.307303692192899e-06,
+      "loss": 0.7912,
+      "step": 25433
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5918873744629716,
+      "learning_rate": 7.306568842023759e-06,
+      "loss": 0.5348,
+      "step": 25434
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6369806213951215,
+      "learning_rate": 7.305834007536632e-06,
+      "loss": 0.6333,
+      "step": 25435
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5648428606421463,
+      "learning_rate": 7.305099188735793e-06,
+      "loss": 0.4934,
+      "step": 25436
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5011898800438073,
+      "learning_rate": 7.3043643856255175e-06,
+      "loss": 0.72,
+      "step": 25437
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.645719005969206,
+      "learning_rate": 7.303629598210084e-06,
+      "loss": 0.5667,
+      "step": 25438
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3789036750889627,
+      "learning_rate": 7.302894826493775e-06,
+      "loss": 0.7212,
+      "step": 25439
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4249945721812436,
+      "learning_rate": 7.302160070480869e-06,
+      "loss": 0.6785,
+      "step": 25440
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.488240858642051,
+      "learning_rate": 7.301425330175635e-06,
+      "loss": 0.7141,
+      "step": 25441
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2863632156459728,
+      "learning_rate": 7.300690605582361e-06,
+      "loss": 0.5887,
+      "step": 25442
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.280312347011802,
+      "learning_rate": 7.299955896705323e-06,
+      "loss": 0.6215,
+      "step": 25443
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.405767860789229,
+      "learning_rate": 7.299221203548794e-06,
+      "loss": 0.7831,
+      "step": 25444
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.498812927430594,
+      "learning_rate": 7.2984865261170525e-06,
+      "loss": 0.7091,
+      "step": 25445
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3117852406494779,
+      "learning_rate": 7.29775186441438e-06,
+      "loss": 0.6282,
+      "step": 25446
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5438962598596937,
+      "learning_rate": 7.2970172184450525e-06,
+      "loss": 0.7917,
+      "step": 25447
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4955642901790271,
+      "learning_rate": 7.296282588213345e-06,
+      "loss": 0.8769,
+      "step": 25448
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2749506896879192,
+      "learning_rate": 7.295547973723533e-06,
+      "loss": 0.7634,
+      "step": 25449
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4340260451601987,
+      "learning_rate": 7.294813374979906e-06,
+      "loss": 0.7585,
+      "step": 25450
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5851986781873615,
+      "learning_rate": 7.294078791986723e-06,
+      "loss": 0.4845,
+      "step": 25451
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5310223718255946,
+      "learning_rate": 7.293344224748274e-06,
+      "loss": 0.9105,
+      "step": 25452
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3686209270953147,
+      "learning_rate": 7.292609673268829e-06,
+      "loss": 0.7588,
+      "step": 25453
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2456632366889429,
+      "learning_rate": 7.291875137552671e-06,
+      "loss": 0.5651,
+      "step": 25454
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2999094577640782,
+      "learning_rate": 7.291140617604072e-06,
+      "loss": 0.6597,
+      "step": 25455
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9970712814614342,
+      "learning_rate": 7.2904061134273105e-06,
+      "loss": 0.7471,
+      "step": 25456
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5843437632552628,
+      "learning_rate": 7.289671625026664e-06,
+      "loss": 0.7638,
+      "step": 25457
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4234531661633285,
+      "learning_rate": 7.288937152406404e-06,
+      "loss": 0.6847,
+      "step": 25458
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.370005156202294,
+      "learning_rate": 7.28820269557081e-06,
+      "loss": 0.634,
+      "step": 25459
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4764066857695943,
+      "learning_rate": 7.2874682545241604e-06,
+      "loss": 0.6575,
+      "step": 25460
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.274157025678442,
+      "learning_rate": 7.286733829270731e-06,
+      "loss": 0.7729,
+      "step": 25461
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.184010414890066,
+      "learning_rate": 7.285999419814792e-06,
+      "loss": 0.808,
+      "step": 25462
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5666211325243848,
+      "learning_rate": 7.285265026160626e-06,
+      "loss": 0.6955,
+      "step": 25463
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5188229015230266,
+      "learning_rate": 7.284530648312509e-06,
+      "loss": 0.6553,
+      "step": 25464
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6399413419147477,
+      "learning_rate": 7.283796286274711e-06,
+      "loss": 0.6678,
+      "step": 25465
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.450591608277789,
+      "learning_rate": 7.283061940051509e-06,
+      "loss": 0.6364,
+      "step": 25466
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4457569533970887,
+      "learning_rate": 7.2823276096471825e-06,
+      "loss": 0.8286,
+      "step": 25467
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2669669483559591,
+      "learning_rate": 7.281593295066008e-06,
+      "loss": 0.5737,
+      "step": 25468
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3816108520820503,
+      "learning_rate": 7.280858996312255e-06,
+      "loss": 0.7213,
+      "step": 25469
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3597389753574836,
+      "learning_rate": 7.280124713390201e-06,
+      "loss": 0.785,
+      "step": 25470
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7482947197238836,
+      "learning_rate": 7.279390446304125e-06,
+      "loss": 0.6816,
+      "step": 25471
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4248226295596345,
+      "learning_rate": 7.278656195058298e-06,
+      "loss": 0.8037,
+      "step": 25472
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.706617825485471,
+      "learning_rate": 7.277921959656995e-06,
+      "loss": 0.7292,
+      "step": 25473
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5564865707430344,
+      "learning_rate": 7.277187740104492e-06,
+      "loss": 0.739,
+      "step": 25474
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.451050150742389,
+      "learning_rate": 7.276453536405066e-06,
+      "loss": 0.6531,
+      "step": 25475
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.357261439467457,
+      "learning_rate": 7.275719348562989e-06,
+      "loss": 0.6835,
+      "step": 25476
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5908558380038105,
+      "learning_rate": 7.274985176582536e-06,
+      "loss": 0.7549,
+      "step": 25477
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3545785705478415,
+      "learning_rate": 7.274251020467984e-06,
+      "loss": 0.7682,
+      "step": 25478
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6164385646317325,
+      "learning_rate": 7.273516880223601e-06,
+      "loss": 0.5458,
+      "step": 25479
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5909371604695571,
+      "learning_rate": 7.272782755853669e-06,
+      "loss": 0.4922,
+      "step": 25480
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3531318633946343,
+      "learning_rate": 7.272048647362457e-06,
+      "loss": 0.7138,
+      "step": 25481
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.555014821805702,
+      "learning_rate": 7.2713145547542475e-06,
+      "loss": 0.7574,
+      "step": 25482
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5965659797690979,
+      "learning_rate": 7.270580478033302e-06,
+      "loss": 0.4976,
+      "step": 25483
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3263714244245886,
+      "learning_rate": 7.269846417203903e-06,
+      "loss": 0.7302,
+      "step": 25484
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4350576247736537,
+      "learning_rate": 7.269112372270325e-06,
+      "loss": 0.704,
+      "step": 25485
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4136700478290567,
+      "learning_rate": 7.268378343236837e-06,
+      "loss": 0.6931,
+      "step": 25486
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3252124940424084,
+      "learning_rate": 7.267644330107714e-06,
+      "loss": 0.6151,
+      "step": 25487
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2733162997744476,
+      "learning_rate": 7.266910332887231e-06,
+      "loss": 0.6119,
+      "step": 25488
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3572358821727193,
+      "learning_rate": 7.266176351579666e-06,
+      "loss": 0.755,
+      "step": 25489
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8097017551150478,
+      "learning_rate": 7.265442386189281e-06,
+      "loss": 0.7631,
+      "step": 25490
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6035501180402238,
+      "learning_rate": 7.26470843672036e-06,
+      "loss": 0.6952,
+      "step": 25491
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4255150163885073,
+      "learning_rate": 7.263974503177174e-06,
+      "loss": 0.8338,
+      "step": 25492
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3766579407443744,
+      "learning_rate": 7.263240585563991e-06,
+      "loss": 0.6992,
+      "step": 25493
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6270607427890531,
+      "learning_rate": 7.262506683885088e-06,
+      "loss": 0.5003,
+      "step": 25494
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3071222762796304,
+      "learning_rate": 7.26177279814474e-06,
+      "loss": 0.6917,
+      "step": 25495
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.555092161039964,
+      "learning_rate": 7.261038928347217e-06,
+      "loss": 0.7226,
+      "step": 25496
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4590406288311188,
+      "learning_rate": 7.260305074496793e-06,
+      "loss": 0.7723,
+      "step": 25497
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1842892141086203,
+      "learning_rate": 7.259571236597737e-06,
+      "loss": 0.7414,
+      "step": 25498
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3289910380065244,
+      "learning_rate": 7.2588374146543315e-06,
+      "loss": 0.7817,
+      "step": 25499
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4382956960091262,
+      "learning_rate": 7.258103608670835e-06,
+      "loss": 0.6144,
+      "step": 25500
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3368801745659329,
+      "learning_rate": 7.2573698186515316e-06,
+      "loss": 0.6772,
+      "step": 25501
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4621312280970165,
+      "learning_rate": 7.256636044600687e-06,
+      "loss": 0.6835,
+      "step": 25502
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4195118950240975,
+      "learning_rate": 7.255902286522578e-06,
+      "loss": 0.5934,
+      "step": 25503
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7662912593318816,
+      "learning_rate": 7.255168544421473e-06,
+      "loss": 0.8011,
+      "step": 25504
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3040803384175002,
+      "learning_rate": 7.254434818301648e-06,
+      "loss": 0.7968,
+      "step": 25505
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3301277232029591,
+      "learning_rate": 7.253701108167373e-06,
+      "loss": 0.6936,
+      "step": 25506
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5382866673957787,
+      "learning_rate": 7.252967414022915e-06,
+      "loss": 0.775,
+      "step": 25507
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5972250339739648,
+      "learning_rate": 7.252233735872554e-06,
+      "loss": 0.5348,
+      "step": 25508
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1990125686554192,
+      "learning_rate": 7.251500073720558e-06,
+      "loss": 0.7378,
+      "step": 25509
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5813064610808847,
+      "learning_rate": 7.2507664275712e-06,
+      "loss": 0.7375,
+      "step": 25510
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5928925577247175,
+      "learning_rate": 7.250032797428746e-06,
+      "loss": 0.879,
+      "step": 25511
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3312973959840935,
+      "learning_rate": 7.249299183297475e-06,
+      "loss": 0.7476,
+      "step": 25512
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3657728341387854,
+      "learning_rate": 7.248565585181655e-06,
+      "loss": 0.7071,
+      "step": 25513
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6405326921840323,
+      "learning_rate": 7.2478320030855576e-06,
+      "loss": 0.5285,
+      "step": 25514
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3590547585051707,
+      "learning_rate": 7.247098437013449e-06,
+      "loss": 0.6122,
+      "step": 25515
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5861405338137553,
+      "learning_rate": 7.246364886969609e-06,
+      "loss": 0.7235,
+      "step": 25516
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9869026537052479,
+      "learning_rate": 7.245631352958306e-06,
+      "loss": 0.6508,
+      "step": 25517
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4308076339548244,
+      "learning_rate": 7.244897834983806e-06,
+      "loss": 0.7261,
+      "step": 25518
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.369575917503248,
+      "learning_rate": 7.244164333050383e-06,
+      "loss": 0.7861,
+      "step": 25519
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4067771241876628,
+      "learning_rate": 7.243430847162308e-06,
+      "loss": 0.7583,
+      "step": 25520
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7674570010668997,
+      "learning_rate": 7.242697377323854e-06,
+      "loss": 0.8277,
+      "step": 25521
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6354124683995022,
+      "learning_rate": 7.241963923539287e-06,
+      "loss": 0.6114,
+      "step": 25522
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3117884072703323,
+      "learning_rate": 7.241230485812877e-06,
+      "loss": 0.6001,
+      "step": 25523
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.785031511624595,
+      "learning_rate": 7.2404970641489015e-06,
+      "loss": 0.7586,
+      "step": 25524
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6552796364758757,
+      "learning_rate": 7.2397636585516215e-06,
+      "loss": 0.5043,
+      "step": 25525
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5315643192437602,
+      "learning_rate": 7.239030269025311e-06,
+      "loss": 0.782,
+      "step": 25526
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0475298636980224,
+      "learning_rate": 7.238296895574242e-06,
+      "loss": 0.6425,
+      "step": 25527
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4942574903732357,
+      "learning_rate": 7.237563538202685e-06,
+      "loss": 0.7149,
+      "step": 25528
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3860343661163588,
+      "learning_rate": 7.2368301969149054e-06,
+      "loss": 0.7137,
+      "step": 25529
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5346453030109042,
+      "learning_rate": 7.236096871715174e-06,
+      "loss": 0.683,
+      "step": 25530
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4672331917525736,
+      "learning_rate": 7.2353635626077664e-06,
+      "loss": 0.8032,
+      "step": 25531
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5169182796276737,
+      "learning_rate": 7.234630269596943e-06,
+      "loss": 0.6763,
+      "step": 25532
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.527012604164372,
+      "learning_rate": 7.233896992686977e-06,
+      "loss": 0.6205,
+      "step": 25533
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3042446815956723,
+      "learning_rate": 7.233163731882139e-06,
+      "loss": 0.6784,
+      "step": 25534
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3643760576362636,
+      "learning_rate": 7.2324304871867036e-06,
+      "loss": 0.7967,
+      "step": 25535
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7234686843938274,
+      "learning_rate": 7.231697258604926e-06,
+      "loss": 0.6743,
+      "step": 25536
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.359241230275816,
+      "learning_rate": 7.230964046141087e-06,
+      "loss": 0.735,
+      "step": 25537
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3529285425077988,
+      "learning_rate": 7.230230849799454e-06,
+      "loss": 0.7181,
+      "step": 25538
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.493903508705966,
+      "learning_rate": 7.229497669584289e-06,
+      "loss": 0.6877,
+      "step": 25539
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.294637982513332,
+      "learning_rate": 7.228764505499869e-06,
+      "loss": 0.7165,
+      "step": 25540
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5118921818380813,
+      "learning_rate": 7.228031357550458e-06,
+      "loss": 0.8445,
+      "step": 25541
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4376738610230555,
+      "learning_rate": 7.227298225740329e-06,
+      "loss": 0.757,
+      "step": 25542
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3644121950799795,
+      "learning_rate": 7.226565110073743e-06,
+      "loss": 0.809,
+      "step": 25543
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2751215153990594,
+      "learning_rate": 7.225832010554976e-06,
+      "loss": 0.6666,
+      "step": 25544
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5856799292858297,
+      "learning_rate": 7.225098927188294e-06,
+      "loss": 0.642,
+      "step": 25545
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4612817884326692,
+      "learning_rate": 7.2243658599779645e-06,
+      "loss": 0.6658,
+      "step": 25546
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6938949284310376,
+      "learning_rate": 7.223632808928252e-06,
+      "loss": 0.6723,
+      "step": 25547
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4462100317153967,
+      "learning_rate": 7.22289977404343e-06,
+      "loss": 0.6066,
+      "step": 25548
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5126398900355091,
+      "learning_rate": 7.222166755327768e-06,
+      "loss": 0.623,
+      "step": 25549
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4424297673023188,
+      "learning_rate": 7.221433752785527e-06,
+      "loss": 0.6227,
+      "step": 25550
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.33673329191142,
+      "learning_rate": 7.2207007664209785e-06,
+      "loss": 0.6091,
+      "step": 25551
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2563586348598519,
+      "learning_rate": 7.2199677962383915e-06,
+      "loss": 0.7742,
+      "step": 25552
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3355163815798283,
+      "learning_rate": 7.219234842242032e-06,
+      "loss": 0.6171,
+      "step": 25553
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4715478778107025,
+      "learning_rate": 7.2185019044361675e-06,
+      "loss": 0.7602,
+      "step": 25554
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2633758482783064,
+      "learning_rate": 7.217768982825062e-06,
+      "loss": 0.6858,
+      "step": 25555
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3961521961540062,
+      "learning_rate": 7.217036077412992e-06,
+      "loss": 0.6991,
+      "step": 25556
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9177003153391883,
+      "learning_rate": 7.216303188204217e-06,
+      "loss": 0.6651,
+      "step": 25557
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.243503624925762,
+      "learning_rate": 7.215570315203006e-06,
+      "loss": 0.6984,
+      "step": 25558
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6687961667661582,
+      "learning_rate": 7.2148374584136274e-06,
+      "loss": 0.6828,
+      "step": 25559
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2994393011612826,
+      "learning_rate": 7.214104617840343e-06,
+      "loss": 0.5846,
+      "step": 25560
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4076367012253528,
+      "learning_rate": 7.213371793487426e-06,
+      "loss": 0.7494,
+      "step": 25561
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.404241865218689,
+      "learning_rate": 7.212638985359139e-06,
+      "loss": 0.5907,
+      "step": 25562
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4889808088984946,
+      "learning_rate": 7.2119061934597545e-06,
+      "loss": 0.7065,
+      "step": 25563
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5874653730358611,
+      "learning_rate": 7.211173417793531e-06,
+      "loss": 0.4891,
+      "step": 25564
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7398582199653303,
+      "learning_rate": 7.210440658364739e-06,
+      "loss": 0.7532,
+      "step": 25565
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3733291990481948,
+      "learning_rate": 7.209707915177648e-06,
+      "loss": 0.7824,
+      "step": 25566
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6263680636307785,
+      "learning_rate": 7.208975188236518e-06,
+      "loss": 0.6416,
+      "step": 25567
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5950768029441912,
+      "learning_rate": 7.208242477545618e-06,
+      "loss": 0.5065,
+      "step": 25568
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5138100170349449,
+      "learning_rate": 7.2075097831092135e-06,
+      "loss": 0.6938,
+      "step": 25569
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.406050566670668,
+      "learning_rate": 7.206777104931575e-06,
+      "loss": 0.6938,
+      "step": 25570
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6098747870092532,
+      "learning_rate": 7.2060444430169615e-06,
+      "loss": 0.5198,
+      "step": 25571
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3292586077492277,
+      "learning_rate": 7.20531179736964e-06,
+      "loss": 0.7762,
+      "step": 25572
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.443869697444008,
+      "learning_rate": 7.204579167993881e-06,
+      "loss": 0.644,
+      "step": 25573
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.689432004582271,
+      "learning_rate": 7.203846554893947e-06,
+      "loss": 0.5567,
+      "step": 25574
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5978545516441045,
+      "learning_rate": 7.203113958074101e-06,
+      "loss": 0.7259,
+      "step": 25575
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3640104412183471,
+      "learning_rate": 7.202381377538612e-06,
+      "loss": 0.7676,
+      "step": 25576
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4439626042610174,
+      "learning_rate": 7.201648813291748e-06,
+      "loss": 0.6752,
+      "step": 25577
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4764687686126738,
+      "learning_rate": 7.200916265337767e-06,
+      "loss": 0.5547,
+      "step": 25578
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.432113501689136,
+      "learning_rate": 7.200183733680937e-06,
+      "loss": 0.7356,
+      "step": 25579
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7297168277140658,
+      "learning_rate": 7.1994512183255284e-06,
+      "loss": 0.7743,
+      "step": 25580
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3589455306846445,
+      "learning_rate": 7.1987187192757965e-06,
+      "loss": 0.6322,
+      "step": 25581
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.472854535227084,
+      "learning_rate": 7.1979862365360135e-06,
+      "loss": 0.7637,
+      "step": 25582
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6218630159434113,
+      "learning_rate": 7.197253770110438e-06,
+      "loss": 0.5585,
+      "step": 25583
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.577458157406441,
+      "learning_rate": 7.196521320003346e-06,
+      "loss": 0.5882,
+      "step": 25584
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2181092988245505,
+      "learning_rate": 7.195788886218987e-06,
+      "loss": 0.5926,
+      "step": 25585
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7466931149285627,
+      "learning_rate": 7.195056468761637e-06,
+      "loss": 0.6601,
+      "step": 25586
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.657509750520343,
+      "learning_rate": 7.194324067635556e-06,
+      "loss": 0.7088,
+      "step": 25587
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6291490647717037,
+      "learning_rate": 7.193591682845006e-06,
+      "loss": 0.6531,
+      "step": 25588
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6106216863960205,
+      "learning_rate": 7.192859314394255e-06,
+      "loss": 0.5089,
+      "step": 25589
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3251316482577076,
+      "learning_rate": 7.1921269622875655e-06,
+      "loss": 0.7584,
+      "step": 25590
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.459147603761279,
+      "learning_rate": 7.191394626529204e-06,
+      "loss": 0.6632,
+      "step": 25591
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.282142570728344,
+      "learning_rate": 7.190662307123428e-06,
+      "loss": 0.704,
+      "step": 25592
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.26907648973934,
+      "learning_rate": 7.189930004074509e-06,
+      "loss": 0.6188,
+      "step": 25593
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5560862290306399,
+      "learning_rate": 7.189197717386709e-06,
+      "loss": 0.5919,
+      "step": 25594
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4195147437897775,
+      "learning_rate": 7.188465447064286e-06,
+      "loss": 0.752,
+      "step": 25595
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7443143287643446,
+      "learning_rate": 7.187733193111505e-06,
+      "loss": 0.69,
+      "step": 25596
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3216348143688155,
+      "learning_rate": 7.187000955532636e-06,
+      "loss": 0.6729,
+      "step": 25597
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4830233559675607,
+      "learning_rate": 7.186268734331938e-06,
+      "loss": 0.7079,
+      "step": 25598
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5801990687046154,
+      "learning_rate": 7.1855365295136746e-06,
+      "loss": 0.5729,
+      "step": 25599
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3724420920788083,
+      "learning_rate": 7.184804341082104e-06,
+      "loss": 0.6327,
+      "step": 25600
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2535484855038348,
+      "learning_rate": 7.184072169041499e-06,
+      "loss": 0.624,
+      "step": 25601
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.34440464479629,
+      "learning_rate": 7.183340013396116e-06,
+      "loss": 0.8029,
+      "step": 25602
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3410591783043075,
+      "learning_rate": 7.182607874150218e-06,
+      "loss": 0.6218,
+      "step": 25603
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8270797938004777,
+      "learning_rate": 7.181875751308067e-06,
+      "loss": 0.8133,
+      "step": 25604
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.572918641964629,
+      "learning_rate": 7.181143644873932e-06,
+      "loss": 0.7612,
+      "step": 25605
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4775647530505536,
+      "learning_rate": 7.1804115548520695e-06,
+      "loss": 0.7932,
+      "step": 25606
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8151704970446512,
+      "learning_rate": 7.179679481246743e-06,
+      "loss": 0.5902,
+      "step": 25607
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3088845722648206,
+      "learning_rate": 7.178947424062217e-06,
+      "loss": 0.7189,
+      "step": 25608
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5199373099666114,
+      "learning_rate": 7.178215383302749e-06,
+      "loss": 0.7053,
+      "step": 25609
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6153494381737397,
+      "learning_rate": 7.177483358972607e-06,
+      "loss": 0.5003,
+      "step": 25610
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3821590062823816,
+      "learning_rate": 7.17675135107605e-06,
+      "loss": 0.7668,
+      "step": 25611
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7238892210104637,
+      "learning_rate": 7.176019359617341e-06,
+      "loss": 0.7343,
+      "step": 25612
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6289041024714632,
+      "learning_rate": 7.17528738460074e-06,
+      "loss": 0.6139,
+      "step": 25613
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4516392117424695,
+      "learning_rate": 7.17455542603051e-06,
+      "loss": 0.6314,
+      "step": 25614
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.371449543502113,
+      "learning_rate": 7.173823483910916e-06,
+      "loss": 0.6487,
+      "step": 25615
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6911176629229852,
+      "learning_rate": 7.173091558246214e-06,
+      "loss": 0.8524,
+      "step": 25616
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.368488034141007,
+      "learning_rate": 7.172359649040666e-06,
+      "loss": 0.5894,
+      "step": 25617
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5255248590567292,
+      "learning_rate": 7.171627756298536e-06,
+      "loss": 0.7615,
+      "step": 25618
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3720506223567297,
+      "learning_rate": 7.170895880024088e-06,
+      "loss": 0.6944,
+      "step": 25619
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4088787116721997,
+      "learning_rate": 7.170164020221577e-06,
+      "loss": 0.7224,
+      "step": 25620
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3586407362063453,
+      "learning_rate": 7.169432176895266e-06,
+      "loss": 0.7491,
+      "step": 25621
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6131290004211492,
+      "learning_rate": 7.168700350049421e-06,
+      "loss": 0.7564,
+      "step": 25622
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6063848998115045,
+      "learning_rate": 7.1679685396882966e-06,
+      "loss": 0.5535,
+      "step": 25623
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2547441603313372,
+      "learning_rate": 7.167236745816154e-06,
+      "loss": 0.7754,
+      "step": 25624
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2561817714097787,
+      "learning_rate": 7.166504968437258e-06,
+      "loss": 0.7357,
+      "step": 25625
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5349706120344808,
+      "learning_rate": 7.165773207555869e-06,
+      "loss": 0.6854,
+      "step": 25626
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5889544883592636,
+      "learning_rate": 7.165041463176243e-06,
+      "loss": 0.5501,
+      "step": 25627
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3734047352436345,
+      "learning_rate": 7.164309735302643e-06,
+      "loss": 0.6021,
+      "step": 25628
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3120765224715338,
+      "learning_rate": 7.163578023939335e-06,
+      "loss": 0.571,
+      "step": 25629
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.824009846321707,
+      "learning_rate": 7.162846329090566e-06,
+      "loss": 0.7244,
+      "step": 25630
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.578788984554327,
+      "learning_rate": 7.162114650760608e-06,
+      "loss": 0.7322,
+      "step": 25631
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2712247034869237,
+      "learning_rate": 7.161382988953714e-06,
+      "loss": 0.744,
+      "step": 25632
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6729150171825033,
+      "learning_rate": 7.160651343674152e-06,
+      "loss": 0.8433,
+      "step": 25633
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3683542548599772,
+      "learning_rate": 7.159919714926171e-06,
+      "loss": 0.7661,
+      "step": 25634
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.907605653749114,
+      "learning_rate": 7.15918810271404e-06,
+      "loss": 0.6678,
+      "step": 25635
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3235200790254158,
+      "learning_rate": 7.158456507042016e-06,
+      "loss": 0.7941,
+      "step": 25636
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2757407827903087,
+      "learning_rate": 7.157724927914355e-06,
+      "loss": 0.6606,
+      "step": 25637
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.385470718416754,
+      "learning_rate": 7.156993365335321e-06,
+      "loss": 0.6482,
+      "step": 25638
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1191561322878651,
+      "learning_rate": 7.156261819309171e-06,
+      "loss": 0.7173,
+      "step": 25639
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5850825093642764,
+      "learning_rate": 7.1555302898401665e-06,
+      "loss": 0.7262,
+      "step": 25640
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2286327877292509,
+      "learning_rate": 7.154798776932563e-06,
+      "loss": 0.6833,
+      "step": 25641
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3870967803890617,
+      "learning_rate": 7.154067280590622e-06,
+      "loss": 0.6324,
+      "step": 25642
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5947662497935582,
+      "learning_rate": 7.153335800818604e-06,
+      "loss": 0.6112,
+      "step": 25643
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3707416980783116,
+      "learning_rate": 7.1526043376207675e-06,
+      "loss": 0.618,
+      "step": 25644
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3712247533285653,
+      "learning_rate": 7.1518728910013655e-06,
+      "loss": 0.6657,
+      "step": 25645
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.539784004282121,
+      "learning_rate": 7.151141460964663e-06,
+      "loss": 0.6523,
+      "step": 25646
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.397536845254486,
+      "learning_rate": 7.15041004751492e-06,
+      "loss": 0.7302,
+      "step": 25647
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5822195852701421,
+      "learning_rate": 7.1496786506563885e-06,
+      "loss": 0.5399,
+      "step": 25648
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4702360030189838,
+      "learning_rate": 7.148947270393329e-06,
+      "loss": 0.6841,
+      "step": 25649
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4805606699819933,
+      "learning_rate": 7.148215906730004e-06,
+      "loss": 0.6696,
+      "step": 25650
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4788327127273138,
+      "learning_rate": 7.14748455967067e-06,
+      "loss": 0.6662,
+      "step": 25651
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.349793635453453,
+      "learning_rate": 7.146753229219583e-06,
+      "loss": 0.7408,
+      "step": 25652
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5993677587174362,
+      "learning_rate": 7.146021915380999e-06,
+      "loss": 0.5132,
+      "step": 25653
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.589700359462435,
+      "learning_rate": 7.145290618159182e-06,
+      "loss": 0.7199,
+      "step": 25654
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.924596116326934,
+      "learning_rate": 7.1445593375583865e-06,
+      "loss": 0.6644,
+      "step": 25655
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.79901299496012,
+      "learning_rate": 7.14382807358287e-06,
+      "loss": 0.5864,
+      "step": 25656
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7723884790200128,
+      "learning_rate": 7.143096826236888e-06,
+      "loss": 0.6305,
+      "step": 25657
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.365147096162333,
+      "learning_rate": 7.142365595524705e-06,
+      "loss": 0.6569,
+      "step": 25658
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.451512585221386,
+      "learning_rate": 7.141634381450573e-06,
+      "loss": 0.884,
+      "step": 25659
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.269523978854411,
+      "learning_rate": 7.14090318401875e-06,
+      "loss": 0.7176,
+      "step": 25660
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4329012858453343,
+      "learning_rate": 7.1401720032334965e-06,
+      "loss": 0.6001,
+      "step": 25661
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5606017262547653,
+      "learning_rate": 7.139440839099063e-06,
+      "loss": 0.6849,
+      "step": 25662
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3455347199219896,
+      "learning_rate": 7.138709691619712e-06,
+      "loss": 0.6388,
+      "step": 25663
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5411828951580893,
+      "learning_rate": 7.137978560799697e-06,
+      "loss": 0.8664,
+      "step": 25664
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.346441121654002,
+      "learning_rate": 7.137247446643283e-06,
+      "loss": 0.6926,
+      "step": 25665
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6094880576606705,
+      "learning_rate": 7.136516349154714e-06,
+      "loss": 0.5046,
+      "step": 25666
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6648425644384397,
+      "learning_rate": 7.1357852683382565e-06,
+      "loss": 0.6054,
+      "step": 25667
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3820157944683578,
+      "learning_rate": 7.135054204198166e-06,
+      "loss": 0.7022,
+      "step": 25668
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.296663357729733,
+      "learning_rate": 7.134323156738694e-06,
+      "loss": 0.6928,
+      "step": 25669
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.785247424037647,
+      "learning_rate": 7.1335921259640996e-06,
+      "loss": 0.8018,
+      "step": 25670
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3718604293839134,
+      "learning_rate": 7.132861111878639e-06,
+      "loss": 0.7259,
+      "step": 25671
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4117774286329883,
+      "learning_rate": 7.132130114486573e-06,
+      "loss": 0.6323,
+      "step": 25672
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.366220468948176,
+      "learning_rate": 7.131399133792147e-06,
+      "loss": 0.745,
+      "step": 25673
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.56569440510621,
+      "learning_rate": 7.130668169799627e-06,
+      "loss": 0.7654,
+      "step": 25674
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4167428667566067,
+      "learning_rate": 7.129937222513268e-06,
+      "loss": 0.6875,
+      "step": 25675
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7056352756888191,
+      "learning_rate": 7.129206291937319e-06,
+      "loss": 0.7443,
+      "step": 25676
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2746608143674498,
+      "learning_rate": 7.128475378076039e-06,
+      "loss": 0.6688,
+      "step": 25677
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.473700946351154,
+      "learning_rate": 7.127744480933687e-06,
+      "loss": 0.7089,
+      "step": 25678
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6246090568474583,
+      "learning_rate": 7.127013600514517e-06,
+      "loss": 0.7188,
+      "step": 25679
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4340281932724888,
+      "learning_rate": 7.126282736822782e-06,
+      "loss": 0.6987,
+      "step": 25680
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3828997728945955,
+      "learning_rate": 7.125551889862736e-06,
+      "loss": 0.7377,
+      "step": 25681
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2031846146702647,
+      "learning_rate": 7.1248210596386444e-06,
+      "loss": 0.7181,
+      "step": 25682
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2336575821966844,
+      "learning_rate": 7.1240902461547475e-06,
+      "loss": 0.6202,
+      "step": 25683
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4182849398416497,
+      "learning_rate": 7.1233594494153104e-06,
+      "loss": 0.7056,
+      "step": 25684
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2795421732522172,
+      "learning_rate": 7.122628669424583e-06,
+      "loss": 0.6885,
+      "step": 25685
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4428239016714202,
+      "learning_rate": 7.121897906186827e-06,
+      "loss": 0.6183,
+      "step": 25686
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.432179141334115,
+      "learning_rate": 7.12116715970629e-06,
+      "loss": 0.6919,
+      "step": 25687
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6999197213945196,
+      "learning_rate": 7.120436429987228e-06,
+      "loss": 0.8329,
+      "step": 25688
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3077265180907995,
+      "learning_rate": 7.1197057170339e-06,
+      "loss": 0.7024,
+      "step": 25689
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.418091682391296,
+      "learning_rate": 7.118975020850554e-06,
+      "loss": 0.783,
+      "step": 25690
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.282449579373558,
+      "learning_rate": 7.1182443414414495e-06,
+      "loss": 0.6877,
+      "step": 25691
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1542921263608057,
+      "learning_rate": 7.117513678810837e-06,
+      "loss": 0.7566,
+      "step": 25692
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5160407150143633,
+      "learning_rate": 7.116783032962976e-06,
+      "loss": 0.7192,
+      "step": 25693
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3371280721097736,
+      "learning_rate": 7.116052403902112e-06,
+      "loss": 0.68,
+      "step": 25694
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3559858533094735,
+      "learning_rate": 7.115321791632507e-06,
+      "loss": 0.7285,
+      "step": 25695
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6639125015352403,
+      "learning_rate": 7.114591196158413e-06,
+      "loss": 0.8256,
+      "step": 25696
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4412595101906929,
+      "learning_rate": 7.1138606174840806e-06,
+      "loss": 0.6547,
+      "step": 25697
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.381029659322912,
+      "learning_rate": 7.113130055613764e-06,
+      "loss": 0.6785,
+      "step": 25698
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3333754306367753,
+      "learning_rate": 7.11239951055172e-06,
+      "loss": 0.6952,
+      "step": 25699
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.092720283458192,
+      "learning_rate": 7.111668982302202e-06,
+      "loss": 0.5042,
+      "step": 25700
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4145717899309962,
+      "learning_rate": 7.11093847086946e-06,
+      "loss": 0.6267,
+      "step": 25701
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5417685938702061,
+      "learning_rate": 7.110207976257746e-06,
+      "loss": 0.778,
+      "step": 25702
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3818382890591983,
+      "learning_rate": 7.10947749847132e-06,
+      "loss": 0.7528,
+      "step": 25703
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.667328282027442,
+      "learning_rate": 7.108747037514431e-06,
+      "loss": 0.6835,
+      "step": 25704
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6559457081186417,
+      "learning_rate": 7.10801659339133e-06,
+      "loss": 0.7257,
+      "step": 25705
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.369904923335844,
+      "learning_rate": 7.107286166106271e-06,
+      "loss": 0.7133,
+      "step": 25706
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.37216924186831,
+      "learning_rate": 7.106555755663511e-06,
+      "loss": 0.7615,
+      "step": 25707
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8575390035689257,
+      "learning_rate": 7.105825362067299e-06,
+      "loss": 0.7335,
+      "step": 25708
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0139899777276367,
+      "learning_rate": 7.105094985321888e-06,
+      "loss": 0.6817,
+      "step": 25709
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3236708094835348,
+      "learning_rate": 7.10436462543153e-06,
+      "loss": 0.7237,
+      "step": 25710
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3982533964594557,
+      "learning_rate": 7.103634282400475e-06,
+      "loss": 0.6091,
+      "step": 25711
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7997612451830693,
+      "learning_rate": 7.102903956232981e-06,
+      "loss": 0.6288,
+      "step": 25712
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2759618769358883,
+      "learning_rate": 7.102173646933295e-06,
+      "loss": 0.7187,
+      "step": 25713
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7140905997437401,
+      "learning_rate": 7.1014433545056785e-06,
+      "loss": 0.5455,
+      "step": 25714
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.546346495198272,
+      "learning_rate": 7.100713078954369e-06,
+      "loss": 0.8342,
+      "step": 25715
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.363988845903791,
+      "learning_rate": 7.099982820283629e-06,
+      "loss": 0.6697,
+      "step": 25716
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3045155611081445,
+      "learning_rate": 7.099252578497709e-06,
+      "loss": 0.6569,
+      "step": 25717
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4207679860501803,
+      "learning_rate": 7.0985223536008555e-06,
+      "loss": 0.6316,
+      "step": 25718
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.055975124907875,
+      "learning_rate": 7.097792145597322e-06,
+      "loss": 0.7341,
+      "step": 25719
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4283055412254921,
+      "learning_rate": 7.097061954491363e-06,
+      "loss": 0.7237,
+      "step": 25720
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5703877715400134,
+      "learning_rate": 7.096331780287231e-06,
+      "loss": 0.4979,
+      "step": 25721
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3418207296816689,
+      "learning_rate": 7.0956016229891715e-06,
+      "loss": 0.9634,
+      "step": 25722
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4893293533769154,
+      "learning_rate": 7.09487148260144e-06,
+      "loss": 0.7834,
+      "step": 25723
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3717503865756975,
+      "learning_rate": 7.094141359128288e-06,
+      "loss": 0.7013,
+      "step": 25724
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4445320029807036,
+      "learning_rate": 7.093411252573964e-06,
+      "loss": 0.8224,
+      "step": 25725
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.352197316591458,
+      "learning_rate": 7.092681162942717e-06,
+      "loss": 0.7039,
+      "step": 25726
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2557804828431005,
+      "learning_rate": 7.0919510902388034e-06,
+      "loss": 0.7494,
+      "step": 25727
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5128930164505736,
+      "learning_rate": 7.0912210344664736e-06,
+      "loss": 0.7186,
+      "step": 25728
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4467391001863017,
+      "learning_rate": 7.090490995629973e-06,
+      "loss": 0.6564,
+      "step": 25729
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7824637484143757,
+      "learning_rate": 7.089760973733553e-06,
+      "loss": 0.6959,
+      "step": 25730
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.586129615825036,
+      "learning_rate": 7.089030968781472e-06,
+      "loss": 0.8974,
+      "step": 25731
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3960365333729892,
+      "learning_rate": 7.088300980777969e-06,
+      "loss": 0.699,
+      "step": 25732
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5978239484578451,
+      "learning_rate": 7.087571009727301e-06,
+      "loss": 0.7553,
+      "step": 25733
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2069065472777758,
+      "learning_rate": 7.086841055633713e-06,
+      "loss": 0.5982,
+      "step": 25734
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3042901093188466,
+      "learning_rate": 7.086111118501465e-06,
+      "loss": 0.6264,
+      "step": 25735
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.981478116653783,
+      "learning_rate": 7.085381198334797e-06,
+      "loss": 0.6429,
+      "step": 25736
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3488525770107938,
+      "learning_rate": 7.084651295137964e-06,
+      "loss": 0.8172,
+      "step": 25737
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3415044126049644,
+      "learning_rate": 7.0839214089152155e-06,
+      "loss": 0.694,
+      "step": 25738
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5352212115079733,
+      "learning_rate": 7.083191539670795e-06,
+      "loss": 0.6818,
+      "step": 25739
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4529483959474794,
+      "learning_rate": 7.08246168740896e-06,
+      "loss": 0.6894,
+      "step": 25740
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4663207681168662,
+      "learning_rate": 7.081731852133958e-06,
+      "loss": 0.717,
+      "step": 25741
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3417200470876343,
+      "learning_rate": 7.081002033850036e-06,
+      "loss": 0.7598,
+      "step": 25742
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3440482931202082,
+      "learning_rate": 7.080272232561441e-06,
+      "loss": 0.7187,
+      "step": 25743
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.503490022640915,
+      "learning_rate": 7.079542448272428e-06,
+      "loss": 0.7073,
+      "step": 25744
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5024279227936435,
+      "learning_rate": 7.078812680987246e-06,
+      "loss": 0.7746,
+      "step": 25745
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6105175964425996,
+      "learning_rate": 7.0780829307101395e-06,
+      "loss": 0.5245,
+      "step": 25746
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.83507452866841,
+      "learning_rate": 7.077353197445356e-06,
+      "loss": 0.7132,
+      "step": 25747
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.530508860115083,
+      "learning_rate": 7.07662348119715e-06,
+      "loss": 0.7406,
+      "step": 25748
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6571250612166084,
+      "learning_rate": 7.075893781969769e-06,
+      "loss": 0.7172,
+      "step": 25749
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2898392570064166,
+      "learning_rate": 7.07516409976746e-06,
+      "loss": 0.6991,
+      "step": 25750
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5796102167810009,
+      "learning_rate": 7.074434434594469e-06,
+      "loss": 0.7257,
+      "step": 25751
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5825017990979119,
+      "learning_rate": 7.073704786455049e-06,
+      "loss": 0.5057,
+      "step": 25752
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.315896118501072,
+      "learning_rate": 7.072975155353446e-06,
+      "loss": 0.7557,
+      "step": 25753
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5085727413509034,
+      "learning_rate": 7.072245541293908e-06,
+      "loss": 0.7726,
+      "step": 25754
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4710414356739976,
+      "learning_rate": 7.0715159442806815e-06,
+      "loss": 0.6337,
+      "step": 25755
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3748391044846486,
+      "learning_rate": 7.0707863643180204e-06,
+      "loss": 0.6859,
+      "step": 25756
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3920261971619845,
+      "learning_rate": 7.070056801410165e-06,
+      "loss": 0.5527,
+      "step": 25757
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7161739294585285,
+      "learning_rate": 7.069327255561369e-06,
+      "loss": 0.7068,
+      "step": 25758
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5878485800014237,
+      "learning_rate": 7.068597726775879e-06,
+      "loss": 0.8069,
+      "step": 25759
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.978247508156525,
+      "learning_rate": 7.0678682150579355e-06,
+      "loss": 0.6632,
+      "step": 25760
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2815101567210663,
+      "learning_rate": 7.067138720411795e-06,
+      "loss": 0.7403,
+      "step": 25761
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3391213626190033,
+      "learning_rate": 7.0664092428416976e-06,
+      "loss": 0.6575,
+      "step": 25762
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6082148488775422,
+      "learning_rate": 7.065679782351902e-06,
+      "loss": 0.5635,
+      "step": 25763
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5082073725536147,
+      "learning_rate": 7.064950338946641e-06,
+      "loss": 0.718,
+      "step": 25764
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.598600363868127,
+      "learning_rate": 7.064220912630171e-06,
+      "loss": 0.7395,
+      "step": 25765
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6857587200516524,
+      "learning_rate": 7.063491503406737e-06,
+      "loss": 0.7112,
+      "step": 25766
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1155719005253824,
+      "learning_rate": 7.062762111280585e-06,
+      "loss": 0.6125,
+      "step": 25767
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4675431221523507,
+      "learning_rate": 7.062032736255959e-06,
+      "loss": 0.8125,
+      "step": 25768
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2828996045474752,
+      "learning_rate": 7.06130337833711e-06,
+      "loss": 0.6515,
+      "step": 25769
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3934027699215947,
+      "learning_rate": 7.060574037528285e-06,
+      "loss": 0.781,
+      "step": 25770
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5614358781703677,
+      "learning_rate": 7.059844713833726e-06,
+      "loss": 0.6357,
+      "step": 25771
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6423501116097041,
+      "learning_rate": 7.059115407257683e-06,
+      "loss": 0.834,
+      "step": 25772
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.522533364548676,
+      "learning_rate": 7.058386117804401e-06,
+      "loss": 0.6813,
+      "step": 25773
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6388866420336553,
+      "learning_rate": 7.057656845478129e-06,
+      "loss": 0.5355,
+      "step": 25774
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3946005835777295,
+      "learning_rate": 7.056927590283104e-06,
+      "loss": 0.7489,
+      "step": 25775
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3763910734240343,
+      "learning_rate": 7.056198352223585e-06,
+      "loss": 0.739,
+      "step": 25776
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5007183405097777,
+      "learning_rate": 7.055469131303809e-06,
+      "loss": 0.6808,
+      "step": 25777
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5502187139175816,
+      "learning_rate": 7.0547399275280235e-06,
+      "loss": 0.8045,
+      "step": 25778
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.01392775686039,
+      "learning_rate": 7.054010740900472e-06,
+      "loss": 0.7288,
+      "step": 25779
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.03096122238538,
+      "learning_rate": 7.053281571425405e-06,
+      "loss": 0.7262,
+      "step": 25780
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.334664197884973,
+      "learning_rate": 7.052552419107068e-06,
+      "loss": 0.8097,
+      "step": 25781
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4400405531925582,
+      "learning_rate": 7.051823283949702e-06,
+      "loss": 0.7973,
+      "step": 25782
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6101535657287777,
+      "learning_rate": 7.051094165957553e-06,
+      "loss": 0.5089,
+      "step": 25783
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3615154207166587,
+      "learning_rate": 7.05036506513487e-06,
+      "loss": 0.7453,
+      "step": 25784
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.292135807789248,
+      "learning_rate": 7.049635981485894e-06,
+      "loss": 0.5829,
+      "step": 25785
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.340351125737063,
+      "learning_rate": 7.048906915014871e-06,
+      "loss": 0.73,
+      "step": 25786
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2941543712625883,
+      "learning_rate": 7.048177865726044e-06,
+      "loss": 0.6904,
+      "step": 25787
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5892660483819018,
+      "learning_rate": 7.0474488336236645e-06,
+      "loss": 0.4967,
+      "step": 25788
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2989079548567763,
+      "learning_rate": 7.0467198187119705e-06,
+      "loss": 0.657,
+      "step": 25789
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.038908105735711,
+      "learning_rate": 7.045990820995208e-06,
+      "loss": 0.6993,
+      "step": 25790
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5998007823274898,
+      "learning_rate": 7.0452618404776255e-06,
+      "loss": 0.7714,
+      "step": 25791
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.167134299538581,
+      "learning_rate": 7.044532877163459e-06,
+      "loss": 0.6608,
+      "step": 25792
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4638621117719854,
+      "learning_rate": 7.043803931056961e-06,
+      "loss": 0.5512,
+      "step": 25793
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5055122899903912,
+      "learning_rate": 7.04307500216237e-06,
+      "loss": 0.7576,
+      "step": 25794
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.255370356145202,
+      "learning_rate": 7.042346090483937e-06,
+      "loss": 0.5599,
+      "step": 25795
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3675969951423532,
+      "learning_rate": 7.041617196025896e-06,
+      "loss": 0.7609,
+      "step": 25796
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4088275641978227,
+      "learning_rate": 7.040888318792499e-06,
+      "loss": 0.755,
+      "step": 25797
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8581431769691252,
+      "learning_rate": 7.04015945878799e-06,
+      "loss": 0.795,
+      "step": 25798
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2367886908503185,
+      "learning_rate": 7.039430616016605e-06,
+      "loss": 0.7327,
+      "step": 25799
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.610145056002885,
+      "learning_rate": 7.038701790482592e-06,
+      "loss": 0.5263,
+      "step": 25800
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6123838873850281,
+      "learning_rate": 7.037972982190197e-06,
+      "loss": 0.5068,
+      "step": 25801
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.389012279678513,
+      "learning_rate": 7.037244191143662e-06,
+      "loss": 0.7687,
+      "step": 25802
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.559360064245633,
+      "learning_rate": 7.036515417347228e-06,
+      "loss": 0.5735,
+      "step": 25803
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3011696822025893,
+      "learning_rate": 7.035786660805137e-06,
+      "loss": 0.7558,
+      "step": 25804
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4429483549586468,
+      "learning_rate": 7.03505792152164e-06,
+      "loss": 0.7353,
+      "step": 25805
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5204613721161684,
+      "learning_rate": 7.034329199500971e-06,
+      "loss": 0.7694,
+      "step": 25806
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2562377547827628,
+      "learning_rate": 7.033600494747377e-06,
+      "loss": 0.7872,
+      "step": 25807
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3386811063666708,
+      "learning_rate": 7.032871807265097e-06,
+      "loss": 0.7841,
+      "step": 25808
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6456262463636353,
+      "learning_rate": 7.032143137058382e-06,
+      "loss": 0.7252,
+      "step": 25809
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4089853706640902,
+      "learning_rate": 7.031414484131467e-06,
+      "loss": 0.7404,
+      "step": 25810
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.516996691012153,
+      "learning_rate": 7.030685848488595e-06,
+      "loss": 0.6821,
+      "step": 25811
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.237179899447956,
+      "learning_rate": 7.029957230134015e-06,
+      "loss": 0.5843,
+      "step": 25812
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4789606168626168,
+      "learning_rate": 7.029228629071958e-06,
+      "loss": 0.7382,
+      "step": 25813
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6669949797936967,
+      "learning_rate": 7.028500045306676e-06,
+      "loss": 0.7266,
+      "step": 25814
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3246321520235507,
+      "learning_rate": 7.027771478842405e-06,
+      "loss": 0.6739,
+      "step": 25815
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5822021145776246,
+      "learning_rate": 7.027042929683394e-06,
+      "loss": 0.6248,
+      "step": 25816
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.459944245392926,
+      "learning_rate": 7.0263143978338756e-06,
+      "loss": 0.7094,
+      "step": 25817
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4310453002574814,
+      "learning_rate": 7.025585883298099e-06,
+      "loss": 0.6573,
+      "step": 25818
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2933485105802993,
+      "learning_rate": 7.0248573860803035e-06,
+      "loss": 0.69,
+      "step": 25819
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4076558601519262,
+      "learning_rate": 7.024128906184726e-06,
+      "loss": 0.7092,
+      "step": 25820
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3823815838826834,
+      "learning_rate": 7.0234004436156155e-06,
+      "loss": 0.6161,
+      "step": 25821
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6004213818347436,
+      "learning_rate": 7.02267199837721e-06,
+      "loss": 0.4712,
+      "step": 25822
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3512975591107135,
+      "learning_rate": 7.021943570473751e-06,
+      "loss": 0.6739,
+      "step": 25823
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5050864586058295,
+      "learning_rate": 7.021215159909478e-06,
+      "loss": 0.7634,
+      "step": 25824
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5519316485151362,
+      "learning_rate": 7.020486766688633e-06,
+      "loss": 0.6537,
+      "step": 25825
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4044559522116922,
+      "learning_rate": 7.019758390815461e-06,
+      "loss": 0.7179,
+      "step": 25826
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4854846627802942,
+      "learning_rate": 7.019030032294196e-06,
+      "loss": 0.7942,
+      "step": 25827
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3786929979777067,
+      "learning_rate": 7.0183016911290816e-06,
+      "loss": 0.6704,
+      "step": 25828
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2395994050602064,
+      "learning_rate": 7.0175733673243586e-06,
+      "loss": 0.7665,
+      "step": 25829
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.472896332889714,
+      "learning_rate": 7.016845060884271e-06,
+      "loss": 0.5918,
+      "step": 25830
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.425387122591079,
+      "learning_rate": 7.016116771813053e-06,
+      "loss": 0.7437,
+      "step": 25831
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6103408997847348,
+      "learning_rate": 7.015388500114947e-06,
+      "loss": 0.8308,
+      "step": 25832
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6702699429569967,
+      "learning_rate": 7.014660245794197e-06,
+      "loss": 0.8375,
+      "step": 25833
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1773454858824657,
+      "learning_rate": 7.013932008855039e-06,
+      "loss": 0.8328,
+      "step": 25834
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.791069641617884,
+      "learning_rate": 7.013203789301713e-06,
+      "loss": 0.6928,
+      "step": 25835
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3034817597542667,
+      "learning_rate": 7.01247558713846e-06,
+      "loss": 0.6982,
+      "step": 25836
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7950982971646765,
+      "learning_rate": 7.011747402369522e-06,
+      "loss": 0.7047,
+      "step": 25837
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.351377153086777,
+      "learning_rate": 7.011019234999137e-06,
+      "loss": 0.6162,
+      "step": 25838
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4384991970138004,
+      "learning_rate": 7.010291085031542e-06,
+      "loss": 0.8079,
+      "step": 25839
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9282306322073852,
+      "learning_rate": 7.0095629524709826e-06,
+      "loss": 0.7515,
+      "step": 25840
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6843982310445282,
+      "learning_rate": 7.008834837321689e-06,
+      "loss": 0.6104,
+      "step": 25841
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.105067584666137,
+      "learning_rate": 7.0081067395879085e-06,
+      "loss": 0.865,
+      "step": 25842
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5695579545309737,
+      "learning_rate": 7.007378659273878e-06,
+      "loss": 0.6769,
+      "step": 25843
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.45677647388772,
+      "learning_rate": 7.006650596383839e-06,
+      "loss": 0.5775,
+      "step": 25844
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6050701577839137,
+      "learning_rate": 7.005922550922023e-06,
+      "loss": 0.6895,
+      "step": 25845
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.279679366118,
+      "learning_rate": 7.005194522892676e-06,
+      "loss": 0.7866,
+      "step": 25846
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.735014871267933,
+      "learning_rate": 7.004466512300037e-06,
+      "loss": 0.7035,
+      "step": 25847
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.570800197083392,
+      "learning_rate": 7.00373851914834e-06,
+      "loss": 0.5121,
+      "step": 25848
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3858039671720108,
+      "learning_rate": 7.003010543441824e-06,
+      "loss": 0.7059,
+      "step": 25849
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4420401015958684,
+      "learning_rate": 7.002282585184731e-06,
+      "loss": 0.7127,
+      "step": 25850
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4703129252612246,
+      "learning_rate": 7.0015546443813e-06,
+      "loss": 0.7574,
+      "step": 25851
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2892331456307975,
+      "learning_rate": 7.000826721035765e-06,
+      "loss": 0.6166,
+      "step": 25852
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2117704838403833,
+      "learning_rate": 7.000098815152365e-06,
+      "loss": 0.6047,
+      "step": 25853
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5954427006961487,
+      "learning_rate": 6.999370926735342e-06,
+      "loss": 0.5006,
+      "step": 25854
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6409017450793681,
+      "learning_rate": 6.998643055788932e-06,
+      "loss": 0.6111,
+      "step": 25855
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5836824072212374,
+      "learning_rate": 6.997915202317372e-06,
+      "loss": 0.6303,
+      "step": 25856
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3295046853520613,
+      "learning_rate": 6.997187366324895e-06,
+      "loss": 0.6955,
+      "step": 25857
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.369069904643296,
+      "learning_rate": 6.99645954781575e-06,
+      "loss": 0.7015,
+      "step": 25858
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.289975926799341,
+      "learning_rate": 6.9957317467941675e-06,
+      "loss": 0.6353,
+      "step": 25859
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3632564368750635,
+      "learning_rate": 6.995003963264382e-06,
+      "loss": 0.6898,
+      "step": 25860
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2164240373238586,
+      "learning_rate": 6.994276197230643e-06,
+      "loss": 0.6571,
+      "step": 25861
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.484058548756807,
+      "learning_rate": 6.993548448697171e-06,
+      "loss": 0.753,
+      "step": 25862
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7030179749425047,
+      "learning_rate": 6.992820717668215e-06,
+      "loss": 0.6901,
+      "step": 25863
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4610249167202007,
+      "learning_rate": 6.992093004148006e-06,
+      "loss": 0.5924,
+      "step": 25864
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6105091734281232,
+      "learning_rate": 6.991365308140792e-06,
+      "loss": 0.5561,
+      "step": 25865
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4260795804832311,
+      "learning_rate": 6.990637629650794e-06,
+      "loss": 0.647,
+      "step": 25866
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5088846496013637,
+      "learning_rate": 6.989909968682258e-06,
+      "loss": 0.7352,
+      "step": 25867
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2235948289754521,
+      "learning_rate": 6.989182325239423e-06,
+      "loss": 0.687,
+      "step": 25868
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5134652949983838,
+      "learning_rate": 6.988454699326515e-06,
+      "loss": 0.7874,
+      "step": 25869
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4317598142033068,
+      "learning_rate": 6.987727090947781e-06,
+      "loss": 0.6341,
+      "step": 25870
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.423265347976796,
+      "learning_rate": 6.986999500107454e-06,
+      "loss": 0.7728,
+      "step": 25871
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5149583556720867,
+      "learning_rate": 6.9862719268097715e-06,
+      "loss": 0.7913,
+      "step": 25872
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4893559228911284,
+      "learning_rate": 6.985544371058963e-06,
+      "loss": 0.6313,
+      "step": 25873
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.384823406504339,
+      "learning_rate": 6.984816832859271e-06,
+      "loss": 0.6725,
+      "step": 25874
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4269491618391605,
+      "learning_rate": 6.9840893122149346e-06,
+      "loss": 0.7616,
+      "step": 25875
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5241714028153137,
+      "learning_rate": 6.98336180913018e-06,
+      "loss": 0.7347,
+      "step": 25876
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4142846779691838,
+      "learning_rate": 6.982634323609249e-06,
+      "loss": 0.6356,
+      "step": 25877
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4448312387158109,
+      "learning_rate": 6.981906855656376e-06,
+      "loss": 0.7696,
+      "step": 25878
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.473407741395967,
+      "learning_rate": 6.9811794052758e-06,
+      "loss": 0.6408,
+      "step": 25879
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.261119257905848,
+      "learning_rate": 6.980451972471751e-06,
+      "loss": 0.6609,
+      "step": 25880
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.108660268519364,
+      "learning_rate": 6.979724557248464e-06,
+      "loss": 0.6792,
+      "step": 25881
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2921968580764283,
+      "learning_rate": 6.978997159610181e-06,
+      "loss": 0.6644,
+      "step": 25882
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3315291280839343,
+      "learning_rate": 6.978269779561131e-06,
+      "loss": 0.7135,
+      "step": 25883
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2264047589776275,
+      "learning_rate": 6.977542417105552e-06,
+      "loss": 0.6265,
+      "step": 25884
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3602638858525726,
+      "learning_rate": 6.9768150722476744e-06,
+      "loss": 0.5685,
+      "step": 25885
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3045378980188134,
+      "learning_rate": 6.976087744991742e-06,
+      "loss": 0.6403,
+      "step": 25886
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4666682376845521,
+      "learning_rate": 6.975360435341981e-06,
+      "loss": 0.6912,
+      "step": 25887
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2944296040773873,
+      "learning_rate": 6.974633143302629e-06,
+      "loss": 0.678,
+      "step": 25888
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.595229935521759,
+      "learning_rate": 6.973905868877923e-06,
+      "loss": 0.7123,
+      "step": 25889
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1960268335839535,
+      "learning_rate": 6.973178612072092e-06,
+      "loss": 0.6814,
+      "step": 25890
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6219442714465417,
+      "learning_rate": 6.972451372889375e-06,
+      "loss": 0.7243,
+      "step": 25891
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3704218102557957,
+      "learning_rate": 6.971724151334005e-06,
+      "loss": 0.6756,
+      "step": 25892
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3669574469703818,
+      "learning_rate": 6.970996947410217e-06,
+      "loss": 0.8945,
+      "step": 25893
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.38332815209273,
+      "learning_rate": 6.9702697611222394e-06,
+      "loss": 0.5981,
+      "step": 25894
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4407708162331787,
+      "learning_rate": 6.9695425924743125e-06,
+      "loss": 0.7474,
+      "step": 25895
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.363734628869347,
+      "learning_rate": 6.968815441470666e-06,
+      "loss": 0.6891,
+      "step": 25896
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4080837121788778,
+      "learning_rate": 6.968088308115543e-06,
+      "loss": 0.7353,
+      "step": 25897
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4323183611855341,
+      "learning_rate": 6.967361192413163e-06,
+      "loss": 0.726,
+      "step": 25898
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8749442399414973,
+      "learning_rate": 6.966634094367768e-06,
+      "loss": 0.7353,
+      "step": 25899
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4048757333504547,
+      "learning_rate": 6.96590701398359e-06,
+      "loss": 0.7357,
+      "step": 25900
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6219191738028292,
+      "learning_rate": 6.965179951264862e-06,
+      "loss": 0.5079,
+      "step": 25901
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7650889519801163,
+      "learning_rate": 6.964452906215815e-06,
+      "loss": 0.6878,
+      "step": 25902
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.259185588023808,
+      "learning_rate": 6.963725878840686e-06,
+      "loss": 0.5738,
+      "step": 25903
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.321162013606955,
+      "learning_rate": 6.9629988691437086e-06,
+      "loss": 0.7571,
+      "step": 25904
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2390627001824275,
+      "learning_rate": 6.962271877129111e-06,
+      "loss": 0.6133,
+      "step": 25905
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6439394568304547,
+      "learning_rate": 6.961544902801126e-06,
+      "loss": 0.5132,
+      "step": 25906
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4047637688446835,
+      "learning_rate": 6.960817946163993e-06,
+      "loss": 0.6807,
+      "step": 25907
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3965450060814535,
+      "learning_rate": 6.960091007221938e-06,
+      "loss": 0.7429,
+      "step": 25908
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.61465213046676,
+      "learning_rate": 6.959364085979192e-06,
+      "loss": 0.6619,
+      "step": 25909
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6007874139984138,
+      "learning_rate": 6.958637182439996e-06,
+      "loss": 0.5262,
+      "step": 25910
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.305359150253357,
+      "learning_rate": 6.957910296608576e-06,
+      "loss": 0.7541,
+      "step": 25911
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.347718763780368,
+      "learning_rate": 6.957183428489165e-06,
+      "loss": 0.6924,
+      "step": 25912
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3668231350432394,
+      "learning_rate": 6.956456578085994e-06,
+      "loss": 0.5528,
+      "step": 25913
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.245758233551976,
+      "learning_rate": 6.955729745403301e-06,
+      "loss": 0.6693,
+      "step": 25914
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.310366141743436,
+      "learning_rate": 6.955002930445308e-06,
+      "loss": 0.7538,
+      "step": 25915
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5263875179636097,
+      "learning_rate": 6.954276133216254e-06,
+      "loss": 0.7278,
+      "step": 25916
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3962987811048087,
+      "learning_rate": 6.953549353720366e-06,
+      "loss": 0.6098,
+      "step": 25917
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1177677155433945,
+      "learning_rate": 6.952822591961883e-06,
+      "loss": 0.7746,
+      "step": 25918
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2976542946930052,
+      "learning_rate": 6.9520958479450285e-06,
+      "loss": 0.7125,
+      "step": 25919
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2841518599768647,
+      "learning_rate": 6.9513691216740385e-06,
+      "loss": 0.7843,
+      "step": 25920
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4662097090491064,
+      "learning_rate": 6.950642413153144e-06,
+      "loss": 0.6715,
+      "step": 25921
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4023678216335094,
+      "learning_rate": 6.949915722386571e-06,
+      "loss": 0.6571,
+      "step": 25922
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2775769805592714,
+      "learning_rate": 6.949189049378555e-06,
+      "loss": 0.6709,
+      "step": 25923
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4165552331377353,
+      "learning_rate": 6.9484623941333274e-06,
+      "loss": 0.778,
+      "step": 25924
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8986609947268636,
+      "learning_rate": 6.94773575665512e-06,
+      "loss": 0.7662,
+      "step": 25925
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2750402508697587,
+      "learning_rate": 6.947009136948157e-06,
+      "loss": 0.6789,
+      "step": 25926
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.321915952973747,
+      "learning_rate": 6.9462825350166744e-06,
+      "loss": 0.7841,
+      "step": 25927
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5103581646164401,
+      "learning_rate": 6.945555950864904e-06,
+      "loss": 0.789,
+      "step": 25928
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3217459810740484,
+      "learning_rate": 6.944829384497073e-06,
+      "loss": 0.7646,
+      "step": 25929
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.448554555732999,
+      "learning_rate": 6.94410283591741e-06,
+      "loss": 0.759,
+      "step": 25930
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5889607336683245,
+      "learning_rate": 6.943376305130149e-06,
+      "loss": 0.6818,
+      "step": 25931
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3096401720362607,
+      "learning_rate": 6.942649792139522e-06,
+      "loss": 0.5929,
+      "step": 25932
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8004358053238336,
+      "learning_rate": 6.941923296949753e-06,
+      "loss": 0.5888,
+      "step": 25933
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.482295677211041,
+      "learning_rate": 6.941196819565074e-06,
+      "loss": 0.7373,
+      "step": 25934
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4471948657397513,
+      "learning_rate": 6.940470359989719e-06,
+      "loss": 0.6484,
+      "step": 25935
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.310790669742609,
+      "learning_rate": 6.939743918227911e-06,
+      "loss": 0.7238,
+      "step": 25936
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.60591838721274,
+      "learning_rate": 6.939017494283884e-06,
+      "loss": 0.7099,
+      "step": 25937
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.639881301308049,
+      "learning_rate": 6.938291088161864e-06,
+      "loss": 0.6841,
+      "step": 25938
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.097800199337968,
+      "learning_rate": 6.937564699866086e-06,
+      "loss": 0.6974,
+      "step": 25939
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4195736087861697,
+      "learning_rate": 6.936838329400774e-06,
+      "loss": 0.7942,
+      "step": 25940
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.463565151544594,
+      "learning_rate": 6.93611197677016e-06,
+      "loss": 0.7773,
+      "step": 25941
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.437805159506806,
+      "learning_rate": 6.9353856419784734e-06,
+      "loss": 0.7772,
+      "step": 25942
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.522296622924194,
+      "learning_rate": 6.934659325029938e-06,
+      "loss": 0.636,
+      "step": 25943
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3695279989240348,
+      "learning_rate": 6.933933025928788e-06,
+      "loss": 0.705,
+      "step": 25944
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2902230818743567,
+      "learning_rate": 6.933206744679249e-06,
+      "loss": 0.8163,
+      "step": 25945
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.199760120662645,
+      "learning_rate": 6.932480481285558e-06,
+      "loss": 0.6981,
+      "step": 25946
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3660535822947228,
+      "learning_rate": 6.931754235751929e-06,
+      "loss": 0.7275,
+      "step": 25947
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4810168131385373,
+      "learning_rate": 6.931028008082602e-06,
+      "loss": 0.8086,
+      "step": 25948
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.263714682524276,
+      "learning_rate": 6.930301798281803e-06,
+      "loss": 0.783,
+      "step": 25949
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2675540507745093,
+      "learning_rate": 6.929575606353757e-06,
+      "loss": 0.7325,
+      "step": 25950
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4177774985756526,
+      "learning_rate": 6.928849432302691e-06,
+      "loss": 0.7364,
+      "step": 25951
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.343457510760118,
+      "learning_rate": 6.9281232761328384e-06,
+      "loss": 0.7656,
+      "step": 25952
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2740594717139846,
+      "learning_rate": 6.927397137848427e-06,
+      "loss": 0.7375,
+      "step": 25953
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.331988946526613,
+      "learning_rate": 6.926671017453678e-06,
+      "loss": 0.6675,
+      "step": 25954
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2459104471435898,
+      "learning_rate": 6.925944914952825e-06,
+      "loss": 0.7567,
+      "step": 25955
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4708141330267688,
+      "learning_rate": 6.925218830350097e-06,
+      "loss": 0.758,
+      "step": 25956
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.292969768894632,
+      "learning_rate": 6.924492763649715e-06,
+      "loss": 0.6927,
+      "step": 25957
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3163250998649612,
+      "learning_rate": 6.923766714855909e-06,
+      "loss": 0.6557,
+      "step": 25958
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5278057801026466,
+      "learning_rate": 6.923040683972908e-06,
+      "loss": 0.888,
+      "step": 25959
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3910600210325454,
+      "learning_rate": 6.922314671004942e-06,
+      "loss": 0.7527,
+      "step": 25960
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4408965765023711,
+      "learning_rate": 6.92158867595623e-06,
+      "loss": 0.7895,
+      "step": 25961
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4953701892415074,
+      "learning_rate": 6.920862698831002e-06,
+      "loss": 0.7589,
+      "step": 25962
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5990368037196762,
+      "learning_rate": 6.920136739633494e-06,
+      "loss": 0.5389,
+      "step": 25963
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3710515686108087,
+      "learning_rate": 6.919410798367917e-06,
+      "loss": 0.7594,
+      "step": 25964
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4243154324687444,
+      "learning_rate": 6.9186848750385085e-06,
+      "loss": 0.7696,
+      "step": 25965
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.317066573769582,
+      "learning_rate": 6.91795896964949e-06,
+      "loss": 0.6723,
+      "step": 25966
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4961312091730552,
+      "learning_rate": 6.917233082205094e-06,
+      "loss": 0.8179,
+      "step": 25967
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4110118038706643,
+      "learning_rate": 6.916507212709541e-06,
+      "loss": 0.7647,
+      "step": 25968
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.434408775675045,
+      "learning_rate": 6.91578136116706e-06,
+      "loss": 0.7747,
+      "step": 25969
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6200059078065001,
+      "learning_rate": 6.915055527581878e-06,
+      "loss": 0.629,
+      "step": 25970
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2724041719699415,
+      "learning_rate": 6.914329711958214e-06,
+      "loss": 0.7316,
+      "step": 25971
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3043117453003286,
+      "learning_rate": 6.913603914300302e-06,
+      "loss": 0.7831,
+      "step": 25972
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4382067128597529,
+      "learning_rate": 6.912878134612367e-06,
+      "loss": 0.7673,
+      "step": 25973
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5540071924248025,
+      "learning_rate": 6.912152372898633e-06,
+      "loss": 0.7688,
+      "step": 25974
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6137135252943315,
+      "learning_rate": 6.911426629163323e-06,
+      "loss": 0.461,
+      "step": 25975
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6011448582663022,
+      "learning_rate": 6.9107009034106655e-06,
+      "loss": 0.6599,
+      "step": 25976
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.519944308978003,
+      "learning_rate": 6.909975195644887e-06,
+      "loss": 0.611,
+      "step": 25977
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4125922419673462,
+      "learning_rate": 6.9092495058702105e-06,
+      "loss": 0.746,
+      "step": 25978
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4578611155342198,
+      "learning_rate": 6.9085238340908585e-06,
+      "loss": 0.6319,
+      "step": 25979
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4458319352415863,
+      "learning_rate": 6.907798180311063e-06,
+      "loss": 0.7329,
+      "step": 25980
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6075537765143229,
+      "learning_rate": 6.907072544535048e-06,
+      "loss": 0.4958,
+      "step": 25981
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6149248905038622,
+      "learning_rate": 6.906346926767031e-06,
+      "loss": 0.7223,
+      "step": 25982
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2910603222430679,
+      "learning_rate": 6.905621327011241e-06,
+      "loss": 0.7694,
+      "step": 25983
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3826912987352613,
+      "learning_rate": 6.904895745271909e-06,
+      "loss": 0.6954,
+      "step": 25984
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3136042712856921,
+      "learning_rate": 6.904170181553251e-06,
+      "loss": 0.7162,
+      "step": 25985
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4921213305995071,
+      "learning_rate": 6.903444635859493e-06,
+      "loss": 0.6089,
+      "step": 25986
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3865128933117652,
+      "learning_rate": 6.902719108194859e-06,
+      "loss": 0.7759,
+      "step": 25987
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.908766656690389,
+      "learning_rate": 6.90199359856358e-06,
+      "loss": 0.7634,
+      "step": 25988
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5769077575211232,
+      "learning_rate": 6.901268106969872e-06,
+      "loss": 0.7488,
+      "step": 25989
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.743870921071418,
+      "learning_rate": 6.9005426334179635e-06,
+      "loss": 0.6784,
+      "step": 25990
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5800938920079212,
+      "learning_rate": 6.899817177912078e-06,
+      "loss": 0.5159,
+      "step": 25991
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5202822576038098,
+      "learning_rate": 6.899091740456434e-06,
+      "loss": 0.6833,
+      "step": 25992
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3251953236346923,
+      "learning_rate": 6.898366321055263e-06,
+      "loss": 0.7725,
+      "step": 25993
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3944379018384179,
+      "learning_rate": 6.897640919712782e-06,
+      "loss": 0.7785,
+      "step": 25994
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5249570862840833,
+      "learning_rate": 6.896915536433224e-06,
+      "loss": 0.6788,
+      "step": 25995
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6330486503038613,
+      "learning_rate": 6.8961901712208005e-06,
+      "loss": 0.7846,
+      "step": 25996
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4271393620146637,
+      "learning_rate": 6.895464824079743e-06,
+      "loss": 0.7775,
+      "step": 25997
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5581903579652514,
+      "learning_rate": 6.894739495014273e-06,
+      "loss": 0.7398,
+      "step": 25998
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6024627500043775,
+      "learning_rate": 6.894014184028611e-06,
+      "loss": 0.5021,
+      "step": 25999
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2556000748393619,
+      "learning_rate": 6.893288891126981e-06,
+      "loss": 0.6491,
+      "step": 26000
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6410449435186902,
+      "learning_rate": 6.892563616313608e-06,
+      "loss": 0.6388,
+      "step": 26001
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4642865542216148,
+      "learning_rate": 6.891838359592714e-06,
+      "loss": 0.6673,
+      "step": 26002
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.283808110546368,
+      "learning_rate": 6.891113120968518e-06,
+      "loss": 0.7495,
+      "step": 26003
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6043153455493645,
+      "learning_rate": 6.890387900445248e-06,
+      "loss": 0.5312,
+      "step": 26004
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5141601098502533,
+      "learning_rate": 6.8896626980271265e-06,
+      "loss": 0.5905,
+      "step": 26005
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.716756767683206,
+      "learning_rate": 6.8889375137183714e-06,
+      "loss": 0.6817,
+      "step": 26006
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3795346107607813,
+      "learning_rate": 6.888212347523204e-06,
+      "loss": 0.6401,
+      "step": 26007
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5079157605461413,
+      "learning_rate": 6.887487199445852e-06,
+      "loss": 0.6762,
+      "step": 26008
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3976967867455408,
+      "learning_rate": 6.886762069490538e-06,
+      "loss": 0.6713,
+      "step": 26009
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.273382782011915,
+      "learning_rate": 6.886036957661477e-06,
+      "loss": 0.7247,
+      "step": 26010
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4670172074679066,
+      "learning_rate": 6.885311863962894e-06,
+      "loss": 0.5646,
+      "step": 26011
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9971946826957945,
+      "learning_rate": 6.884586788399017e-06,
+      "loss": 0.663,
+      "step": 26012
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4150734780872092,
+      "learning_rate": 6.883861730974057e-06,
+      "loss": 0.8637,
+      "step": 26013
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4304250521074657,
+      "learning_rate": 6.883136691692241e-06,
+      "loss": 0.6207,
+      "step": 26014
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.72780805115164,
+      "learning_rate": 6.882411670557788e-06,
+      "loss": 0.7167,
+      "step": 26015
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0872003933597827,
+      "learning_rate": 6.881686667574926e-06,
+      "loss": 0.6587,
+      "step": 26016
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4127167155281433,
+      "learning_rate": 6.8809616827478695e-06,
+      "loss": 0.7081,
+      "step": 26017
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5856211765718214,
+      "learning_rate": 6.880236716080841e-06,
+      "loss": 0.4917,
+      "step": 26018
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4686387023589342,
+      "learning_rate": 6.879511767578061e-06,
+      "loss": 0.8137,
+      "step": 26019
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4707283922941194,
+      "learning_rate": 6.878786837243755e-06,
+      "loss": 0.7962,
+      "step": 26020
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.541646976482345,
+      "learning_rate": 6.878061925082138e-06,
+      "loss": 0.7202,
+      "step": 26021
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6539267574791083,
+      "learning_rate": 6.877337031097433e-06,
+      "loss": 0.7074,
+      "step": 26022
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3765180442547182,
+      "learning_rate": 6.8766121552938625e-06,
+      "loss": 0.7779,
+      "step": 26023
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7048012856941932,
+      "learning_rate": 6.8758872976756406e-06,
+      "loss": 0.719,
+      "step": 26024
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3941096921312635,
+      "learning_rate": 6.875162458246994e-06,
+      "loss": 0.766,
+      "step": 26025
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5157859542768777,
+      "learning_rate": 6.874437637012141e-06,
+      "loss": 0.73,
+      "step": 26026
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4901560113752041,
+      "learning_rate": 6.873712833975304e-06,
+      "loss": 0.6999,
+      "step": 26027
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6278252923829712,
+      "learning_rate": 6.872988049140695e-06,
+      "loss": 0.7857,
+      "step": 26028
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.091881178185917,
+      "learning_rate": 6.872263282512544e-06,
+      "loss": 0.6238,
+      "step": 26029
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4763420957605955,
+      "learning_rate": 6.8715385340950655e-06,
+      "loss": 0.7861,
+      "step": 26030
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6337226738727846,
+      "learning_rate": 6.87081380389248e-06,
+      "loss": 0.7409,
+      "step": 26031
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6868143119793075,
+      "learning_rate": 6.8700890919090044e-06,
+      "loss": 0.6763,
+      "step": 26032
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3409807213827158,
+      "learning_rate": 6.869364398148865e-06,
+      "loss": 0.5698,
+      "step": 26033
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3696077042993857,
+      "learning_rate": 6.868639722616277e-06,
+      "loss": 0.7361,
+      "step": 26034
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7952087630679308,
+      "learning_rate": 6.8679150653154595e-06,
+      "loss": 0.7676,
+      "step": 26035
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4308387719753248,
+      "learning_rate": 6.867190426250629e-06,
+      "loss": 0.7636,
+      "step": 26036
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.639540969075543,
+      "learning_rate": 6.8664658054260124e-06,
+      "loss": 0.711,
+      "step": 26037
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3509708637677744,
+      "learning_rate": 6.865741202845821e-06,
+      "loss": 0.7401,
+      "step": 26038
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3707708606304143,
+      "learning_rate": 6.865016618514277e-06,
+      "loss": 0.6674,
+      "step": 26039
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4184782708829895,
+      "learning_rate": 6.864292052435597e-06,
+      "loss": 0.7062,
+      "step": 26040
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2724418239812103,
+      "learning_rate": 6.863567504614006e-06,
+      "loss": 0.6732,
+      "step": 26041
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3185519732806033,
+      "learning_rate": 6.862842975053714e-06,
+      "loss": 0.7933,
+      "step": 26042
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3141802430730527,
+      "learning_rate": 6.862118463758944e-06,
+      "loss": 0.7041,
+      "step": 26043
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.350455970839347,
+      "learning_rate": 6.8613939707339185e-06,
+      "loss": 0.6825,
+      "step": 26044
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4240527624797241,
+      "learning_rate": 6.860669495982844e-06,
+      "loss": 0.7138,
+      "step": 26045
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6047705099955014,
+      "learning_rate": 6.859945039509948e-06,
+      "loss": 0.6534,
+      "step": 26046
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1788646601388284,
+      "learning_rate": 6.859220601319445e-06,
+      "loss": 0.8254,
+      "step": 26047
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8618935340476004,
+      "learning_rate": 6.8584961814155595e-06,
+      "loss": 0.7013,
+      "step": 26048
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3996221052216213,
+      "learning_rate": 6.857771779802498e-06,
+      "loss": 0.6284,
+      "step": 26049
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5263119261868836,
+      "learning_rate": 6.857047396484486e-06,
+      "loss": 0.5968,
+      "step": 26050
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5621974249773829,
+      "learning_rate": 6.856323031465741e-06,
+      "loss": 0.4732,
+      "step": 26051
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8170291516505208,
+      "learning_rate": 6.855598684750475e-06,
+      "loss": 0.6833,
+      "step": 26052
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4185712030454467,
+      "learning_rate": 6.854874356342911e-06,
+      "loss": 0.7979,
+      "step": 26053
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.399720296999483,
+      "learning_rate": 6.854150046247263e-06,
+      "loss": 0.6224,
+      "step": 26054
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3447351386698676,
+      "learning_rate": 6.853425754467753e-06,
+      "loss": 0.7764,
+      "step": 26055
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6237130385008538,
+      "learning_rate": 6.852701481008589e-06,
+      "loss": 0.5175,
+      "step": 26056
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3368396318654348,
+      "learning_rate": 6.851977225873996e-06,
+      "loss": 0.6771,
+      "step": 26057
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9630239503910958,
+      "learning_rate": 6.851252989068191e-06,
+      "loss": 0.6936,
+      "step": 26058
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8069359322545642,
+      "learning_rate": 6.8505287705953835e-06,
+      "loss": 0.7582,
+      "step": 26059
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6848534487944726,
+      "learning_rate": 6.849804570459795e-06,
+      "loss": 0.6514,
+      "step": 26060
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5168894253079193,
+      "learning_rate": 6.849080388665643e-06,
+      "loss": 0.7132,
+      "step": 26061
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.495698984015666,
+      "learning_rate": 6.848356225217146e-06,
+      "loss": 0.6859,
+      "step": 26062
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5889695492224051,
+      "learning_rate": 6.8476320801185116e-06,
+      "loss": 0.5337,
+      "step": 26063
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6017850837392922,
+      "learning_rate": 6.846907953373962e-06,
+      "loss": 0.6354,
+      "step": 26064
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6064509362246249,
+      "learning_rate": 6.846183844987715e-06,
+      "loss": 0.6872,
+      "step": 26065
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5812157711796191,
+      "learning_rate": 6.845459754963983e-06,
+      "loss": 0.4919,
+      "step": 26066
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2838083923674795,
+      "learning_rate": 6.844735683306982e-06,
+      "loss": 0.7314,
+      "step": 26067
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5699540218250037,
+      "learning_rate": 6.844011630020928e-06,
+      "loss": 0.6256,
+      "step": 26068
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6088134123825136,
+      "learning_rate": 6.843287595110041e-06,
+      "loss": 0.5052,
+      "step": 26069
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4947366395249502,
+      "learning_rate": 6.842563578578531e-06,
+      "loss": 0.7632,
+      "step": 26070
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3846012949747803,
+      "learning_rate": 6.841839580430616e-06,
+      "loss": 0.7304,
+      "step": 26071
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4510174206166848,
+      "learning_rate": 6.8411156006705116e-06,
+      "loss": 0.6478,
+      "step": 26072
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5459778041784584,
+      "learning_rate": 6.8403916393024285e-06,
+      "loss": 0.4797,
+      "step": 26073
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5824996216193794,
+      "learning_rate": 6.839667696330589e-06,
+      "loss": 0.6782,
+      "step": 26074
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.239678608075505,
+      "learning_rate": 6.838943771759204e-06,
+      "loss": 0.6571,
+      "step": 26075
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6283928292935294,
+      "learning_rate": 6.838219865592491e-06,
+      "loss": 0.5034,
+      "step": 26076
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.564217812473414,
+      "learning_rate": 6.837495977834658e-06,
+      "loss": 0.4899,
+      "step": 26077
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2737945344280914,
+      "learning_rate": 6.836772108489928e-06,
+      "loss": 0.6857,
+      "step": 26078
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1924895448749708,
+      "learning_rate": 6.836048257562513e-06,
+      "loss": 0.7042,
+      "step": 26079
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5122242315187613,
+      "learning_rate": 6.835324425056626e-06,
+      "loss": 0.6925,
+      "step": 26080
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3717140064321514,
+      "learning_rate": 6.834600610976478e-06,
+      "loss": 0.8208,
+      "step": 26081
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4517671847841203,
+      "learning_rate": 6.8338768153262925e-06,
+      "loss": 0.6237,
+      "step": 26082
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8529483136954648,
+      "learning_rate": 6.833153038110279e-06,
+      "loss": 0.725,
+      "step": 26083
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2769589551437273,
+      "learning_rate": 6.8324292793326485e-06,
+      "loss": 0.6186,
+      "step": 26084
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5880415673311836,
+      "learning_rate": 6.831705538997617e-06,
+      "loss": 0.5584,
+      "step": 26085
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3344453626485526,
+      "learning_rate": 6.830981817109401e-06,
+      "loss": 0.9127,
+      "step": 26086
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4968894512070254,
+      "learning_rate": 6.830258113672213e-06,
+      "loss": 0.7721,
+      "step": 26087
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3061373445950928,
+      "learning_rate": 6.829534428690263e-06,
+      "loss": 0.6842,
+      "step": 26088
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3320777892059192,
+      "learning_rate": 6.8288107621677655e-06,
+      "loss": 0.7856,
+      "step": 26089
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4812762304943932,
+      "learning_rate": 6.8280871141089415e-06,
+      "loss": 0.7323,
+      "step": 26090
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.330409920375332,
+      "learning_rate": 6.8273634845179955e-06,
+      "loss": 0.6607,
+      "step": 26091
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.536910597782775,
+      "learning_rate": 6.8266398733991415e-06,
+      "loss": 0.6549,
+      "step": 26092
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3178845864698316,
+      "learning_rate": 6.825916280756601e-06,
+      "loss": 0.6278,
+      "step": 26093
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.201752967300295,
+      "learning_rate": 6.8251927065945755e-06,
+      "loss": 0.6472,
+      "step": 26094
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6199386972691425,
+      "learning_rate": 6.824469150917286e-06,
+      "loss": 0.4897,
+      "step": 26095
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4697243855526,
+      "learning_rate": 6.8237456137289385e-06,
+      "loss": 0.7017,
+      "step": 26096
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5835569736137179,
+      "learning_rate": 6.823022095033756e-06,
+      "loss": 0.5385,
+      "step": 26097
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2378672878972599,
+      "learning_rate": 6.822298594835939e-06,
+      "loss": 0.7215,
+      "step": 26098
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3687277657253496,
+      "learning_rate": 6.821575113139707e-06,
+      "loss": 0.7051,
+      "step": 26099
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4507319630959044,
+      "learning_rate": 6.820851649949274e-06,
+      "loss": 0.7351,
+      "step": 26100
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4263174528854166,
+      "learning_rate": 6.820128205268845e-06,
+      "loss": 0.7918,
+      "step": 26101
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4811634228008044,
+      "learning_rate": 6.8194047791026375e-06,
+      "loss": 0.7637,
+      "step": 26102
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9146915089300454,
+      "learning_rate": 6.8186813714548625e-06,
+      "loss": 0.7493,
+      "step": 26103
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3762228134048413,
+      "learning_rate": 6.817957982329735e-06,
+      "loss": 0.8108,
+      "step": 26104
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3490382076878629,
+      "learning_rate": 6.817234611731458e-06,
+      "loss": 0.7032,
+      "step": 26105
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7699952536266284,
+      "learning_rate": 6.816511259664251e-06,
+      "loss": 0.6685,
+      "step": 26106
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2661145435208165,
+      "learning_rate": 6.8157879261323265e-06,
+      "loss": 0.6305,
+      "step": 26107
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5167717087157018,
+      "learning_rate": 6.81506461113989e-06,
+      "loss": 0.8209,
+      "step": 26108
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5562118354451133,
+      "learning_rate": 6.814341314691153e-06,
+      "loss": 0.8145,
+      "step": 26109
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.769499023823783,
+      "learning_rate": 6.813618036790333e-06,
+      "loss": 0.7895,
+      "step": 26110
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4036773311986779,
+      "learning_rate": 6.812894777441638e-06,
+      "loss": 0.5783,
+      "step": 26111
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6028990074251235,
+      "learning_rate": 6.812171536649278e-06,
+      "loss": 0.6805,
+      "step": 26112
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.607564808640395,
+      "learning_rate": 6.811448314417461e-06,
+      "loss": 0.747,
+      "step": 26113
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.458474165511265,
+      "learning_rate": 6.8107251107504065e-06,
+      "loss": 0.7018,
+      "step": 26114
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5212693010948606,
+      "learning_rate": 6.810001925652319e-06,
+      "loss": 0.663,
+      "step": 26115
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6730415496675592,
+      "learning_rate": 6.809278759127409e-06,
+      "loss": 0.6354,
+      "step": 26116
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4573659714519913,
+      "learning_rate": 6.808555611179886e-06,
+      "loss": 0.7764,
+      "step": 26117
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.313363730570249,
+      "learning_rate": 6.807832481813967e-06,
+      "loss": 0.7814,
+      "step": 26118
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5250580063989403,
+      "learning_rate": 6.807109371033856e-06,
+      "loss": 0.5354,
+      "step": 26119
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3368347596838512,
+      "learning_rate": 6.806386278843764e-06,
+      "loss": 0.6228,
+      "step": 26120
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.848382727044131,
+      "learning_rate": 6.8056632052479055e-06,
+      "loss": 0.7745,
+      "step": 26121
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6846119882086203,
+      "learning_rate": 6.804940150250483e-06,
+      "loss": 0.6944,
+      "step": 26122
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5021764312693056,
+      "learning_rate": 6.804217113855712e-06,
+      "loss": 0.701,
+      "step": 26123
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4781808363088347,
+      "learning_rate": 6.803494096067802e-06,
+      "loss": 0.7487,
+      "step": 26124
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4114877299848594,
+      "learning_rate": 6.802771096890962e-06,
+      "loss": 0.6955,
+      "step": 26125
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4003604463007444,
+      "learning_rate": 6.8020481163293975e-06,
+      "loss": 0.7774,
+      "step": 26126
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.438907508855453,
+      "learning_rate": 6.801325154387323e-06,
+      "loss": 0.7693,
+      "step": 26127
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.400976712930673,
+      "learning_rate": 6.800602211068947e-06,
+      "loss": 0.7561,
+      "step": 26128
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.419041666884859,
+      "learning_rate": 6.7998792863784775e-06,
+      "loss": 0.7249,
+      "step": 26129
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3703824344524649,
+      "learning_rate": 6.799156380320122e-06,
+      "loss": 0.7834,
+      "step": 26130
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.613470502467163,
+      "learning_rate": 6.798433492898093e-06,
+      "loss": 0.5575,
+      "step": 26131
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.390817859785538,
+      "learning_rate": 6.797710624116601e-06,
+      "loss": 0.7393,
+      "step": 26132
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.657023401313589,
+      "learning_rate": 6.796987773979846e-06,
+      "loss": 0.8462,
+      "step": 26133
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6057415198779497,
+      "learning_rate": 6.796264942492044e-06,
+      "loss": 0.5414,
+      "step": 26134
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5873942572937045,
+      "learning_rate": 6.7955421296574045e-06,
+      "loss": 0.5319,
+      "step": 26135
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.826925884477658,
+      "learning_rate": 6.7948193354801306e-06,
+      "loss": 0.8571,
+      "step": 26136
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2497694456611956,
+      "learning_rate": 6.794096559964434e-06,
+      "loss": 0.6493,
+      "step": 26137
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2921531266865403,
+      "learning_rate": 6.7933738031145205e-06,
+      "loss": 0.7817,
+      "step": 26138
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.216235065519277,
+      "learning_rate": 6.792651064934602e-06,
+      "loss": 0.6349,
+      "step": 26139
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2731050005551596,
+      "learning_rate": 6.791928345428884e-06,
+      "loss": 0.666,
+      "step": 26140
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3194269006921258,
+      "learning_rate": 6.791205644601573e-06,
+      "loss": 0.6319,
+      "step": 26141
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4079274725236086,
+      "learning_rate": 6.790482962456879e-06,
+      "loss": 0.667,
+      "step": 26142
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5472496907154185,
+      "learning_rate": 6.789760298999013e-06,
+      "loss": 0.6002,
+      "step": 26143
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8433958516990245,
+      "learning_rate": 6.789037654232177e-06,
+      "loss": 0.6818,
+      "step": 26144
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.317635905652594,
+      "learning_rate": 6.788315028160577e-06,
+      "loss": 0.6627,
+      "step": 26145
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.641535817045446,
+      "learning_rate": 6.787592420788431e-06,
+      "loss": 0.5164,
+      "step": 26146
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3925362841783697,
+      "learning_rate": 6.786869832119933e-06,
+      "loss": 0.6329,
+      "step": 26147
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4809299273619294,
+      "learning_rate": 6.786147262159297e-06,
+      "loss": 0.7694,
+      "step": 26148
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5727990250365333,
+      "learning_rate": 6.7854247109107285e-06,
+      "loss": 0.471,
+      "step": 26149
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5622313172515117,
+      "learning_rate": 6.784702178378437e-06,
+      "loss": 0.6542,
+      "step": 26150
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4712026862027927,
+      "learning_rate": 6.783979664566627e-06,
+      "loss": 0.6922,
+      "step": 26151
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8512468773664468,
+      "learning_rate": 6.783257169479504e-06,
+      "loss": 0.6552,
+      "step": 26152
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.393606379181144,
+      "learning_rate": 6.782534693121281e-06,
+      "loss": 0.7171,
+      "step": 26153
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0377055608533174,
+      "learning_rate": 6.781812235496153e-06,
+      "loss": 0.7828,
+      "step": 26154
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2415963420044451,
+      "learning_rate": 6.7810897966083365e-06,
+      "loss": 0.8164,
+      "step": 26155
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.244218316433358,
+      "learning_rate": 6.780367376462033e-06,
+      "loss": 0.7198,
+      "step": 26156
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3756769460683567,
+      "learning_rate": 6.779644975061454e-06,
+      "loss": 0.7634,
+      "step": 26157
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7846703006925495,
+      "learning_rate": 6.778922592410796e-06,
+      "loss": 0.688,
+      "step": 26158
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4329715806343932,
+      "learning_rate": 6.778200228514274e-06,
+      "loss": 0.6497,
+      "step": 26159
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.305077905125565,
+      "learning_rate": 6.777477883376089e-06,
+      "loss": 0.5973,
+      "step": 26160
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4107394105455016,
+      "learning_rate": 6.776755557000449e-06,
+      "loss": 0.7129,
+      "step": 26161
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5083953262462424,
+      "learning_rate": 6.776033249391555e-06,
+      "loss": 0.6679,
+      "step": 26162
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3679424563407392,
+      "learning_rate": 6.775310960553618e-06,
+      "loss": 0.7125,
+      "step": 26163
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3423797649144353,
+      "learning_rate": 6.774588690490844e-06,
+      "loss": 0.6345,
+      "step": 26164
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3030021111446335,
+      "learning_rate": 6.7738664392074345e-06,
+      "loss": 0.6442,
+      "step": 26165
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3722720535261546,
+      "learning_rate": 6.773144206707594e-06,
+      "loss": 0.7397,
+      "step": 26166
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.592383822710416,
+      "learning_rate": 6.772421992995533e-06,
+      "loss": 0.462,
+      "step": 26167
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3212542549444288,
+      "learning_rate": 6.7716997980754505e-06,
+      "loss": 0.7879,
+      "step": 26168
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2551439768543446,
+      "learning_rate": 6.770977621951554e-06,
+      "loss": 0.6992,
+      "step": 26169
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6113245559782903,
+      "learning_rate": 6.770255464628047e-06,
+      "loss": 0.7207,
+      "step": 26170
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5980935299792063,
+      "learning_rate": 6.769533326109139e-06,
+      "loss": 0.5775,
+      "step": 26171
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.603302007163142,
+      "learning_rate": 6.768811206399028e-06,
+      "loss": 0.5065,
+      "step": 26172
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4491693253693296,
+      "learning_rate": 6.7680891055019215e-06,
+      "loss": 0.679,
+      "step": 26173
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.332974989608343,
+      "learning_rate": 6.767367023422025e-06,
+      "loss": 0.685,
+      "step": 26174
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4621830551843904,
+      "learning_rate": 6.766644960163538e-06,
+      "loss": 0.6469,
+      "step": 26175
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3250032473429958,
+      "learning_rate": 6.765922915730669e-06,
+      "loss": 0.6793,
+      "step": 26176
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6610261996101505,
+      "learning_rate": 6.765200890127619e-06,
+      "loss": 0.6663,
+      "step": 26177
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4652830223732929,
+      "learning_rate": 6.7644788833586e-06,
+      "loss": 0.6784,
+      "step": 26178
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5165315542405828,
+      "learning_rate": 6.763756895427802e-06,
+      "loss": 0.6645,
+      "step": 26179
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.436238477741415,
+      "learning_rate": 6.763034926339438e-06,
+      "loss": 0.7735,
+      "step": 26180
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6429387854199557,
+      "learning_rate": 6.762312976097714e-06,
+      "loss": 0.7478,
+      "step": 26181
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3042477187568675,
+      "learning_rate": 6.7615910447068235e-06,
+      "loss": 0.7169,
+      "step": 26182
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6034708948926504,
+      "learning_rate": 6.7608691321709755e-06,
+      "loss": 0.6898,
+      "step": 26183
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7102531106256564,
+      "learning_rate": 6.760147238494374e-06,
+      "loss": 0.6954,
+      "step": 26184
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2116316513776402,
+      "learning_rate": 6.759425363681223e-06,
+      "loss": 0.6751,
+      "step": 26185
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5599537602741813,
+      "learning_rate": 6.758703507735721e-06,
+      "loss": 0.7779,
+      "step": 26186
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3570841411762329,
+      "learning_rate": 6.7579816706620724e-06,
+      "loss": 0.5887,
+      "step": 26187
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6258952649122806,
+      "learning_rate": 6.757259852464484e-06,
+      "loss": 0.5226,
+      "step": 26188
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4669670932208472,
+      "learning_rate": 6.7565380531471545e-06,
+      "loss": 0.6899,
+      "step": 26189
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4325504085177279,
+      "learning_rate": 6.755816272714284e-06,
+      "loss": 0.7211,
+      "step": 26190
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3410361471414651,
+      "learning_rate": 6.755094511170082e-06,
+      "loss": 0.7707,
+      "step": 26191
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0221676939825572,
+      "learning_rate": 6.754372768518746e-06,
+      "loss": 0.7918,
+      "step": 26192
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.727591240751848,
+      "learning_rate": 6.7536510447644795e-06,
+      "loss": 0.6706,
+      "step": 26193
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6520706385947626,
+      "learning_rate": 6.752929339911483e-06,
+      "loss": 0.7756,
+      "step": 26194
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4109542881836439,
+      "learning_rate": 6.752207653963965e-06,
+      "loss": 0.5769,
+      "step": 26195
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.520926862613687,
+      "learning_rate": 6.7514859869261164e-06,
+      "loss": 0.7527,
+      "step": 26196
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5645712392741664,
+      "learning_rate": 6.750764338802147e-06,
+      "loss": 0.6441,
+      "step": 26197
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.351777427832894,
+      "learning_rate": 6.750042709596254e-06,
+      "loss": 0.7184,
+      "step": 26198
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4390644819129375,
+      "learning_rate": 6.749321099312644e-06,
+      "loss": 0.6955,
+      "step": 26199
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4206707836248451,
+      "learning_rate": 6.748599507955516e-06,
+      "loss": 0.697,
+      "step": 26200
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5531983026992502,
+      "learning_rate": 6.74787793552907e-06,
+      "loss": 0.7379,
+      "step": 26201
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4534068649193417,
+      "learning_rate": 6.747156382037511e-06,
+      "loss": 0.7645,
+      "step": 26202
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3798874989336016,
+      "learning_rate": 6.7464348474850324e-06,
+      "loss": 0.6454,
+      "step": 26203
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5868784019424875,
+      "learning_rate": 6.745713331875843e-06,
+      "loss": 0.7625,
+      "step": 26204
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7155683397075563,
+      "learning_rate": 6.74499183521414e-06,
+      "loss": 0.7414,
+      "step": 26205
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5756668915792826,
+      "learning_rate": 6.744270357504128e-06,
+      "loss": 0.7628,
+      "step": 26206
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5563537762242514,
+      "learning_rate": 6.743548898749999e-06,
+      "loss": 0.6783,
+      "step": 26207
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3667267098648312,
+      "learning_rate": 6.7428274589559626e-06,
+      "loss": 0.7737,
+      "step": 26208
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4147143267146627,
+      "learning_rate": 6.742106038126218e-06,
+      "loss": 0.6496,
+      "step": 26209
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3652127085929946,
+      "learning_rate": 6.741384636264961e-06,
+      "loss": 0.7733,
+      "step": 26210
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2422537693894962,
+      "learning_rate": 6.740663253376393e-06,
+      "loss": 0.7165,
+      "step": 26211
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4608952177217536,
+      "learning_rate": 6.7399418894647164e-06,
+      "loss": 0.6266,
+      "step": 26212
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.474457944960141,
+      "learning_rate": 6.739220544534133e-06,
+      "loss": 0.7414,
+      "step": 26213
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5897349079183372,
+      "learning_rate": 6.738499218588839e-06,
+      "loss": 0.7513,
+      "step": 26214
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.284378951388041,
+      "learning_rate": 6.737777911633031e-06,
+      "loss": 0.7979,
+      "step": 26215
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5228969186746593,
+      "learning_rate": 6.737056623670918e-06,
+      "loss": 0.6046,
+      "step": 26216
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.545025900573533,
+      "learning_rate": 6.736335354706691e-06,
+      "loss": 0.6554,
+      "step": 26217
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.444069620572425,
+      "learning_rate": 6.7356141047445545e-06,
+      "loss": 0.7995,
+      "step": 26218
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3414064284615712,
+      "learning_rate": 6.734892873788703e-06,
+      "loss": 0.5978,
+      "step": 26219
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3637092427002109,
+      "learning_rate": 6.7341716618433435e-06,
+      "loss": 0.6575,
+      "step": 26220
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4469867375774743,
+      "learning_rate": 6.733450468912669e-06,
+      "loss": 0.754,
+      "step": 26221
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5452640879416294,
+      "learning_rate": 6.732729295000879e-06,
+      "loss": 0.4958,
+      "step": 26222
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3845376342386648,
+      "learning_rate": 6.732008140112176e-06,
+      "loss": 0.6181,
+      "step": 26223
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.502277071034311,
+      "learning_rate": 6.731287004250752e-06,
+      "loss": 0.6922,
+      "step": 26224
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5333691645174667,
+      "learning_rate": 6.730565887420813e-06,
+      "loss": 0.7242,
+      "step": 26225
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4690954148461346,
+      "learning_rate": 6.729844789626552e-06,
+      "loss": 0.6431,
+      "step": 26226
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4662007167676399,
+      "learning_rate": 6.729123710872175e-06,
+      "loss": 0.7691,
+      "step": 26227
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.703317362717501,
+      "learning_rate": 6.728402651161869e-06,
+      "loss": 0.8641,
+      "step": 26228
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2540052504280146,
+      "learning_rate": 6.727681610499842e-06,
+      "loss": 0.6894,
+      "step": 26229
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8432528010841043,
+      "learning_rate": 6.72696058889029e-06,
+      "loss": 0.6622,
+      "step": 26230
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3260170081526101,
+      "learning_rate": 6.726239586337408e-06,
+      "loss": 0.6861,
+      "step": 26231
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5306988796009182,
+      "learning_rate": 6.725518602845394e-06,
+      "loss": 0.6739,
+      "step": 26232
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6351375791987064,
+      "learning_rate": 6.7247976384184475e-06,
+      "loss": 0.7115,
+      "step": 26233
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4033362858731915,
+      "learning_rate": 6.72407669306077e-06,
+      "loss": 0.7869,
+      "step": 26234
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6323206288454848,
+      "learning_rate": 6.723355766776553e-06,
+      "loss": 0.7332,
+      "step": 26235
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5774739470862262,
+      "learning_rate": 6.722634859569992e-06,
+      "loss": 0.6869,
+      "step": 26236
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5001607924899203,
+      "learning_rate": 6.721913971445295e-06,
+      "loss": 0.6211,
+      "step": 26237
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.424024772438631,
+      "learning_rate": 6.721193102406649e-06,
+      "loss": 0.7403,
+      "step": 26238
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5579798118991184,
+      "learning_rate": 6.720472252458254e-06,
+      "loss": 0.7852,
+      "step": 26239
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2698673598627157,
+      "learning_rate": 6.719751421604309e-06,
+      "loss": 0.6398,
+      "step": 26240
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.371437740676449,
+      "learning_rate": 6.719030609849012e-06,
+      "loss": 0.6454,
+      "step": 26241
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.779158879792803,
+      "learning_rate": 6.718309817196556e-06,
+      "loss": 0.717,
+      "step": 26242
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3521007426988036,
+      "learning_rate": 6.717589043651136e-06,
+      "loss": 0.7169,
+      "step": 26243
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6047057586518775,
+      "learning_rate": 6.716868289216959e-06,
+      "loss": 0.491,
+      "step": 26244
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4446374697604845,
+      "learning_rate": 6.716147553898206e-06,
+      "loss": 0.8054,
+      "step": 26245
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2631697849003485,
+      "learning_rate": 6.715426837699086e-06,
+      "loss": 0.8134,
+      "step": 26246
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6046249796940838,
+      "learning_rate": 6.714706140623789e-06,
+      "loss": 0.5256,
+      "step": 26247
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3088764864038047,
+      "learning_rate": 6.7139854626765155e-06,
+      "loss": 0.8164,
+      "step": 26248
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.212199256368408,
+      "learning_rate": 6.713264803861459e-06,
+      "loss": 0.766,
+      "step": 26249
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3962278652047657,
+      "learning_rate": 6.712544164182813e-06,
+      "loss": 0.6178,
+      "step": 26250
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4374984080246542,
+      "learning_rate": 6.711823543644779e-06,
+      "loss": 0.6803,
+      "step": 26251
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4973425558535864,
+      "learning_rate": 6.711102942251544e-06,
+      "loss": 0.6149,
+      "step": 26252
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6105571219120152,
+      "learning_rate": 6.710382360007313e-06,
+      "loss": 0.5725,
+      "step": 26253
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.631140260702901,
+      "learning_rate": 6.7096617969162756e-06,
+      "loss": 0.7138,
+      "step": 26254
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3062161461898105,
+      "learning_rate": 6.708941252982632e-06,
+      "loss": 0.7075,
+      "step": 26255
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6257279770487902,
+      "learning_rate": 6.70822072821057e-06,
+      "loss": 0.7174,
+      "step": 26256
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3885935800694578,
+      "learning_rate": 6.707500222604291e-06,
+      "loss": 0.723,
+      "step": 26257
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3050848312157994,
+      "learning_rate": 6.70677973616799e-06,
+      "loss": 0.7121,
+      "step": 26258
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.711398312607392,
+      "learning_rate": 6.706059268905858e-06,
+      "loss": 0.6263,
+      "step": 26259
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2948598586804054,
+      "learning_rate": 6.705338820822089e-06,
+      "loss": 0.6247,
+      "step": 26260
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5688752856999055,
+      "learning_rate": 6.704618391920884e-06,
+      "loss": 0.7028,
+      "step": 26261
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4717540230563544,
+      "learning_rate": 6.7038979822064355e-06,
+      "loss": 0.7226,
+      "step": 26262
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5553619560710702,
+      "learning_rate": 6.703177591682935e-06,
+      "loss": 0.4797,
+      "step": 26263
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5661307022921991,
+      "learning_rate": 6.702457220354575e-06,
+      "loss": 0.5195,
+      "step": 26264
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.235289811104162,
+      "learning_rate": 6.701736868225556e-06,
+      "loss": 0.7085,
+      "step": 26265
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.413359747747552,
+      "learning_rate": 6.701016535300071e-06,
+      "loss": 0.728,
+      "step": 26266
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.012915406920626,
+      "learning_rate": 6.7002962215823095e-06,
+      "loss": 0.7373,
+      "step": 26267
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5987209878450303,
+      "learning_rate": 6.699575927076467e-06,
+      "loss": 0.5426,
+      "step": 26268
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.247778958592835,
+      "learning_rate": 6.698855651786742e-06,
+      "loss": 0.7563,
+      "step": 26269
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5959343541692104,
+      "learning_rate": 6.698135395717323e-06,
+      "loss": 0.4901,
+      "step": 26270
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3127554582312282,
+      "learning_rate": 6.697415158872405e-06,
+      "loss": 0.7013,
+      "step": 26271
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5181017856422276,
+      "learning_rate": 6.696694941256179e-06,
+      "loss": 0.7057,
+      "step": 26272
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4991174766716844,
+      "learning_rate": 6.695974742872844e-06,
+      "loss": 0.5993,
+      "step": 26273
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.469152493637012,
+      "learning_rate": 6.69525456372659e-06,
+      "loss": 0.811,
+      "step": 26274
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3995070203846733,
+      "learning_rate": 6.694534403821606e-06,
+      "loss": 0.6971,
+      "step": 26275
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5778875811654418,
+      "learning_rate": 6.6938142631620975e-06,
+      "loss": 0.7255,
+      "step": 26276
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2414983779395166,
+      "learning_rate": 6.693094141752241e-06,
+      "loss": 0.7306,
+      "step": 26277
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3858140444338265,
+      "learning_rate": 6.692374039596241e-06,
+      "loss": 0.6579,
+      "step": 26278
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4959904257293577,
+      "learning_rate": 6.691653956698283e-06,
+      "loss": 0.795,
+      "step": 26279
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5262015731017469,
+      "learning_rate": 6.69093389306257e-06,
+      "loss": 0.7834,
+      "step": 26280
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.570594441689681,
+      "learning_rate": 6.690213848693282e-06,
+      "loss": 0.7568,
+      "step": 26281
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5608675814653135,
+      "learning_rate": 6.689493823594618e-06,
+      "loss": 0.7681,
+      "step": 26282
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6088279839573807,
+      "learning_rate": 6.688773817770772e-06,
+      "loss": 0.6686,
+      "step": 26283
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4051083303282845,
+      "learning_rate": 6.68805383122593e-06,
+      "loss": 0.6587,
+      "step": 26284
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3315094076006728,
+      "learning_rate": 6.687333863964285e-06,
+      "loss": 0.7192,
+      "step": 26285
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2732974034376374,
+      "learning_rate": 6.686613915990034e-06,
+      "loss": 0.7574,
+      "step": 26286
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5554726800254923,
+      "learning_rate": 6.685893987307368e-06,
+      "loss": 0.7124,
+      "step": 26287
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3259247001757397,
+      "learning_rate": 6.685174077920471e-06,
+      "loss": 0.8117,
+      "step": 26288
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5371249441258266,
+      "learning_rate": 6.684454187833543e-06,
+      "loss": 0.8369,
+      "step": 26289
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.294163333287513,
+      "learning_rate": 6.683734317050774e-06,
+      "loss": 0.636,
+      "step": 26290
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5330628240530912,
+      "learning_rate": 6.683014465576353e-06,
+      "loss": 0.6166,
+      "step": 26291
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3243134933969032,
+      "learning_rate": 6.6822946334144675e-06,
+      "loss": 0.7804,
+      "step": 26292
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6061952973776,
+      "learning_rate": 6.681574820569317e-06,
+      "loss": 0.7537,
+      "step": 26293
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4241931630635156,
+      "learning_rate": 6.68085502704509e-06,
+      "loss": 0.6442,
+      "step": 26294
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.444808902531184,
+      "learning_rate": 6.680135252845974e-06,
+      "loss": 0.7599,
+      "step": 26295
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.241250892814119,
+      "learning_rate": 6.6794154979761595e-06,
+      "loss": 0.6494,
+      "step": 26296
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3893084093515393,
+      "learning_rate": 6.678695762439843e-06,
+      "loss": 0.6977,
+      "step": 26297
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5818410383500476,
+      "learning_rate": 6.677976046241209e-06,
+      "loss": 0.667,
+      "step": 26298
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1431601677735799,
+      "learning_rate": 6.677256349384451e-06,
+      "loss": 0.7268,
+      "step": 26299
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.290885292726279,
+      "learning_rate": 6.676536671873755e-06,
+      "loss": 0.6589,
+      "step": 26300
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8404605293174334,
+      "learning_rate": 6.675817013713321e-06,
+      "loss": 0.6155,
+      "step": 26301
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.2302085224508152,
+      "learning_rate": 6.675097374907329e-06,
+      "loss": 0.6519,
+      "step": 26302
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5541295987970651,
+      "learning_rate": 6.6743777554599735e-06,
+      "loss": 0.6956,
+      "step": 26303
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5962162294770332,
+      "learning_rate": 6.673658155375446e-06,
+      "loss": 0.7945,
+      "step": 26304
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6272743254044928,
+      "learning_rate": 6.672938574657928e-06,
+      "loss": 0.5242,
+      "step": 26305
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6772760573619596,
+      "learning_rate": 6.672219013311618e-06,
+      "loss": 0.5675,
+      "step": 26306
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.963647188134349,
+      "learning_rate": 6.671499471340704e-06,
+      "loss": 0.7748,
+      "step": 26307
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.583220432340148,
+      "learning_rate": 6.670779948749373e-06,
+      "loss": 0.7116,
+      "step": 26308
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.276457838476359,
+      "learning_rate": 6.6700604455418125e-06,
+      "loss": 0.7021,
+      "step": 26309
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5835502708718531,
+      "learning_rate": 6.669340961722217e-06,
+      "loss": 0.5743,
+      "step": 26310
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4540181502323417,
+      "learning_rate": 6.668621497294774e-06,
+      "loss": 0.7384,
+      "step": 26311
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.84387836609908,
+      "learning_rate": 6.667902052263671e-06,
+      "loss": 0.5948,
+      "step": 26312
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.471746254613422,
+      "learning_rate": 6.667182626633093e-06,
+      "loss": 0.563,
+      "step": 26313
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.461690642650203,
+      "learning_rate": 6.666463220407235e-06,
+      "loss": 0.7293,
+      "step": 26314
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5845522100526047,
+      "learning_rate": 6.665743833590287e-06,
+      "loss": 0.5218,
+      "step": 26315
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4526173237699926,
+      "learning_rate": 6.665024466186431e-06,
+      "loss": 0.8022,
+      "step": 26316
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.28999026653475,
+      "learning_rate": 6.664305118199857e-06,
+      "loss": 0.7003,
+      "step": 26317
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.398931038921106,
+      "learning_rate": 6.663585789634758e-06,
+      "loss": 0.6795,
+      "step": 26318
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5139631563785907,
+      "learning_rate": 6.662866480495317e-06,
+      "loss": 0.6949,
+      "step": 26319
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.423813439053164,
+      "learning_rate": 6.6621471907857234e-06,
+      "loss": 0.6539,
+      "step": 26320
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4306921752122141,
+      "learning_rate": 6.661427920510164e-06,
+      "loss": 0.6439,
+      "step": 26321
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.468298722183193,
+      "learning_rate": 6.660708669672833e-06,
+      "loss": 0.6413,
+      "step": 26322
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4089933819617768,
+      "learning_rate": 6.6599894382779104e-06,
+      "loss": 0.6791,
+      "step": 26323
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8048111028909868,
+      "learning_rate": 6.659270226329586e-06,
+      "loss": 0.8035,
+      "step": 26324
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4655846198037232,
+      "learning_rate": 6.658551033832052e-06,
+      "loss": 0.6503,
+      "step": 26325
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4657237271686323,
+      "learning_rate": 6.657831860789488e-06,
+      "loss": 0.628,
+      "step": 26326
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4485684719568108,
+      "learning_rate": 6.657112707206086e-06,
+      "loss": 0.7475,
+      "step": 26327
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.623846041754119,
+      "learning_rate": 6.656393573086029e-06,
+      "loss": 0.7392,
+      "step": 26328
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.421544338414859,
+      "learning_rate": 6.655674458433514e-06,
+      "loss": 0.6934,
+      "step": 26329
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.670104538187981,
+      "learning_rate": 6.6549553632527154e-06,
+      "loss": 0.7426,
+      "step": 26330
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.502777366104801,
+      "learning_rate": 6.6542362875478285e-06,
+      "loss": 0.6924,
+      "step": 26331
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4230416863644912,
+      "learning_rate": 6.6535172313230375e-06,
+      "loss": 0.6484,
+      "step": 26332
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0896012102246404,
+      "learning_rate": 6.652798194582528e-06,
+      "loss": 0.6699,
+      "step": 26333
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.256370661219963,
+      "learning_rate": 6.652079177330484e-06,
+      "loss": 0.6545,
+      "step": 26334
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.391270023377734,
+      "learning_rate": 6.6513601795710984e-06,
+      "loss": 0.8197,
+      "step": 26335
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.348124681407903,
+      "learning_rate": 6.650641201308554e-06,
+      "loss": 0.5609,
+      "step": 26336
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1091982277821653,
+      "learning_rate": 6.649922242547033e-06,
+      "loss": 0.6558,
+      "step": 26337
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4977781883593586,
+      "learning_rate": 6.64920330329073e-06,
+      "loss": 0.6259,
+      "step": 26338
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8147412646982946,
+      "learning_rate": 6.648484383543826e-06,
+      "loss": 0.617,
+      "step": 26339
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6362247844207127,
+      "learning_rate": 6.647765483310506e-06,
+      "loss": 0.5092,
+      "step": 26340
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4298929812442054,
+      "learning_rate": 6.647046602594953e-06,
+      "loss": 0.6995,
+      "step": 26341
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6953491071140847,
+      "learning_rate": 6.64632774140136e-06,
+      "loss": 0.6304,
+      "step": 26342
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3566183197399766,
+      "learning_rate": 6.645608899733911e-06,
+      "loss": 0.7751,
+      "step": 26343
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.383105039864992,
+      "learning_rate": 6.644890077596785e-06,
+      "loss": 0.7551,
+      "step": 26344
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3788518649657657,
+      "learning_rate": 6.64417127499417e-06,
+      "loss": 0.6278,
+      "step": 26345
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.124756970405836,
+      "learning_rate": 6.643452491930257e-06,
+      "loss": 0.6108,
+      "step": 26346
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.657217829641375,
+      "learning_rate": 6.642733728409223e-06,
+      "loss": 0.706,
+      "step": 26347
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6865379004429164,
+      "learning_rate": 6.642014984435257e-06,
+      "loss": 0.7214,
+      "step": 26348
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3004536019517432,
+      "learning_rate": 6.641296260012541e-06,
+      "loss": 0.737,
+      "step": 26349
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5753422457112314,
+      "learning_rate": 6.640577555145266e-06,
+      "loss": 0.649,
+      "step": 26350
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.711288639463189,
+      "learning_rate": 6.639858869837608e-06,
+      "loss": 0.6114,
+      "step": 26351
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.467622130488352,
+      "learning_rate": 6.639140204093757e-06,
+      "loss": 0.6825,
+      "step": 26352
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.585430060587158,
+      "learning_rate": 6.638421557917898e-06,
+      "loss": 0.6654,
+      "step": 26353
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.19518305878504,
+      "learning_rate": 6.637702931314208e-06,
+      "loss": 0.7445,
+      "step": 26354
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3353195798521489,
+      "learning_rate": 6.6369843242868795e-06,
+      "loss": 0.6189,
+      "step": 26355
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5055628592193024,
+      "learning_rate": 6.63626573684009e-06,
+      "loss": 0.7324,
+      "step": 26356
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6284345547436936,
+      "learning_rate": 6.635547168978032e-06,
+      "loss": 0.7733,
+      "step": 26357
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4807932121374991,
+      "learning_rate": 6.6348286207048776e-06,
+      "loss": 0.6428,
+      "step": 26358
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3271295597655364,
+      "learning_rate": 6.634110092024819e-06,
+      "loss": 0.7657,
+      "step": 26359
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6493209535942537,
+      "learning_rate": 6.633391582942038e-06,
+      "loss": 0.5496,
+      "step": 26360
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3346563782138616,
+      "learning_rate": 6.632673093460715e-06,
+      "loss": 0.7029,
+      "step": 26361
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3323892935962565,
+      "learning_rate": 6.631954623585035e-06,
+      "loss": 0.7936,
+      "step": 26362
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4558881985309393,
+      "learning_rate": 6.631236173319182e-06,
+      "loss": 0.6225,
+      "step": 26363
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8881056542978547,
+      "learning_rate": 6.630517742667341e-06,
+      "loss": 0.695,
+      "step": 26364
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5319929529204255,
+      "learning_rate": 6.629799331633691e-06,
+      "loss": 0.7614,
+      "step": 26365
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8695228683000316,
+      "learning_rate": 6.629080940222414e-06,
+      "loss": 0.7675,
+      "step": 26366
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3268227600960087,
+      "learning_rate": 6.628362568437698e-06,
+      "loss": 0.6032,
+      "step": 26367
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5652386106770433,
+      "learning_rate": 6.6276442162837215e-06,
+      "loss": 0.7108,
+      "step": 26368
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5560332517216833,
+      "learning_rate": 6.6269258837646675e-06,
+      "loss": 0.7243,
+      "step": 26369
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8106289889604987,
+      "learning_rate": 6.626207570884717e-06,
+      "loss": 0.6284,
+      "step": 26370
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2582009437188533,
+      "learning_rate": 6.625489277648059e-06,
+      "loss": 0.6598,
+      "step": 26371
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4295387637631263,
+      "learning_rate": 6.624771004058869e-06,
+      "loss": 0.7368,
+      "step": 26372
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5582104799376795,
+      "learning_rate": 6.624052750121328e-06,
+      "loss": 0.7088,
+      "step": 26373
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4269941633732228,
+      "learning_rate": 6.623334515839627e-06,
+      "loss": 0.6144,
+      "step": 26374
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6330795666435125,
+      "learning_rate": 6.6226163012179345e-06,
+      "loss": 0.7485,
+      "step": 26375
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4555526653250328,
+      "learning_rate": 6.621898106260443e-06,
+      "loss": 0.7206,
+      "step": 26376
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5093518713282834,
+      "learning_rate": 6.621179930971327e-06,
+      "loss": 0.8309,
+      "step": 26377
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.544374004105235,
+      "learning_rate": 6.620461775354776e-06,
+      "loss": 0.7158,
+      "step": 26378
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8412492823165132,
+      "learning_rate": 6.619743639414961e-06,
+      "loss": 0.7745,
+      "step": 26379
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5509235973470066,
+      "learning_rate": 6.619025523156073e-06,
+      "loss": 0.7257,
+      "step": 26380
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2534350698084744,
+      "learning_rate": 6.618307426582289e-06,
+      "loss": 0.7693,
+      "step": 26381
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5015434239340404,
+      "learning_rate": 6.617589349697786e-06,
+      "loss": 0.5946,
+      "step": 26382
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4421451600383939,
+      "learning_rate": 6.61687129250675e-06,
+      "loss": 0.813,
+      "step": 26383
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9518007219277347,
+      "learning_rate": 6.616153255013363e-06,
+      "loss": 0.6584,
+      "step": 26384
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6112134572070116,
+      "learning_rate": 6.6154352372218015e-06,
+      "loss": 0.8215,
+      "step": 26385
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7942276430625343,
+      "learning_rate": 6.614717239136246e-06,
+      "loss": 0.5787,
+      "step": 26386
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6094466628447408,
+      "learning_rate": 6.6139992607608795e-06,
+      "loss": 0.6633,
+      "step": 26387
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.130028206919289,
+      "learning_rate": 6.613281302099881e-06,
+      "loss": 0.6757,
+      "step": 26388
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.466604998481489,
+      "learning_rate": 6.612563363157434e-06,
+      "loss": 0.7129,
+      "step": 26389
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4438153629812527,
+      "learning_rate": 6.61184544393771e-06,
+      "loss": 0.8842,
+      "step": 26390
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4399709101660538,
+      "learning_rate": 6.611127544444897e-06,
+      "loss": 0.7437,
+      "step": 26391
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.531424694209029,
+      "learning_rate": 6.610409664683176e-06,
+      "loss": 0.7986,
+      "step": 26392
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3932821451132842,
+      "learning_rate": 6.609691804656719e-06,
+      "loss": 0.6393,
+      "step": 26393
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6148035989378069,
+      "learning_rate": 6.608973964369709e-06,
+      "loss": 0.5292,
+      "step": 26394
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2869101418069666,
+      "learning_rate": 6.608256143826328e-06,
+      "loss": 0.6399,
+      "step": 26395
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3037372147799737,
+      "learning_rate": 6.607538343030755e-06,
+      "loss": 0.7474,
+      "step": 26396
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5995466579746476,
+      "learning_rate": 6.6068205619871664e-06,
+      "loss": 0.5612,
+      "step": 26397
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3164679977257305,
+      "learning_rate": 6.606102800699741e-06,
+      "loss": 0.7029,
+      "step": 26398
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3301869014794399,
+      "learning_rate": 6.605385059172662e-06,
+      "loss": 0.6613,
+      "step": 26399
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2326913611269759,
+      "learning_rate": 6.6046673374101054e-06,
+      "loss": 0.5957,
+      "step": 26400
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5844349483174462,
+      "learning_rate": 6.60394963541625e-06,
+      "loss": 0.5118,
+      "step": 26401
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.33377300079404,
+      "learning_rate": 6.603231953195272e-06,
+      "loss": 0.7263,
+      "step": 26402
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3696665549267393,
+      "learning_rate": 6.6025142907513596e-06,
+      "loss": 0.615,
+      "step": 26403
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2755261449677915,
+      "learning_rate": 6.601796648088681e-06,
+      "loss": 0.6768,
+      "step": 26404
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.285212638595402,
+      "learning_rate": 6.601079025211418e-06,
+      "loss": 0.7432,
+      "step": 26405
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7082183048003226,
+      "learning_rate": 6.600361422123751e-06,
+      "loss": 0.6273,
+      "step": 26406
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3339874637693319,
+      "learning_rate": 6.599643838829852e-06,
+      "loss": 0.6057,
+      "step": 26407
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3494903060947858,
+      "learning_rate": 6.598926275333906e-06,
+      "loss": 0.6446,
+      "step": 26408
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5677636542477352,
+      "learning_rate": 6.598208731640085e-06,
+      "loss": 0.5197,
+      "step": 26409
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4768356929772808,
+      "learning_rate": 6.597491207752577e-06,
+      "loss": 0.8176,
+      "step": 26410
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3702537778836936,
+      "learning_rate": 6.596773703675545e-06,
+      "loss": 0.6334,
+      "step": 26411
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2724470315185819,
+      "learning_rate": 6.596056219413176e-06,
+      "loss": 0.7052,
+      "step": 26412
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4995200614846955,
+      "learning_rate": 6.595338754969649e-06,
+      "loss": 0.7595,
+      "step": 26413
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3480492738766938,
+      "learning_rate": 6.594621310349134e-06,
+      "loss": 0.6178,
+      "step": 26414
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5883887272529547,
+      "learning_rate": 6.59390388555581e-06,
+      "loss": 0.6349,
+      "step": 26415
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8481277788989077,
+      "learning_rate": 6.593186480593858e-06,
+      "loss": 0.8143,
+      "step": 26416
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4131655748054321,
+      "learning_rate": 6.592469095467456e-06,
+      "loss": 0.652,
+      "step": 26417
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5995346815495044,
+      "learning_rate": 6.591751730180774e-06,
+      "loss": 0.7592,
+      "step": 26418
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3723525232107538,
+      "learning_rate": 6.591034384737992e-06,
+      "loss": 0.7109,
+      "step": 26419
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5422618850677143,
+      "learning_rate": 6.5903170591432896e-06,
+      "loss": 0.7848,
+      "step": 26420
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2533132453067968,
+      "learning_rate": 6.58959975340084e-06,
+      "loss": 0.797,
+      "step": 26421
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.246009973988368,
+      "learning_rate": 6.588882467514819e-06,
+      "loss": 0.7029,
+      "step": 26422
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6574115727026661,
+      "learning_rate": 6.588165201489405e-06,
+      "loss": 0.5698,
+      "step": 26423
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4776627813367067,
+      "learning_rate": 6.587447955328775e-06,
+      "loss": 0.7636,
+      "step": 26424
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5967859472524824,
+      "learning_rate": 6.586730729037102e-06,
+      "loss": 0.7717,
+      "step": 26425
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6173869692165409,
+      "learning_rate": 6.586013522618562e-06,
+      "loss": 0.5337,
+      "step": 26426
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4144768036247146,
+      "learning_rate": 6.585296336077337e-06,
+      "loss": 0.8978,
+      "step": 26427
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2438541988812044,
+      "learning_rate": 6.584579169417592e-06,
+      "loss": 0.7305,
+      "step": 26428
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5521342673021783,
+      "learning_rate": 6.58386202264351e-06,
+      "loss": 0.7631,
+      "step": 26429
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3034326512276704,
+      "learning_rate": 6.583144895759264e-06,
+      "loss": 0.7163,
+      "step": 26430
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4844931436644704,
+      "learning_rate": 6.582427788769033e-06,
+      "loss": 0.6322,
+      "step": 26431
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8885850436140907,
+      "learning_rate": 6.581710701676989e-06,
+      "loss": 0.6837,
+      "step": 26432
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.511014107033764,
+      "learning_rate": 6.580993634487306e-06,
+      "loss": 0.8368,
+      "step": 26433
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4578394790261946,
+      "learning_rate": 6.580276587204163e-06,
+      "loss": 0.6632,
+      "step": 26434
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3239899626807907,
+      "learning_rate": 6.579559559831729e-06,
+      "loss": 0.7305,
+      "step": 26435
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3746071992267588,
+      "learning_rate": 6.5788425523741825e-06,
+      "loss": 0.7831,
+      "step": 26436
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8794691160476904,
+      "learning_rate": 6.578125564835699e-06,
+      "loss": 0.7502,
+      "step": 26437
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5644231319119652,
+      "learning_rate": 6.5774085972204535e-06,
+      "loss": 0.8917,
+      "step": 26438
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.370737329412908,
+      "learning_rate": 6.576691649532615e-06,
+      "loss": 0.6394,
+      "step": 26439
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.641420140171759,
+      "learning_rate": 6.575974721776364e-06,
+      "loss": 0.6907,
+      "step": 26440
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4550303856767732,
+      "learning_rate": 6.575257813955873e-06,
+      "loss": 0.6878,
+      "step": 26441
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4167659779401944,
+      "learning_rate": 6.574540926075314e-06,
+      "loss": 0.6918,
+      "step": 26442
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2446712741659465,
+      "learning_rate": 6.57382405813886e-06,
+      "loss": 0.7151,
+      "step": 26443
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2968081896746644,
+      "learning_rate": 6.573107210150688e-06,
+      "loss": 0.7069,
+      "step": 26444
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2711588991919585,
+      "learning_rate": 6.5723903821149746e-06,
+      "loss": 0.7795,
+      "step": 26445
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.542853363325433,
+      "learning_rate": 6.5716735740358864e-06,
+      "loss": 0.72,
+      "step": 26446
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3372356265444256,
+      "learning_rate": 6.570956785917599e-06,
+      "loss": 0.7223,
+      "step": 26447
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.480248233909668,
+      "learning_rate": 6.57024001776429e-06,
+      "loss": 0.7339,
+      "step": 26448
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7080765331515246,
+      "learning_rate": 6.569523269580127e-06,
+      "loss": 0.7885,
+      "step": 26449
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4239429350445911,
+      "learning_rate": 6.568806541369287e-06,
+      "loss": 0.7164,
+      "step": 26450
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4778270800296365,
+      "learning_rate": 6.568089833135939e-06,
+      "loss": 0.8015,
+      "step": 26451
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.323354478853942,
+      "learning_rate": 6.567373144884262e-06,
+      "loss": 0.5685,
+      "step": 26452
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.77920758048808,
+      "learning_rate": 6.566656476618424e-06,
+      "loss": 0.7906,
+      "step": 26453
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3127879854220095,
+      "learning_rate": 6.565939828342598e-06,
+      "loss": 0.6869,
+      "step": 26454
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6374640023248054,
+      "learning_rate": 6.565223200060961e-06,
+      "loss": 0.4913,
+      "step": 26455
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6083551037866202,
+      "learning_rate": 6.564506591777677e-06,
+      "loss": 0.5374,
+      "step": 26456
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2557471366284012,
+      "learning_rate": 6.563790003496926e-06,
+      "loss": 0.5575,
+      "step": 26457
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5682365173166394,
+      "learning_rate": 6.563073435222874e-06,
+      "loss": 0.4981,
+      "step": 26458
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3148033436722217,
+      "learning_rate": 6.562356886959704e-06,
+      "loss": 0.6154,
+      "step": 26459
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.671521568131599,
+      "learning_rate": 6.561640358711574e-06,
+      "loss": 0.6875,
+      "step": 26460
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6270025540125483,
+      "learning_rate": 6.560923850482665e-06,
+      "loss": 0.6616,
+      "step": 26461
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.344492831444433,
+      "learning_rate": 6.560207362277146e-06,
+      "loss": 0.6328,
+      "step": 26462
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4954786048073898,
+      "learning_rate": 6.55949089409919e-06,
+      "loss": 0.6452,
+      "step": 26463
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3011030289675531,
+      "learning_rate": 6.558774445952962e-06,
+      "loss": 0.6143,
+      "step": 26464
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5312501154744915,
+      "learning_rate": 6.558058017842643e-06,
+      "loss": 0.7302,
+      "step": 26465
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9182540123063792,
+      "learning_rate": 6.5573416097724e-06,
+      "loss": 0.7304,
+      "step": 26466
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3452580147447533,
+      "learning_rate": 6.5566252217464045e-06,
+      "loss": 0.6517,
+      "step": 26467
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.3855695185457213,
+      "learning_rate": 6.555908853768824e-06,
+      "loss": 0.6425,
+      "step": 26468
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.472719631446548,
+      "learning_rate": 6.555192505843836e-06,
+      "loss": 0.8082,
+      "step": 26469
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.622683169473407,
+      "learning_rate": 6.554476177975606e-06,
+      "loss": 0.7818,
+      "step": 26470
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7865875013380466,
+      "learning_rate": 6.553759870168306e-06,
+      "loss": 0.7359,
+      "step": 26471
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4033197137262112,
+      "learning_rate": 6.553043582426106e-06,
+      "loss": 0.6877,
+      "step": 26472
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2638043713095826,
+      "learning_rate": 6.552327314753182e-06,
+      "loss": 0.6373,
+      "step": 26473
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3887142172614813,
+      "learning_rate": 6.551611067153698e-06,
+      "loss": 0.7462,
+      "step": 26474
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.370608323569564,
+      "learning_rate": 6.550894839631823e-06,
+      "loss": 0.6746,
+      "step": 26475
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5894275813411264,
+      "learning_rate": 6.550178632191737e-06,
+      "loss": 0.5029,
+      "step": 26476
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5302504848610243,
+      "learning_rate": 6.5494624448375964e-06,
+      "loss": 0.8317,
+      "step": 26477
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3935773199245327,
+      "learning_rate": 6.548746277573581e-06,
+      "loss": 0.8262,
+      "step": 26478
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3416809804108039,
+      "learning_rate": 6.548030130403854e-06,
+      "loss": 0.6742,
+      "step": 26479
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.514246463908368,
+      "learning_rate": 6.5473140033325935e-06,
+      "loss": 0.6063,
+      "step": 26480
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6649547787523427,
+      "learning_rate": 6.546597896363962e-06,
+      "loss": 0.6812,
+      "step": 26481
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7981211577537914,
+      "learning_rate": 6.54588180950213e-06,
+      "loss": 0.7744,
+      "step": 26482
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.529396837044541,
+      "learning_rate": 6.545165742751269e-06,
+      "loss": 0.6609,
+      "step": 26483
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.624005033565605,
+      "learning_rate": 6.544449696115545e-06,
+      "loss": 0.828,
+      "step": 26484
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4221615642555852,
+      "learning_rate": 6.543733669599131e-06,
+      "loss": 0.6912,
+      "step": 26485
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6029723792229543,
+      "learning_rate": 6.543017663206192e-06,
+      "loss": 0.4481,
+      "step": 26486
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7084263422775237,
+      "learning_rate": 6.542301676940901e-06,
+      "loss": 0.697,
+      "step": 26487
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.0963135571351015,
+      "learning_rate": 6.541585710807421e-06,
+      "loss": 0.771,
+      "step": 26488
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7218999421226067,
+      "learning_rate": 6.540869764809925e-06,
+      "loss": 0.9078,
+      "step": 26489
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2795073734841882,
+      "learning_rate": 6.5401538389525835e-06,
+      "loss": 0.6336,
+      "step": 26490
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5921459516326288,
+      "learning_rate": 6.539437933239559e-06,
+      "loss": 0.8035,
+      "step": 26491
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3985386973665779,
+      "learning_rate": 6.53872204767502e-06,
+      "loss": 0.6078,
+      "step": 26492
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6304339991817012,
+      "learning_rate": 6.538006182263139e-06,
+      "loss": 0.5716,
+      "step": 26493
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.323735421521744,
+      "learning_rate": 6.537290337008085e-06,
+      "loss": 0.7086,
+      "step": 26494
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.507493753276386,
+      "learning_rate": 6.536574511914021e-06,
+      "loss": 0.7448,
+      "step": 26495
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6020215534570532,
+      "learning_rate": 6.535858706985114e-06,
+      "loss": 0.7835,
+      "step": 26496
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.57719259259084,
+      "learning_rate": 6.5351429222255394e-06,
+      "loss": 0.5433,
+      "step": 26497
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6018766859583424,
+      "learning_rate": 6.5344271576394555e-06,
+      "loss": 0.5129,
+      "step": 26498
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7680478832812176,
+      "learning_rate": 6.533711413231036e-06,
+      "loss": 0.5869,
+      "step": 26499
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3725792128158534,
+      "learning_rate": 6.532995689004444e-06,
+      "loss": 0.7366,
+      "step": 26500
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4139843803363594,
+      "learning_rate": 6.532279984963852e-06,
+      "loss": 0.7536,
+      "step": 26501
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5686036269302923,
+      "learning_rate": 6.531564301113422e-06,
+      "loss": 0.7725,
+      "step": 26502
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5104044439311786,
+      "learning_rate": 6.530848637457322e-06,
+      "loss": 0.7606,
+      "step": 26503
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.861766925749399,
+      "learning_rate": 6.530132993999723e-06,
+      "loss": 0.6409,
+      "step": 26504
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3436078729078682,
+      "learning_rate": 6.529417370744783e-06,
+      "loss": 0.6922,
+      "step": 26505
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5906798553999573,
+      "learning_rate": 6.528701767696676e-06,
+      "loss": 0.6736,
+      "step": 26506
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.434823053774615,
+      "learning_rate": 6.527986184859565e-06,
+      "loss": 0.6159,
+      "step": 26507
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6339128041374626,
+      "learning_rate": 6.527270622237623e-06,
+      "loss": 0.8089,
+      "step": 26508
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6929152113056853,
+      "learning_rate": 6.526555079835004e-06,
+      "loss": 0.7098,
+      "step": 26509
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3756580063986517,
+      "learning_rate": 6.525839557655884e-06,
+      "loss": 0.7248,
+      "step": 26510
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3093144288737657,
+      "learning_rate": 6.525124055704425e-06,
+      "loss": 0.7708,
+      "step": 26511
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5970989550933796,
+      "learning_rate": 6.5244085739848e-06,
+      "loss": 0.6847,
+      "step": 26512
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.400338271171131,
+      "learning_rate": 6.52369311250116e-06,
+      "loss": 0.7573,
+      "step": 26513
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4517664881646777,
+      "learning_rate": 6.522977671257683e-06,
+      "loss": 0.8256,
+      "step": 26514
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5833275215923539,
+      "learning_rate": 6.522262250258533e-06,
+      "loss": 0.5231,
+      "step": 26515
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5379636739584925,
+      "learning_rate": 6.521546849507872e-06,
+      "loss": 0.7687,
+      "step": 26516
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3244383765404328,
+      "learning_rate": 6.520831469009863e-06,
+      "loss": 0.7179,
+      "step": 26517
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3351665425938848,
+      "learning_rate": 6.520116108768677e-06,
+      "loss": 0.6943,
+      "step": 26518
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.514928887581085,
+      "learning_rate": 6.519400768788478e-06,
+      "loss": 0.6477,
+      "step": 26519
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.302564009261434,
+      "learning_rate": 6.518685449073427e-06,
+      "loss": 0.6876,
+      "step": 26520
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3139729385829217,
+      "learning_rate": 6.517970149627692e-06,
+      "loss": 0.6255,
+      "step": 26521
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5282575964628011,
+      "learning_rate": 6.5172548704554405e-06,
+      "loss": 0.6245,
+      "step": 26522
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3373290808888492,
+      "learning_rate": 6.5165396115608315e-06,
+      "loss": 0.5701,
+      "step": 26523
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.635864039782511,
+      "learning_rate": 6.51582437294803e-06,
+      "loss": 0.5317,
+      "step": 26524
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.353422055761939,
+      "learning_rate": 6.5151091546212034e-06,
+      "loss": 0.7593,
+      "step": 26525
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3355461067697383,
+      "learning_rate": 6.5143939565845175e-06,
+      "loss": 0.6828,
+      "step": 26526
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6240087247501683,
+      "learning_rate": 6.513678778842129e-06,
+      "loss": 0.5203,
+      "step": 26527
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3051024819128647,
+      "learning_rate": 6.512963621398207e-06,
+      "loss": 0.5895,
+      "step": 26528
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5989498265225867,
+      "learning_rate": 6.512248484256918e-06,
+      "loss": 0.4899,
+      "step": 26529
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3909560023551568,
+      "learning_rate": 6.51153336742242e-06,
+      "loss": 0.64,
+      "step": 26530
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3445739792306661,
+      "learning_rate": 6.51081827089888e-06,
+      "loss": 0.7076,
+      "step": 26531
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.0054546822769823,
+      "learning_rate": 6.510103194690459e-06,
+      "loss": 0.8193,
+      "step": 26532
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4786249499313744,
+      "learning_rate": 6.509388138801326e-06,
+      "loss": 0.7083,
+      "step": 26533
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.548321804781315,
+      "learning_rate": 6.508673103235637e-06,
+      "loss": 0.7995,
+      "step": 26534
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4340530126776145,
+      "learning_rate": 6.507958087997559e-06,
+      "loss": 0.8166,
+      "step": 26535
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0819848662432867,
+      "learning_rate": 6.507243093091258e-06,
+      "loss": 0.6009,
+      "step": 26536
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8135281900036635,
+      "learning_rate": 6.506528118520888e-06,
+      "loss": 0.8203,
+      "step": 26537
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7788340336174584,
+      "learning_rate": 6.50581316429062e-06,
+      "loss": 0.7262,
+      "step": 26538
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6315039094944741,
+      "learning_rate": 6.505098230404614e-06,
+      "loss": 0.563,
+      "step": 26539
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6130576619937207,
+      "learning_rate": 6.504383316867035e-06,
+      "loss": 0.7313,
+      "step": 26540
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.304076629973347,
+      "learning_rate": 6.503668423682037e-06,
+      "loss": 0.5554,
+      "step": 26541
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4505561725426592,
+      "learning_rate": 6.502953550853792e-06,
+      "loss": 0.7082,
+      "step": 26542
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4859726102496191,
+      "learning_rate": 6.50223869838646e-06,
+      "loss": 0.6847,
+      "step": 26543
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4227979401877304,
+      "learning_rate": 6.501523866284201e-06,
+      "loss": 0.7189,
+      "step": 26544
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6077997305548797,
+      "learning_rate": 6.500809054551173e-06,
+      "loss": 0.6774,
+      "step": 26545
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.376396512602201,
+      "learning_rate": 6.500094263191548e-06,
+      "loss": 0.6612,
+      "step": 26546
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5354272726720775,
+      "learning_rate": 6.499379492209481e-06,
+      "loss": 0.9344,
+      "step": 26547
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6195814170480217,
+      "learning_rate": 6.498664741609135e-06,
+      "loss": 0.5503,
+      "step": 26548
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4690444076497184,
+      "learning_rate": 6.497950011394669e-06,
+      "loss": 0.72,
+      "step": 26549
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8757813428179095,
+      "learning_rate": 6.497235301570251e-06,
+      "loss": 0.6904,
+      "step": 26550
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8457166170533477,
+      "learning_rate": 6.496520612140033e-06,
+      "loss": 0.6171,
+      "step": 26551
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5430550383944772,
+      "learning_rate": 6.4958059431081845e-06,
+      "loss": 0.5057,
+      "step": 26552
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4706854108856402,
+      "learning_rate": 6.4950912944788614e-06,
+      "loss": 0.7234,
+      "step": 26553
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6618233948814811,
+      "learning_rate": 6.494376666256229e-06,
+      "loss": 0.7361,
+      "step": 26554
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3575271217771696,
+      "learning_rate": 6.493662058444443e-06,
+      "loss": 0.632,
+      "step": 26555
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2065939678220212,
+      "learning_rate": 6.492947471047666e-06,
+      "loss": 0.7074,
+      "step": 26556
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5962861369763903,
+      "learning_rate": 6.492232904070065e-06,
+      "loss": 0.6848,
+      "step": 26557
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4780064794270809,
+      "learning_rate": 6.491518357515788e-06,
+      "loss": 0.7953,
+      "step": 26558
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4485552449161239,
+      "learning_rate": 6.490803831389003e-06,
+      "loss": 0.6948,
+      "step": 26559
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5112648684136978,
+      "learning_rate": 6.490089325693869e-06,
+      "loss": 0.7477,
+      "step": 26560
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3808672342125659,
+      "learning_rate": 6.489374840434551e-06,
+      "loss": 0.7121,
+      "step": 26561
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3725395388333952,
+      "learning_rate": 6.4886603756151985e-06,
+      "loss": 0.6718,
+      "step": 26562
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5263579049921863,
+      "learning_rate": 6.487945931239978e-06,
+      "loss": 0.6986,
+      "step": 26563
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.415504235319656,
+      "learning_rate": 6.487231507313052e-06,
+      "loss": 0.6429,
+      "step": 26564
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.261248085987737,
+      "learning_rate": 6.486517103838574e-06,
+      "loss": 0.7489,
+      "step": 26565
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5614992049019896,
+      "learning_rate": 6.485802720820702e-06,
+      "loss": 0.8593,
+      "step": 26566
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3435134857801094,
+      "learning_rate": 6.485088358263603e-06,
+      "loss": 0.7064,
+      "step": 26567
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.530862987209075,
+      "learning_rate": 6.4843740161714334e-06,
+      "loss": 0.7255,
+      "step": 26568
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.607570752469201,
+      "learning_rate": 6.483659694548348e-06,
+      "loss": 0.6589,
+      "step": 26569
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4271914455218864,
+      "learning_rate": 6.4829453933985096e-06,
+      "loss": 0.7097,
+      "step": 26570
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4648644185396225,
+      "learning_rate": 6.482231112726081e-06,
+      "loss": 0.6431,
+      "step": 26571
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2503453810035046,
+      "learning_rate": 6.4815168525352124e-06,
+      "loss": 0.671,
+      "step": 26572
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4610256428591626,
+      "learning_rate": 6.4808026128300665e-06,
+      "loss": 0.7159,
+      "step": 26573
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3506058090273585,
+      "learning_rate": 6.480088393614802e-06,
+      "loss": 0.6177,
+      "step": 26574
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.131712870591743,
+      "learning_rate": 6.4793741948935795e-06,
+      "loss": 0.6098,
+      "step": 26575
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4261038666594783,
+      "learning_rate": 6.478660016670555e-06,
+      "loss": 0.6893,
+      "step": 26576
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6070468844758166,
+      "learning_rate": 6.4779458589498835e-06,
+      "loss": 0.7002,
+      "step": 26577
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.488369799383684,
+      "learning_rate": 6.477231721735729e-06,
+      "loss": 0.6906,
+      "step": 26578
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3386041142600478,
+      "learning_rate": 6.4765176050322465e-06,
+      "loss": 0.722,
+      "step": 26579
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5909943956927917,
+      "learning_rate": 6.4758035088435935e-06,
+      "loss": 0.6482,
+      "step": 26580
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8403160921651727,
+      "learning_rate": 6.475089433173926e-06,
+      "loss": 0.5899,
+      "step": 26581
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5996640792662947,
+      "learning_rate": 6.474375378027408e-06,
+      "loss": 0.4909,
+      "step": 26582
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.437905596139358,
+      "learning_rate": 6.473661343408191e-06,
+      "loss": 0.6131,
+      "step": 26583
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5131191912725883,
+      "learning_rate": 6.472947329320434e-06,
+      "loss": 0.7276,
+      "step": 26584
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9460366733030083,
+      "learning_rate": 6.472233335768295e-06,
+      "loss": 0.7307,
+      "step": 26585
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5065261197309376,
+      "learning_rate": 6.471519362755927e-06,
+      "loss": 0.6969,
+      "step": 26586
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5377481942659751,
+      "learning_rate": 6.470805410287493e-06,
+      "loss": 0.4292,
+      "step": 26587
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3786273954229449,
+      "learning_rate": 6.470091478367147e-06,
+      "loss": 0.6211,
+      "step": 26588
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2931312414205143,
+      "learning_rate": 6.469377566999049e-06,
+      "loss": 0.7028,
+      "step": 26589
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4489374593272288,
+      "learning_rate": 6.468663676187346e-06,
+      "loss": 0.6981,
+      "step": 26590
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3012667095093793,
+      "learning_rate": 6.467949805936206e-06,
+      "loss": 0.7508,
+      "step": 26591
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4760320477825477,
+      "learning_rate": 6.46723595624978e-06,
+      "loss": 0.6763,
+      "step": 26592
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6214110595432526,
+      "learning_rate": 6.4665221271322245e-06,
+      "loss": 0.5073,
+      "step": 26593
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6255364752864458,
+      "learning_rate": 6.465808318587693e-06,
+      "loss": 0.7433,
+      "step": 26594
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5541873106120525,
+      "learning_rate": 6.465094530620346e-06,
+      "loss": 0.7114,
+      "step": 26595
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4342761102827446,
+      "learning_rate": 6.464380763234341e-06,
+      "loss": 0.755,
+      "step": 26596
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2801625146568731,
+      "learning_rate": 6.463667016433828e-06,
+      "loss": 0.7146,
+      "step": 26597
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3981583134861886,
+      "learning_rate": 6.462953290222963e-06,
+      "loss": 0.6751,
+      "step": 26598
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2943827784551183,
+      "learning_rate": 6.462239584605907e-06,
+      "loss": 0.6854,
+      "step": 26599
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4594924480085991,
+      "learning_rate": 6.461525899586812e-06,
+      "loss": 0.6985,
+      "step": 26600
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.062470276019075,
+      "learning_rate": 6.460812235169832e-06,
+      "loss": 0.7853,
+      "step": 26601
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5884347791861272,
+      "learning_rate": 6.460098591359123e-06,
+      "loss": 0.5018,
+      "step": 26602
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2982751771187344,
+      "learning_rate": 6.459384968158844e-06,
+      "loss": 0.715,
+      "step": 26603
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5410610092081756,
+      "learning_rate": 6.458671365573144e-06,
+      "loss": 0.7465,
+      "step": 26604
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3164457776660015,
+      "learning_rate": 6.457957783606179e-06,
+      "loss": 0.7161,
+      "step": 26605
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3100700865214594,
+      "learning_rate": 6.457244222262111e-06,
+      "loss": 0.6844,
+      "step": 26606
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.357266201961414,
+      "learning_rate": 6.456530681545083e-06,
+      "loss": 0.7247,
+      "step": 26607
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2977873224863725,
+      "learning_rate": 6.455817161459257e-06,
+      "loss": 0.6444,
+      "step": 26608
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9701275266829334,
+      "learning_rate": 6.455103662008784e-06,
+      "loss": 0.6562,
+      "step": 26609
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6078436896019874,
+      "learning_rate": 6.4543901831978254e-06,
+      "loss": 0.8158,
+      "step": 26610
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6269049248074595,
+      "learning_rate": 6.453676725030524e-06,
+      "loss": 0.576,
+      "step": 26611
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4551061873162145,
+      "learning_rate": 6.452963287511041e-06,
+      "loss": 0.7345,
+      "step": 26612
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.596593948564543,
+      "learning_rate": 6.452249870643531e-06,
+      "loss": 0.699,
+      "step": 26613
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3997792495543904,
+      "learning_rate": 6.451536474432143e-06,
+      "loss": 0.6384,
+      "step": 26614
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.28603255815111,
+      "learning_rate": 6.450823098881032e-06,
+      "loss": 0.6998,
+      "step": 26615
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.944793373100776,
+      "learning_rate": 6.450109743994353e-06,
+      "loss": 0.5463,
+      "step": 26616
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5766878617512797,
+      "learning_rate": 6.449396409776263e-06,
+      "loss": 0.561,
+      "step": 26617
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4558044369012952,
+      "learning_rate": 6.448683096230906e-06,
+      "loss": 0.7435,
+      "step": 26618
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8629391431951943,
+      "learning_rate": 6.447969803362442e-06,
+      "loss": 0.6798,
+      "step": 26619
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4733964138412292,
+      "learning_rate": 6.447256531175026e-06,
+      "loss": 0.6424,
+      "step": 26620
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4451176963555334,
+      "learning_rate": 6.446543279672803e-06,
+      "loss": 0.791,
+      "step": 26621
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6049606438815336,
+      "learning_rate": 6.445830048859929e-06,
+      "loss": 0.8078,
+      "step": 26622
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1773114142403447,
+      "learning_rate": 6.445116838740559e-06,
+      "loss": 0.8067,
+      "step": 26623
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.45569952752659,
+      "learning_rate": 6.444403649318849e-06,
+      "loss": 0.578,
+      "step": 26624
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3773892345482208,
+      "learning_rate": 6.443690480598942e-06,
+      "loss": 0.625,
+      "step": 26625
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.241069886641756,
+      "learning_rate": 6.4429773325849935e-06,
+      "loss": 0.6228,
+      "step": 26626
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2965737726143078,
+      "learning_rate": 6.442264205281161e-06,
+      "loss": 0.6449,
+      "step": 26627
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.426460797147916,
+      "learning_rate": 6.441551098691591e-06,
+      "loss": 0.6147,
+      "step": 26628
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6439876130712143,
+      "learning_rate": 6.440838012820437e-06,
+      "loss": 0.604,
+      "step": 26629
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2689062787072032,
+      "learning_rate": 6.440124947671848e-06,
+      "loss": 0.6629,
+      "step": 26630
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3792523170475246,
+      "learning_rate": 6.439411903249983e-06,
+      "loss": 0.69,
+      "step": 26631
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.443635058722693,
+      "learning_rate": 6.438698879558989e-06,
+      "loss": 0.731,
+      "step": 26632
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5806014040750288,
+      "learning_rate": 6.437985876603017e-06,
+      "loss": 0.4835,
+      "step": 26633
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5506580365053708,
+      "learning_rate": 6.437272894386219e-06,
+      "loss": 0.7554,
+      "step": 26634
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3348889693719936,
+      "learning_rate": 6.4365599329127435e-06,
+      "loss": 0.6692,
+      "step": 26635
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4646913152695218,
+      "learning_rate": 6.435846992186746e-06,
+      "loss": 0.7302,
+      "step": 26636
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.2657939850536524,
+      "learning_rate": 6.435134072212377e-06,
+      "loss": 0.6001,
+      "step": 26637
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9444140404313373,
+      "learning_rate": 6.4344211729937855e-06,
+      "loss": 0.6444,
+      "step": 26638
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2956228652360544,
+      "learning_rate": 6.43370829453512e-06,
+      "loss": 0.7033,
+      "step": 26639
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7042810874427026,
+      "learning_rate": 6.432995436840537e-06,
+      "loss": 0.7329,
+      "step": 26640
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.702021796617005,
+      "learning_rate": 6.43228259991418e-06,
+      "loss": 0.7146,
+      "step": 26641
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5580309857079397,
+      "learning_rate": 6.43156978376021e-06,
+      "loss": 0.6961,
+      "step": 26642
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.694104249089061,
+      "learning_rate": 6.430856988382766e-06,
+      "loss": 0.7452,
+      "step": 26643
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5666674554152393,
+      "learning_rate": 6.430144213786002e-06,
+      "loss": 0.7107,
+      "step": 26644
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.261492691993402,
+      "learning_rate": 6.4294314599740715e-06,
+      "loss": 0.7176,
+      "step": 26645
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5998041051299736,
+      "learning_rate": 6.42871872695112e-06,
+      "loss": 0.5315,
+      "step": 26646
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4117198037652785,
+      "learning_rate": 6.428006014721295e-06,
+      "loss": 0.7898,
+      "step": 26647
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4096714379032889,
+      "learning_rate": 6.427293323288754e-06,
+      "loss": 0.5548,
+      "step": 26648
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.666489571716614,
+      "learning_rate": 6.426580652657643e-06,
+      "loss": 0.7396,
+      "step": 26649
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6557834736960075,
+      "learning_rate": 6.4258680028321094e-06,
+      "loss": 0.6739,
+      "step": 26650
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4335116426057926,
+      "learning_rate": 6.425155373816303e-06,
+      "loss": 0.6569,
+      "step": 26651
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4785671455069294,
+      "learning_rate": 6.424442765614376e-06,
+      "loss": 0.6977,
+      "step": 26652
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5808502164576752,
+      "learning_rate": 6.423730178230474e-06,
+      "loss": 0.5103,
+      "step": 26653
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6217842276705972,
+      "learning_rate": 6.423017611668745e-06,
+      "loss": 0.7388,
+      "step": 26654
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7417077825729146,
+      "learning_rate": 6.422305065933342e-06,
+      "loss": 0.7865,
+      "step": 26655
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2507371321455807,
+      "learning_rate": 6.421592541028414e-06,
+      "loss": 0.7381,
+      "step": 26656
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.351466406962846,
+      "learning_rate": 6.420880036958103e-06,
+      "loss": 0.742,
+      "step": 26657
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7264850023277543,
+      "learning_rate": 6.4201675537265615e-06,
+      "loss": 0.8231,
+      "step": 26658
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5992957600995934,
+      "learning_rate": 6.419455091337944e-06,
+      "loss": 0.6826,
+      "step": 26659
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6934661080781526,
+      "learning_rate": 6.4187426497963855e-06,
+      "loss": 0.8568,
+      "step": 26660
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.412979238057827,
+      "learning_rate": 6.418030229106045e-06,
+      "loss": 0.7149,
+      "step": 26661
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3287766354855952,
+      "learning_rate": 6.417317829271062e-06,
+      "loss": 0.6207,
+      "step": 26662
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4393764540069696,
+      "learning_rate": 6.4166054502955965e-06,
+      "loss": 0.7804,
+      "step": 26663
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6110647129282623,
+      "learning_rate": 6.415893092183782e-06,
+      "loss": 0.6551,
+      "step": 26664
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5857204107121263,
+      "learning_rate": 6.415180754939775e-06,
+      "loss": 0.4793,
+      "step": 26665
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6507440204884429,
+      "learning_rate": 6.414468438567722e-06,
+      "loss": 0.5115,
+      "step": 26666
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.439856399877325,
+      "learning_rate": 6.413756143071765e-06,
+      "loss": 0.6504,
+      "step": 26667
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8407177194546034,
+      "learning_rate": 6.413043868456057e-06,
+      "loss": 0.7081,
+      "step": 26668
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.045193965558252,
+      "learning_rate": 6.4123316147247435e-06,
+      "loss": 0.6807,
+      "step": 26669
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6007831235842978,
+      "learning_rate": 6.4116193818819735e-06,
+      "loss": 0.7278,
+      "step": 26670
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4667703755752695,
+      "learning_rate": 6.410907169931888e-06,
+      "loss": 0.6848,
+      "step": 26671
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5914572109449124,
+      "learning_rate": 6.410194978878638e-06,
+      "loss": 0.5237,
+      "step": 26672
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.419512272683832,
+      "learning_rate": 6.409482808726372e-06,
+      "loss": 0.6473,
+      "step": 26673
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3416319671524917,
+      "learning_rate": 6.408770659479231e-06,
+      "loss": 0.6059,
+      "step": 26674
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.535898829009146,
+      "learning_rate": 6.408058531141363e-06,
+      "loss": 0.684,
+      "step": 26675
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3565524990332325,
+      "learning_rate": 6.407346423716918e-06,
+      "loss": 0.7021,
+      "step": 26676
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5717715128106345,
+      "learning_rate": 6.406634337210041e-06,
+      "loss": 0.5354,
+      "step": 26677
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4464250489738233,
+      "learning_rate": 6.405922271624874e-06,
+      "loss": 0.8258,
+      "step": 26678
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5879088711887311,
+      "learning_rate": 6.405210226965564e-06,
+      "loss": 0.5698,
+      "step": 26679
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6328986980903604,
+      "learning_rate": 6.404498203236261e-06,
+      "loss": 0.7542,
+      "step": 26680
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.601107823546958,
+      "learning_rate": 6.403786200441106e-06,
+      "loss": 0.667,
+      "step": 26681
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6221969277056115,
+      "learning_rate": 6.403074218584248e-06,
+      "loss": 0.5467,
+      "step": 26682
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3330309348515734,
+      "learning_rate": 6.402362257669827e-06,
+      "loss": 0.5928,
+      "step": 26683
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4219726817595268,
+      "learning_rate": 6.401650317701997e-06,
+      "loss": 0.7531,
+      "step": 26684
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6851705716555225,
+      "learning_rate": 6.400938398684895e-06,
+      "loss": 0.723,
+      "step": 26685
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5002277968938793,
+      "learning_rate": 6.40022650062267e-06,
+      "loss": 0.7202,
+      "step": 26686
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5300847594421307,
+      "learning_rate": 6.399514623519467e-06,
+      "loss": 0.7581,
+      "step": 26687
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2696972139018636,
+      "learning_rate": 6.398802767379426e-06,
+      "loss": 0.7757,
+      "step": 26688
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2222917550655057,
+      "learning_rate": 6.398090932206698e-06,
+      "loss": 0.7256,
+      "step": 26689
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.9999592916771274,
+      "learning_rate": 6.397379118005423e-06,
+      "loss": 0.7895,
+      "step": 26690
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6297630936562297,
+      "learning_rate": 6.396667324779754e-06,
+      "loss": 0.6255,
+      "step": 26691
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8254544004451025,
+      "learning_rate": 6.395955552533821e-06,
+      "loss": 0.7546,
+      "step": 26692
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.619490999139122,
+      "learning_rate": 6.395243801271779e-06,
+      "loss": 0.6197,
+      "step": 26693
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.418661004750706,
+      "learning_rate": 6.39453207099777e-06,
+      "loss": 0.6802,
+      "step": 26694
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6224977659622042,
+      "learning_rate": 6.393820361715935e-06,
+      "loss": 0.7887,
+      "step": 26695
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4342894555729002,
+      "learning_rate": 6.393108673430418e-06,
+      "loss": 0.8619,
+      "step": 26696
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2464992022994166,
+      "learning_rate": 6.392397006145366e-06,
+      "loss": 0.5599,
+      "step": 26697
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3897983664144207,
+      "learning_rate": 6.391685359864923e-06,
+      "loss": 0.6764,
+      "step": 26698
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.395514024445547,
+      "learning_rate": 6.390973734593227e-06,
+      "loss": 0.6951,
+      "step": 26699
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.440299309885523,
+      "learning_rate": 6.390262130334424e-06,
+      "loss": 0.7094,
+      "step": 26700
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3515727676039666,
+      "learning_rate": 6.389550547092661e-06,
+      "loss": 0.8008,
+      "step": 26701
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.738434960204969,
+      "learning_rate": 6.388838984872077e-06,
+      "loss": 0.5913,
+      "step": 26702
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3958182627024343,
+      "learning_rate": 6.388127443676811e-06,
+      "loss": 0.6797,
+      "step": 26703
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3485630894916534,
+      "learning_rate": 6.3874159235110144e-06,
+      "loss": 0.5763,
+      "step": 26704
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2764768486089184,
+      "learning_rate": 6.386704424378828e-06,
+      "loss": 0.719,
+      "step": 26705
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6138470662924251,
+      "learning_rate": 6.38599294628439e-06,
+      "loss": 0.609,
+      "step": 26706
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.486796512222008,
+      "learning_rate": 6.385281489231843e-06,
+      "loss": 0.7062,
+      "step": 26707
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.471980112993225,
+      "learning_rate": 6.384570053225338e-06,
+      "loss": 0.5241,
+      "step": 26708
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5781881021500188,
+      "learning_rate": 6.383858638269005e-06,
+      "loss": 0.7122,
+      "step": 26709
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2404079011151075,
+      "learning_rate": 6.383147244366994e-06,
+      "loss": 0.7314,
+      "step": 26710
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7266876467100578,
+      "learning_rate": 6.382435871523442e-06,
+      "loss": 0.6917,
+      "step": 26711
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.412811849976954,
+      "learning_rate": 6.3817245197425005e-06,
+      "loss": 0.6556,
+      "step": 26712
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.395648714400992,
+      "learning_rate": 6.3810131890282975e-06,
+      "loss": 0.9402,
+      "step": 26713
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5420545959051326,
+      "learning_rate": 6.3803018793849835e-06,
+      "loss": 0.6658,
+      "step": 26714
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.468311375308963,
+      "learning_rate": 6.3795905908167e-06,
+      "loss": 0.7592,
+      "step": 26715
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6156665621898798,
+      "learning_rate": 6.378879323327583e-06,
+      "loss": 0.7421,
+      "step": 26716
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5572493507476644,
+      "learning_rate": 6.378168076921778e-06,
+      "loss": 0.7685,
+      "step": 26717
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6014007667509358,
+      "learning_rate": 6.377456851603426e-06,
+      "loss": 0.7385,
+      "step": 26718
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4689474931823763,
+      "learning_rate": 6.37674564737667e-06,
+      "loss": 0.6689,
+      "step": 26719
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3483386905060442,
+      "learning_rate": 6.376034464245641e-06,
+      "loss": 0.7502,
+      "step": 26720
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6057911523055943,
+      "learning_rate": 6.375323302214491e-06,
+      "loss": 0.5263,
+      "step": 26721
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3427246593372124,
+      "learning_rate": 6.374612161287358e-06,
+      "loss": 0.62,
+      "step": 26722
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6106365680937933,
+      "learning_rate": 6.373901041468378e-06,
+      "loss": 0.4942,
+      "step": 26723
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4522665342659502,
+      "learning_rate": 6.373189942761693e-06,
+      "loss": 0.6779,
+      "step": 26724
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4820783813536005,
+      "learning_rate": 6.372478865171446e-06,
+      "loss": 0.6525,
+      "step": 26725
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2972070137758964,
+      "learning_rate": 6.371767808701778e-06,
+      "loss": 0.7706,
+      "step": 26726
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.264211641412873,
+      "learning_rate": 6.371056773356823e-06,
+      "loss": 0.6191,
+      "step": 26727
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3936484040586776,
+      "learning_rate": 6.370345759140724e-06,
+      "loss": 0.6603,
+      "step": 26728
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2607471984008185,
+      "learning_rate": 6.369634766057624e-06,
+      "loss": 0.7933,
+      "step": 26729
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3092967000417017,
+      "learning_rate": 6.368923794111659e-06,
+      "loss": 0.6802,
+      "step": 26730
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3187722441885843,
+      "learning_rate": 6.368212843306968e-06,
+      "loss": 0.6763,
+      "step": 26731
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.59801035269363,
+      "learning_rate": 6.36750191364769e-06,
+      "loss": 0.7399,
+      "step": 26732
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.416133069961905,
+      "learning_rate": 6.36679100513797e-06,
+      "loss": 0.7733,
+      "step": 26733
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5584468486064451,
+      "learning_rate": 6.366080117781941e-06,
+      "loss": 0.814,
+      "step": 26734
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6108383421494454,
+      "learning_rate": 6.365369251583744e-06,
+      "loss": 0.5091,
+      "step": 26735
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4402859140124986,
+      "learning_rate": 6.364658406547519e-06,
+      "loss": 0.5832,
+      "step": 26736
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5126663237014166,
+      "learning_rate": 6.363947582677401e-06,
+      "loss": 0.6752,
+      "step": 26737
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.2256645780940163,
+      "learning_rate": 6.363236779977533e-06,
+      "loss": 0.7587,
+      "step": 26738
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6310407654029118,
+      "learning_rate": 6.362525998452051e-06,
+      "loss": 0.533,
+      "step": 26739
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3967455300480183,
+      "learning_rate": 6.3618152381050975e-06,
+      "loss": 0.6229,
+      "step": 26740
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5345064049281425,
+      "learning_rate": 6.3611044989408046e-06,
+      "loss": 0.7357,
+      "step": 26741
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4088062119042546,
+      "learning_rate": 6.360393780963313e-06,
+      "loss": 0.8084,
+      "step": 26742
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.506725511957515,
+      "learning_rate": 6.359683084176764e-06,
+      "loss": 0.7882,
+      "step": 26743
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3614529393252337,
+      "learning_rate": 6.358972408585291e-06,
+      "loss": 0.7579,
+      "step": 26744
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2839893313353818,
+      "learning_rate": 6.358261754193032e-06,
+      "loss": 0.7328,
+      "step": 26745
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6412759939792068,
+      "learning_rate": 6.357551121004127e-06,
+      "loss": 0.6711,
+      "step": 26746
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5684205024173807,
+      "learning_rate": 6.3568405090227125e-06,
+      "loss": 0.6548,
+      "step": 26747
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3366455419267933,
+      "learning_rate": 6.356129918252927e-06,
+      "loss": 0.6459,
+      "step": 26748
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2997120381259095,
+      "learning_rate": 6.3554193486989035e-06,
+      "loss": 0.6208,
+      "step": 26749
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6012739912640271,
+      "learning_rate": 6.354708800364786e-06,
+      "loss": 0.5114,
+      "step": 26750
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3917293595204654,
+      "learning_rate": 6.353998273254707e-06,
+      "loss": 0.536,
+      "step": 26751
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.561181653845289,
+      "learning_rate": 6.353287767372802e-06,
+      "loss": 0.9276,
+      "step": 26752
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9977927923603407,
+      "learning_rate": 6.3525772827232115e-06,
+      "loss": 0.6806,
+      "step": 26753
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.435198322310238,
+      "learning_rate": 6.351866819310073e-06,
+      "loss": 0.6034,
+      "step": 26754
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4286356204015696,
+      "learning_rate": 6.35115637713752e-06,
+      "loss": 0.8214,
+      "step": 26755
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.352816273434443,
+      "learning_rate": 6.350445956209687e-06,
+      "loss": 0.7307,
+      "step": 26756
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3435956771063569,
+      "learning_rate": 6.349735556530718e-06,
+      "loss": 0.7619,
+      "step": 26757
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3202798933988698,
+      "learning_rate": 6.34902517810474e-06,
+      "loss": 0.6359,
+      "step": 26758
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4431646831256018,
+      "learning_rate": 6.348314820935895e-06,
+      "loss": 0.6737,
+      "step": 26759
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.215520042875954,
+      "learning_rate": 6.347604485028314e-06,
+      "loss": 0.6787,
+      "step": 26760
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4571657828679025,
+      "learning_rate": 6.346894170386143e-06,
+      "loss": 0.7518,
+      "step": 26761
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.758237429822664,
+      "learning_rate": 6.346183877013505e-06,
+      "loss": 0.7325,
+      "step": 26762
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2504923659128624,
+      "learning_rate": 6.345473604914542e-06,
+      "loss": 0.7028,
+      "step": 26763
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5998698053871956,
+      "learning_rate": 6.344763354093388e-06,
+      "loss": 0.4877,
+      "step": 26764
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3458122274742852,
+      "learning_rate": 6.344053124554183e-06,
+      "loss": 0.7847,
+      "step": 26765
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1302128310387476,
+      "learning_rate": 6.343342916301055e-06,
+      "loss": 0.5645,
+      "step": 26766
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3498983111303533,
+      "learning_rate": 6.342632729338143e-06,
+      "loss": 0.7585,
+      "step": 26767
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8294165109191949,
+      "learning_rate": 6.341922563669582e-06,
+      "loss": 0.6231,
+      "step": 26768
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6551676005487312,
+      "learning_rate": 6.341212419299503e-06,
+      "loss": 0.7948,
+      "step": 26769
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4135103906582216,
+      "learning_rate": 6.340502296232046e-06,
+      "loss": 0.8025,
+      "step": 26770
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5864200122017061,
+      "learning_rate": 6.3397921944713435e-06,
+      "loss": 0.5062,
+      "step": 26771
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6074820222601874,
+      "learning_rate": 6.339082114021531e-06,
+      "loss": 0.5436,
+      "step": 26772
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4657351518556054,
+      "learning_rate": 6.3383720548867366e-06,
+      "loss": 0.7605,
+      "step": 26773
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9275547760613292,
+      "learning_rate": 6.3376620170711026e-06,
+      "loss": 0.6568,
+      "step": 26774
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.308659944810575,
+      "learning_rate": 6.336952000578761e-06,
+      "loss": 0.7415,
+      "step": 26775
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3517087609489655,
+      "learning_rate": 6.336242005413843e-06,
+      "loss": 0.7587,
+      "step": 26776
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3924990028947286,
+      "learning_rate": 6.335532031580482e-06,
+      "loss": 0.7061,
+      "step": 26777
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.81075227515685,
+      "learning_rate": 6.334822079082815e-06,
+      "loss": 0.7942,
+      "step": 26778
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5439771937754339,
+      "learning_rate": 6.334112147924978e-06,
+      "loss": 0.8264,
+      "step": 26779
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4168799010453585,
+      "learning_rate": 6.333402238111097e-06,
+      "loss": 0.6582,
+      "step": 26780
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3795292619929822,
+      "learning_rate": 6.332692349645307e-06,
+      "loss": 0.789,
+      "step": 26781
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2748425567778092,
+      "learning_rate": 6.331982482531748e-06,
+      "loss": 0.6953,
+      "step": 26782
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.426369031030561,
+      "learning_rate": 6.331272636774546e-06,
+      "loss": 0.7099,
+      "step": 26783
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3801512070150457,
+      "learning_rate": 6.330562812377836e-06,
+      "loss": 0.6742,
+      "step": 26784
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4902290779402565,
+      "learning_rate": 6.329853009345748e-06,
+      "loss": 0.7627,
+      "step": 26785
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3876765443343435,
+      "learning_rate": 6.329143227682423e-06,
+      "loss": 0.8405,
+      "step": 26786
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7999113639669344,
+      "learning_rate": 6.328433467391986e-06,
+      "loss": 0.6649,
+      "step": 26787
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8587198676216983,
+      "learning_rate": 6.3277237284785695e-06,
+      "loss": 0.9246,
+      "step": 26788
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8686294028490569,
+      "learning_rate": 6.327014010946315e-06,
+      "loss": 0.6985,
+      "step": 26789
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5875877591302165,
+      "learning_rate": 6.326304314799342e-06,
+      "loss": 0.6456,
+      "step": 26790
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4904372814400637,
+      "learning_rate": 6.32559464004179e-06,
+      "loss": 0.7291,
+      "step": 26791
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.314342614760163,
+      "learning_rate": 6.324884986677786e-06,
+      "loss": 0.7567,
+      "step": 26792
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4319268278881472,
+      "learning_rate": 6.324175354711473e-06,
+      "loss": 0.7215,
+      "step": 26793
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6007468338881417,
+      "learning_rate": 6.323465744146967e-06,
+      "loss": 0.7528,
+      "step": 26794
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5026117848832317,
+      "learning_rate": 6.32275615498841e-06,
+      "loss": 0.7434,
+      "step": 26795
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6100710798260618,
+      "learning_rate": 6.322046587239934e-06,
+      "loss": 0.674,
+      "step": 26796
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.693957777436447,
+      "learning_rate": 6.321337040905664e-06,
+      "loss": 0.5676,
+      "step": 26797
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4795008370135623,
+      "learning_rate": 6.320627515989733e-06,
+      "loss": 0.7802,
+      "step": 26798
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3540424638124424,
+      "learning_rate": 6.319918012496276e-06,
+      "loss": 0.7015,
+      "step": 26799
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.478375879656419,
+      "learning_rate": 6.319208530429422e-06,
+      "loss": 0.7037,
+      "step": 26800
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6500408126444157,
+      "learning_rate": 6.318499069793298e-06,
+      "loss": 0.7564,
+      "step": 26801
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3281467668681513,
+      "learning_rate": 6.317789630592039e-06,
+      "loss": 0.6838,
+      "step": 26802
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.784477573571619,
+      "learning_rate": 6.317080212829778e-06,
+      "loss": 0.7557,
+      "step": 26803
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.704736237415582,
+      "learning_rate": 6.31637081651064e-06,
+      "loss": 0.7115,
+      "step": 26804
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.547503467713914,
+      "learning_rate": 6.315661441638754e-06,
+      "loss": 0.6344,
+      "step": 26805
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4084426046973102,
+      "learning_rate": 6.314952088218256e-06,
+      "loss": 0.6743,
+      "step": 26806
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6819680126081666,
+      "learning_rate": 6.314242756253275e-06,
+      "loss": 0.6233,
+      "step": 26807
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7741790324629825,
+      "learning_rate": 6.313533445747939e-06,
+      "loss": 0.8444,
+      "step": 26808
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6434497285250975,
+      "learning_rate": 6.312824156706376e-06,
+      "loss": 0.5387,
+      "step": 26809
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3857174460379549,
+      "learning_rate": 6.312114889132721e-06,
+      "loss": 0.7224,
+      "step": 26810
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4163039970469666,
+      "learning_rate": 6.311405643031101e-06,
+      "loss": 0.8289,
+      "step": 26811
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.310700784853227,
+      "learning_rate": 6.310696418405643e-06,
+      "loss": 0.6922,
+      "step": 26812
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3365938417783023,
+      "learning_rate": 6.309987215260477e-06,
+      "loss": 0.8601,
+      "step": 26813
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7805566337034875,
+      "learning_rate": 6.309278033599738e-06,
+      "loss": 0.6857,
+      "step": 26814
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5996373004848569,
+      "learning_rate": 6.3085688734275485e-06,
+      "loss": 0.4908,
+      "step": 26815
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.271018025276449,
+      "learning_rate": 6.307859734748039e-06,
+      "loss": 0.655,
+      "step": 26816
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.328550538322507,
+      "learning_rate": 6.307150617565343e-06,
+      "loss": 0.7082,
+      "step": 26817
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3265626228152245,
+      "learning_rate": 6.3064415218835795e-06,
+      "loss": 0.6262,
+      "step": 26818
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6028498077390236,
+      "learning_rate": 6.305732447706885e-06,
+      "loss": 0.7501,
+      "step": 26819
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2681679196724462,
+      "learning_rate": 6.305023395039387e-06,
+      "loss": 0.7707,
+      "step": 26820
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8102926532843435,
+      "learning_rate": 6.3043143638852134e-06,
+      "loss": 0.7312,
+      "step": 26821
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3394370750943072,
+      "learning_rate": 6.303605354248488e-06,
+      "loss": 0.6657,
+      "step": 26822
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7498512156173456,
+      "learning_rate": 6.302896366133344e-06,
+      "loss": 0.6979,
+      "step": 26823
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4455563876802204,
+      "learning_rate": 6.302187399543911e-06,
+      "loss": 0.8751,
+      "step": 26824
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8083895153266936,
+      "learning_rate": 6.301478454484311e-06,
+      "loss": 0.6812,
+      "step": 26825
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2883959938079483,
+      "learning_rate": 6.300769530958673e-06,
+      "loss": 0.6623,
+      "step": 26826
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0166250702039696,
+      "learning_rate": 6.300060628971127e-06,
+      "loss": 0.5407,
+      "step": 26827
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.480727510427269,
+      "learning_rate": 6.299351748525802e-06,
+      "loss": 0.8718,
+      "step": 26828
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3494471086358464,
+      "learning_rate": 6.298642889626821e-06,
+      "loss": 0.599,
+      "step": 26829
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4419615477382393,
+      "learning_rate": 6.297934052278309e-06,
+      "loss": 0.6695,
+      "step": 26830
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.810827808361861,
+      "learning_rate": 6.297225236484402e-06,
+      "loss": 0.7896,
+      "step": 26831
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3511538228670341,
+      "learning_rate": 6.296516442249221e-06,
+      "loss": 0.6764,
+      "step": 26832
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3520245366870325,
+      "learning_rate": 6.295807669576892e-06,
+      "loss": 0.6634,
+      "step": 26833
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.619036951336607,
+      "learning_rate": 6.295098918471543e-06,
+      "loss": 0.7365,
+      "step": 26834
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.296156183237589,
+      "learning_rate": 6.294390188937305e-06,
+      "loss": 0.6365,
+      "step": 26835
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4119169402668685,
+      "learning_rate": 6.293681480978298e-06,
+      "loss": 0.6683,
+      "step": 26836
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.432731369471279,
+      "learning_rate": 6.292972794598649e-06,
+      "loss": 0.7503,
+      "step": 26837
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.335181456973517,
+      "learning_rate": 6.292264129802492e-06,
+      "loss": 0.6412,
+      "step": 26838
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5657402959216407,
+      "learning_rate": 6.291555486593941e-06,
+      "loss": 0.671,
+      "step": 26839
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6110085231049508,
+      "learning_rate": 6.290846864977129e-06,
+      "loss": 0.4903,
+      "step": 26840
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6588778430051636,
+      "learning_rate": 6.290138264956181e-06,
+      "loss": 0.7371,
+      "step": 26841
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5944796436353803,
+      "learning_rate": 6.289429686535226e-06,
+      "loss": 0.6996,
+      "step": 26842
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5439703116392676,
+      "learning_rate": 6.2887211297183825e-06,
+      "loss": 0.5783,
+      "step": 26843
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4299428779848602,
+      "learning_rate": 6.28801259450978e-06,
+      "loss": 0.7331,
+      "step": 26844
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3930651595032866,
+      "learning_rate": 6.2873040809135465e-06,
+      "loss": 0.7725,
+      "step": 26845
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2775955710983442,
+      "learning_rate": 6.286595588933802e-06,
+      "loss": 0.7695,
+      "step": 26846
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3374544978091725,
+      "learning_rate": 6.285887118574671e-06,
+      "loss": 0.704,
+      "step": 26847
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.273687551159587,
+      "learning_rate": 6.285178669840284e-06,
+      "loss": 0.7247,
+      "step": 26848
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3345549380126587,
+      "learning_rate": 6.284470242734764e-06,
+      "loss": 0.7419,
+      "step": 26849
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2557866115937837,
+      "learning_rate": 6.28376183726223e-06,
+      "loss": 0.7359,
+      "step": 26850
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.383521781932008,
+      "learning_rate": 6.283053453426813e-06,
+      "loss": 0.6986,
+      "step": 26851
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8376919537070977,
+      "learning_rate": 6.282345091232639e-06,
+      "loss": 0.6476,
+      "step": 26852
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6305078640358193,
+      "learning_rate": 6.2816367506838265e-06,
+      "loss": 0.7237,
+      "step": 26853
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2388598147123948,
+      "learning_rate": 6.280928431784499e-06,
+      "loss": 0.5933,
+      "step": 26854
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3275877986243978,
+      "learning_rate": 6.280220134538787e-06,
+      "loss": 0.7853,
+      "step": 26855
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5899114239772392,
+      "learning_rate": 6.279511858950813e-06,
+      "loss": 0.6207,
+      "step": 26856
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2919383632381902,
+      "learning_rate": 6.278803605024695e-06,
+      "loss": 0.7475,
+      "step": 26857
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6715245887152685,
+      "learning_rate": 6.278095372764561e-06,
+      "loss": 0.5817,
+      "step": 26858
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5642732347088613,
+      "learning_rate": 6.277387162174536e-06,
+      "loss": 0.7314,
+      "step": 26859
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.583269463816076,
+      "learning_rate": 6.276678973258741e-06,
+      "loss": 0.5025,
+      "step": 26860
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6193093670583323,
+      "learning_rate": 6.2759708060212985e-06,
+      "loss": 0.753,
+      "step": 26861
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.286608040933884,
+      "learning_rate": 6.275262660466333e-06,
+      "loss": 0.6763,
+      "step": 26862
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4862430962287998,
+      "learning_rate": 6.274554536597969e-06,
+      "loss": 0.7184,
+      "step": 26863
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.350491908912509,
+      "learning_rate": 6.273846434420328e-06,
+      "loss": 0.6617,
+      "step": 26864
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3895270973036256,
+      "learning_rate": 6.2731383539375315e-06,
+      "loss": 0.7102,
+      "step": 26865
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5796618530235382,
+      "learning_rate": 6.272430295153705e-06,
+      "loss": 0.5317,
+      "step": 26866
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1903219373102325,
+      "learning_rate": 6.2717222580729665e-06,
+      "loss": 0.7161,
+      "step": 26867
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6571765322595737,
+      "learning_rate": 6.271014242699442e-06,
+      "loss": 0.7044,
+      "step": 26868
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6063399684709752,
+      "learning_rate": 6.270306249037254e-06,
+      "loss": 0.803,
+      "step": 26869
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5028712834956932,
+      "learning_rate": 6.269598277090527e-06,
+      "loss": 0.7488,
+      "step": 26870
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2367623669878707,
+      "learning_rate": 6.268890326863375e-06,
+      "loss": 0.6805,
+      "step": 26871
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2686673525512357,
+      "learning_rate": 6.268182398359927e-06,
+      "loss": 0.7398,
+      "step": 26872
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4150148013093835,
+      "learning_rate": 6.267474491584304e-06,
+      "loss": 0.7521,
+      "step": 26873
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4739057157533557,
+      "learning_rate": 6.266766606540624e-06,
+      "loss": 0.6383,
+      "step": 26874
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.508892377306785,
+      "learning_rate": 6.26605874323301e-06,
+      "loss": 0.757,
+      "step": 26875
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3715738984808477,
+      "learning_rate": 6.265350901665586e-06,
+      "loss": 0.6065,
+      "step": 26876
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3761089753876685,
+      "learning_rate": 6.264643081842472e-06,
+      "loss": 0.6802,
+      "step": 26877
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0908301177612296,
+      "learning_rate": 6.263935283767787e-06,
+      "loss": 0.7291,
+      "step": 26878
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5914991507549442,
+      "learning_rate": 6.2632275074456515e-06,
+      "loss": 0.5027,
+      "step": 26879
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2154681149964373,
+      "learning_rate": 6.262519752880193e-06,
+      "loss": 0.6627,
+      "step": 26880
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.291999210037634,
+      "learning_rate": 6.2618120200755264e-06,
+      "loss": 0.6624,
+      "step": 26881
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6213828232075276,
+      "learning_rate": 6.261104309035772e-06,
+      "loss": 0.8116,
+      "step": 26882
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2716477435942013,
+      "learning_rate": 6.260396619765051e-06,
+      "loss": 0.6161,
+      "step": 26883
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3474123407952243,
+      "learning_rate": 6.259688952267489e-06,
+      "loss": 0.8283,
+      "step": 26884
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3327198081726144,
+      "learning_rate": 6.2589813065472005e-06,
+      "loss": 0.5526,
+      "step": 26885
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2645242848380536,
+      "learning_rate": 6.258273682608304e-06,
+      "loss": 0.443,
+      "step": 26886
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6679187700411466,
+      "learning_rate": 6.257566080454925e-06,
+      "loss": 0.6252,
+      "step": 26887
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4068073365427542,
+      "learning_rate": 6.256858500091184e-06,
+      "loss": 0.6807,
+      "step": 26888
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3714008253931476,
+      "learning_rate": 6.256150941521195e-06,
+      "loss": 0.7665,
+      "step": 26889
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.509905628604715,
+      "learning_rate": 6.25544340474908e-06,
+      "loss": 0.6966,
+      "step": 26890
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2937301927349603,
+      "learning_rate": 6.2547358897789625e-06,
+      "loss": 0.789,
+      "step": 26891
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4539121525221301,
+      "learning_rate": 6.254028396614954e-06,
+      "loss": 0.7686,
+      "step": 26892
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3054961766398214,
+      "learning_rate": 6.2533209252611795e-06,
+      "loss": 0.7951,
+      "step": 26893
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2080915476549712,
+      "learning_rate": 6.252613475721756e-06,
+      "loss": 0.6569,
+      "step": 26894
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4175325554769562,
+      "learning_rate": 6.251906048000807e-06,
+      "loss": 0.5935,
+      "step": 26895
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1586702539308276,
+      "learning_rate": 6.251198642102444e-06,
+      "loss": 0.5985,
+      "step": 26896
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.346891572291453,
+      "learning_rate": 6.2504912580307905e-06,
+      "loss": 0.5342,
+      "step": 26897
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8734589377728008,
+      "learning_rate": 6.2497838957899645e-06,
+      "loss": 0.6669,
+      "step": 26898
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0261238365519993,
+      "learning_rate": 6.249076555384082e-06,
+      "loss": 0.8135,
+      "step": 26899
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5703346967351592,
+      "learning_rate": 6.248369236817265e-06,
+      "loss": 0.6827,
+      "step": 26900
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3346365081538656,
+      "learning_rate": 6.247661940093629e-06,
+      "loss": 0.633,
+      "step": 26901
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5707833687320707,
+      "learning_rate": 6.246954665217295e-06,
+      "loss": 0.7069,
+      "step": 26902
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3749451289138468,
+      "learning_rate": 6.2462474121923765e-06,
+      "loss": 0.6594,
+      "step": 26903
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0531469883183706,
+      "learning_rate": 6.245540181022996e-06,
+      "loss": 0.6969,
+      "step": 26904
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.748309139063683,
+      "learning_rate": 6.2448329717132705e-06,
+      "loss": 0.7402,
+      "step": 26905
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5231675469918022,
+      "learning_rate": 6.244125784267314e-06,
+      "loss": 0.643,
+      "step": 26906
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1580542256645399,
+      "learning_rate": 6.243418618689246e-06,
+      "loss": 0.6049,
+      "step": 26907
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2759365021409053,
+      "learning_rate": 6.242711474983183e-06,
+      "loss": 0.5972,
+      "step": 26908
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7058530554191396,
+      "learning_rate": 6.242004353153248e-06,
+      "loss": 0.7365,
+      "step": 26909
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4515925829083767,
+      "learning_rate": 6.241297253203551e-06,
+      "loss": 0.7983,
+      "step": 26910
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4767503350570976,
+      "learning_rate": 6.240590175138208e-06,
+      "loss": 0.6577,
+      "step": 26911
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.27202028657962,
+      "learning_rate": 6.239883118961344e-06,
+      "loss": 0.7549,
+      "step": 26912
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.253115986375482,
+      "learning_rate": 6.23917608467707e-06,
+      "loss": 0.6641,
+      "step": 26913
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.238962277863931,
+      "learning_rate": 6.2384690722895016e-06,
+      "loss": 0.8022,
+      "step": 26914
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4794393790651463,
+      "learning_rate": 6.237762081802757e-06,
+      "loss": 0.6718,
+      "step": 26915
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4852060972756989,
+      "learning_rate": 6.237055113220955e-06,
+      "loss": 0.6896,
+      "step": 26916
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.246286562340775,
+      "learning_rate": 6.236348166548208e-06,
+      "loss": 0.7446,
+      "step": 26917
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5841895661285842,
+      "learning_rate": 6.2356412417886335e-06,
+      "loss": 0.5156,
+      "step": 26918
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5802257216323101,
+      "learning_rate": 6.23493433894635e-06,
+      "loss": 0.6274,
+      "step": 26919
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4022469337929298,
+      "learning_rate": 6.234227458025466e-06,
+      "loss": 0.7823,
+      "step": 26920
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3546292592898368,
+      "learning_rate": 6.233520599030105e-06,
+      "loss": 0.7078,
+      "step": 26921
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.558451678374872,
+      "learning_rate": 6.232813761964377e-06,
+      "loss": 0.7225,
+      "step": 26922
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2367357781979034,
+      "learning_rate": 6.232106946832406e-06,
+      "loss": 0.7963,
+      "step": 26923
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3225750499482625,
+      "learning_rate": 6.231400153638294e-06,
+      "loss": 0.626,
+      "step": 26924
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6092221475253625,
+      "learning_rate": 6.230693382386167e-06,
+      "loss": 0.6037,
+      "step": 26925
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3738877248425654,
+      "learning_rate": 6.229986633080138e-06,
+      "loss": 0.6685,
+      "step": 26926
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8827239864053944,
+      "learning_rate": 6.229279905724319e-06,
+      "loss": 0.7739,
+      "step": 26927
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5815515361905979,
+      "learning_rate": 6.228573200322823e-06,
+      "loss": 0.6854,
+      "step": 26928
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.475258652296229,
+      "learning_rate": 6.22786651687977e-06,
+      "loss": 0.6943,
+      "step": 26929
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2708003902571197,
+      "learning_rate": 6.227159855399276e-06,
+      "loss": 0.7305,
+      "step": 26930
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5286294041383728,
+      "learning_rate": 6.226453215885448e-06,
+      "loss": 0.7203,
+      "step": 26931
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2590366473429606,
+      "learning_rate": 6.225746598342402e-06,
+      "loss": 0.754,
+      "step": 26932
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4515138321332277,
+      "learning_rate": 6.22504000277426e-06,
+      "loss": 0.6245,
+      "step": 26933
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3519391273458916,
+      "learning_rate": 6.224333429185127e-06,
+      "loss": 0.629,
+      "step": 26934
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4493851614135382,
+      "learning_rate": 6.223626877579119e-06,
+      "loss": 0.6835,
+      "step": 26935
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4757482276839113,
+      "learning_rate": 6.22292034796035e-06,
+      "loss": 0.7318,
+      "step": 26936
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6155018021550553,
+      "learning_rate": 6.22221384033294e-06,
+      "loss": 0.7417,
+      "step": 26937
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7166349401736798,
+      "learning_rate": 6.221507354700993e-06,
+      "loss": 0.7076,
+      "step": 26938
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4239602870763415,
+      "learning_rate": 6.2208008910686255e-06,
+      "loss": 0.7981,
+      "step": 26939
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.529398959870327,
+      "learning_rate": 6.220094449439957e-06,
+      "loss": 0.5187,
+      "step": 26940
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4516466540109805,
+      "learning_rate": 6.21938802981909e-06,
+      "loss": 0.7061,
+      "step": 26941
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5760043929239677,
+      "learning_rate": 6.218681632210144e-06,
+      "loss": 0.4858,
+      "step": 26942
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3513580763113826,
+      "learning_rate": 6.21797525661723e-06,
+      "loss": 0.6464,
+      "step": 26943
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.760086945597844,
+      "learning_rate": 6.217268903044467e-06,
+      "loss": 0.7958,
+      "step": 26944
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.362146819670546,
+      "learning_rate": 6.2165625714959545e-06,
+      "loss": 0.6812,
+      "step": 26945
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3386230879883982,
+      "learning_rate": 6.215856261975816e-06,
+      "loss": 0.6156,
+      "step": 26946
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.437345241338945,
+      "learning_rate": 6.215149974488161e-06,
+      "loss": 0.6146,
+      "step": 26947
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5722270227718746,
+      "learning_rate": 6.214443709037099e-06,
+      "loss": 0.7021,
+      "step": 26948
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4947705380801928,
+      "learning_rate": 6.213737465626746e-06,
+      "loss": 0.7182,
+      "step": 26949
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7422767951752791,
+      "learning_rate": 6.213031244261212e-06,
+      "loss": 0.7434,
+      "step": 26950
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6252442393482678,
+      "learning_rate": 6.212325044944609e-06,
+      "loss": 0.6485,
+      "step": 26951
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4260458225345114,
+      "learning_rate": 6.2116188676810466e-06,
+      "loss": 0.7683,
+      "step": 26952
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.0986104087884905,
+      "learning_rate": 6.2109127124746405e-06,
+      "loss": 0.621,
+      "step": 26953
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5780581049744966,
+      "learning_rate": 6.210206579329502e-06,
+      "loss": 0.5141,
+      "step": 26954
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5988202024162119,
+      "learning_rate": 6.209500468249738e-06,
+      "loss": 0.5305,
+      "step": 26955
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4963057845289423,
+      "learning_rate": 6.208794379239462e-06,
+      "loss": 0.6769,
+      "step": 26956
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3376192905319195,
+      "learning_rate": 6.208088312302785e-06,
+      "loss": 0.593,
+      "step": 26957
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3646749334174257,
+      "learning_rate": 6.207382267443823e-06,
+      "loss": 0.7214,
+      "step": 26958
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.279654252854592,
+      "learning_rate": 6.2066762446666785e-06,
+      "loss": 0.684,
+      "step": 26959
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.34527050415261,
+      "learning_rate": 6.205970243975464e-06,
+      "loss": 0.6322,
+      "step": 26960
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.02000561898886,
+      "learning_rate": 6.205264265374295e-06,
+      "loss": 0.6684,
+      "step": 26961
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3853966592356042,
+      "learning_rate": 6.2045583088672785e-06,
+      "loss": 0.7678,
+      "step": 26962
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9775747821960956,
+      "learning_rate": 6.2038523744585254e-06,
+      "loss": 0.6238,
+      "step": 26963
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4069180599838749,
+      "learning_rate": 6.203146462152142e-06,
+      "loss": 0.7442,
+      "step": 26964
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4647284831601506,
+      "learning_rate": 6.202440571952246e-06,
+      "loss": 0.7616,
+      "step": 26965
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.544838326884574,
+      "learning_rate": 6.201734703862942e-06,
+      "loss": 0.7261,
+      "step": 26966
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3918759547883837,
+      "learning_rate": 6.201028857888342e-06,
+      "loss": 0.7248,
+      "step": 26967
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3243535288254502,
+      "learning_rate": 6.200323034032556e-06,
+      "loss": 0.7061,
+      "step": 26968
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3133910644434768,
+      "learning_rate": 6.1996172322996885e-06,
+      "loss": 0.619,
+      "step": 26969
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3609859634393862,
+      "learning_rate": 6.1989114526938535e-06,
+      "loss": 0.7325,
+      "step": 26970
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4932885419627664,
+      "learning_rate": 6.198205695219157e-06,
+      "loss": 0.6563,
+      "step": 26971
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.281402934312677,
+      "learning_rate": 6.197499959879718e-06,
+      "loss": 0.7292,
+      "step": 26972
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.465532218484749,
+      "learning_rate": 6.1967942466796315e-06,
+      "loss": 0.6751,
+      "step": 26973
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2713784030361175,
+      "learning_rate": 6.196088555623016e-06,
+      "loss": 0.8689,
+      "step": 26974
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4361668592690082,
+      "learning_rate": 6.1953828867139784e-06,
+      "loss": 0.7371,
+      "step": 26975
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5066294408960115,
+      "learning_rate": 6.194677239956625e-06,
+      "loss": 0.6184,
+      "step": 26976
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4680448244112125,
+      "learning_rate": 6.193971615355062e-06,
+      "loss": 0.6396,
+      "step": 26977
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.372149871300511,
+      "learning_rate": 6.1932660129134035e-06,
+      "loss": 0.6221,
+      "step": 26978
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6033181189626193,
+      "learning_rate": 6.1925604326357595e-06,
+      "loss": 0.7571,
+      "step": 26979
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5984568077177892,
+      "learning_rate": 6.191854874526229e-06,
+      "loss": 0.6104,
+      "step": 26980
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4579983875998603,
+      "learning_rate": 6.191149338588924e-06,
+      "loss": 0.6694,
+      "step": 26981
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7030492986757049,
+      "learning_rate": 6.190443824827958e-06,
+      "loss": 0.7499,
+      "step": 26982
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4076821109556978,
+      "learning_rate": 6.189738333247433e-06,
+      "loss": 0.7106,
+      "step": 26983
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4910289183583831,
+      "learning_rate": 6.189032863851454e-06,
+      "loss": 0.682,
+      "step": 26984
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2337110737208097,
+      "learning_rate": 6.1883274166441356e-06,
+      "loss": 0.6816,
+      "step": 26985
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4593890995426317,
+      "learning_rate": 6.187621991629582e-06,
+      "loss": 0.7248,
+      "step": 26986
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.476266831152391,
+      "learning_rate": 6.1869165888119e-06,
+      "loss": 0.6529,
+      "step": 26987
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4373178777076183,
+      "learning_rate": 6.186211208195194e-06,
+      "loss": 0.5676,
+      "step": 26988
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6743379261522868,
+      "learning_rate": 6.18550584978358e-06,
+      "loss": 0.692,
+      "step": 26989
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4057163857468382,
+      "learning_rate": 6.184800513581153e-06,
+      "loss": 0.6904,
+      "step": 26990
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3645659999427708,
+      "learning_rate": 6.184095199592026e-06,
+      "loss": 0.7177,
+      "step": 26991
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4286776214762875,
+      "learning_rate": 6.183389907820303e-06,
+      "loss": 0.6686,
+      "step": 26992
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2729048374712684,
+      "learning_rate": 6.1826846382701e-06,
+      "loss": 0.6693,
+      "step": 26993
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5560511058298052,
+      "learning_rate": 6.1819793909455076e-06,
+      "loss": 0.6837,
+      "step": 26994
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4115343561936338,
+      "learning_rate": 6.181274165850644e-06,
+      "loss": 0.6249,
+      "step": 26995
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3528972841327402,
+      "learning_rate": 6.18056896298961e-06,
+      "loss": 0.6995,
+      "step": 26996
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4008650672462715,
+      "learning_rate": 6.179863782366512e-06,
+      "loss": 0.7111,
+      "step": 26997
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3360178806214502,
+      "learning_rate": 6.179158623985457e-06,
+      "loss": 0.6871,
+      "step": 26998
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5807668146413885,
+      "learning_rate": 6.178453487850551e-06,
+      "loss": 0.7002,
+      "step": 26999
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5927157074185326,
+      "learning_rate": 6.1777483739659e-06,
+      "loss": 0.5017,
+      "step": 27000
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4008105671411193,
+      "learning_rate": 6.177043282335603e-06,
+      "loss": 0.7366,
+      "step": 27001
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6017814653093734,
+      "learning_rate": 6.1763382129637725e-06,
+      "loss": 0.6698,
+      "step": 27002
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3990216275301064,
+      "learning_rate": 6.175633165854515e-06,
+      "loss": 0.7137,
+      "step": 27003
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3991515574916638,
+      "learning_rate": 6.17492814101193e-06,
+      "loss": 0.5666,
+      "step": 27004
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1868801703661167,
+      "learning_rate": 6.174223138440121e-06,
+      "loss": 0.7545,
+      "step": 27005
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3220852271384655,
+      "learning_rate": 6.173518158143199e-06,
+      "loss": 0.6871,
+      "step": 27006
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5261458777623298,
+      "learning_rate": 6.172813200125267e-06,
+      "loss": 0.6431,
+      "step": 27007
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2612888586052107,
+      "learning_rate": 6.172108264390428e-06,
+      "loss": 0.6722,
+      "step": 27008
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4243893900122184,
+      "learning_rate": 6.171403350942783e-06,
+      "loss": 0.7002,
+      "step": 27009
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2817836770692568,
+      "learning_rate": 6.170698459786442e-06,
+      "loss": 0.7687,
+      "step": 27010
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7711739146806018,
+      "learning_rate": 6.169993590925509e-06,
+      "loss": 0.7978,
+      "step": 27011
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.497003555821554,
+      "learning_rate": 6.169288744364084e-06,
+      "loss": 0.6354,
+      "step": 27012
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5075091555395193,
+      "learning_rate": 6.16858392010627e-06,
+      "loss": 0.7252,
+      "step": 27013
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4303878330051087,
+      "learning_rate": 6.167879118156179e-06,
+      "loss": 0.6071,
+      "step": 27014
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3767707530193607,
+      "learning_rate": 6.167174338517905e-06,
+      "loss": 0.7444,
+      "step": 27015
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.436484223399207,
+      "learning_rate": 6.166469581195558e-06,
+      "loss": 0.6634,
+      "step": 27016
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4675483746796587,
+      "learning_rate": 6.165764846193234e-06,
+      "loss": 0.7093,
+      "step": 27017
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3787654041229431,
+      "learning_rate": 6.165060133515046e-06,
+      "loss": 0.7191,
+      "step": 27018
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5308528062536724,
+      "learning_rate": 6.164355443165091e-06,
+      "loss": 0.7677,
+      "step": 27019
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3370820405198118,
+      "learning_rate": 6.163650775147471e-06,
+      "loss": 0.7988,
+      "step": 27020
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8612554471623917,
+      "learning_rate": 6.162946129466295e-06,
+      "loss": 0.7129,
+      "step": 27021
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1853462772174752,
+      "learning_rate": 6.162241506125657e-06,
+      "loss": 0.6505,
+      "step": 27022
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.477768410345997,
+      "learning_rate": 6.161536905129666e-06,
+      "loss": 0.6228,
+      "step": 27023
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.361912852863244,
+      "learning_rate": 6.160832326482418e-06,
+      "loss": 0.7988,
+      "step": 27024
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4014322979767586,
+      "learning_rate": 6.160127770188029e-06,
+      "loss": 0.6811,
+      "step": 27025
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5706791754583008,
+      "learning_rate": 6.1594232362505835e-06,
+      "loss": 0.4789,
+      "step": 27026
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3958912822713938,
+      "learning_rate": 6.158718724674194e-06,
+      "loss": 0.7847,
+      "step": 27027
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.341385986336701,
+      "learning_rate": 6.158014235462964e-06,
+      "loss": 0.5246,
+      "step": 27028
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3366692594848806,
+      "learning_rate": 6.157309768620987e-06,
+      "loss": 0.7877,
+      "step": 27029
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.063544924143433,
+      "learning_rate": 6.156605324152369e-06,
+      "loss": 0.7195,
+      "step": 27030
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5263969587382227,
+      "learning_rate": 6.155900902061212e-06,
+      "loss": 0.7441,
+      "step": 27031
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.443682189182981,
+      "learning_rate": 6.155196502351621e-06,
+      "loss": 0.7115,
+      "step": 27032
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.077098692150945,
+      "learning_rate": 6.154492125027688e-06,
+      "loss": 0.7592,
+      "step": 27033
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2931416627895698,
+      "learning_rate": 6.1537877700935215e-06,
+      "loss": 0.7499,
+      "step": 27034
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.996580951742382,
+      "learning_rate": 6.153083437553222e-06,
+      "loss": 0.6568,
+      "step": 27035
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4864162344133662,
+      "learning_rate": 6.152379127410887e-06,
+      "loss": 0.8117,
+      "step": 27036
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4486416269602589,
+      "learning_rate": 6.151674839670616e-06,
+      "loss": 0.6668,
+      "step": 27037
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.526945843624188,
+      "learning_rate": 6.1509705743365154e-06,
+      "loss": 0.7805,
+      "step": 27038
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4466984058357617,
+      "learning_rate": 6.150266331412684e-06,
+      "loss": 0.5864,
+      "step": 27039
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5918761562901687,
+      "learning_rate": 6.14956211090322e-06,
+      "loss": 0.6782,
+      "step": 27040
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3903135036994867,
+      "learning_rate": 6.148857912812222e-06,
+      "loss": 0.7136,
+      "step": 27041
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.476774778302609,
+      "learning_rate": 6.148153737143798e-06,
+      "loss": 0.7327,
+      "step": 27042
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8236642881051648,
+      "learning_rate": 6.1474495839020365e-06,
+      "loss": 0.6556,
+      "step": 27043
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.37140528613838,
+      "learning_rate": 6.1467454530910454e-06,
+      "loss": 0.7456,
+      "step": 27044
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.592901567988693,
+      "learning_rate": 6.146041344714919e-06,
+      "loss": 0.6581,
+      "step": 27045
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3450199987359033,
+      "learning_rate": 6.145337258777766e-06,
+      "loss": 0.7218,
+      "step": 27046
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4739539540550162,
+      "learning_rate": 6.144633195283677e-06,
+      "loss": 0.8243,
+      "step": 27047
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6690335940171872,
+      "learning_rate": 6.1439291542367544e-06,
+      "loss": 0.6172,
+      "step": 27048
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2653347548080998,
+      "learning_rate": 6.143225135641097e-06,
+      "loss": 0.5621,
+      "step": 27049
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.425051430743048,
+      "learning_rate": 6.142521139500803e-06,
+      "loss": 0.5843,
+      "step": 27050
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4169848611829257,
+      "learning_rate": 6.1418171658199725e-06,
+      "loss": 0.6601,
+      "step": 27051
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2997717075428026,
+      "learning_rate": 6.141113214602705e-06,
+      "loss": 0.6788,
+      "step": 27052
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2470636037796476,
+      "learning_rate": 6.1404092858531e-06,
+      "loss": 0.6657,
+      "step": 27053
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2949931133906798,
+      "learning_rate": 6.139705379575248e-06,
+      "loss": 0.7017,
+      "step": 27054
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5870116622878845,
+      "learning_rate": 6.139001495773258e-06,
+      "loss": 0.8837,
+      "step": 27055
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4564156898385094,
+      "learning_rate": 6.138297634451224e-06,
+      "loss": 0.5448,
+      "step": 27056
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3215288983038305,
+      "learning_rate": 6.137593795613242e-06,
+      "loss": 0.7057,
+      "step": 27057
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6337118040221283,
+      "learning_rate": 6.136889979263409e-06,
+      "loss": 0.646,
+      "step": 27058
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.217779723802113,
+      "learning_rate": 6.13618618540583e-06,
+      "loss": 0.6413,
+      "step": 27059
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3655906799367503,
+      "learning_rate": 6.135482414044598e-06,
+      "loss": 0.7724,
+      "step": 27060
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4024622059286678,
+      "learning_rate": 6.1347786651838105e-06,
+      "loss": 0.7927,
+      "step": 27061
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.610788101355938,
+      "learning_rate": 6.134074938827562e-06,
+      "loss": 0.5385,
+      "step": 27062
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3219550122149197,
+      "learning_rate": 6.133371234979958e-06,
+      "loss": 0.7628,
+      "step": 27063
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.387756144638154,
+      "learning_rate": 6.1326675536450885e-06,
+      "loss": 0.774,
+      "step": 27064
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5024557689442668,
+      "learning_rate": 6.1319638948270536e-06,
+      "loss": 0.7232,
+      "step": 27065
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.296261474192999,
+      "learning_rate": 6.131260258529947e-06,
+      "loss": 0.8003,
+      "step": 27066
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2894097261512798,
+      "learning_rate": 6.130556644757871e-06,
+      "loss": 0.6651,
+      "step": 27067
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4283795823384966,
+      "learning_rate": 6.129853053514918e-06,
+      "loss": 0.8608,
+      "step": 27068
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2041299402789931,
+      "learning_rate": 6.129149484805183e-06,
+      "loss": 0.7396,
+      "step": 27069
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6067478885154605,
+      "learning_rate": 6.1284459386327735e-06,
+      "loss": 0.5147,
+      "step": 27070
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3884679488273939,
+      "learning_rate": 6.1277424150017695e-06,
+      "loss": 0.6827,
+      "step": 27071
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4226643073819978,
+      "learning_rate": 6.127038913916278e-06,
+      "loss": 0.7658,
+      "step": 27072
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3545098854838586,
+      "learning_rate": 6.12633543538039e-06,
+      "loss": 0.9334,
+      "step": 27073
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5623853281187359,
+      "learning_rate": 6.125631979398209e-06,
+      "loss": 0.4991,
+      "step": 27074
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3384051557389731,
+      "learning_rate": 6.124928545973819e-06,
+      "loss": 0.6658,
+      "step": 27075
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5416821201044038,
+      "learning_rate": 6.124225135111325e-06,
+      "loss": 0.7031,
+      "step": 27076
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1982331946701237,
+      "learning_rate": 6.12352174681482e-06,
+      "loss": 0.5653,
+      "step": 27077
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4818760757227005,
+      "learning_rate": 6.122818381088398e-06,
+      "loss": 0.7632,
+      "step": 27078
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.781562828434168,
+      "learning_rate": 6.1221150379361514e-06,
+      "loss": 0.6788,
+      "step": 27079
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2947481167241468,
+      "learning_rate": 6.1214117173621815e-06,
+      "loss": 0.5802,
+      "step": 27080
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.157234296761543,
+      "learning_rate": 6.120708419370583e-06,
+      "loss": 0.7926,
+      "step": 27081
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3982194113577335,
+      "learning_rate": 6.120005143965443e-06,
+      "loss": 0.7332,
+      "step": 27082
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3962568136018971,
+      "learning_rate": 6.1193018911508635e-06,
+      "loss": 0.5625,
+      "step": 27083
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8598926286461333,
+      "learning_rate": 6.118598660930941e-06,
+      "loss": 0.7013,
+      "step": 27084
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3228462585969367,
+      "learning_rate": 6.117895453309762e-06,
+      "loss": 0.7583,
+      "step": 27085
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.506202515608917,
+      "learning_rate": 6.117192268291423e-06,
+      "loss": 0.6494,
+      "step": 27086
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2997292575365984,
+      "learning_rate": 6.116489105880023e-06,
+      "loss": 0.5828,
+      "step": 27087
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.474359682644424,
+      "learning_rate": 6.115785966079653e-06,
+      "loss": 0.8527,
+      "step": 27088
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6191462572842652,
+      "learning_rate": 6.115082848894407e-06,
+      "loss": 0.5295,
+      "step": 27089
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5758178095390514,
+      "learning_rate": 6.114379754328375e-06,
+      "loss": 0.6432,
+      "step": 27090
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3478224159506031,
+      "learning_rate": 6.113676682385661e-06,
+      "loss": 0.605,
+      "step": 27091
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6055552226058577,
+      "learning_rate": 6.112973633070346e-06,
+      "loss": 0.5086,
+      "step": 27092
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.335320914678341,
+      "learning_rate": 6.11227060638653e-06,
+      "loss": 0.6989,
+      "step": 27093
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6344563100339432,
+      "learning_rate": 6.111567602338304e-06,
+      "loss": 0.5149,
+      "step": 27094
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4304304486635468,
+      "learning_rate": 6.110864620929766e-06,
+      "loss": 0.6805,
+      "step": 27095
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5758008032913735,
+      "learning_rate": 6.110161662165005e-06,
+      "loss": 0.7221,
+      "step": 27096
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5579636856748762,
+      "learning_rate": 6.109458726048113e-06,
+      "loss": 0.6771,
+      "step": 27097
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2768932389652612,
+      "learning_rate": 6.108755812583187e-06,
+      "loss": 0.6887,
+      "step": 27098
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1508469626483184,
+      "learning_rate": 6.108052921774312e-06,
+      "loss": 0.5832,
+      "step": 27099
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2367496656132717,
+      "learning_rate": 6.107350053625587e-06,
+      "loss": 0.7148,
+      "step": 27100
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5945598313252898,
+      "learning_rate": 6.106647208141101e-06,
+      "loss": 0.7864,
+      "step": 27101
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6790633610183952,
+      "learning_rate": 6.105944385324951e-06,
+      "loss": 0.7111,
+      "step": 27102
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6426236910646725,
+      "learning_rate": 6.105241585181222e-06,
+      "loss": 0.7798,
+      "step": 27103
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4192450060685426,
+      "learning_rate": 6.104538807714011e-06,
+      "loss": 0.7711,
+      "step": 27104
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4485604683075843,
+      "learning_rate": 6.10383605292741e-06,
+      "loss": 0.6762,
+      "step": 27105
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4204093513514355,
+      "learning_rate": 6.103133320825507e-06,
+      "loss": 0.8259,
+      "step": 27106
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.460886078873224,
+      "learning_rate": 6.102430611412395e-06,
+      "loss": 0.7517,
+      "step": 27107
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2996850476574093,
+      "learning_rate": 6.101727924692167e-06,
+      "loss": 0.5869,
+      "step": 27108
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2389259187624266,
+      "learning_rate": 6.101025260668916e-06,
+      "loss": 0.6991,
+      "step": 27109
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.36674979456997,
+      "learning_rate": 6.100322619346726e-06,
+      "loss": 0.7271,
+      "step": 27110
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4485571531672907,
+      "learning_rate": 6.099620000729692e-06,
+      "loss": 0.723,
+      "step": 27111
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5771590900048235,
+      "learning_rate": 6.098917404821909e-06,
+      "loss": 0.5375,
+      "step": 27112
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1530349870539895,
+      "learning_rate": 6.098214831627462e-06,
+      "loss": 0.7505,
+      "step": 27113
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3621694764658803,
+      "learning_rate": 6.097512281150444e-06,
+      "loss": 0.7728,
+      "step": 27114
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4102183369218,
+      "learning_rate": 6.096809753394941e-06,
+      "loss": 0.6552,
+      "step": 27115
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8729566180536246,
+      "learning_rate": 6.0961072483650526e-06,
+      "loss": 0.6715,
+      "step": 27116
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.584205925613229,
+      "learning_rate": 6.095404766064863e-06,
+      "loss": 0.5331,
+      "step": 27117
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3385952440035198,
+      "learning_rate": 6.094702306498459e-06,
+      "loss": 0.7569,
+      "step": 27118
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4262767433529722,
+      "learning_rate": 6.093999869669942e-06,
+      "loss": 0.6877,
+      "step": 27119
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2567284483269825,
+      "learning_rate": 6.0932974555833875e-06,
+      "loss": 0.764,
+      "step": 27120
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2129268001727265,
+      "learning_rate": 6.092595064242894e-06,
+      "loss": 0.6698,
+      "step": 27121
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2415028269155142,
+      "learning_rate": 6.091892695652548e-06,
+      "loss": 0.703,
+      "step": 27122
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3393009915391452,
+      "learning_rate": 6.091190349816446e-06,
+      "loss": 0.6569,
+      "step": 27123
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3841786189696352,
+      "learning_rate": 6.090488026738666e-06,
+      "loss": 0.7569,
+      "step": 27124
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3978512266223313,
+      "learning_rate": 6.089785726423303e-06,
+      "loss": 0.7035,
+      "step": 27125
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7327653973344646,
+      "learning_rate": 6.089083448874446e-06,
+      "loss": 0.6793,
+      "step": 27126
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.236475012849336,
+      "learning_rate": 6.088381194096184e-06,
+      "loss": 0.7252,
+      "step": 27127
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.620885231281148,
+      "learning_rate": 6.087678962092603e-06,
+      "loss": 0.4625,
+      "step": 27128
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2438873975990907,
+      "learning_rate": 6.0869767528677944e-06,
+      "loss": 0.6532,
+      "step": 27129
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2904281496451198,
+      "learning_rate": 6.086274566425848e-06,
+      "loss": 0.6759,
+      "step": 27130
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4454918395516327,
+      "learning_rate": 6.085572402770848e-06,
+      "loss": 0.7845,
+      "step": 27131
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0075481707823597,
+      "learning_rate": 6.084870261906883e-06,
+      "loss": 0.6896,
+      "step": 27132
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2590918123938524,
+      "learning_rate": 6.084168143838046e-06,
+      "loss": 0.8814,
+      "step": 27133
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3472007711587521,
+      "learning_rate": 6.083466048568423e-06,
+      "loss": 0.7564,
+      "step": 27134
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6819106934888866,
+      "learning_rate": 6.082763976102097e-06,
+      "loss": 0.6964,
+      "step": 27135
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7384548620089275,
+      "learning_rate": 6.082061926443162e-06,
+      "loss": 0.7352,
+      "step": 27136
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1804630867941868,
+      "learning_rate": 6.081359899595702e-06,
+      "loss": 0.7244,
+      "step": 27137
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.487731764191838,
+      "learning_rate": 6.080657895563807e-06,
+      "loss": 0.7635,
+      "step": 27138
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3386356990669557,
+      "learning_rate": 6.079955914351558e-06,
+      "loss": 0.7422,
+      "step": 27139
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5179441877431379,
+      "learning_rate": 6.079253955963049e-06,
+      "loss": 0.6929,
+      "step": 27140
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5114869183146358,
+      "learning_rate": 6.078552020402367e-06,
+      "loss": 0.7523,
+      "step": 27141
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5530355662611878,
+      "learning_rate": 6.0778501076735956e-06,
+      "loss": 0.4968,
+      "step": 27142
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3185191457570642,
+      "learning_rate": 6.077148217780818e-06,
+      "loss": 0.6187,
+      "step": 27143
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3575566046132852,
+      "learning_rate": 6.076446350728131e-06,
+      "loss": 0.7392,
+      "step": 27144
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4149085186452897,
+      "learning_rate": 6.0757445065196145e-06,
+      "loss": 0.8005,
+      "step": 27145
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5825981853477566,
+      "learning_rate": 6.075042685159356e-06,
+      "loss": 0.4835,
+      "step": 27146
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4743297245419382,
+      "learning_rate": 6.074340886651438e-06,
+      "loss": 0.5873,
+      "step": 27147
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.354860598603777,
+      "learning_rate": 6.073639110999955e-06,
+      "loss": 0.7207,
+      "step": 27148
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3465133373246345,
+      "learning_rate": 6.072937358208986e-06,
+      "loss": 0.6829,
+      "step": 27149
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6589161277533475,
+      "learning_rate": 6.072235628282619e-06,
+      "loss": 0.575,
+      "step": 27150
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.337072415250522,
+      "learning_rate": 6.071533921224941e-06,
+      "loss": 0.7262,
+      "step": 27151
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3938438006021467,
+      "learning_rate": 6.0708322370400345e-06,
+      "loss": 0.5576,
+      "step": 27152
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5955692084084778,
+      "learning_rate": 6.0701305757319866e-06,
+      "loss": 0.7073,
+      "step": 27153
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.445645179551321,
+      "learning_rate": 6.069428937304883e-06,
+      "loss": 0.6766,
+      "step": 27154
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3358965912637537,
+      "learning_rate": 6.06872732176281e-06,
+      "loss": 0.7033,
+      "step": 27155
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.662435816957485,
+      "learning_rate": 6.068025729109849e-06,
+      "loss": 0.6578,
+      "step": 27156
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5275826520945814,
+      "learning_rate": 6.067324159350088e-06,
+      "loss": 0.6276,
+      "step": 27157
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4268380880496605,
+      "learning_rate": 6.066622612487611e-06,
+      "loss": 0.6739,
+      "step": 27158
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.386555875664689,
+      "learning_rate": 6.065921088526503e-06,
+      "loss": 0.5783,
+      "step": 27159
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6541068666351426,
+      "learning_rate": 6.065219587470845e-06,
+      "loss": 0.4923,
+      "step": 27160
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5028774376210279,
+      "learning_rate": 6.064518109324727e-06,
+      "loss": 0.5641,
+      "step": 27161
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5357847343623887,
+      "learning_rate": 6.063816654092232e-06,
+      "loss": 0.653,
+      "step": 27162
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3384275927694387,
+      "learning_rate": 6.06311522177744e-06,
+      "loss": 0.7255,
+      "step": 27163
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.358212347826197,
+      "learning_rate": 6.062413812384436e-06,
+      "loss": 0.8263,
+      "step": 27164
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5044336001170577,
+      "learning_rate": 6.06171242591731e-06,
+      "loss": 0.7487,
+      "step": 27165
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6190397894680582,
+      "learning_rate": 6.061011062380139e-06,
+      "loss": 0.681,
+      "step": 27166
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.382347349495827,
+      "learning_rate": 6.060309721777006e-06,
+      "loss": 0.7382,
+      "step": 27167
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6365349164042878,
+      "learning_rate": 6.059608404111998e-06,
+      "loss": 0.5315,
+      "step": 27168
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4488236134717734,
+      "learning_rate": 6.058907109389202e-06,
+      "loss": 0.7898,
+      "step": 27169
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4311795392343483,
+      "learning_rate": 6.058205837612694e-06,
+      "loss": 0.6025,
+      "step": 27170
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3200426596788537,
+      "learning_rate": 6.057504588786556e-06,
+      "loss": 0.6701,
+      "step": 27171
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3029939244681346,
+      "learning_rate": 6.0568033629148814e-06,
+      "loss": 0.7243,
+      "step": 27172
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3457507341896742,
+      "learning_rate": 6.05610216000174e-06,
+      "loss": 0.7593,
+      "step": 27173
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4445410626057622,
+      "learning_rate": 6.055400980051223e-06,
+      "loss": 0.7102,
+      "step": 27174
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3154716025707087,
+      "learning_rate": 6.054699823067407e-06,
+      "loss": 0.6874,
+      "step": 27175
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3291900168785546,
+      "learning_rate": 6.053998689054384e-06,
+      "loss": 0.6768,
+      "step": 27176
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.512341756505623,
+      "learning_rate": 6.053297578016224e-06,
+      "loss": 0.5506,
+      "step": 27177
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5959314165189853,
+      "learning_rate": 6.052596489957017e-06,
+      "loss": 0.6589,
+      "step": 27178
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3581509129705622,
+      "learning_rate": 6.051895424880847e-06,
+      "loss": 0.7469,
+      "step": 27179
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4343553011148058,
+      "learning_rate": 6.051194382791785e-06,
+      "loss": 0.5202,
+      "step": 27180
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.500726654893941,
+      "learning_rate": 6.050493363693924e-06,
+      "loss": 0.5205,
+      "step": 27181
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4678350204369357,
+      "learning_rate": 6.049792367591338e-06,
+      "loss": 0.6718,
+      "step": 27182
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3711598683604114,
+      "learning_rate": 6.049091394488116e-06,
+      "loss": 0.6899,
+      "step": 27183
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6597522642405316,
+      "learning_rate": 6.048390444388329e-06,
+      "loss": 0.5527,
+      "step": 27184
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4541770237450267,
+      "learning_rate": 6.047689517296067e-06,
+      "loss": 0.6359,
+      "step": 27185
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6670364883691047,
+      "learning_rate": 6.04698861321541e-06,
+      "loss": 0.6907,
+      "step": 27186
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.389477978997974,
+      "learning_rate": 6.046287732150433e-06,
+      "loss": 0.6927,
+      "step": 27187
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4258611820890363,
+      "learning_rate": 6.045586874105221e-06,
+      "loss": 0.6473,
+      "step": 27188
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6381970316197891,
+      "learning_rate": 6.044886039083855e-06,
+      "loss": 0.8243,
+      "step": 27189
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6153899034424665,
+      "learning_rate": 6.044185227090415e-06,
+      "loss": 0.5354,
+      "step": 27190
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4765770012323054,
+      "learning_rate": 6.0434844381289814e-06,
+      "loss": 0.9189,
+      "step": 27191
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5103340315416829,
+      "learning_rate": 6.042783672203632e-06,
+      "loss": 0.6674,
+      "step": 27192
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4720221859381295,
+      "learning_rate": 6.042082929318452e-06,
+      "loss": 0.6839,
+      "step": 27193
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6011514587987024,
+      "learning_rate": 6.041382209477516e-06,
+      "loss": 0.7435,
+      "step": 27194
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4822404989853324,
+      "learning_rate": 6.040681512684907e-06,
+      "loss": 0.7146,
+      "step": 27195
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4315742972566237,
+      "learning_rate": 6.039980838944701e-06,
+      "loss": 0.6191,
+      "step": 27196
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3973849034696795,
+      "learning_rate": 6.039280188260985e-06,
+      "loss": 0.7113,
+      "step": 27197
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2390594392460355,
+      "learning_rate": 6.0385795606378315e-06,
+      "loss": 0.705,
+      "step": 27198
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3427114301908019,
+      "learning_rate": 6.037878956079322e-06,
+      "loss": 0.7914,
+      "step": 27199
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6310176322153984,
+      "learning_rate": 6.037178374589536e-06,
+      "loss": 0.7206,
+      "step": 27200
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3820727228017131,
+      "learning_rate": 6.0364778161725505e-06,
+      "loss": 0.6355,
+      "step": 27201
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.585205614330327,
+      "learning_rate": 6.035777280832447e-06,
+      "loss": 0.5225,
+      "step": 27202
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.237696298867593,
+      "learning_rate": 6.035076768573303e-06,
+      "loss": 0.753,
+      "step": 27203
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.435401739821131,
+      "learning_rate": 6.034376279399199e-06,
+      "loss": 0.6328,
+      "step": 27204
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8035467411662935,
+      "learning_rate": 6.033675813314208e-06,
+      "loss": 0.7474,
+      "step": 27205
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3127109308014544,
+      "learning_rate": 6.032975370322414e-06,
+      "loss": 0.6826,
+      "step": 27206
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8416548863170308,
+      "learning_rate": 6.032274950427896e-06,
+      "loss": 0.7396,
+      "step": 27207
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3429199051086336,
+      "learning_rate": 6.031574553634727e-06,
+      "loss": 0.604,
+      "step": 27208
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.448897624762624,
+      "learning_rate": 6.030874179946986e-06,
+      "loss": 0.817,
+      "step": 27209
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.56267648269261,
+      "learning_rate": 6.0301738293687525e-06,
+      "loss": 0.7077,
+      "step": 27210
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3281432007452334,
+      "learning_rate": 6.029473501904108e-06,
+      "loss": 0.6483,
+      "step": 27211
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5988453803830498,
+      "learning_rate": 6.028773197557121e-06,
+      "loss": 0.5287,
+      "step": 27212
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3847467343076079,
+      "learning_rate": 6.028072916331874e-06,
+      "loss": 0.7137,
+      "step": 27213
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6769917183801273,
+      "learning_rate": 6.027372658232448e-06,
+      "loss": 0.7038,
+      "step": 27214
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5778257562289641,
+      "learning_rate": 6.026672423262914e-06,
+      "loss": 0.6632,
+      "step": 27215
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4387010346512912,
+      "learning_rate": 6.025972211427346e-06,
+      "loss": 0.6201,
+      "step": 27216
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5658958908629477,
+      "learning_rate": 6.025272022729831e-06,
+      "loss": 0.6795,
+      "step": 27217
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.558689447958537,
+      "learning_rate": 6.024571857174443e-06,
+      "loss": 0.695,
+      "step": 27218
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3777952190804403,
+      "learning_rate": 6.0238717147652525e-06,
+      "loss": 0.7229,
+      "step": 27219
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2465413948059019,
+      "learning_rate": 6.023171595506338e-06,
+      "loss": 0.7486,
+      "step": 27220
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5490182785827695,
+      "learning_rate": 6.022471499401785e-06,
+      "loss": 0.6463,
+      "step": 27221
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.626041665855543,
+      "learning_rate": 6.021771426455656e-06,
+      "loss": 0.7494,
+      "step": 27222
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.803186315302757,
+      "learning_rate": 6.0210713766720344e-06,
+      "loss": 0.6595,
+      "step": 27223
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4888918249394048,
+      "learning_rate": 6.020371350054992e-06,
+      "loss": 0.742,
+      "step": 27224
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3688834516250314,
+      "learning_rate": 6.019671346608615e-06,
+      "loss": 0.7444,
+      "step": 27225
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3932343094981872,
+      "learning_rate": 6.018971366336965e-06,
+      "loss": 0.5803,
+      "step": 27226
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3922911126639888,
+      "learning_rate": 6.018271409244127e-06,
+      "loss": 0.6124,
+      "step": 27227
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6560736470056163,
+      "learning_rate": 6.017571475334174e-06,
+      "loss": 0.7479,
+      "step": 27228
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7068416739760106,
+      "learning_rate": 6.016871564611177e-06,
+      "loss": 0.6186,
+      "step": 27229
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5289723312578387,
+      "learning_rate": 6.016171677079218e-06,
+      "loss": 0.7502,
+      "step": 27230
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.497252357112121,
+      "learning_rate": 6.0154718127423685e-06,
+      "loss": 0.6362,
+      "step": 27231
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8114332505998665,
+      "learning_rate": 6.014771971604705e-06,
+      "loss": 0.763,
+      "step": 27232
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3946070501156205,
+      "learning_rate": 6.014072153670296e-06,
+      "loss": 0.8083,
+      "step": 27233
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2845300564753788,
+      "learning_rate": 6.0133723589432235e-06,
+      "loss": 0.7026,
+      "step": 27234
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3525151737802714,
+      "learning_rate": 6.012672587427562e-06,
+      "loss": 0.7119,
+      "step": 27235
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2912684947515065,
+      "learning_rate": 6.01197283912738e-06,
+      "loss": 0.6471,
+      "step": 27236
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.790902926238996,
+      "learning_rate": 6.011273114046754e-06,
+      "loss": 0.7523,
+      "step": 27237
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5226841313395538,
+      "learning_rate": 6.010573412189759e-06,
+      "loss": 0.8313,
+      "step": 27238
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3161378906716317,
+      "learning_rate": 6.009873733560472e-06,
+      "loss": 0.7431,
+      "step": 27239
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4263882708296372,
+      "learning_rate": 6.00917407816296e-06,
+      "loss": 0.6613,
+      "step": 27240
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4280213845802903,
+      "learning_rate": 6.0084744460013e-06,
+      "loss": 0.8058,
+      "step": 27241
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8030945404482879,
+      "learning_rate": 6.007774837079567e-06,
+      "loss": 0.7394,
+      "step": 27242
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3580919735924128,
+      "learning_rate": 6.007075251401833e-06,
+      "loss": 0.7068,
+      "step": 27243
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5324547146772474,
+      "learning_rate": 6.00637568897217e-06,
+      "loss": 0.6987,
+      "step": 27244
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.427160333344121,
+      "learning_rate": 6.0056761497946505e-06,
+      "loss": 0.683,
+      "step": 27245
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4388303361643848,
+      "learning_rate": 6.004976633873354e-06,
+      "loss": 0.888,
+      "step": 27246
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.106869988777094,
+      "learning_rate": 6.004277141212345e-06,
+      "loss": 0.6797,
+      "step": 27247
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2621197380823077,
+      "learning_rate": 6.003577671815699e-06,
+      "loss": 0.6462,
+      "step": 27248
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5962305524387852,
+      "learning_rate": 6.002878225687492e-06,
+      "loss": 0.4545,
+      "step": 27249
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6456420981424447,
+      "learning_rate": 6.002178802831789e-06,
+      "loss": 0.6946,
+      "step": 27250
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5731803489657493,
+      "learning_rate": 6.001479403252669e-06,
+      "loss": 0.5059,
+      "step": 27251
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.40244494995063,
+      "learning_rate": 6.000780026954199e-06,
+      "loss": 0.5657,
+      "step": 27252
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3866162810026166,
+      "learning_rate": 6.000080673940461e-06,
+      "loss": 0.6721,
+      "step": 27253
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8153296701268076,
+      "learning_rate": 5.999381344215514e-06,
+      "loss": 0.8359,
+      "step": 27254
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6823055022985354,
+      "learning_rate": 5.998682037783437e-06,
+      "loss": 0.72,
+      "step": 27255
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4539157029492513,
+      "learning_rate": 5.997982754648297e-06,
+      "loss": 0.7441,
+      "step": 27256
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2493316070155125,
+      "learning_rate": 5.997283494814175e-06,
+      "loss": 0.7814,
+      "step": 27257
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.347423785549244,
+      "learning_rate": 5.99658425828513e-06,
+      "loss": 0.7171,
+      "step": 27258
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5575130127543938,
+      "learning_rate": 5.99588504506524e-06,
+      "loss": 0.4841,
+      "step": 27259
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3682600047933513,
+      "learning_rate": 5.995185855158577e-06,
+      "loss": 0.8147,
+      "step": 27260
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3978109374624976,
+      "learning_rate": 5.994486688569209e-06,
+      "loss": 0.6939,
+      "step": 27261
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5973805413900716,
+      "learning_rate": 5.993787545301204e-06,
+      "loss": 0.513,
+      "step": 27262
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5624161851464327,
+      "learning_rate": 5.99308842535864e-06,
+      "loss": 0.8191,
+      "step": 27263
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.078323285446034,
+      "learning_rate": 5.9923893287455845e-06,
+      "loss": 0.6709,
+      "step": 27264
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5997510199081564,
+      "learning_rate": 5.991690255466105e-06,
+      "loss": 0.4975,
+      "step": 27265
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5570217770656258,
+      "learning_rate": 5.990991205524274e-06,
+      "loss": 0.7238,
+      "step": 27266
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4491550179990507,
+      "learning_rate": 5.990292178924164e-06,
+      "loss": 0.7694,
+      "step": 27267
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5523337717218084,
+      "learning_rate": 5.98959317566984e-06,
+      "loss": 0.7227,
+      "step": 27268
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3381904649573748,
+      "learning_rate": 5.988894195765372e-06,
+      "loss": 0.6661,
+      "step": 27269
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3372298708217942,
+      "learning_rate": 5.988195239214835e-06,
+      "loss": 0.7486,
+      "step": 27270
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3540156635429288,
+      "learning_rate": 5.987496306022297e-06,
+      "loss": 0.5673,
+      "step": 27271
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7051366114898603,
+      "learning_rate": 5.986797396191824e-06,
+      "loss": 0.627,
+      "step": 27272
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.285688984771002,
+      "learning_rate": 5.986098509727485e-06,
+      "loss": 0.6309,
+      "step": 27273
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5937715357577412,
+      "learning_rate": 5.985399646633356e-06,
+      "loss": 0.5589,
+      "step": 27274
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3613555688675179,
+      "learning_rate": 5.984700806913497e-06,
+      "loss": 0.6841,
+      "step": 27275
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5065560962640314,
+      "learning_rate": 5.9840019905719815e-06,
+      "loss": 0.7286,
+      "step": 27276
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7209181258943975,
+      "learning_rate": 5.983303197612878e-06,
+      "loss": 0.7271,
+      "step": 27277
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2525471632209,
+      "learning_rate": 5.9826044280402574e-06,
+      "loss": 0.6608,
+      "step": 27278
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.502863415046381,
+      "learning_rate": 5.981905681858184e-06,
+      "loss": 0.7418,
+      "step": 27279
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7830486665506609,
+      "learning_rate": 5.981206959070728e-06,
+      "loss": 0.7318,
+      "step": 27280
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.364905406803133,
+      "learning_rate": 5.980508259681958e-06,
+      "loss": 0.6342,
+      "step": 27281
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2691420039542674,
+      "learning_rate": 5.9798095836959405e-06,
+      "loss": 0.7087,
+      "step": 27282
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.318406473318531,
+      "learning_rate": 5.979110931116744e-06,
+      "loss": 0.7202,
+      "step": 27283
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3190779746917525,
+      "learning_rate": 5.978412301948438e-06,
+      "loss": 0.7332,
+      "step": 27284
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5642981758338743,
+      "learning_rate": 5.977713696195089e-06,
+      "loss": 0.7153,
+      "step": 27285
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.523396917551104,
+      "learning_rate": 5.977015113860762e-06,
+      "loss": 0.7075,
+      "step": 27286
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3319237052673267,
+      "learning_rate": 5.9763165549495286e-06,
+      "loss": 0.6784,
+      "step": 27287
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.260677399599114,
+      "learning_rate": 5.975618019465454e-06,
+      "loss": 0.6859,
+      "step": 27288
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3181280762374876,
+      "learning_rate": 5.974919507412607e-06,
+      "loss": 0.7342,
+      "step": 27289
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.668619934818777,
+      "learning_rate": 5.974221018795048e-06,
+      "loss": 0.6486,
+      "step": 27290
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8655884181885536,
+      "learning_rate": 5.9735225536168505e-06,
+      "loss": 0.6171,
+      "step": 27291
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5633608662832106,
+      "learning_rate": 5.972824111882083e-06,
+      "loss": 0.6681,
+      "step": 27292
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3983740282497448,
+      "learning_rate": 5.972125693594806e-06,
+      "loss": 0.6679,
+      "step": 27293
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5962682788864714,
+      "learning_rate": 5.971427298759087e-06,
+      "loss": 0.4953,
+      "step": 27294
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3780269137347978,
+      "learning_rate": 5.970728927378997e-06,
+      "loss": 0.7047,
+      "step": 27295
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4370407245320762,
+      "learning_rate": 5.970030579458597e-06,
+      "loss": 0.7424,
+      "step": 27296
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3681929035689995,
+      "learning_rate": 5.969332255001955e-06,
+      "loss": 0.7184,
+      "step": 27297
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.250834107779669,
+      "learning_rate": 5.968633954013134e-06,
+      "loss": 0.6067,
+      "step": 27298
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4716153383702222,
+      "learning_rate": 5.967935676496208e-06,
+      "loss": 0.6458,
+      "step": 27299
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7759706842150473,
+      "learning_rate": 5.967237422455234e-06,
+      "loss": 0.7771,
+      "step": 27300
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.337249437148123,
+      "learning_rate": 5.966539191894279e-06,
+      "loss": 0.6284,
+      "step": 27301
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3696693841967225,
+      "learning_rate": 5.965840984817415e-06,
+      "loss": 0.7563,
+      "step": 27302
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4288852604795772,
+      "learning_rate": 5.965142801228696e-06,
+      "loss": 0.7323,
+      "step": 27303
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3666414988335591,
+      "learning_rate": 5.964444641132194e-06,
+      "loss": 0.6729,
+      "step": 27304
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2981013573648867,
+      "learning_rate": 5.9637465045319734e-06,
+      "loss": 0.7413,
+      "step": 27305
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4457731639661862,
+      "learning_rate": 5.963048391432102e-06,
+      "loss": 0.6158,
+      "step": 27306
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2357386952435876,
+      "learning_rate": 5.962350301836637e-06,
+      "loss": 0.7941,
+      "step": 27307
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5631663486973273,
+      "learning_rate": 5.961652235749648e-06,
+      "loss": 0.7314,
+      "step": 27308
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6281917267880837,
+      "learning_rate": 5.9609541931752015e-06,
+      "loss": 0.5291,
+      "step": 27309
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.390187449322235,
+      "learning_rate": 5.960256174117355e-06,
+      "loss": 0.5919,
+      "step": 27310
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5349481032470436,
+      "learning_rate": 5.959558178580174e-06,
+      "loss": 0.5986,
+      "step": 27311
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5709998379476806,
+      "learning_rate": 5.958860206567727e-06,
+      "loss": 0.5116,
+      "step": 27312
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5033046536688217,
+      "learning_rate": 5.958162258084077e-06,
+      "loss": 0.5171,
+      "step": 27313
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4084443543344214,
+      "learning_rate": 5.957464333133282e-06,
+      "loss": 0.7803,
+      "step": 27314
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3149235729527649,
+      "learning_rate": 5.9567664317194115e-06,
+      "loss": 0.6354,
+      "step": 27315
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5990612725718282,
+      "learning_rate": 5.956068553846528e-06,
+      "loss": 0.4971,
+      "step": 27316
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6649503346455974,
+      "learning_rate": 5.955370699518694e-06,
+      "loss": 0.7134,
+      "step": 27317
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.552349102232897,
+      "learning_rate": 5.954672868739968e-06,
+      "loss": 0.6901,
+      "step": 27318
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3958591855498579,
+      "learning_rate": 5.95397506151442e-06,
+      "loss": 0.6616,
+      "step": 27319
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5520555171893997,
+      "learning_rate": 5.953277277846112e-06,
+      "loss": 0.6206,
+      "step": 27320
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7569906665225825,
+      "learning_rate": 5.952579517739103e-06,
+      "loss": 0.6778,
+      "step": 27321
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3693328766998767,
+      "learning_rate": 5.951881781197456e-06,
+      "loss": 0.6427,
+      "step": 27322
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4722965808912367,
+      "learning_rate": 5.95118406822524e-06,
+      "loss": 0.6533,
+      "step": 27323
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4242994924172176,
+      "learning_rate": 5.950486378826508e-06,
+      "loss": 0.8271,
+      "step": 27324
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2801958985907393,
+      "learning_rate": 5.949788713005326e-06,
+      "loss": 0.5608,
+      "step": 27325
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.360783989971945,
+      "learning_rate": 5.949091070765755e-06,
+      "loss": 0.7966,
+      "step": 27326
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6159461641324099,
+      "learning_rate": 5.948393452111863e-06,
+      "loss": 0.4399,
+      "step": 27327
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2137251685005281,
+      "learning_rate": 5.9476958570477036e-06,
+      "loss": 0.5752,
+      "step": 27328
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1982778838597208,
+      "learning_rate": 5.946998285577343e-06,
+      "loss": 0.5899,
+      "step": 27329
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5320837998757595,
+      "learning_rate": 5.946300737704845e-06,
+      "loss": 0.685,
+      "step": 27330
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2326441464053806,
+      "learning_rate": 5.945603213434261e-06,
+      "loss": 0.7702,
+      "step": 27331
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3747551377891891,
+      "learning_rate": 5.9449057127696616e-06,
+      "loss": 0.7538,
+      "step": 27332
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4171599011427558,
+      "learning_rate": 5.944208235715105e-06,
+      "loss": 0.5959,
+      "step": 27333
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4514241573660425,
+      "learning_rate": 5.943510782274655e-06,
+      "loss": 0.6266,
+      "step": 27334
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6893160086147176,
+      "learning_rate": 5.942813352452365e-06,
+      "loss": 0.7562,
+      "step": 27335
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4135580556537168,
+      "learning_rate": 5.9421159462523e-06,
+      "loss": 0.7351,
+      "step": 27336
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4503977128594632,
+      "learning_rate": 5.941418563678525e-06,
+      "loss": 0.7642,
+      "step": 27337
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5019528886644644,
+      "learning_rate": 5.940721204735094e-06,
+      "loss": 0.7786,
+      "step": 27338
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4136961665489844,
+      "learning_rate": 5.940023869426067e-06,
+      "loss": 0.6587,
+      "step": 27339
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3863489598296632,
+      "learning_rate": 5.939326557755508e-06,
+      "loss": 0.7455,
+      "step": 27340
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2656994247082407,
+      "learning_rate": 5.938629269727477e-06,
+      "loss": 0.6639,
+      "step": 27341
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.491065413153557,
+      "learning_rate": 5.937932005346031e-06,
+      "loss": 0.7958,
+      "step": 27342
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5276612194677996,
+      "learning_rate": 5.93723476461523e-06,
+      "loss": 0.7505,
+      "step": 27343
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3194528528305671,
+      "learning_rate": 5.936537547539136e-06,
+      "loss": 0.6265,
+      "step": 27344
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5899497751973128,
+      "learning_rate": 5.9358403541218055e-06,
+      "loss": 0.4943,
+      "step": 27345
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4282279796480755,
+      "learning_rate": 5.9351431843673e-06,
+      "loss": 0.6143,
+      "step": 27346
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3268114941657758,
+      "learning_rate": 5.934446038279676e-06,
+      "loss": 0.7201,
+      "step": 27347
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3871841696277314,
+      "learning_rate": 5.933748915862997e-06,
+      "loss": 0.7096,
+      "step": 27348
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3981631223205166,
+      "learning_rate": 5.933051817121318e-06,
+      "loss": 0.6551,
+      "step": 27349
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9562363456303438,
+      "learning_rate": 5.9323547420586965e-06,
+      "loss": 0.6998,
+      "step": 27350
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4625454371327724,
+      "learning_rate": 5.9316576906792e-06,
+      "loss": 0.6655,
+      "step": 27351
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4132801308152672,
+      "learning_rate": 5.930960662986873e-06,
+      "loss": 0.7492,
+      "step": 27352
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.577089478039649,
+      "learning_rate": 5.930263658985785e-06,
+      "loss": 0.7159,
+      "step": 27353
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4027632156677974,
+      "learning_rate": 5.9295666786799874e-06,
+      "loss": 0.7622,
+      "step": 27354
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4443758896103653,
+      "learning_rate": 5.928869722073549e-06,
+      "loss": 0.6504,
+      "step": 27355
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.380636052217967,
+      "learning_rate": 5.928172789170512e-06,
+      "loss": 0.6887,
+      "step": 27356
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.616618633951516,
+      "learning_rate": 5.927475879974946e-06,
+      "loss": 0.5337,
+      "step": 27357
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5205280566980999,
+      "learning_rate": 5.926778994490905e-06,
+      "loss": 0.7884,
+      "step": 27358
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.414998676228942,
+      "learning_rate": 5.926082132722445e-06,
+      "loss": 0.7308,
+      "step": 27359
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7079743740180793,
+      "learning_rate": 5.925385294673623e-06,
+      "loss": 0.7383,
+      "step": 27360
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5856281650714354,
+      "learning_rate": 5.924688480348498e-06,
+      "loss": 0.8269,
+      "step": 27361
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.467280999906115,
+      "learning_rate": 5.92399168975113e-06,
+      "loss": 0.7181,
+      "step": 27362
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4652908167597074,
+      "learning_rate": 5.923294922885569e-06,
+      "loss": 0.6673,
+      "step": 27363
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6007571251134984,
+      "learning_rate": 5.922598179755879e-06,
+      "loss": 0.659,
+      "step": 27364
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7513266755727106,
+      "learning_rate": 5.921901460366113e-06,
+      "loss": 0.6961,
+      "step": 27365
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4194848323466256,
+      "learning_rate": 5.921204764720327e-06,
+      "loss": 0.6869,
+      "step": 27366
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4266034085529864,
+      "learning_rate": 5.920508092822575e-06,
+      "loss": 0.635,
+      "step": 27367
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6390471838983889,
+      "learning_rate": 5.919811444676919e-06,
+      "loss": 0.5394,
+      "step": 27368
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6068369355393411,
+      "learning_rate": 5.919114820287416e-06,
+      "loss": 0.5444,
+      "step": 27369
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.815306916633385,
+      "learning_rate": 5.918418219658115e-06,
+      "loss": 0.7429,
+      "step": 27370
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2909527918343322,
+      "learning_rate": 5.917721642793074e-06,
+      "loss": 0.7149,
+      "step": 27371
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2914622510390195,
+      "learning_rate": 5.917025089696355e-06,
+      "loss": 0.7114,
+      "step": 27372
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6594954111192726,
+      "learning_rate": 5.916328560372004e-06,
+      "loss": 0.7916,
+      "step": 27373
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.750521798009164,
+      "learning_rate": 5.915632054824083e-06,
+      "loss": 0.6417,
+      "step": 27374
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6140772246285067,
+      "learning_rate": 5.914935573056641e-06,
+      "loss": 0.5529,
+      "step": 27375
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5859532515574235,
+      "learning_rate": 5.914239115073743e-06,
+      "loss": 0.4779,
+      "step": 27376
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4089590344931884,
+      "learning_rate": 5.913542680879437e-06,
+      "loss": 0.7578,
+      "step": 27377
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.617326031217511,
+      "learning_rate": 5.912846270477779e-06,
+      "loss": 0.5194,
+      "step": 27378
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4368995569830711,
+      "learning_rate": 5.912149883872822e-06,
+      "loss": 0.6464,
+      "step": 27379
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3571493786031195,
+      "learning_rate": 5.911453521068626e-06,
+      "loss": 0.6951,
+      "step": 27380
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6064585707892199,
+      "learning_rate": 5.91075718206924e-06,
+      "loss": 0.5529,
+      "step": 27381
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2145012611199812,
+      "learning_rate": 5.910060866878719e-06,
+      "loss": 0.7272,
+      "step": 27382
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3925686360001794,
+      "learning_rate": 5.909364575501123e-06,
+      "loss": 0.6968,
+      "step": 27383
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5084378564519252,
+      "learning_rate": 5.908668307940497e-06,
+      "loss": 0.6926,
+      "step": 27384
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3472643518889456,
+      "learning_rate": 5.9079720642009004e-06,
+      "loss": 0.7811,
+      "step": 27385
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.953757845320988,
+      "learning_rate": 5.907275844286386e-06,
+      "loss": 0.7273,
+      "step": 27386
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4724393568757874,
+      "learning_rate": 5.9065796482010104e-06,
+      "loss": 0.7006,
+      "step": 27387
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4870376392371527,
+      "learning_rate": 5.905883475948819e-06,
+      "loss": 0.6529,
+      "step": 27388
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3758415688709986,
+      "learning_rate": 5.905187327533873e-06,
+      "loss": 0.7351,
+      "step": 27389
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2495530745352663,
+      "learning_rate": 5.904491202960225e-06,
+      "loss": 0.7153,
+      "step": 27390
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3451806403093824,
+      "learning_rate": 5.903795102231923e-06,
+      "loss": 0.7735,
+      "step": 27391
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3277445676479835,
+      "learning_rate": 5.903099025353021e-06,
+      "loss": 0.6447,
+      "step": 27392
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4886779635206842,
+      "learning_rate": 5.902402972327575e-06,
+      "loss": 0.6557,
+      "step": 27393
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.963722411684426,
+      "learning_rate": 5.9017069431596395e-06,
+      "loss": 0.7394,
+      "step": 27394
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4056189990007306,
+      "learning_rate": 5.901010937853261e-06,
+      "loss": 0.663,
+      "step": 27395
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5909916203417604,
+      "learning_rate": 5.900314956412492e-06,
+      "loss": 0.5276,
+      "step": 27396
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7223528167966962,
+      "learning_rate": 5.8996189988413925e-06,
+      "loss": 0.5892,
+      "step": 27397
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5208836699825967,
+      "learning_rate": 5.898923065144006e-06,
+      "loss": 0.7611,
+      "step": 27398
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.768871443528494,
+      "learning_rate": 5.8982271553243866e-06,
+      "loss": 0.6996,
+      "step": 27399
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3924336847336176,
+      "learning_rate": 5.89753126938659e-06,
+      "loss": 0.782,
+      "step": 27400
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4887861112433074,
+      "learning_rate": 5.896835407334665e-06,
+      "loss": 0.7949,
+      "step": 27401
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.227295635725487,
+      "learning_rate": 5.8961395691726635e-06,
+      "loss": 0.583,
+      "step": 27402
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.537700646125286,
+      "learning_rate": 5.8954437549046325e-06,
+      "loss": 0.63,
+      "step": 27403
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5955245614770517,
+      "learning_rate": 5.894747964534635e-06,
+      "loss": 0.4955,
+      "step": 27404
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.695660506505971,
+      "learning_rate": 5.8940521980667086e-06,
+      "loss": 0.7026,
+      "step": 27405
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.426844955745929,
+      "learning_rate": 5.893356455504911e-06,
+      "loss": 0.5975,
+      "step": 27406
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5647539300623554,
+      "learning_rate": 5.892660736853291e-06,
+      "loss": 0.8127,
+      "step": 27407
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.508937453446293,
+      "learning_rate": 5.891965042115907e-06,
+      "loss": 0.8752,
+      "step": 27408
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0427018565874304,
+      "learning_rate": 5.891269371296798e-06,
+      "loss": 0.7705,
+      "step": 27409
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4256542857493824,
+      "learning_rate": 5.89057372440002e-06,
+      "loss": 0.7701,
+      "step": 27410
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.648137909771297,
+      "learning_rate": 5.889878101429624e-06,
+      "loss": 0.6747,
+      "step": 27411
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.023496287477565,
+      "learning_rate": 5.889182502389657e-06,
+      "loss": 0.7857,
+      "step": 27412
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.59250542978291,
+      "learning_rate": 5.88848692728417e-06,
+      "loss": 0.6823,
+      "step": 27413
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.364352312987842,
+      "learning_rate": 5.887791376117216e-06,
+      "loss": 0.6704,
+      "step": 27414
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.366057283476365,
+      "learning_rate": 5.887095848892846e-06,
+      "loss": 0.7085,
+      "step": 27415
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6040965634031898,
+      "learning_rate": 5.886400345615098e-06,
+      "loss": 0.7313,
+      "step": 27416
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6108648117492931,
+      "learning_rate": 5.885704866288034e-06,
+      "loss": 0.5327,
+      "step": 27417
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8950469280942543,
+      "learning_rate": 5.885009410915701e-06,
+      "loss": 0.6877,
+      "step": 27418
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2048555180061127,
+      "learning_rate": 5.8843139795021446e-06,
+      "loss": 0.7554,
+      "step": 27419
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6084255735119014,
+      "learning_rate": 5.883618572051411e-06,
+      "loss": 0.5601,
+      "step": 27420
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5238837718578064,
+      "learning_rate": 5.882923188567558e-06,
+      "loss": 0.545,
+      "step": 27421
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.627005944206406,
+      "learning_rate": 5.88222782905463e-06,
+      "loss": 0.6825,
+      "step": 27422
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.632261881519479,
+      "learning_rate": 5.881532493516673e-06,
+      "loss": 0.5376,
+      "step": 27423
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2929192825557687,
+      "learning_rate": 5.880837181957737e-06,
+      "loss": 0.6473,
+      "step": 27424
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3296400839126847,
+      "learning_rate": 5.880141894381873e-06,
+      "loss": 0.7552,
+      "step": 27425
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4674166081974884,
+      "learning_rate": 5.879446630793126e-06,
+      "loss": 0.7092,
+      "step": 27426
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6034984724829079,
+      "learning_rate": 5.878751391195546e-06,
+      "loss": 0.4821,
+      "step": 27427
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2248398954602426,
+      "learning_rate": 5.8780561755931785e-06,
+      "loss": 0.6968,
+      "step": 27428
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.326939677183078,
+      "learning_rate": 5.877360983990077e-06,
+      "loss": 0.767,
+      "step": 27429
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5045457675807294,
+      "learning_rate": 5.876665816390281e-06,
+      "loss": 0.6754,
+      "step": 27430
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2395575709588271,
+      "learning_rate": 5.875970672797844e-06,
+      "loss": 0.6848,
+      "step": 27431
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6619758145883925,
+      "learning_rate": 5.8752755532168125e-06,
+      "loss": 0.5654,
+      "step": 27432
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3006908191263373,
+      "learning_rate": 5.874580457651229e-06,
+      "loss": 0.7514,
+      "step": 27433
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3234093954351278,
+      "learning_rate": 5.8738853861051455e-06,
+      "loss": 0.6922,
+      "step": 27434
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2401008350791172,
+      "learning_rate": 5.873190338582608e-06,
+      "loss": 0.583,
+      "step": 27435
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.381764801860899,
+      "learning_rate": 5.8724953150876655e-06,
+      "loss": 0.7253,
+      "step": 27436
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9991569649228347,
+      "learning_rate": 5.871800315624357e-06,
+      "loss": 0.8249,
+      "step": 27437
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5957777905461108,
+      "learning_rate": 5.871105340196737e-06,
+      "loss": 0.6395,
+      "step": 27438
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5497342151328062,
+      "learning_rate": 5.87041038880885e-06,
+      "loss": 0.6501,
+      "step": 27439
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8037378361382246,
+      "learning_rate": 5.86971546146474e-06,
+      "loss": 0.6065,
+      "step": 27440
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4983269203548846,
+      "learning_rate": 5.869020558168452e-06,
+      "loss": 0.7265,
+      "step": 27441
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8498572804840112,
+      "learning_rate": 5.868325678924036e-06,
+      "loss": 0.6702,
+      "step": 27442
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3793360941142632,
+      "learning_rate": 5.867630823735537e-06,
+      "loss": 0.7683,
+      "step": 27443
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5033463906951512,
+      "learning_rate": 5.866935992606999e-06,
+      "loss": 0.6719,
+      "step": 27444
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.11025717757898,
+      "learning_rate": 5.866241185542466e-06,
+      "loss": 0.6891,
+      "step": 27445
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4256926010730684,
+      "learning_rate": 5.865546402545989e-06,
+      "loss": 0.7192,
+      "step": 27446
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.193067213890712,
+      "learning_rate": 5.864851643621607e-06,
+      "loss": 0.7312,
+      "step": 27447
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.460365389395997,
+      "learning_rate": 5.864156908773369e-06,
+      "loss": 0.6263,
+      "step": 27448
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.432116158745216,
+      "learning_rate": 5.863462198005318e-06,
+      "loss": 0.6027,
+      "step": 27449
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6351384353505622,
+      "learning_rate": 5.862767511321502e-06,
+      "loss": 0.5158,
+      "step": 27450
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4502545274311296,
+      "learning_rate": 5.862072848725961e-06,
+      "loss": 0.6207,
+      "step": 27451
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5436102452526408,
+      "learning_rate": 5.8613782102227416e-06,
+      "loss": 0.7724,
+      "step": 27452
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5147497665702319,
+      "learning_rate": 5.860683595815894e-06,
+      "loss": 0.6824,
+      "step": 27453
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3649142546855473,
+      "learning_rate": 5.85998900550945e-06,
+      "loss": 0.7142,
+      "step": 27454
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5140281794045625,
+      "learning_rate": 5.8592944393074634e-06,
+      "loss": 0.5875,
+      "step": 27455
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3951507732211879,
+      "learning_rate": 5.858599897213975e-06,
+      "loss": 0.7286,
+      "step": 27456
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.20694730682673,
+      "learning_rate": 5.857905379233033e-06,
+      "loss": 0.6641,
+      "step": 27457
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.483671112360524,
+      "learning_rate": 5.857210885368672e-06,
+      "loss": 0.549,
+      "step": 27458
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4644146228298525,
+      "learning_rate": 5.856516415624942e-06,
+      "loss": 0.8168,
+      "step": 27459
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6490641047508241,
+      "learning_rate": 5.855821970005888e-06,
+      "loss": 0.5181,
+      "step": 27460
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.2501187234088036,
+      "learning_rate": 5.855127548515547e-06,
+      "loss": 0.8066,
+      "step": 27461
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.619438392032533,
+      "learning_rate": 5.8544331511579664e-06,
+      "loss": 0.7888,
+      "step": 27462
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4658833421930324,
+      "learning_rate": 5.85373877793719e-06,
+      "loss": 0.7199,
+      "step": 27463
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3688157520049797,
+      "learning_rate": 5.8530444288572595e-06,
+      "loss": 0.7105,
+      "step": 27464
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3892819312049827,
+      "learning_rate": 5.852350103922214e-06,
+      "loss": 0.6124,
+      "step": 27465
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5740001023948931,
+      "learning_rate": 5.851655803136101e-06,
+      "loss": 0.5185,
+      "step": 27466
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3206724789180901,
+      "learning_rate": 5.850961526502964e-06,
+      "loss": 0.6322,
+      "step": 27467
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3935966739748569,
+      "learning_rate": 5.850267274026839e-06,
+      "loss": 0.5768,
+      "step": 27468
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5885676981936714,
+      "learning_rate": 5.849573045711771e-06,
+      "loss": 0.9012,
+      "step": 27469
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7984208007902094,
+      "learning_rate": 5.848878841561805e-06,
+      "loss": 0.5507,
+      "step": 27470
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.354543728883897,
+      "learning_rate": 5.848184661580982e-06,
+      "loss": 0.6673,
+      "step": 27471
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6235060824752414,
+      "learning_rate": 5.847490505773341e-06,
+      "loss": 0.509,
+      "step": 27472
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3447206954119906,
+      "learning_rate": 5.846796374142923e-06,
+      "loss": 0.6433,
+      "step": 27473
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5208011464186373,
+      "learning_rate": 5.8461022666937735e-06,
+      "loss": 0.6066,
+      "step": 27474
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.476122660047883,
+      "learning_rate": 5.8454081834299324e-06,
+      "loss": 0.6785,
+      "step": 27475
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8439582200524642,
+      "learning_rate": 5.844714124355438e-06,
+      "loss": 0.8108,
+      "step": 27476
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5829973516940048,
+      "learning_rate": 5.844020089474334e-06,
+      "loss": 0.7516,
+      "step": 27477
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4229339456284253,
+      "learning_rate": 5.8433260787906624e-06,
+      "loss": 0.6099,
+      "step": 27478
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2899648143519387,
+      "learning_rate": 5.84263209230846e-06,
+      "loss": 0.8321,
+      "step": 27479
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.452004768167644,
+      "learning_rate": 5.841938130031772e-06,
+      "loss": 0.7428,
+      "step": 27480
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4292593978817858,
+      "learning_rate": 5.841244191964637e-06,
+      "loss": 0.6376,
+      "step": 27481
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.094207578212418,
+      "learning_rate": 5.8405502781110925e-06,
+      "loss": 0.6739,
+      "step": 27482
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2233634534620128,
+      "learning_rate": 5.839856388475182e-06,
+      "loss": 0.6345,
+      "step": 27483
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4455325506045962,
+      "learning_rate": 5.839162523060944e-06,
+      "loss": 0.6424,
+      "step": 27484
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.51599162912985,
+      "learning_rate": 5.8384686818724205e-06,
+      "loss": 0.75,
+      "step": 27485
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.220879214428202,
+      "learning_rate": 5.83777486491365e-06,
+      "loss": 0.827,
+      "step": 27486
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6139113724568128,
+      "learning_rate": 5.837081072188669e-06,
+      "loss": 0.5064,
+      "step": 27487
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3549069789662203,
+      "learning_rate": 5.836387303701525e-06,
+      "loss": 0.8882,
+      "step": 27488
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3770407041384944,
+      "learning_rate": 5.835693559456248e-06,
+      "loss": 0.7001,
+      "step": 27489
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.5968375624682425,
+      "learning_rate": 5.8349998394568845e-06,
+      "loss": 0.7622,
+      "step": 27490
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4712346404461987,
+      "learning_rate": 5.834306143707467e-06,
+      "loss": 0.8284,
+      "step": 27491
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4544798701135857,
+      "learning_rate": 5.833612472212043e-06,
+      "loss": 0.8144,
+      "step": 27492
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4394216688307016,
+      "learning_rate": 5.83291882497464e-06,
+      "loss": 0.5853,
+      "step": 27493
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3795196449772629,
+      "learning_rate": 5.832225201999308e-06,
+      "loss": 0.743,
+      "step": 27494
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3452814064502499,
+      "learning_rate": 5.83153160329008e-06,
+      "loss": 0.6754,
+      "step": 27495
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3531309491649743,
+      "learning_rate": 5.830838028850992e-06,
+      "loss": 0.7581,
+      "step": 27496
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4313437331195267,
+      "learning_rate": 5.830144478686084e-06,
+      "loss": 0.7038,
+      "step": 27497
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3704688244908025,
+      "learning_rate": 5.829450952799398e-06,
+      "loss": 0.6523,
+      "step": 27498
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.688145669959995,
+      "learning_rate": 5.828757451194971e-06,
+      "loss": 0.6837,
+      "step": 27499
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.291211721652013,
+      "learning_rate": 5.828063973876834e-06,
+      "loss": 0.7817,
+      "step": 27500
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3141683863744045,
+      "learning_rate": 5.827370520849029e-06,
+      "loss": 0.8005,
+      "step": 27501
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4040148840758835,
+      "learning_rate": 5.826677092115598e-06,
+      "loss": 0.5705,
+      "step": 27502
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4070157463410613,
+      "learning_rate": 5.8259836876805745e-06,
+      "loss": 0.8583,
+      "step": 27503
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4059304196986266,
+      "learning_rate": 5.82529030754799e-06,
+      "loss": 0.7081,
+      "step": 27504
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4453942607314145,
+      "learning_rate": 5.824596951721888e-06,
+      "loss": 0.8758,
+      "step": 27505
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2834950018784286,
+      "learning_rate": 5.823903620206308e-06,
+      "loss": 0.6595,
+      "step": 27506
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6982258750716546,
+      "learning_rate": 5.8232103130052855e-06,
+      "loss": 0.7253,
+      "step": 27507
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4107103116222162,
+      "learning_rate": 5.8225170301228485e-06,
+      "loss": 0.6462,
+      "step": 27508
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4509601922763902,
+      "learning_rate": 5.8218237715630396e-06,
+      "loss": 0.6468,
+      "step": 27509
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2713148158787235,
+      "learning_rate": 5.821130537329902e-06,
+      "loss": 0.6872,
+      "step": 27510
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6890800025889101,
+      "learning_rate": 5.820437327427464e-06,
+      "loss": 0.6779,
+      "step": 27511
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4407662081909487,
+      "learning_rate": 5.819744141859759e-06,
+      "loss": 0.6481,
+      "step": 27512
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3166428440010522,
+      "learning_rate": 5.8190509806308316e-06,
+      "loss": 0.5473,
+      "step": 27513
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.524239381624361,
+      "learning_rate": 5.8183578437447085e-06,
+      "loss": 0.6492,
+      "step": 27514
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.387042809227752,
+      "learning_rate": 5.817664731205429e-06,
+      "loss": 0.7853,
+      "step": 27515
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2956703084458057,
+      "learning_rate": 5.816971643017036e-06,
+      "loss": 0.6544,
+      "step": 27516
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3143514749794158,
+      "learning_rate": 5.816278579183556e-06,
+      "loss": 0.6781,
+      "step": 27517
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2835665334901314,
+      "learning_rate": 5.815585539709023e-06,
+      "loss": 0.7364,
+      "step": 27518
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.399495013775873,
+      "learning_rate": 5.814892524597477e-06,
+      "loss": 0.6013,
+      "step": 27519
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5709498517070665,
+      "learning_rate": 5.814199533852959e-06,
+      "loss": 0.5012,
+      "step": 27520
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.177715467004485,
+      "learning_rate": 5.8135065674794885e-06,
+      "loss": 0.6849,
+      "step": 27521
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.94635133569441,
+      "learning_rate": 5.812813625481108e-06,
+      "loss": 0.6474,
+      "step": 27522
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7059024632404276,
+      "learning_rate": 5.812120707861852e-06,
+      "loss": 0.7031,
+      "step": 27523
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5992032490637672,
+      "learning_rate": 5.811427814625765e-06,
+      "loss": 0.5427,
+      "step": 27524
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3028833564562923,
+      "learning_rate": 5.8107349457768615e-06,
+      "loss": 0.5737,
+      "step": 27525
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3337073444363423,
+      "learning_rate": 5.810042101319185e-06,
+      "loss": 0.6834,
+      "step": 27526
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4423146782821787,
+      "learning_rate": 5.8093492812567754e-06,
+      "loss": 0.7137,
+      "step": 27527
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4434022574045648,
+      "learning_rate": 5.808656485593657e-06,
+      "loss": 0.6724,
+      "step": 27528
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.46507788222503,
+      "learning_rate": 5.80796371433387e-06,
+      "loss": 0.7856,
+      "step": 27529
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.436158322000714,
+      "learning_rate": 5.807270967481442e-06,
+      "loss": 0.6096,
+      "step": 27530
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.275260152464568,
+      "learning_rate": 5.8065782450404155e-06,
+      "loss": 0.7377,
+      "step": 27531
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.521867095115403,
+      "learning_rate": 5.805885547014812e-06,
+      "loss": 0.6401,
+      "step": 27532
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4208184784663385,
+      "learning_rate": 5.805192873408676e-06,
+      "loss": 0.6656,
+      "step": 27533
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9568184204818024,
+      "learning_rate": 5.804500224226034e-06,
+      "loss": 0.615,
+      "step": 27534
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.352653015766192,
+      "learning_rate": 5.803807599470917e-06,
+      "loss": 0.6621,
+      "step": 27535
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4052122098007735,
+      "learning_rate": 5.80311499914736e-06,
+      "loss": 0.7324,
+      "step": 27536
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5764766424214485,
+      "learning_rate": 5.8024224232594e-06,
+      "loss": 0.5391,
+      "step": 27537
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2528601494467742,
+      "learning_rate": 5.801729871811066e-06,
+      "loss": 0.8037,
+      "step": 27538
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3878396980584633,
+      "learning_rate": 5.801037344806386e-06,
+      "loss": 0.7791,
+      "step": 27539
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.459693320541034,
+      "learning_rate": 5.800344842249396e-06,
+      "loss": 0.7965,
+      "step": 27540
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4915738403589736,
+      "learning_rate": 5.79965236414413e-06,
+      "loss": 0.7232,
+      "step": 27541
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6039175226068495,
+      "learning_rate": 5.798959910494622e-06,
+      "loss": 0.4955,
+      "step": 27542
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.413709055886406,
+      "learning_rate": 5.798267481304891e-06,
+      "loss": 0.693,
+      "step": 27543
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3959191071727817,
+      "learning_rate": 5.797575076578978e-06,
+      "loss": 0.6026,
+      "step": 27544
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5647958895292422,
+      "learning_rate": 5.79688269632092e-06,
+      "loss": 0.6243,
+      "step": 27545
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6378726147605192,
+      "learning_rate": 5.79619034053474e-06,
+      "loss": 0.6669,
+      "step": 27546
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.355207921346134,
+      "learning_rate": 5.795498009224465e-06,
+      "loss": 0.6344,
+      "step": 27547
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2029485852028208,
+      "learning_rate": 5.794805702394138e-06,
+      "loss": 0.6951,
+      "step": 27548
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5833459638543297,
+      "learning_rate": 5.794113420047779e-06,
+      "loss": 0.4889,
+      "step": 27549
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.674287912623322,
+      "learning_rate": 5.793421162189427e-06,
+      "loss": 0.7534,
+      "step": 27550
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5064613510844418,
+      "learning_rate": 5.792728928823105e-06,
+      "loss": 0.6925,
+      "step": 27551
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5936337014678024,
+      "learning_rate": 5.792036719952851e-06,
+      "loss": 0.5328,
+      "step": 27552
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.401326129339685,
+      "learning_rate": 5.791344535582687e-06,
+      "loss": 0.5032,
+      "step": 27553
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4115573023704309,
+      "learning_rate": 5.790652375716653e-06,
+      "loss": 0.812,
+      "step": 27554
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4795192756057571,
+      "learning_rate": 5.789960240358772e-06,
+      "loss": 0.7397,
+      "step": 27555
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7165454452480877,
+      "learning_rate": 5.78926812951307e-06,
+      "loss": 0.6238,
+      "step": 27556
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2770293368527479,
+      "learning_rate": 5.788576043183585e-06,
+      "loss": 0.6386,
+      "step": 27557
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5944756984820063,
+      "learning_rate": 5.787883981374347e-06,
+      "loss": 0.514,
+      "step": 27558
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2301239953231036,
+      "learning_rate": 5.787191944089381e-06,
+      "loss": 0.7656,
+      "step": 27559
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.757332558362362,
+      "learning_rate": 5.786499931332714e-06,
+      "loss": 0.7083,
+      "step": 27560
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.271468020818912,
+      "learning_rate": 5.785807943108377e-06,
+      "loss": 0.6366,
+      "step": 27561
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6967960035461003,
+      "learning_rate": 5.785115979420406e-06,
+      "loss": 0.8057,
+      "step": 27562
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1086694504011154,
+      "learning_rate": 5.784424040272823e-06,
+      "loss": 0.6919,
+      "step": 27563
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6456663429513738,
+      "learning_rate": 5.7837321256696555e-06,
+      "loss": 0.6719,
+      "step": 27564
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7021310722325114,
+      "learning_rate": 5.783040235614933e-06,
+      "loss": 0.6698,
+      "step": 27565
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5640452363746924,
+      "learning_rate": 5.782348370112689e-06,
+      "loss": 0.6737,
+      "step": 27566
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5967352611217298,
+      "learning_rate": 5.7816565291669504e-06,
+      "loss": 0.5112,
+      "step": 27567
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.58932595110677,
+      "learning_rate": 5.780964712781736e-06,
+      "loss": 0.4877,
+      "step": 27568
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3472335342242832,
+      "learning_rate": 5.780272920961087e-06,
+      "loss": 0.6503,
+      "step": 27569
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5880608601787515,
+      "learning_rate": 5.779581153709019e-06,
+      "loss": 0.4753,
+      "step": 27570
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2861091786678227,
+      "learning_rate": 5.778889411029571e-06,
+      "loss": 0.6229,
+      "step": 27571
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3554108334350543,
+      "learning_rate": 5.778197692926761e-06,
+      "loss": 0.6453,
+      "step": 27572
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.406220854599577,
+      "learning_rate": 5.777505999404625e-06,
+      "loss": 0.6928,
+      "step": 27573
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3776189479911582,
+      "learning_rate": 5.776814330467182e-06,
+      "loss": 0.7067,
+      "step": 27574
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.420824983312619,
+      "learning_rate": 5.776122686118466e-06,
+      "loss": 0.7131,
+      "step": 27575
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4736555859493925,
+      "learning_rate": 5.775431066362503e-06,
+      "loss": 0.5771,
+      "step": 27576
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5868441413180853,
+      "learning_rate": 5.774739471203311e-06,
+      "loss": 0.4784,
+      "step": 27577
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4282058302589185,
+      "learning_rate": 5.774047900644926e-06,
+      "loss": 0.5654,
+      "step": 27578
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.381867616278949,
+      "learning_rate": 5.773356354691374e-06,
+      "loss": 0.7275,
+      "step": 27579
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5936609399365174,
+      "learning_rate": 5.772664833346681e-06,
+      "loss": 0.6474,
+      "step": 27580
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4211440729607836,
+      "learning_rate": 5.771973336614867e-06,
+      "loss": 0.733,
+      "step": 27581
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4904691984770364,
+      "learning_rate": 5.7712818644999624e-06,
+      "loss": 0.689,
+      "step": 27582
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5299058503095138,
+      "learning_rate": 5.7705904170059976e-06,
+      "loss": 0.8229,
+      "step": 27583
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4265137108633534,
+      "learning_rate": 5.769898994136996e-06,
+      "loss": 0.8458,
+      "step": 27584
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0669052480331485,
+      "learning_rate": 5.769207595896975e-06,
+      "loss": 0.8491,
+      "step": 27585
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6326009834334148,
+      "learning_rate": 5.768516222289968e-06,
+      "loss": 0.7273,
+      "step": 27586
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.43897041801736,
+      "learning_rate": 5.767824873320005e-06,
+      "loss": 0.7637,
+      "step": 27587
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2835374370455415,
+      "learning_rate": 5.767133548991104e-06,
+      "loss": 0.6301,
+      "step": 27588
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3810418184874524,
+      "learning_rate": 5.766442249307287e-06,
+      "loss": 0.8223,
+      "step": 27589
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4321935244372563,
+      "learning_rate": 5.765750974272588e-06,
+      "loss": 0.6711,
+      "step": 27590
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6033007252081625,
+      "learning_rate": 5.765059723891024e-06,
+      "loss": 0.5034,
+      "step": 27591
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5756965908093669,
+      "learning_rate": 5.764368498166626e-06,
+      "loss": 0.4815,
+      "step": 27592
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4580211438831812,
+      "learning_rate": 5.763677297103411e-06,
+      "loss": 0.7101,
+      "step": 27593
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4318602061767138,
+      "learning_rate": 5.7629861207054135e-06,
+      "loss": 0.707,
+      "step": 27594
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4103274754110418,
+      "learning_rate": 5.7622949689766496e-06,
+      "loss": 0.6009,
+      "step": 27595
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6053961438725911,
+      "learning_rate": 5.761603841921147e-06,
+      "loss": 0.513,
+      "step": 27596
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5815560565527234,
+      "learning_rate": 5.76091273954293e-06,
+      "loss": 0.7058,
+      "step": 27597
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5576275070849551,
+      "learning_rate": 5.760221661846017e-06,
+      "loss": 0.525,
+      "step": 27598
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4141296059517956,
+      "learning_rate": 5.7595306088344364e-06,
+      "loss": 0.793,
+      "step": 27599
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3355188581621775,
+      "learning_rate": 5.758839580512214e-06,
+      "loss": 0.6055,
+      "step": 27600
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4476478484038007,
+      "learning_rate": 5.7581485768833715e-06,
+      "loss": 0.5544,
+      "step": 27601
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4197674998119276,
+      "learning_rate": 5.757457597951925e-06,
+      "loss": 0.7318,
+      "step": 27602
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3250427092122976,
+      "learning_rate": 5.756766643721905e-06,
+      "loss": 0.7432,
+      "step": 27603
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3767506545998511,
+      "learning_rate": 5.756075714197338e-06,
+      "loss": 0.6702,
+      "step": 27604
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5536155666355431,
+      "learning_rate": 5.755384809382239e-06,
+      "loss": 0.7066,
+      "step": 27605
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4276342565398952,
+      "learning_rate": 5.754693929280633e-06,
+      "loss": 0.7725,
+      "step": 27606
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.488013069077554,
+      "learning_rate": 5.754003073896539e-06,
+      "loss": 0.6315,
+      "step": 27607
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.600844724083682,
+      "learning_rate": 5.75331224323399e-06,
+      "loss": 0.8288,
+      "step": 27608
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.465906913306619,
+      "learning_rate": 5.7526214372970014e-06,
+      "loss": 0.6642,
+      "step": 27609
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5251193097294042,
+      "learning_rate": 5.751930656089591e-06,
+      "loss": 0.5728,
+      "step": 27610
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4003491754246689,
+      "learning_rate": 5.751239899615789e-06,
+      "loss": 0.6484,
+      "step": 27611
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.270754968921773,
+      "learning_rate": 5.75054916787961e-06,
+      "loss": 0.6124,
+      "step": 27612
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4284003533364171,
+      "learning_rate": 5.7498584608850786e-06,
+      "loss": 0.7516,
+      "step": 27613
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.38874304005351,
+      "learning_rate": 5.749167778636221e-06,
+      "loss": 0.7781,
+      "step": 27614
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.616219778223299,
+      "learning_rate": 5.748477121137055e-06,
+      "loss": 0.7581,
+      "step": 27615
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4420721211733387,
+      "learning_rate": 5.7477864883915956e-06,
+      "loss": 0.7585,
+      "step": 27616
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6540774315753304,
+      "learning_rate": 5.74709588040387e-06,
+      "loss": 0.5516,
+      "step": 27617
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4072759559319568,
+      "learning_rate": 5.746405297177908e-06,
+      "loss": 0.6969,
+      "step": 27618
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.031653040766001,
+      "learning_rate": 5.745714738717711e-06,
+      "loss": 0.668,
+      "step": 27619
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3737769639342996,
+      "learning_rate": 5.745024205027312e-06,
+      "loss": 0.7068,
+      "step": 27620
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5624534131719539,
+      "learning_rate": 5.744333696110729e-06,
+      "loss": 0.7986,
+      "step": 27621
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.621636451791478,
+      "learning_rate": 5.743643211971989e-06,
+      "loss": 0.7762,
+      "step": 27622
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3665877310507955,
+      "learning_rate": 5.742952752615099e-06,
+      "loss": 0.5861,
+      "step": 27623
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3429660148591713,
+      "learning_rate": 5.742262318044084e-06,
+      "loss": 0.7837,
+      "step": 27624
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2595664361504648,
+      "learning_rate": 5.74157190826297e-06,
+      "loss": 0.6552,
+      "step": 27625
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2698169337295948,
+      "learning_rate": 5.740881523275771e-06,
+      "loss": 0.6877,
+      "step": 27626
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.322249673683954,
+      "learning_rate": 5.74019116308651e-06,
+      "loss": 0.6751,
+      "step": 27627
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.280404140074809,
+      "learning_rate": 5.7395008276992015e-06,
+      "loss": 0.7215,
+      "step": 27628
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.698617757291471,
+      "learning_rate": 5.738810517117872e-06,
+      "loss": 0.7503,
+      "step": 27629
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.420537006604186,
+      "learning_rate": 5.7381202313465334e-06,
+      "loss": 0.6651,
+      "step": 27630
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2117835562425712,
+      "learning_rate": 5.73742997038921e-06,
+      "loss": 0.612,
+      "step": 27631
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.440835788991958,
+      "learning_rate": 5.736739734249916e-06,
+      "loss": 0.7134,
+      "step": 27632
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4267952748941208,
+      "learning_rate": 5.736049522932679e-06,
+      "loss": 0.8649,
+      "step": 27633
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5002802122239902,
+      "learning_rate": 5.735359336441504e-06,
+      "loss": 0.5872,
+      "step": 27634
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4633105693113873,
+      "learning_rate": 5.734669174780423e-06,
+      "loss": 0.6665,
+      "step": 27635
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3910912542303913,
+      "learning_rate": 5.733979037953449e-06,
+      "loss": 0.7874,
+      "step": 27636
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.516568575547228,
+      "learning_rate": 5.733288925964595e-06,
+      "loss": 0.6133,
+      "step": 27637
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.333028440318043,
+      "learning_rate": 5.732598838817883e-06,
+      "loss": 0.6361,
+      "step": 27638
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3049217795150185,
+      "learning_rate": 5.731908776517336e-06,
+      "loss": 0.7327,
+      "step": 27639
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4606735481530706,
+      "learning_rate": 5.731218739066968e-06,
+      "loss": 0.6914,
+      "step": 27640
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3827616466319101,
+      "learning_rate": 5.730528726470792e-06,
+      "loss": 0.7083,
+      "step": 27641
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.338148318647758,
+      "learning_rate": 5.7298387387328295e-06,
+      "loss": 0.6053,
+      "step": 27642
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8606046571820822,
+      "learning_rate": 5.729148775857101e-06,
+      "loss": 0.7697,
+      "step": 27643
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1977087628143697,
+      "learning_rate": 5.72845883784762e-06,
+      "loss": 0.6159,
+      "step": 27644
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3310031584019617,
+      "learning_rate": 5.7277689247084016e-06,
+      "loss": 0.6325,
+      "step": 27645
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2650793412264683,
+      "learning_rate": 5.727079036443464e-06,
+      "loss": 0.6221,
+      "step": 27646
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9796697419979494,
+      "learning_rate": 5.726389173056829e-06,
+      "loss": 0.6778,
+      "step": 27647
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2039670741003476,
+      "learning_rate": 5.725699334552509e-06,
+      "loss": 0.7345,
+      "step": 27648
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4062303246381658,
+      "learning_rate": 5.725009520934517e-06,
+      "loss": 0.8086,
+      "step": 27649
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.778092726448735,
+      "learning_rate": 5.724319732206878e-06,
+      "loss": 0.7276,
+      "step": 27650
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3469801135694395,
+      "learning_rate": 5.723629968373598e-06,
+      "loss": 0.679,
+      "step": 27651
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4022778549359176,
+      "learning_rate": 5.722940229438703e-06,
+      "loss": 0.8186,
+      "step": 27652
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6075366093579916,
+      "learning_rate": 5.722250515406199e-06,
+      "loss": 0.5363,
+      "step": 27653
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5128108584440985,
+      "learning_rate": 5.721560826280112e-06,
+      "loss": 0.8028,
+      "step": 27654
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5776249854338358,
+      "learning_rate": 5.720871162064447e-06,
+      "loss": 0.4826,
+      "step": 27655
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3430208802572678,
+      "learning_rate": 5.7201815227632305e-06,
+      "loss": 0.6713,
+      "step": 27656
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3028943055679814,
+      "learning_rate": 5.719491908380471e-06,
+      "loss": 0.6723,
+      "step": 27657
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2031237343621932,
+      "learning_rate": 5.718802318920181e-06,
+      "loss": 0.7348,
+      "step": 27658
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5616206827861518,
+      "learning_rate": 5.71811275438638e-06,
+      "loss": 0.7622,
+      "step": 27659
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5732522727696625,
+      "learning_rate": 5.717423214783085e-06,
+      "loss": 0.5793,
+      "step": 27660
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1795409533656929,
+      "learning_rate": 5.7167337001143085e-06,
+      "loss": 0.6463,
+      "step": 27661
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3105462632305747,
+      "learning_rate": 5.716044210384061e-06,
+      "loss": 0.8701,
+      "step": 27662
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2711052119466983,
+      "learning_rate": 5.7153547455963595e-06,
+      "loss": 0.728,
+      "step": 27663
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3229655065184671,
+      "learning_rate": 5.714665305755223e-06,
+      "loss": 0.825,
+      "step": 27664
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2889072647656883,
+      "learning_rate": 5.7139758908646625e-06,
+      "loss": 0.7844,
+      "step": 27665
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.566623666440481,
+      "learning_rate": 5.7132865009286875e-06,
+      "loss": 0.7202,
+      "step": 27666
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4789479195058097,
+      "learning_rate": 5.712597135951314e-06,
+      "loss": 0.6566,
+      "step": 27667
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.621321536099634,
+      "learning_rate": 5.711907795936564e-06,
+      "loss": 0.6354,
+      "step": 27668
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5697979908927966,
+      "learning_rate": 5.711218480888443e-06,
+      "loss": 0.8213,
+      "step": 27669
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3807217075852067,
+      "learning_rate": 5.710529190810962e-06,
+      "loss": 0.7838,
+      "step": 27670
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2806433813497753,
+      "learning_rate": 5.709839925708141e-06,
+      "loss": 0.6724,
+      "step": 27671
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3552891792203166,
+      "learning_rate": 5.709150685583987e-06,
+      "loss": 0.6723,
+      "step": 27672
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4075869227107387,
+      "learning_rate": 5.708461470442521e-06,
+      "loss": 0.8468,
+      "step": 27673
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.229162163306644,
+      "learning_rate": 5.7077722802877465e-06,
+      "loss": 0.5962,
+      "step": 27674
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.302787133168449,
+      "learning_rate": 5.7070831151236836e-06,
+      "loss": 0.8027,
+      "step": 27675
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4429437844225337,
+      "learning_rate": 5.706393974954338e-06,
+      "loss": 0.6404,
+      "step": 27676
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.773592444892416,
+      "learning_rate": 5.705704859783732e-06,
+      "loss": 0.7999,
+      "step": 27677
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3691664594584845,
+      "learning_rate": 5.70501576961587e-06,
+      "loss": 0.7516,
+      "step": 27678
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.1087089803604897,
+      "learning_rate": 5.7043267044547636e-06,
+      "loss": 0.7144,
+      "step": 27679
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1813613107920677,
+      "learning_rate": 5.703637664304425e-06,
+      "loss": 0.5563,
+      "step": 27680
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4849157615458939,
+      "learning_rate": 5.702948649168874e-06,
+      "loss": 0.6196,
+      "step": 27681
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3929255045555686,
+      "learning_rate": 5.702259659052115e-06,
+      "loss": 0.6384,
+      "step": 27682
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6152776522572891,
+      "learning_rate": 5.701570693958159e-06,
+      "loss": 0.7937,
+      "step": 27683
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3649339446069884,
+      "learning_rate": 5.700881753891017e-06,
+      "loss": 0.6832,
+      "step": 27684
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3782318407119076,
+      "learning_rate": 5.700192838854707e-06,
+      "loss": 0.7598,
+      "step": 27685
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.23939240166786,
+      "learning_rate": 5.6995039488532355e-06,
+      "loss": 0.7137,
+      "step": 27686
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4312520053569315,
+      "learning_rate": 5.6988150838906096e-06,
+      "loss": 0.7587,
+      "step": 27687
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6391505833201516,
+      "learning_rate": 5.698126243970845e-06,
+      "loss": 0.7533,
+      "step": 27688
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7764596705871616,
+      "learning_rate": 5.697437429097954e-06,
+      "loss": 0.7275,
+      "step": 27689
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.013889390345372,
+      "learning_rate": 5.6967486392759455e-06,
+      "loss": 0.7278,
+      "step": 27690
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3906590058757768,
+      "learning_rate": 5.696059874508823e-06,
+      "loss": 0.7389,
+      "step": 27691
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6066394839657765,
+      "learning_rate": 5.695371134800609e-06,
+      "loss": 0.5132,
+      "step": 27692
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3702444136125385,
+      "learning_rate": 5.694682420155302e-06,
+      "loss": 0.6523,
+      "step": 27693
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5937898618905216,
+      "learning_rate": 5.693993730576921e-06,
+      "loss": 0.539,
+      "step": 27694
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3828301626039856,
+      "learning_rate": 5.6933050660694674e-06,
+      "loss": 0.7129,
+      "step": 27695
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.500826066053566,
+      "learning_rate": 5.69261642663696e-06,
+      "loss": 0.7411,
+      "step": 27696
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3088746416814732,
+      "learning_rate": 5.6919278122834e-06,
+      "loss": 0.7748,
+      "step": 27697
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5070781831730964,
+      "learning_rate": 5.691239223012802e-06,
+      "loss": 0.6464,
+      "step": 27698
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.483029029397111,
+      "learning_rate": 5.690550658829175e-06,
+      "loss": 0.5755,
+      "step": 27699
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.613396940550984,
+      "learning_rate": 5.689862119736522e-06,
+      "loss": 0.7282,
+      "step": 27700
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3486264776249228,
+      "learning_rate": 5.689173605738856e-06,
+      "loss": 0.6403,
+      "step": 27701
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2134670137131949,
+      "learning_rate": 5.688485116840187e-06,
+      "loss": 0.7655,
+      "step": 27702
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5972743405385856,
+      "learning_rate": 5.68779665304453e-06,
+      "loss": 0.7173,
+      "step": 27703
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6459829283468281,
+      "learning_rate": 5.687108214355877e-06,
+      "loss": 0.6926,
+      "step": 27704
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5640007387187647,
+      "learning_rate": 5.686419800778246e-06,
+      "loss": 0.6659,
+      "step": 27705
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6231434603323018,
+      "learning_rate": 5.685731412315649e-06,
+      "loss": 0.8112,
+      "step": 27706
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2459345498027408,
+      "learning_rate": 5.685043048972091e-06,
+      "loss": 0.6015,
+      "step": 27707
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3369155684634018,
+      "learning_rate": 5.684354710751573e-06,
+      "loss": 0.6866,
+      "step": 27708
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2394614419436805,
+      "learning_rate": 5.683666397658108e-06,
+      "loss": 0.701,
+      "step": 27709
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3946054466261255,
+      "learning_rate": 5.6829781096957086e-06,
+      "loss": 0.6766,
+      "step": 27710
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.52769055624235,
+      "learning_rate": 5.682289846868372e-06,
+      "loss": 0.8132,
+      "step": 27711
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.26497717680111,
+      "learning_rate": 5.681601609180117e-06,
+      "loss": 0.5938,
+      "step": 27712
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6668314023883943,
+      "learning_rate": 5.680913396634943e-06,
+      "loss": 0.6598,
+      "step": 27713
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3207674943467473,
+      "learning_rate": 5.680225209236855e-06,
+      "loss": 0.78,
+      "step": 27714
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4197031485327927,
+      "learning_rate": 5.6795370469898625e-06,
+      "loss": 0.8972,
+      "step": 27715
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3054798402176877,
+      "learning_rate": 5.678848909897979e-06,
+      "loss": 0.6766,
+      "step": 27716
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6081521684258406,
+      "learning_rate": 5.678160797965204e-06,
+      "loss": 0.6389,
+      "step": 27717
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5070530567113574,
+      "learning_rate": 5.677472711195542e-06,
+      "loss": 0.6974,
+      "step": 27718
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4325931333804771,
+      "learning_rate": 5.676784649593001e-06,
+      "loss": 0.7024,
+      "step": 27719
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4416643444237103,
+      "learning_rate": 5.676096613161597e-06,
+      "loss": 0.8145,
+      "step": 27720
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3925128979445311,
+      "learning_rate": 5.67540860190532e-06,
+      "loss": 0.6426,
+      "step": 27721
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4041597091639724,
+      "learning_rate": 5.674720615828182e-06,
+      "loss": 0.775,
+      "step": 27722
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.60955700119129,
+      "learning_rate": 5.674032654934192e-06,
+      "loss": 0.5189,
+      "step": 27723
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9865036380747674,
+      "learning_rate": 5.673344719227356e-06,
+      "loss": 0.7003,
+      "step": 27724
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3524187947062916,
+      "learning_rate": 5.672656808711677e-06,
+      "loss": 0.7255,
+      "step": 27725
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6615480171726182,
+      "learning_rate": 5.671968923391156e-06,
+      "loss": 0.7649,
+      "step": 27726
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.458374492082499,
+      "learning_rate": 5.671281063269807e-06,
+      "loss": 0.6397,
+      "step": 27727
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6183378056311661,
+      "learning_rate": 5.670593228351625e-06,
+      "loss": 0.6789,
+      "step": 27728
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6926065212894497,
+      "learning_rate": 5.669905418640624e-06,
+      "loss": 0.6632,
+      "step": 27729
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3778948782117924,
+      "learning_rate": 5.669217634140802e-06,
+      "loss": 0.6088,
+      "step": 27730
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3603292123503368,
+      "learning_rate": 5.668529874856168e-06,
+      "loss": 0.7309,
+      "step": 27731
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.602955187057266,
+      "learning_rate": 5.667842140790722e-06,
+      "loss": 0.543,
+      "step": 27732
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3442522662066128,
+      "learning_rate": 5.667154431948474e-06,
+      "loss": 0.7381,
+      "step": 27733
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4652969517091103,
+      "learning_rate": 5.666466748333424e-06,
+      "loss": 0.802,
+      "step": 27734
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6116984517843593,
+      "learning_rate": 5.665779089949574e-06,
+      "loss": 0.5385,
+      "step": 27735
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2566291742100306,
+      "learning_rate": 5.665091456800929e-06,
+      "loss": 0.6073,
+      "step": 27736
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3322966588714147,
+      "learning_rate": 5.664403848891499e-06,
+      "loss": 0.7711,
+      "step": 27737
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3766706824513333,
+      "learning_rate": 5.663716266225281e-06,
+      "loss": 0.7529,
+      "step": 27738
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2671758896882344,
+      "learning_rate": 5.663028708806277e-06,
+      "loss": 0.6901,
+      "step": 27739
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.616154388503769,
+      "learning_rate": 5.662341176638493e-06,
+      "loss": 0.7355,
+      "step": 27740
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4355677881509716,
+      "learning_rate": 5.661653669725935e-06,
+      "loss": 0.8088,
+      "step": 27741
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.281063434855921,
+      "learning_rate": 5.660966188072603e-06,
+      "loss": 0.7335,
+      "step": 27742
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7773859878465779,
+      "learning_rate": 5.660278731682496e-06,
+      "loss": 0.5717,
+      "step": 27743
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3348807138434184,
+      "learning_rate": 5.65959130055962e-06,
+      "loss": 0.7488,
+      "step": 27744
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.110497306201815,
+      "learning_rate": 5.658903894707981e-06,
+      "loss": 0.6159,
+      "step": 27745
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.487395842790125,
+      "learning_rate": 5.65821651413158e-06,
+      "loss": 0.6054,
+      "step": 27746
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2425021888534382,
+      "learning_rate": 5.65752915883441e-06,
+      "loss": 0.6981,
+      "step": 27747
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.385566232039058,
+      "learning_rate": 5.656841828820486e-06,
+      "loss": 0.6395,
+      "step": 27748
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8501509689342546,
+      "learning_rate": 5.6561545240937995e-06,
+      "loss": 0.6621,
+      "step": 27749
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5783878105288417,
+      "learning_rate": 5.655467244658361e-06,
+      "loss": 0.4912,
+      "step": 27750
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4261406993171004,
+      "learning_rate": 5.654779990518163e-06,
+      "loss": 0.8019,
+      "step": 27751
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3453763395371268,
+      "learning_rate": 5.6540927616772165e-06,
+      "loss": 0.6661,
+      "step": 27752
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3987818478099856,
+      "learning_rate": 5.653405558139513e-06,
+      "loss": 0.786,
+      "step": 27753
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6295515963757479,
+      "learning_rate": 5.652718379909062e-06,
+      "loss": 0.5399,
+      "step": 27754
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.403698441288719,
+      "learning_rate": 5.652031226989858e-06,
+      "loss": 0.5852,
+      "step": 27755
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4975349976749557,
+      "learning_rate": 5.651344099385908e-06,
+      "loss": 0.7164,
+      "step": 27756
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4272524760412224,
+      "learning_rate": 5.650656997101205e-06,
+      "loss": 0.7151,
+      "step": 27757
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3141257633582852,
+      "learning_rate": 5.64996992013976e-06,
+      "loss": 0.7447,
+      "step": 27758
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3795600238728531,
+      "learning_rate": 5.6492828685055656e-06,
+      "loss": 0.7388,
+      "step": 27759
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5365367771958993,
+      "learning_rate": 5.64859584220262e-06,
+      "loss": 0.6974,
+      "step": 27760
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5746158444498505,
+      "learning_rate": 5.647908841234927e-06,
+      "loss": 0.4896,
+      "step": 27761
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.235227782606757,
+      "learning_rate": 5.64722186560649e-06,
+      "loss": 0.6275,
+      "step": 27762
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3384305931560525,
+      "learning_rate": 5.6465349153213065e-06,
+      "loss": 0.6828,
+      "step": 27763
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.388561913002585,
+      "learning_rate": 5.6458479903833705e-06,
+      "loss": 0.9573,
+      "step": 27764
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4201027507762838,
+      "learning_rate": 5.645161090796687e-06,
+      "loss": 0.6197,
+      "step": 27765
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4478447030078758,
+      "learning_rate": 5.644474216565257e-06,
+      "loss": 0.8471,
+      "step": 27766
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5444641398578904,
+      "learning_rate": 5.643787367693077e-06,
+      "loss": 0.6442,
+      "step": 27767
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4609316519797737,
+      "learning_rate": 5.643100544184142e-06,
+      "loss": 0.7103,
+      "step": 27768
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5800308375739932,
+      "learning_rate": 5.642413746042457e-06,
+      "loss": 0.4897,
+      "step": 27769
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.378542316472128,
+      "learning_rate": 5.6417269732720204e-06,
+      "loss": 0.6231,
+      "step": 27770
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3800738674461408,
+      "learning_rate": 5.64104022587683e-06,
+      "loss": 0.7375,
+      "step": 27771
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2524388153333337,
+      "learning_rate": 5.64035350386088e-06,
+      "loss": 0.771,
+      "step": 27772
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3445918462356024,
+      "learning_rate": 5.6396668072281745e-06,
+      "loss": 0.5332,
+      "step": 27773
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5902251479015255,
+      "learning_rate": 5.6389801359827075e-06,
+      "loss": 0.6162,
+      "step": 27774
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4459520982720744,
+      "learning_rate": 5.638293490128481e-06,
+      "loss": 0.7578,
+      "step": 27775
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6057999352605488,
+      "learning_rate": 5.6376068696694875e-06,
+      "loss": 0.506,
+      "step": 27776
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.31298791646243,
+      "learning_rate": 5.636920274609731e-06,
+      "loss": 0.6514,
+      "step": 27777
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8312786383209443,
+      "learning_rate": 5.636233704953205e-06,
+      "loss": 0.6815,
+      "step": 27778
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4092125733284584,
+      "learning_rate": 5.635547160703909e-06,
+      "loss": 0.6166,
+      "step": 27779
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.641889372372502,
+      "learning_rate": 5.63486064186584e-06,
+      "loss": 0.7328,
+      "step": 27780
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3025731510697138,
+      "learning_rate": 5.63417414844299e-06,
+      "loss": 0.7062,
+      "step": 27781
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.327525644081442,
+      "learning_rate": 5.633487680439362e-06,
+      "loss": 0.8365,
+      "step": 27782
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.324260821672386,
+      "learning_rate": 5.632801237858954e-06,
+      "loss": 0.7357,
+      "step": 27783
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7147577035680048,
+      "learning_rate": 5.632114820705759e-06,
+      "loss": 0.8842,
+      "step": 27784
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5751241390292086,
+      "learning_rate": 5.6314284289837705e-06,
+      "loss": 0.7732,
+      "step": 27785
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2727249665977554,
+      "learning_rate": 5.630742062696989e-06,
+      "loss": 0.6365,
+      "step": 27786
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.716449397681893,
+      "learning_rate": 5.630055721849414e-06,
+      "loss": 0.6242,
+      "step": 27787
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4023799074609236,
+      "learning_rate": 5.62936940644504e-06,
+      "loss": 0.581,
+      "step": 27788
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4037696821583405,
+      "learning_rate": 5.6286831164878556e-06,
+      "loss": 0.7729,
+      "step": 27789
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4373089371663876,
+      "learning_rate": 5.627996851981863e-06,
+      "loss": 0.6537,
+      "step": 27790
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.411033020720061,
+      "learning_rate": 5.627310612931058e-06,
+      "loss": 0.6358,
+      "step": 27791
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3793871871153522,
+      "learning_rate": 5.626624399339438e-06,
+      "loss": 0.6107,
+      "step": 27792
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3170117672407127,
+      "learning_rate": 5.62593821121099e-06,
+      "loss": 0.5928,
+      "step": 27793
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4976044549642424,
+      "learning_rate": 5.62525204854972e-06,
+      "loss": 0.7494,
+      "step": 27794
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4188654406330945,
+      "learning_rate": 5.624565911359612e-06,
+      "loss": 0.6309,
+      "step": 27795
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5067809322347163,
+      "learning_rate": 5.6238797996446714e-06,
+      "loss": 0.7226,
+      "step": 27796
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5257899851447947,
+      "learning_rate": 5.623193713408883e-06,
+      "loss": 0.6914,
+      "step": 27797
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3161329588760358,
+      "learning_rate": 5.6225076526562525e-06,
+      "loss": 0.7651,
+      "step": 27798
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2610294842766951,
+      "learning_rate": 5.621821617390763e-06,
+      "loss": 0.6055,
+      "step": 27799
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.515535146829408,
+      "learning_rate": 5.621135607616413e-06,
+      "loss": 0.7653,
+      "step": 27800
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.085256175309635,
+      "learning_rate": 5.620449623337208e-06,
+      "loss": 0.6814,
+      "step": 27801
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5355276012478665,
+      "learning_rate": 5.619763664557123e-06,
+      "loss": 0.6673,
+      "step": 27802
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2843527406909045,
+      "learning_rate": 5.61907773128016e-06,
+      "loss": 0.8036,
+      "step": 27803
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2062431580323303,
+      "learning_rate": 5.6183918235103144e-06,
+      "loss": 0.7095,
+      "step": 27804
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3045338749181468,
+      "learning_rate": 5.617705941251586e-06,
+      "loss": 0.7319,
+      "step": 27805
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4331774342207315,
+      "learning_rate": 5.617020084507954e-06,
+      "loss": 0.711,
+      "step": 27806
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4796386716067074,
+      "learning_rate": 5.616334253283418e-06,
+      "loss": 0.6529,
+      "step": 27807
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1302955611410161,
+      "learning_rate": 5.615648447581976e-06,
+      "loss": 0.625,
+      "step": 27808
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4496382434087247,
+      "learning_rate": 5.614962667407613e-06,
+      "loss": 0.8015,
+      "step": 27809
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5785038416263006,
+      "learning_rate": 5.614276912764329e-06,
+      "loss": 0.5335,
+      "step": 27810
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5976708837248987,
+      "learning_rate": 5.61359118365611e-06,
+      "loss": 0.6475,
+      "step": 27811
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4502543733190911,
+      "learning_rate": 5.612905480086954e-06,
+      "loss": 0.6982,
+      "step": 27812
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.437261966028487,
+      "learning_rate": 5.61221980206085e-06,
+      "loss": 0.7676,
+      "step": 27813
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8231788951918169,
+      "learning_rate": 5.611534149581794e-06,
+      "loss": 0.7133,
+      "step": 27814
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6142086421251196,
+      "learning_rate": 5.610848522653776e-06,
+      "loss": 0.577,
+      "step": 27815
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.612574022304928,
+      "learning_rate": 5.610162921280783e-06,
+      "loss": 0.5242,
+      "step": 27816
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.604819460069161,
+      "learning_rate": 5.609477345466811e-06,
+      "loss": 0.7232,
+      "step": 27817
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1716643743178845,
+      "learning_rate": 5.608791795215857e-06,
+      "loss": 0.6772,
+      "step": 27818
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6838527633571003,
+      "learning_rate": 5.608106270531907e-06,
+      "loss": 0.7477,
+      "step": 27819
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5549290139287222,
+      "learning_rate": 5.6074207714189475e-06,
+      "loss": 0.6495,
+      "step": 27820
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.219443667070929,
+      "learning_rate": 5.606735297880977e-06,
+      "loss": 0.6656,
+      "step": 27821
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.530626494260015,
+      "learning_rate": 5.606049849921987e-06,
+      "loss": 0.5782,
+      "step": 27822
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2614742874960345,
+      "learning_rate": 5.6053644275459665e-06,
+      "loss": 0.6634,
+      "step": 27823
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7063694183841283,
+      "learning_rate": 5.604679030756901e-06,
+      "loss": 0.7899,
+      "step": 27824
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4524112749667704,
+      "learning_rate": 5.603993659558786e-06,
+      "loss": 0.8043,
+      "step": 27825
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.099360172382575,
+      "learning_rate": 5.603308313955617e-06,
+      "loss": 0.7421,
+      "step": 27826
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4097583305109425,
+      "learning_rate": 5.602622993951377e-06,
+      "loss": 0.846,
+      "step": 27827
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5683521445520834,
+      "learning_rate": 5.601937699550055e-06,
+      "loss": 0.6627,
+      "step": 27828
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.586199372793028,
+      "learning_rate": 5.601252430755648e-06,
+      "loss": 0.7034,
+      "step": 27829
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.246877610521053,
+      "learning_rate": 5.600567187572139e-06,
+      "loss": 0.7523,
+      "step": 27830
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.581342341953432,
+      "learning_rate": 5.599881970003522e-06,
+      "loss": 0.6335,
+      "step": 27831
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3182107980439501,
+      "learning_rate": 5.599196778053784e-06,
+      "loss": 0.6424,
+      "step": 27832
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.152486931489305,
+      "learning_rate": 5.5985116117269175e-06,
+      "loss": 0.6951,
+      "step": 27833
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7986139615397554,
+      "learning_rate": 5.5978264710269085e-06,
+      "loss": 0.7276,
+      "step": 27834
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5446654431361422,
+      "learning_rate": 5.597141355957749e-06,
+      "loss": 0.8004,
+      "step": 27835
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7652591934639557,
+      "learning_rate": 5.5964562665234275e-06,
+      "loss": 0.7399,
+      "step": 27836
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5229707839361781,
+      "learning_rate": 5.595771202727928e-06,
+      "loss": 0.6948,
+      "step": 27837
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3887227124407397,
+      "learning_rate": 5.595086164575243e-06,
+      "loss": 0.756,
+      "step": 27838
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5926072309331576,
+      "learning_rate": 5.5944011520693645e-06,
+      "loss": 0.859,
+      "step": 27839
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8827248621214439,
+      "learning_rate": 5.593716165214276e-06,
+      "loss": 0.757,
+      "step": 27840
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.9753638617998055,
+      "learning_rate": 5.593031204013965e-06,
+      "loss": 0.633,
+      "step": 27841
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.648549439239514,
+      "learning_rate": 5.59234626847242e-06,
+      "loss": 0.6956,
+      "step": 27842
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7814933644575086,
+      "learning_rate": 5.591661358593636e-06,
+      "loss": 0.7099,
+      "step": 27843
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.453854499492863,
+      "learning_rate": 5.590976474381595e-06,
+      "loss": 0.6689,
+      "step": 27844
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4872468809467783,
+      "learning_rate": 5.59029161584028e-06,
+      "loss": 0.5898,
+      "step": 27845
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.298852589686506,
+      "learning_rate": 5.589606782973683e-06,
+      "loss": 0.6616,
+      "step": 27846
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3380914013633707,
+      "learning_rate": 5.588921975785797e-06,
+      "loss": 0.6891,
+      "step": 27847
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4041882019217167,
+      "learning_rate": 5.588237194280603e-06,
+      "loss": 0.6996,
+      "step": 27848
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.449921958670881,
+      "learning_rate": 5.5875524384620846e-06,
+      "loss": 0.5857,
+      "step": 27849
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5666344714671301,
+      "learning_rate": 5.586867708334236e-06,
+      "loss": 0.5157,
+      "step": 27850
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6865709385663459,
+      "learning_rate": 5.5861830039010375e-06,
+      "loss": 0.6912,
+      "step": 27851
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3686690952213738,
+      "learning_rate": 5.585498325166484e-06,
+      "loss": 0.7375,
+      "step": 27852
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5927975684458482,
+      "learning_rate": 5.584813672134553e-06,
+      "loss": 0.5613,
+      "step": 27853
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2897077992470758,
+      "learning_rate": 5.584129044809237e-06,
+      "loss": 0.6471,
+      "step": 27854
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6087451912091688,
+      "learning_rate": 5.583444443194517e-06,
+      "loss": 0.7221,
+      "step": 27855
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3441390970641163,
+      "learning_rate": 5.582759867294385e-06,
+      "loss": 0.7587,
+      "step": 27856
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5740382246555708,
+      "learning_rate": 5.5820753171128235e-06,
+      "loss": 0.4783,
+      "step": 27857
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.579021817903338,
+      "learning_rate": 5.581390792653815e-06,
+      "loss": 0.6205,
+      "step": 27858
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9262307867609607,
+      "learning_rate": 5.580706293921348e-06,
+      "loss": 0.7705,
+      "step": 27859
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4183310379317484,
+      "learning_rate": 5.580021820919411e-06,
+      "loss": 0.6469,
+      "step": 27860
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.595484293853474,
+      "learning_rate": 5.5793373736519876e-06,
+      "loss": 0.4875,
+      "step": 27861
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2188387364082978,
+      "learning_rate": 5.578652952123056e-06,
+      "loss": 0.6394,
+      "step": 27862
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3296713414050265,
+      "learning_rate": 5.577968556336608e-06,
+      "loss": 0.661,
+      "step": 27863
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5611234456927118,
+      "learning_rate": 5.577284186296631e-06,
+      "loss": 0.4858,
+      "step": 27864
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5934978993677227,
+      "learning_rate": 5.576599842007105e-06,
+      "loss": 0.5223,
+      "step": 27865
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6807489446610673,
+      "learning_rate": 5.575915523472012e-06,
+      "loss": 0.6925,
+      "step": 27866
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3814601153656823,
+      "learning_rate": 5.575231230695339e-06,
+      "loss": 0.7321,
+      "step": 27867
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0023914797321245,
+      "learning_rate": 5.5745469636810746e-06,
+      "loss": 0.7596,
+      "step": 27868
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2940914022212586,
+      "learning_rate": 5.573862722433199e-06,
+      "loss": 0.6786,
+      "step": 27869
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4013775764783867,
+      "learning_rate": 5.573178506955692e-06,
+      "loss": 0.676,
+      "step": 27870
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3124087323921496,
+      "learning_rate": 5.572494317252545e-06,
+      "loss": 0.8253,
+      "step": 27871
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.448524126982705,
+      "learning_rate": 5.571810153327734e-06,
+      "loss": 0.7309,
+      "step": 27872
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3691536810864797,
+      "learning_rate": 5.5711260151852494e-06,
+      "loss": 0.6472,
+      "step": 27873
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2460762610990739,
+      "learning_rate": 5.570441902829069e-06,
+      "loss": 0.585,
+      "step": 27874
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2549297008006453,
+      "learning_rate": 5.569757816263181e-06,
+      "loss": 0.5877,
+      "step": 27875
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.029534403994797,
+      "learning_rate": 5.569073755491561e-06,
+      "loss": 0.75,
+      "step": 27876
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5178686406665742,
+      "learning_rate": 5.568389720518201e-06,
+      "loss": 0.661,
+      "step": 27877
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3401355744914825,
+      "learning_rate": 5.567705711347075e-06,
+      "loss": 0.6269,
+      "step": 27878
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0338448003741134,
+      "learning_rate": 5.5670217279821745e-06,
+      "loss": 0.6643,
+      "step": 27879
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.52184840601331,
+      "learning_rate": 5.566337770427472e-06,
+      "loss": 0.6663,
+      "step": 27880
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0429671518783286,
+      "learning_rate": 5.565653838686959e-06,
+      "loss": 0.7838,
+      "step": 27881
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.38959214291426,
+      "learning_rate": 5.564969932764615e-06,
+      "loss": 0.7497,
+      "step": 27882
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8555354501553005,
+      "learning_rate": 5.5642860526644135e-06,
+      "loss": 0.6947,
+      "step": 27883
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7295125325810645,
+      "learning_rate": 5.563602198390343e-06,
+      "loss": 0.7767,
+      "step": 27884
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7108768995444454,
+      "learning_rate": 5.562918369946386e-06,
+      "loss": 0.7747,
+      "step": 27885
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.8939638154660945,
+      "learning_rate": 5.562234567336531e-06,
+      "loss": 0.6855,
+      "step": 27886
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5328479008817444,
+      "learning_rate": 5.561550790564744e-06,
+      "loss": 0.7027,
+      "step": 27887
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5957399744986702,
+      "learning_rate": 5.560867039635012e-06,
+      "loss": 0.7201,
+      "step": 27888
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4273121100157937,
+      "learning_rate": 5.560183314551323e-06,
+      "loss": 0.7676,
+      "step": 27889
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4930187904350678,
+      "learning_rate": 5.559499615317652e-06,
+      "loss": 0.7322,
+      "step": 27890
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4123410396325753,
+      "learning_rate": 5.558815941937975e-06,
+      "loss": 0.6401,
+      "step": 27891
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5233467402458898,
+      "learning_rate": 5.558132294416278e-06,
+      "loss": 0.6772,
+      "step": 27892
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3661427385353022,
+      "learning_rate": 5.557448672756546e-06,
+      "loss": 0.7129,
+      "step": 27893
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3230746356587324,
+      "learning_rate": 5.556765076962749e-06,
+      "loss": 0.6748,
+      "step": 27894
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6042361438984802,
+      "learning_rate": 5.556081507038877e-06,
+      "loss": 0.6412,
+      "step": 27895
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.333245273666174,
+      "learning_rate": 5.555397962988905e-06,
+      "loss": 0.6784,
+      "step": 27896
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2571992274997794,
+      "learning_rate": 5.5547144448168115e-06,
+      "loss": 0.6455,
+      "step": 27897
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4147366833075974,
+      "learning_rate": 5.554030952526574e-06,
+      "loss": 0.8094,
+      "step": 27898
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3925000949314124,
+      "learning_rate": 5.553347486122182e-06,
+      "loss": 0.6864,
+      "step": 27899
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3853748988256516,
+      "learning_rate": 5.55266404560761e-06,
+      "loss": 0.8025,
+      "step": 27900
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6445061531450361,
+      "learning_rate": 5.5519806309868304e-06,
+      "loss": 0.5244,
+      "step": 27901
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4083854875738266,
+      "learning_rate": 5.551297242263827e-06,
+      "loss": 0.6305,
+      "step": 27902
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6727684478995128,
+      "learning_rate": 5.550613879442589e-06,
+      "loss": 0.6065,
+      "step": 27903
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3252849652850351,
+      "learning_rate": 5.5499305425270765e-06,
+      "loss": 0.644,
+      "step": 27904
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8620648163976221,
+      "learning_rate": 5.549247231521278e-06,
+      "loss": 0.6631,
+      "step": 27905
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2779055642295154,
+      "learning_rate": 5.54856394642917e-06,
+      "loss": 0.709,
+      "step": 27906
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5090184206378174,
+      "learning_rate": 5.547880687254739e-06,
+      "loss": 0.5225,
+      "step": 27907
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6510047637469418,
+      "learning_rate": 5.5471974540019536e-06,
+      "loss": 0.7424,
+      "step": 27908
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5941143963670266,
+      "learning_rate": 5.54651424667479e-06,
+      "loss": 0.5071,
+      "step": 27909
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4309330840596608,
+      "learning_rate": 5.545831065277236e-06,
+      "loss": 0.6158,
+      "step": 27910
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2118944356026056,
+      "learning_rate": 5.545147909813259e-06,
+      "loss": 0.7516,
+      "step": 27911
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3547634578166914,
+      "learning_rate": 5.544464780286845e-06,
+      "loss": 0.7445,
+      "step": 27912
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3087883774075384,
+      "learning_rate": 5.543781676701964e-06,
+      "loss": 0.7542,
+      "step": 27913
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5537827054262305,
+      "learning_rate": 5.543098599062602e-06,
+      "loss": 0.5268,
+      "step": 27914
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6778897220670548,
+      "learning_rate": 5.542415547372724e-06,
+      "loss": 0.6484,
+      "step": 27915
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.523826182109656,
+      "learning_rate": 5.54173252163632e-06,
+      "loss": 0.6917,
+      "step": 27916
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3591405841051387,
+      "learning_rate": 5.541049521857361e-06,
+      "loss": 0.6025,
+      "step": 27917
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5236499469939047,
+      "learning_rate": 5.5403665480398196e-06,
+      "loss": 0.645,
+      "step": 27918
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.330139984474037,
+      "learning_rate": 5.539683600187675e-06,
+      "loss": 0.6499,
+      "step": 27919
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3410707702964073,
+      "learning_rate": 5.539000678304909e-06,
+      "loss": 0.7185,
+      "step": 27920
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2694547660075972,
+      "learning_rate": 5.538317782395493e-06,
+      "loss": 0.685,
+      "step": 27921
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6128596940724786,
+      "learning_rate": 5.537634912463399e-06,
+      "loss": 0.4991,
+      "step": 27922
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3335092099320203,
+      "learning_rate": 5.536952068512609e-06,
+      "loss": 0.7582,
+      "step": 27923
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2871257497117792,
+      "learning_rate": 5.536269250547099e-06,
+      "loss": 0.4964,
+      "step": 27924
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5851826131768941,
+      "learning_rate": 5.535586458570844e-06,
+      "loss": 0.4918,
+      "step": 27925
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.377528613367255,
+      "learning_rate": 5.534903692587814e-06,
+      "loss": 0.6506,
+      "step": 27926
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4024408776557726,
+      "learning_rate": 5.534220952601987e-06,
+      "loss": 0.6932,
+      "step": 27927
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5881627747987878,
+      "learning_rate": 5.533538238617345e-06,
+      "loss": 0.4914,
+      "step": 27928
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4184766569540312,
+      "learning_rate": 5.532855550637856e-06,
+      "loss": 0.5845,
+      "step": 27929
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3374234908276725,
+      "learning_rate": 5.532172888667494e-06,
+      "loss": 0.7327,
+      "step": 27930
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7594924556645488,
+      "learning_rate": 5.531490252710239e-06,
+      "loss": 0.772,
+      "step": 27931
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3179855843415529,
+      "learning_rate": 5.53080764277006e-06,
+      "loss": 0.6146,
+      "step": 27932
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4653643610107896,
+      "learning_rate": 5.530125058850935e-06,
+      "loss": 0.6977,
+      "step": 27933
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5899463809480198,
+      "learning_rate": 5.529442500956834e-06,
+      "loss": 0.5445,
+      "step": 27934
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3437036685238744,
+      "learning_rate": 5.528759969091739e-06,
+      "loss": 0.7442,
+      "step": 27935
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6103056314709819,
+      "learning_rate": 5.5280774632596145e-06,
+      "loss": 0.491,
+      "step": 27936
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4070750837333883,
+      "learning_rate": 5.5273949834644425e-06,
+      "loss": 0.7627,
+      "step": 27937
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.207954222362879,
+      "learning_rate": 5.526712529710194e-06,
+      "loss": 0.6115,
+      "step": 27938
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3110101133024594,
+      "learning_rate": 5.526030102000836e-06,
+      "loss": 0.6851,
+      "step": 27939
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.594252547480309,
+      "learning_rate": 5.525347700340349e-06,
+      "loss": 0.6906,
+      "step": 27940
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.065528721716439,
+      "learning_rate": 5.524665324732706e-06,
+      "loss": 0.6677,
+      "step": 27941
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9753327404548224,
+      "learning_rate": 5.52398297518188e-06,
+      "loss": 0.7078,
+      "step": 27942
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.425961960810655,
+      "learning_rate": 5.523300651691837e-06,
+      "loss": 0.5486,
+      "step": 27943
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4301183908071828,
+      "learning_rate": 5.522618354266556e-06,
+      "loss": 0.6496,
+      "step": 27944
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1951687460746914,
+      "learning_rate": 5.521936082910012e-06,
+      "loss": 0.5962,
+      "step": 27945
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4699289246844407,
+      "learning_rate": 5.521253837626175e-06,
+      "loss": 0.6778,
+      "step": 27946
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.317387762470875,
+      "learning_rate": 5.5205716184190106e-06,
+      "loss": 0.745,
+      "step": 27947
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3404215157125976,
+      "learning_rate": 5.519889425292496e-06,
+      "loss": 0.7676,
+      "step": 27948
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6026119439825782,
+      "learning_rate": 5.5192072582506086e-06,
+      "loss": 0.5019,
+      "step": 27949
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5622265717309178,
+      "learning_rate": 5.5185251172973155e-06,
+      "loss": 0.812,
+      "step": 27950
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.33694141141314,
+      "learning_rate": 5.517843002436583e-06,
+      "loss": 0.6962,
+      "step": 27951
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3623420304410934,
+      "learning_rate": 5.517160913672393e-06,
+      "loss": 0.617,
+      "step": 27952
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6649272132404986,
+      "learning_rate": 5.516478851008705e-06,
+      "loss": 0.6554,
+      "step": 27953
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4460694396850606,
+      "learning_rate": 5.515796814449502e-06,
+      "loss": 0.614,
+      "step": 27954
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6198231231776647,
+      "learning_rate": 5.515114803998745e-06,
+      "loss": 0.7021,
+      "step": 27955
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4463579223312255,
+      "learning_rate": 5.5144328196604145e-06,
+      "loss": 0.7291,
+      "step": 27956
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9603527647051848,
+      "learning_rate": 5.513750861438472e-06,
+      "loss": 0.5842,
+      "step": 27957
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4050519595899433,
+      "learning_rate": 5.5130689293368955e-06,
+      "loss": 0.8182,
+      "step": 27958
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3790998318969196,
+      "learning_rate": 5.5123870233596536e-06,
+      "loss": 0.6881,
+      "step": 27959
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3676969522261144,
+      "learning_rate": 5.511705143510711e-06,
+      "loss": 0.7337,
+      "step": 27960
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2554529244465542,
+      "learning_rate": 5.51102328979404e-06,
+      "loss": 0.6682,
+      "step": 27961
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.398412118846575,
+      "learning_rate": 5.510341462213619e-06,
+      "loss": 0.6182,
+      "step": 27962
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6036967284410577,
+      "learning_rate": 5.50965966077341e-06,
+      "loss": 0.7361,
+      "step": 27963
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3594917691976796,
+      "learning_rate": 5.50897788547738e-06,
+      "loss": 0.607,
+      "step": 27964
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.589820574163001,
+      "learning_rate": 5.508296136329504e-06,
+      "loss": 0.6129,
+      "step": 27965
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4711708405453425,
+      "learning_rate": 5.507614413333753e-06,
+      "loss": 0.6942,
+      "step": 27966
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.452559218693446,
+      "learning_rate": 5.506932716494093e-06,
+      "loss": 0.7035,
+      "step": 27967
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.57458554168875,
+      "learning_rate": 5.50625104581449e-06,
+      "loss": 0.53,
+      "step": 27968
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6609816153933286,
+      "learning_rate": 5.505569401298916e-06,
+      "loss": 0.6904,
+      "step": 27969
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3823944974507132,
+      "learning_rate": 5.504887782951343e-06,
+      "loss": 0.6294,
+      "step": 27970
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0059652205963467,
+      "learning_rate": 5.504206190775736e-06,
+      "loss": 0.66,
+      "step": 27971
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3736000528972023,
+      "learning_rate": 5.503524624776061e-06,
+      "loss": 0.7085,
+      "step": 27972
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.038495429688477,
+      "learning_rate": 5.502843084956293e-06,
+      "loss": 0.7834,
+      "step": 27973
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2205425360191209,
+      "learning_rate": 5.502161571320393e-06,
+      "loss": 0.7183,
+      "step": 27974
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3711788073289928,
+      "learning_rate": 5.501480083872335e-06,
+      "loss": 0.6545,
+      "step": 27975
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.354157677093437,
+      "learning_rate": 5.500798622616081e-06,
+      "loss": 0.8012,
+      "step": 27976
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.232326874830056,
+      "learning_rate": 5.500117187555604e-06,
+      "loss": 0.6796,
+      "step": 27977
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3122798894079255,
+      "learning_rate": 5.499435778694868e-06,
+      "loss": 0.7166,
+      "step": 27978
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5723472149038002,
+      "learning_rate": 5.4987543960378445e-06,
+      "loss": 0.53,
+      "step": 27979
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4423129936641872,
+      "learning_rate": 5.498073039588498e-06,
+      "loss": 0.7505,
+      "step": 27980
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4932978511478523,
+      "learning_rate": 5.497391709350792e-06,
+      "loss": 0.6373,
+      "step": 27981
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3429897552911962,
+      "learning_rate": 5.496710405328698e-06,
+      "loss": 0.5876,
+      "step": 27982
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5279679785674523,
+      "learning_rate": 5.496029127526179e-06,
+      "loss": 0.7548,
+      "step": 27983
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5015053176658328,
+      "learning_rate": 5.495347875947215e-06,
+      "loss": 0.7133,
+      "step": 27984
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5777018125233113,
+      "learning_rate": 5.494666650595752e-06,
+      "loss": 0.638,
+      "step": 27985
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.157547841097106,
+      "learning_rate": 5.4939854514757675e-06,
+      "loss": 0.6973,
+      "step": 27986
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3258929232470478,
+      "learning_rate": 5.493304278591229e-06,
+      "loss": 0.7406,
+      "step": 27987
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6239153551630339,
+      "learning_rate": 5.4926231319461e-06,
+      "loss": 0.6949,
+      "step": 27988
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3200202071264595,
+      "learning_rate": 5.491942011544342e-06,
+      "loss": 0.7665,
+      "step": 27989
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6138597972367663,
+      "learning_rate": 5.4912609173899255e-06,
+      "loss": 0.714,
+      "step": 27990
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4090109511364244,
+      "learning_rate": 5.490579849486819e-06,
+      "loss": 0.6539,
+      "step": 27991
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3211915171646833,
+      "learning_rate": 5.48989880783898e-06,
+      "loss": 0.6523,
+      "step": 27992
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5944909734644509,
+      "learning_rate": 5.489217792450383e-06,
+      "loss": 0.5177,
+      "step": 27993
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6717320538561726,
+      "learning_rate": 5.488536803324989e-06,
+      "loss": 0.7815,
+      "step": 27994
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3732122299376888,
+      "learning_rate": 5.487855840466756e-06,
+      "loss": 0.6858,
+      "step": 27995
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4857773292969627,
+      "learning_rate": 5.4871749038796554e-06,
+      "loss": 0.7079,
+      "step": 27996
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.289227987227831,
+      "learning_rate": 5.486493993567654e-06,
+      "loss": 0.6617,
+      "step": 27997
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5943132826198443,
+      "learning_rate": 5.485813109534716e-06,
+      "loss": 0.5351,
+      "step": 27998
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.871441498118176,
+      "learning_rate": 5.485132251784797e-06,
+      "loss": 0.7018,
+      "step": 27999
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3428339203412543,
+      "learning_rate": 5.484451420321869e-06,
+      "loss": 0.7391,
+      "step": 28000
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3178430297779606,
+      "learning_rate": 5.483770615149898e-06,
+      "loss": 0.7052,
+      "step": 28001
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5858385951289171,
+      "learning_rate": 5.483089836272844e-06,
+      "loss": 0.7063,
+      "step": 28002
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4028354554942821,
+      "learning_rate": 5.482409083694667e-06,
+      "loss": 0.709,
+      "step": 28003
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6188062187727389,
+      "learning_rate": 5.481728357419336e-06,
+      "loss": 0.5314,
+      "step": 28004
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3657035311656154,
+      "learning_rate": 5.481047657450816e-06,
+      "loss": 0.6894,
+      "step": 28005
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5730448022448922,
+      "learning_rate": 5.480366983793066e-06,
+      "loss": 0.8793,
+      "step": 28006
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.543496552074209,
+      "learning_rate": 5.479686336450047e-06,
+      "loss": 0.7441,
+      "step": 28007
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5100685904858935,
+      "learning_rate": 5.479005715425725e-06,
+      "loss": 0.697,
+      "step": 28008
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3958929643150726,
+      "learning_rate": 5.4783251207240685e-06,
+      "loss": 0.6907,
+      "step": 28009
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5060191262723424,
+      "learning_rate": 5.477644552349033e-06,
+      "loss": 0.8096,
+      "step": 28010
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3923736310476689,
+      "learning_rate": 5.476964010304579e-06,
+      "loss": 0.662,
+      "step": 28011
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5125157706059877,
+      "learning_rate": 5.476283494594677e-06,
+      "loss": 0.7167,
+      "step": 28012
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.597977190124621,
+      "learning_rate": 5.475603005223281e-06,
+      "loss": 0.7112,
+      "step": 28013
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3500222251223188,
+      "learning_rate": 5.474922542194361e-06,
+      "loss": 0.7012,
+      "step": 28014
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3876625646883793,
+      "learning_rate": 5.4742421055118686e-06,
+      "loss": 0.6795,
+      "step": 28015
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.374384120594536,
+      "learning_rate": 5.473561695179777e-06,
+      "loss": 0.7764,
+      "step": 28016
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.598596629476866,
+      "learning_rate": 5.472881311202039e-06,
+      "loss": 0.6794,
+      "step": 28017
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4916781152201712,
+      "learning_rate": 5.472200953582622e-06,
+      "loss": 0.688,
+      "step": 28018
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3217873324995133,
+      "learning_rate": 5.471520622325484e-06,
+      "loss": 0.818,
+      "step": 28019
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3985604335260837,
+      "learning_rate": 5.470840317434584e-06,
+      "loss": 0.7702,
+      "step": 28020
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.330379179481612,
+      "learning_rate": 5.470160038913884e-06,
+      "loss": 0.7032,
+      "step": 28021
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4715654575284598,
+      "learning_rate": 5.469479786767351e-06,
+      "loss": 0.8224,
+      "step": 28022
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3976089691243438,
+      "learning_rate": 5.468799560998942e-06,
+      "loss": 0.6528,
+      "step": 28023
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.392311335261885,
+      "learning_rate": 5.468119361612612e-06,
+      "loss": 0.6611,
+      "step": 28024
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4910449129669274,
+      "learning_rate": 5.467439188612326e-06,
+      "loss": 0.7718,
+      "step": 28025
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4812610057426063,
+      "learning_rate": 5.4667590420020475e-06,
+      "loss": 0.6698,
+      "step": 28026
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3908231914227072,
+      "learning_rate": 5.466078921785733e-06,
+      "loss": 0.6539,
+      "step": 28027
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2848093335329347,
+      "learning_rate": 5.465398827967339e-06,
+      "loss": 0.6237,
+      "step": 28028
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5919488629888725,
+      "learning_rate": 5.464718760550829e-06,
+      "loss": 0.7095,
+      "step": 28029
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9441002434328831,
+      "learning_rate": 5.464038719540166e-06,
+      "loss": 0.8372,
+      "step": 28030
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5909810154557384,
+      "learning_rate": 5.463358704939305e-06,
+      "loss": 0.5117,
+      "step": 28031
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3976247002674032,
+      "learning_rate": 5.462678716752201e-06,
+      "loss": 0.7087,
+      "step": 28032
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.974805721907066,
+      "learning_rate": 5.461998754982825e-06,
+      "loss": 0.6646,
+      "step": 28033
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.510258793418277,
+      "learning_rate": 5.461318819635122e-06,
+      "loss": 0.6898,
+      "step": 28034
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4945798424392311,
+      "learning_rate": 5.4606389107130634e-06,
+      "loss": 0.8147,
+      "step": 28035
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3713094326507667,
+      "learning_rate": 5.459959028220597e-06,
+      "loss": 0.5238,
+      "step": 28036
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6100477989923714,
+      "learning_rate": 5.459279172161691e-06,
+      "loss": 0.5336,
+      "step": 28037
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3211749405089117,
+      "learning_rate": 5.458599342540296e-06,
+      "loss": 0.7552,
+      "step": 28038
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2572529096774834,
+      "learning_rate": 5.457919539360377e-06,
+      "loss": 0.6827,
+      "step": 28039
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1693430282648738,
+      "learning_rate": 5.457239762625888e-06,
+      "loss": 0.6234,
+      "step": 28040
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.284513697251861,
+      "learning_rate": 5.456560012340784e-06,
+      "loss": 0.6219,
+      "step": 28041
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3999958872664013,
+      "learning_rate": 5.4558802885090255e-06,
+      "loss": 0.7374,
+      "step": 28042
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.276474272049803,
+      "learning_rate": 5.455200591134574e-06,
+      "loss": 0.7551,
+      "step": 28043
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2837753052700631,
+      "learning_rate": 5.4545209202213845e-06,
+      "loss": 0.6416,
+      "step": 28044
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.538286856913804,
+      "learning_rate": 5.453841275773408e-06,
+      "loss": 0.7438,
+      "step": 28045
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3634704160301603,
+      "learning_rate": 5.453161657794609e-06,
+      "loss": 0.7099,
+      "step": 28046
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3599454530907253,
+      "learning_rate": 5.452482066288945e-06,
+      "loss": 0.8886,
+      "step": 28047
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3532877764963338,
+      "learning_rate": 5.45180250126037e-06,
+      "loss": 0.751,
+      "step": 28048
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.362579981400417,
+      "learning_rate": 5.451122962712838e-06,
+      "loss": 0.7106,
+      "step": 28049
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6154759947996785,
+      "learning_rate": 5.450443450650306e-06,
+      "loss": 0.4965,
+      "step": 28050
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3996921441745258,
+      "learning_rate": 5.449763965076739e-06,
+      "loss": 0.5405,
+      "step": 28051
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3778197755097177,
+      "learning_rate": 5.449084505996085e-06,
+      "loss": 0.6342,
+      "step": 28052
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.478984370231298,
+      "learning_rate": 5.4484050734122995e-06,
+      "loss": 0.6008,
+      "step": 28053
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3507893402514124,
+      "learning_rate": 5.447725667329343e-06,
+      "loss": 0.7281,
+      "step": 28054
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9422293271576256,
+      "learning_rate": 5.447046287751165e-06,
+      "loss": 0.6803,
+      "step": 28055
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.345443903460922,
+      "learning_rate": 5.446366934681731e-06,
+      "loss": 0.6064,
+      "step": 28056
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.392650665756545,
+      "learning_rate": 5.445687608124984e-06,
+      "loss": 0.7629,
+      "step": 28057
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2662328038851502,
+      "learning_rate": 5.445008308084892e-06,
+      "loss": 0.7315,
+      "step": 28058
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2576776407265542,
+      "learning_rate": 5.444329034565397e-06,
+      "loss": 0.626,
+      "step": 28059
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.434354405429624,
+      "learning_rate": 5.443649787570467e-06,
+      "loss": 0.7096,
+      "step": 28060
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4248683319917612,
+      "learning_rate": 5.4429705671040486e-06,
+      "loss": 0.766,
+      "step": 28061
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2702996215016837,
+      "learning_rate": 5.442291373170095e-06,
+      "loss": 0.7508,
+      "step": 28062
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4459105384635709,
+      "learning_rate": 5.441612205772564e-06,
+      "loss": 0.6953,
+      "step": 28063
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5901739853229335,
+      "learning_rate": 5.440933064915414e-06,
+      "loss": 0.6953,
+      "step": 28064
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4385058342536388,
+      "learning_rate": 5.440253950602594e-06,
+      "loss": 0.7017,
+      "step": 28065
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2810017976647396,
+      "learning_rate": 5.439574862838057e-06,
+      "loss": 0.7112,
+      "step": 28066
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2884527217939612,
+      "learning_rate": 5.438895801625758e-06,
+      "loss": 0.6968,
+      "step": 28067
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.268544188355331,
+      "learning_rate": 5.438216766969655e-06,
+      "loss": 0.6804,
+      "step": 28068
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5670757639519608,
+      "learning_rate": 5.437537758873699e-06,
+      "loss": 0.5223,
+      "step": 28069
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6555636098530142,
+      "learning_rate": 5.436858777341839e-06,
+      "loss": 0.5396,
+      "step": 28070
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1816884582169185,
+      "learning_rate": 5.436179822378031e-06,
+      "loss": 0.7095,
+      "step": 28071
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6409230399867862,
+      "learning_rate": 5.435500893986234e-06,
+      "loss": 0.7444,
+      "step": 28072
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.608957268788176,
+      "learning_rate": 5.434821992170396e-06,
+      "loss": 0.6234,
+      "step": 28073
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4209451361115206,
+      "learning_rate": 5.434143116934465e-06,
+      "loss": 0.6303,
+      "step": 28074
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4019942735076816,
+      "learning_rate": 5.433464268282401e-06,
+      "loss": 0.597,
+      "step": 28075
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3523315968505774,
+      "learning_rate": 5.4327854462181515e-06,
+      "loss": 0.6158,
+      "step": 28076
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5426465493308892,
+      "learning_rate": 5.432106650745675e-06,
+      "loss": 0.6965,
+      "step": 28077
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3690267035190347,
+      "learning_rate": 5.431427881868914e-06,
+      "loss": 0.7378,
+      "step": 28078
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6738844155719876,
+      "learning_rate": 5.430749139591833e-06,
+      "loss": 0.786,
+      "step": 28079
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5074684471447506,
+      "learning_rate": 5.430070423918371e-06,
+      "loss": 0.7432,
+      "step": 28080
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4620671098498137,
+      "learning_rate": 5.429391734852487e-06,
+      "loss": 0.7009,
+      "step": 28081
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3881010650286087,
+      "learning_rate": 5.428713072398139e-06,
+      "loss": 0.7083,
+      "step": 28082
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.34887613390341,
+      "learning_rate": 5.4280344365592615e-06,
+      "loss": 0.6657,
+      "step": 28083
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8036958484500647,
+      "learning_rate": 5.427355827339818e-06,
+      "loss": 0.6526,
+      "step": 28084
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6283378900407919,
+      "learning_rate": 5.4266772447437535e-06,
+      "loss": 0.6333,
+      "step": 28085
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2118425387060938,
+      "learning_rate": 5.4259986887750315e-06,
+      "loss": 0.566,
+      "step": 28086
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4864718381168858,
+      "learning_rate": 5.4253201594375835e-06,
+      "loss": 0.6678,
+      "step": 28087
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4535956275964823,
+      "learning_rate": 5.424641656735372e-06,
+      "loss": 0.6412,
+      "step": 28088
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3918904080768315,
+      "learning_rate": 5.423963180672349e-06,
+      "loss": 0.7912,
+      "step": 28089
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3484578687501738,
+      "learning_rate": 5.4232847312524565e-06,
+      "loss": 0.8419,
+      "step": 28090
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6132029067762068,
+      "learning_rate": 5.4226063084796536e-06,
+      "loss": 0.6863,
+      "step": 28091
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3378167445147706,
+      "learning_rate": 5.421927912357881e-06,
+      "loss": 0.5829,
+      "step": 28092
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4173707092200782,
+      "learning_rate": 5.421249542891098e-06,
+      "loss": 0.8076,
+      "step": 28093
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7514595052793558,
+      "learning_rate": 5.420571200083246e-06,
+      "loss": 0.6812,
+      "step": 28094
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3635522266878757,
+      "learning_rate": 5.419892883938282e-06,
+      "loss": 0.5898,
+      "step": 28095
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2984749626679597,
+      "learning_rate": 5.419214594460152e-06,
+      "loss": 0.7528,
+      "step": 28096
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.229270686037722,
+      "learning_rate": 5.418536331652802e-06,
+      "loss": 0.6271,
+      "step": 28097
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5870727818878247,
+      "learning_rate": 5.417858095520182e-06,
+      "loss": 0.5019,
+      "step": 28098
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4243147522661224,
+      "learning_rate": 5.417179886066249e-06,
+      "loss": 0.721,
+      "step": 28099
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5345204886214956,
+      "learning_rate": 5.416501703294943e-06,
+      "loss": 0.63,
+      "step": 28100
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6372276086791501,
+      "learning_rate": 5.415823547210213e-06,
+      "loss": 0.543,
+      "step": 28101
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.821123554182305,
+      "learning_rate": 5.415145417816008e-06,
+      "loss": 0.6634,
+      "step": 28102
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5875420517786798,
+      "learning_rate": 5.414467315116283e-06,
+      "loss": 0.7692,
+      "step": 28103
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2185387364641986,
+      "learning_rate": 5.4137892391149815e-06,
+      "loss": 0.5397,
+      "step": 28104
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3500323824124179,
+      "learning_rate": 5.413111189816047e-06,
+      "loss": 0.7925,
+      "step": 28105
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3618949563667853,
+      "learning_rate": 5.41243316722343e-06,
+      "loss": 0.7268,
+      "step": 28106
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3846703105930318,
+      "learning_rate": 5.411755171341085e-06,
+      "loss": 0.6903,
+      "step": 28107
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3967093316846786,
+      "learning_rate": 5.411077202172953e-06,
+      "loss": 0.7491,
+      "step": 28108
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3805406597724406,
+      "learning_rate": 5.410399259722978e-06,
+      "loss": 0.8485,
+      "step": 28109
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6175293628025313,
+      "learning_rate": 5.409721343995116e-06,
+      "loss": 0.5579,
+      "step": 28110
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.866130579016354,
+      "learning_rate": 5.409043454993307e-06,
+      "loss": 0.5948,
+      "step": 28111
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6616396684362267,
+      "learning_rate": 5.408365592721503e-06,
+      "loss": 0.7412,
+      "step": 28112
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2898437843475823,
+      "learning_rate": 5.407687757183645e-06,
+      "loss": 0.7722,
+      "step": 28113
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2855477151395094,
+      "learning_rate": 5.407009948383686e-06,
+      "loss": 0.658,
+      "step": 28114
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5453516922531636,
+      "learning_rate": 5.406332166325565e-06,
+      "loss": 0.6548,
+      "step": 28115
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5583322446700698,
+      "learning_rate": 5.405654411013237e-06,
+      "loss": 0.741,
+      "step": 28116
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3466446976709852,
+      "learning_rate": 5.404976682450644e-06,
+      "loss": 0.7507,
+      "step": 28117
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.69223988776836,
+      "learning_rate": 5.404298980641727e-06,
+      "loss": 0.6518,
+      "step": 28118
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.638248953531669,
+      "learning_rate": 5.403621305590436e-06,
+      "loss": 0.794,
+      "step": 28119
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.024020903555593,
+      "learning_rate": 5.402943657300722e-06,
+      "loss": 0.6526,
+      "step": 28120
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.319387080569785,
+      "learning_rate": 5.402266035776525e-06,
+      "loss": 0.7332,
+      "step": 28121
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9081344290748157,
+      "learning_rate": 5.401588441021786e-06,
+      "loss": 0.6847,
+      "step": 28122
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.421209813543306,
+      "learning_rate": 5.400910873040456e-06,
+      "loss": 0.6654,
+      "step": 28123
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.387748099640769,
+      "learning_rate": 5.400233331836482e-06,
+      "loss": 0.6932,
+      "step": 28124
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4358330249336637,
+      "learning_rate": 5.399555817413806e-06,
+      "loss": 0.6725,
+      "step": 28125
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2875988446693267,
+      "learning_rate": 5.398878329776369e-06,
+      "loss": 0.6421,
+      "step": 28126
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6352050724978973,
+      "learning_rate": 5.398200868928118e-06,
+      "loss": 0.6733,
+      "step": 28127
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3297440154354856,
+      "learning_rate": 5.397523434873004e-06,
+      "loss": 0.6668,
+      "step": 28128
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.282776504549256,
+      "learning_rate": 5.3968460276149645e-06,
+      "loss": 0.6806,
+      "step": 28129
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6676129013726266,
+      "learning_rate": 5.396168647157942e-06,
+      "loss": 0.6926,
+      "step": 28130
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3608952685249605,
+      "learning_rate": 5.395491293505881e-06,
+      "loss": 0.6787,
+      "step": 28131
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.651916566107379,
+      "learning_rate": 5.3948139666627335e-06,
+      "loss": 0.7267,
+      "step": 28132
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7284688824392354,
+      "learning_rate": 5.394136666632437e-06,
+      "loss": 0.602,
+      "step": 28133
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4860847331489813,
+      "learning_rate": 5.3934593934189295e-06,
+      "loss": 0.7064,
+      "step": 28134
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3181934895613618,
+      "learning_rate": 5.392782147026166e-06,
+      "loss": 0.5488,
+      "step": 28135
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4918318386425138,
+      "learning_rate": 5.392104927458077e-06,
+      "loss": 0.74,
+      "step": 28136
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5946982016515525,
+      "learning_rate": 5.3914277347186174e-06,
+      "loss": 0.7955,
+      "step": 28137
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5280492967909316,
+      "learning_rate": 5.39075056881172e-06,
+      "loss": 0.5689,
+      "step": 28138
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3106948741479105,
+      "learning_rate": 5.390073429741336e-06,
+      "loss": 0.5906,
+      "step": 28139
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.550496767004136,
+      "learning_rate": 5.389396317511401e-06,
+      "loss": 0.485,
+      "step": 28140
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9272920636351358,
+      "learning_rate": 5.388719232125863e-06,
+      "loss": 0.7147,
+      "step": 28141
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.411493785848107,
+      "learning_rate": 5.388042173588662e-06,
+      "loss": 0.6617,
+      "step": 28142
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.345114125849023,
+      "learning_rate": 5.387365141903735e-06,
+      "loss": 0.6841,
+      "step": 28143
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.66128303908717,
+      "learning_rate": 5.386688137075031e-06,
+      "loss": 0.6617,
+      "step": 28144
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2336365005890129,
+      "learning_rate": 5.38601115910649e-06,
+      "loss": 0.7373,
+      "step": 28145
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5373734955457092,
+      "learning_rate": 5.385334208002055e-06,
+      "loss": 0.7297,
+      "step": 28146
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2274822009243607,
+      "learning_rate": 5.384657283765661e-06,
+      "loss": 0.6743,
+      "step": 28147
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5187842499144162,
+      "learning_rate": 5.383980386401253e-06,
+      "loss": 0.7633,
+      "step": 28148
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1789911916052611,
+      "learning_rate": 5.383303515912777e-06,
+      "loss": 0.6469,
+      "step": 28149
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8261646122870887,
+      "learning_rate": 5.382626672304168e-06,
+      "loss": 0.65,
+      "step": 28150
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3978227976404083,
+      "learning_rate": 5.381949855579365e-06,
+      "loss": 0.7392,
+      "step": 28151
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4905353884035304,
+      "learning_rate": 5.381273065742313e-06,
+      "loss": 0.6888,
+      "step": 28152
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.59329371374641,
+      "learning_rate": 5.380596302796954e-06,
+      "loss": 0.4844,
+      "step": 28153
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.623483439675859,
+      "learning_rate": 5.379919566747227e-06,
+      "loss": 0.7018,
+      "step": 28154
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4357234370127072,
+      "learning_rate": 5.379242857597067e-06,
+      "loss": 0.5525,
+      "step": 28155
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4255752454661952,
+      "learning_rate": 5.378566175350421e-06,
+      "loss": 0.918,
+      "step": 28156
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.355994718150767,
+      "learning_rate": 5.377889520011224e-06,
+      "loss": 0.6439,
+      "step": 28157
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5128214906938233,
+      "learning_rate": 5.377212891583419e-06,
+      "loss": 0.805,
+      "step": 28158
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.602902861442564,
+      "learning_rate": 5.376536290070942e-06,
+      "loss": 0.7026,
+      "step": 28159
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.432708467053694,
+      "learning_rate": 5.375859715477738e-06,
+      "loss": 0.7772,
+      "step": 28160
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5843990933276302,
+      "learning_rate": 5.3751831678077385e-06,
+      "loss": 0.6222,
+      "step": 28161
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3560459483445026,
+      "learning_rate": 5.3745066470648895e-06,
+      "loss": 0.5808,
+      "step": 28162
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2903513395400854,
+      "learning_rate": 5.373830153253129e-06,
+      "loss": 0.7202,
+      "step": 28163
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3174621281096728,
+      "learning_rate": 5.373153686376389e-06,
+      "loss": 0.684,
+      "step": 28164
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2080680773518708,
+      "learning_rate": 5.3724772464386145e-06,
+      "loss": 0.7366,
+      "step": 28165
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2856278572292053,
+      "learning_rate": 5.371800833443741e-06,
+      "loss": 0.617,
+      "step": 28166
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.530106968445339,
+      "learning_rate": 5.371124447395717e-06,
+      "loss": 0.7496,
+      "step": 28167
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.397845108813634,
+      "learning_rate": 5.370448088298463e-06,
+      "loss": 0.7135,
+      "step": 28168
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5535101040339065,
+      "learning_rate": 5.369771756155926e-06,
+      "loss": 0.7927,
+      "step": 28169
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.34892493047768,
+      "learning_rate": 5.3690954509720484e-06,
+      "loss": 0.7848,
+      "step": 28170
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3712875109985383,
+      "learning_rate": 5.368419172750763e-06,
+      "loss": 0.7419,
+      "step": 28171
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3547916583660193,
+      "learning_rate": 5.367742921496004e-06,
+      "loss": 0.8502,
+      "step": 28172
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6058577106773168,
+      "learning_rate": 5.367066697211711e-06,
+      "loss": 0.5236,
+      "step": 28173
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2476345480512119,
+      "learning_rate": 5.366390499901828e-06,
+      "loss": 0.7312,
+      "step": 28174
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7763326581362973,
+      "learning_rate": 5.365714329570286e-06,
+      "loss": 0.5019,
+      "step": 28175
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4198916097529222,
+      "learning_rate": 5.3650381862210185e-06,
+      "loss": 0.591,
+      "step": 28176
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5724932162125151,
+      "learning_rate": 5.364362069857969e-06,
+      "loss": 0.697,
+      "step": 28177
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.306619408247059,
+      "learning_rate": 5.363685980485069e-06,
+      "loss": 0.6541,
+      "step": 28178
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5175622731693277,
+      "learning_rate": 5.363009918106256e-06,
+      "loss": 0.7398,
+      "step": 28179
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3924906482669066,
+      "learning_rate": 5.362333882725471e-06,
+      "loss": 0.7531,
+      "step": 28180
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6284079809907719,
+      "learning_rate": 5.361657874346647e-06,
+      "loss": 0.503,
+      "step": 28181
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3247640870014101,
+      "learning_rate": 5.360981892973716e-06,
+      "loss": 0.6082,
+      "step": 28182
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3316077787652036,
+      "learning_rate": 5.3603059386106145e-06,
+      "loss": 0.6723,
+      "step": 28183
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3927872085683286,
+      "learning_rate": 5.359630011261291e-06,
+      "loss": 0.7745,
+      "step": 28184
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3198136882453617,
+      "learning_rate": 5.3589541109296615e-06,
+      "loss": 0.7195,
+      "step": 28185
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7644167652615936,
+      "learning_rate": 5.358278237619672e-06,
+      "loss": 0.7003,
+      "step": 28186
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.422494599341666,
+      "learning_rate": 5.357602391335255e-06,
+      "loss": 0.6864,
+      "step": 28187
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4601933891898131,
+      "learning_rate": 5.35692657208035e-06,
+      "loss": 0.7058,
+      "step": 28188
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6250429669876201,
+      "learning_rate": 5.356250779858889e-06,
+      "loss": 0.5698,
+      "step": 28189
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7102895153583688,
+      "learning_rate": 5.3555750146748034e-06,
+      "loss": 0.7899,
+      "step": 28190
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3278768689043368,
+      "learning_rate": 5.354899276532033e-06,
+      "loss": 0.6629,
+      "step": 28191
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3892828192674838,
+      "learning_rate": 5.354223565434507e-06,
+      "loss": 0.7807,
+      "step": 28192
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3325734481299043,
+      "learning_rate": 5.353547881386165e-06,
+      "loss": 0.7649,
+      "step": 28193
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.19781496797945,
+      "learning_rate": 5.3528722243909346e-06,
+      "loss": 0.71,
+      "step": 28194
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4707261403658891,
+      "learning_rate": 5.352196594452756e-06,
+      "loss": 0.7747,
+      "step": 28195
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4352641488164197,
+      "learning_rate": 5.351520991575558e-06,
+      "loss": 0.7787,
+      "step": 28196
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5487574605031698,
+      "learning_rate": 5.3508454157632815e-06,
+      "loss": 0.9247,
+      "step": 28197
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6051350931808297,
+      "learning_rate": 5.350169867019853e-06,
+      "loss": 0.5358,
+      "step": 28198
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2804269643640975,
+      "learning_rate": 5.349494345349204e-06,
+      "loss": 0.7657,
+      "step": 28199
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4086362973929893,
+      "learning_rate": 5.348818850755272e-06,
+      "loss": 0.726,
+      "step": 28200
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4620469625735524,
+      "learning_rate": 5.348143383241992e-06,
+      "loss": 0.6983,
+      "step": 28201
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3959106689762795,
+      "learning_rate": 5.347467942813296e-06,
+      "loss": 0.6775,
+      "step": 28202
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3670587736862387,
+      "learning_rate": 5.346792529473108e-06,
+      "loss": 0.7313,
+      "step": 28203
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9558158238989987,
+      "learning_rate": 5.346117143225369e-06,
+      "loss": 0.6794,
+      "step": 28204
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3130191080380065,
+      "learning_rate": 5.3454417840740125e-06,
+      "loss": 0.6816,
+      "step": 28205
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2888155500556513,
+      "learning_rate": 5.344766452022967e-06,
+      "loss": 0.7985,
+      "step": 28206
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4053133271456193,
+      "learning_rate": 5.344091147076162e-06,
+      "loss": 0.7107,
+      "step": 28207
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.403491425934746,
+      "learning_rate": 5.343415869237533e-06,
+      "loss": 0.5757,
+      "step": 28208
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6767224522357491,
+      "learning_rate": 5.342740618511013e-06,
+      "loss": 0.7077,
+      "step": 28209
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5386465804163478,
+      "learning_rate": 5.342065394900533e-06,
+      "loss": 0.8108,
+      "step": 28210
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4610997911564447,
+      "learning_rate": 5.3413901984100195e-06,
+      "loss": 0.472,
+      "step": 28211
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2103612121238554,
+      "learning_rate": 5.340715029043411e-06,
+      "loss": 0.7558,
+      "step": 28212
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3980921552059038,
+      "learning_rate": 5.3400398868046285e-06,
+      "loss": 0.5637,
+      "step": 28213
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3047598734344845,
+      "learning_rate": 5.339364771697615e-06,
+      "loss": 0.7958,
+      "step": 28214
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.264620914537152,
+      "learning_rate": 5.338689683726291e-06,
+      "loss": 0.7273,
+      "step": 28215
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.375918217146469,
+      "learning_rate": 5.338014622894596e-06,
+      "loss": 0.8071,
+      "step": 28216
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4041315492005102,
+      "learning_rate": 5.337339589206452e-06,
+      "loss": 0.7112,
+      "step": 28217
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7853654681070994,
+      "learning_rate": 5.3366645826657956e-06,
+      "loss": 0.6708,
+      "step": 28218
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.609463581656125,
+      "learning_rate": 5.3359896032765545e-06,
+      "loss": 0.7803,
+      "step": 28219
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.020084248740641,
+      "learning_rate": 5.335314651042656e-06,
+      "loss": 0.7407,
+      "step": 28220
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4409248974551763,
+      "learning_rate": 5.334639725968031e-06,
+      "loss": 0.7427,
+      "step": 28221
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.301924285050996,
+      "learning_rate": 5.333964828056615e-06,
+      "loss": 0.6174,
+      "step": 28222
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5523506387169168,
+      "learning_rate": 5.333289957312333e-06,
+      "loss": 0.7106,
+      "step": 28223
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1177046334706682,
+      "learning_rate": 5.332615113739111e-06,
+      "loss": 0.7062,
+      "step": 28224
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3841152821434137,
+      "learning_rate": 5.3319402973408805e-06,
+      "loss": 0.6208,
+      "step": 28225
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4729922534017927,
+      "learning_rate": 5.331265508121577e-06,
+      "loss": 0.6673,
+      "step": 28226
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.338494639489297,
+      "learning_rate": 5.330590746085122e-06,
+      "loss": 0.6528,
+      "step": 28227
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2769437417389646,
+      "learning_rate": 5.329916011235443e-06,
+      "loss": 0.6012,
+      "step": 28228
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5346375483940442,
+      "learning_rate": 5.329241303576471e-06,
+      "loss": 0.7042,
+      "step": 28229
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8365266845059522,
+      "learning_rate": 5.328566623112139e-06,
+      "loss": 0.7413,
+      "step": 28230
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8334651461365947,
+      "learning_rate": 5.327891969846371e-06,
+      "loss": 0.6108,
+      "step": 28231
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1215078066263846,
+      "learning_rate": 5.327217343783091e-06,
+      "loss": 0.6968,
+      "step": 28232
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6218903071875923,
+      "learning_rate": 5.326542744926235e-06,
+      "loss": 0.6566,
+      "step": 28233
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2151232383837625,
+      "learning_rate": 5.325868173279725e-06,
+      "loss": 0.644,
+      "step": 28234
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3563394754322584,
+      "learning_rate": 5.325193628847492e-06,
+      "loss": 0.7136,
+      "step": 28235
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4493861797363932,
+      "learning_rate": 5.324519111633457e-06,
+      "loss": 0.8001,
+      "step": 28236
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4600602334580666,
+      "learning_rate": 5.323844621641556e-06,
+      "loss": 0.7797,
+      "step": 28237
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5465368574430409,
+      "learning_rate": 5.32317015887571e-06,
+      "loss": 0.4999,
+      "step": 28238
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4618833225942809,
+      "learning_rate": 5.322495723339851e-06,
+      "loss": 0.6509,
+      "step": 28239
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5343195177693516,
+      "learning_rate": 5.321821315037903e-06,
+      "loss": 0.7386,
+      "step": 28240
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2811636509147557,
+      "learning_rate": 5.3211469339737885e-06,
+      "loss": 0.6685,
+      "step": 28241
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5145605313851482,
+      "learning_rate": 5.320472580151437e-06,
+      "loss": 0.7585,
+      "step": 28242
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4724213471171896,
+      "learning_rate": 5.31979825357478e-06,
+      "loss": 0.759,
+      "step": 28243
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5882559999238426,
+      "learning_rate": 5.31912395424774e-06,
+      "loss": 0.6239,
+      "step": 28244
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5262985669873244,
+      "learning_rate": 5.318449682174238e-06,
+      "loss": 0.6645,
+      "step": 28245
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.417943810239809,
+      "learning_rate": 5.317775437358203e-06,
+      "loss": 0.703,
+      "step": 28246
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6318439874066686,
+      "learning_rate": 5.317101219803566e-06,
+      "loss": 0.6386,
+      "step": 28247
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.279871732780615,
+      "learning_rate": 5.316427029514251e-06,
+      "loss": 0.7388,
+      "step": 28248
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6638505804156336,
+      "learning_rate": 5.315752866494174e-06,
+      "loss": 0.5966,
+      "step": 28249
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8166205492716423,
+      "learning_rate": 5.315078730747268e-06,
+      "loss": 0.8196,
+      "step": 28250
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.397370661412728,
+      "learning_rate": 5.3144046222774595e-06,
+      "loss": 0.6114,
+      "step": 28251
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.442938514589007,
+      "learning_rate": 5.313730541088671e-06,
+      "loss": 0.7222,
+      "step": 28252
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4190539747343056,
+      "learning_rate": 5.313056487184824e-06,
+      "loss": 0.6672,
+      "step": 28253
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5297596208591442,
+      "learning_rate": 5.3123824605698444e-06,
+      "loss": 0.7441,
+      "step": 28254
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4162022470536542,
+      "learning_rate": 5.311708461247664e-06,
+      "loss": 0.631,
+      "step": 28255
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.629800928285562,
+      "learning_rate": 5.3110344892222e-06,
+      "loss": 0.7993,
+      "step": 28256
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4459812442160753,
+      "learning_rate": 5.3103605444973735e-06,
+      "loss": 0.7866,
+      "step": 28257
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3645864526235154,
+      "learning_rate": 5.309686627077117e-06,
+      "loss": 0.6756,
+      "step": 28258
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3615138896557566,
+      "learning_rate": 5.3090127369653465e-06,
+      "loss": 0.743,
+      "step": 28259
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.617315514212919,
+      "learning_rate": 5.308338874165993e-06,
+      "loss": 0.8008,
+      "step": 28260
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5962541245868147,
+      "learning_rate": 5.307665038682972e-06,
+      "loss": 0.4696,
+      "step": 28261
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.348756671349856,
+      "learning_rate": 5.306991230520213e-06,
+      "loss": 0.7043,
+      "step": 28262
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5024914211142242,
+      "learning_rate": 5.306317449681635e-06,
+      "loss": 0.8388,
+      "step": 28263
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6650595442953868,
+      "learning_rate": 5.305643696171161e-06,
+      "loss": 0.6255,
+      "step": 28264
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6179840492948238,
+      "learning_rate": 5.304969969992725e-06,
+      "loss": 0.4915,
+      "step": 28265
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.429126406492183,
+      "learning_rate": 5.304296271150231e-06,
+      "loss": 0.647,
+      "step": 28266
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.275299832587262,
+      "learning_rate": 5.303622599647613e-06,
+      "loss": 0.7615,
+      "step": 28267
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.810042740243931,
+      "learning_rate": 5.302948955488789e-06,
+      "loss": 0.6903,
+      "step": 28268
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4411175825735718,
+      "learning_rate": 5.302275338677692e-06,
+      "loss": 0.7338,
+      "step": 28269
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7600436127590668,
+      "learning_rate": 5.301601749218227e-06,
+      "loss": 0.7846,
+      "step": 28270
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.9225947532241014,
+      "learning_rate": 5.300928187114323e-06,
+      "loss": 0.6604,
+      "step": 28271
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4830280301030523,
+      "learning_rate": 5.30025465236991e-06,
+      "loss": 0.6998,
+      "step": 28272
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.248696822310044,
+      "learning_rate": 5.299581144988899e-06,
+      "loss": 0.6421,
+      "step": 28273
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2362414677322964,
+      "learning_rate": 5.298907664975213e-06,
+      "loss": 0.68,
+      "step": 28274
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3072881591465433,
+      "learning_rate": 5.298234212332773e-06,
+      "loss": 0.7526,
+      "step": 28275
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3665685865961186,
+      "learning_rate": 5.297560787065506e-06,
+      "loss": 0.601,
+      "step": 28276
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9361998383676948,
+      "learning_rate": 5.296887389177326e-06,
+      "loss": 0.6804,
+      "step": 28277
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1612319539017082,
+      "learning_rate": 5.296214018672161e-06,
+      "loss": 0.6995,
+      "step": 28278
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3681228576906104,
+      "learning_rate": 5.295540675553926e-06,
+      "loss": 0.6316,
+      "step": 28279
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4348094887198892,
+      "learning_rate": 5.294867359826539e-06,
+      "loss": 0.8273,
+      "step": 28280
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5228407002994648,
+      "learning_rate": 5.294194071493925e-06,
+      "loss": 0.6236,
+      "step": 28281
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.287018276681531,
+      "learning_rate": 5.293520810560006e-06,
+      "loss": 0.7525,
+      "step": 28282
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4218887876521533,
+      "learning_rate": 5.292847577028698e-06,
+      "loss": 0.7411,
+      "step": 28283
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5614391013412694,
+      "learning_rate": 5.292174370903919e-06,
+      "loss": 0.7383,
+      "step": 28284
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8335133895394764,
+      "learning_rate": 5.29150119218959e-06,
+      "loss": 0.728,
+      "step": 28285
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.236329313543588,
+      "learning_rate": 5.290828040889637e-06,
+      "loss": 0.6788,
+      "step": 28286
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5415872560890953,
+      "learning_rate": 5.2901549170079736e-06,
+      "loss": 0.613,
+      "step": 28287
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2912880468120398,
+      "learning_rate": 5.2894818205485144e-06,
+      "loss": 0.7233,
+      "step": 28288
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4110106649010183,
+      "learning_rate": 5.288808751515184e-06,
+      "loss": 0.8257,
+      "step": 28289
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3890008864524392,
+      "learning_rate": 5.288135709911906e-06,
+      "loss": 0.6928,
+      "step": 28290
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4113900983933934,
+      "learning_rate": 5.287462695742593e-06,
+      "loss": 0.6395,
+      "step": 28291
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4884995317508405,
+      "learning_rate": 5.2867897090111594e-06,
+      "loss": 0.6971,
+      "step": 28292
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.361108741321272,
+      "learning_rate": 5.286116749721533e-06,
+      "loss": 0.6878,
+      "step": 28293
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4365137762373525,
+      "learning_rate": 5.285443817877622e-06,
+      "loss": 0.6044,
+      "step": 28294
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5336644843215625,
+      "learning_rate": 5.284770913483355e-06,
+      "loss": 0.7923,
+      "step": 28295
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3697741111496062,
+      "learning_rate": 5.28409803654264e-06,
+      "loss": 0.6957,
+      "step": 28296
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6268585508689495,
+      "learning_rate": 5.283425187059403e-06,
+      "loss": 0.6938,
+      "step": 28297
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1350315587342286,
+      "learning_rate": 5.282752365037554e-06,
+      "loss": 0.6883,
+      "step": 28298
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.620047473562656,
+      "learning_rate": 5.282079570481018e-06,
+      "loss": 0.8555,
+      "step": 28299
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3471821527336694,
+      "learning_rate": 5.2814068033937096e-06,
+      "loss": 0.7939,
+      "step": 28300
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.403199974864157,
+      "learning_rate": 5.28073406377954e-06,
+      "loss": 0.5861,
+      "step": 28301
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4357407443030623,
+      "learning_rate": 5.280061351642429e-06,
+      "loss": 0.6629,
+      "step": 28302
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4173404008929924,
+      "learning_rate": 5.279388666986301e-06,
+      "loss": 0.7726,
+      "step": 28303
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3714534993861953,
+      "learning_rate": 5.278716009815067e-06,
+      "loss": 0.6252,
+      "step": 28304
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5480154020615908,
+      "learning_rate": 5.278043380132638e-06,
+      "loss": 0.633,
+      "step": 28305
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3790096011190935,
+      "learning_rate": 5.277370777942936e-06,
+      "loss": 0.5609,
+      "step": 28306
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1286407100925562,
+      "learning_rate": 5.276698203249882e-06,
+      "loss": 0.6054,
+      "step": 28307
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3416479061882174,
+      "learning_rate": 5.276025656057384e-06,
+      "loss": 0.6218,
+      "step": 28308
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2898594730489747,
+      "learning_rate": 5.275353136369357e-06,
+      "loss": 0.622,
+      "step": 28309
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.29356135702499,
+      "learning_rate": 5.27468064418972e-06,
+      "loss": 0.6213,
+      "step": 28310
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3299102881790794,
+      "learning_rate": 5.274008179522392e-06,
+      "loss": 0.7057,
+      "step": 28311
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5391114188365311,
+      "learning_rate": 5.273335742371284e-06,
+      "loss": 0.7529,
+      "step": 28312
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.900239901367205,
+      "learning_rate": 5.272663332740308e-06,
+      "loss": 0.6796,
+      "step": 28313
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5931390842355375,
+      "learning_rate": 5.271990950633388e-06,
+      "loss": 0.4482,
+      "step": 28314
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8077269064595098,
+      "learning_rate": 5.271318596054429e-06,
+      "loss": 0.6729,
+      "step": 28315
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4133571204596462,
+      "learning_rate": 5.270646269007355e-06,
+      "loss": 0.6892,
+      "step": 28316
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5642389173121418,
+      "learning_rate": 5.269973969496069e-06,
+      "loss": 0.4754,
+      "step": 28317
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.092637688362982,
+      "learning_rate": 5.269301697524498e-06,
+      "loss": 0.6445,
+      "step": 28318
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.513175605283234,
+      "learning_rate": 5.268629453096547e-06,
+      "loss": 0.6862,
+      "step": 28319
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3084798628530452,
+      "learning_rate": 5.267957236216137e-06,
+      "loss": 0.7815,
+      "step": 28320
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6380279737141534,
+      "learning_rate": 5.267285046887176e-06,
+      "loss": 0.6621,
+      "step": 28321
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4719726269309048,
+      "learning_rate": 5.2666128851135775e-06,
+      "loss": 0.6614,
+      "step": 28322
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3829772899310018,
+      "learning_rate": 5.265940750899256e-06,
+      "loss": 0.7087,
+      "step": 28323
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5428638960575476,
+      "learning_rate": 5.265268644248133e-06,
+      "loss": 0.6381,
+      "step": 28324
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6005822147510251,
+      "learning_rate": 5.264596565164113e-06,
+      "loss": 0.5571,
+      "step": 28325
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.240171867653032,
+      "learning_rate": 5.263924513651107e-06,
+      "loss": 0.5985,
+      "step": 28326
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.900738568769719,
+      "learning_rate": 5.263252489713032e-06,
+      "loss": 0.7344,
+      "step": 28327
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4511095861977055,
+      "learning_rate": 5.262580493353805e-06,
+      "loss": 0.8742,
+      "step": 28328
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5009087020958818,
+      "learning_rate": 5.261908524577334e-06,
+      "loss": 0.6737,
+      "step": 28329
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.60788210833814,
+      "learning_rate": 5.261236583387528e-06,
+      "loss": 0.5141,
+      "step": 28330
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5290691407152763,
+      "learning_rate": 5.260564669788302e-06,
+      "loss": 0.7309,
+      "step": 28331
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5908544829471418,
+      "learning_rate": 5.259892783783573e-06,
+      "loss": 0.752,
+      "step": 28332
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.318188159032524,
+      "learning_rate": 5.259220925377249e-06,
+      "loss": 0.6533,
+      "step": 28333
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4706163844252471,
+      "learning_rate": 5.258549094573239e-06,
+      "loss": 0.6773,
+      "step": 28334
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.30830898040979,
+      "learning_rate": 5.257877291375459e-06,
+      "loss": 0.6243,
+      "step": 28335
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3158716653332474,
+      "learning_rate": 5.257205515787817e-06,
+      "loss": 0.6064,
+      "step": 28336
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6398837612991852,
+      "learning_rate": 5.256533767814229e-06,
+      "loss": 0.6488,
+      "step": 28337
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4491751154496095,
+      "learning_rate": 5.2558620474586e-06,
+      "loss": 0.6793,
+      "step": 28338
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.837085638797741,
+      "learning_rate": 5.255190354724846e-06,
+      "loss": 0.6536,
+      "step": 28339
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.462658412239539,
+      "learning_rate": 5.254518689616874e-06,
+      "loss": 0.7623,
+      "step": 28340
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3213995765748958,
+      "learning_rate": 5.2538470521385996e-06,
+      "loss": 0.7451,
+      "step": 28341
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5487424711234483,
+      "learning_rate": 5.253175442293929e-06,
+      "loss": 0.7774,
+      "step": 28342
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3647381217494243,
+      "learning_rate": 5.252503860086773e-06,
+      "loss": 0.6555,
+      "step": 28343
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3786701016306733,
+      "learning_rate": 5.2518323055210405e-06,
+      "loss": 0.7148,
+      "step": 28344
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2553525288176148,
+      "learning_rate": 5.251160778600648e-06,
+      "loss": 0.6889,
+      "step": 28345
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4409504831915352,
+      "learning_rate": 5.250489279329501e-06,
+      "loss": 0.6539,
+      "step": 28346
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.549073737074632,
+      "learning_rate": 5.249817807711505e-06,
+      "loss": 0.7372,
+      "step": 28347
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.666566278942237,
+      "learning_rate": 5.249146363750574e-06,
+      "loss": 0.7206,
+      "step": 28348
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4238859809146844,
+      "learning_rate": 5.24847494745062e-06,
+      "loss": 0.6746,
+      "step": 28349
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3706794676070189,
+      "learning_rate": 5.24780355881555e-06,
+      "loss": 0.6698,
+      "step": 28350
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2566569042010136,
+      "learning_rate": 5.247132197849267e-06,
+      "loss": 0.6371,
+      "step": 28351
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7129045359973052,
+      "learning_rate": 5.246460864555687e-06,
+      "loss": 0.7632,
+      "step": 28352
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4643670981086259,
+      "learning_rate": 5.245789558938719e-06,
+      "loss": 0.6574,
+      "step": 28353
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4205721555201114,
+      "learning_rate": 5.245118281002271e-06,
+      "loss": 0.7002,
+      "step": 28354
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2762951862847738,
+      "learning_rate": 5.244447030750244e-06,
+      "loss": 0.6629,
+      "step": 28355
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.60100183819764,
+      "learning_rate": 5.243775808186557e-06,
+      "loss": 0.7073,
+      "step": 28356
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4569186234198075,
+      "learning_rate": 5.2431046133151085e-06,
+      "loss": 0.5211,
+      "step": 28357
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.37087017606082,
+      "learning_rate": 5.242433446139816e-06,
+      "loss": 0.7443,
+      "step": 28358
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.341779868144895,
+      "learning_rate": 5.241762306664578e-06,
+      "loss": 0.7587,
+      "step": 28359
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5343849414481954,
+      "learning_rate": 5.241091194893309e-06,
+      "loss": 0.7707,
+      "step": 28360
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3471179215548827,
+      "learning_rate": 5.24042011082991e-06,
+      "loss": 0.7278,
+      "step": 28361
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.268035286339333,
+      "learning_rate": 5.2397490544782935e-06,
+      "loss": 0.6752,
+      "step": 28362
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4728421631854618,
+      "learning_rate": 5.2390780258423715e-06,
+      "loss": 0.7279,
+      "step": 28363
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.324845061312501,
+      "learning_rate": 5.238407024926038e-06,
+      "loss": 0.5897,
+      "step": 28364
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3655244808173659,
+      "learning_rate": 5.237736051733205e-06,
+      "loss": 0.6843,
+      "step": 28365
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7301579877445816,
+      "learning_rate": 5.23706510626778e-06,
+      "loss": 0.6361,
+      "step": 28366
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4246288776762484,
+      "learning_rate": 5.236394188533679e-06,
+      "loss": 0.6757,
+      "step": 28367
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3333441913801432,
+      "learning_rate": 5.2357232985347905e-06,
+      "loss": 0.7148,
+      "step": 28368
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3966503632685194,
+      "learning_rate": 5.23505243627503e-06,
+      "loss": 0.7282,
+      "step": 28369
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5778988773364787,
+      "learning_rate": 5.234381601758306e-06,
+      "loss": 0.4724,
+      "step": 28370
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6210357912676295,
+      "learning_rate": 5.233710794988518e-06,
+      "loss": 0.5585,
+      "step": 28371
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2816134611160757,
+      "learning_rate": 5.233040015969577e-06,
+      "loss": 0.7398,
+      "step": 28372
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4610245241040196,
+      "learning_rate": 5.232369264705384e-06,
+      "loss": 0.6274,
+      "step": 28373
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5941611251544617,
+      "learning_rate": 5.23169854119985e-06,
+      "loss": 0.4797,
+      "step": 28374
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4958253407291866,
+      "learning_rate": 5.231027845456873e-06,
+      "loss": 0.6771,
+      "step": 28375
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4712200567447322,
+      "learning_rate": 5.230357177480365e-06,
+      "loss": 0.6219,
+      "step": 28376
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5648884150165732,
+      "learning_rate": 5.229686537274223e-06,
+      "loss": 0.4415,
+      "step": 28377
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4056410200321034,
+      "learning_rate": 5.229015924842362e-06,
+      "loss": 0.6981,
+      "step": 28378
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2419220085194755,
+      "learning_rate": 5.228345340188677e-06,
+      "loss": 0.7175,
+      "step": 28379
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4331748746981081,
+      "learning_rate": 5.227674783317078e-06,
+      "loss": 0.66,
+      "step": 28380
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5956317179315453,
+      "learning_rate": 5.227004254231468e-06,
+      "loss": 0.7731,
+      "step": 28381
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.585127921867215,
+      "learning_rate": 5.226333752935747e-06,
+      "loss": 0.6431,
+      "step": 28382
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.395677766078183,
+      "learning_rate": 5.2256632794338216e-06,
+      "loss": 0.6824,
+      "step": 28383
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9982003064458715,
+      "learning_rate": 5.224992833729601e-06,
+      "loss": 0.5278,
+      "step": 28384
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2494170311227177,
+      "learning_rate": 5.2243224158269835e-06,
+      "loss": 0.7531,
+      "step": 28385
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4339733287859253,
+      "learning_rate": 5.223652025729868e-06,
+      "loss": 0.7207,
+      "step": 28386
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5008166325363832,
+      "learning_rate": 5.222981663442164e-06,
+      "loss": 0.7103,
+      "step": 28387
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5557504973945373,
+      "learning_rate": 5.222311328967775e-06,
+      "loss": 0.6967,
+      "step": 28388
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6021043587349417,
+      "learning_rate": 5.221641022310604e-06,
+      "loss": 0.7574,
+      "step": 28389
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5427655150693274,
+      "learning_rate": 5.220970743474547e-06,
+      "loss": 0.7749,
+      "step": 28390
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2960267690752725,
+      "learning_rate": 5.220300492463511e-06,
+      "loss": 0.6083,
+      "step": 28391
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.438667862834299,
+      "learning_rate": 5.219630269281404e-06,
+      "loss": 0.7619,
+      "step": 28392
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5910724582363915,
+      "learning_rate": 5.218960073932122e-06,
+      "loss": 0.762,
+      "step": 28393
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4597015037124694,
+      "learning_rate": 5.218289906419564e-06,
+      "loss": 0.6427,
+      "step": 28394
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5498323899796387,
+      "learning_rate": 5.217619766747641e-06,
+      "loss": 0.4913,
+      "step": 28395
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2253471865291294,
+      "learning_rate": 5.2169496549202444e-06,
+      "loss": 0.6237,
+      "step": 28396
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.112305640009551,
+      "learning_rate": 5.216279570941287e-06,
+      "loss": 0.6345,
+      "step": 28397
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.424597478169668,
+      "learning_rate": 5.2156095148146594e-06,
+      "loss": 0.6166,
+      "step": 28398
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.428582592382979,
+      "learning_rate": 5.214939486544271e-06,
+      "loss": 0.6382,
+      "step": 28399
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3138406808991194,
+      "learning_rate": 5.214269486134017e-06,
+      "loss": 0.7518,
+      "step": 28400
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2816183967609842,
+      "learning_rate": 5.213599513587805e-06,
+      "loss": 0.6809,
+      "step": 28401
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6138163332660671,
+      "learning_rate": 5.212929568909532e-06,
+      "loss": 0.5461,
+      "step": 28402
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3360536661666758,
+      "learning_rate": 5.212259652103095e-06,
+      "loss": 0.6763,
+      "step": 28403
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8342218241246795,
+      "learning_rate": 5.211589763172398e-06,
+      "loss": 0.605,
+      "step": 28404
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5502257324360411,
+      "learning_rate": 5.210919902121344e-06,
+      "loss": 0.6538,
+      "step": 28405
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3412516259380203,
+      "learning_rate": 5.2102500689538316e-06,
+      "loss": 0.747,
+      "step": 28406
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.497622233134083,
+      "learning_rate": 5.2095802636737556e-06,
+      "loss": 0.7789,
+      "step": 28407
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.627105849099144,
+      "learning_rate": 5.20891048628502e-06,
+      "loss": 0.6484,
+      "step": 28408
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3653794259949803,
+      "learning_rate": 5.208240736791528e-06,
+      "loss": 0.7119,
+      "step": 28409
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.138502017062702,
+      "learning_rate": 5.207571015197177e-06,
+      "loss": 0.7288,
+      "step": 28410
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.369475697615544,
+      "learning_rate": 5.206901321505859e-06,
+      "loss": 0.6352,
+      "step": 28411
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3230028606645052,
+      "learning_rate": 5.2062316557214785e-06,
+      "loss": 0.6034,
+      "step": 28412
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3427063942016224,
+      "learning_rate": 5.2055620178479405e-06,
+      "loss": 0.7026,
+      "step": 28413
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3465968360781848,
+      "learning_rate": 5.204892407889138e-06,
+      "loss": 0.7324,
+      "step": 28414
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.11245786473953,
+      "learning_rate": 5.204222825848966e-06,
+      "loss": 0.8749,
+      "step": 28415
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4279423356873855,
+      "learning_rate": 5.203553271731332e-06,
+      "loss": 0.7494,
+      "step": 28416
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5301492321842487,
+      "learning_rate": 5.202883745540124e-06,
+      "loss": 0.6791,
+      "step": 28417
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2768930401503178,
+      "learning_rate": 5.2022142472792494e-06,
+      "loss": 0.7527,
+      "step": 28418
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3084289405630107,
+      "learning_rate": 5.2015447769525985e-06,
+      "loss": 0.6744,
+      "step": 28419
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5936379700987666,
+      "learning_rate": 5.2008753345640775e-06,
+      "loss": 0.5196,
+      "step": 28420
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6674235736710767,
+      "learning_rate": 5.200205920117576e-06,
+      "loss": 0.7033,
+      "step": 28421
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4896265483316817,
+      "learning_rate": 5.199536533616999e-06,
+      "loss": 0.7185,
+      "step": 28422
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4166752901578856,
+      "learning_rate": 5.19886717506624e-06,
+      "loss": 0.5402,
+      "step": 28423
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5393320067158738,
+      "learning_rate": 5.198197844469193e-06,
+      "loss": 0.6681,
+      "step": 28424
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8065567748645026,
+      "learning_rate": 5.197528541829758e-06,
+      "loss": 0.6436,
+      "step": 28425
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5099682097645521,
+      "learning_rate": 5.196859267151835e-06,
+      "loss": 0.7481,
+      "step": 28426
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.395587597053068,
+      "learning_rate": 5.196190020439319e-06,
+      "loss": 0.6466,
+      "step": 28427
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6256539901882474,
+      "learning_rate": 5.195520801696101e-06,
+      "loss": 0.714,
+      "step": 28428
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5084215552031253,
+      "learning_rate": 5.194851610926083e-06,
+      "loss": 0.6489,
+      "step": 28429
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5902789062938938,
+      "learning_rate": 5.194182448133163e-06,
+      "loss": 0.5221,
+      "step": 28430
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4904452895460885,
+      "learning_rate": 5.1935133133212345e-06,
+      "loss": 0.6652,
+      "step": 28431
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.374401654959623,
+      "learning_rate": 5.19284420649419e-06,
+      "loss": 0.5668,
+      "step": 28432
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.678651326587063,
+      "learning_rate": 5.192175127655929e-06,
+      "loss": 0.7398,
+      "step": 28433
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.448438575588563,
+      "learning_rate": 5.191506076810349e-06,
+      "loss": 0.6144,
+      "step": 28434
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.407202880725959,
+      "learning_rate": 5.190837053961344e-06,
+      "loss": 0.6444,
+      "step": 28435
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.407196199551912,
+      "learning_rate": 5.190168059112804e-06,
+      "loss": 0.706,
+      "step": 28436
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4017828436436448,
+      "learning_rate": 5.189499092268632e-06,
+      "loss": 0.7465,
+      "step": 28437
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2919027840059805,
+      "learning_rate": 5.188830153432716e-06,
+      "loss": 0.6558,
+      "step": 28438
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4167540835605774,
+      "learning_rate": 5.188161242608957e-06,
+      "loss": 0.7541,
+      "step": 28439
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6406555357834781,
+      "learning_rate": 5.187492359801243e-06,
+      "loss": 0.5353,
+      "step": 28440
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.453460885527931,
+      "learning_rate": 5.186823505013476e-06,
+      "loss": 0.6884,
+      "step": 28441
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4621535992884296,
+      "learning_rate": 5.186154678249542e-06,
+      "loss": 0.6588,
+      "step": 28442
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4767184109754772,
+      "learning_rate": 5.185485879513344e-06,
+      "loss": 0.6228,
+      "step": 28443
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8211669358469467,
+      "learning_rate": 5.184817108808773e-06,
+      "loss": 0.7342,
+      "step": 28444
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3352051658045303,
+      "learning_rate": 5.184148366139715e-06,
+      "loss": 0.6206,
+      "step": 28445
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6378973216174088,
+      "learning_rate": 5.18347965151007e-06,
+      "loss": 0.5023,
+      "step": 28446
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3411196667885084,
+      "learning_rate": 5.182810964923733e-06,
+      "loss": 0.6344,
+      "step": 28447
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3746965402428912,
+      "learning_rate": 5.182142306384602e-06,
+      "loss": 0.7447,
+      "step": 28448
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4945103538488604,
+      "learning_rate": 5.181473675896557e-06,
+      "loss": 0.8126,
+      "step": 28449
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8049484593762208,
+      "learning_rate": 5.180805073463497e-06,
+      "loss": 0.836,
+      "step": 28450
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.430272145364646,
+      "learning_rate": 5.18013649908932e-06,
+      "loss": 0.7109,
+      "step": 28451
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7086038267027102,
+      "learning_rate": 5.179467952777915e-06,
+      "loss": 0.8023,
+      "step": 28452
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1921157089369798,
+      "learning_rate": 5.178799434533168e-06,
+      "loss": 0.6028,
+      "step": 28453
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6870387065166885,
+      "learning_rate": 5.178130944358978e-06,
+      "loss": 0.7825,
+      "step": 28454
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2097354351820293,
+      "learning_rate": 5.17746248225924e-06,
+      "loss": 0.7108,
+      "step": 28455
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2226231496097255,
+      "learning_rate": 5.176794048237842e-06,
+      "loss": 0.6794,
+      "step": 28456
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5652730520137791,
+      "learning_rate": 5.176125642298673e-06,
+      "loss": 0.4923,
+      "step": 28457
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3729244019281457,
+      "learning_rate": 5.1754572644456315e-06,
+      "loss": 0.6617,
+      "step": 28458
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.158961956680246,
+      "learning_rate": 5.1747889146826e-06,
+      "loss": 0.6191,
+      "step": 28459
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2304052412204165,
+      "learning_rate": 5.174120593013476e-06,
+      "loss": 0.6931,
+      "step": 28460
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2426506296036033,
+      "learning_rate": 5.173452299442154e-06,
+      "loss": 0.6293,
+      "step": 28461
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.380703758948219,
+      "learning_rate": 5.17278403397252e-06,
+      "loss": 0.7353,
+      "step": 28462
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3907309233607792,
+      "learning_rate": 5.172115796608462e-06,
+      "loss": 0.7954,
+      "step": 28463
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6035757789973109,
+      "learning_rate": 5.171447587353875e-06,
+      "loss": 0.6949,
+      "step": 28464
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.299364202327705,
+      "learning_rate": 5.170779406212656e-06,
+      "loss": 0.6616,
+      "step": 28465
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.459910406977318,
+      "learning_rate": 5.170111253188681e-06,
+      "loss": 0.6936,
+      "step": 28466
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6854760289532604,
+      "learning_rate": 5.169443128285846e-06,
+      "loss": 0.6274,
+      "step": 28467
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.695819589560405,
+      "learning_rate": 5.1687750315080446e-06,
+      "loss": 0.6881,
+      "step": 28468
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2546876644866334,
+      "learning_rate": 5.168106962859166e-06,
+      "loss": 0.6362,
+      "step": 28469
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5933848729562189,
+      "learning_rate": 5.1674389223431e-06,
+      "loss": 0.478,
+      "step": 28470
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4620088925144645,
+      "learning_rate": 5.16677090996373e-06,
+      "loss": 0.7387,
+      "step": 28471
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4784410186825507,
+      "learning_rate": 5.166102925724954e-06,
+      "loss": 0.6443,
+      "step": 28472
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5897477165769988,
+      "learning_rate": 5.165434969630654e-06,
+      "loss": 0.5074,
+      "step": 28473
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4998078278255211,
+      "learning_rate": 5.164767041684724e-06,
+      "loss": 0.7271,
+      "step": 28474
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3729241058603567,
+      "learning_rate": 5.1640991418910494e-06,
+      "loss": 0.7026,
+      "step": 28475
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6083537278579643,
+      "learning_rate": 5.163431270253523e-06,
+      "loss": 0.5925,
+      "step": 28476
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4852049873190838,
+      "learning_rate": 5.162763426776027e-06,
+      "loss": 0.7743,
+      "step": 28477
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3818284238507408,
+      "learning_rate": 5.162095611462458e-06,
+      "loss": 0.6524,
+      "step": 28478
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5995266410910113,
+      "learning_rate": 5.161427824316703e-06,
+      "loss": 0.5008,
+      "step": 28479
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3481722251703117,
+      "learning_rate": 5.160760065342639e-06,
+      "loss": 0.6612,
+      "step": 28480
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9670907125946981,
+      "learning_rate": 5.160092334544165e-06,
+      "loss": 0.7869,
+      "step": 28481
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4617888984169605,
+      "learning_rate": 5.159424631925168e-06,
+      "loss": 0.7981,
+      "step": 28482
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5350057168714482,
+      "learning_rate": 5.158756957489533e-06,
+      "loss": 0.7225,
+      "step": 28483
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.372734678536357,
+      "learning_rate": 5.158089311241143e-06,
+      "loss": 0.7002,
+      "step": 28484
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6238203631934418,
+      "learning_rate": 5.157421693183893e-06,
+      "loss": 0.5188,
+      "step": 28485
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4200285152242011,
+      "learning_rate": 5.156754103321669e-06,
+      "loss": 0.7295,
+      "step": 28486
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.235934750646987,
+      "learning_rate": 5.156086541658356e-06,
+      "loss": 0.657,
+      "step": 28487
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6546865513216626,
+      "learning_rate": 5.155419008197837e-06,
+      "loss": 0.7349,
+      "step": 28488
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4005646125050975,
+      "learning_rate": 5.154751502944002e-06,
+      "loss": 0.6658,
+      "step": 28489
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.5866293268087235,
+      "learning_rate": 5.154084025900742e-06,
+      "loss": 0.7788,
+      "step": 28490
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3744077644243484,
+      "learning_rate": 5.153416577071939e-06,
+      "loss": 0.7802,
+      "step": 28491
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4883234477107434,
+      "learning_rate": 5.152749156461475e-06,
+      "loss": 0.7795,
+      "step": 28492
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8309732859796306,
+      "learning_rate": 5.152081764073244e-06,
+      "loss": 0.6437,
+      "step": 28493
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5866623450289816,
+      "learning_rate": 5.1514143999111234e-06,
+      "loss": 0.511,
+      "step": 28494
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4273913399830902,
+      "learning_rate": 5.150747063979008e-06,
+      "loss": 0.6146,
+      "step": 28495
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.319891910255053,
+      "learning_rate": 5.150079756280774e-06,
+      "loss": 0.6593,
+      "step": 28496
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2147281395176213,
+      "learning_rate": 5.149412476820314e-06,
+      "loss": 0.6758,
+      "step": 28497
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5969932017184241,
+      "learning_rate": 5.148745225601509e-06,
+      "loss": 0.5314,
+      "step": 28498
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3111442800101445,
+      "learning_rate": 5.148078002628247e-06,
+      "loss": 0.7058,
+      "step": 28499
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.486723160892994,
+      "learning_rate": 5.147410807904408e-06,
+      "loss": 0.6313,
+      "step": 28500
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5606085778288334,
+      "learning_rate": 5.146743641433884e-06,
+      "loss": 0.6501,
+      "step": 28501
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.093061327045977,
+      "learning_rate": 5.146076503220549e-06,
+      "loss": 0.5779,
+      "step": 28502
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4714581937559839,
+      "learning_rate": 5.145409393268299e-06,
+      "loss": 0.7051,
+      "step": 28503
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3151516125182414,
+      "learning_rate": 5.144742311581013e-06,
+      "loss": 0.6054,
+      "step": 28504
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7350483913308063,
+      "learning_rate": 5.144075258162569e-06,
+      "loss": 0.7789,
+      "step": 28505
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.643890377579004,
+      "learning_rate": 5.143408233016855e-06,
+      "loss": 0.5891,
+      "step": 28506
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5609095217759648,
+      "learning_rate": 5.142741236147762e-06,
+      "loss": 0.4932,
+      "step": 28507
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4562446256700556,
+      "learning_rate": 5.142074267559167e-06,
+      "loss": 0.7543,
+      "step": 28508
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4701032189374494,
+      "learning_rate": 5.141407327254951e-06,
+      "loss": 0.7199,
+      "step": 28509
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4967634222778923,
+      "learning_rate": 5.140740415238996e-06,
+      "loss": 0.7894,
+      "step": 28510
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2583244304716088,
+      "learning_rate": 5.140073531515195e-06,
+      "loss": 0.7127,
+      "step": 28511
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.619209474045925,
+      "learning_rate": 5.139406676087424e-06,
+      "loss": 0.6598,
+      "step": 28512
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4100603046809093,
+      "learning_rate": 5.138739848959563e-06,
+      "loss": 0.7492,
+      "step": 28513
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.566761201234818,
+      "learning_rate": 5.138073050135497e-06,
+      "loss": 0.5061,
+      "step": 28514
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3375041567072716,
+      "learning_rate": 5.137406279619113e-06,
+      "loss": 0.6211,
+      "step": 28515
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3535403133111394,
+      "learning_rate": 5.136739537414291e-06,
+      "loss": 0.7071,
+      "step": 28516
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.43959271385658,
+      "learning_rate": 5.136072823524905e-06,
+      "loss": 0.7831,
+      "step": 28517
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.582805322184844,
+      "learning_rate": 5.1354061379548455e-06,
+      "loss": 0.8106,
+      "step": 28518
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.433891975730388,
+      "learning_rate": 5.134739480707991e-06,
+      "loss": 0.5261,
+      "step": 28519
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.533162372762774,
+      "learning_rate": 5.1340728517882256e-06,
+      "loss": 0.7866,
+      "step": 28520
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.237457704572216,
+      "learning_rate": 5.133406251199424e-06,
+      "loss": 0.7936,
+      "step": 28521
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3429243087107423,
+      "learning_rate": 5.132739678945478e-06,
+      "loss": 0.6871,
+      "step": 28522
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3288888523367175,
+      "learning_rate": 5.132073135030255e-06,
+      "loss": 0.6489,
+      "step": 28523
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2922554205633658,
+      "learning_rate": 5.13140661945765e-06,
+      "loss": 0.7244,
+      "step": 28524
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6304530717460533,
+      "learning_rate": 5.1307401322315375e-06,
+      "loss": 0.6093,
+      "step": 28525
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3629205448122281,
+      "learning_rate": 5.1300736733557936e-06,
+      "loss": 0.6454,
+      "step": 28526
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5550608102794516,
+      "learning_rate": 5.129407242834301e-06,
+      "loss": 0.6906,
+      "step": 28527
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.44828181639002,
+      "learning_rate": 5.128740840670946e-06,
+      "loss": 0.7262,
+      "step": 28528
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.387135477742036,
+      "learning_rate": 5.128074466869603e-06,
+      "loss": 0.5596,
+      "step": 28529
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3997803516039449,
+      "learning_rate": 5.127408121434152e-06,
+      "loss": 0.7139,
+      "step": 28530
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2287074807582987,
+      "learning_rate": 5.126741804368471e-06,
+      "loss": 0.7362,
+      "step": 28531
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.582585923601692,
+      "learning_rate": 5.1260755156764474e-06,
+      "loss": 0.6786,
+      "step": 28532
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4060078085355767,
+      "learning_rate": 5.125409255361954e-06,
+      "loss": 0.782,
+      "step": 28533
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2730389385491392,
+      "learning_rate": 5.124743023428867e-06,
+      "loss": 0.7028,
+      "step": 28534
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.372702485496691,
+      "learning_rate": 5.124076819881071e-06,
+      "loss": 0.7401,
+      "step": 28535
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3509080948940981,
+      "learning_rate": 5.123410644722446e-06,
+      "loss": 0.7466,
+      "step": 28536
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4719184185066443,
+      "learning_rate": 5.1227444979568694e-06,
+      "loss": 0.6082,
+      "step": 28537
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.434118847087047,
+      "learning_rate": 5.122078379588214e-06,
+      "loss": 0.7918,
+      "step": 28538
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6138417821936395,
+      "learning_rate": 5.121412289620367e-06,
+      "loss": 0.7418,
+      "step": 28539
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6721368409146036,
+      "learning_rate": 5.120746228057197e-06,
+      "loss": 0.5275,
+      "step": 28540
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6699660148280413,
+      "learning_rate": 5.120080194902593e-06,
+      "loss": 0.7276,
+      "step": 28541
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3333360527191123,
+      "learning_rate": 5.119414190160421e-06,
+      "loss": 0.6403,
+      "step": 28542
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2406471687760081,
+      "learning_rate": 5.11874821383457e-06,
+      "loss": 0.5618,
+      "step": 28543
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6747745055559773,
+      "learning_rate": 5.118082265928908e-06,
+      "loss": 0.7575,
+      "step": 28544
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4458997087530698,
+      "learning_rate": 5.117416346447317e-06,
+      "loss": 0.7038,
+      "step": 28545
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6884458200559123,
+      "learning_rate": 5.116750455393681e-06,
+      "loss": 0.5793,
+      "step": 28546
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2799840362092127,
+      "learning_rate": 5.116084592771861e-06,
+      "loss": 0.7608,
+      "step": 28547
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5090049079822954,
+      "learning_rate": 5.115418758585745e-06,
+      "loss": 0.7071,
+      "step": 28548
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1733157279867013,
+      "learning_rate": 5.114752952839207e-06,
+      "loss": 0.7566,
+      "step": 28549
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2085741815842075,
+      "learning_rate": 5.114087175536131e-06,
+      "loss": 0.6933,
+      "step": 28550
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8741607100682607,
+      "learning_rate": 5.113421426680378e-06,
+      "loss": 0.736,
+      "step": 28551
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3452056025751888,
+      "learning_rate": 5.112755706275833e-06,
+      "loss": 0.7835,
+      "step": 28552
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5653530886841653,
+      "learning_rate": 5.112090014326376e-06,
+      "loss": 0.7803,
+      "step": 28553
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.679004103958661,
+      "learning_rate": 5.111424350835879e-06,
+      "loss": 0.7819,
+      "step": 28554
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.410342218965118,
+      "learning_rate": 5.110758715808213e-06,
+      "loss": 0.6504,
+      "step": 28555
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3743036765702592,
+      "learning_rate": 5.1100931092472564e-06,
+      "loss": 0.7412,
+      "step": 28556
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.290765024870339,
+      "learning_rate": 5.10942753115689e-06,
+      "loss": 0.6398,
+      "step": 28557
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0492164087071107,
+      "learning_rate": 5.1087619815409815e-06,
+      "loss": 0.6944,
+      "step": 28558
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5912688100877284,
+      "learning_rate": 5.1080964604034136e-06,
+      "loss": 0.6729,
+      "step": 28559
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.324505454458412,
+      "learning_rate": 5.107430967748056e-06,
+      "loss": 0.7859,
+      "step": 28560
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6184633422816384,
+      "learning_rate": 5.106765503578781e-06,
+      "loss": 0.6251,
+      "step": 28561
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5693803347565864,
+      "learning_rate": 5.1061000678994665e-06,
+      "loss": 0.6713,
+      "step": 28562
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4394050527378839,
+      "learning_rate": 5.10543466071399e-06,
+      "loss": 0.7516,
+      "step": 28563
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3420761815793911,
+      "learning_rate": 5.104769282026223e-06,
+      "loss": 0.6657,
+      "step": 28564
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5058999260024977,
+      "learning_rate": 5.104103931840035e-06,
+      "loss": 0.6817,
+      "step": 28565
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6222554403139359,
+      "learning_rate": 5.103438610159303e-06,
+      "loss": 0.5259,
+      "step": 28566
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8091230297961005,
+      "learning_rate": 5.102773316987905e-06,
+      "loss": 0.6096,
+      "step": 28567
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5823806036604565,
+      "learning_rate": 5.102108052329714e-06,
+      "loss": 0.4916,
+      "step": 28568
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3266874378951021,
+      "learning_rate": 5.1014428161885956e-06,
+      "loss": 0.64,
+      "step": 28569
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4777109834352453,
+      "learning_rate": 5.1007776085684265e-06,
+      "loss": 0.6413,
+      "step": 28570
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3244823728613975,
+      "learning_rate": 5.100112429473087e-06,
+      "loss": 0.6473,
+      "step": 28571
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.700749299054773,
+      "learning_rate": 5.0994472789064434e-06,
+      "loss": 0.735,
+      "step": 28572
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5859052677726282,
+      "learning_rate": 5.098782156872365e-06,
+      "loss": 0.6506,
+      "step": 28573
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3670290724996537,
+      "learning_rate": 5.0981170633747326e-06,
+      "loss": 0.6387,
+      "step": 28574
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6092358860598344,
+      "learning_rate": 5.097451998417411e-06,
+      "loss": 0.5567,
+      "step": 28575
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6017491625740625,
+      "learning_rate": 5.09678696200428e-06,
+      "loss": 0.7092,
+      "step": 28576
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2636153543685813,
+      "learning_rate": 5.096121954139204e-06,
+      "loss": 0.7504,
+      "step": 28577
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.202871969068032,
+      "learning_rate": 5.095456974826061e-06,
+      "loss": 0.6569,
+      "step": 28578
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.039230548092801,
+      "learning_rate": 5.094792024068718e-06,
+      "loss": 0.7237,
+      "step": 28579
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6269416641382878,
+      "learning_rate": 5.0941271018710514e-06,
+      "loss": 0.5149,
+      "step": 28580
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5839695144908728,
+      "learning_rate": 5.093462208236931e-06,
+      "loss": 0.5338,
+      "step": 28581
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2550448943825459,
+      "learning_rate": 5.092797343170223e-06,
+      "loss": 0.7258,
+      "step": 28582
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.6610945667681385,
+      "learning_rate": 5.092132506674801e-06,
+      "loss": 0.7632,
+      "step": 28583
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2221242063934707,
+      "learning_rate": 5.091467698754543e-06,
+      "loss": 0.6825,
+      "step": 28584
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5092406074427431,
+      "learning_rate": 5.090802919413314e-06,
+      "loss": 0.6769,
+      "step": 28585
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6060754984900295,
+      "learning_rate": 5.09013816865498e-06,
+      "loss": 0.5287,
+      "step": 28586
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9460960918832695,
+      "learning_rate": 5.089473446483416e-06,
+      "loss": 0.7511,
+      "step": 28587
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.610820285561574,
+      "learning_rate": 5.0888087529024965e-06,
+      "loss": 0.5538,
+      "step": 28588
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3406730716092714,
+      "learning_rate": 5.088144087916087e-06,
+      "loss": 0.6846,
+      "step": 28589
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2385358610007644,
+      "learning_rate": 5.087479451528055e-06,
+      "loss": 0.6056,
+      "step": 28590
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3775283807315764,
+      "learning_rate": 5.086814843742272e-06,
+      "loss": 0.8271,
+      "step": 28591
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3956285426796458,
+      "learning_rate": 5.086150264562612e-06,
+      "loss": 0.8251,
+      "step": 28592
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.674766882036685,
+      "learning_rate": 5.08548571399294e-06,
+      "loss": 0.675,
+      "step": 28593
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3643830980358753,
+      "learning_rate": 5.0848211920371236e-06,
+      "loss": 0.7901,
+      "step": 28594
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2397918353510509,
+      "learning_rate": 5.084156698699038e-06,
+      "loss": 0.71,
+      "step": 28595
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0710569768318936,
+      "learning_rate": 5.0834922339825445e-06,
+      "loss": 0.5925,
+      "step": 28596
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.230441932052473,
+      "learning_rate": 5.08282779789152e-06,
+      "loss": 0.7066,
+      "step": 28597
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2463024059861647,
+      "learning_rate": 5.082163390429824e-06,
+      "loss": 0.6633,
+      "step": 28598
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.345402842171718,
+      "learning_rate": 5.081499011601334e-06,
+      "loss": 0.6763,
+      "step": 28599
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4440902526755313,
+      "learning_rate": 5.0808346614099106e-06,
+      "loss": 0.6679,
+      "step": 28600
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.184426091596998,
+      "learning_rate": 5.080170339859429e-06,
+      "loss": 0.6317,
+      "step": 28601
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5924354936076444,
+      "learning_rate": 5.079506046953753e-06,
+      "loss": 0.7267,
+      "step": 28602
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5018575567047974,
+      "learning_rate": 5.078841782696746e-06,
+      "loss": 0.6988,
+      "step": 28603
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5181508366900824,
+      "learning_rate": 5.07817754709228e-06,
+      "loss": 0.7618,
+      "step": 28604
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4094398785020337,
+      "learning_rate": 5.0775133401442275e-06,
+      "loss": 0.6354,
+      "step": 28605
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.647497751056639,
+      "learning_rate": 5.0768491618564504e-06,
+      "loss": 0.7034,
+      "step": 28606
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4882678833930538,
+      "learning_rate": 5.076185012232813e-06,
+      "loss": 0.5847,
+      "step": 28607
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6561567676019515,
+      "learning_rate": 5.0755208912771835e-06,
+      "loss": 0.7346,
+      "step": 28608
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.401827911171612,
+      "learning_rate": 5.074856798993436e-06,
+      "loss": 0.7047,
+      "step": 28609
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.335077448366656,
+      "learning_rate": 5.0741927353854305e-06,
+      "loss": 0.7397,
+      "step": 28610
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5009578680382492,
+      "learning_rate": 5.073528700457031e-06,
+      "loss": 0.6447,
+      "step": 28611
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4340992405115045,
+      "learning_rate": 5.0728646942121074e-06,
+      "loss": 0.8353,
+      "step": 28612
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6657154072450158,
+      "learning_rate": 5.0722007166545294e-06,
+      "loss": 0.7301,
+      "step": 28613
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.781754552320829,
+      "learning_rate": 5.071536767788158e-06,
+      "loss": 0.6467,
+      "step": 28614
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.743698152714447,
+      "learning_rate": 5.070872847616857e-06,
+      "loss": 0.5318,
+      "step": 28615
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.474457327090471,
+      "learning_rate": 5.070208956144498e-06,
+      "loss": 0.6425,
+      "step": 28616
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4456391476483468,
+      "learning_rate": 5.0695450933749395e-06,
+      "loss": 0.6377,
+      "step": 28617
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5352217523345817,
+      "learning_rate": 5.068881259312054e-06,
+      "loss": 0.6883,
+      "step": 28618
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5487500890728911,
+      "learning_rate": 5.068217453959698e-06,
+      "loss": 0.7633,
+      "step": 28619
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1644305920033329,
+      "learning_rate": 5.067553677321748e-06,
+      "loss": 0.6321,
+      "step": 28620
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4793450001504422,
+      "learning_rate": 5.066889929402057e-06,
+      "loss": 0.7149,
+      "step": 28621
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.576884729924891,
+      "learning_rate": 5.066226210204498e-06,
+      "loss": 0.5248,
+      "step": 28622
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.956439987289428,
+      "learning_rate": 5.065562519732929e-06,
+      "loss": 0.6372,
+      "step": 28623
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.397459609692404,
+      "learning_rate": 5.064898857991221e-06,
+      "loss": 0.6696,
+      "step": 28624
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3289996940283215,
+      "learning_rate": 5.0642352249832295e-06,
+      "loss": 0.7588,
+      "step": 28625
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.259672679794536,
+      "learning_rate": 5.0635716207128274e-06,
+      "loss": 0.7098,
+      "step": 28626
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3877846889992587,
+      "learning_rate": 5.0629080451838745e-06,
+      "loss": 0.6273,
+      "step": 28627
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4730825442565347,
+      "learning_rate": 5.062244498400228e-06,
+      "loss": 0.7731,
+      "step": 28628
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4536634091933347,
+      "learning_rate": 5.0615809803657615e-06,
+      "loss": 0.7349,
+      "step": 28629
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4596786919327411,
+      "learning_rate": 5.060917491084331e-06,
+      "loss": 0.6154,
+      "step": 28630
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3229607839274455,
+      "learning_rate": 5.06025403055981e-06,
+      "loss": 0.7378,
+      "step": 28631
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3600458163458073,
+      "learning_rate": 5.059590598796047e-06,
+      "loss": 0.7276,
+      "step": 28632
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4506505350143022,
+      "learning_rate": 5.058927195796912e-06,
+      "loss": 0.7336,
+      "step": 28633
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.489044070099684,
+      "learning_rate": 5.0582638215662715e-06,
+      "loss": 0.6514,
+      "step": 28634
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6146157286865617,
+      "learning_rate": 5.057600476107982e-06,
+      "loss": 0.6557,
+      "step": 28635
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4102864990180475,
+      "learning_rate": 5.0569371594259056e-06,
+      "loss": 0.6614,
+      "step": 28636
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4457938179774728,
+      "learning_rate": 5.056273871523904e-06,
+      "loss": 0.7778,
+      "step": 28637
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3079145390251916,
+      "learning_rate": 5.055610612405846e-06,
+      "loss": 0.7518,
+      "step": 28638
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5727336375554893,
+      "learning_rate": 5.054947382075588e-06,
+      "loss": 0.5292,
+      "step": 28639
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.390754731452752,
+      "learning_rate": 5.054284180536988e-06,
+      "loss": 0.7147,
+      "step": 28640
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2891798543297472,
+      "learning_rate": 5.053621007793916e-06,
+      "loss": 0.6421,
+      "step": 28641
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3006209863387757,
+      "learning_rate": 5.052957863850223e-06,
+      "loss": 0.6516,
+      "step": 28642
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.28206597371749,
+      "learning_rate": 5.052294748709776e-06,
+      "loss": 0.6953,
+      "step": 28643
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.575071358634623,
+      "learning_rate": 5.051631662376441e-06,
+      "loss": 0.5284,
+      "step": 28644
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4744381566090983,
+      "learning_rate": 5.050968604854071e-06,
+      "loss": 0.8707,
+      "step": 28645
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3753420459008006,
+      "learning_rate": 5.050305576146526e-06,
+      "loss": 0.6454,
+      "step": 28646
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.997186471699976,
+      "learning_rate": 5.0496425762576685e-06,
+      "loss": 0.8375,
+      "step": 28647
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5783785201034152,
+      "learning_rate": 5.048979605191367e-06,
+      "loss": 0.4698,
+      "step": 28648
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6079330624459657,
+      "learning_rate": 5.0483166629514654e-06,
+      "loss": 0.5073,
+      "step": 28649
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3958926737860864,
+      "learning_rate": 5.047653749541832e-06,
+      "loss": 0.6021,
+      "step": 28650
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3754258442267042,
+      "learning_rate": 5.046990864966328e-06,
+      "loss": 0.7184,
+      "step": 28651
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2926245609364233,
+      "learning_rate": 5.046328009228817e-06,
+      "loss": 0.6452,
+      "step": 28652
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3571480541647412,
+      "learning_rate": 5.045665182333146e-06,
+      "loss": 0.6758,
+      "step": 28653
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.393404244805768,
+      "learning_rate": 5.045002384283179e-06,
+      "loss": 0.7033,
+      "step": 28654
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5443610668188255,
+      "learning_rate": 5.044339615082783e-06,
+      "loss": 0.7142,
+      "step": 28655
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9290587333167635,
+      "learning_rate": 5.043676874735807e-06,
+      "loss": 0.7412,
+      "step": 28656
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7637265334274943,
+      "learning_rate": 5.043014163246116e-06,
+      "loss": 0.6706,
+      "step": 28657
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1668687201844852,
+      "learning_rate": 5.042351480617562e-06,
+      "loss": 0.7259,
+      "step": 28658
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9477271886694982,
+      "learning_rate": 5.041688826854012e-06,
+      "loss": 0.7608,
+      "step": 28659
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3253045309632459,
+      "learning_rate": 5.041026201959314e-06,
+      "loss": 0.6648,
+      "step": 28660
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.475366487224251,
+      "learning_rate": 5.040363605937337e-06,
+      "loss": 0.7131,
+      "step": 28661
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9282837885984383,
+      "learning_rate": 5.039701038791932e-06,
+      "loss": 0.7582,
+      "step": 28662
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3159371660828052,
+      "learning_rate": 5.039038500526955e-06,
+      "loss": 0.7809,
+      "step": 28663
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5930209849169641,
+      "learning_rate": 5.038375991146266e-06,
+      "loss": 0.5162,
+      "step": 28664
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3860542092445474,
+      "learning_rate": 5.037713510653727e-06,
+      "loss": 0.7455,
+      "step": 28665
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3450317198023765,
+      "learning_rate": 5.03705105905319e-06,
+      "loss": 0.6878,
+      "step": 28666
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4914654289639813,
+      "learning_rate": 5.0363886363485105e-06,
+      "loss": 0.8158,
+      "step": 28667
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0138412646661736,
+      "learning_rate": 5.035726242543546e-06,
+      "loss": 0.6498,
+      "step": 28668
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.341414390534334,
+      "learning_rate": 5.0350638776421615e-06,
+      "loss": 0.6408,
+      "step": 28669
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.210612860403683,
+      "learning_rate": 5.034401541648204e-06,
+      "loss": 0.6139,
+      "step": 28670
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7702024890296015,
+      "learning_rate": 5.033739234565531e-06,
+      "loss": 0.711,
+      "step": 28671
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3007604602864073,
+      "learning_rate": 5.033076956398e-06,
+      "loss": 0.7259,
+      "step": 28672
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4737910675962163,
+      "learning_rate": 5.032414707149471e-06,
+      "loss": 0.7844,
+      "step": 28673
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.812204848266552,
+      "learning_rate": 5.031752486823797e-06,
+      "loss": 0.6599,
+      "step": 28674
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7817090528215702,
+      "learning_rate": 5.031090295424828e-06,
+      "loss": 0.775,
+      "step": 28675
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6080504232919939,
+      "learning_rate": 5.030428132956428e-06,
+      "loss": 0.5763,
+      "step": 28676
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6635068569384956,
+      "learning_rate": 5.029765999422447e-06,
+      "loss": 0.7461,
+      "step": 28677
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3392910060574463,
+      "learning_rate": 5.029103894826745e-06,
+      "loss": 0.729,
+      "step": 28678
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.381589791770767,
+      "learning_rate": 5.0284418191731684e-06,
+      "loss": 0.7171,
+      "step": 28679
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6519050291188941,
+      "learning_rate": 5.027779772465583e-06,
+      "loss": 0.6342,
+      "step": 28680
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4653920187802,
+      "learning_rate": 5.027117754707833e-06,
+      "loss": 0.7374,
+      "step": 28681
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4764807670664732,
+      "learning_rate": 5.026455765903783e-06,
+      "loss": 0.6504,
+      "step": 28682
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7260872361782889,
+      "learning_rate": 5.025793806057282e-06,
+      "loss": 0.6958,
+      "step": 28683
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2409977260585634,
+      "learning_rate": 5.025131875172179e-06,
+      "loss": 0.7434,
+      "step": 28684
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.43516103879611,
+      "learning_rate": 5.024469973252334e-06,
+      "loss": 0.6804,
+      "step": 28685
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3531862650182123,
+      "learning_rate": 5.023808100301604e-06,
+      "loss": 0.6419,
+      "step": 28686
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5985788532673006,
+      "learning_rate": 5.023146256323839e-06,
+      "loss": 0.5134,
+      "step": 28687
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3102393605888452,
+      "learning_rate": 5.022484441322888e-06,
+      "loss": 0.5249,
+      "step": 28688
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6555297101880284,
+      "learning_rate": 5.021822655302608e-06,
+      "loss": 0.8063,
+      "step": 28689
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.259135453439861,
+      "learning_rate": 5.021160898266857e-06,
+      "loss": 0.7572,
+      "step": 28690
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2952111979889989,
+      "learning_rate": 5.020499170219483e-06,
+      "loss": 0.6997,
+      "step": 28691
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.7233335850898888,
+      "learning_rate": 5.019837471164335e-06,
+      "loss": 0.6525,
+      "step": 28692
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.482608118853271,
+      "learning_rate": 5.019175801105272e-06,
+      "loss": 0.6429,
+      "step": 28693
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7205085953028663,
+      "learning_rate": 5.018514160046146e-06,
+      "loss": 0.5842,
+      "step": 28694
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3850415842529888,
+      "learning_rate": 5.017852547990809e-06,
+      "loss": 0.6783,
+      "step": 28695
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5841266709421434,
+      "learning_rate": 5.0171909649431075e-06,
+      "loss": 0.5195,
+      "step": 28696
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3279097681897345,
+      "learning_rate": 5.0165294109069e-06,
+      "loss": 0.8126,
+      "step": 28697
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5394572226694383,
+      "learning_rate": 5.015867885886035e-06,
+      "loss": 0.7292,
+      "step": 28698
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.535319867794247,
+      "learning_rate": 5.0152063898843685e-06,
+      "loss": 0.7422,
+      "step": 28699
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6324699076838805,
+      "learning_rate": 5.014544922905744e-06,
+      "loss": 0.4937,
+      "step": 28700
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.384357625162453,
+      "learning_rate": 5.013883484954021e-06,
+      "loss": 0.6917,
+      "step": 28701
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.533952203432041,
+      "learning_rate": 5.013222076033043e-06,
+      "loss": 0.6888,
+      "step": 28702
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6091652698845379,
+      "learning_rate": 5.012560696146669e-06,
+      "loss": 0.5496,
+      "step": 28703
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.507109335041127,
+      "learning_rate": 5.011899345298745e-06,
+      "loss": 0.7012,
+      "step": 28704
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7795183598935005,
+      "learning_rate": 5.011238023493119e-06,
+      "loss": 0.7882,
+      "step": 28705
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4126865218374045,
+      "learning_rate": 5.010576730733645e-06,
+      "loss": 0.6718,
+      "step": 28706
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.513557509160717,
+      "learning_rate": 5.0099154670241756e-06,
+      "loss": 0.7249,
+      "step": 28707
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5128983262844786,
+      "learning_rate": 5.009254232368559e-06,
+      "loss": 0.762,
+      "step": 28708
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5010440536739127,
+      "learning_rate": 5.0085930267706404e-06,
+      "loss": 0.695,
+      "step": 28709
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.331643795740837,
+      "learning_rate": 5.007931850234273e-06,
+      "loss": 0.7697,
+      "step": 28710
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4254206297701846,
+      "learning_rate": 5.0072707027633115e-06,
+      "loss": 0.6702,
+      "step": 28711
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6147744315310663,
+      "learning_rate": 5.006609584361601e-06,
+      "loss": 0.5063,
+      "step": 28712
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.350580306217693,
+      "learning_rate": 5.005948495032985e-06,
+      "loss": 0.6724,
+      "step": 28713
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.417510936549154,
+      "learning_rate": 5.0052874347813186e-06,
+      "loss": 0.6977,
+      "step": 28714
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.344449825250622,
+      "learning_rate": 5.004626403610454e-06,
+      "loss": 0.596,
+      "step": 28715
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3338642994642418,
+      "learning_rate": 5.0039654015242355e-06,
+      "loss": 0.7085,
+      "step": 28716
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2047562790560975,
+      "learning_rate": 5.00330442852651e-06,
+      "loss": 0.6662,
+      "step": 28717
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4727875628716889,
+      "learning_rate": 5.00264348462113e-06,
+      "loss": 0.6383,
+      "step": 28718
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3755151233696352,
+      "learning_rate": 5.001982569811938e-06,
+      "loss": 0.7626,
+      "step": 28719
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3127238234758063,
+      "learning_rate": 5.001321684102791e-06,
+      "loss": 0.6783,
+      "step": 28720
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5071732750279898,
+      "learning_rate": 5.000660827497527e-06,
+      "loss": 0.7966,
+      "step": 28721
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6051595313451269,
+      "learning_rate": 5.000000000000003e-06,
+      "loss": 0.4798,
+      "step": 28722
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5333383596874977,
+      "learning_rate": 4.999339201614057e-06,
+      "loss": 0.653,
+      "step": 28723
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5954247144815774,
+      "learning_rate": 4.998678432343545e-06,
+      "loss": 0.5296,
+      "step": 28724
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2668755165488497,
+      "learning_rate": 4.998017692192312e-06,
+      "loss": 0.739,
+      "step": 28725
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6231039787711761,
+      "learning_rate": 4.9973569811641985e-06,
+      "loss": 0.5077,
+      "step": 28726
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4204274440120181,
+      "learning_rate": 4.996696299263057e-06,
+      "loss": 0.5814,
+      "step": 28727
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6245333572146102,
+      "learning_rate": 4.996035646492734e-06,
+      "loss": 0.5259,
+      "step": 28728
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4391488777151824,
+      "learning_rate": 4.995375022857083e-06,
+      "loss": 0.7938,
+      "step": 28729
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4785909954750036,
+      "learning_rate": 4.994714428359936e-06,
+      "loss": 0.565,
+      "step": 28730
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4412391717279949,
+      "learning_rate": 4.994053863005145e-06,
+      "loss": 0.623,
+      "step": 28731
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4032494292768765,
+      "learning_rate": 4.993393326796562e-06,
+      "loss": 0.7262,
+      "step": 28732
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.55760048828172,
+      "learning_rate": 4.992732819738028e-06,
+      "loss": 0.7608,
+      "step": 28733
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.745329916377965,
+      "learning_rate": 4.992072341833385e-06,
+      "loss": 0.7116,
+      "step": 28734
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2912943693258683,
+      "learning_rate": 4.991411893086482e-06,
+      "loss": 0.6118,
+      "step": 28735
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.375505605356802,
+      "learning_rate": 4.990751473501168e-06,
+      "loss": 0.6202,
+      "step": 28736
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1937310063943851,
+      "learning_rate": 4.990091083081285e-06,
+      "loss": 0.7633,
+      "step": 28737
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4967611210864897,
+      "learning_rate": 4.989430721830676e-06,
+      "loss": 0.6471,
+      "step": 28738
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.313346659917707,
+      "learning_rate": 4.988770389753189e-06,
+      "loss": 0.6401,
+      "step": 28739
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5656857502724466,
+      "learning_rate": 4.988110086852664e-06,
+      "loss": 0.52,
+      "step": 28740
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2624168584924094,
+      "learning_rate": 4.9874498131329495e-06,
+      "loss": 0.6672,
+      "step": 28741
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4032359008871176,
+      "learning_rate": 4.986789568597893e-06,
+      "loss": 0.6451,
+      "step": 28742
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4949705150572172,
+      "learning_rate": 4.986129353251335e-06,
+      "loss": 0.7834,
+      "step": 28743
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2980324945540602,
+      "learning_rate": 4.9854691670971145e-06,
+      "loss": 0.6954,
+      "step": 28744
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5000947490580867,
+      "learning_rate": 4.9848090101390815e-06,
+      "loss": 0.7204,
+      "step": 28745
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.266439471359335,
+      "learning_rate": 4.9841488823810805e-06,
+      "loss": 0.7272,
+      "step": 28746
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3783096054949864,
+      "learning_rate": 4.983488783826953e-06,
+      "loss": 0.6299,
+      "step": 28747
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3854038516764613,
+      "learning_rate": 4.9828287144805385e-06,
+      "loss": 0.6416,
+      "step": 28748
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2632575150554277,
+      "learning_rate": 4.9821686743456835e-06,
+      "loss": 0.6609,
+      "step": 28749
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.650125402847339,
+      "learning_rate": 4.981508663426236e-06,
+      "loss": 0.8745,
+      "step": 28750
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.270863528991978,
+      "learning_rate": 4.9808486817260314e-06,
+      "loss": 0.645,
+      "step": 28751
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5366792959394038,
+      "learning_rate": 4.980188729248913e-06,
+      "loss": 0.6789,
+      "step": 28752
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3852693948410961,
+      "learning_rate": 4.979528805998724e-06,
+      "loss": 0.6374,
+      "step": 28753
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.384345493834049,
+      "learning_rate": 4.978868911979312e-06,
+      "loss": 0.5091,
+      "step": 28754
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.368972460127659,
+      "learning_rate": 4.9782090471945155e-06,
+      "loss": 0.6855,
+      "step": 28755
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1995611013375753,
+      "learning_rate": 4.977549211648172e-06,
+      "loss": 0.5853,
+      "step": 28756
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.313699883515082,
+      "learning_rate": 4.976889405344129e-06,
+      "loss": 0.6246,
+      "step": 28757
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.943250597911863,
+      "learning_rate": 4.9762296282862245e-06,
+      "loss": 0.5871,
+      "step": 28758
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7983963264286014,
+      "learning_rate": 4.9755698804783045e-06,
+      "loss": 0.6047,
+      "step": 28759
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7956568147338885,
+      "learning_rate": 4.974910161924204e-06,
+      "loss": 0.7186,
+      "step": 28760
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4636278511903893,
+      "learning_rate": 4.974250472627771e-06,
+      "loss": 0.7351,
+      "step": 28761
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4620712816924304,
+      "learning_rate": 4.973590812592839e-06,
+      "loss": 0.6717,
+      "step": 28762
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2827186618336253,
+      "learning_rate": 4.9729311818232565e-06,
+      "loss": 0.6292,
+      "step": 28763
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5441768005949659,
+      "learning_rate": 4.97227158032286e-06,
+      "loss": 0.7118,
+      "step": 28764
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4263954863313784,
+      "learning_rate": 4.971612008095486e-06,
+      "loss": 0.7084,
+      "step": 28765
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4615017056378805,
+      "learning_rate": 4.970952465144979e-06,
+      "loss": 0.7281,
+      "step": 28766
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5487491341381558,
+      "learning_rate": 4.970292951475183e-06,
+      "loss": 0.698,
+      "step": 28767
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2628071811641,
+      "learning_rate": 4.969633467089934e-06,
+      "loss": 0.7114,
+      "step": 28768
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4987075931314606,
+      "learning_rate": 4.968974011993067e-06,
+      "loss": 0.7716,
+      "step": 28769
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3599336898483754,
+      "learning_rate": 4.968314586188427e-06,
+      "loss": 0.5784,
+      "step": 28770
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5624447017243593,
+      "learning_rate": 4.9676551896798565e-06,
+      "loss": 0.4717,
+      "step": 28771
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3829763154413917,
+      "learning_rate": 4.96699582247119e-06,
+      "loss": 0.6069,
+      "step": 28772
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.454834880044789,
+      "learning_rate": 4.9663364845662635e-06,
+      "loss": 0.8539,
+      "step": 28773
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3068807673976588,
+      "learning_rate": 4.965677175968919e-06,
+      "loss": 0.6751,
+      "step": 28774
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2760196440555678,
+      "learning_rate": 4.965017896683001e-06,
+      "loss": 0.6662,
+      "step": 28775
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5284362238280111,
+      "learning_rate": 4.9643586467123416e-06,
+      "loss": 0.6671,
+      "step": 28776
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6204604996998713,
+      "learning_rate": 4.963699426060777e-06,
+      "loss": 0.4729,
+      "step": 28777
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3257084520127451,
+      "learning_rate": 4.963040234732154e-06,
+      "loss": 0.6523,
+      "step": 28778
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4211648541240425,
+      "learning_rate": 4.9623810727303e-06,
+      "loss": 0.7256,
+      "step": 28779
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4054609225560197,
+      "learning_rate": 4.961721940059062e-06,
+      "loss": 0.7596,
+      "step": 28780
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3439507144587057,
+      "learning_rate": 4.961062836722271e-06,
+      "loss": 0.534,
+      "step": 28781
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3680998152253299,
+      "learning_rate": 4.960403762723771e-06,
+      "loss": 0.7043,
+      "step": 28782
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4618126311965558,
+      "learning_rate": 4.9597447180673916e-06,
+      "loss": 0.8366,
+      "step": 28783
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3339047786073197,
+      "learning_rate": 4.959085702756979e-06,
+      "loss": 0.6551,
+      "step": 28784
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2620520225624368,
+      "learning_rate": 4.958426716796364e-06,
+      "loss": 0.6414,
+      "step": 28785
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5721512281038204,
+      "learning_rate": 4.957767760189382e-06,
+      "loss": 0.7549,
+      "step": 28786
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5690774825155162,
+      "learning_rate": 4.957108832939872e-06,
+      "loss": 0.5472,
+      "step": 28787
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3495456422854133,
+      "learning_rate": 4.956449935051675e-06,
+      "loss": 0.7219,
+      "step": 28788
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4816759882871562,
+      "learning_rate": 4.9557910665286236e-06,
+      "loss": 0.599,
+      "step": 28789
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.504073672827601,
+      "learning_rate": 4.955132227374548e-06,
+      "loss": 0.7416,
+      "step": 28790
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3836475766072946,
+      "learning_rate": 4.954473417593292e-06,
+      "loss": 0.6747,
+      "step": 28791
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3168614092152626,
+      "learning_rate": 4.953814637188692e-06,
+      "loss": 0.6793,
+      "step": 28792
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6452704543017638,
+      "learning_rate": 4.95315588616458e-06,
+      "loss": 0.6542,
+      "step": 28793
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.306376420379643,
+      "learning_rate": 4.95249716452479e-06,
+      "loss": 0.643,
+      "step": 28794
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3019855566032217,
+      "learning_rate": 4.951838472273158e-06,
+      "loss": 0.6824,
+      "step": 28795
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3870226398940468,
+      "learning_rate": 4.951179809413526e-06,
+      "loss": 0.6393,
+      "step": 28796
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4823518116575107,
+      "learning_rate": 4.950521175949722e-06,
+      "loss": 0.811,
+      "step": 28797
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3570672432380682,
+      "learning_rate": 4.9498625718855786e-06,
+      "loss": 0.7299,
+      "step": 28798
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.276308537160101,
+      "learning_rate": 4.94920399722494e-06,
+      "loss": 0.6916,
+      "step": 28799
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.679853267957549,
+      "learning_rate": 4.948545451971629e-06,
+      "loss": 0.8132,
+      "step": 28800
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.7689192693537614,
+      "learning_rate": 4.94788693612949e-06,
+      "loss": 0.7629,
+      "step": 28801
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4730573718242106,
+      "learning_rate": 4.947228449702348e-06,
+      "loss": 0.7494,
+      "step": 28802
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5874977564049326,
+      "learning_rate": 4.946569992694047e-06,
+      "loss": 0.7167,
+      "step": 28803
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4970579237419102,
+      "learning_rate": 4.945911565108411e-06,
+      "loss": 0.8384,
+      "step": 28804
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5149786672252934,
+      "learning_rate": 4.945253166949282e-06,
+      "loss": 0.7373,
+      "step": 28805
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.643310589318182,
+      "learning_rate": 4.944594798220489e-06,
+      "loss": 0.665,
+      "step": 28806
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5785615058738021,
+      "learning_rate": 4.943936458925862e-06,
+      "loss": 0.4721,
+      "step": 28807
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6590570973977263,
+      "learning_rate": 4.943278149069237e-06,
+      "loss": 0.7219,
+      "step": 28808
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7018130443432158,
+      "learning_rate": 4.942619868654453e-06,
+      "loss": 0.8213,
+      "step": 28809
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.647130447128783,
+      "learning_rate": 4.9419616176853355e-06,
+      "loss": 0.7452,
+      "step": 28810
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.435408985429272,
+      "learning_rate": 4.941303396165716e-06,
+      "loss": 0.7256,
+      "step": 28811
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6150363313772225,
+      "learning_rate": 4.940645204099429e-06,
+      "loss": 0.6005,
+      "step": 28812
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5460083506479287,
+      "learning_rate": 4.939987041490312e-06,
+      "loss": 0.6563,
+      "step": 28813
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.600025609728117,
+      "learning_rate": 4.939328908342192e-06,
+      "loss": 0.6711,
+      "step": 28814
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3788580037233504,
+      "learning_rate": 4.938670804658897e-06,
+      "loss": 0.7306,
+      "step": 28815
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4206264861346736,
+      "learning_rate": 4.938012730444264e-06,
+      "loss": 0.6512,
+      "step": 28816
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3181606638350643,
+      "learning_rate": 4.937354685702126e-06,
+      "loss": 0.7869,
+      "step": 28817
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0410179299801836,
+      "learning_rate": 4.936696670436313e-06,
+      "loss": 0.7114,
+      "step": 28818
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5761689379636203,
+      "learning_rate": 4.93603868465065e-06,
+      "loss": 0.6829,
+      "step": 28819
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8606438604581494,
+      "learning_rate": 4.935380728348978e-06,
+      "loss": 0.7125,
+      "step": 28820
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4121260402943872,
+      "learning_rate": 4.9347228015351185e-06,
+      "loss": 0.6702,
+      "step": 28821
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.304548201226797,
+      "learning_rate": 4.934064904212911e-06,
+      "loss": 0.6591,
+      "step": 28822
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5559265422896238,
+      "learning_rate": 4.9334070363861765e-06,
+      "loss": 0.7271,
+      "step": 28823
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.662251462608918,
+      "learning_rate": 4.932749198058755e-06,
+      "loss": 0.6466,
+      "step": 28824
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6847662471853813,
+      "learning_rate": 4.932091389234467e-06,
+      "loss": 0.6374,
+      "step": 28825
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3139461694998111,
+      "learning_rate": 4.931433609917149e-06,
+      "loss": 0.6566,
+      "step": 28826
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4759771331068985,
+      "learning_rate": 4.930775860110636e-06,
+      "loss": 0.7846,
+      "step": 28827
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6356558659180895,
+      "learning_rate": 4.930118139818742e-06,
+      "loss": 0.694,
+      "step": 28828
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.259993626660355,
+      "learning_rate": 4.929460449045307e-06,
+      "loss": 0.6396,
+      "step": 28829
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5995469756524955,
+      "learning_rate": 4.928802787794158e-06,
+      "loss": 0.513,
+      "step": 28830
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4651316747819814,
+      "learning_rate": 4.9281451560691325e-06,
+      "loss": 0.6627,
+      "step": 28831
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.028608973540018,
+      "learning_rate": 4.927487553874044e-06,
+      "loss": 0.7189,
+      "step": 28832
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.436274341980864,
+      "learning_rate": 4.926829981212728e-06,
+      "loss": 0.7463,
+      "step": 28833
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4430509866400316,
+      "learning_rate": 4.9261724380890195e-06,
+      "loss": 0.644,
+      "step": 28834
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.585745722526115,
+      "learning_rate": 4.92551492450674e-06,
+      "loss": 0.7277,
+      "step": 28835
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.353884104781485,
+      "learning_rate": 4.924857440469716e-06,
+      "loss": 0.5701,
+      "step": 28836
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.311262575651332,
+      "learning_rate": 4.9241999859817794e-06,
+      "loss": 0.6591,
+      "step": 28837
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7593424074436195,
+      "learning_rate": 4.923542561046759e-06,
+      "loss": 0.8593,
+      "step": 28838
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2855996490066468,
+      "learning_rate": 4.9228851656684794e-06,
+      "loss": 0.733,
+      "step": 28839
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3621719934714296,
+      "learning_rate": 4.922227799850772e-06,
+      "loss": 0.7108,
+      "step": 28840
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5667751022731955,
+      "learning_rate": 4.921570463597464e-06,
+      "loss": 0.8344,
+      "step": 28841
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.422975848173837,
+      "learning_rate": 4.920913156912374e-06,
+      "loss": 0.5993,
+      "step": 28842
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7637649040849626,
+      "learning_rate": 4.920255879799338e-06,
+      "loss": 0.71,
+      "step": 28843
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.694363520621722,
+      "learning_rate": 4.919598632262183e-06,
+      "loss": 0.6689,
+      "step": 28844
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3602398985206456,
+      "learning_rate": 4.918941414304733e-06,
+      "loss": 0.8024,
+      "step": 28845
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3728473628155151,
+      "learning_rate": 4.918284225930812e-06,
+      "loss": 0.7171,
+      "step": 28846
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6025870480608857,
+      "learning_rate": 4.917627067144248e-06,
+      "loss": 0.7096,
+      "step": 28847
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2525061664513824,
+      "learning_rate": 4.916969937948874e-06,
+      "loss": 0.6657,
+      "step": 28848
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.337678243885278,
+      "learning_rate": 4.916312838348508e-06,
+      "loss": 0.6175,
+      "step": 28849
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5547591532500116,
+      "learning_rate": 4.915655768346975e-06,
+      "loss": 0.668,
+      "step": 28850
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4891633515124658,
+      "learning_rate": 4.914998727948105e-06,
+      "loss": 0.8029,
+      "step": 28851
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.185762711291053,
+      "learning_rate": 4.9143417171557255e-06,
+      "loss": 0.8187,
+      "step": 28852
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.585189206824038,
+      "learning_rate": 4.913684735973657e-06,
+      "loss": 0.7357,
+      "step": 28853
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3923609844633638,
+      "learning_rate": 4.913027784405724e-06,
+      "loss": 0.7548,
+      "step": 28854
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3094953803838787,
+      "learning_rate": 4.912370862455757e-06,
+      "loss": 0.6873,
+      "step": 28855
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3178652326945386,
+      "learning_rate": 4.911713970127574e-06,
+      "loss": 0.7341,
+      "step": 28856
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.603273386312977,
+      "learning_rate": 4.911057107425007e-06,
+      "loss": 0.7581,
+      "step": 28857
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7978623491762171,
+      "learning_rate": 4.910400274351873e-06,
+      "loss": 0.615,
+      "step": 28858
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2103520521260402,
+      "learning_rate": 4.909743470912004e-06,
+      "loss": 0.7215,
+      "step": 28859
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5247246925510243,
+      "learning_rate": 4.909086697109215e-06,
+      "loss": 0.6287,
+      "step": 28860
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1692260234440024,
+      "learning_rate": 4.90842995294734e-06,
+      "loss": 0.5262,
+      "step": 28861
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.241259861831362,
+      "learning_rate": 4.907773238430198e-06,
+      "loss": 0.677,
+      "step": 28862
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8506974921645793,
+      "learning_rate": 4.907116553561608e-06,
+      "loss": 0.7177,
+      "step": 28863
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2731878014840492,
+      "learning_rate": 4.906459898345397e-06,
+      "loss": 0.647,
+      "step": 28864
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4936439158700916,
+      "learning_rate": 4.905803272785394e-06,
+      "loss": 0.6749,
+      "step": 28865
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5941820904043165,
+      "learning_rate": 4.905146676885417e-06,
+      "loss": 0.7604,
+      "step": 28866
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3163704235062266,
+      "learning_rate": 4.904490110649285e-06,
+      "loss": 0.693,
+      "step": 28867
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3954602218427097,
+      "learning_rate": 4.903833574080825e-06,
+      "loss": 0.7497,
+      "step": 28868
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7410777749296178,
+      "learning_rate": 4.903177067183863e-06,
+      "loss": 0.6706,
+      "step": 28869
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4771286188159611,
+      "learning_rate": 4.902520589962218e-06,
+      "loss": 0.6789,
+      "step": 28870
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5841678229633778,
+      "learning_rate": 4.9018641424197075e-06,
+      "loss": 0.4812,
+      "step": 28871
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5580986393988363,
+      "learning_rate": 4.901207724560158e-06,
+      "loss": 0.6744,
+      "step": 28872
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2276958328815322,
+      "learning_rate": 4.900551336387395e-06,
+      "loss": 0.669,
+      "step": 28873
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.218079566835727,
+      "learning_rate": 4.899894977905239e-06,
+      "loss": 0.7226,
+      "step": 28874
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5022682323357288,
+      "learning_rate": 4.8992386491175025e-06,
+      "loss": 0.8257,
+      "step": 28875
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.334596004433571,
+      "learning_rate": 4.898582350028015e-06,
+      "loss": 0.698,
+      "step": 28876
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4931170308293376,
+      "learning_rate": 4.897926080640599e-06,
+      "loss": 0.7242,
+      "step": 28877
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3170681568861546,
+      "learning_rate": 4.897269840959072e-06,
+      "loss": 0.8103,
+      "step": 28878
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4541398533866703,
+      "learning_rate": 4.896613630987254e-06,
+      "loss": 0.6851,
+      "step": 28879
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8956294688463176,
+      "learning_rate": 4.89595745072897e-06,
+      "loss": 0.7231,
+      "step": 28880
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5896783602635344,
+      "learning_rate": 4.8953013001880324e-06,
+      "loss": 0.489,
+      "step": 28881
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.232544417072813,
+      "learning_rate": 4.894645179368272e-06,
+      "loss": 0.6116,
+      "step": 28882
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2916263301569655,
+      "learning_rate": 4.8939890882735e-06,
+      "loss": 0.6689,
+      "step": 28883
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.369930829007966,
+      "learning_rate": 4.893333026907543e-06,
+      "loss": 0.7014,
+      "step": 28884
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4756198182195006,
+      "learning_rate": 4.8926769952742145e-06,
+      "loss": 0.6189,
+      "step": 28885
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.502558021795447,
+      "learning_rate": 4.892020993377341e-06,
+      "loss": 0.7341,
+      "step": 28886
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6307975443418384,
+      "learning_rate": 4.891365021220739e-06,
+      "loss": 0.5322,
+      "step": 28887
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5927114930085522,
+      "learning_rate": 4.8907090788082225e-06,
+      "loss": 0.5371,
+      "step": 28888
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5747368048470263,
+      "learning_rate": 4.890053166143617e-06,
+      "loss": 0.874,
+      "step": 28889
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2475846286413772,
+      "learning_rate": 4.889397283230742e-06,
+      "loss": 0.6364,
+      "step": 28890
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6301822384417269,
+      "learning_rate": 4.888741430073414e-06,
+      "loss": 0.477,
+      "step": 28891
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7139435474829592,
+      "learning_rate": 4.888085606675449e-06,
+      "loss": 0.5839,
+      "step": 28892
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2918521775179472,
+      "learning_rate": 4.8874298130406665e-06,
+      "loss": 0.7636,
+      "step": 28893
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2814708295065218,
+      "learning_rate": 4.886774049172891e-06,
+      "loss": 0.7284,
+      "step": 28894
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1587129496751054,
+      "learning_rate": 4.886118315075935e-06,
+      "loss": 0.63,
+      "step": 28895
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5308075588931995,
+      "learning_rate": 4.885462610753614e-06,
+      "loss": 0.7754,
+      "step": 28896
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9180901209279413,
+      "learning_rate": 4.884806936209749e-06,
+      "loss": 0.6683,
+      "step": 28897
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3050008555836097,
+      "learning_rate": 4.884151291448161e-06,
+      "loss": 0.6104,
+      "step": 28898
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.473775859785532,
+      "learning_rate": 4.883495676472665e-06,
+      "loss": 0.7432,
+      "step": 28899
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3542926071739185,
+      "learning_rate": 4.882840091287072e-06,
+      "loss": 0.6693,
+      "step": 28900
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9084413652821715,
+      "learning_rate": 4.882184535895208e-06,
+      "loss": 0.6953,
+      "step": 28901
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2215859423717328,
+      "learning_rate": 4.881529010300882e-06,
+      "loss": 0.5197,
+      "step": 28902
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6066126411655514,
+      "learning_rate": 4.880873514507919e-06,
+      "loss": 0.5314,
+      "step": 28903
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5678852344380612,
+      "learning_rate": 4.880218048520127e-06,
+      "loss": 0.6976,
+      "step": 28904
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3938685457433004,
+      "learning_rate": 4.87956261234133e-06,
+      "loss": 0.5646,
+      "step": 28905
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2656938897691339,
+      "learning_rate": 4.878907205975336e-06,
+      "loss": 0.6159,
+      "step": 28906
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6384537786000544,
+      "learning_rate": 4.87825182942597e-06,
+      "loss": 0.7167,
+      "step": 28907
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4563378103999094,
+      "learning_rate": 4.877596482697045e-06,
+      "loss": 0.6991,
+      "step": 28908
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4989028766468213,
+      "learning_rate": 4.87694116579237e-06,
+      "loss": 0.69,
+      "step": 28909
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.780417331601687,
+      "learning_rate": 4.876285878715764e-06,
+      "loss": 0.6898,
+      "step": 28910
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5952915863349488,
+      "learning_rate": 4.875630621471045e-06,
+      "loss": 0.5164,
+      "step": 28911
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4374688276576864,
+      "learning_rate": 4.874975394062034e-06,
+      "loss": 0.7488,
+      "step": 28912
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5489716307023984,
+      "learning_rate": 4.874320196492533e-06,
+      "loss": 0.8383,
+      "step": 28913
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5972307365136021,
+      "learning_rate": 4.873665028766361e-06,
+      "loss": 0.4938,
+      "step": 28914
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5947533017023371,
+      "learning_rate": 4.873009890887338e-06,
+      "loss": 0.5681,
+      "step": 28915
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4205377063582425,
+      "learning_rate": 4.872354782859274e-06,
+      "loss": 0.7614,
+      "step": 28916
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3553919146438773,
+      "learning_rate": 4.8716997046859816e-06,
+      "loss": 0.7049,
+      "step": 28917
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3940871386829143,
+      "learning_rate": 4.871044656371276e-06,
+      "loss": 0.6958,
+      "step": 28918
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2433393454599386,
+      "learning_rate": 4.870389637918976e-06,
+      "loss": 0.7933,
+      "step": 28919
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3669831940934594,
+      "learning_rate": 4.869734649332892e-06,
+      "loss": 0.5673,
+      "step": 28920
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1272214911477596,
+      "learning_rate": 4.8690796906168334e-06,
+      "loss": 0.5242,
+      "step": 28921
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5005576287303721,
+      "learning_rate": 4.868424761774622e-06,
+      "loss": 0.6873,
+      "step": 28922
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3806242091399863,
+      "learning_rate": 4.867769862810061e-06,
+      "loss": 0.8788,
+      "step": 28923
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4202599876150581,
+      "learning_rate": 4.8671149937269694e-06,
+      "loss": 0.6082,
+      "step": 28924
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.27862457002398,
+      "learning_rate": 4.866460154529163e-06,
+      "loss": 0.6798,
+      "step": 28925
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6543193560285818,
+      "learning_rate": 4.86580534522045e-06,
+      "loss": 0.6446,
+      "step": 28926
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3346319849182726,
+      "learning_rate": 4.86515056580464e-06,
+      "loss": 0.5989,
+      "step": 28927
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.92894210378497,
+      "learning_rate": 4.864495816285549e-06,
+      "loss": 0.7431,
+      "step": 28928
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.459791025041198,
+      "learning_rate": 4.863841096666998e-06,
+      "loss": 0.8013,
+      "step": 28929
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5399954018756656,
+      "learning_rate": 4.863186406952782e-06,
+      "loss": 0.7747,
+      "step": 28930
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9395928075666484,
+      "learning_rate": 4.862531747146722e-06,
+      "loss": 0.601,
+      "step": 28931
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2447814524937986,
+      "learning_rate": 4.861877117252627e-06,
+      "loss": 0.6647,
+      "step": 28932
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4349272712152057,
+      "learning_rate": 4.861222517274319e-06,
+      "loss": 0.6748,
+      "step": 28933
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.21148083553729,
+      "learning_rate": 4.860567947215592e-06,
+      "loss": 0.6567,
+      "step": 28934
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3511233566946135,
+      "learning_rate": 4.8599134070802656e-06,
+      "loss": 0.678,
+      "step": 28935
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4395762627573487,
+      "learning_rate": 4.859258896872155e-06,
+      "loss": 0.6961,
+      "step": 28936
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4181587795029433,
+      "learning_rate": 4.858604416595063e-06,
+      "loss": 0.7263,
+      "step": 28937
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.063431183247304,
+      "learning_rate": 4.857949966252806e-06,
+      "loss": 0.5697,
+      "step": 28938
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1249906922527813,
+      "learning_rate": 4.85729554584919e-06,
+      "loss": 0.5629,
+      "step": 28939
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4349816344847908,
+      "learning_rate": 4.856641155388031e-06,
+      "loss": 0.6313,
+      "step": 28940
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3820299828140212,
+      "learning_rate": 4.8559867948731325e-06,
+      "loss": 0.5397,
+      "step": 28941
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4660616561805795,
+      "learning_rate": 4.85533246430831e-06,
+      "loss": 0.7924,
+      "step": 28942
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3624925282335425,
+      "learning_rate": 4.854678163697371e-06,
+      "loss": 0.5962,
+      "step": 28943
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.494532800961593,
+      "learning_rate": 4.854023893044121e-06,
+      "loss": 0.6896,
+      "step": 28944
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3746137313100968,
+      "learning_rate": 4.853369652352373e-06,
+      "loss": 0.7357,
+      "step": 28945
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3489892838059785,
+      "learning_rate": 4.85271544162594e-06,
+      "loss": 0.6298,
+      "step": 28946
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5972574703696301,
+      "learning_rate": 4.8520612608686265e-06,
+      "loss": 0.508,
+      "step": 28947
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6086659049228331,
+      "learning_rate": 4.8514071100842395e-06,
+      "loss": 0.4888,
+      "step": 28948
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3651105328352833,
+      "learning_rate": 4.85075298927659e-06,
+      "loss": 0.678,
+      "step": 28949
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6850848261009452,
+      "learning_rate": 4.850098898449491e-06,
+      "loss": 0.6405,
+      "step": 28950
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4753177511908389,
+      "learning_rate": 4.849444837606747e-06,
+      "loss": 0.7307,
+      "step": 28951
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.903481073245615,
+      "learning_rate": 4.848790806752162e-06,
+      "loss": 0.6974,
+      "step": 28952
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.294807403715141,
+      "learning_rate": 4.848136805889546e-06,
+      "loss": 0.677,
+      "step": 28953
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0203375745060286,
+      "learning_rate": 4.8474828350227135e-06,
+      "loss": 0.7743,
+      "step": 28954
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5404248874965454,
+      "learning_rate": 4.846828894155468e-06,
+      "loss": 0.6873,
+      "step": 28955
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4041703754382715,
+      "learning_rate": 4.8461749832916104e-06,
+      "loss": 0.7366,
+      "step": 28956
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5800156157762979,
+      "learning_rate": 4.845521102434959e-06,
+      "loss": 0.5283,
+      "step": 28957
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4487118300146005,
+      "learning_rate": 4.844867251589311e-06,
+      "loss": 0.5895,
+      "step": 28958
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3085683635461696,
+      "learning_rate": 4.844213430758483e-06,
+      "loss": 0.7038,
+      "step": 28959
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.320104932904526,
+      "learning_rate": 4.8435596399462725e-06,
+      "loss": 0.7764,
+      "step": 28960
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5449271035373426,
+      "learning_rate": 4.842905879156493e-06,
+      "loss": 0.6523,
+      "step": 28961
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4672214735729987,
+      "learning_rate": 4.842252148392945e-06,
+      "loss": 0.698,
+      "step": 28962
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2320909526663035,
+      "learning_rate": 4.841598447659443e-06,
+      "loss": 0.5963,
+      "step": 28963
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5770903679621155,
+      "learning_rate": 4.840944776959786e-06,
+      "loss": 0.8063,
+      "step": 28964
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3914813411590596,
+      "learning_rate": 4.840291136297779e-06,
+      "loss": 0.7659,
+      "step": 28965
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8027701927897426,
+      "learning_rate": 4.83963752567723e-06,
+      "loss": 0.6782,
+      "step": 28966
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.354363602166582,
+      "learning_rate": 4.838983945101949e-06,
+      "loss": 0.6908,
+      "step": 28967
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4090520817366936,
+      "learning_rate": 4.8383303945757374e-06,
+      "loss": 0.7357,
+      "step": 28968
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4922959812692345,
+      "learning_rate": 4.837676874102397e-06,
+      "loss": 0.7916,
+      "step": 28969
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4800872805683913,
+      "learning_rate": 4.837023383685736e-06,
+      "loss": 0.6788,
+      "step": 28970
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2855159086265184,
+      "learning_rate": 4.836369923329564e-06,
+      "loss": 0.6125,
+      "step": 28971
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5058375069523056,
+      "learning_rate": 4.83571649303768e-06,
+      "loss": 0.6842,
+      "step": 28972
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3607783562881255,
+      "learning_rate": 4.835063092813886e-06,
+      "loss": 0.6861,
+      "step": 28973
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.369370403370052,
+      "learning_rate": 4.83440972266199e-06,
+      "loss": 0.5957,
+      "step": 28974
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2291716457056956,
+      "learning_rate": 4.833756382585799e-06,
+      "loss": 0.7621,
+      "step": 28975
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.29936156387946,
+      "learning_rate": 4.8331030725891145e-06,
+      "loss": 0.7735,
+      "step": 28976
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7809144136485116,
+      "learning_rate": 4.832449792675737e-06,
+      "loss": 0.7447,
+      "step": 28977
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3151164649370732,
+      "learning_rate": 4.831796542849476e-06,
+      "loss": 0.6681,
+      "step": 28978
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.148641919225582,
+      "learning_rate": 4.831143323114127e-06,
+      "loss": 0.6752,
+      "step": 28979
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3051647623383165,
+      "learning_rate": 4.8304901334735035e-06,
+      "loss": 0.6734,
+      "step": 28980
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5570004655127487,
+      "learning_rate": 4.8298369739313975e-06,
+      "loss": 0.6175,
+      "step": 28981
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.457425072568817,
+      "learning_rate": 4.829183844491623e-06,
+      "loss": 0.7584,
+      "step": 28982
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4251385977758468,
+      "learning_rate": 4.828530745157972e-06,
+      "loss": 0.6513,
+      "step": 28983
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3497228033498634,
+      "learning_rate": 4.827877675934256e-06,
+      "loss": 0.584,
+      "step": 28984
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5940653879912303,
+      "learning_rate": 4.827224636824275e-06,
+      "loss": 0.6941,
+      "step": 28985
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3747793703642255,
+      "learning_rate": 4.826571627831824e-06,
+      "loss": 0.7306,
+      "step": 28986
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.815181250036409,
+      "learning_rate": 4.825918648960712e-06,
+      "loss": 0.686,
+      "step": 28987
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3723665379115093,
+      "learning_rate": 4.825265700214744e-06,
+      "loss": 0.7194,
+      "step": 28988
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4259661675324806,
+      "learning_rate": 4.824612781597717e-06,
+      "loss": 0.7142,
+      "step": 28989
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2303405850959985,
+      "learning_rate": 4.823959893113429e-06,
+      "loss": 0.7547,
+      "step": 28990
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.694114141434383,
+      "learning_rate": 4.823307034765686e-06,
+      "loss": 0.7503,
+      "step": 28991
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4924688727630782,
+      "learning_rate": 4.822654206558292e-06,
+      "loss": 0.823,
+      "step": 28992
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5623046916660106,
+      "learning_rate": 4.822001408495043e-06,
+      "loss": 0.4634,
+      "step": 28993
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5578744704687242,
+      "learning_rate": 4.821348640579738e-06,
+      "loss": 0.7027,
+      "step": 28994
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4014569433640311,
+      "learning_rate": 4.820695902816181e-06,
+      "loss": 0.8183,
+      "step": 28995
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4044213345112346,
+      "learning_rate": 4.820043195208176e-06,
+      "loss": 0.7411,
+      "step": 28996
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3584287452239105,
+      "learning_rate": 4.819390517759519e-06,
+      "loss": 0.7543,
+      "step": 28997
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5957143803880639,
+      "learning_rate": 4.818737870474007e-06,
+      "loss": 0.5638,
+      "step": 28998
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3549698639145393,
+      "learning_rate": 4.818085253355443e-06,
+      "loss": 0.7814,
+      "step": 28999
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3206874243927202,
+      "learning_rate": 4.8174326664076324e-06,
+      "loss": 0.6103,
+      "step": 29000
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6079262516498811,
+      "learning_rate": 4.816780109634368e-06,
+      "loss": 0.613,
+      "step": 29001
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.358310742353366,
+      "learning_rate": 4.816127583039448e-06,
+      "loss": 0.6548,
+      "step": 29002
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5902641049405761,
+      "learning_rate": 4.8154750866266785e-06,
+      "loss": 0.5046,
+      "step": 29003
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0419707598883474,
+      "learning_rate": 4.814822620399849e-06,
+      "loss": 0.6326,
+      "step": 29004
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.404081278128425,
+      "learning_rate": 4.81417018436277e-06,
+      "loss": 0.6371,
+      "step": 29005
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6025730630199693,
+      "learning_rate": 4.8135177785192275e-06,
+      "loss": 0.5534,
+      "step": 29006
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5611708571890182,
+      "learning_rate": 4.812865402873032e-06,
+      "loss": 0.6652,
+      "step": 29007
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5729001804648667,
+      "learning_rate": 4.812213057427972e-06,
+      "loss": 0.509,
+      "step": 29008
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6207276309134265,
+      "learning_rate": 4.811560742187851e-06,
+      "loss": 0.557,
+      "step": 29009
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4135950789658351,
+      "learning_rate": 4.810908457156472e-06,
+      "loss": 0.8182,
+      "step": 29010
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.344032293400959,
+      "learning_rate": 4.810256202337621e-06,
+      "loss": 0.6946,
+      "step": 29011
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4309243573569852,
+      "learning_rate": 4.8096039777351e-06,
+      "loss": 0.6122,
+      "step": 29012
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.5042633280538906,
+      "learning_rate": 4.808951783352709e-06,
+      "loss": 0.7571,
+      "step": 29013
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.440719171980786,
+      "learning_rate": 4.808299619194251e-06,
+      "loss": 0.6676,
+      "step": 29014
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3547208517749048,
+      "learning_rate": 4.807647485263508e-06,
+      "loss": 0.8353,
+      "step": 29015
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.825884178399604,
+      "learning_rate": 4.8069953815642865e-06,
+      "loss": 0.6928,
+      "step": 29016
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2820133989347766,
+      "learning_rate": 4.806343308100385e-06,
+      "loss": 0.6155,
+      "step": 29017
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4046894714698963,
+      "learning_rate": 4.8056912648755975e-06,
+      "loss": 0.6679,
+      "step": 29018
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.398373983440168,
+      "learning_rate": 4.805039251893716e-06,
+      "loss": 0.8297,
+      "step": 29019
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5199267562399443,
+      "learning_rate": 4.80438726915854e-06,
+      "loss": 0.749,
+      "step": 29020
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6670866604284187,
+      "learning_rate": 4.803735316673871e-06,
+      "loss": 0.8081,
+      "step": 29021
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6809489893657188,
+      "learning_rate": 4.803083394443497e-06,
+      "loss": 0.713,
+      "step": 29022
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4499681986913664,
+      "learning_rate": 4.802431502471219e-06,
+      "loss": 0.6991,
+      "step": 29023
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4709632270192339,
+      "learning_rate": 4.801779640760831e-06,
+      "loss": 0.6752,
+      "step": 29024
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4406770935487838,
+      "learning_rate": 4.801127809316123e-06,
+      "loss": 0.6313,
+      "step": 29025
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4776065572699637,
+      "learning_rate": 4.8004760081408966e-06,
+      "loss": 0.7216,
+      "step": 29026
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4221299743035039,
+      "learning_rate": 4.799824237238947e-06,
+      "loss": 0.8052,
+      "step": 29027
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.487940444311986,
+      "learning_rate": 4.7991724966140674e-06,
+      "loss": 0.7938,
+      "step": 29028
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4910178120637252,
+      "learning_rate": 4.798520786270049e-06,
+      "loss": 0.631,
+      "step": 29029
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5518047020041452,
+      "learning_rate": 4.79786910621069e-06,
+      "loss": 0.8534,
+      "step": 29030
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2857462200926386,
+      "learning_rate": 4.79721745643979e-06,
+      "loss": 0.6657,
+      "step": 29031
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3928479752269864,
+      "learning_rate": 4.79656583696113e-06,
+      "loss": 0.689,
+      "step": 29032
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.077705224061478,
+      "learning_rate": 4.795914247778511e-06,
+      "loss": 0.8424,
+      "step": 29033
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5941060907222446,
+      "learning_rate": 4.795262688895728e-06,
+      "loss": 0.7273,
+      "step": 29034
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4192317232216054,
+      "learning_rate": 4.794611160316576e-06,
+      "loss": 0.7299,
+      "step": 29035
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5180175299420378,
+      "learning_rate": 4.793959662044846e-06,
+      "loss": 0.586,
+      "step": 29036
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4570352245797062,
+      "learning_rate": 4.793308194084327e-06,
+      "loss": 0.6671,
+      "step": 29037
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1662807320635937,
+      "learning_rate": 4.792656756438821e-06,
+      "loss": 0.6393,
+      "step": 29038
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3751176024595908,
+      "learning_rate": 4.7920053491121114e-06,
+      "loss": 0.8111,
+      "step": 29039
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3755912564692545,
+      "learning_rate": 4.791353972107999e-06,
+      "loss": 0.662,
+      "step": 29040
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4097912200277372,
+      "learning_rate": 4.790702625430269e-06,
+      "loss": 0.6429,
+      "step": 29041
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5480552920583848,
+      "learning_rate": 4.7900513090827225e-06,
+      "loss": 0.8178,
+      "step": 29042
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7291105601784529,
+      "learning_rate": 4.789400023069142e-06,
+      "loss": 0.6621,
+      "step": 29043
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1995183202201736,
+      "learning_rate": 4.788748767393328e-06,
+      "loss": 0.751,
+      "step": 29044
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0728623922680853,
+      "learning_rate": 4.788097542059069e-06,
+      "loss": 0.6315,
+      "step": 29045
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4528229088528757,
+      "learning_rate": 4.787446347070152e-06,
+      "loss": 0.8356,
+      "step": 29046
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2928756851962722,
+      "learning_rate": 4.7867951824303725e-06,
+      "loss": 0.6865,
+      "step": 29047
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3375324232011874,
+      "learning_rate": 4.786144048143526e-06,
+      "loss": 0.632,
+      "step": 29048
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5186512370460756,
+      "learning_rate": 4.785492944213399e-06,
+      "loss": 0.8136,
+      "step": 29049
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2029631815933404,
+      "learning_rate": 4.784841870643781e-06,
+      "loss": 0.6338,
+      "step": 29050
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4813785230764445,
+      "learning_rate": 4.784190827438462e-06,
+      "loss": 0.768,
+      "step": 29051
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3998865955404693,
+      "learning_rate": 4.783539814601242e-06,
+      "loss": 0.7972,
+      "step": 29052
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.530255187843021,
+      "learning_rate": 4.782888832135903e-06,
+      "loss": 0.7524,
+      "step": 29053
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3953058592796255,
+      "learning_rate": 4.782237880046232e-06,
+      "loss": 0.6867,
+      "step": 29054
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4386021218522578,
+      "learning_rate": 4.781586958336025e-06,
+      "loss": 0.6842,
+      "step": 29055
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.48861425425501,
+      "learning_rate": 4.780936067009075e-06,
+      "loss": 0.8068,
+      "step": 29056
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4085443918313967,
+      "learning_rate": 4.7802852060691676e-06,
+      "loss": 0.5996,
+      "step": 29057
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2632804193627412,
+      "learning_rate": 4.779634375520088e-06,
+      "loss": 0.6606,
+      "step": 29058
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1922085046480204,
+      "learning_rate": 4.778983575365634e-06,
+      "loss": 0.6359,
+      "step": 29059
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2601845255497948,
+      "learning_rate": 4.778332805609587e-06,
+      "loss": 0.6308,
+      "step": 29060
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.518530140173165,
+      "learning_rate": 4.777682066255743e-06,
+      "loss": 0.8359,
+      "step": 29061
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4311452741661435,
+      "learning_rate": 4.777031357307884e-06,
+      "loss": 0.7497,
+      "step": 29062
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4893852017346911,
+      "learning_rate": 4.776380678769805e-06,
+      "loss": 0.7514,
+      "step": 29063
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.828691430882495,
+      "learning_rate": 4.775730030645288e-06,
+      "loss": 0.7734,
+      "step": 29064
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2355589202697326,
+      "learning_rate": 4.7750794129381285e-06,
+      "loss": 0.7066,
+      "step": 29065
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2445106951495637,
+      "learning_rate": 4.774428825652111e-06,
+      "loss": 0.6902,
+      "step": 29066
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6534828497918457,
+      "learning_rate": 4.77377826879102e-06,
+      "loss": 0.7709,
+      "step": 29067
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.6211264004504513,
+      "learning_rate": 4.773127742358645e-06,
+      "loss": 0.7416,
+      "step": 29068
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3490543962619637,
+      "learning_rate": 4.772477246358781e-06,
+      "loss": 0.6426,
+      "step": 29069
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.019983067293188,
+      "learning_rate": 4.771826780795208e-06,
+      "loss": 0.7561,
+      "step": 29070
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.611983794413432,
+      "learning_rate": 4.771176345671711e-06,
+      "loss": 0.7456,
+      "step": 29071
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6992471191590799,
+      "learning_rate": 4.770525940992081e-06,
+      "loss": 0.6563,
+      "step": 29072
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5756966670150498,
+      "learning_rate": 4.769875566760108e-06,
+      "loss": 0.6016,
+      "step": 29073
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6809498070331028,
+      "learning_rate": 4.769225222979575e-06,
+      "loss": 0.6461,
+      "step": 29074
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4433776934921783,
+      "learning_rate": 4.768574909654267e-06,
+      "loss": 0.6656,
+      "step": 29075
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0433124069623863,
+      "learning_rate": 4.76792462678797e-06,
+      "loss": 0.7976,
+      "step": 29076
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4816362985780493,
+      "learning_rate": 4.767274374384476e-06,
+      "loss": 0.7155,
+      "step": 29077
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4489727408108637,
+      "learning_rate": 4.766624152447569e-06,
+      "loss": 0.5445,
+      "step": 29078
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.469822306253079,
+      "learning_rate": 4.765973960981029e-06,
+      "loss": 0.5851,
+      "step": 29079
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3557951331213385,
+      "learning_rate": 4.765323799988648e-06,
+      "loss": 0.6038,
+      "step": 29080
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.324510584502085,
+      "learning_rate": 4.764673669474206e-06,
+      "loss": 0.6529,
+      "step": 29081
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6076346276885177,
+      "learning_rate": 4.764023569441495e-06,
+      "loss": 0.5307,
+      "step": 29082
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5574519936414042,
+      "learning_rate": 4.763373499894293e-06,
+      "loss": 0.8384,
+      "step": 29083
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5618970799986707,
+      "learning_rate": 4.762723460836392e-06,
+      "loss": 0.6815,
+      "step": 29084
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5589828758698437,
+      "learning_rate": 4.762073452271569e-06,
+      "loss": 0.4738,
+      "step": 29085
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5996466323724674,
+      "learning_rate": 4.761423474203617e-06,
+      "loss": 0.7039,
+      "step": 29086
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3515664928592979,
+      "learning_rate": 4.760773526636315e-06,
+      "loss": 0.6485,
+      "step": 29087
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7258056139338538,
+      "learning_rate": 4.760123609573446e-06,
+      "loss": 0.6988,
+      "step": 29088
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.618422122804252,
+      "learning_rate": 4.759473723018796e-06,
+      "loss": 0.6477,
+      "step": 29089
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6230368908864545,
+      "learning_rate": 4.758823866976152e-06,
+      "loss": 0.7091,
+      "step": 29090
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.175002646082361,
+      "learning_rate": 4.758174041449295e-06,
+      "loss": 0.6555,
+      "step": 29091
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4448059845286538,
+      "learning_rate": 4.7575242464420045e-06,
+      "loss": 0.8679,
+      "step": 29092
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3263937809642985,
+      "learning_rate": 4.756874481958068e-06,
+      "loss": 0.7451,
+      "step": 29093
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.402146007517653,
+      "learning_rate": 4.756224748001272e-06,
+      "loss": 0.7994,
+      "step": 29094
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.548161460014661,
+      "learning_rate": 4.755575044575396e-06,
+      "loss": 0.6589,
+      "step": 29095
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3409931455079573,
+      "learning_rate": 4.754925371684217e-06,
+      "loss": 0.5947,
+      "step": 29096
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4221958812178606,
+      "learning_rate": 4.754275729331524e-06,
+      "loss": 0.5304,
+      "step": 29097
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.396060188557615,
+      "learning_rate": 4.753626117521103e-06,
+      "loss": 0.8143,
+      "step": 29098
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5711376397817445,
+      "learning_rate": 4.752976536256731e-06,
+      "loss": 0.607,
+      "step": 29099
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1444339571241484,
+      "learning_rate": 4.752326985542187e-06,
+      "loss": 0.6818,
+      "step": 29100
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3635247156569263,
+      "learning_rate": 4.75167746538126e-06,
+      "loss": 0.776,
+      "step": 29101
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.5717213778257,
+      "learning_rate": 4.751027975777726e-06,
+      "loss": 0.7864,
+      "step": 29102
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.370305297572773,
+      "learning_rate": 4.750378516735372e-06,
+      "loss": 0.7289,
+      "step": 29103
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.333201439114389,
+      "learning_rate": 4.7497290882579724e-06,
+      "loss": 0.6156,
+      "step": 29104
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3840634910268392,
+      "learning_rate": 4.749079690349316e-06,
+      "loss": 0.6336,
+      "step": 29105
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3370532064765661,
+      "learning_rate": 4.748430323013176e-06,
+      "loss": 0.8079,
+      "step": 29106
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3904160970270034,
+      "learning_rate": 4.747780986253337e-06,
+      "loss": 0.7751,
+      "step": 29107
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3825673220874923,
+      "learning_rate": 4.747131680073589e-06,
+      "loss": 0.7443,
+      "step": 29108
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2253464534151928,
+      "learning_rate": 4.746482404477695e-06,
+      "loss": 0.615,
+      "step": 29109
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.842935051018127,
+      "learning_rate": 4.745833159469443e-06,
+      "loss": 0.7103,
+      "step": 29110
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.435538256198245,
+      "learning_rate": 4.745183945052615e-06,
+      "loss": 0.6764,
+      "step": 29111
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3803216605159583,
+      "learning_rate": 4.744534761230996e-06,
+      "loss": 0.5957,
+      "step": 29112
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3918402770007243,
+      "learning_rate": 4.7438856080083525e-06,
+      "loss": 0.6521,
+      "step": 29113
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3930293215138618,
+      "learning_rate": 4.743236485388471e-06,
+      "loss": 0.6511,
+      "step": 29114
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6736888337210976,
+      "learning_rate": 4.742587393375131e-06,
+      "loss": 0.7398,
+      "step": 29115
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3304593991638678,
+      "learning_rate": 4.741938331972119e-06,
+      "loss": 0.6397,
+      "step": 29116
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5042119030923642,
+      "learning_rate": 4.7412893011831995e-06,
+      "loss": 0.6908,
+      "step": 29117
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1535527085407775,
+      "learning_rate": 4.7406403010121595e-06,
+      "loss": 0.5847,
+      "step": 29118
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.380642993639078,
+      "learning_rate": 4.7399913314627785e-06,
+      "loss": 0.7913,
+      "step": 29119
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6614802276495344,
+      "learning_rate": 4.739342392538831e-06,
+      "loss": 0.813,
+      "step": 29120
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3476658552326797,
+      "learning_rate": 4.738693484244101e-06,
+      "loss": 0.6904,
+      "step": 29121
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4129851045516129,
+      "learning_rate": 4.738044606582359e-06,
+      "loss": 0.8026,
+      "step": 29122
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3195165776848135,
+      "learning_rate": 4.737395759557392e-06,
+      "loss": 0.715,
+      "step": 29123
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.488458818207237,
+      "learning_rate": 4.736746943172969e-06,
+      "loss": 0.6558,
+      "step": 29124
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3610140683730203,
+      "learning_rate": 4.736098157432874e-06,
+      "loss": 0.7042,
+      "step": 29125
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.28496663890265,
+      "learning_rate": 4.735449402340882e-06,
+      "loss": 0.7735,
+      "step": 29126
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1199119955265728,
+      "learning_rate": 4.7348006779007685e-06,
+      "loss": 0.5123,
+      "step": 29127
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4270689430433174,
+      "learning_rate": 4.734151984116311e-06,
+      "loss": 0.6698,
+      "step": 29128
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3291808000352499,
+      "learning_rate": 4.733503320991292e-06,
+      "loss": 0.6683,
+      "step": 29129
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2401477529447122,
+      "learning_rate": 4.732854688529484e-06,
+      "loss": 0.5987,
+      "step": 29130
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3273478990600331,
+      "learning_rate": 4.73220608673466e-06,
+      "loss": 0.5785,
+      "step": 29131
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5835845884234891,
+      "learning_rate": 4.7315575156106e-06,
+      "loss": 0.6726,
+      "step": 29132
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2869549694823215,
+      "learning_rate": 4.730908975161083e-06,
+      "loss": 0.6687,
+      "step": 29133
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.788576041564226,
+      "learning_rate": 4.730260465389882e-06,
+      "loss": 0.6575,
+      "step": 29134
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.605268954664034,
+      "learning_rate": 4.729611986300771e-06,
+      "loss": 0.5567,
+      "step": 29135
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3399403658955042,
+      "learning_rate": 4.728963537897526e-06,
+      "loss": 0.7491,
+      "step": 29136
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0955201445424967,
+      "learning_rate": 4.728315120183928e-06,
+      "loss": 0.605,
+      "step": 29137
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4865114503365755,
+      "learning_rate": 4.727666733163748e-06,
+      "loss": 0.6952,
+      "step": 29138
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5498670450511467,
+      "learning_rate": 4.7270183768407585e-06,
+      "loss": 0.7175,
+      "step": 29139
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5785998612280615,
+      "learning_rate": 4.726370051218741e-06,
+      "loss": 0.5147,
+      "step": 29140
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8902072981776645,
+      "learning_rate": 4.725721756301463e-06,
+      "loss": 0.7765,
+      "step": 29141
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.429534198477086,
+      "learning_rate": 4.725073492092707e-06,
+      "loss": 0.8159,
+      "step": 29142
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4825357558914307,
+      "learning_rate": 4.724425258596238e-06,
+      "loss": 0.6806,
+      "step": 29143
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4681881990300376,
+      "learning_rate": 4.72377705581584e-06,
+      "loss": 0.714,
+      "step": 29144
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2912694603308479,
+      "learning_rate": 4.723128883755279e-06,
+      "loss": 0.5825,
+      "step": 29145
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.363212561665938,
+      "learning_rate": 4.722480742418335e-06,
+      "loss": 0.689,
+      "step": 29146
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.392070217942904,
+      "learning_rate": 4.721832631808781e-06,
+      "loss": 0.7427,
+      "step": 29147
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3405661315212856,
+      "learning_rate": 4.721184551930382e-06,
+      "loss": 0.7484,
+      "step": 29148
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3294412337105885,
+      "learning_rate": 4.720536502786919e-06,
+      "loss": 0.6732,
+      "step": 29149
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1054765753205826,
+      "learning_rate": 4.719888484382168e-06,
+      "loss": 0.6586,
+      "step": 29150
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3282196715040702,
+      "learning_rate": 4.719240496719897e-06,
+      "loss": 0.6507,
+      "step": 29151
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.416050782522555,
+      "learning_rate": 4.718592539803877e-06,
+      "loss": 0.6772,
+      "step": 29152
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3768394072498544,
+      "learning_rate": 4.717944613637883e-06,
+      "loss": 0.6602,
+      "step": 29153
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5042986100896671,
+      "learning_rate": 4.717296718225691e-06,
+      "loss": 0.7315,
+      "step": 29154
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.256212424321999,
+      "learning_rate": 4.7166488535710696e-06,
+      "loss": 0.7687,
+      "step": 29155
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6246969250846324,
+      "learning_rate": 4.716001019677788e-06,
+      "loss": 0.7146,
+      "step": 29156
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.44599963077313,
+      "learning_rate": 4.715353216549622e-06,
+      "loss": 0.7224,
+      "step": 29157
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6198553322576089,
+      "learning_rate": 4.714705444190346e-06,
+      "loss": 0.5295,
+      "step": 29158
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5902870178330633,
+      "learning_rate": 4.714057702603728e-06,
+      "loss": 0.7661,
+      "step": 29159
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9994149976274869,
+      "learning_rate": 4.713409991793536e-06,
+      "loss": 0.7461,
+      "step": 29160
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5248280893652233,
+      "learning_rate": 4.712762311763549e-06,
+      "loss": 0.702,
+      "step": 29161
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2225294788449241,
+      "learning_rate": 4.71211466251753e-06,
+      "loss": 0.6595,
+      "step": 29162
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3124314723275294,
+      "learning_rate": 4.711467044059256e-06,
+      "loss": 0.7302,
+      "step": 29163
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.551129781218192,
+      "learning_rate": 4.710819456392492e-06,
+      "loss": 0.6631,
+      "step": 29164
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4400893493538152,
+      "learning_rate": 4.710171899521015e-06,
+      "loss": 0.6161,
+      "step": 29165
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7841291856733754,
+      "learning_rate": 4.709524373448589e-06,
+      "loss": 0.7374,
+      "step": 29166
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4419839895161792,
+      "learning_rate": 4.708876878178991e-06,
+      "loss": 0.7304,
+      "step": 29167
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6272687675923825,
+      "learning_rate": 4.708229413715987e-06,
+      "loss": 0.6456,
+      "step": 29168
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4096349050649948,
+      "learning_rate": 4.707581980063342e-06,
+      "loss": 0.6288,
+      "step": 29169
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5012727129133323,
+      "learning_rate": 4.706934577224831e-06,
+      "loss": 0.8317,
+      "step": 29170
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6617678643446568,
+      "learning_rate": 4.706287205204225e-06,
+      "loss": 0.5084,
+      "step": 29171
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6455540025684592,
+      "learning_rate": 4.705639864005292e-06,
+      "loss": 0.8029,
+      "step": 29172
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.328012120430031,
+      "learning_rate": 4.704992553631796e-06,
+      "loss": 0.6819,
+      "step": 29173
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1130507266406355,
+      "learning_rate": 4.704345274087511e-06,
+      "loss": 0.7726,
+      "step": 29174
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4483462222540113,
+      "learning_rate": 4.703698025376207e-06,
+      "loss": 0.6225,
+      "step": 29175
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3723212143810004,
+      "learning_rate": 4.703050807501649e-06,
+      "loss": 0.6385,
+      "step": 29176
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.46436177725643,
+      "learning_rate": 4.702403620467604e-06,
+      "loss": 0.6948,
+      "step": 29177
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3479188506540962,
+      "learning_rate": 4.701756464277841e-06,
+      "loss": 0.7286,
+      "step": 29178
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.294269788646955,
+      "learning_rate": 4.701109338936134e-06,
+      "loss": 0.6168,
+      "step": 29179
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2848486153598688,
+      "learning_rate": 4.700462244446246e-06,
+      "loss": 0.6834,
+      "step": 29180
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.350236498831391,
+      "learning_rate": 4.699815180811939e-06,
+      "loss": 0.7614,
+      "step": 29181
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4443036015678636,
+      "learning_rate": 4.699168148036992e-06,
+      "loss": 0.7243,
+      "step": 29182
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6951305092598559,
+      "learning_rate": 4.698521146125161e-06,
+      "loss": 0.6238,
+      "step": 29183
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.253808407821702,
+      "learning_rate": 4.697874175080224e-06,
+      "loss": 0.6637,
+      "step": 29184
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5035878450699376,
+      "learning_rate": 4.6972272349059365e-06,
+      "loss": 0.6366,
+      "step": 29185
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4050166768216514,
+      "learning_rate": 4.696580325606075e-06,
+      "loss": 0.7019,
+      "step": 29186
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4033682223788648,
+      "learning_rate": 4.695933447184398e-06,
+      "loss": 0.6235,
+      "step": 29187
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.349705371217143,
+      "learning_rate": 4.695286599644681e-06,
+      "loss": 0.6848,
+      "step": 29188
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4353524462008354,
+      "learning_rate": 4.694639782990683e-06,
+      "loss": 0.6582,
+      "step": 29189
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5985242858529989,
+      "learning_rate": 4.69399299722617e-06,
+      "loss": 0.7415,
+      "step": 29190
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4102528754467067,
+      "learning_rate": 4.693346242354908e-06,
+      "loss": 0.6828,
+      "step": 29191
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6582583411145282,
+      "learning_rate": 4.692699518380664e-06,
+      "loss": 0.7321,
+      "step": 29192
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4711135103928976,
+      "learning_rate": 4.692052825307213e-06,
+      "loss": 0.6949,
+      "step": 29193
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5525371189018435,
+      "learning_rate": 4.691406163138302e-06,
+      "loss": 0.7139,
+      "step": 29194
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4382340656712103,
+      "learning_rate": 4.690759531877706e-06,
+      "loss": 0.5603,
+      "step": 29195
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4395056787984604,
+      "learning_rate": 4.690112931529191e-06,
+      "loss": 0.6786,
+      "step": 29196
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7330749249281732,
+      "learning_rate": 4.689466362096521e-06,
+      "loss": 0.6087,
+      "step": 29197
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6589694145112532,
+      "learning_rate": 4.688819823583454e-06,
+      "loss": 0.5941,
+      "step": 29198
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4707617819247956,
+      "learning_rate": 4.688173315993761e-06,
+      "loss": 0.7605,
+      "step": 29199
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2530361586834027,
+      "learning_rate": 4.687526839331208e-06,
+      "loss": 0.5855,
+      "step": 29200
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2794931141339803,
+      "learning_rate": 4.686880393599556e-06,
+      "loss": 0.6485,
+      "step": 29201
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.519121593566395,
+      "learning_rate": 4.686233978802563e-06,
+      "loss": 0.7781,
+      "step": 29202
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7533335704424113,
+      "learning_rate": 4.685587594944003e-06,
+      "loss": 0.7393,
+      "step": 29203
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3271134233004906,
+      "learning_rate": 4.684941242027631e-06,
+      "loss": 0.6755,
+      "step": 29204
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.389538531948468,
+      "learning_rate": 4.684294920057214e-06,
+      "loss": 0.7548,
+      "step": 29205
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4582364586291645,
+      "learning_rate": 4.683648629036518e-06,
+      "loss": 0.6941,
+      "step": 29206
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4967912458969743,
+      "learning_rate": 4.683002368969303e-06,
+      "loss": 0.7994,
+      "step": 29207
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1912113164091782,
+      "learning_rate": 4.682356139859327e-06,
+      "loss": 0.5935,
+      "step": 29208
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3920368178491334,
+      "learning_rate": 4.681709941710358e-06,
+      "loss": 0.7239,
+      "step": 29209
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3009955063036027,
+      "learning_rate": 4.681063774526166e-06,
+      "loss": 0.6259,
+      "step": 29210
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3788293162823198,
+      "learning_rate": 4.680417638310496e-06,
+      "loss": 0.7688,
+      "step": 29211
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5075713305212388,
+      "learning_rate": 4.679771533067119e-06,
+      "loss": 0.7653,
+      "step": 29212
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.411077522921773,
+      "learning_rate": 4.679125458799796e-06,
+      "loss": 0.7075,
+      "step": 29213
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5330078944410066,
+      "learning_rate": 4.678479415512297e-06,
+      "loss": 0.768,
+      "step": 29214
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4731010550951424,
+      "learning_rate": 4.677833403208367e-06,
+      "loss": 0.7994,
+      "step": 29215
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5586806649294516,
+      "learning_rate": 4.6771874218917765e-06,
+      "loss": 0.6161,
+      "step": 29216
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2963866974562124,
+      "learning_rate": 4.676541471566291e-06,
+      "loss": 0.7394,
+      "step": 29217
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.627695237782834,
+      "learning_rate": 4.675895552235662e-06,
+      "loss": 0.4627,
+      "step": 29218
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5833104038899748,
+      "learning_rate": 4.675249663903658e-06,
+      "loss": 0.5058,
+      "step": 29219
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4476047892994814,
+      "learning_rate": 4.674603806574033e-06,
+      "loss": 0.6795,
+      "step": 29220
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5661199310406166,
+      "learning_rate": 4.673957980250553e-06,
+      "loss": 0.4488,
+      "step": 29221
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2289358086142692,
+      "learning_rate": 4.673312184936973e-06,
+      "loss": 0.6277,
+      "step": 29222
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.183919068028116,
+      "learning_rate": 4.672666420637061e-06,
+      "loss": 0.6979,
+      "step": 29223
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3806336934042904,
+      "learning_rate": 4.67202068735457e-06,
+      "loss": 0.6958,
+      "step": 29224
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3420745998360282,
+      "learning_rate": 4.6713749850932576e-06,
+      "loss": 0.5972,
+      "step": 29225
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0660198112494923,
+      "learning_rate": 4.670729313856887e-06,
+      "loss": 0.6279,
+      "step": 29226
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3183706127676786,
+      "learning_rate": 4.670083673649223e-06,
+      "loss": 0.6981,
+      "step": 29227
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.575467198644236,
+      "learning_rate": 4.669438064474018e-06,
+      "loss": 0.6186,
+      "step": 29228
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4252078256989575,
+      "learning_rate": 4.668792486335029e-06,
+      "loss": 0.5865,
+      "step": 29229
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.445066175118394,
+      "learning_rate": 4.668146939236017e-06,
+      "loss": 0.5986,
+      "step": 29230
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5713373419415934,
+      "learning_rate": 4.667501423180746e-06,
+      "loss": 0.4728,
+      "step": 29231
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6214408994754989,
+      "learning_rate": 4.666855938172969e-06,
+      "loss": 0.5211,
+      "step": 29232
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4103722712215647,
+      "learning_rate": 4.666210484216444e-06,
+      "loss": 0.5443,
+      "step": 29233
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7846852201343577,
+      "learning_rate": 4.665565061314927e-06,
+      "loss": 0.6969,
+      "step": 29234
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.606228307505145,
+      "learning_rate": 4.6649196694721845e-06,
+      "loss": 0.7828,
+      "step": 29235
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.321471137860456,
+      "learning_rate": 4.664274308691968e-06,
+      "loss": 0.6616,
+      "step": 29236
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5527376471416017,
+      "learning_rate": 4.663628978978033e-06,
+      "loss": 0.7017,
+      "step": 29237
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5952651300499271,
+      "learning_rate": 4.662983680334144e-06,
+      "loss": 0.4914,
+      "step": 29238
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4679398276008746,
+      "learning_rate": 4.662338412764048e-06,
+      "loss": 0.6254,
+      "step": 29239
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9689222530070754,
+      "learning_rate": 4.661693176271513e-06,
+      "loss": 0.7563,
+      "step": 29240
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.375973919238749,
+      "learning_rate": 4.661047970860285e-06,
+      "loss": 0.6742,
+      "step": 29241
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.3938653797608715,
+      "learning_rate": 4.66040279653413e-06,
+      "loss": 0.6256,
+      "step": 29242
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3399256769684642,
+      "learning_rate": 4.659757653296797e-06,
+      "loss": 0.6159,
+      "step": 29243
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3104496192736965,
+      "learning_rate": 4.65911254115205e-06,
+      "loss": 0.6027,
+      "step": 29244
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8521484127976677,
+      "learning_rate": 4.658467460103636e-06,
+      "loss": 0.6275,
+      "step": 29245
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.45710999341808,
+      "learning_rate": 4.657822410155321e-06,
+      "loss": 0.7404,
+      "step": 29246
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5058469840826672,
+      "learning_rate": 4.65717739131085e-06,
+      "loss": 0.7199,
+      "step": 29247
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3207674302122945,
+      "learning_rate": 4.656532403573987e-06,
+      "loss": 0.7579,
+      "step": 29248
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8573088342658364,
+      "learning_rate": 4.655887446948485e-06,
+      "loss": 0.6564,
+      "step": 29249
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4009887533243788,
+      "learning_rate": 4.655242521438095e-06,
+      "loss": 0.8026,
+      "step": 29250
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4460505276291498,
+      "learning_rate": 4.654597627046574e-06,
+      "loss": 0.7174,
+      "step": 29251
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3934608775960378,
+      "learning_rate": 4.653952763777682e-06,
+      "loss": 0.8178,
+      "step": 29252
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3178549735652918,
+      "learning_rate": 4.653307931635169e-06,
+      "loss": 0.6452,
+      "step": 29253
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6304960652152614,
+      "learning_rate": 4.652663130622788e-06,
+      "loss": 0.4648,
+      "step": 29254
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6062958765219195,
+      "learning_rate": 4.652018360744294e-06,
+      "loss": 0.568,
+      "step": 29255
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4265573305329993,
+      "learning_rate": 4.651373622003445e-06,
+      "loss": 0.6221,
+      "step": 29256
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2548851814229105,
+      "learning_rate": 4.650728914403994e-06,
+      "loss": 0.5615,
+      "step": 29257
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5810122540562952,
+      "learning_rate": 4.650084237949688e-06,
+      "loss": 0.614,
+      "step": 29258
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3987582597718744,
+      "learning_rate": 4.649439592644286e-06,
+      "loss": 0.5459,
+      "step": 29259
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2200541856091836,
+      "learning_rate": 4.648794978491543e-06,
+      "loss": 0.6615,
+      "step": 29260
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5367642096182055,
+      "learning_rate": 4.6481503954952114e-06,
+      "loss": 0.787,
+      "step": 29261
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.249824044029289,
+      "learning_rate": 4.647505843659038e-06,
+      "loss": 0.7322,
+      "step": 29262
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6757744706005346,
+      "learning_rate": 4.6468613229867845e-06,
+      "loss": 0.6168,
+      "step": 29263
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.833380023395764,
+      "learning_rate": 4.646216833482195e-06,
+      "loss": 0.6214,
+      "step": 29264
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4489680420530684,
+      "learning_rate": 4.645572375149029e-06,
+      "loss": 0.6279,
+      "step": 29265
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2667084598271263,
+      "learning_rate": 4.644927947991034e-06,
+      "loss": 0.6954,
+      "step": 29266
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6609321293745163,
+      "learning_rate": 4.644283552011965e-06,
+      "loss": 0.6548,
+      "step": 29267
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5210428424374138,
+      "learning_rate": 4.6436391872155714e-06,
+      "loss": 0.7052,
+      "step": 29268
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5341159073167856,
+      "learning_rate": 4.642994853605609e-06,
+      "loss": 0.7423,
+      "step": 29269
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6536273370472054,
+      "learning_rate": 4.642350551185827e-06,
+      "loss": 0.7055,
+      "step": 29270
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3233866554111426,
+      "learning_rate": 4.6417062799599725e-06,
+      "loss": 0.7147,
+      "step": 29271
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3987725728207228,
+      "learning_rate": 4.641062039931799e-06,
+      "loss": 0.6125,
+      "step": 29272
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2788031995751674,
+      "learning_rate": 4.640417831105064e-06,
+      "loss": 0.699,
+      "step": 29273
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2265009004459189,
+      "learning_rate": 4.6397736534835136e-06,
+      "loss": 0.6101,
+      "step": 29274
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0764444823138684,
+      "learning_rate": 4.639129507070893e-06,
+      "loss": 0.6209,
+      "step": 29275
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6027677160202667,
+      "learning_rate": 4.638485391870958e-06,
+      "loss": 0.6605,
+      "step": 29276
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5346194099001342,
+      "learning_rate": 4.637841307887463e-06,
+      "loss": 0.6473,
+      "step": 29277
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.730666896113659,
+      "learning_rate": 4.637197255124153e-06,
+      "loss": 0.7562,
+      "step": 29278
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5236485015338033,
+      "learning_rate": 4.636553233584775e-06,
+      "loss": 0.7227,
+      "step": 29279
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3158957670268359,
+      "learning_rate": 4.635909243273082e-06,
+      "loss": 0.6718,
+      "step": 29280
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7158858461653141,
+      "learning_rate": 4.635265284192827e-06,
+      "loss": 0.6574,
+      "step": 29281
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5565270395577631,
+      "learning_rate": 4.634621356347756e-06,
+      "loss": 0.5304,
+      "step": 29282
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5039593100573396,
+      "learning_rate": 4.633977459741614e-06,
+      "loss": 0.7877,
+      "step": 29283
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5097613763124846,
+      "learning_rate": 4.6333335943781586e-06,
+      "loss": 0.6372,
+      "step": 29284
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.503425734291761,
+      "learning_rate": 4.632689760261131e-06,
+      "loss": 0.6844,
+      "step": 29285
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4516130721776321,
+      "learning_rate": 4.632045957394286e-06,
+      "loss": 0.6568,
+      "step": 29286
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.551068356844459,
+      "learning_rate": 4.631402185781364e-06,
+      "loss": 0.7398,
+      "step": 29287
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.392441907053197,
+      "learning_rate": 4.630758445426123e-06,
+      "loss": 0.7307,
+      "step": 29288
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7375661916576746,
+      "learning_rate": 4.630114736332302e-06,
+      "loss": 0.6667,
+      "step": 29289
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5008112038244885,
+      "learning_rate": 4.6294710585036536e-06,
+      "loss": 0.8198,
+      "step": 29290
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0056901083684404,
+      "learning_rate": 4.628827411943933e-06,
+      "loss": 0.6335,
+      "step": 29291
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.168213143412195,
+      "learning_rate": 4.628183796656871e-06,
+      "loss": 0.6582,
+      "step": 29292
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3858786819669697,
+      "learning_rate": 4.627540212646223e-06,
+      "loss": 0.6619,
+      "step": 29293
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5158287368827272,
+      "learning_rate": 4.626896659915739e-06,
+      "loss": 0.6237,
+      "step": 29294
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5091503999363411,
+      "learning_rate": 4.626253138469169e-06,
+      "loss": 0.7524,
+      "step": 29295
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4431474951821661,
+      "learning_rate": 4.625609648310248e-06,
+      "loss": 0.5859,
+      "step": 29296
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.549287843809127,
+      "learning_rate": 4.62496618944273e-06,
+      "loss": 0.7094,
+      "step": 29297
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4553896906751043,
+      "learning_rate": 4.624322761870363e-06,
+      "loss": 0.7665,
+      "step": 29298
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8572955295937885,
+      "learning_rate": 4.623679365596891e-06,
+      "loss": 0.7735,
+      "step": 29299
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3737398325039698,
+      "learning_rate": 4.623036000626056e-06,
+      "loss": 0.6468,
+      "step": 29300
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3089077191002998,
+      "learning_rate": 4.62239266696161e-06,
+      "loss": 0.6556,
+      "step": 29301
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3466292627986223,
+      "learning_rate": 4.621749364607297e-06,
+      "loss": 0.7078,
+      "step": 29302
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6562644539139262,
+      "learning_rate": 4.6211060935668595e-06,
+      "loss": 0.6983,
+      "step": 29303
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4352057403697023,
+      "learning_rate": 4.620462853844049e-06,
+      "loss": 0.7069,
+      "step": 29304
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3458226158480675,
+      "learning_rate": 4.619819645442607e-06,
+      "loss": 0.6207,
+      "step": 29305
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6069759773104527,
+      "learning_rate": 4.619176468366274e-06,
+      "loss": 0.5384,
+      "step": 29306
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6422899547997488,
+      "learning_rate": 4.6185333226188e-06,
+      "loss": 0.5638,
+      "step": 29307
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.427268220747587,
+      "learning_rate": 4.617890208203932e-06,
+      "loss": 0.7862,
+      "step": 29308
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4598633588440892,
+      "learning_rate": 4.617247125125411e-06,
+      "loss": 0.7469,
+      "step": 29309
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2704471948711398,
+      "learning_rate": 4.616604073386978e-06,
+      "loss": 0.6454,
+      "step": 29310
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3844405060436498,
+      "learning_rate": 4.6159610529923805e-06,
+      "loss": 0.656,
+      "step": 29311
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7772639184700827,
+      "learning_rate": 4.61531806394537e-06,
+      "loss": 0.5952,
+      "step": 29312
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3872221967098282,
+      "learning_rate": 4.614675106249674e-06,
+      "loss": 0.673,
+      "step": 29313
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.244179781804908,
+      "learning_rate": 4.614032179909045e-06,
+      "loss": 0.6121,
+      "step": 29314
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2622218139223649,
+      "learning_rate": 4.613389284927225e-06,
+      "loss": 0.58,
+      "step": 29315
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3469800518390214,
+      "learning_rate": 4.612746421307963e-06,
+      "loss": 0.6902,
+      "step": 29316
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7258279187221592,
+      "learning_rate": 4.612103589054997e-06,
+      "loss": 0.5822,
+      "step": 29317
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6099227647151809,
+      "learning_rate": 4.611460788172065e-06,
+      "loss": 0.4936,
+      "step": 29318
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2371856528274563,
+      "learning_rate": 4.610818018662919e-06,
+      "loss": 0.6432,
+      "step": 29319
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.21917734221134,
+      "learning_rate": 4.6101752805312926e-06,
+      "loss": 0.6602,
+      "step": 29320
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4522702971897004,
+      "learning_rate": 4.609532573780936e-06,
+      "loss": 0.692,
+      "step": 29321
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2230277616999174,
+      "learning_rate": 4.608889898415583e-06,
+      "loss": 0.7799,
+      "step": 29322
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3470860798274082,
+      "learning_rate": 4.608247254438983e-06,
+      "loss": 0.6335,
+      "step": 29323
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3866980782100569,
+      "learning_rate": 4.607604641854872e-06,
+      "loss": 0.6428,
+      "step": 29324
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.569548011954587,
+      "learning_rate": 4.606962060666997e-06,
+      "loss": 0.4896,
+      "step": 29325
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6678367809856818,
+      "learning_rate": 4.606319510879097e-06,
+      "loss": 0.7094,
+      "step": 29326
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5940972197025831,
+      "learning_rate": 4.605676992494909e-06,
+      "loss": 0.4873,
+      "step": 29327
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5156717278041785,
+      "learning_rate": 4.605034505518176e-06,
+      "loss": 0.8517,
+      "step": 29328
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.362781909363339,
+      "learning_rate": 4.604392049952645e-06,
+      "loss": 0.6972,
+      "step": 29329
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4802869804630945,
+      "learning_rate": 4.603749625802051e-06,
+      "loss": 0.7177,
+      "step": 29330
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6222248257676414,
+      "learning_rate": 4.603107233070131e-06,
+      "loss": 0.5155,
+      "step": 29331
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6828838844709377,
+      "learning_rate": 4.60246487176063e-06,
+      "loss": 0.8499,
+      "step": 29332
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5744051923438047,
+      "learning_rate": 4.601822541877291e-06,
+      "loss": 0.5416,
+      "step": 29333
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2766437583164731,
+      "learning_rate": 4.601180243423851e-06,
+      "loss": 0.7873,
+      "step": 29334
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2617383504799613,
+      "learning_rate": 4.600537976404044e-06,
+      "loss": 0.6633,
+      "step": 29335
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5597219430887093,
+      "learning_rate": 4.5998957408216135e-06,
+      "loss": 0.4806,
+      "step": 29336
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5388078718202143,
+      "learning_rate": 4.599253536680305e-06,
+      "loss": 0.6383,
+      "step": 29337
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.252186183870878,
+      "learning_rate": 4.598611363983853e-06,
+      "loss": 0.7191,
+      "step": 29338
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4641954276248754,
+      "learning_rate": 4.59796922273599e-06,
+      "loss": 0.7687,
+      "step": 29339
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.259693427775511,
+      "learning_rate": 4.597327112940465e-06,
+      "loss": 0.6099,
+      "step": 29340
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.388799628059154,
+      "learning_rate": 4.596685034601008e-06,
+      "loss": 0.7507,
+      "step": 29341
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1751218859716268,
+      "learning_rate": 4.596042987721365e-06,
+      "loss": 0.8471,
+      "step": 29342
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4068049225365926,
+      "learning_rate": 4.595400972305267e-06,
+      "loss": 0.7026,
+      "step": 29343
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2846450711475266,
+      "learning_rate": 4.594758988356459e-06,
+      "loss": 0.7296,
+      "step": 29344
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4790914329038607,
+      "learning_rate": 4.594117035878672e-06,
+      "loss": 0.7241,
+      "step": 29345
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4209737107343656,
+      "learning_rate": 4.593475114875652e-06,
+      "loss": 0.7176,
+      "step": 29346
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.9444716214611386,
+      "learning_rate": 4.592833225351131e-06,
+      "loss": 0.7418,
+      "step": 29347
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2723426533320359,
+      "learning_rate": 4.592191367308842e-06,
+      "loss": 0.7926,
+      "step": 29348
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.417725354253239,
+      "learning_rate": 4.5915495407525275e-06,
+      "loss": 0.8152,
+      "step": 29349
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4008794604858301,
+      "learning_rate": 4.590907745685927e-06,
+      "loss": 0.7769,
+      "step": 29350
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.508144034018287,
+      "learning_rate": 4.590265982112776e-06,
+      "loss": 0.7123,
+      "step": 29351
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6201042007671078,
+      "learning_rate": 4.589624250036804e-06,
+      "loss": 0.6925,
+      "step": 29352
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1878598826297835,
+      "learning_rate": 4.588982549461752e-06,
+      "loss": 0.5754,
+      "step": 29353
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5996602004977598,
+      "learning_rate": 4.588340880391362e-06,
+      "loss": 0.7295,
+      "step": 29354
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6364424284892913,
+      "learning_rate": 4.587699242829364e-06,
+      "loss": 0.5709,
+      "step": 29355
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.553497570739611,
+      "learning_rate": 4.58705763677949e-06,
+      "loss": 0.7821,
+      "step": 29356
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3675032290694416,
+      "learning_rate": 4.586416062245481e-06,
+      "loss": 0.8075,
+      "step": 29357
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4384301196866054,
+      "learning_rate": 4.5857745192310755e-06,
+      "loss": 0.6297,
+      "step": 29358
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3100691920927705,
+      "learning_rate": 4.585133007740005e-06,
+      "loss": 0.7546,
+      "step": 29359
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3162607369050694,
+      "learning_rate": 4.584491527775999e-06,
+      "loss": 0.698,
+      "step": 29360
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2755115270175745,
+      "learning_rate": 4.583850079342803e-06,
+      "loss": 0.6347,
+      "step": 29361
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7855480286653913,
+      "learning_rate": 4.583208662444143e-06,
+      "loss": 0.6609,
+      "step": 29362
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4181849386749337,
+      "learning_rate": 4.58256727708376e-06,
+      "loss": 0.7661,
+      "step": 29363
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8737721980532023,
+      "learning_rate": 4.5819259232653824e-06,
+      "loss": 0.7003,
+      "step": 29364
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4977960943056985,
+      "learning_rate": 4.581284600992751e-06,
+      "loss": 0.5793,
+      "step": 29365
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6665885872447832,
+      "learning_rate": 4.580643310269592e-06,
+      "loss": 0.7609,
+      "step": 29366
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.528348098257028,
+      "learning_rate": 4.580002051099648e-06,
+      "loss": 0.6446,
+      "step": 29367
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.524837336136904,
+      "learning_rate": 4.579360823486644e-06,
+      "loss": 0.7255,
+      "step": 29368
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.658329651355899,
+      "learning_rate": 4.57871962743432e-06,
+      "loss": 0.7479,
+      "step": 29369
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2713140819156123,
+      "learning_rate": 4.578078462946404e-06,
+      "loss": 0.6365,
+      "step": 29370
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0946468925066775,
+      "learning_rate": 4.577437330026635e-06,
+      "loss": 0.6426,
+      "step": 29371
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5090275180886534,
+      "learning_rate": 4.576796228678744e-06,
+      "loss": 0.5014,
+      "step": 29372
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.486371924183659,
+      "learning_rate": 4.576155158906457e-06,
+      "loss": 0.8643,
+      "step": 29373
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3438845810466817,
+      "learning_rate": 4.575514120713513e-06,
+      "loss": 0.6,
+      "step": 29374
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6026304147270573,
+      "learning_rate": 4.574873114103645e-06,
+      "loss": 0.4913,
+      "step": 29375
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6237647010469565,
+      "learning_rate": 4.574232139080586e-06,
+      "loss": 0.7221,
+      "step": 29376
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2167134848139574,
+      "learning_rate": 4.573591195648059e-06,
+      "loss": 0.6585,
+      "step": 29377
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5811348968680455,
+      "learning_rate": 4.572950283809803e-06,
+      "loss": 0.5281,
+      "step": 29378
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2588679004335122,
+      "learning_rate": 4.572309403569552e-06,
+      "loss": 0.6532,
+      "step": 29379
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4646882753714325,
+      "learning_rate": 4.571668554931035e-06,
+      "loss": 0.589,
+      "step": 29380
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6237689412182825,
+      "learning_rate": 4.571027737897979e-06,
+      "loss": 0.639,
+      "step": 29381
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4233786273403113,
+      "learning_rate": 4.570386952474116e-06,
+      "loss": 0.7199,
+      "step": 29382
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6162937779836705,
+      "learning_rate": 4.569746198663184e-06,
+      "loss": 0.5189,
+      "step": 29383
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.77401269672034,
+      "learning_rate": 4.569105476468909e-06,
+      "loss": 0.6263,
+      "step": 29384
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3836800242518297,
+      "learning_rate": 4.568464785895018e-06,
+      "loss": 0.6558,
+      "step": 29385
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5284972434358441,
+      "learning_rate": 4.567824126945248e-06,
+      "loss": 0.6435,
+      "step": 29386
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4815999078806745,
+      "learning_rate": 4.567183499623322e-06,
+      "loss": 0.6492,
+      "step": 29387
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4322985202643845,
+      "learning_rate": 4.5665429039329735e-06,
+      "loss": 0.692,
+      "step": 29388
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2729045228725704,
+      "learning_rate": 4.565902339877936e-06,
+      "loss": 0.5882,
+      "step": 29389
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2353129327853973,
+      "learning_rate": 4.5652618074619355e-06,
+      "loss": 0.621,
+      "step": 29390
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.661547683412181,
+      "learning_rate": 4.564621306688698e-06,
+      "loss": 0.774,
+      "step": 29391
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2340855711235996,
+      "learning_rate": 4.563980837561955e-06,
+      "loss": 0.6805,
+      "step": 29392
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4859701956330398,
+      "learning_rate": 4.563340400085445e-06,
+      "loss": 0.6623,
+      "step": 29393
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3937245012914554,
+      "learning_rate": 4.562699994262881e-06,
+      "loss": 0.6676,
+      "step": 29394
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3743672020629536,
+      "learning_rate": 4.5620596200979996e-06,
+      "loss": 0.7561,
+      "step": 29395
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4007002154502661,
+      "learning_rate": 4.561419277594528e-06,
+      "loss": 0.8137,
+      "step": 29396
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4372803470872169,
+      "learning_rate": 4.560778966756203e-06,
+      "loss": 0.6595,
+      "step": 29397
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4708428572959533,
+      "learning_rate": 4.560138687586737e-06,
+      "loss": 0.7824,
+      "step": 29398
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7573673237406686,
+      "learning_rate": 4.5594984400898655e-06,
+      "loss": 0.7054,
+      "step": 29399
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3252673226066642,
+      "learning_rate": 4.5588582242693215e-06,
+      "loss": 0.6819,
+      "step": 29400
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.581730585911167,
+      "learning_rate": 4.558218040128822e-06,
+      "loss": 0.6601,
+      "step": 29401
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5833508169632545,
+      "learning_rate": 4.5575778876721055e-06,
+      "loss": 0.4753,
+      "step": 29402
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.742946658381581,
+      "learning_rate": 4.556937766902888e-06,
+      "loss": 0.6916,
+      "step": 29403
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.404529423473554,
+      "learning_rate": 4.556297677824907e-06,
+      "loss": 0.6488,
+      "step": 29404
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6820428563183398,
+      "learning_rate": 4.555657620441881e-06,
+      "loss": 0.6758,
+      "step": 29405
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6349328076713376,
+      "learning_rate": 4.555017594757543e-06,
+      "loss": 0.7838,
+      "step": 29406
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4406556106041652,
+      "learning_rate": 4.554377600775617e-06,
+      "loss": 0.8041,
+      "step": 29407
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.374195776826694,
+      "learning_rate": 4.553737638499824e-06,
+      "loss": 0.6972,
+      "step": 29408
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6131591236190231,
+      "learning_rate": 4.553097707933894e-06,
+      "loss": 0.5124,
+      "step": 29409
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3878699606313107,
+      "learning_rate": 4.55245780908156e-06,
+      "loss": 0.7059,
+      "step": 29410
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.530680471422826,
+      "learning_rate": 4.551817941946539e-06,
+      "loss": 0.7416,
+      "step": 29411
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.428402675436166,
+      "learning_rate": 4.5511781065325564e-06,
+      "loss": 0.7325,
+      "step": 29412
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4179844282952763,
+      "learning_rate": 4.550538302843339e-06,
+      "loss": 0.6753,
+      "step": 29413
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4711943103563478,
+      "learning_rate": 4.5498985308826174e-06,
+      "loss": 0.6946,
+      "step": 29414
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4406410433994032,
+      "learning_rate": 4.549258790654112e-06,
+      "loss": 0.778,
+      "step": 29415
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6005975412371094,
+      "learning_rate": 4.548619082161543e-06,
+      "loss": 0.508,
+      "step": 29416
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8871620040918966,
+      "learning_rate": 4.54797940540864e-06,
+      "loss": 0.6829,
+      "step": 29417
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5636064061155461,
+      "learning_rate": 4.547339760399131e-06,
+      "loss": 0.6546,
+      "step": 29418
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2371097399852335,
+      "learning_rate": 4.5467001471367355e-06,
+      "loss": 0.6388,
+      "step": 29419
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.942001936381676,
+      "learning_rate": 4.546060565625175e-06,
+      "loss": 0.6606,
+      "step": 29420
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4233748219430524,
+      "learning_rate": 4.545421015868181e-06,
+      "loss": 0.5634,
+      "step": 29421
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4662061639271748,
+      "learning_rate": 4.5447814978694685e-06,
+      "loss": 0.59,
+      "step": 29422
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.71500715354182,
+      "learning_rate": 4.544142011632769e-06,
+      "loss": 0.7265,
+      "step": 29423
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5961387414046506,
+      "learning_rate": 4.543502557161797e-06,
+      "loss": 0.6691,
+      "step": 29424
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.608593676311458,
+      "learning_rate": 4.542863134460286e-06,
+      "loss": 0.528,
+      "step": 29425
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6090997945875498,
+      "learning_rate": 4.542223743531949e-06,
+      "loss": 0.6386,
+      "step": 29426
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4019804155169098,
+      "learning_rate": 4.541584384380518e-06,
+      "loss": 0.5771,
+      "step": 29427
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.313859844278506,
+      "learning_rate": 4.540945057009709e-06,
+      "loss": 0.6003,
+      "step": 29428
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6064017283967136,
+      "learning_rate": 4.540305761423244e-06,
+      "loss": 0.5221,
+      "step": 29429
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4143183575141056,
+      "learning_rate": 4.5396664976248476e-06,
+      "loss": 0.7528,
+      "step": 29430
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3717307341890057,
+      "learning_rate": 4.539027265618244e-06,
+      "loss": 0.6569,
+      "step": 29431
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.318544417084399,
+      "learning_rate": 4.5383880654071525e-06,
+      "loss": 0.7439,
+      "step": 29432
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.340554643277315,
+      "learning_rate": 4.537748896995292e-06,
+      "loss": 0.6909,
+      "step": 29433
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3911830644881023,
+      "learning_rate": 4.537109760386387e-06,
+      "loss": 0.6349,
+      "step": 29434
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.404298550146507,
+      "learning_rate": 4.536470655584161e-06,
+      "loss": 0.7054,
+      "step": 29435
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.645619902467825,
+      "learning_rate": 4.535831582592333e-06,
+      "loss": 0.5034,
+      "step": 29436
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6586703925142887,
+      "learning_rate": 4.53519254141462e-06,
+      "loss": 0.6179,
+      "step": 29437
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5977683506252676,
+      "learning_rate": 4.5345535320547465e-06,
+      "loss": 0.5103,
+      "step": 29438
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.9488783186927265,
+      "learning_rate": 4.533914554516435e-06,
+      "loss": 0.5589,
+      "step": 29439
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4466572157599957,
+      "learning_rate": 4.533275608803405e-06,
+      "loss": 0.8615,
+      "step": 29440
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3990790328640925,
+      "learning_rate": 4.532636694919371e-06,
+      "loss": 0.7877,
+      "step": 29441
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3898622935367173,
+      "learning_rate": 4.531997812868061e-06,
+      "loss": 0.7312,
+      "step": 29442
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6288007792217567,
+      "learning_rate": 4.5313589626531864e-06,
+      "loss": 0.8242,
+      "step": 29443
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6815150685455962,
+      "learning_rate": 4.5307201442784755e-06,
+      "loss": 0.5732,
+      "step": 29444
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4754544147590438,
+      "learning_rate": 4.53008135774764e-06,
+      "loss": 0.6691,
+      "step": 29445
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4513533048300444,
+      "learning_rate": 4.529442603064406e-06,
+      "loss": 0.7796,
+      "step": 29446
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5724930476061246,
+      "learning_rate": 4.528803880232485e-06,
+      "loss": 0.5605,
+      "step": 29447
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3853254866023477,
+      "learning_rate": 4.528165189255605e-06,
+      "loss": 0.6791,
+      "step": 29448
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7897202710252764,
+      "learning_rate": 4.527526530137478e-06,
+      "loss": 0.7227,
+      "step": 29449
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4203070221070146,
+      "learning_rate": 4.526887902881822e-06,
+      "loss": 0.6165,
+      "step": 29450
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5898016052597108,
+      "learning_rate": 4.526249307492355e-06,
+      "loss": 0.7746,
+      "step": 29451
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7369123892202734,
+      "learning_rate": 4.525610743972804e-06,
+      "loss": 0.6645,
+      "step": 29452
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5357043092146467,
+      "learning_rate": 4.524972212326878e-06,
+      "loss": 0.7461,
+      "step": 29453
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3071063012100952,
+      "learning_rate": 4.524333712558294e-06,
+      "loss": 0.5668,
+      "step": 29454
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3932455461542315,
+      "learning_rate": 4.5236952446707735e-06,
+      "loss": 0.8046,
+      "step": 29455
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.62086284429936,
+      "learning_rate": 4.523056808668037e-06,
+      "loss": 0.8341,
+      "step": 29456
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4411655002220003,
+      "learning_rate": 4.5224184045537964e-06,
+      "loss": 0.7912,
+      "step": 29457
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8690503775042155,
+      "learning_rate": 4.521780032331768e-06,
+      "loss": 0.7527,
+      "step": 29458
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9415751537534385,
+      "learning_rate": 4.521141692005669e-06,
+      "loss": 0.6909,
+      "step": 29459
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4580734332021053,
+      "learning_rate": 4.5205033835792226e-06,
+      "loss": 0.7392,
+      "step": 29460
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.462057710179495,
+      "learning_rate": 4.519865107056139e-06,
+      "loss": 0.6681,
+      "step": 29461
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.022520160560557,
+      "learning_rate": 4.519226862440134e-06,
+      "loss": 0.6751,
+      "step": 29462
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5647731244747702,
+      "learning_rate": 4.518588649734927e-06,
+      "loss": 0.6769,
+      "step": 29463
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7361384198879224,
+      "learning_rate": 4.51795046894423e-06,
+      "loss": 0.5934,
+      "step": 29464
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6440975214999705,
+      "learning_rate": 4.517312320071764e-06,
+      "loss": 0.5215,
+      "step": 29465
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3643937517301528,
+      "learning_rate": 4.5166742031212375e-06,
+      "loss": 0.7132,
+      "step": 29466
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3470869109617187,
+      "learning_rate": 4.516036118096374e-06,
+      "loss": 0.7223,
+      "step": 29467
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.495166783435638,
+      "learning_rate": 4.515398065000881e-06,
+      "loss": 0.6884,
+      "step": 29468
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.800345182558868,
+      "learning_rate": 4.51476004383848e-06,
+      "loss": 0.7317,
+      "step": 29469
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.551309613476827,
+      "learning_rate": 4.514122054612883e-06,
+      "loss": 0.7177,
+      "step": 29470
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7322053561736932,
+      "learning_rate": 4.5134840973278e-06,
+      "loss": 0.8502,
+      "step": 29471
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2536790490282825,
+      "learning_rate": 4.51284617198695e-06,
+      "loss": 0.568,
+      "step": 29472
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.632779232317668,
+      "learning_rate": 4.51220827859405e-06,
+      "loss": 0.511,
+      "step": 29473
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2946901196618965,
+      "learning_rate": 4.5115704171528105e-06,
+      "loss": 0.6296,
+      "step": 29474
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6491022726922604,
+      "learning_rate": 4.510932587666943e-06,
+      "loss": 0.8751,
+      "step": 29475
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2765560924666124,
+      "learning_rate": 4.510294790140164e-06,
+      "loss": 0.7161,
+      "step": 29476
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3197904756066319,
+      "learning_rate": 4.509657024576189e-06,
+      "loss": 0.81,
+      "step": 29477
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9621056120191787,
+      "learning_rate": 4.50901929097873e-06,
+      "loss": 0.6806,
+      "step": 29478
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.514734898842334,
+      "learning_rate": 4.508381589351496e-06,
+      "loss": 0.6362,
+      "step": 29479
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4718591738562843,
+      "learning_rate": 4.507743919698202e-06,
+      "loss": 0.6834,
+      "step": 29480
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4317162278059474,
+      "learning_rate": 4.507106282022565e-06,
+      "loss": 0.7248,
+      "step": 29481
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.638965984811859,
+      "learning_rate": 4.506468676328295e-06,
+      "loss": 0.6782,
+      "step": 29482
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2748447539299,
+      "learning_rate": 4.5058311026191e-06,
+      "loss": 0.5725,
+      "step": 29483
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.534092922471099,
+      "learning_rate": 4.5051935608986994e-06,
+      "loss": 0.8968,
+      "step": 29484
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2753044830761944,
+      "learning_rate": 4.504556051170797e-06,
+      "loss": 0.7492,
+      "step": 29485
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.507745255447182,
+      "learning_rate": 4.50391857343911e-06,
+      "loss": 0.7075,
+      "step": 29486
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.593528163278786,
+      "learning_rate": 4.5032811277073525e-06,
+      "loss": 0.4569,
+      "step": 29487
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8743672426245392,
+      "learning_rate": 4.502643713979232e-06,
+      "loss": 0.7923,
+      "step": 29488
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4445061652758386,
+      "learning_rate": 4.5020063322584585e-06,
+      "loss": 0.763,
+      "step": 29489
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.381072575262787,
+      "learning_rate": 4.501368982548743e-06,
+      "loss": 0.724,
+      "step": 29490
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.60584652649866,
+      "learning_rate": 4.500731664853801e-06,
+      "loss": 0.6842,
+      "step": 29491
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4406123649809224,
+      "learning_rate": 4.500094379177342e-06,
+      "loss": 0.7342,
+      "step": 29492
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4614620067834903,
+      "learning_rate": 4.499457125523071e-06,
+      "loss": 0.8623,
+      "step": 29493
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.419075920509989,
+      "learning_rate": 4.498819903894701e-06,
+      "loss": 0.6012,
+      "step": 29494
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7069636470032563,
+      "learning_rate": 4.498182714295951e-06,
+      "loss": 0.5997,
+      "step": 29495
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.476830808125202,
+      "learning_rate": 4.497545556730516e-06,
+      "loss": 0.8178,
+      "step": 29496
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6897663861529613,
+      "learning_rate": 4.496908431202113e-06,
+      "loss": 0.6694,
+      "step": 29497
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6152051113228374,
+      "learning_rate": 4.49627133771445e-06,
+      "loss": 0.6508,
+      "step": 29498
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7109159488302517,
+      "learning_rate": 4.495634276271244e-06,
+      "loss": 0.7261,
+      "step": 29499
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2787519442089934,
+      "learning_rate": 4.494997246876196e-06,
+      "loss": 0.6367,
+      "step": 29500
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.698759183211977,
+      "learning_rate": 4.494360249533013e-06,
+      "loss": 0.7709,
+      "step": 29501
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4902045193163493,
+      "learning_rate": 4.493723284245413e-06,
+      "loss": 0.8419,
+      "step": 29502
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3735178845884592,
+      "learning_rate": 4.493086351017093e-06,
+      "loss": 0.7667,
+      "step": 29503
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5986521182838522,
+      "learning_rate": 4.4924494498517735e-06,
+      "loss": 0.5097,
+      "step": 29504
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5626865944018178,
+      "learning_rate": 4.491812580753152e-06,
+      "loss": 0.491,
+      "step": 29505
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5594463763691255,
+      "learning_rate": 4.491175743724947e-06,
+      "loss": 0.52,
+      "step": 29506
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4355506849583035,
+      "learning_rate": 4.490538938770855e-06,
+      "loss": 0.8078,
+      "step": 29507
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2458191658635611,
+      "learning_rate": 4.489902165894595e-06,
+      "loss": 0.6983,
+      "step": 29508
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7236743875215967,
+      "learning_rate": 4.489265425099868e-06,
+      "loss": 0.6425,
+      "step": 29509
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3744089466755707,
+      "learning_rate": 4.48862871639038e-06,
+      "loss": 0.7372,
+      "step": 29510
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.292512559677001,
+      "learning_rate": 4.487992039769838e-06,
+      "loss": 0.7857,
+      "step": 29511
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.704660597980463,
+      "learning_rate": 4.487355395241958e-06,
+      "loss": 0.6984,
+      "step": 29512
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.380257712762612,
+      "learning_rate": 4.486718782810439e-06,
+      "loss": 0.7086,
+      "step": 29513
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4201610259123283,
+      "learning_rate": 4.486082202478984e-06,
+      "loss": 0.6107,
+      "step": 29514
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2171427722826755,
+      "learning_rate": 4.485445654251305e-06,
+      "loss": 0.6256,
+      "step": 29515
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7434070034026978,
+      "learning_rate": 4.484809138131111e-06,
+      "loss": 0.6511,
+      "step": 29516
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4668897514437287,
+      "learning_rate": 4.484172654122104e-06,
+      "loss": 0.7459,
+      "step": 29517
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5872450038822423,
+      "learning_rate": 4.483536202227987e-06,
+      "loss": 0.6505,
+      "step": 29518
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4656749705861636,
+      "learning_rate": 4.482899782452468e-06,
+      "loss": 0.6771,
+      "step": 29519
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0842486090170205,
+      "learning_rate": 4.482263394799258e-06,
+      "loss": 0.7108,
+      "step": 29520
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5494242830316876,
+      "learning_rate": 4.481627039272056e-06,
+      "loss": 0.7252,
+      "step": 29521
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.351863915729129,
+      "learning_rate": 4.480990715874564e-06,
+      "loss": 0.5609,
+      "step": 29522
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5823832407022652,
+      "learning_rate": 4.480354424610498e-06,
+      "loss": 0.5077,
+      "step": 29523
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4740988324316784,
+      "learning_rate": 4.47971816548355e-06,
+      "loss": 0.7221,
+      "step": 29524
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3578858506150306,
+      "learning_rate": 4.479081938497435e-06,
+      "loss": 0.6105,
+      "step": 29525
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3865074654085197,
+      "learning_rate": 4.478445743655849e-06,
+      "loss": 0.6107,
+      "step": 29526
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6092908958927625,
+      "learning_rate": 4.477809580962503e-06,
+      "loss": 0.7127,
+      "step": 29527
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.507292372561299,
+      "learning_rate": 4.477173450421096e-06,
+      "loss": 0.5797,
+      "step": 29528
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.636563750410774,
+      "learning_rate": 4.476537352035336e-06,
+      "loss": 0.6502,
+      "step": 29529
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.365330274034365,
+      "learning_rate": 4.475901285808924e-06,
+      "loss": 0.7015,
+      "step": 29530
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.493480656167568,
+      "learning_rate": 4.4752652517455605e-06,
+      "loss": 0.5594,
+      "step": 29531
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5529422589173019,
+      "learning_rate": 4.4746292498489515e-06,
+      "loss": 0.6336,
+      "step": 29532
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2749966588361676,
+      "learning_rate": 4.4739932801228046e-06,
+      "loss": 0.5857,
+      "step": 29533
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.199574535580676,
+      "learning_rate": 4.473357342570819e-06,
+      "loss": 0.6752,
+      "step": 29534
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1521654221192217,
+      "learning_rate": 4.472721437196692e-06,
+      "loss": 0.7189,
+      "step": 29535
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2698007442922468,
+      "learning_rate": 4.472085564004131e-06,
+      "loss": 0.6457,
+      "step": 29536
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4069374768451357,
+      "learning_rate": 4.47144972299684e-06,
+      "loss": 0.679,
+      "step": 29537
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4460759359332525,
+      "learning_rate": 4.470813914178521e-06,
+      "loss": 0.7502,
+      "step": 29538
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5161649442841223,
+      "learning_rate": 4.47017813755287e-06,
+      "loss": 0.7374,
+      "step": 29539
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.607186368814528,
+      "learning_rate": 4.4695423931235915e-06,
+      "loss": 0.4737,
+      "step": 29540
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.400257960629786,
+      "learning_rate": 4.4689066808943925e-06,
+      "loss": 0.6726,
+      "step": 29541
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6054606600574729,
+      "learning_rate": 4.46827100086897e-06,
+      "loss": 0.5592,
+      "step": 29542
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5517781487382052,
+      "learning_rate": 4.467635353051021e-06,
+      "loss": 0.8082,
+      "step": 29543
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5670208939789168,
+      "learning_rate": 4.466999737444254e-06,
+      "loss": 0.4625,
+      "step": 29544
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.313376292359693,
+      "learning_rate": 4.466364154052364e-06,
+      "loss": 0.6108,
+      "step": 29545
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3052146252508794,
+      "learning_rate": 4.4657286028790555e-06,
+      "loss": 0.641,
+      "step": 29546
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2785694009952109,
+      "learning_rate": 4.465093083928025e-06,
+      "loss": 0.6167,
+      "step": 29547
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2536669886527994,
+      "learning_rate": 4.464457597202978e-06,
+      "loss": 0.6745,
+      "step": 29548
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5368406056946695,
+      "learning_rate": 4.4638221427076075e-06,
+      "loss": 0.6094,
+      "step": 29549
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5194787647043546,
+      "learning_rate": 4.4631867204456215e-06,
+      "loss": 0.6827,
+      "step": 29550
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4229108846781175,
+      "learning_rate": 4.462551330420715e-06,
+      "loss": 0.6218,
+      "step": 29551
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.764183996231113,
+      "learning_rate": 4.461915972636584e-06,
+      "loss": 0.6012,
+      "step": 29552
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2293416771408623,
+      "learning_rate": 4.461280647096932e-06,
+      "loss": 0.6791,
+      "step": 29553
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5899113198612556,
+      "learning_rate": 4.46064535380546e-06,
+      "loss": 0.4957,
+      "step": 29554
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7692768402224015,
+      "learning_rate": 4.460010092765866e-06,
+      "loss": 0.6013,
+      "step": 29555
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4411297213233238,
+      "learning_rate": 4.4593748639818425e-06,
+      "loss": 0.6644,
+      "step": 29556
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5770890502997847,
+      "learning_rate": 4.458739667457094e-06,
+      "loss": 0.565,
+      "step": 29557
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4123349399342446,
+      "learning_rate": 4.4581045031953195e-06,
+      "loss": 0.5931,
+      "step": 29558
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0219948532735534,
+      "learning_rate": 4.457469371200216e-06,
+      "loss": 0.7162,
+      "step": 29559
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.411547639886769,
+      "learning_rate": 4.456834271475477e-06,
+      "loss": 0.6235,
+      "step": 29560
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3655097576021786,
+      "learning_rate": 4.456199204024803e-06,
+      "loss": 0.7267,
+      "step": 29561
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2572057679198585,
+      "learning_rate": 4.455564168851897e-06,
+      "loss": 0.6316,
+      "step": 29562
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.599609518143932,
+      "learning_rate": 4.454929165960451e-06,
+      "loss": 0.7182,
+      "step": 29563
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4239478854198664,
+      "learning_rate": 4.45429419535416e-06,
+      "loss": 0.6572,
+      "step": 29564
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4543204935337404,
+      "learning_rate": 4.453659257036727e-06,
+      "loss": 0.7405,
+      "step": 29565
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2702046803278397,
+      "learning_rate": 4.453024351011842e-06,
+      "loss": 0.7982,
+      "step": 29566
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4268270242059982,
+      "learning_rate": 4.452389477283209e-06,
+      "loss": 0.7945,
+      "step": 29567
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4663919351986976,
+      "learning_rate": 4.4517546358545175e-06,
+      "loss": 0.7234,
+      "step": 29568
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5789806279860058,
+      "learning_rate": 4.451119826729471e-06,
+      "loss": 0.5149,
+      "step": 29569
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6387558965287761,
+      "learning_rate": 4.450485049911757e-06,
+      "loss": 0.5576,
+      "step": 29570
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3243432217583309,
+      "learning_rate": 4.449850305405082e-06,
+      "loss": 0.7628,
+      "step": 29571
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.198388822895281,
+      "learning_rate": 4.449215593213133e-06,
+      "loss": 0.6971,
+      "step": 29572
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4374932839540258,
+      "learning_rate": 4.448580913339605e-06,
+      "loss": 0.8362,
+      "step": 29573
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3303454809229296,
+      "learning_rate": 4.447946265788197e-06,
+      "loss": 0.7106,
+      "step": 29574
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9367096812913556,
+      "learning_rate": 4.447311650562604e-06,
+      "loss": 0.5961,
+      "step": 29575
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6278958248684336,
+      "learning_rate": 4.446677067666527e-06,
+      "loss": 0.5153,
+      "step": 29576
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4250268404687265,
+      "learning_rate": 4.446042517103647e-06,
+      "loss": 0.6689,
+      "step": 29577
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3634321900867585,
+      "learning_rate": 4.445407998877667e-06,
+      "loss": 0.7291,
+      "step": 29578
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.335622157718983,
+      "learning_rate": 4.444773512992282e-06,
+      "loss": 0.6957,
+      "step": 29579
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4339971313370135,
+      "learning_rate": 4.4441390594511855e-06,
+      "loss": 0.6986,
+      "step": 29580
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3892561830531922,
+      "learning_rate": 4.443504638258066e-06,
+      "loss": 0.8102,
+      "step": 29581
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5584691637457282,
+      "learning_rate": 4.442870249416622e-06,
+      "loss": 0.8095,
+      "step": 29582
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3299638155780342,
+      "learning_rate": 4.442235892930551e-06,
+      "loss": 0.6194,
+      "step": 29583
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9688396977360672,
+      "learning_rate": 4.441601568803538e-06,
+      "loss": 0.6404,
+      "step": 29584
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2635797322104543,
+      "learning_rate": 4.440967277039284e-06,
+      "loss": 0.7145,
+      "step": 29585
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4973196134613622,
+      "learning_rate": 4.4403330176414795e-06,
+      "loss": 0.6396,
+      "step": 29586
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4974611975455503,
+      "learning_rate": 4.439698790613811e-06,
+      "loss": 0.7431,
+      "step": 29587
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3226254492637632,
+      "learning_rate": 4.439064595959978e-06,
+      "loss": 0.59,
+      "step": 29588
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4355974446463364,
+      "learning_rate": 4.4384304336836745e-06,
+      "loss": 0.7396,
+      "step": 29589
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.655373456153497,
+      "learning_rate": 4.437796303788589e-06,
+      "loss": 0.5583,
+      "step": 29590
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4530941239752826,
+      "learning_rate": 4.437162206278413e-06,
+      "loss": 0.6996,
+      "step": 29591
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4132507735521092,
+      "learning_rate": 4.436528141156838e-06,
+      "loss": 0.676,
+      "step": 29592
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4104796220805842,
+      "learning_rate": 4.435894108427564e-06,
+      "loss": 0.6509,
+      "step": 29593
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8352943811069982,
+      "learning_rate": 4.43526010809427e-06,
+      "loss": 0.7755,
+      "step": 29594
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5855046927223676,
+      "learning_rate": 4.434626140160654e-06,
+      "loss": 0.5259,
+      "step": 29595
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5720079500682852,
+      "learning_rate": 4.4339922046304055e-06,
+      "loss": 0.4702,
+      "step": 29596
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3665377539966124,
+      "learning_rate": 4.43335830150722e-06,
+      "loss": 0.5709,
+      "step": 29597
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5440022408608396,
+      "learning_rate": 4.432724430794786e-06,
+      "loss": 0.7116,
+      "step": 29598
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.290420493231117,
+      "learning_rate": 4.432090592496788e-06,
+      "loss": 0.7352,
+      "step": 29599
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5406549043578726,
+      "learning_rate": 4.431456786616927e-06,
+      "loss": 0.7264,
+      "step": 29600
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4134262121126935,
+      "learning_rate": 4.4308230131588826e-06,
+      "loss": 0.7904,
+      "step": 29601
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3500701081464397,
+      "learning_rate": 4.430189272126353e-06,
+      "loss": 0.6466,
+      "step": 29602
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2591633764071888,
+      "learning_rate": 4.429555563523021e-06,
+      "loss": 0.6851,
+      "step": 29603
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4308316808982848,
+      "learning_rate": 4.4289218873525855e-06,
+      "loss": 0.6723,
+      "step": 29604
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2413008078098913,
+      "learning_rate": 4.428288243618725e-06,
+      "loss": 0.7018,
+      "step": 29605
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1651098989492292,
+      "learning_rate": 4.427654632325139e-06,
+      "loss": 0.6175,
+      "step": 29606
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5883209089007221,
+      "learning_rate": 4.427021053475512e-06,
+      "loss": 0.525,
+      "step": 29607
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.314180217206529,
+      "learning_rate": 4.426387507073529e-06,
+      "loss": 0.7041,
+      "step": 29608
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1981091118526095,
+      "learning_rate": 4.425753993122882e-06,
+      "loss": 0.7232,
+      "step": 29609
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3947685711191649,
+      "learning_rate": 4.425120511627263e-06,
+      "loss": 0.6169,
+      "step": 29610
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4008186310599064,
+      "learning_rate": 4.424487062590359e-06,
+      "loss": 0.6175,
+      "step": 29611
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4807522962398612,
+      "learning_rate": 4.423853646015852e-06,
+      "loss": 0.79,
+      "step": 29612
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.35955751023028,
+      "learning_rate": 4.423220261907435e-06,
+      "loss": 0.6785,
+      "step": 29613
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.694325796632555,
+      "learning_rate": 4.422586910268798e-06,
+      "loss": 0.7584,
+      "step": 29614
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3407573139679725,
+      "learning_rate": 4.421953591103627e-06,
+      "loss": 0.6745,
+      "step": 29615
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4018612321917359,
+      "learning_rate": 4.421320304415604e-06,
+      "loss": 0.6632,
+      "step": 29616
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.633137835593009,
+      "learning_rate": 4.42068705020842e-06,
+      "loss": 0.7707,
+      "step": 29617
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.485626306606491,
+      "learning_rate": 4.420053828485765e-06,
+      "loss": 0.7138,
+      "step": 29618
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3702087369739682,
+      "learning_rate": 4.419420639251325e-06,
+      "loss": 0.7623,
+      "step": 29619
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3849162329027929,
+      "learning_rate": 4.4187874825087806e-06,
+      "loss": 0.6399,
+      "step": 29620
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3333787128409154,
+      "learning_rate": 4.418154358261822e-06,
+      "loss": 0.7042,
+      "step": 29621
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2978914403116186,
+      "learning_rate": 4.417521266514141e-06,
+      "loss": 0.7281,
+      "step": 29622
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1886390873201385,
+      "learning_rate": 4.416888207269418e-06,
+      "loss": 0.7197,
+      "step": 29623
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4930232039991003,
+      "learning_rate": 4.4162551805313345e-06,
+      "loss": 0.6727,
+      "step": 29624
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.259215913482873,
+      "learning_rate": 4.4156221863035854e-06,
+      "loss": 0.7132,
+      "step": 29625
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6022440597332126,
+      "learning_rate": 4.414989224589848e-06,
+      "loss": 0.6319,
+      "step": 29626
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6101935001031369,
+      "learning_rate": 4.414356295393816e-06,
+      "loss": 0.5017,
+      "step": 29627
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7825383483047068,
+      "learning_rate": 4.413723398719167e-06,
+      "loss": 0.765,
+      "step": 29628
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6236384372467914,
+      "learning_rate": 4.413090534569591e-06,
+      "loss": 0.7865,
+      "step": 29629
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4282778614431852,
+      "learning_rate": 4.412457702948768e-06,
+      "loss": 0.7525,
+      "step": 29630
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4106973119704738,
+      "learning_rate": 4.411824903860387e-06,
+      "loss": 0.7738,
+      "step": 29631
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.8536338453705,
+      "learning_rate": 4.411192137308134e-06,
+      "loss": 0.7359,
+      "step": 29632
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6048769990320118,
+      "learning_rate": 4.410559403295683e-06,
+      "loss": 0.7074,
+      "step": 29633
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3894861783513002,
+      "learning_rate": 4.409926701826726e-06,
+      "loss": 0.7242,
+      "step": 29634
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5464164860431702,
+      "learning_rate": 4.4092940329049495e-06,
+      "loss": 0.7703,
+      "step": 29635
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4572449463534338,
+      "learning_rate": 4.408661396534033e-06,
+      "loss": 0.6794,
+      "step": 29636
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.625437345230232,
+      "learning_rate": 4.4080287927176565e-06,
+      "loss": 0.7643,
+      "step": 29637
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7315559049158007,
+      "learning_rate": 4.407396221459507e-06,
+      "loss": 0.6701,
+      "step": 29638
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6386905762986824,
+      "learning_rate": 4.40676368276327e-06,
+      "loss": 0.4782,
+      "step": 29639
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5110095602555635,
+      "learning_rate": 4.406131176632626e-06,
+      "loss": 0.6393,
+      "step": 29640
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6041064687517163,
+      "learning_rate": 4.405498703071254e-06,
+      "loss": 0.4891,
+      "step": 29641
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.540490612077272,
+      "learning_rate": 4.404866262082839e-06,
+      "loss": 0.7889,
+      "step": 29642
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6064054965024392,
+      "learning_rate": 4.404233853671069e-06,
+      "loss": 0.7935,
+      "step": 29643
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2608349780714063,
+      "learning_rate": 4.403601477839619e-06,
+      "loss": 0.6282,
+      "step": 29644
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5106486837498903,
+      "learning_rate": 4.402969134592171e-06,
+      "loss": 0.7015,
+      "step": 29645
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4577902604549868,
+      "learning_rate": 4.402336823932412e-06,
+      "loss": 0.6394,
+      "step": 29646
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6679092831516489,
+      "learning_rate": 4.401704545864016e-06,
+      "loss": 0.7014,
+      "step": 29647
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3632046897402093,
+      "learning_rate": 4.401072300390672e-06,
+      "loss": 0.7088,
+      "step": 29648
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5054136775343903,
+      "learning_rate": 4.4004400875160545e-06,
+      "loss": 0.6673,
+      "step": 29649
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4115069141982246,
+      "learning_rate": 4.399807907243851e-06,
+      "loss": 0.6562,
+      "step": 29650
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.729751712987075,
+      "learning_rate": 4.399175759577737e-06,
+      "loss": 0.697,
+      "step": 29651
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3994942578773426,
+      "learning_rate": 4.398543644521396e-06,
+      "loss": 0.7405,
+      "step": 29652
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.348023747254022,
+      "learning_rate": 4.397911562078508e-06,
+      "loss": 0.806,
+      "step": 29653
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6746142614216438,
+      "learning_rate": 4.39727951225275e-06,
+      "loss": 0.5262,
+      "step": 29654
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6236736353008574,
+      "learning_rate": 4.3966474950478036e-06,
+      "loss": 0.8164,
+      "step": 29655
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.627640460222627,
+      "learning_rate": 4.396015510467353e-06,
+      "loss": 0.612,
+      "step": 29656
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2112455733354222,
+      "learning_rate": 4.395383558515075e-06,
+      "loss": 0.6159,
+      "step": 29657
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.440716819623934,
+      "learning_rate": 4.394751639194644e-06,
+      "loss": 0.8144,
+      "step": 29658
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3396235386717967,
+      "learning_rate": 4.394119752509743e-06,
+      "loss": 0.7922,
+      "step": 29659
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.85854730102882,
+      "learning_rate": 4.393487898464056e-06,
+      "loss": 0.6002,
+      "step": 29660
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2159264078968353,
+      "learning_rate": 4.392856077061258e-06,
+      "loss": 0.6561,
+      "step": 29661
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.391365207700788,
+      "learning_rate": 4.3922242883050226e-06,
+      "loss": 0.7327,
+      "step": 29662
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5932408008424734,
+      "learning_rate": 4.391592532199034e-06,
+      "loss": 0.5023,
+      "step": 29663
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6165900322758099,
+      "learning_rate": 4.390960808746971e-06,
+      "loss": 0.5405,
+      "step": 29664
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6557163465665126,
+      "learning_rate": 4.390329117952511e-06,
+      "loss": 0.8487,
+      "step": 29665
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5857335748246284,
+      "learning_rate": 4.389697459819327e-06,
+      "loss": 0.5016,
+      "step": 29666
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8344920245636405,
+      "learning_rate": 4.389065834351105e-06,
+      "loss": 0.7715,
+      "step": 29667
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2529766876102304,
+      "learning_rate": 4.3884342415515135e-06,
+      "loss": 0.6961,
+      "step": 29668
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3965922738078034,
+      "learning_rate": 4.387802681424234e-06,
+      "loss": 0.6054,
+      "step": 29669
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6744469526109318,
+      "learning_rate": 4.387171153972949e-06,
+      "loss": 0.644,
+      "step": 29670
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4716996650456806,
+      "learning_rate": 4.38653965920133e-06,
+      "loss": 0.6638,
+      "step": 29671
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6288515472224696,
+      "learning_rate": 4.3859081971130516e-06,
+      "loss": 0.6904,
+      "step": 29672
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3322429537472582,
+      "learning_rate": 4.385276767711792e-06,
+      "loss": 0.6377,
+      "step": 29673
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5816730262855416,
+      "learning_rate": 4.384645371001236e-06,
+      "loss": 0.7791,
+      "step": 29674
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6534809656524347,
+      "learning_rate": 4.384014006985044e-06,
+      "loss": 0.7697,
+      "step": 29675
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5774559331453091,
+      "learning_rate": 4.383382675666903e-06,
+      "loss": 0.4666,
+      "step": 29676
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.270434293140391,
+      "learning_rate": 4.382751377050485e-06,
+      "loss": 0.6828,
+      "step": 29677
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.312501992141824,
+      "learning_rate": 4.382120111139474e-06,
+      "loss": 0.7982,
+      "step": 29678
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5296437599499606,
+      "learning_rate": 4.3814888779375305e-06,
+      "loss": 0.7607,
+      "step": 29679
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5266842740028583,
+      "learning_rate": 4.380857677448338e-06,
+      "loss": 0.8027,
+      "step": 29680
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.385899173708467,
+      "learning_rate": 4.380226509675576e-06,
+      "loss": 0.6416,
+      "step": 29681
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6215370765082922,
+      "learning_rate": 4.379595374622909e-06,
+      "loss": 0.5341,
+      "step": 29682
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5345061847204935,
+      "learning_rate": 4.378964272294022e-06,
+      "loss": 0.5579,
+      "step": 29683
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2881129198424162,
+      "learning_rate": 4.37833320269258e-06,
+      "loss": 0.6137,
+      "step": 29684
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5220483215334635,
+      "learning_rate": 4.377702165822266e-06,
+      "loss": 0.7448,
+      "step": 29685
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4326697751197062,
+      "learning_rate": 4.377071161686745e-06,
+      "loss": 0.7443,
+      "step": 29686
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5850360692305373,
+      "learning_rate": 4.376440190289703e-06,
+      "loss": 0.4682,
+      "step": 29687
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6310120012017624,
+      "learning_rate": 4.375809251634804e-06,
+      "loss": 0.7658,
+      "step": 29688
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3276820107994762,
+      "learning_rate": 4.37517834572572e-06,
+      "loss": 0.7508,
+      "step": 29689
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5963500916537163,
+      "learning_rate": 4.374547472566129e-06,
+      "loss": 0.6873,
+      "step": 29690
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.315636632442725,
+      "learning_rate": 4.373916632159708e-06,
+      "loss": 0.7137,
+      "step": 29691
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3240110859918708,
+      "learning_rate": 4.373285824510125e-06,
+      "loss": 0.7268,
+      "step": 29692
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.305605586888062,
+      "learning_rate": 4.372655049621051e-06,
+      "loss": 0.6922,
+      "step": 29693
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4841541083184575,
+      "learning_rate": 4.372024307496159e-06,
+      "loss": 0.725,
+      "step": 29694
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2172812083264846,
+      "learning_rate": 4.371393598139128e-06,
+      "loss": 0.5867,
+      "step": 29695
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.602727474237576,
+      "learning_rate": 4.370762921553625e-06,
+      "loss": 0.5534,
+      "step": 29696
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.45904948020521,
+      "learning_rate": 4.370132277743319e-06,
+      "loss": 0.6733,
+      "step": 29697
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2640524477432904,
+      "learning_rate": 4.369501666711885e-06,
+      "loss": 0.6489,
+      "step": 29698
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8416873247177175,
+      "learning_rate": 4.368871088462998e-06,
+      "loss": 0.6455,
+      "step": 29699
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.532650779684517,
+      "learning_rate": 4.368240543000327e-06,
+      "loss": 0.6942,
+      "step": 29700
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3639720907017587,
+      "learning_rate": 4.36761003032754e-06,
+      "loss": 0.8092,
+      "step": 29701
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4019231886100012,
+      "learning_rate": 4.366979550448313e-06,
+      "loss": 0.6319,
+      "step": 29702
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2866719770481154,
+      "learning_rate": 4.366349103366311e-06,
+      "loss": 0.6245,
+      "step": 29703
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3774517629248264,
+      "learning_rate": 4.365718689085211e-06,
+      "loss": 0.7166,
+      "step": 29704
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.340892377940252,
+      "learning_rate": 4.365088307608678e-06,
+      "loss": 0.631,
+      "step": 29705
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8276507691653,
+      "learning_rate": 4.364457958940388e-06,
+      "loss": 0.7157,
+      "step": 29706
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6012820896655151,
+      "learning_rate": 4.3638276430840045e-06,
+      "loss": 0.5043,
+      "step": 29707
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3854183986221345,
+      "learning_rate": 4.363197360043204e-06,
+      "loss": 0.6177,
+      "step": 29708
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6247573525289624,
+      "learning_rate": 4.362567109821654e-06,
+      "loss": 0.4968,
+      "step": 29709
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5132050858417232,
+      "learning_rate": 4.3619368924230185e-06,
+      "loss": 0.8013,
+      "step": 29710
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4534630489692706,
+      "learning_rate": 4.361306707850972e-06,
+      "loss": 0.6772,
+      "step": 29711
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3714679106620722,
+      "learning_rate": 4.360676556109186e-06,
+      "loss": 0.744,
+      "step": 29712
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4175222636991787,
+      "learning_rate": 4.360046437201327e-06,
+      "loss": 0.6343,
+      "step": 29713
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3361899608651506,
+      "learning_rate": 4.3594163511310596e-06,
+      "loss": 0.7849,
+      "step": 29714
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.394859912635597,
+      "learning_rate": 4.358786297902056e-06,
+      "loss": 0.6819,
+      "step": 29715
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2246113436509505,
+      "learning_rate": 4.3581562775179875e-06,
+      "loss": 0.629,
+      "step": 29716
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6409100339422105,
+      "learning_rate": 4.35752628998252e-06,
+      "loss": 0.4915,
+      "step": 29717
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.481320531787208,
+      "learning_rate": 4.356896335299317e-06,
+      "loss": 0.7708,
+      "step": 29718
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.488916814460727,
+      "learning_rate": 4.35626641347205e-06,
+      "loss": 0.5849,
+      "step": 29719
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3750129892639265,
+      "learning_rate": 4.3556365245043905e-06,
+      "loss": 0.7557,
+      "step": 29720
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7313061714894131,
+      "learning_rate": 4.355006668400003e-06,
+      "loss": 0.6654,
+      "step": 29721
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4011590076125033,
+      "learning_rate": 4.354376845162549e-06,
+      "loss": 0.7249,
+      "step": 29722
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3115048289132794,
+      "learning_rate": 4.3537470547957054e-06,
+      "loss": 0.7544,
+      "step": 29723
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.166001985174452,
+      "learning_rate": 4.353117297303129e-06,
+      "loss": 0.5987,
+      "step": 29724
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4057022037015892,
+      "learning_rate": 4.352487572688495e-06,
+      "loss": 0.6767,
+      "step": 29725
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5149613601891003,
+      "learning_rate": 4.351857880955464e-06,
+      "loss": 0.6859,
+      "step": 29726
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5821669814143096,
+      "learning_rate": 4.351228222107706e-06,
+      "loss": 0.5443,
+      "step": 29727
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5303423880698463,
+      "learning_rate": 4.350598596148884e-06,
+      "loss": 0.7091,
+      "step": 29728
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5801826321675218,
+      "learning_rate": 4.349969003082669e-06,
+      "loss": 0.5141,
+      "step": 29729
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3478767209711797,
+      "learning_rate": 4.349339442912723e-06,
+      "loss": 0.6468,
+      "step": 29730
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4153860015593465,
+      "learning_rate": 4.3487099156427075e-06,
+      "loss": 0.7381,
+      "step": 29731
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2655182258820932,
+      "learning_rate": 4.348080421276293e-06,
+      "loss": 0.6433,
+      "step": 29732
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3226260517382922,
+      "learning_rate": 4.347450959817146e-06,
+      "loss": 0.6964,
+      "step": 29733
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1868290895113844,
+      "learning_rate": 4.34682153126893e-06,
+      "loss": 0.6993,
+      "step": 29734
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5932184581417599,
+      "learning_rate": 4.346192135635305e-06,
+      "loss": 0.6078,
+      "step": 29735
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3622164897030047,
+      "learning_rate": 4.345562772919939e-06,
+      "loss": 0.7638,
+      "step": 29736
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3020789218495064,
+      "learning_rate": 4.344933443126501e-06,
+      "loss": 0.785,
+      "step": 29737
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2300443454175443,
+      "learning_rate": 4.344304146258651e-06,
+      "loss": 0.6073,
+      "step": 29738
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6007285121153418,
+      "learning_rate": 4.343674882320048e-06,
+      "loss": 0.6317,
+      "step": 29739
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5993447628162144,
+      "learning_rate": 4.343045651314361e-06,
+      "loss": 0.4744,
+      "step": 29740
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5231131717537836,
+      "learning_rate": 4.342416453245259e-06,
+      "loss": 0.6862,
+      "step": 29741
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.632198217156286,
+      "learning_rate": 4.341787288116397e-06,
+      "loss": 0.6466,
+      "step": 29742
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.31910877696295,
+      "learning_rate": 4.341158155931437e-06,
+      "loss": 0.7036,
+      "step": 29743
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.483216781663773,
+      "learning_rate": 4.3405290566940475e-06,
+      "loss": 0.6884,
+      "step": 29744
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5945334732071648,
+      "learning_rate": 4.339899990407892e-06,
+      "loss": 0.6411,
+      "step": 29745
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3659043185925626,
+      "learning_rate": 4.339270957076631e-06,
+      "loss": 0.7539,
+      "step": 29746
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0052139372030964,
+      "learning_rate": 4.338641956703923e-06,
+      "loss": 0.6273,
+      "step": 29747
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.240189511362288,
+      "learning_rate": 4.338012989293438e-06,
+      "loss": 0.7067,
+      "step": 29748
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5828986543845969,
+      "learning_rate": 4.33738405484883e-06,
+      "loss": 0.6538,
+      "step": 29749
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3788751887600763,
+      "learning_rate": 4.3367551533737685e-06,
+      "loss": 0.6492,
+      "step": 29750
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3982362820847054,
+      "learning_rate": 4.336126284871908e-06,
+      "loss": 0.7203,
+      "step": 29751
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.328770968301973,
+      "learning_rate": 4.335497449346916e-06,
+      "loss": 0.6179,
+      "step": 29752
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8785582517507242,
+      "learning_rate": 4.334868646802448e-06,
+      "loss": 0.7373,
+      "step": 29753
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3937477425315548,
+      "learning_rate": 4.334239877242172e-06,
+      "loss": 0.7456,
+      "step": 29754
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3678823933269477,
+      "learning_rate": 4.333611140669746e-06,
+      "loss": 0.7731,
+      "step": 29755
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4868779959316871,
+      "learning_rate": 4.332982437088825e-06,
+      "loss": 0.6555,
+      "step": 29756
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.43758259429803,
+      "learning_rate": 4.332353766503075e-06,
+      "loss": 0.7629,
+      "step": 29757
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4597436774250823,
+      "learning_rate": 4.331725128916157e-06,
+      "loss": 0.5991,
+      "step": 29758
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.863329033178328,
+      "learning_rate": 4.331096524331735e-06,
+      "loss": 0.6172,
+      "step": 29759
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0079876261014302,
+      "learning_rate": 4.330467952753458e-06,
+      "loss": 0.7027,
+      "step": 29760
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7654705714746184,
+      "learning_rate": 4.329839414184991e-06,
+      "loss": 0.591,
+      "step": 29761
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6594275784867532,
+      "learning_rate": 4.329210908629997e-06,
+      "loss": 0.6634,
+      "step": 29762
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3783865375512816,
+      "learning_rate": 4.328582436092133e-06,
+      "loss": 0.7288,
+      "step": 29763
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2980645051386919,
+      "learning_rate": 4.3279539965750535e-06,
+      "loss": 0.6306,
+      "step": 29764
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5862441495189779,
+      "learning_rate": 4.327325590082423e-06,
+      "loss": 0.4764,
+      "step": 29765
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5027712011793808,
+      "learning_rate": 4.326697216617901e-06,
+      "loss": 0.6128,
+      "step": 29766
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.324476422117698,
+      "learning_rate": 4.326068876185141e-06,
+      "loss": 0.6081,
+      "step": 29767
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.409200753722349,
+      "learning_rate": 4.3254405687878075e-06,
+      "loss": 0.7022,
+      "step": 29768
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3500264076878072,
+      "learning_rate": 4.324812294429557e-06,
+      "loss": 0.6574,
+      "step": 29769
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8531007118379472,
+      "learning_rate": 4.324184053114041e-06,
+      "loss": 0.8234,
+      "step": 29770
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4947074651323724,
+      "learning_rate": 4.323555844844922e-06,
+      "loss": 0.7817,
+      "step": 29771
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6290116193850717,
+      "learning_rate": 4.3229276696258625e-06,
+      "loss": 0.4837,
+      "step": 29772
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5052220010273123,
+      "learning_rate": 4.3222995274605155e-06,
+      "loss": 0.6503,
+      "step": 29773
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3891102287843662,
+      "learning_rate": 4.321671418352534e-06,
+      "loss": 0.7241,
+      "step": 29774
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3569672746320114,
+      "learning_rate": 4.3210433423055795e-06,
+      "loss": 0.5988,
+      "step": 29775
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2359784956220092,
+      "learning_rate": 4.320415299323316e-06,
+      "loss": 0.699,
+      "step": 29776
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8942857035222262,
+      "learning_rate": 4.319787289409385e-06,
+      "loss": 0.7012,
+      "step": 29777
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4178876795569797,
+      "learning_rate": 4.319159312567451e-06,
+      "loss": 0.7387,
+      "step": 29778
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6310541875524933,
+      "learning_rate": 4.318531368801171e-06,
+      "loss": 0.6348,
+      "step": 29779
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2303641630148385,
+      "learning_rate": 4.317903458114202e-06,
+      "loss": 0.6595,
+      "step": 29780
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4478287051013687,
+      "learning_rate": 4.317275580510199e-06,
+      "loss": 0.7747,
+      "step": 29781
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3423087425915918,
+      "learning_rate": 4.316647735992814e-06,
+      "loss": 0.8178,
+      "step": 29782
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3075854610183166,
+      "learning_rate": 4.316019924565708e-06,
+      "loss": 0.6552,
+      "step": 29783
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4219174163925867,
+      "learning_rate": 4.315392146232529e-06,
+      "loss": 0.6684,
+      "step": 29784
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.389844250900475,
+      "learning_rate": 4.31476440099694e-06,
+      "loss": 0.7129,
+      "step": 29785
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4959418871363095,
+      "learning_rate": 4.31413668886259e-06,
+      "loss": 0.663,
+      "step": 29786
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3649830532597222,
+      "learning_rate": 4.31350900983314e-06,
+      "loss": 0.6793,
+      "step": 29787
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4701909080837103,
+      "learning_rate": 4.312881363912237e-06,
+      "loss": 0.7264,
+      "step": 29788
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4165003282922761,
+      "learning_rate": 4.312253751103544e-06,
+      "loss": 0.7395,
+      "step": 29789
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.611907037728475,
+      "learning_rate": 4.311626171410707e-06,
+      "loss": 0.6546,
+      "step": 29790
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6187846818413056,
+      "learning_rate": 4.310998624837383e-06,
+      "loss": 0.8155,
+      "step": 29791
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2881177440369258,
+      "learning_rate": 4.310371111387225e-06,
+      "loss": 0.6693,
+      "step": 29792
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3813022908292052,
+      "learning_rate": 4.309743631063892e-06,
+      "loss": 0.6751,
+      "step": 29793
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1975262153732238,
+      "learning_rate": 4.309116183871032e-06,
+      "loss": 0.6435,
+      "step": 29794
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3796136762915359,
+      "learning_rate": 4.308488769812296e-06,
+      "loss": 0.6475,
+      "step": 29795
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3289392909264974,
+      "learning_rate": 4.30786138889134e-06,
+      "loss": 0.5933,
+      "step": 29796
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4376768831504243,
+      "learning_rate": 4.30723404111182e-06,
+      "loss": 0.7588,
+      "step": 29797
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4554323444313984,
+      "learning_rate": 4.306606726477388e-06,
+      "loss": 0.7013,
+      "step": 29798
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5150028690616035,
+      "learning_rate": 4.3059794449916896e-06,
+      "loss": 0.7622,
+      "step": 29799
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4630604247199948,
+      "learning_rate": 4.305352196658381e-06,
+      "loss": 0.738,
+      "step": 29800
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2901144842654195,
+      "learning_rate": 4.3047249814811185e-06,
+      "loss": 0.712,
+      "step": 29801
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6809939212489016,
+      "learning_rate": 4.30409779946355e-06,
+      "loss": 0.704,
+      "step": 29802
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1594958823074313,
+      "learning_rate": 4.303470650609325e-06,
+      "loss": 0.6874,
+      "step": 29803
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4621379097871696,
+      "learning_rate": 4.3028435349221e-06,
+      "loss": 0.5675,
+      "step": 29804
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6185563024864006,
+      "learning_rate": 4.302216452405521e-06,
+      "loss": 0.5428,
+      "step": 29805
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4430422351205505,
+      "learning_rate": 4.301589403063244e-06,
+      "loss": 0.582,
+      "step": 29806
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6153444888900624,
+      "learning_rate": 4.300962386898915e-06,
+      "loss": 0.5073,
+      "step": 29807
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.357778075556922,
+      "learning_rate": 4.3003354039161906e-06,
+      "loss": 0.72,
+      "step": 29808
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2083916792893301,
+      "learning_rate": 4.299708454118714e-06,
+      "loss": 0.604,
+      "step": 29809
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2678234563973798,
+      "learning_rate": 4.299081537510143e-06,
+      "loss": 0.6978,
+      "step": 29810
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3961868773806874,
+      "learning_rate": 4.2984546540941245e-06,
+      "loss": 0.7356,
+      "step": 29811
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2682565698368051,
+      "learning_rate": 4.297827803874304e-06,
+      "loss": 0.6439,
+      "step": 29812
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.765210285234947,
+      "learning_rate": 4.297200986854335e-06,
+      "loss": 0.6677,
+      "step": 29813
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5979576182996177,
+      "learning_rate": 4.296574203037872e-06,
+      "loss": 0.5558,
+      "step": 29814
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.443255475857962,
+      "learning_rate": 4.295947452428559e-06,
+      "loss": 0.7782,
+      "step": 29815
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3171014596558632,
+      "learning_rate": 4.295320735030043e-06,
+      "loss": 0.7322,
+      "step": 29816
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5173232928368934,
+      "learning_rate": 4.294694050845976e-06,
+      "loss": 0.7938,
+      "step": 29817
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3550751296369155,
+      "learning_rate": 4.294067399880009e-06,
+      "loss": 0.6985,
+      "step": 29818
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4199190047301007,
+      "learning_rate": 4.293440782135788e-06,
+      "loss": 0.6485,
+      "step": 29819
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4809657038379855,
+      "learning_rate": 4.292814197616958e-06,
+      "loss": 0.7196,
+      "step": 29820
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6259587989595958,
+      "learning_rate": 4.29218764632717e-06,
+      "loss": 0.5071,
+      "step": 29821
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.482181937974094,
+      "learning_rate": 4.2915611282700775e-06,
+      "loss": 0.7488,
+      "step": 29822
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2006446451248851,
+      "learning_rate": 4.290934643449324e-06,
+      "loss": 0.6777,
+      "step": 29823
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7680085701814239,
+      "learning_rate": 4.290308191868551e-06,
+      "loss": 0.6681,
+      "step": 29824
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7945591387086801,
+      "learning_rate": 4.289681773531415e-06,
+      "loss": 0.6996,
+      "step": 29825
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4096445041447099,
+      "learning_rate": 4.289055388441558e-06,
+      "loss": 0.6507,
+      "step": 29826
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5143047136976493,
+      "learning_rate": 4.28842903660263e-06,
+      "loss": 0.6599,
+      "step": 29827
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3597711760706044,
+      "learning_rate": 4.2878027180182745e-06,
+      "loss": 0.6896,
+      "step": 29828
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3649852981718638,
+      "learning_rate": 4.287176432692143e-06,
+      "loss": 0.83,
+      "step": 29829
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6443810907788299,
+      "learning_rate": 4.286550180627874e-06,
+      "loss": 0.5275,
+      "step": 29830
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4753177289090107,
+      "learning_rate": 4.285923961829124e-06,
+      "loss": 0.6992,
+      "step": 29831
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.618174721318555,
+      "learning_rate": 4.285297776299533e-06,
+      "loss": 0.6809,
+      "step": 29832
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5839845413672049,
+      "learning_rate": 4.284671624042744e-06,
+      "loss": 0.5025,
+      "step": 29833
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7385887846044334,
+      "learning_rate": 4.284045505062406e-06,
+      "loss": 0.724,
+      "step": 29834
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2263844636417365,
+      "learning_rate": 4.283419419362169e-06,
+      "loss": 0.7137,
+      "step": 29835
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.9526978964192185,
+      "learning_rate": 4.282793366945675e-06,
+      "loss": 0.6954,
+      "step": 29836
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.424211310066194,
+      "learning_rate": 4.282167347816562e-06,
+      "loss": 0.6748,
+      "step": 29837
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.369555641161617,
+      "learning_rate": 4.281541361978483e-06,
+      "loss": 0.6565,
+      "step": 29838
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6450610669164153,
+      "learning_rate": 4.280915409435084e-06,
+      "loss": 0.5457,
+      "step": 29839
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4238221440824104,
+      "learning_rate": 4.280289490190005e-06,
+      "loss": 0.702,
+      "step": 29840
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1779188400749447,
+      "learning_rate": 4.279663604246888e-06,
+      "loss": 0.6917,
+      "step": 29841
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6797979037243093,
+      "learning_rate": 4.27903775160938e-06,
+      "loss": 0.5883,
+      "step": 29842
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3742410527220221,
+      "learning_rate": 4.27841193228113e-06,
+      "loss": 0.7076,
+      "step": 29843
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6096769188281075,
+      "learning_rate": 4.277786146265778e-06,
+      "loss": 0.5186,
+      "step": 29844
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.233809624752604,
+      "learning_rate": 4.277160393566962e-06,
+      "loss": 0.8513,
+      "step": 29845
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5783877582726759,
+      "learning_rate": 4.2765346741883325e-06,
+      "loss": 0.4905,
+      "step": 29846
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4539573065199127,
+      "learning_rate": 4.275908988133528e-06,
+      "loss": 0.6861,
+      "step": 29847
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.206968235696648,
+      "learning_rate": 4.275283335406197e-06,
+      "loss": 0.6183,
+      "step": 29848
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.601785333835043,
+      "learning_rate": 4.274657716009976e-06,
+      "loss": 0.7602,
+      "step": 29849
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4308368073457594,
+      "learning_rate": 4.274032129948512e-06,
+      "loss": 0.7233,
+      "step": 29850
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6187074339361728,
+      "learning_rate": 4.273406577225444e-06,
+      "loss": 0.5303,
+      "step": 29851
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4389337012097512,
+      "learning_rate": 4.272781057844418e-06,
+      "loss": 0.7626,
+      "step": 29852
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3659980897434003,
+      "learning_rate": 4.272155571809076e-06,
+      "loss": 0.7705,
+      "step": 29853
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4953230009227523,
+      "learning_rate": 4.271530119123052e-06,
+      "loss": 0.703,
+      "step": 29854
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.594772250649003,
+      "learning_rate": 4.270904699789994e-06,
+      "loss": 0.5661,
+      "step": 29855
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5867254270623845,
+      "learning_rate": 4.270279313813542e-06,
+      "loss": 0.8448,
+      "step": 29856
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2581115784365704,
+      "learning_rate": 4.269653961197346e-06,
+      "loss": 0.6711,
+      "step": 29857
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.428682966076229,
+      "learning_rate": 4.269028641945032e-06,
+      "loss": 0.617,
+      "step": 29858
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6623565662000008,
+      "learning_rate": 4.2684033560602465e-06,
+      "loss": 0.6017,
+      "step": 29859
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4241726528420708,
+      "learning_rate": 4.267778103546632e-06,
+      "loss": 0.7275,
+      "step": 29860
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9237814864551226,
+      "learning_rate": 4.267152884407835e-06,
+      "loss": 0.7129,
+      "step": 29861
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.284041546215039,
+      "learning_rate": 4.266527698647482e-06,
+      "loss": 0.7061,
+      "step": 29862
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.071406317648011,
+      "learning_rate": 4.265902546269219e-06,
+      "loss": 0.6734,
+      "step": 29863
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6076793241963778,
+      "learning_rate": 4.265277427276692e-06,
+      "loss": 0.7177,
+      "step": 29864
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3276196035462713,
+      "learning_rate": 4.264652341673531e-06,
+      "loss": 0.7277,
+      "step": 29865
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6540427732695575,
+      "learning_rate": 4.2640272894633825e-06,
+      "loss": 0.7255,
+      "step": 29866
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.593313217518136,
+      "learning_rate": 4.2634022706498815e-06,
+      "loss": 0.6458,
+      "step": 29867
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6151769787073813,
+      "learning_rate": 4.26277728523667e-06,
+      "loss": 0.6838,
+      "step": 29868
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2920011810459524,
+      "learning_rate": 4.262152333227383e-06,
+      "loss": 0.6713,
+      "step": 29869
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3664005165391486,
+      "learning_rate": 4.2615274146256654e-06,
+      "loss": 0.8352,
+      "step": 29870
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4939716022328915,
+      "learning_rate": 4.2609025294351525e-06,
+      "loss": 0.6793,
+      "step": 29871
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5124936481077036,
+      "learning_rate": 4.260277677659477e-06,
+      "loss": 0.6742,
+      "step": 29872
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4577351152006475,
+      "learning_rate": 4.2596528593022835e-06,
+      "loss": 0.535,
+      "step": 29873
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3364501838482339,
+      "learning_rate": 4.259028074367211e-06,
+      "loss": 0.7622,
+      "step": 29874
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6468396547424751,
+      "learning_rate": 4.258403322857895e-06,
+      "loss": 0.5321,
+      "step": 29875
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3174357379533832,
+      "learning_rate": 4.257778604777969e-06,
+      "loss": 0.6751,
+      "step": 29876
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.490997809023855,
+      "learning_rate": 4.257153920131074e-06,
+      "loss": 0.7555,
+      "step": 29877
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.33295259031097,
+      "learning_rate": 4.256529268920851e-06,
+      "loss": 0.6661,
+      "step": 29878
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4543228043951104,
+      "learning_rate": 4.255904651150931e-06,
+      "loss": 0.6752,
+      "step": 29879
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.483420645204306,
+      "learning_rate": 4.255280066824952e-06,
+      "loss": 0.6141,
+      "step": 29880
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4799486364301166,
+      "learning_rate": 4.254655515946549e-06,
+      "loss": 0.6865,
+      "step": 29881
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4166963000697907,
+      "learning_rate": 4.254030998519365e-06,
+      "loss": 0.6602,
+      "step": 29882
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5535059473408406,
+      "learning_rate": 4.253406514547031e-06,
+      "loss": 0.7145,
+      "step": 29883
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3585280153847923,
+      "learning_rate": 4.25278206403318e-06,
+      "loss": 0.5668,
+      "step": 29884
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7831571306263154,
+      "learning_rate": 4.252157646981457e-06,
+      "loss": 0.6718,
+      "step": 29885
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2305437917511997,
+      "learning_rate": 4.251533263395486e-06,
+      "loss": 0.642,
+      "step": 29886
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6524323823804996,
+      "learning_rate": 4.2509089132789126e-06,
+      "loss": 0.685,
+      "step": 29887
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8908301691366767,
+      "learning_rate": 4.250284596635365e-06,
+      "loss": 0.667,
+      "step": 29888
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.371643538713236,
+      "learning_rate": 4.249660313468483e-06,
+      "loss": 0.6588,
+      "step": 29889
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2515495399388576,
+      "learning_rate": 4.2490360637818965e-06,
+      "loss": 0.6504,
+      "step": 29890
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5950303123763925,
+      "learning_rate": 4.248411847579245e-06,
+      "loss": 0.7552,
+      "step": 29891
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3575862261449136,
+      "learning_rate": 4.247787664864163e-06,
+      "loss": 0.7732,
+      "step": 29892
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2977209724237744,
+      "learning_rate": 4.247163515640277e-06,
+      "loss": 0.624,
+      "step": 29893
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3036604515669374,
+      "learning_rate": 4.2465393999112256e-06,
+      "loss": 0.7165,
+      "step": 29894
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2524416383014338,
+      "learning_rate": 4.245915317680649e-06,
+      "loss": 0.6592,
+      "step": 29895
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.293069125140176,
+      "learning_rate": 4.245291268952174e-06,
+      "loss": 0.6348,
+      "step": 29896
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3895551913030777,
+      "learning_rate": 4.244667253729431e-06,
+      "loss": 0.7296,
+      "step": 29897
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3333265485268257,
+      "learning_rate": 4.244043272016059e-06,
+      "loss": 0.6841,
+      "step": 29898
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2666479701732107,
+      "learning_rate": 4.243419323815692e-06,
+      "loss": 0.6384,
+      "step": 29899
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4032007978166336,
+      "learning_rate": 4.24279540913196e-06,
+      "loss": 0.6704,
+      "step": 29900
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.297259830303271,
+      "learning_rate": 4.242171527968492e-06,
+      "loss": 0.7047,
+      "step": 29901
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3479983449297128,
+      "learning_rate": 4.241547680328924e-06,
+      "loss": 0.7009,
+      "step": 29902
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4454682703082822,
+      "learning_rate": 4.240923866216894e-06,
+      "loss": 0.6882,
+      "step": 29903
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3938655295555213,
+      "learning_rate": 4.240300085636027e-06,
+      "loss": 0.6578,
+      "step": 29904
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5064035652773275,
+      "learning_rate": 4.239676338589953e-06,
+      "loss": 0.6403,
+      "step": 29905
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1089076478534505,
+      "learning_rate": 4.239052625082311e-06,
+      "loss": 0.6876,
+      "step": 29906
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.509561520369769,
+      "learning_rate": 4.238428945116725e-06,
+      "loss": 0.6662,
+      "step": 29907
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.371358126013847,
+      "learning_rate": 4.237805298696832e-06,
+      "loss": 0.6789,
+      "step": 29908
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5995049144518474,
+      "learning_rate": 4.23718168582626e-06,
+      "loss": 0.4956,
+      "step": 29909
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1298418384385287,
+      "learning_rate": 4.236558106508641e-06,
+      "loss": 0.5797,
+      "step": 29910
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7389154724961047,
+      "learning_rate": 4.235934560747604e-06,
+      "loss": 0.6898,
+      "step": 29911
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6789755056401043,
+      "learning_rate": 4.2353110485467845e-06,
+      "loss": 0.7321,
+      "step": 29912
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6307252114273239,
+      "learning_rate": 4.234687569909808e-06,
+      "loss": 0.4317,
+      "step": 29913
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4567666280482052,
+      "learning_rate": 4.234064124840302e-06,
+      "loss": 0.7631,
+      "step": 29914
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3158540531537692,
+      "learning_rate": 4.233440713341901e-06,
+      "loss": 0.7762,
+      "step": 29915
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3792918069217708,
+      "learning_rate": 4.232817335418236e-06,
+      "loss": 0.6343,
+      "step": 29916
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3439350299889858,
+      "learning_rate": 4.232193991072937e-06,
+      "loss": 0.7343,
+      "step": 29917
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.531275845275778,
+      "learning_rate": 4.231570680309626e-06,
+      "loss": 0.6337,
+      "step": 29918
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.498106513204368,
+      "learning_rate": 4.230947403131935e-06,
+      "loss": 0.6585,
+      "step": 29919
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6237095467460391,
+      "learning_rate": 4.2303241595435e-06,
+      "loss": 0.5407,
+      "step": 29920
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3654717342876557,
+      "learning_rate": 4.229700949547944e-06,
+      "loss": 0.6758,
+      "step": 29921
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3791597447978763,
+      "learning_rate": 4.229077773148892e-06,
+      "loss": 0.628,
+      "step": 29922
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2361052782481905,
+      "learning_rate": 4.228454630349977e-06,
+      "loss": 0.5616,
+      "step": 29923
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2733241747462674,
+      "learning_rate": 4.227831521154829e-06,
+      "loss": 0.7078,
+      "step": 29924
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3757674177443893,
+      "learning_rate": 4.227208445567075e-06,
+      "loss": 0.6824,
+      "step": 29925
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2108002748698576,
+      "learning_rate": 4.226585403590335e-06,
+      "loss": 0.6583,
+      "step": 29926
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.585493977368275,
+      "learning_rate": 4.225962395228248e-06,
+      "loss": 0.4928,
+      "step": 29927
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0608450229747515,
+      "learning_rate": 4.225339420484432e-06,
+      "loss": 0.7546,
+      "step": 29928
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4400633999965247,
+      "learning_rate": 4.224716479362521e-06,
+      "loss": 0.6696,
+      "step": 29929
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.009839089075504,
+      "learning_rate": 4.2240935718661365e-06,
+      "loss": 0.7509,
+      "step": 29930
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3575701177432538,
+      "learning_rate": 4.223470697998911e-06,
+      "loss": 0.5505,
+      "step": 29931
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4560391456158628,
+      "learning_rate": 4.222847857764464e-06,
+      "loss": 0.7401,
+      "step": 29932
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8190685758579235,
+      "learning_rate": 4.22222505116643e-06,
+      "loss": 0.7564,
+      "step": 29933
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7003468811944338,
+      "learning_rate": 4.221602278208431e-06,
+      "loss": 0.6911,
+      "step": 29934
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.366944089278871,
+      "learning_rate": 4.220979538894089e-06,
+      "loss": 0.6393,
+      "step": 29935
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5885578600981238,
+      "learning_rate": 4.220356833227034e-06,
+      "loss": 0.5335,
+      "step": 29936
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5155633541882683,
+      "learning_rate": 4.2197341612108944e-06,
+      "loss": 0.6668,
+      "step": 29937
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3862578316097862,
+      "learning_rate": 4.219111522849293e-06,
+      "loss": 0.6481,
+      "step": 29938
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.339633661281078,
+      "learning_rate": 4.21848891814585e-06,
+      "loss": 0.6088,
+      "step": 29939
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7528395644961363,
+      "learning_rate": 4.217866347104196e-06,
+      "loss": 0.7198,
+      "step": 29940
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.282323911518001,
+      "learning_rate": 4.217243809727958e-06,
+      "loss": 0.739,
+      "step": 29941
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.437275606652009,
+      "learning_rate": 4.216621306020757e-06,
+      "loss": 0.7614,
+      "step": 29942
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4269092600174365,
+      "learning_rate": 4.215998835986215e-06,
+      "loss": 0.7173,
+      "step": 29943
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5346496560191756,
+      "learning_rate": 4.215376399627959e-06,
+      "loss": 0.7733,
+      "step": 29944
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4903549733421413,
+      "learning_rate": 4.214753996949616e-06,
+      "loss": 0.6196,
+      "step": 29945
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.607041105947859,
+      "learning_rate": 4.2141316279548065e-06,
+      "loss": 0.4916,
+      "step": 29946
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5375740275777503,
+      "learning_rate": 4.21350929264715e-06,
+      "loss": 0.7254,
+      "step": 29947
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3978650311601946,
+      "learning_rate": 4.21288699103028e-06,
+      "loss": 0.7289,
+      "step": 29948
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2598606310653566,
+      "learning_rate": 4.21226472310781e-06,
+      "loss": 0.7195,
+      "step": 29949
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.50344242975325,
+      "learning_rate": 4.211642488883369e-06,
+      "loss": 0.5632,
+      "step": 29950
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4452415726791663,
+      "learning_rate": 4.211020288360576e-06,
+      "loss": 0.6442,
+      "step": 29951
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3913275383224928,
+      "learning_rate": 4.210398121543059e-06,
+      "loss": 0.5598,
+      "step": 29952
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.7886218487478045,
+      "learning_rate": 4.2097759884344335e-06,
+      "loss": 0.7942,
+      "step": 29953
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3038262600425468,
+      "learning_rate": 4.209153889038325e-06,
+      "loss": 0.6974,
+      "step": 29954
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4324228597043154,
+      "learning_rate": 4.208531823358364e-06,
+      "loss": 0.8036,
+      "step": 29955
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6382296779652141,
+      "learning_rate": 4.207909791398156e-06,
+      "loss": 0.6092,
+      "step": 29956
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6171787365224453,
+      "learning_rate": 4.207287793161332e-06,
+      "loss": 0.5332,
+      "step": 29957
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.299341755571432,
+      "learning_rate": 4.2066658286515115e-06,
+      "loss": 0.5653,
+      "step": 29958
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3814839477246126,
+      "learning_rate": 4.206043897872324e-06,
+      "loss": 0.6076,
+      "step": 29959
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5704192171713323,
+      "learning_rate": 4.2054220008273765e-06,
+      "loss": 0.5153,
+      "step": 29960
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.331029386760525,
+      "learning_rate": 4.204800137520296e-06,
+      "loss": 0.7506,
+      "step": 29961
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8353260275606134,
+      "learning_rate": 4.204178307954708e-06,
+      "loss": 0.755,
+      "step": 29962
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6271902437623769,
+      "learning_rate": 4.203556512134225e-06,
+      "loss": 0.7284,
+      "step": 29963
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2968998298881151,
+      "learning_rate": 4.202934750062475e-06,
+      "loss": 0.7078,
+      "step": 29964
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6114679953159855,
+      "learning_rate": 4.202313021743071e-06,
+      "loss": 0.5521,
+      "step": 29965
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.391151818895334,
+      "learning_rate": 4.201691327179639e-06,
+      "loss": 0.6673,
+      "step": 29966
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2449281764949083,
+      "learning_rate": 4.201069666375792e-06,
+      "loss": 0.69,
+      "step": 29967
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6004392081705571,
+      "learning_rate": 4.2004480393351565e-06,
+      "loss": 0.5152,
+      "step": 29968
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4927896047796874,
+      "learning_rate": 4.19982644606135e-06,
+      "loss": 0.6889,
+      "step": 29969
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.285413505719597,
+      "learning_rate": 4.199204886557986e-06,
+      "loss": 0.7227,
+      "step": 29970
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3011439604698263,
+      "learning_rate": 4.198583360828688e-06,
+      "loss": 0.7201,
+      "step": 29971
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4526365450049545,
+      "learning_rate": 4.197961868877078e-06,
+      "loss": 0.7049,
+      "step": 29972
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5612791870186251,
+      "learning_rate": 4.19734041070677e-06,
+      "loss": 0.4889,
+      "step": 29973
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4014747464334383,
+      "learning_rate": 4.19671898632138e-06,
+      "loss": 0.7365,
+      "step": 29974
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.844488412435815,
+      "learning_rate": 4.1960975957245285e-06,
+      "loss": 0.7152,
+      "step": 29975
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.485183494540969,
+      "learning_rate": 4.19547623891984e-06,
+      "loss": 0.6694,
+      "step": 29976
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4646424017053563,
+      "learning_rate": 4.194854915910925e-06,
+      "loss": 0.7311,
+      "step": 29977
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3250818090631937,
+      "learning_rate": 4.194233626701399e-06,
+      "loss": 0.6624,
+      "step": 29978
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7536464933656761,
+      "learning_rate": 4.193612371294883e-06,
+      "loss": 0.6807,
+      "step": 29979
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2373608095984794,
+      "learning_rate": 4.1929911496949985e-06,
+      "loss": 0.6687,
+      "step": 29980
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8720447206339887,
+      "learning_rate": 4.192369961905357e-06,
+      "loss": 0.7939,
+      "step": 29981
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.372164038822348,
+      "learning_rate": 4.1917488079295735e-06,
+      "loss": 0.7866,
+      "step": 29982
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3265038713660866,
+      "learning_rate": 4.191127687771266e-06,
+      "loss": 0.7646,
+      "step": 29983
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4137756691037133,
+      "learning_rate": 4.190506601434057e-06,
+      "loss": 0.6747,
+      "step": 29984
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3698515514825633,
+      "learning_rate": 4.1898855489215574e-06,
+      "loss": 0.6211,
+      "step": 29985
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.455988090199081,
+      "learning_rate": 4.18926453023738e-06,
+      "loss": 0.7461,
+      "step": 29986
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0141602324066725,
+      "learning_rate": 4.188643545385148e-06,
+      "loss": 0.7228,
+      "step": 29987
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4458072681673608,
+      "learning_rate": 4.1880225943684695e-06,
+      "loss": 0.8671,
+      "step": 29988
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.297133040543281,
+      "learning_rate": 4.187401677190966e-06,
+      "loss": 0.7628,
+      "step": 29989
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.481365466763657,
+      "learning_rate": 4.186780793856247e-06,
+      "loss": 0.7079,
+      "step": 29990
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9907859516850237,
+      "learning_rate": 4.186159944367936e-06,
+      "loss": 0.5523,
+      "step": 29991
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.609687254735402,
+      "learning_rate": 4.185539128729636e-06,
+      "loss": 0.7149,
+      "step": 29992
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3172400380744222,
+      "learning_rate": 4.184918346944974e-06,
+      "loss": 0.5586,
+      "step": 29993
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.498919818969349,
+      "learning_rate": 4.1842975990175575e-06,
+      "loss": 0.8822,
+      "step": 29994
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3265862295899056,
+      "learning_rate": 4.183676884950998e-06,
+      "loss": 0.6312,
+      "step": 29995
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3019919424568076,
+      "learning_rate": 4.183056204748913e-06,
+      "loss": 0.7213,
+      "step": 29996
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3204184671629497,
+      "learning_rate": 4.18243555841492e-06,
+      "loss": 0.7161,
+      "step": 29997
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.9870066192234654,
+      "learning_rate": 4.181814945952629e-06,
+      "loss": 0.6595,
+      "step": 29998
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2916059576514187,
+      "learning_rate": 4.18119436736565e-06,
+      "loss": 0.6738,
+      "step": 29999
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4015922540832124,
+      "learning_rate": 4.1805738226575995e-06,
+      "loss": 0.5182,
+      "step": 30000
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2514562145700119,
+      "learning_rate": 4.179953311832094e-06,
+      "loss": 0.6387,
+      "step": 30001
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3989875903276596,
+      "learning_rate": 4.179332834892743e-06,
+      "loss": 0.75,
+      "step": 30002
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6364504484240116,
+      "learning_rate": 4.178712391843156e-06,
+      "loss": 0.5314,
+      "step": 30003
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4024454238687427,
+      "learning_rate": 4.178091982686948e-06,
+      "loss": 0.5561,
+      "step": 30004
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.501306365292641,
+      "learning_rate": 4.177471607427736e-06,
+      "loss": 0.6734,
+      "step": 30005
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.588605740731968,
+      "learning_rate": 4.176851266069126e-06,
+      "loss": 0.7183,
+      "step": 30006
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4015338598798937,
+      "learning_rate": 4.176230958614729e-06,
+      "loss": 0.7361,
+      "step": 30007
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4513983881836836,
+      "learning_rate": 4.1756106850681624e-06,
+      "loss": 0.7895,
+      "step": 30008
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4925628942266547,
+      "learning_rate": 4.174990445433032e-06,
+      "loss": 0.6825,
+      "step": 30009
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2469819190224825,
+      "learning_rate": 4.174370239712952e-06,
+      "loss": 0.5741,
+      "step": 30010
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.935837470645075,
+      "learning_rate": 4.173750067911532e-06,
+      "loss": 0.6595,
+      "step": 30011
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3086454591658538,
+      "learning_rate": 4.173129930032387e-06,
+      "loss": 0.7769,
+      "step": 30012
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4812946156591256,
+      "learning_rate": 4.17250982607912e-06,
+      "loss": 0.6731,
+      "step": 30013
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2123314141823411,
+      "learning_rate": 4.17188975605535e-06,
+      "loss": 0.7059,
+      "step": 30014
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.277731532113125,
+      "learning_rate": 4.171269719964681e-06,
+      "loss": 0.6712,
+      "step": 30015
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4213687739018388,
+      "learning_rate": 4.170649717810723e-06,
+      "loss": 0.7001,
+      "step": 30016
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4415635260633273,
+      "learning_rate": 4.170029749597087e-06,
+      "loss": 0.762,
+      "step": 30017
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8159601240584107,
+      "learning_rate": 4.169409815327388e-06,
+      "loss": 0.7143,
+      "step": 30018
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4332223664604582,
+      "learning_rate": 4.168789915005233e-06,
+      "loss": 0.5862,
+      "step": 30019
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3306269577667307,
+      "learning_rate": 4.168170048634223e-06,
+      "loss": 0.6849,
+      "step": 30020
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5755862417050776,
+      "learning_rate": 4.167550216217974e-06,
+      "loss": 0.4814,
+      "step": 30021
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2717672377896962,
+      "learning_rate": 4.1669304177600964e-06,
+      "loss": 0.6172,
+      "step": 30022
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3140665450734457,
+      "learning_rate": 4.166310653264199e-06,
+      "loss": 0.781,
+      "step": 30023
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.41555871436112,
+      "learning_rate": 4.1656909227338826e-06,
+      "loss": 0.6899,
+      "step": 30024
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5579577382935135,
+      "learning_rate": 4.1650712261727614e-06,
+      "loss": 0.7164,
+      "step": 30025
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4566545345214066,
+      "learning_rate": 4.164451563584446e-06,
+      "loss": 0.772,
+      "step": 30026
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2100271530930409,
+      "learning_rate": 4.163831934972541e-06,
+      "loss": 0.5616,
+      "step": 30027
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3226144157533455,
+      "learning_rate": 4.163212340340651e-06,
+      "loss": 0.7304,
+      "step": 30028
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.390150525321723,
+      "learning_rate": 4.16259277969239e-06,
+      "loss": 0.7733,
+      "step": 30029
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3599959606449874,
+      "learning_rate": 4.161973253031359e-06,
+      "loss": 0.6951,
+      "step": 30030
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3707756118470673,
+      "learning_rate": 4.161353760361171e-06,
+      "loss": 0.5725,
+      "step": 30031
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2429647211877795,
+      "learning_rate": 4.160734301685427e-06,
+      "loss": 0.5839,
+      "step": 30032
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4168640971620763,
+      "learning_rate": 4.160114877007739e-06,
+      "loss": 0.5626,
+      "step": 30033
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8469789245131558,
+      "learning_rate": 4.159495486331708e-06,
+      "loss": 0.6455,
+      "step": 30034
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2841835378749584,
+      "learning_rate": 4.1588761296609485e-06,
+      "loss": 0.6876,
+      "step": 30035
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3545385235710092,
+      "learning_rate": 4.158256806999059e-06,
+      "loss": 0.6469,
+      "step": 30036
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2973456327038724,
+      "learning_rate": 4.157637518349646e-06,
+      "loss": 0.6225,
+      "step": 30037
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4263895437342466,
+      "learning_rate": 4.1570182637163155e-06,
+      "loss": 0.6142,
+      "step": 30038
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4526686381591731,
+      "learning_rate": 4.156399043102675e-06,
+      "loss": 0.6436,
+      "step": 30039
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.962583626848756,
+      "learning_rate": 4.155779856512338e-06,
+      "loss": 0.699,
+      "step": 30040
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6250310791908134,
+      "learning_rate": 4.155160703948892e-06,
+      "loss": 0.6779,
+      "step": 30041
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3706278732882204,
+      "learning_rate": 4.154541585415951e-06,
+      "loss": 0.714,
+      "step": 30042
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6654831480057835,
+      "learning_rate": 4.153922500917123e-06,
+      "loss": 0.7462,
+      "step": 30043
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4495191435277435,
+      "learning_rate": 4.15330345045601e-06,
+      "loss": 0.6902,
+      "step": 30044
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8124196469747864,
+      "learning_rate": 4.152684434036212e-06,
+      "loss": 0.5543,
+      "step": 30045
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4068696327586003,
+      "learning_rate": 4.152065451661334e-06,
+      "loss": 0.7216,
+      "step": 30046
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.962772827207844,
+      "learning_rate": 4.151446503334988e-06,
+      "loss": 0.6499,
+      "step": 30047
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3137681125638738,
+      "learning_rate": 4.150827589060769e-06,
+      "loss": 0.6502,
+      "step": 30048
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5898041617014321,
+      "learning_rate": 4.150208708842286e-06,
+      "loss": 0.4931,
+      "step": 30049
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3821804971004814,
+      "learning_rate": 4.149589862683141e-06,
+      "loss": 0.4893,
+      "step": 30050
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6067951784391105,
+      "learning_rate": 4.1489710505869305e-06,
+      "loss": 0.4977,
+      "step": 30051
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3036476041737042,
+      "learning_rate": 4.148352272557264e-06,
+      "loss": 0.6777,
+      "step": 30052
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5483832726339783,
+      "learning_rate": 4.147733528597748e-06,
+      "loss": 0.7254,
+      "step": 30053
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.860717934092665,
+      "learning_rate": 4.147114818711979e-06,
+      "loss": 0.4993,
+      "step": 30054
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5882019969164652,
+      "learning_rate": 4.146496142903557e-06,
+      "loss": 0.5289,
+      "step": 30055
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5653260730807999,
+      "learning_rate": 4.145877501176088e-06,
+      "loss": 0.6228,
+      "step": 30056
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6750530874843539,
+      "learning_rate": 4.145258893533179e-06,
+      "loss": 0.781,
+      "step": 30057
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3855184305262853,
+      "learning_rate": 4.144640319978422e-06,
+      "loss": 0.6727,
+      "step": 30058
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3250108827635898,
+      "learning_rate": 4.144021780515421e-06,
+      "loss": 0.6311,
+      "step": 30059
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4966696997443296,
+      "learning_rate": 4.1434032751477794e-06,
+      "loss": 0.6351,
+      "step": 30060
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5945791954708055,
+      "learning_rate": 4.142784803879102e-06,
+      "loss": 0.6904,
+      "step": 30061
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4303063008482166,
+      "learning_rate": 4.142166366712985e-06,
+      "loss": 0.7457,
+      "step": 30062
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7067266502668608,
+      "learning_rate": 4.141547963653027e-06,
+      "loss": 0.7177,
+      "step": 30063
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3120910270098303,
+      "learning_rate": 4.140929594702836e-06,
+      "loss": 0.71,
+      "step": 30064
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3337713508443139,
+      "learning_rate": 4.140311259866003e-06,
+      "loss": 0.7759,
+      "step": 30065
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.436442601217575,
+      "learning_rate": 4.139692959146138e-06,
+      "loss": 0.6677,
+      "step": 30066
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.387166313683159,
+      "learning_rate": 4.139074692546831e-06,
+      "loss": 0.7144,
+      "step": 30067
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2290284672031717,
+      "learning_rate": 4.138456460071691e-06,
+      "loss": 0.5901,
+      "step": 30068
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4160234908798037,
+      "learning_rate": 4.137838261724308e-06,
+      "loss": 0.6283,
+      "step": 30069
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2705075707140332,
+      "learning_rate": 4.137220097508292e-06,
+      "loss": 0.6679,
+      "step": 30070
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4520034550614527,
+      "learning_rate": 4.136601967427236e-06,
+      "loss": 0.7164,
+      "step": 30071
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5849785751589732,
+      "learning_rate": 4.135983871484735e-06,
+      "loss": 0.5341,
+      "step": 30072
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3136320808338964,
+      "learning_rate": 4.135365809684394e-06,
+      "loss": 0.6371,
+      "step": 30073
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6106734290944604,
+      "learning_rate": 4.134747782029814e-06,
+      "loss": 0.6487,
+      "step": 30074
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4480368468584321,
+      "learning_rate": 4.134129788524588e-06,
+      "loss": 0.731,
+      "step": 30075
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3433079588410406,
+      "learning_rate": 4.133511829172313e-06,
+      "loss": 0.7187,
+      "step": 30076
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0460885169616727,
+      "learning_rate": 4.132893903976588e-06,
+      "loss": 0.8157,
+      "step": 30077
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3373609875300156,
+      "learning_rate": 4.132276012941016e-06,
+      "loss": 0.6866,
+      "step": 30078
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2548027116795812,
+      "learning_rate": 4.1316581560691924e-06,
+      "loss": 0.7817,
+      "step": 30079
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1662058779568387,
+      "learning_rate": 4.131040333364708e-06,
+      "loss": 0.6926,
+      "step": 30080
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4804354207205093,
+      "learning_rate": 4.130422544831165e-06,
+      "loss": 0.7754,
+      "step": 30081
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5482826633343727,
+      "learning_rate": 4.129804790472164e-06,
+      "loss": 0.7803,
+      "step": 30082
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4227949644220594,
+      "learning_rate": 4.129187070291298e-06,
+      "loss": 0.7766,
+      "step": 30083
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.336804269827972,
+      "learning_rate": 4.128569384292161e-06,
+      "loss": 0.5954,
+      "step": 30084
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3920066325935663,
+      "learning_rate": 4.127951732478353e-06,
+      "loss": 0.6535,
+      "step": 30085
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4112445995843594,
+      "learning_rate": 4.127334114853467e-06,
+      "loss": 0.7205,
+      "step": 30086
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.508749740895073,
+      "learning_rate": 4.126716531421105e-06,
+      "loss": 0.657,
+      "step": 30087
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7705748314955628,
+      "learning_rate": 4.126098982184854e-06,
+      "loss": 0.5059,
+      "step": 30088
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2684722601042813,
+      "learning_rate": 4.125481467148319e-06,
+      "loss": 0.665,
+      "step": 30089
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5247685350613391,
+      "learning_rate": 4.124863986315087e-06,
+      "loss": 0.6637,
+      "step": 30090
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3287733770185404,
+      "learning_rate": 4.1242465396887595e-06,
+      "loss": 0.6576,
+      "step": 30091
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3936375720603045,
+      "learning_rate": 4.12362912727293e-06,
+      "loss": 0.6094,
+      "step": 30092
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5723899514055097,
+      "learning_rate": 4.123011749071188e-06,
+      "loss": 0.5156,
+      "step": 30093
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5985202950958716,
+      "learning_rate": 4.122394405087132e-06,
+      "loss": 0.5111,
+      "step": 30094
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.195505051095464,
+      "learning_rate": 4.121777095324361e-06,
+      "loss": 0.6456,
+      "step": 30095
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3984328645865438,
+      "learning_rate": 4.121159819786465e-06,
+      "loss": 0.78,
+      "step": 30096
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5902380106716352,
+      "learning_rate": 4.120542578477033e-06,
+      "loss": 0.7345,
+      "step": 30097
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3526440001258546,
+      "learning_rate": 4.119925371399665e-06,
+      "loss": 0.5145,
+      "step": 30098
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5873890884168047,
+      "learning_rate": 4.119308198557955e-06,
+      "loss": 0.5471,
+      "step": 30099
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4900541480912446,
+      "learning_rate": 4.1186910599554954e-06,
+      "loss": 0.693,
+      "step": 30100
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.757171813407195,
+      "learning_rate": 4.118073955595873e-06,
+      "loss": 0.6478,
+      "step": 30101
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2751014958031541,
+      "learning_rate": 4.117456885482689e-06,
+      "loss": 0.7028,
+      "step": 30102
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7030900463574723,
+      "learning_rate": 4.116839849619536e-06,
+      "loss": 0.7886,
+      "step": 30103
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3561793061183898,
+      "learning_rate": 4.116222848010004e-06,
+      "loss": 0.7223,
+      "step": 30104
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4989349547542157,
+      "learning_rate": 4.1156058806576815e-06,
+      "loss": 0.6244,
+      "step": 30105
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.272026867302253,
+      "learning_rate": 4.114988947566167e-06,
+      "loss": 0.7322,
+      "step": 30106
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.31835453338308,
+      "learning_rate": 4.114372048739048e-06,
+      "loss": 0.7642,
+      "step": 30107
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.424045261419761,
+      "learning_rate": 4.113755184179922e-06,
+      "loss": 0.7195,
+      "step": 30108
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8876571681252707,
+      "learning_rate": 4.113138353892373e-06,
+      "loss": 0.5877,
+      "step": 30109
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3058778645934155,
+      "learning_rate": 4.11252155788e-06,
+      "loss": 0.6714,
+      "step": 30110
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5192786316418263,
+      "learning_rate": 4.1119047961463855e-06,
+      "loss": 0.7381,
+      "step": 30111
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4365617592198283,
+      "learning_rate": 4.1112880686951306e-06,
+      "loss": 0.6643,
+      "step": 30112
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4949096915619935,
+      "learning_rate": 4.110671375529817e-06,
+      "loss": 0.7197,
+      "step": 30113
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6223979716683729,
+      "learning_rate": 4.110054716654042e-06,
+      "loss": 0.5314,
+      "step": 30114
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4538692014566927,
+      "learning_rate": 4.109438092071389e-06,
+      "loss": 0.7068,
+      "step": 30115
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.388347857581723,
+      "learning_rate": 4.108821501785458e-06,
+      "loss": 0.8222,
+      "step": 30116
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2783915428243193,
+      "learning_rate": 4.1082049457998316e-06,
+      "loss": 0.745,
+      "step": 30117
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3990318728914866,
+      "learning_rate": 4.107588424118098e-06,
+      "loss": 0.6392,
+      "step": 30118
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4222446179801385,
+      "learning_rate": 4.106971936743851e-06,
+      "loss": 0.772,
+      "step": 30119
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5503146546829527,
+      "learning_rate": 4.106355483680683e-06,
+      "loss": 0.6144,
+      "step": 30120
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.324514490355172,
+      "learning_rate": 4.105739064932178e-06,
+      "loss": 0.616,
+      "step": 30121
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5709107704988987,
+      "learning_rate": 4.105122680501923e-06,
+      "loss": 0.6857,
+      "step": 30122
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3757335434467344,
+      "learning_rate": 4.1045063303935105e-06,
+      "loss": 0.735,
+      "step": 30123
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5478458513773512,
+      "learning_rate": 4.103890014610532e-06,
+      "loss": 0.6147,
+      "step": 30124
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5672041547380633,
+      "learning_rate": 4.103273733156572e-06,
+      "loss": 0.6011,
+      "step": 30125
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5736653404478429,
+      "learning_rate": 4.102657486035216e-06,
+      "loss": 0.4712,
+      "step": 30126
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.407001616146228,
+      "learning_rate": 4.102041273250055e-06,
+      "loss": 0.7121,
+      "step": 30127
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3760789810424745,
+      "learning_rate": 4.1014250948046805e-06,
+      "loss": 0.6581,
+      "step": 30128
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3326707611657103,
+      "learning_rate": 4.100808950702677e-06,
+      "loss": 0.5797,
+      "step": 30129
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6270407261478936,
+      "learning_rate": 4.100192840947628e-06,
+      "loss": 0.6418,
+      "step": 30130
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4167856291850032,
+      "learning_rate": 4.099576765543128e-06,
+      "loss": 0.8095,
+      "step": 30131
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9037700450837831,
+      "learning_rate": 4.098960724492756e-06,
+      "loss": 0.6351,
+      "step": 30132
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3645878765886286,
+      "learning_rate": 4.098344717800105e-06,
+      "loss": 0.6965,
+      "step": 30133
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4590585558023232,
+      "learning_rate": 4.0977287454687575e-06,
+      "loss": 0.7621,
+      "step": 30134
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3339331686202809,
+      "learning_rate": 4.097112807502306e-06,
+      "loss": 0.769,
+      "step": 30135
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4860382514665869,
+      "learning_rate": 4.096496903904328e-06,
+      "loss": 0.6201,
+      "step": 30136
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2688902936779565,
+      "learning_rate": 4.095881034678414e-06,
+      "loss": 0.7682,
+      "step": 30137
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6008007019806421,
+      "learning_rate": 4.095265199828157e-06,
+      "loss": 0.4994,
+      "step": 30138
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3759621915610811,
+      "learning_rate": 4.094649399357129e-06,
+      "loss": 0.7545,
+      "step": 30139
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.413215967123639,
+      "learning_rate": 4.0940336332689216e-06,
+      "loss": 0.7986,
+      "step": 30140
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2771174056984178,
+      "learning_rate": 4.093417901567119e-06,
+      "loss": 0.6711,
+      "step": 30141
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5598955421802059,
+      "learning_rate": 4.092802204255316e-06,
+      "loss": 0.8177,
+      "step": 30142
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3784558730299215,
+      "learning_rate": 4.092186541337081e-06,
+      "loss": 0.8583,
+      "step": 30143
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4785005707152317,
+      "learning_rate": 4.091570912816006e-06,
+      "loss": 0.7503,
+      "step": 30144
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4153779888977724,
+      "learning_rate": 4.090955318695679e-06,
+      "loss": 0.6177,
+      "step": 30145
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.384525494907287,
+      "learning_rate": 4.090339758979679e-06,
+      "loss": 0.7532,
+      "step": 30146
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2287847246359094,
+      "learning_rate": 4.089724233671594e-06,
+      "loss": 0.7525,
+      "step": 30147
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5974373140082773,
+      "learning_rate": 4.089108742775003e-06,
+      "loss": 0.5034,
+      "step": 30148
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4804954650246935,
+      "learning_rate": 4.088493286293495e-06,
+      "loss": 0.7198,
+      "step": 30149
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1961437450917434,
+      "learning_rate": 4.0878778642306475e-06,
+      "loss": 0.5594,
+      "step": 30150
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6617219886816115,
+      "learning_rate": 4.08726247659005e-06,
+      "loss": 0.554,
+      "step": 30151
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4440959237878925,
+      "learning_rate": 4.086647123375282e-06,
+      "loss": 0.7349,
+      "step": 30152
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6885318910011782,
+      "learning_rate": 4.086031804589925e-06,
+      "loss": 0.6263,
+      "step": 30153
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3684555320632683,
+      "learning_rate": 4.085416520237562e-06,
+      "loss": 0.6764,
+      "step": 30154
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3021195339433214,
+      "learning_rate": 4.0848012703217785e-06,
+      "loss": 0.6497,
+      "step": 30155
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6153425643657162,
+      "learning_rate": 4.0841860548461565e-06,
+      "loss": 0.6226,
+      "step": 30156
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5804067124167899,
+      "learning_rate": 4.083570873814273e-06,
+      "loss": 0.5084,
+      "step": 30157
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7930101016738949,
+      "learning_rate": 4.082955727229712e-06,
+      "loss": 0.7326,
+      "step": 30158
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3468349606749088,
+      "learning_rate": 4.08234061509606e-06,
+      "loss": 0.6689,
+      "step": 30159
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3765914725496846,
+      "learning_rate": 4.081725537416894e-06,
+      "loss": 0.7084,
+      "step": 30160
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.333045367198337,
+      "learning_rate": 4.0811104941957924e-06,
+      "loss": 0.6415,
+      "step": 30161
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5906225446464615,
+      "learning_rate": 4.0804954854363395e-06,
+      "loss": 0.638,
+      "step": 30162
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5563977217257727,
+      "learning_rate": 4.079880511142119e-06,
+      "loss": 0.4956,
+      "step": 30163
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5843487513766189,
+      "learning_rate": 4.07926557131671e-06,
+      "loss": 0.5368,
+      "step": 30164
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5906662437723793,
+      "learning_rate": 4.078650665963686e-06,
+      "loss": 0.7305,
+      "step": 30165
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0544802775298665,
+      "learning_rate": 4.078035795086636e-06,
+      "loss": 0.651,
+      "step": 30166
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.327112718376738,
+      "learning_rate": 4.077420958689134e-06,
+      "loss": 0.6971,
+      "step": 30167
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6042271619496433,
+      "learning_rate": 4.076806156774765e-06,
+      "loss": 0.5255,
+      "step": 30168
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.398334834431645,
+      "learning_rate": 4.076191389347103e-06,
+      "loss": 0.6059,
+      "step": 30169
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.640647226984701,
+      "learning_rate": 4.075576656409733e-06,
+      "loss": 0.7169,
+      "step": 30170
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4287901859107663,
+      "learning_rate": 4.074961957966228e-06,
+      "loss": 0.7014,
+      "step": 30171
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4074181434812556,
+      "learning_rate": 4.0743472940201746e-06,
+      "loss": 0.7625,
+      "step": 30172
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3389310204117209,
+      "learning_rate": 4.073732664575146e-06,
+      "loss": 0.7374,
+      "step": 30173
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5741594260844052,
+      "learning_rate": 4.073118069634719e-06,
+      "loss": 0.4705,
+      "step": 30174
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3701550427610436,
+      "learning_rate": 4.072503509202476e-06,
+      "loss": 0.6694,
+      "step": 30175
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.268773406862081,
+      "learning_rate": 4.071888983281996e-06,
+      "loss": 0.6171,
+      "step": 30176
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1870488642198531,
+      "learning_rate": 4.071274491876856e-06,
+      "loss": 0.6026,
+      "step": 30177
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4388679088915906,
+      "learning_rate": 4.07066003499063e-06,
+      "loss": 0.7466,
+      "step": 30178
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5144471349335138,
+      "learning_rate": 4.070045612626898e-06,
+      "loss": 0.5884,
+      "step": 30179
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.34466952772906,
+      "learning_rate": 4.069431224789242e-06,
+      "loss": 0.7379,
+      "step": 30180
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.222093800187586,
+      "learning_rate": 4.068816871481234e-06,
+      "loss": 0.577,
+      "step": 30181
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4142640217255376,
+      "learning_rate": 4.0682025527064486e-06,
+      "loss": 0.6009,
+      "step": 30182
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4331777859821238,
+      "learning_rate": 4.067588268468467e-06,
+      "loss": 0.5887,
+      "step": 30183
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5418104784899045,
+      "learning_rate": 4.066974018770866e-06,
+      "loss": 0.602,
+      "step": 30184
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8941407478225636,
+      "learning_rate": 4.0663598036172225e-06,
+      "loss": 0.7514,
+      "step": 30185
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.485090315116043,
+      "learning_rate": 4.065745623011106e-06,
+      "loss": 0.5514,
+      "step": 30186
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2479289809586882,
+      "learning_rate": 4.065131476956101e-06,
+      "loss": 0.6878,
+      "step": 30187
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3132543111022212,
+      "learning_rate": 4.0645173654557766e-06,
+      "loss": 0.7248,
+      "step": 30188
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3279760482082579,
+      "learning_rate": 4.063903288513714e-06,
+      "loss": 0.6641,
+      "step": 30189
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.339255099002969,
+      "learning_rate": 4.063289246133484e-06,
+      "loss": 0.6929,
+      "step": 30190
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4277186140964853,
+      "learning_rate": 4.062675238318665e-06,
+      "loss": 0.6592,
+      "step": 30191
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5095217611803717,
+      "learning_rate": 4.062061265072829e-06,
+      "loss": 0.6215,
+      "step": 30192
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.541905795829097,
+      "learning_rate": 4.061447326399554e-06,
+      "loss": 0.8777,
+      "step": 30193
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3583023604134865,
+      "learning_rate": 4.060833422302412e-06,
+      "loss": 0.6765,
+      "step": 30194
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.473333543431154,
+      "learning_rate": 4.060219552784976e-06,
+      "loss": 0.7832,
+      "step": 30195
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6103619783347207,
+      "learning_rate": 4.059605717850823e-06,
+      "loss": 0.5146,
+      "step": 30196
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3195400817545822,
+      "learning_rate": 4.058991917503529e-06,
+      "loss": 0.598,
+      "step": 30197
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4600213138828726,
+      "learning_rate": 4.058378151746665e-06,
+      "loss": 0.6813,
+      "step": 30198
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4201311193767365,
+      "learning_rate": 4.057764420583801e-06,
+      "loss": 0.6738,
+      "step": 30199
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4380869042367423,
+      "learning_rate": 4.057150724018514e-06,
+      "loss": 0.7437,
+      "step": 30200
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4349918107180548,
+      "learning_rate": 4.0565370620543795e-06,
+      "loss": 0.8331,
+      "step": 30201
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5589045407662447,
+      "learning_rate": 4.055923434694968e-06,
+      "loss": 0.7514,
+      "step": 30202
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0010143961978715,
+      "learning_rate": 4.05530984194385e-06,
+      "loss": 0.5901,
+      "step": 30203
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5943177532090074,
+      "learning_rate": 4.054696283804599e-06,
+      "loss": 0.8404,
+      "step": 30204
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.316089606625441,
+      "learning_rate": 4.054082760280792e-06,
+      "loss": 0.7228,
+      "step": 30205
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8405329261138923,
+      "learning_rate": 4.053469271375998e-06,
+      "loss": 0.6922,
+      "step": 30206
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.362216796943146,
+      "learning_rate": 4.052855817093785e-06,
+      "loss": 0.7027,
+      "step": 30207
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4971219376013987,
+      "learning_rate": 4.0522423974377325e-06,
+      "loss": 0.7929,
+      "step": 30208
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4600472421756914,
+      "learning_rate": 4.051629012411403e-06,
+      "loss": 0.7669,
+      "step": 30209
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3020158380719942,
+      "learning_rate": 4.051015662018376e-06,
+      "loss": 0.609,
+      "step": 30210
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3790890673615774,
+      "learning_rate": 4.050402346262217e-06,
+      "loss": 0.7767,
+      "step": 30211
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.266718776676019,
+      "learning_rate": 4.049789065146502e-06,
+      "loss": 0.6715,
+      "step": 30212
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3862536700679715,
+      "learning_rate": 4.049175818674795e-06,
+      "loss": 0.7228,
+      "step": 30213
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.232842469800609,
+      "learning_rate": 4.048562606850674e-06,
+      "loss": 0.6425,
+      "step": 30214
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.457499642162548,
+      "learning_rate": 4.047949429677706e-06,
+      "loss": 0.7194,
+      "step": 30215
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6041231193833452,
+      "learning_rate": 4.0473362871594576e-06,
+      "loss": 0.6327,
+      "step": 30216
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2256167690276412,
+      "learning_rate": 4.0467231792995024e-06,
+      "loss": 0.6918,
+      "step": 30217
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.421851864551378,
+      "learning_rate": 4.046110106101412e-06,
+      "loss": 0.6423,
+      "step": 30218
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3775400915894842,
+      "learning_rate": 4.045497067568753e-06,
+      "loss": 0.6617,
+      "step": 30219
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5247522404248888,
+      "learning_rate": 4.0448840637050934e-06,
+      "loss": 0.622,
+      "step": 30220
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.417066157049028,
+      "learning_rate": 4.044271094514003e-06,
+      "loss": 0.8313,
+      "step": 30221
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4759425561945145,
+      "learning_rate": 4.043658159999055e-06,
+      "loss": 0.7369,
+      "step": 30222
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3216797897615544,
+      "learning_rate": 4.043045260163816e-06,
+      "loss": 0.6899,
+      "step": 30223
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.232255040716951,
+      "learning_rate": 4.0424323950118495e-06,
+      "loss": 0.7556,
+      "step": 30224
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5831600726258108,
+      "learning_rate": 4.041819564546727e-06,
+      "loss": 0.483,
+      "step": 30225
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.489565390450296,
+      "learning_rate": 4.041206768772023e-06,
+      "loss": 0.7094,
+      "step": 30226
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4303802005705055,
+      "learning_rate": 4.040594007691297e-06,
+      "loss": 0.6979,
+      "step": 30227
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.502269584893676,
+      "learning_rate": 4.039981281308117e-06,
+      "loss": 0.6812,
+      "step": 30228
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2629886624892328,
+      "learning_rate": 4.039368589626057e-06,
+      "loss": 0.6655,
+      "step": 30229
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4901344992177017,
+      "learning_rate": 4.038755932648677e-06,
+      "loss": 0.7011,
+      "step": 30230
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4431535062408358,
+      "learning_rate": 4.03814331037955e-06,
+      "loss": 0.6745,
+      "step": 30231
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3309849587891849,
+      "learning_rate": 4.037530722822236e-06,
+      "loss": 0.6861,
+      "step": 30232
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5703192819875458,
+      "learning_rate": 4.036918169980311e-06,
+      "loss": 0.7271,
+      "step": 30233
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2087071596771057,
+      "learning_rate": 4.03630565185733e-06,
+      "loss": 0.6929,
+      "step": 30234
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.393622175856312,
+      "learning_rate": 4.035693168456869e-06,
+      "loss": 0.7553,
+      "step": 30235
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3215229930730812,
+      "learning_rate": 4.035080719782491e-06,
+      "loss": 0.555,
+      "step": 30236
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4737477825901852,
+      "learning_rate": 4.0344683058377645e-06,
+      "loss": 0.6182,
+      "step": 30237
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3765767327671528,
+      "learning_rate": 4.033855926626247e-06,
+      "loss": 0.6783,
+      "step": 30238
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3090112706761934,
+      "learning_rate": 4.033243582151509e-06,
+      "loss": 0.635,
+      "step": 30239
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3971472899126067,
+      "learning_rate": 4.032631272417123e-06,
+      "loss": 0.6007,
+      "step": 30240
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4587659816983563,
+      "learning_rate": 4.03201899742664e-06,
+      "loss": 0.6927,
+      "step": 30241
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.458479004726844,
+      "learning_rate": 4.031406757183632e-06,
+      "loss": 0.8637,
+      "step": 30242
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2832119024400117,
+      "learning_rate": 4.030794551691663e-06,
+      "loss": 0.6487,
+      "step": 30243
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.433481566441253,
+      "learning_rate": 4.030182380954303e-06,
+      "loss": 0.733,
+      "step": 30244
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3607014662532977,
+      "learning_rate": 4.02957024497511e-06,
+      "loss": 0.6064,
+      "step": 30245
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.544616556687638,
+      "learning_rate": 4.028958143757647e-06,
+      "loss": 0.8009,
+      "step": 30246
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4733300169716106,
+      "learning_rate": 4.028346077305482e-06,
+      "loss": 0.6582,
+      "step": 30247
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4948508136612306,
+      "learning_rate": 4.027734045622172e-06,
+      "loss": 0.7334,
+      "step": 30248
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4700389781798964,
+      "learning_rate": 4.027122048711292e-06,
+      "loss": 0.6108,
+      "step": 30249
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.803111039805684,
+      "learning_rate": 4.026510086576393e-06,
+      "loss": 0.6221,
+      "step": 30250
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8105188354993353,
+      "learning_rate": 4.0258981592210464e-06,
+      "loss": 0.6879,
+      "step": 30251
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5876321834304055,
+      "learning_rate": 4.0252862666488105e-06,
+      "loss": 0.6946,
+      "step": 30252
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3315078853917002,
+      "learning_rate": 4.024674408863252e-06,
+      "loss": 0.6451,
+      "step": 30253
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2803596119931948,
+      "learning_rate": 4.02406258586793e-06,
+      "loss": 0.5482,
+      "step": 30254
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.339973108168437,
+      "learning_rate": 4.023450797666405e-06,
+      "loss": 0.7111,
+      "step": 30255
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3528211314781402,
+      "learning_rate": 4.022839044262241e-06,
+      "loss": 0.7732,
+      "step": 30256
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.465880935940651,
+      "learning_rate": 4.022227325659005e-06,
+      "loss": 0.6954,
+      "step": 30257
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3977621579071577,
+      "learning_rate": 4.0216156418602535e-06,
+      "loss": 0.7009,
+      "step": 30258
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4365768478402874,
+      "learning_rate": 4.021003992869545e-06,
+      "loss": 0.7956,
+      "step": 30259
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.867406512636975,
+      "learning_rate": 4.020392378690444e-06,
+      "loss": 0.7477,
+      "step": 30260
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5974029018601018,
+      "learning_rate": 4.019780799326515e-06,
+      "loss": 0.4992,
+      "step": 30261
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6273623419692025,
+      "learning_rate": 4.019169254781315e-06,
+      "loss": 0.5117,
+      "step": 30262
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5611130817325191,
+      "learning_rate": 4.018557745058403e-06,
+      "loss": 0.5123,
+      "step": 30263
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5179064058381955,
+      "learning_rate": 4.01794627016134e-06,
+      "loss": 0.6213,
+      "step": 30264
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4196359220080994,
+      "learning_rate": 4.017334830093693e-06,
+      "loss": 0.7348,
+      "step": 30265
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3476376725726988,
+      "learning_rate": 4.016723424859016e-06,
+      "loss": 0.7704,
+      "step": 30266
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5429573923025675,
+      "learning_rate": 4.016112054460866e-06,
+      "loss": 0.8359,
+      "step": 30267
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.469362933262852,
+      "learning_rate": 4.0155007189028086e-06,
+      "loss": 0.6304,
+      "step": 30268
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.165659162929355,
+      "learning_rate": 4.014889418188398e-06,
+      "loss": 0.7129,
+      "step": 30269
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5203993352973788,
+      "learning_rate": 4.014278152321199e-06,
+      "loss": 0.7297,
+      "step": 30270
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6650475796657864,
+      "learning_rate": 4.013666921304765e-06,
+      "loss": 0.7212,
+      "step": 30271
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1631780018380602,
+      "learning_rate": 4.013055725142661e-06,
+      "loss": 0.5184,
+      "step": 30272
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5100075148902592,
+      "learning_rate": 4.012444563838437e-06,
+      "loss": 0.6311,
+      "step": 30273
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.54013744447748,
+      "learning_rate": 4.011833437395661e-06,
+      "loss": 0.649,
+      "step": 30274
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3668479452964624,
+      "learning_rate": 4.011222345817886e-06,
+      "loss": 0.7893,
+      "step": 30275
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3510737872247969,
+      "learning_rate": 4.0106112891086675e-06,
+      "loss": 0.6992,
+      "step": 30276
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5874435287418722,
+      "learning_rate": 4.010000267271566e-06,
+      "loss": 0.8226,
+      "step": 30277
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0501872278092983,
+      "learning_rate": 4.009389280310144e-06,
+      "loss": 0.6924,
+      "step": 30278
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5799933415198267,
+      "learning_rate": 4.008778328227953e-06,
+      "loss": 0.5138,
+      "step": 30279
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4266429404846384,
+      "learning_rate": 4.008167411028548e-06,
+      "loss": 0.8687,
+      "step": 30280
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.331049874809808,
+      "learning_rate": 4.007556528715489e-06,
+      "loss": 0.6745,
+      "step": 30281
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.344668664983133,
+      "learning_rate": 4.006945681292338e-06,
+      "loss": 0.6663,
+      "step": 30282
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.347836753995971,
+      "learning_rate": 4.0063348687626455e-06,
+      "loss": 0.617,
+      "step": 30283
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4236251439202539,
+      "learning_rate": 4.005724091129965e-06,
+      "loss": 0.6994,
+      "step": 30284
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4544308888452422,
+      "learning_rate": 4.005113348397858e-06,
+      "loss": 0.7803,
+      "step": 30285
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3571004300635985,
+      "learning_rate": 4.004502640569882e-06,
+      "loss": 0.6328,
+      "step": 30286
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.597581942893921,
+      "learning_rate": 4.003891967649589e-06,
+      "loss": 0.6352,
+      "step": 30287
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9914618795024346,
+      "learning_rate": 4.003281329640533e-06,
+      "loss": 0.7669,
+      "step": 30288
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5719359336545444,
+      "learning_rate": 4.002670726546274e-06,
+      "loss": 0.6883,
+      "step": 30289
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2642693537813374,
+      "learning_rate": 4.002060158370361e-06,
+      "loss": 0.689,
+      "step": 30290
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.215885570156729,
+      "learning_rate": 4.001449625116357e-06,
+      "loss": 0.6831,
+      "step": 30291
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.301390990577009,
+      "learning_rate": 4.000839126787809e-06,
+      "loss": 0.6553,
+      "step": 30292
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7494495234997816,
+      "learning_rate": 4.000228663388278e-06,
+      "loss": 0.7082,
+      "step": 30293
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.498947402255873,
+      "learning_rate": 3.999618234921312e-06,
+      "loss": 0.8648,
+      "step": 30294
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5792384653295113,
+      "learning_rate": 3.999007841390472e-06,
+      "loss": 0.4967,
+      "step": 30295
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3465397697927104,
+      "learning_rate": 3.998397482799307e-06,
+      "loss": 0.6629,
+      "step": 30296
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3762464677290553,
+      "learning_rate": 3.9977871591513705e-06,
+      "loss": 0.6738,
+      "step": 30297
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3950805391503653,
+      "learning_rate": 3.9971768704502156e-06,
+      "loss": 0.677,
+      "step": 30298
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2906639347163376,
+      "learning_rate": 3.996566616699402e-06,
+      "loss": 0.6785,
+      "step": 30299
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3137419936850703,
+      "learning_rate": 3.995956397902478e-06,
+      "loss": 0.6799,
+      "step": 30300
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5089747900115136,
+      "learning_rate": 3.995346214062993e-06,
+      "loss": 0.7549,
+      "step": 30301
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3943467229525723,
+      "learning_rate": 3.994736065184503e-06,
+      "loss": 0.614,
+      "step": 30302
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6156732413548335,
+      "learning_rate": 3.9941259512705654e-06,
+      "loss": 0.4864,
+      "step": 30303
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.8247247905036925,
+      "learning_rate": 3.993515872324727e-06,
+      "loss": 0.6996,
+      "step": 30304
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.6650855453631244,
+      "learning_rate": 3.992905828350538e-06,
+      "loss": 0.5271,
+      "step": 30305
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4087038550021045,
+      "learning_rate": 3.992295819351553e-06,
+      "loss": 0.704,
+      "step": 30306
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.34800280490631,
+      "learning_rate": 3.991685845331326e-06,
+      "loss": 0.6932,
+      "step": 30307
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6110673216134155,
+      "learning_rate": 3.9910759062934065e-06,
+      "loss": 0.5194,
+      "step": 30308
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4300043835315972,
+      "learning_rate": 3.990466002241344e-06,
+      "loss": 0.6479,
+      "step": 30309
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3374644048902116,
+      "learning_rate": 3.989856133178692e-06,
+      "loss": 0.6645,
+      "step": 30310
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5026025397544673,
+      "learning_rate": 3.989246299108999e-06,
+      "loss": 0.564,
+      "step": 30311
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6091544646216829,
+      "learning_rate": 3.988636500035819e-06,
+      "loss": 0.7986,
+      "step": 30312
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.38561392555806,
+      "learning_rate": 3.988026735962698e-06,
+      "loss": 0.6314,
+      "step": 30313
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3203674414661655,
+      "learning_rate": 3.987417006893191e-06,
+      "loss": 0.7754,
+      "step": 30314
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2860708563638257,
+      "learning_rate": 3.986807312830843e-06,
+      "loss": 0.7377,
+      "step": 30315
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6077466043166835,
+      "learning_rate": 3.98619765377921e-06,
+      "loss": 0.6989,
+      "step": 30316
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3347059002684947,
+      "learning_rate": 3.985588029741838e-06,
+      "loss": 0.7494,
+      "step": 30317
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5323615613202208,
+      "learning_rate": 3.984978440722274e-06,
+      "loss": 0.7214,
+      "step": 30318
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4394445769346205,
+      "learning_rate": 3.9843688867240695e-06,
+      "loss": 0.6992,
+      "step": 30319
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.284857617795983,
+      "learning_rate": 3.983759367750772e-06,
+      "loss": 0.7812,
+      "step": 30320
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4459335293240716,
+      "learning_rate": 3.9831498838059415e-06,
+      "loss": 0.7331,
+      "step": 30321
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6141408468216374,
+      "learning_rate": 3.982540434893109e-06,
+      "loss": 0.6231,
+      "step": 30322
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.346486519651051,
+      "learning_rate": 3.9819310210158324e-06,
+      "loss": 0.5993,
+      "step": 30323
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3663415232890357,
+      "learning_rate": 3.981321642177662e-06,
+      "loss": 0.6698,
+      "step": 30324
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5258434583075051,
+      "learning_rate": 3.980712298382141e-06,
+      "loss": 0.7654,
+      "step": 30325
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6434339079288296,
+      "learning_rate": 3.980102989632816e-06,
+      "loss": 0.681,
+      "step": 30326
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3113892561535192,
+      "learning_rate": 3.979493715933238e-06,
+      "loss": 0.6502,
+      "step": 30327
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.435174157701661,
+      "learning_rate": 3.978884477286956e-06,
+      "loss": 0.6429,
+      "step": 30328
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.442151305441366,
+      "learning_rate": 3.9782752736975165e-06,
+      "loss": 0.7763,
+      "step": 30329
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7093524878196484,
+      "learning_rate": 3.97766610516846e-06,
+      "loss": 0.7223,
+      "step": 30330
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6765799925568514,
+      "learning_rate": 3.977056971703342e-06,
+      "loss": 0.6156,
+      "step": 30331
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6085457165574437,
+      "learning_rate": 3.9764478733057e-06,
+      "loss": 0.6651,
+      "step": 30332
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.606344816450882,
+      "learning_rate": 3.975838809979087e-06,
+      "loss": 0.5896,
+      "step": 30333
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.594192115807523,
+      "learning_rate": 3.975229781727052e-06,
+      "loss": 0.7709,
+      "step": 30334
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5255530392776873,
+      "learning_rate": 3.974620788553135e-06,
+      "loss": 0.8102,
+      "step": 30335
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.709551034969662,
+      "learning_rate": 3.97401183046088e-06,
+      "loss": 0.6193,
+      "step": 30336
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2099276891249346,
+      "learning_rate": 3.973402907453837e-06,
+      "loss": 0.7636,
+      "step": 30337
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4863336769253612,
+      "learning_rate": 3.972794019535556e-06,
+      "loss": 0.6478,
+      "step": 30338
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.150444444681566,
+      "learning_rate": 3.97218516670957e-06,
+      "loss": 0.7222,
+      "step": 30339
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2961165754641943,
+      "learning_rate": 3.971576348979431e-06,
+      "loss": 0.6536,
+      "step": 30340
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2554379733180168,
+      "learning_rate": 3.970967566348682e-06,
+      "loss": 0.6804,
+      "step": 30341
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.47228169853983,
+      "learning_rate": 3.970358818820873e-06,
+      "loss": 0.6103,
+      "step": 30342
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4919634744853894,
+      "learning_rate": 3.9697501063995435e-06,
+      "loss": 0.813,
+      "step": 30343
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5197549044327543,
+      "learning_rate": 3.969141429088235e-06,
+      "loss": 0.6493,
+      "step": 30344
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.341639383062948,
+      "learning_rate": 3.968532786890496e-06,
+      "loss": 0.7094,
+      "step": 30345
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3857556617281173,
+      "learning_rate": 3.967924179809868e-06,
+      "loss": 0.6783,
+      "step": 30346
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.387979081690266,
+      "learning_rate": 3.9673156078498964e-06,
+      "loss": 0.5817,
+      "step": 30347
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.153674094581514,
+      "learning_rate": 3.966707071014121e-06,
+      "loss": 0.6482,
+      "step": 30348
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3583603236514872,
+      "learning_rate": 3.966098569306091e-06,
+      "loss": 0.7204,
+      "step": 30349
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4428283111910907,
+      "learning_rate": 3.9654901027293405e-06,
+      "loss": 0.7115,
+      "step": 30350
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5565317038410664,
+      "learning_rate": 3.964881671287422e-06,
+      "loss": 0.5113,
+      "step": 30351
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6053811048781559,
+      "learning_rate": 3.964273274983872e-06,
+      "loss": 0.6412,
+      "step": 30352
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.763847351602225,
+      "learning_rate": 3.963664913822231e-06,
+      "loss": 0.6519,
+      "step": 30353
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.57761734110748,
+      "learning_rate": 3.963056587806044e-06,
+      "loss": 0.6869,
+      "step": 30354
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4915430574008446,
+      "learning_rate": 3.962448296938854e-06,
+      "loss": 0.7185,
+      "step": 30355
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3535521091521214,
+      "learning_rate": 3.961840041224204e-06,
+      "loss": 0.6516,
+      "step": 30356
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.004855121506305,
+      "learning_rate": 3.96123182066563e-06,
+      "loss": 0.7575,
+      "step": 30357
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6116486632075953,
+      "learning_rate": 3.960623635266675e-06,
+      "loss": 0.7267,
+      "step": 30358
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6322045759496765,
+      "learning_rate": 3.960015485030885e-06,
+      "loss": 0.4863,
+      "step": 30359
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2599167409546144,
+      "learning_rate": 3.959407369961796e-06,
+      "loss": 0.6367,
+      "step": 30360
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4625980640671752,
+      "learning_rate": 3.958799290062947e-06,
+      "loss": 0.5878,
+      "step": 30361
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6439848449798802,
+      "learning_rate": 3.958191245337881e-06,
+      "loss": 0.6,
+      "step": 30362
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3769421320418918,
+      "learning_rate": 3.957583235790142e-06,
+      "loss": 0.7228,
+      "step": 30363
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8772934654908209,
+      "learning_rate": 3.956975261423267e-06,
+      "loss": 0.8199,
+      "step": 30364
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3675886639782373,
+      "learning_rate": 3.956367322240792e-06,
+      "loss": 0.6206,
+      "step": 30365
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4213697338227884,
+      "learning_rate": 3.955759418246259e-06,
+      "loss": 0.7426,
+      "step": 30366
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.340686049375965,
+      "learning_rate": 3.955151549443211e-06,
+      "loss": 0.761,
+      "step": 30367
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.65100212191438,
+      "learning_rate": 3.954543715835185e-06,
+      "loss": 0.7217,
+      "step": 30368
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5636746983178196,
+      "learning_rate": 3.953935917425716e-06,
+      "loss": 0.6011,
+      "step": 30369
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3973368154782124,
+      "learning_rate": 3.953328154218351e-06,
+      "loss": 0.598,
+      "step": 30370
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5822424655878331,
+      "learning_rate": 3.952720426216618e-06,
+      "loss": 0.4875,
+      "step": 30371
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6189497245511772,
+      "learning_rate": 3.952112733424067e-06,
+      "loss": 0.4877,
+      "step": 30372
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6213302300108344,
+      "learning_rate": 3.951505075844225e-06,
+      "loss": 0.551,
+      "step": 30373
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3166157253705282,
+      "learning_rate": 3.95089745348064e-06,
+      "loss": 0.7632,
+      "step": 30374
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2247747660196238,
+      "learning_rate": 3.950289866336842e-06,
+      "loss": 0.6877,
+      "step": 30375
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5147241765177961,
+      "learning_rate": 3.949682314416374e-06,
+      "loss": 0.6888,
+      "step": 30376
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4889315894723558,
+      "learning_rate": 3.9490747977227715e-06,
+      "loss": 0.6832,
+      "step": 30377
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6348016546671638,
+      "learning_rate": 3.948467316259569e-06,
+      "loss": 0.7224,
+      "step": 30378
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5407406049873003,
+      "learning_rate": 3.947859870030304e-06,
+      "loss": 0.6895,
+      "step": 30379
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5383986540058847,
+      "learning_rate": 3.947252459038519e-06,
+      "loss": 0.6584,
+      "step": 30380
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4490090504877973,
+      "learning_rate": 3.946645083287747e-06,
+      "loss": 0.8541,
+      "step": 30381
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3579150455474105,
+      "learning_rate": 3.946037742781519e-06,
+      "loss": 0.8104,
+      "step": 30382
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7407267789470866,
+      "learning_rate": 3.945430437523376e-06,
+      "loss": 0.6678,
+      "step": 30383
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5530442530500301,
+      "learning_rate": 3.9448231675168584e-06,
+      "loss": 0.5796,
+      "step": 30384
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6344179113953019,
+      "learning_rate": 3.944215932765498e-06,
+      "loss": 0.4768,
+      "step": 30385
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2851569958731044,
+      "learning_rate": 3.943608733272824e-06,
+      "loss": 0.6715,
+      "step": 30386
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.249602981983969,
+      "learning_rate": 3.943001569042378e-06,
+      "loss": 0.5162,
+      "step": 30387
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3269155442841176,
+      "learning_rate": 3.942394440077699e-06,
+      "loss": 0.6568,
+      "step": 30388
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6109554644991246,
+      "learning_rate": 3.9417873463823165e-06,
+      "loss": 0.7402,
+      "step": 30389
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5922867779122512,
+      "learning_rate": 3.941180287959763e-06,
+      "loss": 0.7746,
+      "step": 30390
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4215115709408621,
+      "learning_rate": 3.94057326481358e-06,
+      "loss": 0.718,
+      "step": 30391
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3371928261627861,
+      "learning_rate": 3.939966276947293e-06,
+      "loss": 0.6596,
+      "step": 30392
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5144309206319184,
+      "learning_rate": 3.939359324364445e-06,
+      "loss": 0.5799,
+      "step": 30393
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1953978037254402,
+      "learning_rate": 3.938752407068562e-06,
+      "loss": 0.673,
+      "step": 30394
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4748106968263066,
+      "learning_rate": 3.9381455250631864e-06,
+      "loss": 0.7213,
+      "step": 30395
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5969500378449903,
+      "learning_rate": 3.937538678351842e-06,
+      "loss": 0.448,
+      "step": 30396
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2852940856302013,
+      "learning_rate": 3.936931866938071e-06,
+      "loss": 0.8069,
+      "step": 30397
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4482406388233982,
+      "learning_rate": 3.9363250908254015e-06,
+      "loss": 0.7381,
+      "step": 30398
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5260372512673694,
+      "learning_rate": 3.9357183500173635e-06,
+      "loss": 0.7934,
+      "step": 30399
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5436853634130274,
+      "learning_rate": 3.935111644517493e-06,
+      "loss": 0.7482,
+      "step": 30400
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5313949320224292,
+      "learning_rate": 3.934504974329326e-06,
+      "loss": 0.7425,
+      "step": 30401
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4429729031715284,
+      "learning_rate": 3.933898339456392e-06,
+      "loss": 0.7227,
+      "step": 30402
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2348824877369997,
+      "learning_rate": 3.9332917399022196e-06,
+      "loss": 0.7239,
+      "step": 30403
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5431153247421494,
+      "learning_rate": 3.9326851756703425e-06,
+      "loss": 0.795,
+      "step": 30404
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3221224641825953,
+      "learning_rate": 3.932078646764297e-06,
+      "loss": 0.6506,
+      "step": 30405
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.558215491920308,
+      "learning_rate": 3.93147215318761e-06,
+      "loss": 0.6645,
+      "step": 30406
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4840193858548478,
+      "learning_rate": 3.930865694943809e-06,
+      "loss": 0.6827,
+      "step": 30407
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7545520981850076,
+      "learning_rate": 3.930259272036432e-06,
+      "loss": 0.6527,
+      "step": 30408
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.3721872911587414,
+      "learning_rate": 3.9296528844690085e-06,
+      "loss": 0.8274,
+      "step": 30409
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3785974625425184,
+      "learning_rate": 3.9290465322450685e-06,
+      "loss": 0.5948,
+      "step": 30410
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4747202087309286,
+      "learning_rate": 3.9284402153681385e-06,
+      "loss": 0.7965,
+      "step": 30411
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6151482851104099,
+      "learning_rate": 3.927833933841755e-06,
+      "loss": 0.4936,
+      "step": 30412
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6581975749751214,
+      "learning_rate": 3.927227687669442e-06,
+      "loss": 0.7885,
+      "step": 30413
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4747866249291868,
+      "learning_rate": 3.9266214768547335e-06,
+      "loss": 0.6849,
+      "step": 30414
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.473794895346751,
+      "learning_rate": 3.926015301401156e-06,
+      "loss": 0.7648,
+      "step": 30415
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7061634071556748,
+      "learning_rate": 3.925409161312243e-06,
+      "loss": 0.6388,
+      "step": 30416
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9434314780301858,
+      "learning_rate": 3.924803056591517e-06,
+      "loss": 0.6357,
+      "step": 30417
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8308450203858664,
+      "learning_rate": 3.924196987242512e-06,
+      "loss": 0.5991,
+      "step": 30418
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4640965409102236,
+      "learning_rate": 3.923590953268762e-06,
+      "loss": 0.5688,
+      "step": 30419
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6368001094382298,
+      "learning_rate": 3.922984954673783e-06,
+      "loss": 0.6879,
+      "step": 30420
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5578160878976848,
+      "learning_rate": 3.92237899146111e-06,
+      "loss": 0.5369,
+      "step": 30421
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2544619035184914,
+      "learning_rate": 3.921773063634269e-06,
+      "loss": 0.7473,
+      "step": 30422
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5994964544633128,
+      "learning_rate": 3.921167171196798e-06,
+      "loss": 0.6012,
+      "step": 30423
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4455201566658684,
+      "learning_rate": 3.920561314152208e-06,
+      "loss": 0.6793,
+      "step": 30424
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5743025972359315,
+      "learning_rate": 3.919955492504036e-06,
+      "loss": 0.5213,
+      "step": 30425
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3806749892889707,
+      "learning_rate": 3.91934970625581e-06,
+      "loss": 0.7243,
+      "step": 30426
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2814202084633854,
+      "learning_rate": 3.918743955411057e-06,
+      "loss": 0.6508,
+      "step": 30427
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.639434623925554,
+      "learning_rate": 3.918138239973297e-06,
+      "loss": 0.5663,
+      "step": 30428
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4324081296458122,
+      "learning_rate": 3.917532559946062e-06,
+      "loss": 0.6901,
+      "step": 30429
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.212853879493628,
+      "learning_rate": 3.916926915332882e-06,
+      "loss": 0.687,
+      "step": 30430
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4347915447990354,
+      "learning_rate": 3.916321306137276e-06,
+      "loss": 0.6759,
+      "step": 30431
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.383343962226587,
+      "learning_rate": 3.915715732362777e-06,
+      "loss": 0.6627,
+      "step": 30432
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4314093747998906,
+      "learning_rate": 3.915110194012907e-06,
+      "loss": 0.6449,
+      "step": 30433
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5554754931619803,
+      "learning_rate": 3.914504691091188e-06,
+      "loss": 0.6805,
+      "step": 30434
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4416524231268133,
+      "learning_rate": 3.913899223601149e-06,
+      "loss": 0.8047,
+      "step": 30435
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.496712455873202,
+      "learning_rate": 3.91329379154632e-06,
+      "loss": 0.6022,
+      "step": 30436
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4829757966153243,
+      "learning_rate": 3.912688394930221e-06,
+      "loss": 0.7447,
+      "step": 30437
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3567841529659537,
+      "learning_rate": 3.912083033756374e-06,
+      "loss": 0.6571,
+      "step": 30438
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6420455716330337,
+      "learning_rate": 3.911477708028308e-06,
+      "loss": 0.4772,
+      "step": 30439
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.833319802990236,
+      "learning_rate": 3.910872417749548e-06,
+      "loss": 0.6889,
+      "step": 30440
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8165869751505046,
+      "learning_rate": 3.910267162923618e-06,
+      "loss": 0.6136,
+      "step": 30441
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7910885664699647,
+      "learning_rate": 3.909661943554036e-06,
+      "loss": 0.6257,
+      "step": 30442
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3852044248025555,
+      "learning_rate": 3.909056759644331e-06,
+      "loss": 0.6149,
+      "step": 30443
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4752248705453483,
+      "learning_rate": 3.9084516111980285e-06,
+      "loss": 0.6821,
+      "step": 30444
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5920637647483165,
+      "learning_rate": 3.9078464982186505e-06,
+      "loss": 0.7593,
+      "step": 30445
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2742356568161333,
+      "learning_rate": 3.907241420709716e-06,
+      "loss": 0.6994,
+      "step": 30446
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3313131851975246,
+      "learning_rate": 3.906636378674752e-06,
+      "loss": 0.6448,
+      "step": 30447
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6042641512054194,
+      "learning_rate": 3.906031372117278e-06,
+      "loss": 0.5279,
+      "step": 30448
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1936656551353049,
+      "learning_rate": 3.905426401040823e-06,
+      "loss": 0.6182,
+      "step": 30449
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3317390457806433,
+      "learning_rate": 3.9048214654489e-06,
+      "loss": 0.5994,
+      "step": 30450
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.284511379583186,
+      "learning_rate": 3.90421656534504e-06,
+      "loss": 0.7645,
+      "step": 30451
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5877861653618058,
+      "learning_rate": 3.903611700732758e-06,
+      "loss": 0.4819,
+      "step": 30452
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4710836052232212,
+      "learning_rate": 3.903006871615582e-06,
+      "loss": 0.6399,
+      "step": 30453
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5012507891491986,
+      "learning_rate": 3.9024020779970305e-06,
+      "loss": 0.5416,
+      "step": 30454
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6323804638055014,
+      "learning_rate": 3.901797319880622e-06,
+      "loss": 0.5038,
+      "step": 30455
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3080523534364763,
+      "learning_rate": 3.901192597269878e-06,
+      "loss": 0.7027,
+      "step": 30456
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7202544326216864,
+      "learning_rate": 3.9005879101683266e-06,
+      "loss": 0.6833,
+      "step": 30457
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6272932458842408,
+      "learning_rate": 3.899983258579483e-06,
+      "loss": 0.6859,
+      "step": 30458
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5974253889830687,
+      "learning_rate": 3.899378642506863e-06,
+      "loss": 0.7545,
+      "step": 30459
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.459721134559444,
+      "learning_rate": 3.898774061953995e-06,
+      "loss": 0.7302,
+      "step": 30460
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3598087288291767,
+      "learning_rate": 3.8981695169243986e-06,
+      "loss": 0.79,
+      "step": 30461
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2520607255781278,
+      "learning_rate": 3.89756500742159e-06,
+      "loss": 0.7095,
+      "step": 30462
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3924743978530016,
+      "learning_rate": 3.896960533449087e-06,
+      "loss": 0.6248,
+      "step": 30463
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.202718830710422,
+      "learning_rate": 3.896356095010412e-06,
+      "loss": 0.6363,
+      "step": 30464
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.59188838295119,
+      "learning_rate": 3.895751692109087e-06,
+      "loss": 0.6783,
+      "step": 30465
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.566357332997391,
+      "learning_rate": 3.895147324748629e-06,
+      "loss": 0.7352,
+      "step": 30466
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.274734026209157,
+      "learning_rate": 3.894542992932553e-06,
+      "loss": 0.6762,
+      "step": 30467
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3458100304430378,
+      "learning_rate": 3.893938696664383e-06,
+      "loss": 0.7737,
+      "step": 30468
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4185082648821177,
+      "learning_rate": 3.893334435947632e-06,
+      "loss": 0.614,
+      "step": 30469
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.595926639512736,
+      "learning_rate": 3.892730210785825e-06,
+      "loss": 0.6691,
+      "step": 30470
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.224288927188668,
+      "learning_rate": 3.892126021182473e-06,
+      "loss": 0.4905,
+      "step": 30471
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4316481664364717,
+      "learning_rate": 3.891521867141099e-06,
+      "loss": 0.6178,
+      "step": 30472
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5414912985416613,
+      "learning_rate": 3.890917748665217e-06,
+      "loss": 0.6755,
+      "step": 30473
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4799160208225959,
+      "learning_rate": 3.890313665758348e-06,
+      "loss": 0.7688,
+      "step": 30474
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4194882393723638,
+      "learning_rate": 3.889709618424008e-06,
+      "loss": 0.625,
+      "step": 30475
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3477068548618214,
+      "learning_rate": 3.889105606665709e-06,
+      "loss": 0.5776,
+      "step": 30476
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6001471917640203,
+      "learning_rate": 3.888501630486973e-06,
+      "loss": 0.4809,
+      "step": 30477
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4965380935905022,
+      "learning_rate": 3.887897689891318e-06,
+      "loss": 0.5138,
+      "step": 30478
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3487489741661443,
+      "learning_rate": 3.887293784882258e-06,
+      "loss": 0.6848,
+      "step": 30479
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.474110480807659,
+      "learning_rate": 3.886689915463305e-06,
+      "loss": 0.8327,
+      "step": 30480
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8197583838371862,
+      "learning_rate": 3.886086081637979e-06,
+      "loss": 0.7027,
+      "step": 30481
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.817155048151333,
+      "learning_rate": 3.8854822834097985e-06,
+      "loss": 0.7558,
+      "step": 30482
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6231608067559249,
+      "learning_rate": 3.884878520782277e-06,
+      "loss": 0.4978,
+      "step": 30483
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4700439268398413,
+      "learning_rate": 3.884274793758924e-06,
+      "loss": 0.7023,
+      "step": 30484
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3481393814344758,
+      "learning_rate": 3.88367110234326e-06,
+      "loss": 0.7767,
+      "step": 30485
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.653605620677375,
+      "learning_rate": 3.883067446538803e-06,
+      "loss": 0.8269,
+      "step": 30486
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.584896516694395,
+      "learning_rate": 3.8824638263490635e-06,
+      "loss": 0.7544,
+      "step": 30487
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5399236318045957,
+      "learning_rate": 3.881860241777553e-06,
+      "loss": 0.7838,
+      "step": 30488
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3902851505302412,
+      "learning_rate": 3.88125669282779e-06,
+      "loss": 0.6998,
+      "step": 30489
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3442509010291028,
+      "learning_rate": 3.88065317950329e-06,
+      "loss": 0.6645,
+      "step": 30490
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7382673137252138,
+      "learning_rate": 3.8800497018075645e-06,
+      "loss": 0.607,
+      "step": 30491
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.408625755926876,
+      "learning_rate": 3.879446259744124e-06,
+      "loss": 0.7643,
+      "step": 30492
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.406605324665395,
+      "learning_rate": 3.878842853316489e-06,
+      "loss": 0.6541,
+      "step": 30493
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.439387074461661,
+      "learning_rate": 3.878239482528164e-06,
+      "loss": 0.6847,
+      "step": 30494
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.76844436045767,
+      "learning_rate": 3.877636147382672e-06,
+      "loss": 0.6801,
+      "step": 30495
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.470644181187305,
+      "learning_rate": 3.877032847883516e-06,
+      "loss": 0.7394,
+      "step": 30496
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.193821318446558,
+      "learning_rate": 3.876429584034218e-06,
+      "loss": 0.6168,
+      "step": 30497
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0344466321992516,
+      "learning_rate": 3.875826355838282e-06,
+      "loss": 0.7708,
+      "step": 30498
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3714191709121957,
+      "learning_rate": 3.8752231632992254e-06,
+      "loss": 0.6403,
+      "step": 30499
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3685023279895134,
+      "learning_rate": 3.87462000642056e-06,
+      "loss": 0.7023,
+      "step": 30500
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.306803186675632,
+      "learning_rate": 3.874016885205793e-06,
+      "loss": 0.8134,
+      "step": 30501
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4397906179058462,
+      "learning_rate": 3.873413799658438e-06,
+      "loss": 0.7879,
+      "step": 30502
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9252713322633035,
+      "learning_rate": 3.872810749782009e-06,
+      "loss": 0.7654,
+      "step": 30503
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4383739035638556,
+      "learning_rate": 3.8722077355800225e-06,
+      "loss": 0.7037,
+      "step": 30504
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.746717304176774,
+      "learning_rate": 3.871604757055974e-06,
+      "loss": 0.7265,
+      "step": 30505
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2799268616655983,
+      "learning_rate": 3.8710018142133844e-06,
+      "loss": 0.6054,
+      "step": 30506
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2954541832657729,
+      "learning_rate": 3.870398907055766e-06,
+      "loss": 0.7726,
+      "step": 30507
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9902870243103141,
+      "learning_rate": 3.869796035586625e-06,
+      "loss": 0.7017,
+      "step": 30508
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7225931691379697,
+      "learning_rate": 3.869193199809469e-06,
+      "loss": 0.7322,
+      "step": 30509
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.3275333927608575,
+      "learning_rate": 3.8685903997278115e-06,
+      "loss": 0.6718,
+      "step": 30510
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9241804692978075,
+      "learning_rate": 3.867987635345165e-06,
+      "loss": 0.7563,
+      "step": 30511
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8579203711656855,
+      "learning_rate": 3.867384906665036e-06,
+      "loss": 0.555,
+      "step": 30512
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4177211957604337,
+      "learning_rate": 3.8667822136909304e-06,
+      "loss": 0.5526,
+      "step": 30513
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.187962173687857,
+      "learning_rate": 3.866179556426364e-06,
+      "loss": 0.7702,
+      "step": 30514
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2296375139012716,
+      "learning_rate": 3.865576934874839e-06,
+      "loss": 0.6284,
+      "step": 30515
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3913173897988904,
+      "learning_rate": 3.864974349039866e-06,
+      "loss": 0.6262,
+      "step": 30516
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5586138145644937,
+      "learning_rate": 3.8643717989249585e-06,
+      "loss": 0.8027,
+      "step": 30517
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3065801180910954,
+      "learning_rate": 3.863769284533622e-06,
+      "loss": 0.7493,
+      "step": 30518
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4455158121250775,
+      "learning_rate": 3.8631668058693585e-06,
+      "loss": 0.6984,
+      "step": 30519
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5383661014034808,
+      "learning_rate": 3.862564362935681e-06,
+      "loss": 0.8536,
+      "step": 30520
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1024694549439387,
+      "learning_rate": 3.8619619557361044e-06,
+      "loss": 0.6483,
+      "step": 30521
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3779123512786278,
+      "learning_rate": 3.861359584274122e-06,
+      "loss": 0.6958,
+      "step": 30522
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7502904132050587,
+      "learning_rate": 3.860757248553247e-06,
+      "loss": 0.8611,
+      "step": 30523
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5642882663394593,
+      "learning_rate": 3.8601549485769865e-06,
+      "loss": 0.7167,
+      "step": 30524
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3160723289146101,
+      "learning_rate": 3.8595526843488515e-06,
+      "loss": 0.7187,
+      "step": 30525
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4079139962470264,
+      "learning_rate": 3.858950455872344e-06,
+      "loss": 0.7946,
+      "step": 30526
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6455976061525409,
+      "learning_rate": 3.8583482631509695e-06,
+      "loss": 0.5405,
+      "step": 30527
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2516974971226276,
+      "learning_rate": 3.857746106188238e-06,
+      "loss": 0.6345,
+      "step": 30528
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.0442081549773112,
+      "learning_rate": 3.85714398498765e-06,
+      "loss": 0.742,
+      "step": 30529
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2546532656261042,
+      "learning_rate": 3.8565418995527185e-06,
+      "loss": 0.6842,
+      "step": 30530
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.368958001757695,
+      "learning_rate": 3.8559398498869416e-06,
+      "loss": 0.7488,
+      "step": 30531
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3843256974100888,
+      "learning_rate": 3.855337835993831e-06,
+      "loss": 0.7299,
+      "step": 30532
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6459763849610272,
+      "learning_rate": 3.8547358578768854e-06,
+      "loss": 0.7438,
+      "step": 30533
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3782092823383527,
+      "learning_rate": 3.854133915539617e-06,
+      "loss": 0.5665,
+      "step": 30534
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3437606133318054,
+      "learning_rate": 3.853532008985527e-06,
+      "loss": 0.6863,
+      "step": 30535
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.562320378920311,
+      "learning_rate": 3.852930138218115e-06,
+      "loss": 0.7093,
+      "step": 30536
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5122932730646825,
+      "learning_rate": 3.852328303240891e-06,
+      "loss": 0.6114,
+      "step": 30537
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5966389045790271,
+      "learning_rate": 3.8517265040573606e-06,
+      "loss": 0.5056,
+      "step": 30538
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4563213202411507,
+      "learning_rate": 3.851124740671026e-06,
+      "loss": 0.8333,
+      "step": 30539
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3090493900300966,
+      "learning_rate": 3.850523013085385e-06,
+      "loss": 0.711,
+      "step": 30540
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5434257816838497,
+      "learning_rate": 3.849921321303947e-06,
+      "loss": 0.712,
+      "step": 30541
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7211854849453117,
+      "learning_rate": 3.849319665330218e-06,
+      "loss": 0.63,
+      "step": 30542
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3437324840690048,
+      "learning_rate": 3.848718045167695e-06,
+      "loss": 0.6689,
+      "step": 30543
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9565684752215453,
+      "learning_rate": 3.848116460819882e-06,
+      "loss": 0.6746,
+      "step": 30544
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2980194354803671,
+      "learning_rate": 3.8475149122902815e-06,
+      "loss": 0.6262,
+      "step": 30545
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3135762864638982,
+      "learning_rate": 3.846913399582401e-06,
+      "loss": 0.535,
+      "step": 30546
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.358049027703328,
+      "learning_rate": 3.846311922699738e-06,
+      "loss": 0.6783,
+      "step": 30547
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3200962973097508,
+      "learning_rate": 3.845710481645793e-06,
+      "loss": 0.5806,
+      "step": 30548
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5028539672154697,
+      "learning_rate": 3.845109076424071e-06,
+      "loss": 0.7114,
+      "step": 30549
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6056888474429318,
+      "learning_rate": 3.844507707038071e-06,
+      "loss": 0.7561,
+      "step": 30550
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5704583287248826,
+      "learning_rate": 3.843906373491299e-06,
+      "loss": 0.7456,
+      "step": 30551
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5994075022078413,
+      "learning_rate": 3.843305075787249e-06,
+      "loss": 0.7241,
+      "step": 30552
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.764869953503038,
+      "learning_rate": 3.8427038139294304e-06,
+      "loss": 0.751,
+      "step": 30553
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4037373888159963,
+      "learning_rate": 3.842102587921336e-06,
+      "loss": 0.7248,
+      "step": 30554
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.32918875805996,
+      "learning_rate": 3.841501397766474e-06,
+      "loss": 0.5411,
+      "step": 30555
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5282258494345786,
+      "learning_rate": 3.84090024346834e-06,
+      "loss": 0.7393,
+      "step": 30556
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5683384575642496,
+      "learning_rate": 3.84029912503043e-06,
+      "loss": 0.8858,
+      "step": 30557
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4778692323285803,
+      "learning_rate": 3.83969804245625e-06,
+      "loss": 0.6428,
+      "step": 30558
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3755946456568753,
+      "learning_rate": 3.839096995749302e-06,
+      "loss": 0.8012,
+      "step": 30559
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5805337900474583,
+      "learning_rate": 3.838495984913081e-06,
+      "loss": 0.635,
+      "step": 30560
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2161785086114785,
+      "learning_rate": 3.8378950099510835e-06,
+      "loss": 0.578,
+      "step": 30561
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6237351261230889,
+      "learning_rate": 3.837294070866812e-06,
+      "loss": 0.5081,
+      "step": 30562
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.483473269557928,
+      "learning_rate": 3.83669316766377e-06,
+      "loss": 0.6622,
+      "step": 30563
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5063942076470258,
+      "learning_rate": 3.8360923003454505e-06,
+      "loss": 0.6227,
+      "step": 30564
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5582623900992407,
+      "learning_rate": 3.835491468915351e-06,
+      "loss": 0.7289,
+      "step": 30565
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2088637751271132,
+      "learning_rate": 3.834890673376971e-06,
+      "loss": 0.7412,
+      "step": 30566
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3101032304814928,
+      "learning_rate": 3.834289913733813e-06,
+      "loss": 0.6058,
+      "step": 30567
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6473055293517822,
+      "learning_rate": 3.83368918998937e-06,
+      "loss": 0.711,
+      "step": 30568
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2857534299380606,
+      "learning_rate": 3.833088502147139e-06,
+      "loss": 0.7469,
+      "step": 30569
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5693010652355629,
+      "learning_rate": 3.832487850210622e-06,
+      "loss": 0.5109,
+      "step": 30570
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4823957913185677,
+      "learning_rate": 3.83188723418331e-06,
+      "loss": 0.6943,
+      "step": 30571
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3316587747778583,
+      "learning_rate": 3.831286654068706e-06,
+      "loss": 0.599,
+      "step": 30572
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6142651052839945,
+      "learning_rate": 3.830686109870302e-06,
+      "loss": 0.477,
+      "step": 30573
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.347624243718146,
+      "learning_rate": 3.830085601591599e-06,
+      "loss": 0.666,
+      "step": 30574
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5823698689468157,
+      "learning_rate": 3.829485129236088e-06,
+      "loss": 0.6076,
+      "step": 30575
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.797090829167408,
+      "learning_rate": 3.8288846928072705e-06,
+      "loss": 0.5864,
+      "step": 30576
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7108177707300898,
+      "learning_rate": 3.8282842923086415e-06,
+      "loss": 0.7019,
+      "step": 30577
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.723956522138819,
+      "learning_rate": 3.827683927743692e-06,
+      "loss": 0.7233,
+      "step": 30578
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3244514732703747,
+      "learning_rate": 3.8270835991159195e-06,
+      "loss": 0.5272,
+      "step": 30579
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4831892779526377,
+      "learning_rate": 3.826483306428824e-06,
+      "loss": 0.7955,
+      "step": 30580
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3568737841897418,
+      "learning_rate": 3.8258830496858975e-06,
+      "loss": 0.7025,
+      "step": 30581
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2780772339985869,
+      "learning_rate": 3.825282828890632e-06,
+      "loss": 0.7782,
+      "step": 30582
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5047380221295579,
+      "learning_rate": 3.824682644046524e-06,
+      "loss": 0.6401,
+      "step": 30583
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4356791345537845,
+      "learning_rate": 3.824082495157073e-06,
+      "loss": 0.7151,
+      "step": 30584
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6412878432503146,
+      "learning_rate": 3.823482382225768e-06,
+      "loss": 0.7453,
+      "step": 30585
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3121190677805017,
+      "learning_rate": 3.822882305256101e-06,
+      "loss": 0.6374,
+      "step": 30586
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2876125874320874,
+      "learning_rate": 3.822282264251569e-06,
+      "loss": 0.7274,
+      "step": 30587
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3670028446020275,
+      "learning_rate": 3.8216822592156685e-06,
+      "loss": 0.5981,
+      "step": 30588
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.437412653309235,
+      "learning_rate": 3.8210822901518904e-06,
+      "loss": 0.649,
+      "step": 30589
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4933288193114636,
+      "learning_rate": 3.820482357063724e-06,
+      "loss": 0.7023,
+      "step": 30590
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4763895394538404,
+      "learning_rate": 3.819882459954668e-06,
+      "loss": 0.7178,
+      "step": 30591
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4419732338564735,
+      "learning_rate": 3.819282598828211e-06,
+      "loss": 0.6679,
+      "step": 30592
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4580720988940024,
+      "learning_rate": 3.8186827736878505e-06,
+      "loss": 0.6371,
+      "step": 30593
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3967169662057923,
+      "learning_rate": 3.8180829845370714e-06,
+      "loss": 0.801,
+      "step": 30594
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4876090095110477,
+      "learning_rate": 3.817483231379375e-06,
+      "loss": 0.7239,
+      "step": 30595
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.959713109670435,
+      "learning_rate": 3.816883514218245e-06,
+      "loss": 0.7558,
+      "step": 30596
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3960709010787322,
+      "learning_rate": 3.816283833057179e-06,
+      "loss": 0.7323,
+      "step": 30597
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5818749056697999,
+      "learning_rate": 3.815684187899668e-06,
+      "loss": 0.5168,
+      "step": 30598
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.008968216590698,
+      "learning_rate": 3.815084578749197e-06,
+      "loss": 0.6504,
+      "step": 30599
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7384549133542047,
+      "learning_rate": 3.814485005609262e-06,
+      "loss": 0.6488,
+      "step": 30600
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4114234355794881,
+      "learning_rate": 3.8138854684833537e-06,
+      "loss": 0.9922,
+      "step": 30601
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2835331614680379,
+      "learning_rate": 3.8132859673749688e-06,
+      "loss": 0.6566,
+      "step": 30602
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4772904913056797,
+      "learning_rate": 3.8126865022875858e-06,
+      "loss": 0.6645,
+      "step": 30603
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3412747560516727,
+      "learning_rate": 3.812087073224701e-06,
+      "loss": 0.5871,
+      "step": 30604
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4201016725467286,
+      "learning_rate": 3.8114876801898036e-06,
+      "loss": 0.6647,
+      "step": 30605
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6129836335324244,
+      "learning_rate": 3.810888323186391e-06,
+      "loss": 0.6778,
+      "step": 30606
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4591272390921335,
+      "learning_rate": 3.8102890022179393e-06,
+      "loss": 0.5667,
+      "step": 30607
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2641250964222808,
+      "learning_rate": 3.809689717287944e-06,
+      "loss": 0.6937,
+      "step": 30608
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.610580543743041,
+      "learning_rate": 3.8090904683998997e-06,
+      "loss": 0.5174,
+      "step": 30609
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4885807952053318,
+      "learning_rate": 3.8084912555572907e-06,
+      "loss": 0.6718,
+      "step": 30610
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5199070656755354,
+      "learning_rate": 3.807892078763602e-06,
+      "loss": 0.7093,
+      "step": 30611
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2948702171502193,
+      "learning_rate": 3.8072929380223256e-06,
+      "loss": 0.7659,
+      "step": 30612
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.434795593876439,
+      "learning_rate": 3.8066938333369542e-06,
+      "loss": 0.6807,
+      "step": 30613
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4332993205179785,
+      "learning_rate": 3.806094764710968e-06,
+      "loss": 0.7364,
+      "step": 30614
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.341565986998254,
+      "learning_rate": 3.8054957321478623e-06,
+      "loss": 0.683,
+      "step": 30615
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.463231582235523,
+      "learning_rate": 3.8048967356511234e-06,
+      "loss": 0.7424,
+      "step": 30616
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.636860392380084,
+      "learning_rate": 3.8042977752242315e-06,
+      "loss": 0.7408,
+      "step": 30617
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3968794938754634,
+      "learning_rate": 3.803698850870681e-06,
+      "loss": 0.7865,
+      "step": 30618
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6240836908060414,
+      "learning_rate": 3.8030999625939604e-06,
+      "loss": 0.4814,
+      "step": 30619
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5509400850631343,
+      "learning_rate": 3.802501110397553e-06,
+      "loss": 0.6888,
+      "step": 30620
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4581555183826942,
+      "learning_rate": 3.8019022942849436e-06,
+      "loss": 0.6228,
+      "step": 30621
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3764222852594123,
+      "learning_rate": 3.8013035142596213e-06,
+      "loss": 0.6551,
+      "step": 30622
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7864993768378228,
+      "learning_rate": 3.8007047703250756e-06,
+      "loss": 0.6704,
+      "step": 30623
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.371161453420614,
+      "learning_rate": 3.8001060624847896e-06,
+      "loss": 0.7592,
+      "step": 30624
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3252681028599929,
+      "learning_rate": 3.7995073907422454e-06,
+      "loss": 0.7786,
+      "step": 30625
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5426014385953755,
+      "learning_rate": 3.798908755100932e-06,
+      "loss": 0.6457,
+      "step": 30626
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2910933269421687,
+      "learning_rate": 3.798310155564339e-06,
+      "loss": 0.5969,
+      "step": 30627
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.367740293132861,
+      "learning_rate": 3.7977115921359477e-06,
+      "loss": 0.6477,
+      "step": 30628
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4734999058741116,
+      "learning_rate": 3.797113064819239e-06,
+      "loss": 0.6861,
+      "step": 30629
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2727328549927126,
+      "learning_rate": 3.7965145736177058e-06,
+      "loss": 0.7496,
+      "step": 30630
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7162708143841916,
+      "learning_rate": 3.795916118534826e-06,
+      "loss": 0.7497,
+      "step": 30631
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.435469752019257,
+      "learning_rate": 3.795317699574089e-06,
+      "loss": 0.657,
+      "step": 30632
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4287760135954117,
+      "learning_rate": 3.7947193167389728e-06,
+      "loss": 0.672,
+      "step": 30633
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4536339055821588,
+      "learning_rate": 3.7941209700329706e-06,
+      "loss": 0.6515,
+      "step": 30634
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3623395808940009,
+      "learning_rate": 3.7935226594595564e-06,
+      "loss": 0.6377,
+      "step": 30635
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3152735368955697,
+      "learning_rate": 3.7929243850222206e-06,
+      "loss": 0.6492,
+      "step": 30636
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6644979070675802,
+      "learning_rate": 3.7923261467244455e-06,
+      "loss": 0.5655,
+      "step": 30637
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3509263442863368,
+      "learning_rate": 3.79172794456971e-06,
+      "loss": 0.5984,
+      "step": 30638
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.287446104262168,
+      "learning_rate": 3.791129778561499e-06,
+      "loss": 0.7906,
+      "step": 30639
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6407338273938907,
+      "learning_rate": 3.7905316487032995e-06,
+      "loss": 0.6822,
+      "step": 30640
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3238675318238757,
+      "learning_rate": 3.7899335549985903e-06,
+      "loss": 0.6308,
+      "step": 30641
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.637535985772727,
+      "learning_rate": 3.7893354974508513e-06,
+      "loss": 0.8296,
+      "step": 30642
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5725598880478757,
+      "learning_rate": 3.7887374760635675e-06,
+      "loss": 0.7652,
+      "step": 30643
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5975392488979602,
+      "learning_rate": 3.7881394908402237e-06,
+      "loss": 0.4396,
+      "step": 30644
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6458527357902883,
+      "learning_rate": 3.7875415417842974e-06,
+      "loss": 0.6602,
+      "step": 30645
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3225354716690112,
+      "learning_rate": 3.786943628899269e-06,
+      "loss": 0.6671,
+      "step": 30646
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5671782626539051,
+      "learning_rate": 3.7863457521886215e-06,
+      "loss": 0.4775,
+      "step": 30647
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3550060672103255,
+      "learning_rate": 3.7857479116558393e-06,
+      "loss": 0.6056,
+      "step": 30648
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5274610275447005,
+      "learning_rate": 3.7851501073044005e-06,
+      "loss": 0.6462,
+      "step": 30649
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6570034518357464,
+      "learning_rate": 3.7845523391377815e-06,
+      "loss": 0.6753,
+      "step": 30650
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.6972730883727536,
+      "learning_rate": 3.7839546071594703e-06,
+      "loss": 0.6567,
+      "step": 30651
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.497977504515468,
+      "learning_rate": 3.7833569113729394e-06,
+      "loss": 0.7032,
+      "step": 30652
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.677021913906687,
+      "learning_rate": 3.782759251781677e-06,
+      "loss": 0.7693,
+      "step": 30653
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3401442023226748,
+      "learning_rate": 3.782161628389155e-06,
+      "loss": 0.5942,
+      "step": 30654
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.33670115778409,
+      "learning_rate": 3.78156404119886e-06,
+      "loss": 0.7469,
+      "step": 30655
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7841497920130676,
+      "learning_rate": 3.780966490214264e-06,
+      "loss": 0.7066,
+      "step": 30656
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2720383889791687,
+      "learning_rate": 3.7803689754388538e-06,
+      "loss": 0.7803,
+      "step": 30657
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5472862654019435,
+      "learning_rate": 3.7797714968761045e-06,
+      "loss": 0.7758,
+      "step": 30658
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3743281013117477,
+      "learning_rate": 3.7791740545294907e-06,
+      "loss": 0.6149,
+      "step": 30659
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5116736565558433,
+      "learning_rate": 3.7785766484024954e-06,
+      "loss": 0.6194,
+      "step": 30660
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3440878219051502,
+      "learning_rate": 3.7779792784986e-06,
+      "loss": 0.6008,
+      "step": 30661
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6077847424896134,
+      "learning_rate": 3.77738194482128e-06,
+      "loss": 0.5278,
+      "step": 30662
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3947862427283635,
+      "learning_rate": 3.7767846473740077e-06,
+      "loss": 0.6208,
+      "step": 30663
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.358684154947113,
+      "learning_rate": 3.776187386160266e-06,
+      "loss": 0.5995,
+      "step": 30664
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4519664042279379,
+      "learning_rate": 3.7755901611835356e-06,
+      "loss": 0.7174,
+      "step": 30665
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5924318757820535,
+      "learning_rate": 3.774992972447289e-06,
+      "loss": 0.5417,
+      "step": 30666
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.400509614092379,
+      "learning_rate": 3.7743958199550013e-06,
+      "loss": 0.5784,
+      "step": 30667
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2569310219757441,
+      "learning_rate": 3.773798703710152e-06,
+      "loss": 0.6762,
+      "step": 30668
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4822245709766309,
+      "learning_rate": 3.7732016237162216e-06,
+      "loss": 0.6614,
+      "step": 30669
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9704170589331549,
+      "learning_rate": 3.772604579976682e-06,
+      "loss": 0.6842,
+      "step": 30670
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.334438630135068,
+      "learning_rate": 3.772007572495007e-06,
+      "loss": 0.7497,
+      "step": 30671
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6024359035325658,
+      "learning_rate": 3.7714106012746786e-06,
+      "loss": 0.5184,
+      "step": 30672
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3525720679756732,
+      "learning_rate": 3.7708136663191673e-06,
+      "loss": 0.8944,
+      "step": 30673
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3771335908493114,
+      "learning_rate": 3.7702167676319534e-06,
+      "loss": 0.7911,
+      "step": 30674
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6810683820279035,
+      "learning_rate": 3.769619905216507e-06,
+      "loss": 0.6884,
+      "step": 30675
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.498056264674052,
+      "learning_rate": 3.7690230790763094e-06,
+      "loss": 0.6246,
+      "step": 30676
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4865225794826287,
+      "learning_rate": 3.768426289214828e-06,
+      "loss": 0.6628,
+      "step": 30677
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4078585545322813,
+      "learning_rate": 3.7678295356355456e-06,
+      "loss": 0.6471,
+      "step": 30678
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.377055519541849,
+      "learning_rate": 3.7672328183419327e-06,
+      "loss": 0.6133,
+      "step": 30679
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5766010001985682,
+      "learning_rate": 3.7666361373374607e-06,
+      "loss": 0.4858,
+      "step": 30680
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4906670697532056,
+      "learning_rate": 3.766039492625605e-06,
+      "loss": 0.6449,
+      "step": 30681
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8769734877837387,
+      "learning_rate": 3.7654428842098457e-06,
+      "loss": 0.6494,
+      "step": 30682
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4908961159396468,
+      "learning_rate": 3.764846312093652e-06,
+      "loss": 0.6663,
+      "step": 30683
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8405512691326202,
+      "learning_rate": 3.764249776280493e-06,
+      "loss": 0.6236,
+      "step": 30684
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6058877863644321,
+      "learning_rate": 3.7636532767738455e-06,
+      "loss": 0.5335,
+      "step": 30685
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4715913529011901,
+      "learning_rate": 3.7630568135771873e-06,
+      "loss": 0.6698,
+      "step": 30686
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5456874981037654,
+      "learning_rate": 3.762460386693987e-06,
+      "loss": 0.7265,
+      "step": 30687
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8735233544698624,
+      "learning_rate": 3.7618639961277124e-06,
+      "loss": 0.5902,
+      "step": 30688
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4068052807267357,
+      "learning_rate": 3.7612676418818415e-06,
+      "loss": 0.6468,
+      "step": 30689
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.336179479249708,
+      "learning_rate": 3.760671323959849e-06,
+      "loss": 0.5785,
+      "step": 30690
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6117115352457768,
+      "learning_rate": 3.7600750423652043e-06,
+      "loss": 0.5444,
+      "step": 30691
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.389780774422035,
+      "learning_rate": 3.7594787971013736e-06,
+      "loss": 0.5666,
+      "step": 30692
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5949629910497385,
+      "learning_rate": 3.7588825881718372e-06,
+      "loss": 0.5683,
+      "step": 30693
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8853326141536597,
+      "learning_rate": 3.758286415580058e-06,
+      "loss": 0.7574,
+      "step": 30694
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.759981683527087,
+      "learning_rate": 3.7576902793295158e-06,
+      "loss": 0.7104,
+      "step": 30695
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3886100641736117,
+      "learning_rate": 3.7570941794236725e-06,
+      "loss": 0.7411,
+      "step": 30696
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8701493861421394,
+      "learning_rate": 3.7564981158660073e-06,
+      "loss": 0.6952,
+      "step": 30697
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3679010768692181,
+      "learning_rate": 3.755902088659983e-06,
+      "loss": 0.6976,
+      "step": 30698
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.232015791051058,
+      "learning_rate": 3.755306097809074e-06,
+      "loss": 0.8104,
+      "step": 30699
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.669339545295768,
+      "learning_rate": 3.7547101433167567e-06,
+      "loss": 0.5783,
+      "step": 30700
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4961301268716214,
+      "learning_rate": 3.7541142251864883e-06,
+      "loss": 0.713,
+      "step": 30701
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7350153046655816,
+      "learning_rate": 3.7535183434217427e-06,
+      "loss": 0.6912,
+      "step": 30702
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6019328972283513,
+      "learning_rate": 3.7529224980259914e-06,
+      "loss": 0.6334,
+      "step": 30703
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3964888541724574,
+      "learning_rate": 3.75232668900271e-06,
+      "loss": 0.6813,
+      "step": 30704
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6008574005902049,
+      "learning_rate": 3.751730916355354e-06,
+      "loss": 0.516,
+      "step": 30705
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8681345229304822,
+      "learning_rate": 3.7511351800873987e-06,
+      "loss": 0.6722,
+      "step": 30706
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3936208565559398,
+      "learning_rate": 3.7505394802023163e-06,
+      "loss": 0.6495,
+      "step": 30707
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4032183339475466,
+      "learning_rate": 3.749943816703572e-06,
+      "loss": 0.6285,
+      "step": 30708
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8371747427968925,
+      "learning_rate": 3.7493481895946295e-06,
+      "loss": 0.6959,
+      "step": 30709
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.438609105793473,
+      "learning_rate": 3.7487525988789595e-06,
+      "loss": 0.6146,
+      "step": 30710
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3297293501425789,
+      "learning_rate": 3.7481570445600357e-06,
+      "loss": 0.7599,
+      "step": 30711
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6063810465644786,
+      "learning_rate": 3.7475615266413178e-06,
+      "loss": 0.7565,
+      "step": 30712
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4011555843601677,
+      "learning_rate": 3.746966045126279e-06,
+      "loss": 0.6346,
+      "step": 30713
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.045125956293571,
+      "learning_rate": 3.7463706000183843e-06,
+      "loss": 0.7629,
+      "step": 30714
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6072493968343871,
+      "learning_rate": 3.745775191321096e-06,
+      "loss": 0.5205,
+      "step": 30715
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3587370734136546,
+      "learning_rate": 3.745179819037884e-06,
+      "loss": 0.7897,
+      "step": 30716
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4621945168213084,
+      "learning_rate": 3.74458448317222e-06,
+      "loss": 0.7137,
+      "step": 30717
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5048078143149501,
+      "learning_rate": 3.7439891837275644e-06,
+      "loss": 0.7605,
+      "step": 30718
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3398141653727884,
+      "learning_rate": 3.743393920707381e-06,
+      "loss": 0.6652,
+      "step": 30719
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2248232018862781,
+      "learning_rate": 3.7427986941151394e-06,
+      "loss": 0.5839,
+      "step": 30720
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7575711654938617,
+      "learning_rate": 3.742203503954308e-06,
+      "loss": 0.7571,
+      "step": 30721
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.249210703375929,
+      "learning_rate": 3.7416083502283494e-06,
+      "loss": 0.6673,
+      "step": 30722
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.332298072538763,
+      "learning_rate": 3.7410132329407258e-06,
+      "loss": 0.6632,
+      "step": 30723
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1512261350790833,
+      "learning_rate": 3.7404181520949033e-06,
+      "loss": 0.7176,
+      "step": 30724
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3529791099018278,
+      "learning_rate": 3.7398231076943513e-06,
+      "loss": 0.7299,
+      "step": 30725
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5701164052288293,
+      "learning_rate": 3.7392280997425325e-06,
+      "loss": 0.4895,
+      "step": 30726
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5807811622444212,
+      "learning_rate": 3.7386331282429045e-06,
+      "loss": 0.4989,
+      "step": 30727
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.495504462914102,
+      "learning_rate": 3.7380381931989384e-06,
+      "loss": 0.6023,
+      "step": 30728
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3418979731516063,
+      "learning_rate": 3.7374432946140994e-06,
+      "loss": 0.6186,
+      "step": 30729
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3633169721625689,
+      "learning_rate": 3.7368484324918473e-06,
+      "loss": 0.5906,
+      "step": 30730
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.932545278245709,
+      "learning_rate": 3.7362536068356448e-06,
+      "loss": 0.5305,
+      "step": 30731
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4191155312173587,
+      "learning_rate": 3.7356588176489595e-06,
+      "loss": 0.738,
+      "step": 30732
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5850263243654459,
+      "learning_rate": 3.7350640649352477e-06,
+      "loss": 0.6505,
+      "step": 30733
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3322259949786361,
+      "learning_rate": 3.73446934869798e-06,
+      "loss": 0.7717,
+      "step": 30734
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5526525030099748,
+      "learning_rate": 3.7338746689406123e-06,
+      "loss": 0.4689,
+      "step": 30735
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.285586825686451,
+      "learning_rate": 3.7332800256666134e-06,
+      "loss": 0.6702,
+      "step": 30736
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5811213576937395,
+      "learning_rate": 3.7326854188794393e-06,
+      "loss": 0.4987,
+      "step": 30737
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.335455325379,
+      "learning_rate": 3.732090848582558e-06,
+      "loss": 0.6751,
+      "step": 30738
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6274872522433956,
+      "learning_rate": 3.731496314779428e-06,
+      "loss": 0.6959,
+      "step": 30739
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8018711838245094,
+      "learning_rate": 3.7309018174735078e-06,
+      "loss": 0.729,
+      "step": 30740
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5124840486122895,
+      "learning_rate": 3.7303073566682625e-06,
+      "loss": 0.6718,
+      "step": 30741
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8302489434468834,
+      "learning_rate": 3.7297129323671553e-06,
+      "loss": 0.7601,
+      "step": 30742
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5067726124439245,
+      "learning_rate": 3.7291185445736445e-06,
+      "loss": 0.7458,
+      "step": 30743
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4371321264600165,
+      "learning_rate": 3.7285241932911874e-06,
+      "loss": 0.5538,
+      "step": 30744
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4469056539659244,
+      "learning_rate": 3.7279298785232488e-06,
+      "loss": 0.7064,
+      "step": 30745
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4522592545502446,
+      "learning_rate": 3.7273356002732908e-06,
+      "loss": 0.752,
+      "step": 30746
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4379746674644516,
+      "learning_rate": 3.726741358544771e-06,
+      "loss": 0.7178,
+      "step": 30747
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2634093109575164,
+      "learning_rate": 3.7261471533411464e-06,
+      "loss": 0.6136,
+      "step": 30748
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8276350327164734,
+      "learning_rate": 3.7255529846658777e-06,
+      "loss": 0.7261,
+      "step": 30749
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5608495501762856,
+      "learning_rate": 3.724958852522431e-06,
+      "loss": 0.4713,
+      "step": 30750
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7254754660816671,
+      "learning_rate": 3.7243647569142606e-06,
+      "loss": 0.8456,
+      "step": 30751
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.69796822913897,
+      "learning_rate": 3.7237706978448206e-06,
+      "loss": 0.8371,
+      "step": 30752
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5684337444961813,
+      "learning_rate": 3.723176675317579e-06,
+      "loss": 0.6716,
+      "step": 30753
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8875727954629116,
+      "learning_rate": 3.722582689335986e-06,
+      "loss": 0.6684,
+      "step": 30754
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.436109600400129,
+      "learning_rate": 3.721988739903508e-06,
+      "loss": 0.7321,
+      "step": 30755
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1645585043993953,
+      "learning_rate": 3.7213948270235954e-06,
+      "loss": 0.6465,
+      "step": 30756
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3763848529862153,
+      "learning_rate": 3.7208009506997135e-06,
+      "loss": 0.5627,
+      "step": 30757
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6970644849584862,
+      "learning_rate": 3.720207110935312e-06,
+      "loss": 0.6474,
+      "step": 30758
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5581533878873717,
+      "learning_rate": 3.719613307733858e-06,
+      "loss": 0.725,
+      "step": 30759
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3596821380855708,
+      "learning_rate": 3.719019541098803e-06,
+      "loss": 0.7805,
+      "step": 30760
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7287859103426924,
+      "learning_rate": 3.7184258110336e-06,
+      "loss": 0.6717,
+      "step": 30761
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.691559235471377,
+      "learning_rate": 3.717832117541712e-06,
+      "loss": 0.6079,
+      "step": 30762
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.440923873627258,
+      "learning_rate": 3.7172384606265976e-06,
+      "loss": 0.7471,
+      "step": 30763
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6984759135843372,
+      "learning_rate": 3.71664484029171e-06,
+      "loss": 0.6586,
+      "step": 30764
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4035474162744026,
+      "learning_rate": 3.7160512565405016e-06,
+      "loss": 0.7139,
+      "step": 30765
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3227645637881817,
+      "learning_rate": 3.7154577093764334e-06,
+      "loss": 0.6915,
+      "step": 30766
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3677267152901054,
+      "learning_rate": 3.714864198802962e-06,
+      "loss": 0.7014,
+      "step": 30767
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7222400060051415,
+      "learning_rate": 3.714270724823542e-06,
+      "loss": 0.6853,
+      "step": 30768
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6331358444505181,
+      "learning_rate": 3.713677287441624e-06,
+      "loss": 0.5209,
+      "step": 30769
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2943815983891784,
+      "learning_rate": 3.7130838866606665e-06,
+      "loss": 0.7537,
+      "step": 30770
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3842979757913385,
+      "learning_rate": 3.7124905224841303e-06,
+      "loss": 0.6529,
+      "step": 30771
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6241280723590665,
+      "learning_rate": 3.7118971949154624e-06,
+      "loss": 0.8419,
+      "step": 30772
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2376477493442564,
+      "learning_rate": 3.7113039039581177e-06,
+      "loss": 0.6451,
+      "step": 30773
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5770705304586878,
+      "learning_rate": 3.7107106496155565e-06,
+      "loss": 0.5255,
+      "step": 30774
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4318867445994667,
+      "learning_rate": 3.7101174318912257e-06,
+      "loss": 0.7226,
+      "step": 30775
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.299657392796724,
+      "learning_rate": 3.7095242507885854e-06,
+      "loss": 0.6101,
+      "step": 30776
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4156377503152324,
+      "learning_rate": 3.7089311063110823e-06,
+      "loss": 0.727,
+      "step": 30777
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7919465338045746,
+      "learning_rate": 3.708337998462178e-06,
+      "loss": 0.6946,
+      "step": 30778
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3553946696225814,
+      "learning_rate": 3.707744927245319e-06,
+      "loss": 0.69,
+      "step": 30779
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4643572421013886,
+      "learning_rate": 3.707151892663964e-06,
+      "loss": 0.7103,
+      "step": 30780
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.571381549455665,
+      "learning_rate": 3.7065588947215636e-06,
+      "loss": 0.6,
+      "step": 30781
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.363930656433358,
+      "learning_rate": 3.705965933421566e-06,
+      "loss": 0.718,
+      "step": 30782
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.596190775043312,
+      "learning_rate": 3.7053730087674267e-06,
+      "loss": 0.524,
+      "step": 30783
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4647113124563336,
+      "learning_rate": 3.704780120762599e-06,
+      "loss": 0.7361,
+      "step": 30784
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6433659777815812,
+      "learning_rate": 3.704187269410542e-06,
+      "loss": 0.5863,
+      "step": 30785
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3507646786337129,
+      "learning_rate": 3.7035944547146917e-06,
+      "loss": 0.7751,
+      "step": 30786
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7075140668912758,
+      "learning_rate": 3.7030016766785095e-06,
+      "loss": 0.661,
+      "step": 30787
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5149711737581888,
+      "learning_rate": 3.702408935305447e-06,
+      "loss": 0.6426,
+      "step": 30788
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.54428152823829,
+      "learning_rate": 3.701816230598955e-06,
+      "loss": 0.7786,
+      "step": 30789
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3025673876491266,
+      "learning_rate": 3.701223562562478e-06,
+      "loss": 0.6839,
+      "step": 30790
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5120096012926456,
+      "learning_rate": 3.7006309311994727e-06,
+      "loss": 0.7076,
+      "step": 30791
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5044375482705394,
+      "learning_rate": 3.7000383365133916e-06,
+      "loss": 0.6657,
+      "step": 30792
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4510047773355041,
+      "learning_rate": 3.6994457785076823e-06,
+      "loss": 0.7135,
+      "step": 30793
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4823005294362153,
+      "learning_rate": 3.6988532571857895e-06,
+      "loss": 0.6196,
+      "step": 30794
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4408730940349423,
+      "learning_rate": 3.698260772551172e-06,
+      "loss": 0.6887,
+      "step": 30795
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2805606432624292,
+      "learning_rate": 3.697668324607272e-06,
+      "loss": 0.5689,
+      "step": 30796
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4895664616270547,
+      "learning_rate": 3.697075913357543e-06,
+      "loss": 0.5418,
+      "step": 30797
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4484418770857126,
+      "learning_rate": 3.6964835388054367e-06,
+      "loss": 0.7871,
+      "step": 30798
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.357529147479702,
+      "learning_rate": 3.695891200954398e-06,
+      "loss": 0.7189,
+      "step": 30799
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.20096780928398,
+      "learning_rate": 3.695298899807874e-06,
+      "loss": 0.7982,
+      "step": 30800
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.31207185855806,
+      "learning_rate": 3.694706635369316e-06,
+      "loss": 0.7549,
+      "step": 30801
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2776992377509733,
+      "learning_rate": 3.6941144076421786e-06,
+      "loss": 0.7078,
+      "step": 30802
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2407348138138683,
+      "learning_rate": 3.693522216629897e-06,
+      "loss": 0.6524,
+      "step": 30803
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4739449977352663,
+      "learning_rate": 3.6929300623359255e-06,
+      "loss": 0.6999,
+      "step": 30804
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.597566274774316,
+      "learning_rate": 3.692337944763711e-06,
+      "loss": 0.786,
+      "step": 30805
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4876459833014959,
+      "learning_rate": 3.6917458639167105e-06,
+      "loss": 0.7172,
+      "step": 30806
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4782212041227116,
+      "learning_rate": 3.691153819798354e-06,
+      "loss": 0.6277,
+      "step": 30807
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3907409080480848,
+      "learning_rate": 3.690561812412099e-06,
+      "loss": 0.7887,
+      "step": 30808
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6019401552226971,
+      "learning_rate": 3.689969841761393e-06,
+      "loss": 0.5093,
+      "step": 30809
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.424085104507751,
+      "learning_rate": 3.689377907849677e-06,
+      "loss": 0.6201,
+      "step": 30810
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2886059134519179,
+      "learning_rate": 3.688786010680405e-06,
+      "loss": 0.703,
+      "step": 30811
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.293246768270835,
+      "learning_rate": 3.688194150257015e-06,
+      "loss": 0.6724,
+      "step": 30812
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4543000213657757,
+      "learning_rate": 3.687602326582961e-06,
+      "loss": 0.7447,
+      "step": 30813
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2738091822208766,
+      "learning_rate": 3.6870105396616795e-06,
+      "loss": 0.6538,
+      "step": 30814
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2914224267049703,
+      "learning_rate": 3.6864187894966263e-06,
+      "loss": 0.5356,
+      "step": 30815
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4733377769453442,
+      "learning_rate": 3.685827076091242e-06,
+      "loss": 0.6811,
+      "step": 30816
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2561065708430015,
+      "learning_rate": 3.6852353994489665e-06,
+      "loss": 0.5809,
+      "step": 30817
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.182952535695797,
+      "learning_rate": 3.684643759573251e-06,
+      "loss": 0.5832,
+      "step": 30818
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2065989996383586,
+      "learning_rate": 3.6840521564675424e-06,
+      "loss": 0.6951,
+      "step": 30819
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.480465836800462,
+      "learning_rate": 3.6834605901352804e-06,
+      "loss": 0.6255,
+      "step": 30820
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3963819072873378,
+      "learning_rate": 3.6828690605799087e-06,
+      "loss": 0.8443,
+      "step": 30821
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1834604494182714,
+      "learning_rate": 3.6822775678048727e-06,
+      "loss": 0.6335,
+      "step": 30822
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4932759757138134,
+      "learning_rate": 3.6816861118136204e-06,
+      "loss": 0.6824,
+      "step": 30823
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3271475384958646,
+      "learning_rate": 3.6810946926095914e-06,
+      "loss": 0.6468,
+      "step": 30824
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3371697635792548,
+      "learning_rate": 3.680503310196227e-06,
+      "loss": 0.7314,
+      "step": 30825
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9736997304749613,
+      "learning_rate": 3.679911964576972e-06,
+      "loss": 0.6464,
+      "step": 30826
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.386644414551413,
+      "learning_rate": 3.6793206557552753e-06,
+      "loss": 0.6708,
+      "step": 30827
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6812090851072308,
+      "learning_rate": 3.678729383734574e-06,
+      "loss": 0.6587,
+      "step": 30828
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5657937683808149,
+      "learning_rate": 3.678138148518308e-06,
+      "loss": 0.4909,
+      "step": 30829
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4031915505325594,
+      "learning_rate": 3.677546950109927e-06,
+      "loss": 0.8415,
+      "step": 30830
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0481274549246837,
+      "learning_rate": 3.676955788512866e-06,
+      "loss": 0.5493,
+      "step": 30831
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.358995709921254,
+      "learning_rate": 3.6763646637305726e-06,
+      "loss": 0.6993,
+      "step": 30832
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6302268775662572,
+      "learning_rate": 3.6757735757664835e-06,
+      "loss": 0.6948,
+      "step": 30833
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5827562751338559,
+      "learning_rate": 3.675182524624047e-06,
+      "loss": 0.5027,
+      "step": 30834
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6229997546925141,
+      "learning_rate": 3.6745915103066952e-06,
+      "loss": 0.6852,
+      "step": 30835
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5720076083827688,
+      "learning_rate": 3.6740005328178786e-06,
+      "loss": 0.4865,
+      "step": 30836
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4538453540413256,
+      "learning_rate": 3.6734095921610336e-06,
+      "loss": 0.769,
+      "step": 30837
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5167341572916402,
+      "learning_rate": 3.6728186883395978e-06,
+      "loss": 0.6564,
+      "step": 30838
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5375450397328516,
+      "learning_rate": 3.672227821357014e-06,
+      "loss": 0.6472,
+      "step": 30839
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9974076145348243,
+      "learning_rate": 3.671636991216727e-06,
+      "loss": 0.742,
+      "step": 30840
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3782591278833543,
+      "learning_rate": 3.6710461979221735e-06,
+      "loss": 0.6905,
+      "step": 30841
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4151601005664123,
+      "learning_rate": 3.670455441476788e-06,
+      "loss": 0.5846,
+      "step": 30842
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3576648936295328,
+      "learning_rate": 3.6698647218840157e-06,
+      "loss": 0.6543,
+      "step": 30843
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2947524704572628,
+      "learning_rate": 3.669274039147299e-06,
+      "loss": 0.5641,
+      "step": 30844
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2830613214231794,
+      "learning_rate": 3.6686833932700717e-06,
+      "loss": 0.7185,
+      "step": 30845
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4808289024478145,
+      "learning_rate": 3.6680927842557724e-06,
+      "loss": 0.599,
+      "step": 30846
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5934652853731257,
+      "learning_rate": 3.667502212107841e-06,
+      "loss": 0.691,
+      "step": 30847
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4535039653844803,
+      "learning_rate": 3.6669116768297187e-06,
+      "loss": 0.6592,
+      "step": 30848
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5547833171672714,
+      "learning_rate": 3.6663211784248433e-06,
+      "loss": 0.755,
+      "step": 30849
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5777618399965017,
+      "learning_rate": 3.665730716896647e-06,
+      "loss": 0.5569,
+      "step": 30850
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4197077521765276,
+      "learning_rate": 3.6651402922485715e-06,
+      "loss": 0.6087,
+      "step": 30851
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.20426193969197,
+      "learning_rate": 3.6645499044840594e-06,
+      "loss": 0.6699,
+      "step": 30852
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4872595969055906,
+      "learning_rate": 3.6639595536065418e-06,
+      "loss": 0.6481,
+      "step": 30853
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3523898899545705,
+      "learning_rate": 3.6633692396194554e-06,
+      "loss": 0.6667,
+      "step": 30854
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.695413486101234,
+      "learning_rate": 3.6627789625262434e-06,
+      "loss": 0.7179,
+      "step": 30855
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5287541005601568,
+      "learning_rate": 3.662188722330334e-06,
+      "loss": 0.6052,
+      "step": 30856
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2216748970055118,
+      "learning_rate": 3.6615985190351722e-06,
+      "loss": 0.7751,
+      "step": 30857
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.370938277041448,
+      "learning_rate": 3.6610083526441876e-06,
+      "loss": 0.6601,
+      "step": 30858
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6678759398452245,
+      "learning_rate": 3.660418223160822e-06,
+      "loss": 0.6602,
+      "step": 30859
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0064472928798813,
+      "learning_rate": 3.6598281305885052e-06,
+      "loss": 0.7638,
+      "step": 30860
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4415025697251553,
+      "learning_rate": 3.65923807493068e-06,
+      "loss": 0.7776,
+      "step": 30861
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3037052250799033,
+      "learning_rate": 3.658648056190778e-06,
+      "loss": 0.6487,
+      "step": 30862
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3350400121759138,
+      "learning_rate": 3.6580580743722315e-06,
+      "loss": 0.5851,
+      "step": 30863
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.275712420995139,
+      "learning_rate": 3.6574681294784788e-06,
+      "loss": 0.8225,
+      "step": 30864
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.270652223254675,
+      "learning_rate": 3.6568782215129574e-06,
+      "loss": 0.6338,
+      "step": 30865
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3998799544281366,
+      "learning_rate": 3.6562883504791004e-06,
+      "loss": 0.627,
+      "step": 30866
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3816265335386804,
+      "learning_rate": 3.655698516380336e-06,
+      "loss": 0.7956,
+      "step": 30867
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.8484168536413086,
+      "learning_rate": 3.655108719220104e-06,
+      "loss": 0.6698,
+      "step": 30868
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5318284728580633,
+      "learning_rate": 3.6545189590018416e-06,
+      "loss": 0.718,
+      "step": 30869
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3837158942446788,
+      "learning_rate": 3.653929235728979e-06,
+      "loss": 0.7679,
+      "step": 30870
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0260668712717997,
+      "learning_rate": 3.653339549404946e-06,
+      "loss": 0.7389,
+      "step": 30871
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4493819530432792,
+      "learning_rate": 3.652749900033179e-06,
+      "loss": 0.678,
+      "step": 30872
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3808908308770778,
+      "learning_rate": 3.6521602876171147e-06,
+      "loss": 0.686,
+      "step": 30873
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6194299544900272,
+      "learning_rate": 3.651570712160184e-06,
+      "loss": 0.5142,
+      "step": 30874
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4191505122293147,
+      "learning_rate": 3.650981173665814e-06,
+      "loss": 0.6737,
+      "step": 30875
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6063243799667405,
+      "learning_rate": 3.6503916721374456e-06,
+      "loss": 0.4945,
+      "step": 30876
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2666976532611374,
+      "learning_rate": 3.6498022075785034e-06,
+      "loss": 0.7226,
+      "step": 30877
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6235449544083922,
+      "learning_rate": 3.649212779992427e-06,
+      "loss": 0.6023,
+      "step": 30878
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2480240927180282,
+      "learning_rate": 3.6486233893826405e-06,
+      "loss": 0.7379,
+      "step": 30879
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4848002766571708,
+      "learning_rate": 3.6480340357525832e-06,
+      "loss": 0.8346,
+      "step": 30880
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5806921030245668,
+      "learning_rate": 3.6474447191056793e-06,
+      "loss": 0.4971,
+      "step": 30881
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.531413543883804,
+      "learning_rate": 3.6468554394453636e-06,
+      "loss": 0.7156,
+      "step": 30882
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6734294681950146,
+      "learning_rate": 3.6462661967750744e-06,
+      "loss": 0.6627,
+      "step": 30883
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5235734839721373,
+      "learning_rate": 3.645676991098227e-06,
+      "loss": 0.6878,
+      "step": 30884
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3496064165987096,
+      "learning_rate": 3.6450878224182608e-06,
+      "loss": 0.7033,
+      "step": 30885
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6285015285476397,
+      "learning_rate": 3.6444986907386035e-06,
+      "loss": 0.5228,
+      "step": 30886
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3736530544270107,
+      "learning_rate": 3.6439095960626957e-06,
+      "loss": 0.5715,
+      "step": 30887
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.341454185577203,
+      "learning_rate": 3.643320538393951e-06,
+      "loss": 0.6761,
+      "step": 30888
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4400786242982218,
+      "learning_rate": 3.642731517735807e-06,
+      "loss": 0.6856,
+      "step": 30889
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5286848752203974,
+      "learning_rate": 3.642142534091695e-06,
+      "loss": 0.6442,
+      "step": 30890
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5632395223401276,
+      "learning_rate": 3.641553587465043e-06,
+      "loss": 0.4618,
+      "step": 30891
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5956769164199383,
+      "learning_rate": 3.6409646778592756e-06,
+      "loss": 0.4476,
+      "step": 30892
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.268034282554411,
+      "learning_rate": 3.640375805277825e-06,
+      "loss": 0.7257,
+      "step": 30893
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5958612207800449,
+      "learning_rate": 3.639786969724123e-06,
+      "loss": 0.4991,
+      "step": 30894
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.382358587332581,
+      "learning_rate": 3.6391981712015922e-06,
+      "loss": 0.6147,
+      "step": 30895
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3538524360773367,
+      "learning_rate": 3.638609409713666e-06,
+      "loss": 0.7313,
+      "step": 30896
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.41611946404509,
+      "learning_rate": 3.63802068526377e-06,
+      "loss": 0.7156,
+      "step": 30897
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4463398318472307,
+      "learning_rate": 3.6374319978553284e-06,
+      "loss": 0.691,
+      "step": 30898
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2998080395012528,
+      "learning_rate": 3.636843347491772e-06,
+      "loss": 0.8084,
+      "step": 30899
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3439495833734718,
+      "learning_rate": 3.636254734176532e-06,
+      "loss": 0.6502,
+      "step": 30900
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2836006615061346,
+      "learning_rate": 3.6356661579130314e-06,
+      "loss": 0.7328,
+      "step": 30901
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4556648073531355,
+      "learning_rate": 3.635077618704694e-06,
+      "loss": 0.5872,
+      "step": 30902
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4494432548928082,
+      "learning_rate": 3.63448911655495e-06,
+      "loss": 0.626,
+      "step": 30903
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5915045591952142,
+      "learning_rate": 3.633900651467228e-06,
+      "loss": 0.7482,
+      "step": 30904
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3751000435875032,
+      "learning_rate": 3.6333122234449523e-06,
+      "loss": 0.666,
+      "step": 30905
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.347767524405644,
+      "learning_rate": 3.632723832491546e-06,
+      "loss": 0.6637,
+      "step": 30906
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5905917277512276,
+      "learning_rate": 3.632135478610437e-06,
+      "loss": 0.7946,
+      "step": 30907
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3460742630872273,
+      "learning_rate": 3.6315471618050546e-06,
+      "loss": 0.612,
+      "step": 30908
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2606133730946594,
+      "learning_rate": 3.6309588820788213e-06,
+      "loss": 0.6815,
+      "step": 30909
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6952499600033815,
+      "learning_rate": 3.6303706394351567e-06,
+      "loss": 0.76,
+      "step": 30910
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3541138257897716,
+      "learning_rate": 3.6297824338774957e-06,
+      "loss": 0.7301,
+      "step": 30911
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.559287386206068,
+      "learning_rate": 3.6291942654092547e-06,
+      "loss": 0.8114,
+      "step": 30912
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.375796453120599,
+      "learning_rate": 3.628606134033865e-06,
+      "loss": 0.7558,
+      "step": 30913
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4008346619209906,
+      "learning_rate": 3.6280180397547438e-06,
+      "loss": 0.6514,
+      "step": 30914
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.286293820115557,
+      "learning_rate": 3.6274299825753225e-06,
+      "loss": 0.7517,
+      "step": 30915
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.117223175168792,
+      "learning_rate": 3.626841962499018e-06,
+      "loss": 0.7268,
+      "step": 30916
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3608027885327707,
+      "learning_rate": 3.6262539795292604e-06,
+      "loss": 0.7202,
+      "step": 30917
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8034635183180097,
+      "learning_rate": 3.6256660336694705e-06,
+      "loss": 0.6915,
+      "step": 30918
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3944601077565317,
+      "learning_rate": 3.625078124923067e-06,
+      "loss": 0.7462,
+      "step": 30919
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.404941057438733,
+      "learning_rate": 3.624490253293477e-06,
+      "loss": 0.6238,
+      "step": 30920
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3694808206723303,
+      "learning_rate": 3.6239024187841266e-06,
+      "loss": 0.8606,
+      "step": 30921
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3602983950378287,
+      "learning_rate": 3.623314621398436e-06,
+      "loss": 0.6289,
+      "step": 30922
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.544958354966703,
+      "learning_rate": 3.6227268611398214e-06,
+      "loss": 0.7214,
+      "step": 30923
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4801701856079097,
+      "learning_rate": 3.622139138011711e-06,
+      "loss": 0.6601,
+      "step": 30924
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6099073122220069,
+      "learning_rate": 3.62155145201753e-06,
+      "loss": 0.5366,
+      "step": 30925
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6024361124019053,
+      "learning_rate": 3.6209638031606942e-06,
+      "loss": 0.5235,
+      "step": 30926
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4659695319432373,
+      "learning_rate": 3.6203761914446243e-06,
+      "loss": 0.756,
+      "step": 30927
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3074861645737066,
+      "learning_rate": 3.6197886168727436e-06,
+      "loss": 0.7008,
+      "step": 30928
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.373150919444015,
+      "learning_rate": 3.6192010794484766e-06,
+      "loss": 0.6388,
+      "step": 30929
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4349491565850816,
+      "learning_rate": 3.618613579175242e-06,
+      "loss": 0.7592,
+      "step": 30930
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1933732454372599,
+      "learning_rate": 3.618026116056456e-06,
+      "loss": 0.7148,
+      "step": 30931
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.569959828532168,
+      "learning_rate": 3.617438690095545e-06,
+      "loss": 0.5958,
+      "step": 30932
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6928995762010737,
+      "learning_rate": 3.6168513012959227e-06,
+      "loss": 0.5646,
+      "step": 30933
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.359857908474566,
+      "learning_rate": 3.6162639496610173e-06,
+      "loss": 0.6739,
+      "step": 30934
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.448171801646819,
+      "learning_rate": 3.615676635194242e-06,
+      "loss": 0.6557,
+      "step": 30935
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.283339962461126,
+      "learning_rate": 3.6150893578990198e-06,
+      "loss": 0.6958,
+      "step": 30936
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1613541980050943,
+      "learning_rate": 3.6145021177787666e-06,
+      "loss": 0.5856,
+      "step": 30937
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6114495832252066,
+      "learning_rate": 3.613914914836907e-06,
+      "loss": 0.5197,
+      "step": 30938
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4372902213608798,
+      "learning_rate": 3.613327749076856e-06,
+      "loss": 0.6716,
+      "step": 30939
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5725430033110681,
+      "learning_rate": 3.6127406205020298e-06,
+      "loss": 0.511,
+      "step": 30940
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5466532061474616,
+      "learning_rate": 3.6121535291158506e-06,
+      "loss": 0.8756,
+      "step": 30941
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5134545949183642,
+      "learning_rate": 3.6115664749217383e-06,
+      "loss": 0.7055,
+      "step": 30942
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5442942074820387,
+      "learning_rate": 3.6109794579231094e-06,
+      "loss": 0.6543,
+      "step": 30943
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3694359539320808,
+      "learning_rate": 3.610392478123377e-06,
+      "loss": 0.6408,
+      "step": 30944
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5070823054039975,
+      "learning_rate": 3.609805535525962e-06,
+      "loss": 0.6142,
+      "step": 30945
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4826275126248822,
+      "learning_rate": 3.6092186301342858e-06,
+      "loss": 0.6753,
+      "step": 30946
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4776072109902403,
+      "learning_rate": 3.608631761951763e-06,
+      "loss": 0.7982,
+      "step": 30947
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3304405040954415,
+      "learning_rate": 3.6080449309818045e-06,
+      "loss": 0.7809,
+      "step": 30948
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.493259924998336,
+      "learning_rate": 3.607458137227834e-06,
+      "loss": 0.6236,
+      "step": 30949
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4303583129269617,
+      "learning_rate": 3.6068713806932678e-06,
+      "loss": 0.6713,
+      "step": 30950
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3080518965712709,
+      "learning_rate": 3.6062846613815217e-06,
+      "loss": 0.6303,
+      "step": 30951
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.622504189994361,
+      "learning_rate": 3.6056979792960067e-06,
+      "loss": 0.7215,
+      "step": 30952
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.313799529924334,
+      "learning_rate": 3.6051113344401457e-06,
+      "loss": 0.5102,
+      "step": 30953
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3691292503819918,
+      "learning_rate": 3.604524726817348e-06,
+      "loss": 0.7087,
+      "step": 30954
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9884163521672689,
+      "learning_rate": 3.603938156431035e-06,
+      "loss": 0.8016,
+      "step": 30955
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.244467309473289,
+      "learning_rate": 3.6033516232846154e-06,
+      "loss": 0.6123,
+      "step": 30956
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4244705930872856,
+      "learning_rate": 3.602765127381511e-06,
+      "loss": 0.6429,
+      "step": 30957
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3796452948411788,
+      "learning_rate": 3.6021786687251315e-06,
+      "loss": 0.8356,
+      "step": 30958
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2596131564024158,
+      "learning_rate": 3.601592247318896e-06,
+      "loss": 0.5472,
+      "step": 30959
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3706891719249807,
+      "learning_rate": 3.601005863166216e-06,
+      "loss": 0.7258,
+      "step": 30960
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.061011693844532,
+      "learning_rate": 3.600419516270501e-06,
+      "loss": 0.6769,
+      "step": 30961
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5825104612030516,
+      "learning_rate": 3.5998332066351716e-06,
+      "loss": 0.6525,
+      "step": 30962
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4659008260428208,
+      "learning_rate": 3.599246934263643e-06,
+      "loss": 0.7036,
+      "step": 30963
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4564565236345255,
+      "learning_rate": 3.598660699159324e-06,
+      "loss": 0.7155,
+      "step": 30964
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3227708387122816,
+      "learning_rate": 3.5980745013256258e-06,
+      "loss": 0.7054,
+      "step": 30965
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4829497422730453,
+      "learning_rate": 3.5974883407659655e-06,
+      "loss": 0.8354,
+      "step": 30966
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3186926638999366,
+      "learning_rate": 3.5969022174837576e-06,
+      "loss": 0.745,
+      "step": 30967
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9638023551157273,
+      "learning_rate": 3.5963161314824123e-06,
+      "loss": 0.698,
+      "step": 30968
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3244076142389605,
+      "learning_rate": 3.5957300827653385e-06,
+      "loss": 0.7441,
+      "step": 30969
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.26564484327009,
+      "learning_rate": 3.595144071335952e-06,
+      "loss": 0.5797,
+      "step": 30970
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.41964281242975,
+      "learning_rate": 3.594558097197668e-06,
+      "loss": 0.7281,
+      "step": 30971
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.605359872523824,
+      "learning_rate": 3.5939721603538946e-06,
+      "loss": 0.7542,
+      "step": 30972
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5994733153833286,
+      "learning_rate": 3.5933862608080405e-06,
+      "loss": 0.5152,
+      "step": 30973
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.71127635844573,
+      "learning_rate": 3.592800398563522e-06,
+      "loss": 0.6514,
+      "step": 30974
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2997468939143126,
+      "learning_rate": 3.5922145736237457e-06,
+      "loss": 0.7352,
+      "step": 30975
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.439872051634756,
+      "learning_rate": 3.5916287859921273e-06,
+      "loss": 0.6633,
+      "step": 30976
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4000310962318636,
+      "learning_rate": 3.5910430356720727e-06,
+      "loss": 0.8117,
+      "step": 30977
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4592008637541056,
+      "learning_rate": 3.590457322666997e-06,
+      "loss": 0.6211,
+      "step": 30978
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3527270225099923,
+      "learning_rate": 3.589871646980305e-06,
+      "loss": 0.5405,
+      "step": 30979
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6038301252302399,
+      "learning_rate": 3.5892860086154103e-06,
+      "loss": 0.533,
+      "step": 30980
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.353123652738199,
+      "learning_rate": 3.588700407575725e-06,
+      "loss": 0.7106,
+      "step": 30981
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.146637736712843,
+      "learning_rate": 3.588114843864655e-06,
+      "loss": 0.629,
+      "step": 30982
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2340763214233579,
+      "learning_rate": 3.5875293174856073e-06,
+      "loss": 0.552,
+      "step": 30983
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5961432514434608,
+      "learning_rate": 3.5869438284419946e-06,
+      "loss": 0.7275,
+      "step": 30984
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4892100594243805,
+      "learning_rate": 3.586358376737231e-06,
+      "loss": 0.7192,
+      "step": 30985
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5093893626758035,
+      "learning_rate": 3.585772962374714e-06,
+      "loss": 0.7149,
+      "step": 30986
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5789697795726978,
+      "learning_rate": 3.5851875853578568e-06,
+      "loss": 0.7369,
+      "step": 30987
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4093656176958838,
+      "learning_rate": 3.584602245690069e-06,
+      "loss": 0.7373,
+      "step": 30988
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4851780792811629,
+      "learning_rate": 3.584016943374764e-06,
+      "loss": 0.618,
+      "step": 30989
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8554035311141357,
+      "learning_rate": 3.583431678415337e-06,
+      "loss": 0.7548,
+      "step": 30990
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3791714392395598,
+      "learning_rate": 3.582846450815202e-06,
+      "loss": 0.7199,
+      "step": 30991
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5297668719545954,
+      "learning_rate": 3.5822612605777706e-06,
+      "loss": 0.8567,
+      "step": 30992
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.551950579699927,
+      "learning_rate": 3.581676107706442e-06,
+      "loss": 0.8183,
+      "step": 30993
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3420211415088346,
+      "learning_rate": 3.581090992204631e-06,
+      "loss": 0.7508,
+      "step": 30994
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6276004314697639,
+      "learning_rate": 3.580505914075737e-06,
+      "loss": 0.6285,
+      "step": 30995
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5624380634929196,
+      "learning_rate": 3.5799208733231727e-06,
+      "loss": 0.7671,
+      "step": 30996
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.543424325013695,
+      "learning_rate": 3.579335869950339e-06,
+      "loss": 0.7658,
+      "step": 30997
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2151450885162398,
+      "learning_rate": 3.578750903960647e-06,
+      "loss": 0.5979,
+      "step": 30998
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9736858106253048,
+      "learning_rate": 3.5781659753575015e-06,
+      "loss": 0.7343,
+      "step": 30999
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4649966731516586,
+      "learning_rate": 3.5775810841443025e-06,
+      "loss": 0.6207,
+      "step": 31000
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5410837788067847,
+      "learning_rate": 3.5769962303244597e-06,
+      "loss": 0.4461,
+      "step": 31001
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.513441801294249,
+      "learning_rate": 3.5764114139013828e-06,
+      "loss": 0.5773,
+      "step": 31002
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4702058107679963,
+      "learning_rate": 3.575826634878471e-06,
+      "loss": 0.5644,
+      "step": 31003
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.343988599909805,
+      "learning_rate": 3.575241893259128e-06,
+      "loss": 0.7461,
+      "step": 31004
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.5661772367459985,
+      "learning_rate": 3.5746571890467607e-06,
+      "loss": 0.7664,
+      "step": 31005
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3370185526734555,
+      "learning_rate": 3.574072522244776e-06,
+      "loss": 0.6089,
+      "step": 31006
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.467665502884204,
+      "learning_rate": 3.5734878928565763e-06,
+      "loss": 0.7223,
+      "step": 31007
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6117816995335635,
+      "learning_rate": 3.572903300885561e-06,
+      "loss": 0.5192,
+      "step": 31008
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5863964097758008,
+      "learning_rate": 3.572318746335137e-06,
+      "loss": 0.6446,
+      "step": 31009
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5616056134804068,
+      "learning_rate": 3.571734229208712e-06,
+      "loss": 0.5917,
+      "step": 31010
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3391792730882768,
+      "learning_rate": 3.571149749509686e-06,
+      "loss": 0.5954,
+      "step": 31011
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6359748184783512,
+      "learning_rate": 3.570565307241457e-06,
+      "loss": 0.4833,
+      "step": 31012
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.613311843083729,
+      "learning_rate": 3.5699809024074373e-06,
+      "loss": 0.7534,
+      "step": 31013
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.337313486278335,
+      "learning_rate": 3.5693965350110194e-06,
+      "loss": 0.7035,
+      "step": 31014
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5175172451711016,
+      "learning_rate": 3.568812205055615e-06,
+      "loss": 0.7061,
+      "step": 31015
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.831272068193375,
+      "learning_rate": 3.5682279125446185e-06,
+      "loss": 0.6669,
+      "step": 31016
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8782525148660276,
+      "learning_rate": 3.567643657481439e-06,
+      "loss": 0.6875,
+      "step": 31017
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6661368541745452,
+      "learning_rate": 3.5670594398694713e-06,
+      "loss": 0.6613,
+      "step": 31018
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2532041614476135,
+      "learning_rate": 3.5664752597121233e-06,
+      "loss": 0.7966,
+      "step": 31019
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2317304998560579,
+      "learning_rate": 3.565891117012793e-06,
+      "loss": 0.6672,
+      "step": 31020
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2702489255281166,
+      "learning_rate": 3.5653070117748777e-06,
+      "loss": 0.6635,
+      "step": 31021
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.588980563404746,
+      "learning_rate": 3.5647229440017827e-06,
+      "loss": 0.7302,
+      "step": 31022
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.007425690945737,
+      "learning_rate": 3.564138913696912e-06,
+      "loss": 0.6971,
+      "step": 31023
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4108142919608597,
+      "learning_rate": 3.5635549208636623e-06,
+      "loss": 0.7234,
+      "step": 31024
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3426932492046062,
+      "learning_rate": 3.562970965505429e-06,
+      "loss": 0.584,
+      "step": 31025
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.317458532667727,
+      "learning_rate": 3.5623870476256174e-06,
+      "loss": 0.7652,
+      "step": 31026
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4936934252139469,
+      "learning_rate": 3.56180316722763e-06,
+      "loss": 0.6978,
+      "step": 31027
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5921872430561551,
+      "learning_rate": 3.5612193243148617e-06,
+      "loss": 0.4816,
+      "step": 31028
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3770422158752817,
+      "learning_rate": 3.5606355188907116e-06,
+      "loss": 0.7301,
+      "step": 31029
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3463738768003806,
+      "learning_rate": 3.5600517509585784e-06,
+      "loss": 0.7474,
+      "step": 31030
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3944335946904667,
+      "learning_rate": 3.5594680205218667e-06,
+      "loss": 0.7773,
+      "step": 31031
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.178121525540627,
+      "learning_rate": 3.558884327583971e-06,
+      "loss": 0.6074,
+      "step": 31032
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5579904799463693,
+      "learning_rate": 3.5583006721482873e-06,
+      "loss": 0.589,
+      "step": 31033
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.5881784225841744,
+      "learning_rate": 3.5577170542182195e-06,
+      "loss": 0.7603,
+      "step": 31034
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6908419324691177,
+      "learning_rate": 3.5571334737971596e-06,
+      "loss": 0.7743,
+      "step": 31035
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.602828288134221,
+      "learning_rate": 3.556549930888511e-06,
+      "loss": 0.6815,
+      "step": 31036
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2955020295304234,
+      "learning_rate": 3.555966425495666e-06,
+      "loss": 0.6019,
+      "step": 31037
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.498409004011635,
+      "learning_rate": 3.5553829576220277e-06,
+      "loss": 0.788,
+      "step": 31038
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4413998145718825,
+      "learning_rate": 3.554799527270987e-06,
+      "loss": 0.6742,
+      "step": 31039
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6250109221763231,
+      "learning_rate": 3.554216134445947e-06,
+      "loss": 0.506,
+      "step": 31040
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.061210437584272,
+      "learning_rate": 3.553632779150301e-06,
+      "loss": 0.6216,
+      "step": 31041
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5602250355989158,
+      "learning_rate": 3.5530494613874433e-06,
+      "loss": 0.5178,
+      "step": 31042
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7210433018403866,
+      "learning_rate": 3.552466181160772e-06,
+      "loss": 0.793,
+      "step": 31043
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4886786949099844,
+      "learning_rate": 3.5518829384736888e-06,
+      "loss": 0.6609,
+      "step": 31044
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3566606715419371,
+      "learning_rate": 3.551299733329583e-06,
+      "loss": 0.5376,
+      "step": 31045
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5948575603178303,
+      "learning_rate": 3.5507165657318487e-06,
+      "loss": 0.6914,
+      "step": 31046
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4298737124786542,
+      "learning_rate": 3.5501334356838846e-06,
+      "loss": 0.6722,
+      "step": 31047
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4044039144482545,
+      "learning_rate": 3.549550343189089e-06,
+      "loss": 0.6663,
+      "step": 31048
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4099263162743474,
+      "learning_rate": 3.548967288250853e-06,
+      "loss": 0.7869,
+      "step": 31049
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4445741133367684,
+      "learning_rate": 3.548384270872568e-06,
+      "loss": 0.6729,
+      "step": 31050
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4638077442986206,
+      "learning_rate": 3.5478012910576332e-06,
+      "loss": 0.7453,
+      "step": 31051
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.548736009314105,
+      "learning_rate": 3.547218348809446e-06,
+      "loss": 0.6782,
+      "step": 31052
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4597577056441549,
+      "learning_rate": 3.5466354441313954e-06,
+      "loss": 0.6712,
+      "step": 31053
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4986977051288526,
+      "learning_rate": 3.5460525770268726e-06,
+      "loss": 0.692,
+      "step": 31054
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3015939897806028,
+      "learning_rate": 3.5454697474992796e-06,
+      "loss": 0.7108,
+      "step": 31055
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.370993881306506,
+      "learning_rate": 3.5448869555520016e-06,
+      "loss": 0.675,
+      "step": 31056
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.530670323497275,
+      "learning_rate": 3.5443042011884377e-06,
+      "loss": 0.6771,
+      "step": 31057
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.593508924477214,
+      "learning_rate": 3.543721484411976e-06,
+      "loss": 0.6045,
+      "step": 31058
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3189805279721516,
+      "learning_rate": 3.543138805226015e-06,
+      "loss": 0.6725,
+      "step": 31059
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5645451118405809,
+      "learning_rate": 3.542556163633941e-06,
+      "loss": 0.4792,
+      "step": 31060
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.2924964586716934,
+      "learning_rate": 3.5419735596391537e-06,
+      "loss": 0.6761,
+      "step": 31061
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.231770825899817,
+      "learning_rate": 3.5413909932450398e-06,
+      "loss": 0.6558,
+      "step": 31062
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.206622095832049,
+      "learning_rate": 3.540808464454989e-06,
+      "loss": 0.7265,
+      "step": 31063
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6451012343357139,
+      "learning_rate": 3.5402259732723953e-06,
+      "loss": 0.9165,
+      "step": 31064
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3141348888241562,
+      "learning_rate": 3.539643519700653e-06,
+      "loss": 0.7469,
+      "step": 31065
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.308554851186027,
+      "learning_rate": 3.5390611037431564e-06,
+      "loss": 0.6739,
+      "step": 31066
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3501673283155644,
+      "learning_rate": 3.538478725403286e-06,
+      "loss": 0.6877,
+      "step": 31067
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6197784658451283,
+      "learning_rate": 3.5378963846844383e-06,
+      "loss": 0.4964,
+      "step": 31068
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4713004429821548,
+      "learning_rate": 3.5373140815900065e-06,
+      "loss": 0.6864,
+      "step": 31069
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6580700201429017,
+      "learning_rate": 3.5367318161233777e-06,
+      "loss": 0.5797,
+      "step": 31070
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3767698953595253,
+      "learning_rate": 3.5361495882879406e-06,
+      "loss": 0.7783,
+      "step": 31071
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4769713470285302,
+      "learning_rate": 3.5355673980870853e-06,
+      "loss": 0.6675,
+      "step": 31072
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3441914613217454,
+      "learning_rate": 3.5349852455242086e-06,
+      "loss": 0.6365,
+      "step": 31073
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5216322200935573,
+      "learning_rate": 3.534403130602693e-06,
+      "loss": 0.5818,
+      "step": 31074
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.544808161984976,
+      "learning_rate": 3.5338210533259264e-06,
+      "loss": 0.8359,
+      "step": 31075
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3931312389572712,
+      "learning_rate": 3.5332390136973036e-06,
+      "loss": 0.763,
+      "step": 31076
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5621856792675608,
+      "learning_rate": 3.5326570117202076e-06,
+      "loss": 0.7903,
+      "step": 31077
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4045585943576135,
+      "learning_rate": 3.5320750473980304e-06,
+      "loss": 0.704,
+      "step": 31078
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4080415644624686,
+      "learning_rate": 3.531493120734163e-06,
+      "loss": 0.6364,
+      "step": 31079
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6216790715596763,
+      "learning_rate": 3.5309112317319916e-06,
+      "loss": 0.5762,
+      "step": 31080
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4774398871435523,
+      "learning_rate": 3.530329380394898e-06,
+      "loss": 0.8514,
+      "step": 31081
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3046540979615515,
+      "learning_rate": 3.529747566726276e-06,
+      "loss": 0.6935,
+      "step": 31082
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.154058814221893,
+      "learning_rate": 3.5291657907295184e-06,
+      "loss": 0.6604,
+      "step": 31083
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4610756200635064,
+      "learning_rate": 3.5285840524079994e-06,
+      "loss": 0.6343,
+      "step": 31084
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6016471906179236,
+      "learning_rate": 3.528002351765114e-06,
+      "loss": 0.5028,
+      "step": 31085
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6192104993241228,
+      "learning_rate": 3.527420688804247e-06,
+      "loss": 0.7022,
+      "step": 31086
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.520978497409908,
+      "learning_rate": 3.526839063528793e-06,
+      "loss": 0.7154,
+      "step": 31087
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3674852810927696,
+      "learning_rate": 3.5262574759421243e-06,
+      "loss": 0.9185,
+      "step": 31088
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.370629870211449,
+      "learning_rate": 3.525675926047634e-06,
+      "loss": 0.5581,
+      "step": 31089
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5762584880559974,
+      "learning_rate": 3.525094413848712e-06,
+      "loss": 0.4524,
+      "step": 31090
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6644184837194962,
+      "learning_rate": 3.5245129393487365e-06,
+      "loss": 0.7158,
+      "step": 31091
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6262207453013847,
+      "learning_rate": 3.5239315025511e-06,
+      "loss": 0.5169,
+      "step": 31092
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6552725443291192,
+      "learning_rate": 3.523350103459182e-06,
+      "loss": 0.5003,
+      "step": 31093
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7801831028976485,
+      "learning_rate": 3.522768742076372e-06,
+      "loss": 0.769,
+      "step": 31094
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5844624692783364,
+      "learning_rate": 3.5221874184060513e-06,
+      "loss": 0.7169,
+      "step": 31095
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9002908897303377,
+      "learning_rate": 3.5216061324516093e-06,
+      "loss": 0.6687,
+      "step": 31096
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2719950657630494,
+      "learning_rate": 3.5210248842164274e-06,
+      "loss": 0.778,
+      "step": 31097
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.548187809639309,
+      "learning_rate": 3.520443673703886e-06,
+      "loss": 0.6301,
+      "step": 31098
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4616565454045225,
+      "learning_rate": 3.5198625009173726e-06,
+      "loss": 0.7864,
+      "step": 31099
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5827921889423116,
+      "learning_rate": 3.5192813658602754e-06,
+      "loss": 0.8115,
+      "step": 31100
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3055974804525585,
+      "learning_rate": 3.518700268535974e-06,
+      "loss": 0.6629,
+      "step": 31101
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4165192251156224,
+      "learning_rate": 3.5181192089478477e-06,
+      "loss": 0.6658,
+      "step": 31102
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4429941909905142,
+      "learning_rate": 3.517538187099284e-06,
+      "loss": 0.5743,
+      "step": 31103
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.417319688945195,
+      "learning_rate": 3.5169572029936682e-06,
+      "loss": 0.7433,
+      "step": 31104
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3793351149421866,
+      "learning_rate": 3.516376256634381e-06,
+      "loss": 0.7168,
+      "step": 31105
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5654368325873839,
+      "learning_rate": 3.5157953480247985e-06,
+      "loss": 0.493,
+      "step": 31106
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5853542844509407,
+      "learning_rate": 3.5152144771683107e-06,
+      "loss": 0.644,
+      "step": 31107
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3073938523813315,
+      "learning_rate": 3.5146336440682995e-06,
+      "loss": 0.5919,
+      "step": 31108
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2686385438521401,
+      "learning_rate": 3.5140528487281443e-06,
+      "loss": 0.7405,
+      "step": 31109
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6009054152626536,
+      "learning_rate": 3.5134720911512233e-06,
+      "loss": 0.6454,
+      "step": 31110
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3708171795319368,
+      "learning_rate": 3.512891371340922e-06,
+      "loss": 0.7044,
+      "step": 31111
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9450160533252947,
+      "learning_rate": 3.5123106893006242e-06,
+      "loss": 0.7299,
+      "step": 31112
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4105539282074675,
+      "learning_rate": 3.5117300450337066e-06,
+      "loss": 0.5355,
+      "step": 31113
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.534313646163942,
+      "learning_rate": 3.5111494385435484e-06,
+      "loss": 0.7497,
+      "step": 31114
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4075276265498458,
+      "learning_rate": 3.5105688698335363e-06,
+      "loss": 0.6598,
+      "step": 31115
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2987731088373076,
+      "learning_rate": 3.509988338907042e-06,
+      "loss": 0.792,
+      "step": 31116
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3679056579293636,
+      "learning_rate": 3.5094078457674553e-06,
+      "loss": 0.719,
+      "step": 31117
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4597048938317432,
+      "learning_rate": 3.5088273904181457e-06,
+      "loss": 0.6151,
+      "step": 31118
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3530674621468037,
+      "learning_rate": 3.5082469728625034e-06,
+      "loss": 0.7117,
+      "step": 31119
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.502779367312449,
+      "learning_rate": 3.5076665931038976e-06,
+      "loss": 0.7338,
+      "step": 31120
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8935119366695226,
+      "learning_rate": 3.5070862511457172e-06,
+      "loss": 0.7483,
+      "step": 31121
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5977032294917293,
+      "learning_rate": 3.5065059469913356e-06,
+      "loss": 0.6317,
+      "step": 31122
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3143864118164905,
+      "learning_rate": 3.5059256806441276e-06,
+      "loss": 0.5532,
+      "step": 31123
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3861082072616908,
+      "learning_rate": 3.505345452107478e-06,
+      "loss": 0.6974,
+      "step": 31124
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.65192768319359,
+      "learning_rate": 3.504765261384766e-06,
+      "loss": 0.6304,
+      "step": 31125
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.293600689833717,
+      "learning_rate": 3.5041851084793665e-06,
+      "loss": 0.6313,
+      "step": 31126
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5077918344570698,
+      "learning_rate": 3.503604993394656e-06,
+      "loss": 0.6313,
+      "step": 31127
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3919057958244674,
+      "learning_rate": 3.5030249161340124e-06,
+      "loss": 0.6281,
+      "step": 31128
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5162143285739804,
+      "learning_rate": 3.502444876700819e-06,
+      "loss": 0.7213,
+      "step": 31129
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5059275686491012,
+      "learning_rate": 3.5018648750984473e-06,
+      "loss": 0.696,
+      "step": 31130
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2594078356847762,
+      "learning_rate": 3.5012849113302737e-06,
+      "loss": 0.7475,
+      "step": 31131
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3687572497308274,
+      "learning_rate": 3.5007049853996767e-06,
+      "loss": 0.7602,
+      "step": 31132
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4141900301697414,
+      "learning_rate": 3.5001250973100353e-06,
+      "loss": 0.6134,
+      "step": 31133
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.383539283266218,
+      "learning_rate": 3.499545247064724e-06,
+      "loss": 0.5641,
+      "step": 31134
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.510725737154578,
+      "learning_rate": 3.4989654346671143e-06,
+      "loss": 0.6963,
+      "step": 31135
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.317963915859677,
+      "learning_rate": 3.4983856601205903e-06,
+      "loss": 0.709,
+      "step": 31136
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2956456800835225,
+      "learning_rate": 3.49780592342852e-06,
+      "loss": 0.6762,
+      "step": 31137
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2989549587461036,
+      "learning_rate": 3.4972262245942847e-06,
+      "loss": 0.6621,
+      "step": 31138
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6094301175473458,
+      "learning_rate": 3.4966465636212543e-06,
+      "loss": 0.5341,
+      "step": 31139
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5961068055011106,
+      "learning_rate": 3.49606694051281e-06,
+      "loss": 0.5214,
+      "step": 31140
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2881235236903235,
+      "learning_rate": 3.4954873552723188e-06,
+      "loss": 0.5659,
+      "step": 31141
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3390339843279018,
+      "learning_rate": 3.494907807903164e-06,
+      "loss": 0.5786,
+      "step": 31142
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.0350743400499796,
+      "learning_rate": 3.4943282984087147e-06,
+      "loss": 0.647,
+      "step": 31143
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4976859798410125,
+      "learning_rate": 3.4937488267923426e-06,
+      "loss": 0.6317,
+      "step": 31144
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3612794635439351,
+      "learning_rate": 3.493169393057425e-06,
+      "loss": 0.6807,
+      "step": 31145
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2588604655098954,
+      "learning_rate": 3.4925899972073387e-06,
+      "loss": 0.6919,
+      "step": 31146
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.562105049919253,
+      "learning_rate": 3.4920106392454523e-06,
+      "loss": 0.5128,
+      "step": 31147
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4516346381305025,
+      "learning_rate": 3.491431319175138e-06,
+      "loss": 0.6124,
+      "step": 31148
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.23507508805861,
+      "learning_rate": 3.490852036999771e-06,
+      "loss": 0.6821,
+      "step": 31149
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5272713081638363,
+      "learning_rate": 3.4902727927227277e-06,
+      "loss": 0.6794,
+      "step": 31150
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2724925346597733,
+      "learning_rate": 3.489693586347378e-06,
+      "loss": 0.6637,
+      "step": 31151
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5182650483548916,
+      "learning_rate": 3.4891144178770885e-06,
+      "loss": 0.6655,
+      "step": 31152
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3371567362753265,
+      "learning_rate": 3.488535287315237e-06,
+      "loss": 0.6804,
+      "step": 31153
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3583914322971555,
+      "learning_rate": 3.4879561946651974e-06,
+      "loss": 0.6919,
+      "step": 31154
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3678010808697207,
+      "learning_rate": 3.487377139930339e-06,
+      "loss": 0.5802,
+      "step": 31155
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3400272607748436,
+      "learning_rate": 3.486798123114029e-06,
+      "loss": 0.7304,
+      "step": 31156
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.588437569183132,
+      "learning_rate": 3.486219144219646e-06,
+      "loss": 0.7265,
+      "step": 31157
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6201882361607196,
+      "learning_rate": 3.485640203250553e-06,
+      "loss": 0.5604,
+      "step": 31158
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4303007597188448,
+      "learning_rate": 3.4850613002101295e-06,
+      "loss": 0.5476,
+      "step": 31159
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2561345542531197,
+      "learning_rate": 3.4844824351017394e-06,
+      "loss": 0.5766,
+      "step": 31160
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4057118095119805,
+      "learning_rate": 3.483903607928757e-06,
+      "loss": 0.6552,
+      "step": 31161
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6060953524880022,
+      "learning_rate": 3.4833248186945477e-06,
+      "loss": 0.4895,
+      "step": 31162
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5883943085316511,
+      "learning_rate": 3.4827460674024847e-06,
+      "loss": 0.519,
+      "step": 31163
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5145161773579903,
+      "learning_rate": 3.4821673540559442e-06,
+      "loss": 0.6612,
+      "step": 31164
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2960267729265103,
+      "learning_rate": 3.481588678658282e-06,
+      "loss": 0.6657,
+      "step": 31165
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6153518279705879,
+      "learning_rate": 3.4810100412128743e-06,
+      "loss": 0.6446,
+      "step": 31166
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9882577826038605,
+      "learning_rate": 3.48043144172309e-06,
+      "loss": 0.6813,
+      "step": 31167
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2679146740308394,
+      "learning_rate": 3.479852880192305e-06,
+      "loss": 0.7053,
+      "step": 31168
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6098374775922755,
+      "learning_rate": 3.479274356623874e-06,
+      "loss": 0.748,
+      "step": 31169
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3922949496834547,
+      "learning_rate": 3.4786958710211714e-06,
+      "loss": 0.6976,
+      "step": 31170
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.468361544454452,
+      "learning_rate": 3.4781174233875705e-06,
+      "loss": 0.5995,
+      "step": 31171
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1985270424568542,
+      "learning_rate": 3.477539013726433e-06,
+      "loss": 0.6302,
+      "step": 31172
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3411244687330106,
+      "learning_rate": 3.4769606420411273e-06,
+      "loss": 0.6966,
+      "step": 31173
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.343475389704803,
+      "learning_rate": 3.4763823083350202e-06,
+      "loss": 0.8412,
+      "step": 31174
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.622643446617459,
+      "learning_rate": 3.4758040126114845e-06,
+      "loss": 0.7329,
+      "step": 31175
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.575222904053984,
+      "learning_rate": 3.47522575487388e-06,
+      "loss": 0.7348,
+      "step": 31176
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5710749166281077,
+      "learning_rate": 3.4746475351255806e-06,
+      "loss": 0.6645,
+      "step": 31177
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4898784927766513,
+      "learning_rate": 3.474069353369949e-06,
+      "loss": 0.7394,
+      "step": 31178
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4980881373207353,
+      "learning_rate": 3.473491209610347e-06,
+      "loss": 0.6467,
+      "step": 31179
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.56862133254423,
+      "learning_rate": 3.4729131038501475e-06,
+      "loss": 0.6507,
+      "step": 31180
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5541525265020438,
+      "learning_rate": 3.4723350360927165e-06,
+      "loss": 0.4324,
+      "step": 31181
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3470536827870168,
+      "learning_rate": 3.4717570063414174e-06,
+      "loss": 0.6401,
+      "step": 31182
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5637995916559915,
+      "learning_rate": 3.4711790145996127e-06,
+      "loss": 0.5933,
+      "step": 31183
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5153565850905664,
+      "learning_rate": 3.470601060870671e-06,
+      "loss": 0.6256,
+      "step": 31184
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4446597008055513,
+      "learning_rate": 3.4700231451579634e-06,
+      "loss": 0.7078,
+      "step": 31185
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6572996869434955,
+      "learning_rate": 3.4694452674648425e-06,
+      "loss": 0.685,
+      "step": 31186
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7365179896403298,
+      "learning_rate": 3.4688674277946775e-06,
+      "loss": 0.7726,
+      "step": 31187
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4912055183174309,
+      "learning_rate": 3.468289626150836e-06,
+      "loss": 0.7123,
+      "step": 31188
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7083856909078512,
+      "learning_rate": 3.4677118625366825e-06,
+      "loss": 0.7844,
+      "step": 31189
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.644747944797217,
+      "learning_rate": 3.4671341369555776e-06,
+      "loss": 0.6327,
+      "step": 31190
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3387394233475398,
+      "learning_rate": 3.4665564494108837e-06,
+      "loss": 0.762,
+      "step": 31191
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6694124923259048,
+      "learning_rate": 3.46597879990597e-06,
+      "loss": 0.662,
+      "step": 31192
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7365782094791646,
+      "learning_rate": 3.465401188444193e-06,
+      "loss": 0.8056,
+      "step": 31193
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.574189043557676,
+      "learning_rate": 3.464823615028923e-06,
+      "loss": 0.4899,
+      "step": 31194
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.961705903381725,
+      "learning_rate": 3.4642460796635146e-06,
+      "loss": 0.6427,
+      "step": 31195
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.261560435271115,
+      "learning_rate": 3.4636685823513384e-06,
+      "loss": 0.7399,
+      "step": 31196
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3525726624043997,
+      "learning_rate": 3.4630911230957486e-06,
+      "loss": 0.643,
+      "step": 31197
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3903024616507094,
+      "learning_rate": 3.4625137019001164e-06,
+      "loss": 0.7177,
+      "step": 31198
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.276365232926757,
+      "learning_rate": 3.461936318767799e-06,
+      "loss": 0.4958,
+      "step": 31199
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.89408957401488,
+      "learning_rate": 3.4613589737021557e-06,
+      "loss": 0.7265,
+      "step": 31200
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.293572110538053,
+      "learning_rate": 3.4607816667065497e-06,
+      "loss": 0.5982,
+      "step": 31201
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6861172503282555,
+      "learning_rate": 3.460204397784347e-06,
+      "loss": 0.631,
+      "step": 31202
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4189260483162287,
+      "learning_rate": 3.459627166938904e-06,
+      "loss": 0.6136,
+      "step": 31203
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.296816808728683,
+      "learning_rate": 3.4590499741735796e-06,
+      "loss": 0.6451,
+      "step": 31204
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2304209137527526,
+      "learning_rate": 3.4584728194917363e-06,
+      "loss": 0.7302,
+      "step": 31205
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3453966414257432,
+      "learning_rate": 3.4578957028967388e-06,
+      "loss": 0.7488,
+      "step": 31206
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3220959719457839,
+      "learning_rate": 3.4573186243919443e-06,
+      "loss": 0.7953,
+      "step": 31207
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4158710054526298,
+      "learning_rate": 3.4567415839807083e-06,
+      "loss": 0.7766,
+      "step": 31208
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5188575468013115,
+      "learning_rate": 3.4561645816663945e-06,
+      "loss": 0.6829,
+      "step": 31209
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6086899307085352,
+      "learning_rate": 3.455587617452365e-06,
+      "loss": 0.5475,
+      "step": 31210
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7550273479607341,
+      "learning_rate": 3.455010691341978e-06,
+      "loss": 0.6278,
+      "step": 31211
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2284506030574047,
+      "learning_rate": 3.454433803338586e-06,
+      "loss": 0.4356,
+      "step": 31212
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3906686093538443,
+      "learning_rate": 3.453856953445557e-06,
+      "loss": 0.8067,
+      "step": 31213
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.884803969613633,
+      "learning_rate": 3.453280141666241e-06,
+      "loss": 0.7656,
+      "step": 31214
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5909485037170834,
+      "learning_rate": 3.4527033680040047e-06,
+      "loss": 0.6153,
+      "step": 31215
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2997507067267464,
+      "learning_rate": 3.4521266324621983e-06,
+      "loss": 0.626,
+      "step": 31216
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3982620426378274,
+      "learning_rate": 3.451549935044187e-06,
+      "loss": 0.8638,
+      "step": 31217
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3089766110720054,
+      "learning_rate": 3.450973275753322e-06,
+      "loss": 0.7225,
+      "step": 31218
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.396875750715799,
+      "learning_rate": 3.450396654592968e-06,
+      "loss": 0.6607,
+      "step": 31219
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3014650950054463,
+      "learning_rate": 3.4498200715664775e-06,
+      "loss": 0.7196,
+      "step": 31220
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.464577030715113,
+      "learning_rate": 3.449243526677205e-06,
+      "loss": 0.7457,
+      "step": 31221
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.541752307529935,
+      "learning_rate": 3.448667019928511e-06,
+      "loss": 0.6469,
+      "step": 31222
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2421828388096294,
+      "learning_rate": 3.448090551323755e-06,
+      "loss": 0.6166,
+      "step": 31223
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3864313321747008,
+      "learning_rate": 3.44751412086629e-06,
+      "loss": 0.621,
+      "step": 31224
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.45209525524386,
+      "learning_rate": 3.4469377285594687e-06,
+      "loss": 0.6466,
+      "step": 31225
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5900209475806506,
+      "learning_rate": 3.4463613744066516e-06,
+      "loss": 0.7223,
+      "step": 31226
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8499200119580586,
+      "learning_rate": 3.445785058411195e-06,
+      "loss": 0.6111,
+      "step": 31227
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6074253743157212,
+      "learning_rate": 3.4452087805764544e-06,
+      "loss": 0.731,
+      "step": 31228
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3965179309906532,
+      "learning_rate": 3.4446325409057792e-06,
+      "loss": 0.7589,
+      "step": 31229
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4931315044868476,
+      "learning_rate": 3.4440563394025294e-06,
+      "loss": 0.6444,
+      "step": 31230
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3043497165479239,
+      "learning_rate": 3.443480176070062e-06,
+      "loss": 0.6575,
+      "step": 31231
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.555782286065675,
+      "learning_rate": 3.4429040509117295e-06,
+      "loss": 0.7077,
+      "step": 31232
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3900821608278382,
+      "learning_rate": 3.4423279639308815e-06,
+      "loss": 0.6927,
+      "step": 31233
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3809288133291775,
+      "learning_rate": 3.4417519151308763e-06,
+      "loss": 0.674,
+      "step": 31234
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3518319533621823,
+      "learning_rate": 3.441175904515072e-06,
+      "loss": 0.6184,
+      "step": 31235
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5934385539082843,
+      "learning_rate": 3.4405999320868165e-06,
+      "loss": 0.7679,
+      "step": 31236
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.480939207531976,
+      "learning_rate": 3.440023997849462e-06,
+      "loss": 0.759,
+      "step": 31237
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3219870654091965,
+      "learning_rate": 3.4394481018063685e-06,
+      "loss": 0.7508,
+      "step": 31238
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.370426833189612,
+      "learning_rate": 3.4388722439608822e-06,
+      "loss": 0.7255,
+      "step": 31239
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3165995316949737,
+      "learning_rate": 3.438296424316362e-06,
+      "loss": 0.614,
+      "step": 31240
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4564817187342638,
+      "learning_rate": 3.4377206428761534e-06,
+      "loss": 0.6014,
+      "step": 31241
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6230485129577756,
+      "learning_rate": 3.4371448996436163e-06,
+      "loss": 0.5078,
+      "step": 31242
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6233843879689025,
+      "learning_rate": 3.436569194622096e-06,
+      "loss": 0.499,
+      "step": 31243
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3064318448058732,
+      "learning_rate": 3.4359935278149513e-06,
+      "loss": 0.7274,
+      "step": 31244
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4774319131103728,
+      "learning_rate": 3.4354178992255306e-06,
+      "loss": 0.6792,
+      "step": 31245
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3001470013945444,
+      "learning_rate": 3.4348423088571815e-06,
+      "loss": 0.6982,
+      "step": 31246
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4324569804649823,
+      "learning_rate": 3.4342667567132593e-06,
+      "loss": 0.7495,
+      "step": 31247
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5079326238598796,
+      "learning_rate": 3.433691242797118e-06,
+      "loss": 0.6804,
+      "step": 31248
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5692513525452589,
+      "learning_rate": 3.433115767112105e-06,
+      "loss": 0.5217,
+      "step": 31249
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5760849087962612,
+      "learning_rate": 3.4325403296615677e-06,
+      "loss": 0.6086,
+      "step": 31250
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7728723287571135,
+      "learning_rate": 3.43196493044886e-06,
+      "loss": 0.6797,
+      "step": 31251
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5694595494195624,
+      "learning_rate": 3.4313895694773357e-06,
+      "loss": 0.6618,
+      "step": 31252
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2293239013890316,
+      "learning_rate": 3.4308142467503404e-06,
+      "loss": 0.6618,
+      "step": 31253
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4894218109853692,
+      "learning_rate": 3.4302389622712207e-06,
+      "loss": 0.6464,
+      "step": 31254
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5929752211128123,
+      "learning_rate": 3.42966371604333e-06,
+      "loss": 0.5917,
+      "step": 31255
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.817548471592082,
+      "learning_rate": 3.429088508070021e-06,
+      "loss": 0.6689,
+      "step": 31256
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6011923131546582,
+      "learning_rate": 3.428513338354639e-06,
+      "loss": 0.7135,
+      "step": 31257
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3410968691202292,
+      "learning_rate": 3.4279382069005283e-06,
+      "loss": 0.7604,
+      "step": 31258
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.950228143727505,
+      "learning_rate": 3.4273631137110464e-06,
+      "loss": 0.7519,
+      "step": 31259
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5470973537558779,
+      "learning_rate": 3.4267880587895343e-06,
+      "loss": 0.744,
+      "step": 31260
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5134132873093598,
+      "learning_rate": 3.4262130421393424e-06,
+      "loss": 0.6597,
+      "step": 31261
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.576179694415719,
+      "learning_rate": 3.425638063763823e-06,
+      "loss": 0.809,
+      "step": 31262
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.557449800124461,
+      "learning_rate": 3.42506312366632e-06,
+      "loss": 0.659,
+      "step": 31263
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4847232376491268,
+      "learning_rate": 3.4244882218501775e-06,
+      "loss": 0.7213,
+      "step": 31264
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2743745373049207,
+      "learning_rate": 3.423913358318747e-06,
+      "loss": 0.7551,
+      "step": 31265
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6141407999692796,
+      "learning_rate": 3.423338533075381e-06,
+      "loss": 0.7756,
+      "step": 31266
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6411035804368186,
+      "learning_rate": 3.422763746123414e-06,
+      "loss": 0.6774,
+      "step": 31267
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1626828477909052,
+      "learning_rate": 3.422188997466199e-06,
+      "loss": 0.662,
+      "step": 31268
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2528327995628812,
+      "learning_rate": 3.421614287107082e-06,
+      "loss": 0.6681,
+      "step": 31269
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4628850957038,
+      "learning_rate": 3.4210396150494152e-06,
+      "loss": 0.703,
+      "step": 31270
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8886129993787217,
+      "learning_rate": 3.4204649812965327e-06,
+      "loss": 0.739,
+      "step": 31271
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5611438253823626,
+      "learning_rate": 3.4198903858517863e-06,
+      "loss": 0.7633,
+      "step": 31272
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3942124440486436,
+      "learning_rate": 3.4193158287185248e-06,
+      "loss": 0.6117,
+      "step": 31273
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6452892360545828,
+      "learning_rate": 3.4187413099000854e-06,
+      "loss": 0.6967,
+      "step": 31274
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.501361007976705,
+      "learning_rate": 3.4181668293998216e-06,
+      "loss": 0.6858,
+      "step": 31275
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.403454875676846,
+      "learning_rate": 3.4175923872210715e-06,
+      "loss": 0.6942,
+      "step": 31276
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.565963953465843,
+      "learning_rate": 3.4170179833671847e-06,
+      "loss": 0.5038,
+      "step": 31277
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.32349128718697,
+      "learning_rate": 3.4164436178415007e-06,
+      "loss": 0.6693,
+      "step": 31278
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.233900814283461,
+      "learning_rate": 3.41586929064737e-06,
+      "loss": 0.6358,
+      "step": 31279
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4155794130008568,
+      "learning_rate": 3.4152950017881325e-06,
+      "loss": 0.6435,
+      "step": 31280
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5392302109010394,
+      "learning_rate": 3.414720751267129e-06,
+      "loss": 0.5824,
+      "step": 31281
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6384492013920715,
+      "learning_rate": 3.414146539087706e-06,
+      "loss": 0.7719,
+      "step": 31282
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4161680074514267,
+      "learning_rate": 3.4135723652532093e-06,
+      "loss": 0.7681,
+      "step": 31283
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4966997753995026,
+      "learning_rate": 3.4129982297669807e-06,
+      "loss": 0.6899,
+      "step": 31284
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4245153207063916,
+      "learning_rate": 3.412424132632358e-06,
+      "loss": 0.7476,
+      "step": 31285
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3552727859834526,
+      "learning_rate": 3.411850073852688e-06,
+      "loss": 0.8048,
+      "step": 31286
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4436500167304576,
+      "learning_rate": 3.411276053431315e-06,
+      "loss": 0.752,
+      "step": 31287
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4177371927557398,
+      "learning_rate": 3.41070207137158e-06,
+      "loss": 0.587,
+      "step": 31288
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4782567603892374,
+      "learning_rate": 3.4101281276768194e-06,
+      "loss": 0.6696,
+      "step": 31289
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5842418009568655,
+      "learning_rate": 3.4095542223503795e-06,
+      "loss": 0.6516,
+      "step": 31290
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4718728337428884,
+      "learning_rate": 3.4089803553956058e-06,
+      "loss": 0.5976,
+      "step": 31291
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.207550876043572,
+      "learning_rate": 3.4084065268158338e-06,
+      "loss": 0.5933,
+      "step": 31292
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4332459126082568,
+      "learning_rate": 3.4078327366144025e-06,
+      "loss": 0.7157,
+      "step": 31293
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3039591189904636,
+      "learning_rate": 3.4072589847946603e-06,
+      "loss": 0.6284,
+      "step": 31294
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3789686114902084,
+      "learning_rate": 3.40668527135994e-06,
+      "loss": 0.6072,
+      "step": 31295
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4916692622863517,
+      "learning_rate": 3.406111596313588e-06,
+      "loss": 0.6802,
+      "step": 31296
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5527634326203485,
+      "learning_rate": 3.4055379596589387e-06,
+      "loss": 0.7242,
+      "step": 31297
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6731348235800192,
+      "learning_rate": 3.404964361399339e-06,
+      "loss": 0.818,
+      "step": 31298
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6369576070136296,
+      "learning_rate": 3.4043908015381212e-06,
+      "loss": 0.7694,
+      "step": 31299
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7672828807429886,
+      "learning_rate": 3.4038172800786317e-06,
+      "loss": 0.8221,
+      "step": 31300
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3198729210756988,
+      "learning_rate": 3.403243797024206e-06,
+      "loss": 0.6799,
+      "step": 31301
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5947589256936797,
+      "learning_rate": 3.40267035237818e-06,
+      "loss": 0.5157,
+      "step": 31302
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4744793975937667,
+      "learning_rate": 3.4020969461438966e-06,
+      "loss": 0.8577,
+      "step": 31303
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.314461626505655,
+      "learning_rate": 3.401523578324697e-06,
+      "loss": 0.7115,
+      "step": 31304
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2239204864632636,
+      "learning_rate": 3.4009502489239157e-06,
+      "loss": 0.5157,
+      "step": 31305
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3258031374948793,
+      "learning_rate": 3.4003769579448875e-06,
+      "loss": 0.7032,
+      "step": 31306
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3846091078294098,
+      "learning_rate": 3.399803705390955e-06,
+      "loss": 0.6709,
+      "step": 31307
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2813179345577301,
+      "learning_rate": 3.3992304912654573e-06,
+      "loss": 0.7269,
+      "step": 31308
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1922438910352784,
+      "learning_rate": 3.3986573155717307e-06,
+      "loss": 0.5646,
+      "step": 31309
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3230037111959898,
+      "learning_rate": 3.3980841783131067e-06,
+      "loss": 0.7319,
+      "step": 31310
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2595450573350508,
+      "learning_rate": 3.3975110794929267e-06,
+      "loss": 0.684,
+      "step": 31311
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.42813717590374,
+      "learning_rate": 3.396938019114532e-06,
+      "loss": 0.6552,
+      "step": 31312
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5714450848174437,
+      "learning_rate": 3.396364997181254e-06,
+      "loss": 0.5162,
+      "step": 31313
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4053732410016493,
+      "learning_rate": 3.395792013696426e-06,
+      "loss": 0.8215,
+      "step": 31314
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5163434778363747,
+      "learning_rate": 3.3952190686633914e-06,
+      "loss": 0.6892,
+      "step": 31315
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3372658862733324,
+      "learning_rate": 3.3946461620854788e-06,
+      "loss": 0.7713,
+      "step": 31316
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4363154611807825,
+      "learning_rate": 3.3940732939660303e-06,
+      "loss": 0.5356,
+      "step": 31317
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3340020637719152,
+      "learning_rate": 3.393500464308377e-06,
+      "loss": 0.738,
+      "step": 31318
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1670521174651003,
+      "learning_rate": 3.392927673115857e-06,
+      "loss": 0.6989,
+      "step": 31319
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7303106072953183,
+      "learning_rate": 3.3923549203918017e-06,
+      "loss": 0.7247,
+      "step": 31320
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9974807152157592,
+      "learning_rate": 3.391782206139551e-06,
+      "loss": 0.6112,
+      "step": 31321
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4597889469978977,
+      "learning_rate": 3.391209530362436e-06,
+      "loss": 0.731,
+      "step": 31322
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7315269387839596,
+      "learning_rate": 3.390636893063789e-06,
+      "loss": 0.6221,
+      "step": 31323
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3209484337979664,
+      "learning_rate": 3.390064294246945e-06,
+      "loss": 0.6605,
+      "step": 31324
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7903830725101695,
+      "learning_rate": 3.3894917339152444e-06,
+      "loss": 0.7106,
+      "step": 31325
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4168817150115052,
+      "learning_rate": 3.388919212072015e-06,
+      "loss": 0.708,
+      "step": 31326
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3473153868643897,
+      "learning_rate": 3.3883467287205874e-06,
+      "loss": 0.5984,
+      "step": 31327
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2877434642706476,
+      "learning_rate": 3.387774283864298e-06,
+      "loss": 0.725,
+      "step": 31328
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3841216202848519,
+      "learning_rate": 3.387201877506484e-06,
+      "loss": 0.7281,
+      "step": 31329
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3918966643410697,
+      "learning_rate": 3.3866295096504744e-06,
+      "loss": 0.6378,
+      "step": 31330
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6209410970857168,
+      "learning_rate": 3.386057180299599e-06,
+      "loss": 0.6599,
+      "step": 31331
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2669013575184498,
+      "learning_rate": 3.3854848894571913e-06,
+      "loss": 0.7803,
+      "step": 31332
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3205569685662106,
+      "learning_rate": 3.384912637126588e-06,
+      "loss": 0.7242,
+      "step": 31333
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4096495650506087,
+      "learning_rate": 3.3843404233111187e-06,
+      "loss": 0.5945,
+      "step": 31334
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6298131622735756,
+      "learning_rate": 3.38376824801411e-06,
+      "loss": 0.5056,
+      "step": 31335
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4479795285652577,
+      "learning_rate": 3.3831961112389e-06,
+      "loss": 0.521,
+      "step": 31336
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4924767442884976,
+      "learning_rate": 3.3826240129888143e-06,
+      "loss": 0.6421,
+      "step": 31337
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5343640760507207,
+      "learning_rate": 3.3820519532671905e-06,
+      "loss": 0.7947,
+      "step": 31338
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8293502647241082,
+      "learning_rate": 3.3814799320773517e-06,
+      "loss": 0.6167,
+      "step": 31339
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6034842603443937,
+      "learning_rate": 3.3809079494226347e-06,
+      "loss": 0.6833,
+      "step": 31340
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5493925450486543,
+      "learning_rate": 3.380336005306365e-06,
+      "loss": 0.7007,
+      "step": 31341
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4673069273345896,
+      "learning_rate": 3.379764099731877e-06,
+      "loss": 0.8249,
+      "step": 31342
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4708026124336717,
+      "learning_rate": 3.3791922327024994e-06,
+      "loss": 0.7013,
+      "step": 31343
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7239737885968829,
+      "learning_rate": 3.3786204042215576e-06,
+      "loss": 0.6801,
+      "step": 31344
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2812737593064285,
+      "learning_rate": 3.378048614292383e-06,
+      "loss": 0.6273,
+      "step": 31345
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3810484941683903,
+      "learning_rate": 3.3774768629183065e-06,
+      "loss": 0.5733,
+      "step": 31346
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6909587752015725,
+      "learning_rate": 3.376905150102663e-06,
+      "loss": 0.6909,
+      "step": 31347
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3385424817995561,
+      "learning_rate": 3.3763334758487674e-06,
+      "loss": 0.698,
+      "step": 31348
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4766411098407548,
+      "learning_rate": 3.375761840159956e-06,
+      "loss": 0.6957,
+      "step": 31349
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.799734843496735,
+      "learning_rate": 3.3751902430395558e-06,
+      "loss": 0.6742,
+      "step": 31350
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4484351851509825,
+      "learning_rate": 3.3746186844909025e-06,
+      "loss": 0.7237,
+      "step": 31351
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.921954701417891,
+      "learning_rate": 3.374047164517309e-06,
+      "loss": 0.8221,
+      "step": 31352
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.319368232167498,
+      "learning_rate": 3.3734756831221116e-06,
+      "loss": 0.6401,
+      "step": 31353
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4275461939964862,
+      "learning_rate": 3.37290424030864e-06,
+      "loss": 0.7815,
+      "step": 31354
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.422946971252235,
+      "learning_rate": 3.3723328360802187e-06,
+      "loss": 0.5984,
+      "step": 31355
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.409491394425776,
+      "learning_rate": 3.371761470440169e-06,
+      "loss": 0.657,
+      "step": 31356
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4206504184193367,
+      "learning_rate": 3.371190143391824e-06,
+      "loss": 0.6398,
+      "step": 31357
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3085312058075196,
+      "learning_rate": 3.370618854938511e-06,
+      "loss": 0.6607,
+      "step": 31358
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6771872562183179,
+      "learning_rate": 3.3700476050835506e-06,
+      "loss": 0.6523,
+      "step": 31359
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2502241087645656,
+      "learning_rate": 3.3694763938302764e-06,
+      "loss": 0.6212,
+      "step": 31360
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3941827390402428,
+      "learning_rate": 3.368905221182008e-06,
+      "loss": 0.6931,
+      "step": 31361
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3036967240699064,
+      "learning_rate": 3.3683340871420712e-06,
+      "loss": 0.751,
+      "step": 31362
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5494401373959883,
+      "learning_rate": 3.367762991713792e-06,
+      "loss": 0.7428,
+      "step": 31363
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8394736137232124,
+      "learning_rate": 3.3671919349005e-06,
+      "loss": 0.7653,
+      "step": 31364
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4300889560346746,
+      "learning_rate": 3.3666209167055163e-06,
+      "loss": 0.6931,
+      "step": 31365
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3666070084889597,
+      "learning_rate": 3.366049937132162e-06,
+      "loss": 0.6516,
+      "step": 31366
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2811046834720374,
+      "learning_rate": 3.3654789961837666e-06,
+      "loss": 0.5706,
+      "step": 31367
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4426305098404248,
+      "learning_rate": 3.364908093863658e-06,
+      "loss": 0.7251,
+      "step": 31368
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3678158366949928,
+      "learning_rate": 3.3643372301751498e-06,
+      "loss": 0.6937,
+      "step": 31369
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2518339034084824,
+      "learning_rate": 3.3637664051215703e-06,
+      "loss": 0.5955,
+      "step": 31370
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4466781444204642,
+      "learning_rate": 3.363195618706243e-06,
+      "loss": 0.7363,
+      "step": 31371
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2838274431917525,
+      "learning_rate": 3.3626248709324973e-06,
+      "loss": 0.701,
+      "step": 31372
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5012418620327916,
+      "learning_rate": 3.3620541618036496e-06,
+      "loss": 0.6723,
+      "step": 31373
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2535058067810034,
+      "learning_rate": 3.361483491323021e-06,
+      "loss": 0.6118,
+      "step": 31374
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6236891579680796,
+      "learning_rate": 3.36091285949394e-06,
+      "loss": 0.4974,
+      "step": 31375
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.527655960270509,
+      "learning_rate": 3.3603422663197236e-06,
+      "loss": 0.635,
+      "step": 31376
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3016932607012903,
+      "learning_rate": 3.3597717118037e-06,
+      "loss": 0.7477,
+      "step": 31377
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6233019592841682,
+      "learning_rate": 3.3592011959491844e-06,
+      "loss": 0.7624,
+      "step": 31378
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3107251947246752,
+      "learning_rate": 3.3586307187595045e-06,
+      "loss": 0.6033,
+      "step": 31379
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.617505349189157,
+      "learning_rate": 3.358060280237977e-06,
+      "loss": 0.7801,
+      "step": 31380
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2527459408157884,
+      "learning_rate": 3.3574898803879286e-06,
+      "loss": 0.6802,
+      "step": 31381
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8113458240455467,
+      "learning_rate": 3.3569195192126758e-06,
+      "loss": 0.6859,
+      "step": 31382
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1787164951879263,
+      "learning_rate": 3.356349196715538e-06,
+      "loss": 0.5234,
+      "step": 31383
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4099287030866032,
+      "learning_rate": 3.35577891289984e-06,
+      "loss": 0.6734,
+      "step": 31384
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2501536889166251,
+      "learning_rate": 3.355208667768902e-06,
+      "loss": 0.6777,
+      "step": 31385
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8124685218907275,
+      "learning_rate": 3.354638461326043e-06,
+      "loss": 0.6822,
+      "step": 31386
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4512253099111676,
+      "learning_rate": 3.3540682935745804e-06,
+      "loss": 0.7508,
+      "step": 31387
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2896834881691883,
+      "learning_rate": 3.3534981645178356e-06,
+      "loss": 0.7106,
+      "step": 31388
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5519648253817946,
+      "learning_rate": 3.352928074159133e-06,
+      "loss": 0.8965,
+      "step": 31389
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5259078062849267,
+      "learning_rate": 3.3523580225017872e-06,
+      "loss": 0.6445,
+      "step": 31390
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.528107125456284,
+      "learning_rate": 3.351788009549114e-06,
+      "loss": 0.7297,
+      "step": 31391
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2139750076937246,
+      "learning_rate": 3.3512180353044356e-06,
+      "loss": 0.6965,
+      "step": 31392
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4197968357776034,
+      "learning_rate": 3.3506480997710743e-06,
+      "loss": 0.7989,
+      "step": 31393
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4017657513490496,
+      "learning_rate": 3.350078202952345e-06,
+      "loss": 0.6801,
+      "step": 31394
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5821403582014361,
+      "learning_rate": 3.349508344851562e-06,
+      "loss": 0.5151,
+      "step": 31395
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3652451564272037,
+      "learning_rate": 3.3489385254720507e-06,
+      "loss": 0.5699,
+      "step": 31396
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5627713900834015,
+      "learning_rate": 3.348368744817121e-06,
+      "loss": 0.776,
+      "step": 31397
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3620311720452833,
+      "learning_rate": 3.3477990028900986e-06,
+      "loss": 0.8058,
+      "step": 31398
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.831661039862546,
+      "learning_rate": 3.347229299694292e-06,
+      "loss": 0.6765,
+      "step": 31399
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2006622358572012,
+      "learning_rate": 3.346659635233027e-06,
+      "loss": 0.7271,
+      "step": 31400
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4632719189824461,
+      "learning_rate": 3.346090009509613e-06,
+      "loss": 0.7784,
+      "step": 31401
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7056292134922642,
+      "learning_rate": 3.3455204225273718e-06,
+      "loss": 0.7319,
+      "step": 31402
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4409141543405581,
+      "learning_rate": 3.3449508742896176e-06,
+      "loss": 0.5959,
+      "step": 31403
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5124061271804032,
+      "learning_rate": 3.3443813647996638e-06,
+      "loss": 0.688,
+      "step": 31404
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4672996867391854,
+      "learning_rate": 3.343811894060829e-06,
+      "loss": 0.701,
+      "step": 31405
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2286360383437118,
+      "learning_rate": 3.3432424620764304e-06,
+      "loss": 0.7204,
+      "step": 31406
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2141355649798373,
+      "learning_rate": 3.342673068849783e-06,
+      "loss": 0.7444,
+      "step": 31407
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.421052930382724,
+      "learning_rate": 3.342103714384197e-06,
+      "loss": 0.6964,
+      "step": 31408
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5992210090715192,
+      "learning_rate": 3.3415343986829908e-06,
+      "loss": 0.4898,
+      "step": 31409
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2619384882890095,
+      "learning_rate": 3.340965121749483e-06,
+      "loss": 0.7836,
+      "step": 31410
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.081620240193606,
+      "learning_rate": 3.3403958835869843e-06,
+      "loss": 0.7114,
+      "step": 31411
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.369035957708814,
+      "learning_rate": 3.339826684198806e-06,
+      "loss": 0.6676,
+      "step": 31412
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.576255993882755,
+      "learning_rate": 3.3392575235882654e-06,
+      "loss": 0.5282,
+      "step": 31413
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2874552193321347,
+      "learning_rate": 3.3386884017586794e-06,
+      "loss": 0.7026,
+      "step": 31414
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3444377600341793,
+      "learning_rate": 3.3381193187133587e-06,
+      "loss": 0.6366,
+      "step": 31415
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6338625759394266,
+      "learning_rate": 3.3375502744556134e-06,
+      "loss": 0.5465,
+      "step": 31416
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.368670453734911,
+      "learning_rate": 3.3369812689887617e-06,
+      "loss": 0.6067,
+      "step": 31417
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5281390502818166,
+      "learning_rate": 3.336412302316112e-06,
+      "loss": 0.8046,
+      "step": 31418
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5137457803307732,
+      "learning_rate": 3.3358433744409835e-06,
+      "loss": 0.7418,
+      "step": 31419
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3960046541828641,
+      "learning_rate": 3.3352744853666807e-06,
+      "loss": 0.7306,
+      "step": 31420
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6388745773690389,
+      "learning_rate": 3.3347056350965235e-06,
+      "loss": 0.5359,
+      "step": 31421
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3316976168832693,
+      "learning_rate": 3.334136823633817e-06,
+      "loss": 0.6252,
+      "step": 31422
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6538284939616639,
+      "learning_rate": 3.333568050981879e-06,
+      "loss": 0.6112,
+      "step": 31423
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3342053021569344,
+      "learning_rate": 3.332999317144019e-06,
+      "loss": 0.6451,
+      "step": 31424
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3370828109184079,
+      "learning_rate": 3.3324306221235435e-06,
+      "loss": 0.6427,
+      "step": 31425
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1928687389860362,
+      "learning_rate": 3.3318619659237684e-06,
+      "loss": 0.5454,
+      "step": 31426
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6869607912034967,
+      "learning_rate": 3.331293348548007e-06,
+      "loss": 0.5916,
+      "step": 31427
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3085192524375682,
+      "learning_rate": 3.330724769999568e-06,
+      "loss": 0.7426,
+      "step": 31428
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2569110339596208,
+      "learning_rate": 3.3301562302817568e-06,
+      "loss": 0.65,
+      "step": 31429
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3116342762079674,
+      "learning_rate": 3.3295877293978875e-06,
+      "loss": 0.6288,
+      "step": 31430
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3124653774714132,
+      "learning_rate": 3.3290192673512747e-06,
+      "loss": 0.6282,
+      "step": 31431
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5389617063514232,
+      "learning_rate": 3.3284508441452225e-06,
+      "loss": 0.704,
+      "step": 31432
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.32503069976697,
+      "learning_rate": 3.3278824597830383e-06,
+      "loss": 0.7869,
+      "step": 31433
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2817093898718643,
+      "learning_rate": 3.3273141142680356e-06,
+      "loss": 0.7018,
+      "step": 31434
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4987308456859276,
+      "learning_rate": 3.3267458076035255e-06,
+      "loss": 0.6694,
+      "step": 31435
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4503608007123197,
+      "learning_rate": 3.3261775397928143e-06,
+      "loss": 0.7158,
+      "step": 31436
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6319191941473534,
+      "learning_rate": 3.3256093108392064e-06,
+      "loss": 0.6875,
+      "step": 31437
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3038893775318992,
+      "learning_rate": 3.3250411207460188e-06,
+      "loss": 0.7162,
+      "step": 31438
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3408391598076985,
+      "learning_rate": 3.32447296951655e-06,
+      "loss": 0.7095,
+      "step": 31439
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4030425370489261,
+      "learning_rate": 3.323904857154118e-06,
+      "loss": 0.654,
+      "step": 31440
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6615453894085391,
+      "learning_rate": 3.323336783662021e-06,
+      "loss": 0.5263,
+      "step": 31441
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6177790674048753,
+      "learning_rate": 3.3227687490435746e-06,
+      "loss": 0.5099,
+      "step": 31442
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.762582651456709,
+      "learning_rate": 3.3222007533020793e-06,
+      "loss": 0.7717,
+      "step": 31443
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6685069510503439,
+      "learning_rate": 3.3216327964408456e-06,
+      "loss": 0.6211,
+      "step": 31444
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4228989549306066,
+      "learning_rate": 3.321064878463187e-06,
+      "loss": 0.7226,
+      "step": 31445
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5759371891883952,
+      "learning_rate": 3.3204969993723955e-06,
+      "loss": 0.4772,
+      "step": 31446
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3584360317310447,
+      "learning_rate": 3.3199291591717865e-06,
+      "loss": 0.7353,
+      "step": 31447
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6055703852377656,
+      "learning_rate": 3.3193613578646633e-06,
+      "loss": 0.7881,
+      "step": 31448
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4655493118079717,
+      "learning_rate": 3.3187935954543417e-06,
+      "loss": 0.7027,
+      "step": 31449
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.412989600412758,
+      "learning_rate": 3.3182258719441107e-06,
+      "loss": 0.8811,
+      "step": 31450
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7419794288839225,
+      "learning_rate": 3.3176581873372848e-06,
+      "loss": 0.737,
+      "step": 31451
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5620111233648846,
+      "learning_rate": 3.3170905416371723e-06,
+      "loss": 0.6857,
+      "step": 31452
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2119597406625147,
+      "learning_rate": 3.316522934847074e-06,
+      "loss": 0.677,
+      "step": 31453
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1942269873997473,
+      "learning_rate": 3.3159553669702917e-06,
+      "loss": 0.6527,
+      "step": 31454
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.558302647556335,
+      "learning_rate": 3.3153878380101332e-06,
+      "loss": 0.7943,
+      "step": 31455
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4040475724333004,
+      "learning_rate": 3.314820347969907e-06,
+      "loss": 0.7534,
+      "step": 31456
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.385139561133872,
+      "learning_rate": 3.3142528968529097e-06,
+      "loss": 0.5716,
+      "step": 31457
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3761490297291221,
+      "learning_rate": 3.313685484662452e-06,
+      "loss": 0.6185,
+      "step": 31458
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.308272366437807,
+      "learning_rate": 3.313118111401834e-06,
+      "loss": 0.6548,
+      "step": 31459
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4461518814581913,
+      "learning_rate": 3.312550777074356e-06,
+      "loss": 0.7481,
+      "step": 31460
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3727921309189677,
+      "learning_rate": 3.311983481683325e-06,
+      "loss": 0.726,
+      "step": 31461
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.020369898316959,
+      "learning_rate": 3.311416225232047e-06,
+      "loss": 0.7593,
+      "step": 31462
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5702619428266893,
+      "learning_rate": 3.3108490077238207e-06,
+      "loss": 0.6269,
+      "step": 31463
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.507806249808865,
+      "learning_rate": 3.3102818291619467e-06,
+      "loss": 0.7297,
+      "step": 31464
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2339656754189632,
+      "learning_rate": 3.309714689549729e-06,
+      "loss": 0.5797,
+      "step": 31465
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5897344871513448,
+      "learning_rate": 3.3091475888904777e-06,
+      "loss": 0.8014,
+      "step": 31466
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5358582868679647,
+      "learning_rate": 3.3085805271874803e-06,
+      "loss": 0.6153,
+      "step": 31467
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1667173588660251,
+      "learning_rate": 3.3080135044440454e-06,
+      "loss": 0.6569,
+      "step": 31468
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3583439232878183,
+      "learning_rate": 3.3074465206634755e-06,
+      "loss": 0.7529,
+      "step": 31469
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6265509310998606,
+      "learning_rate": 3.3068795758490736e-06,
+      "loss": 0.5005,
+      "step": 31470
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2779963892907935,
+      "learning_rate": 3.306312670004137e-06,
+      "loss": 0.7009,
+      "step": 31471
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3462112286264145,
+      "learning_rate": 3.3057458031319654e-06,
+      "loss": 0.6713,
+      "step": 31472
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.49340051279057,
+      "learning_rate": 3.30517897523586e-06,
+      "loss": 0.595,
+      "step": 31473
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3709984470755563,
+      "learning_rate": 3.304612186319126e-06,
+      "loss": 0.6135,
+      "step": 31474
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5689178087547981,
+      "learning_rate": 3.304045436385059e-06,
+      "loss": 0.5586,
+      "step": 31475
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5857786575953705,
+      "learning_rate": 3.303478725436956e-06,
+      "loss": 0.6964,
+      "step": 31476
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.275140545881827,
+      "learning_rate": 3.3029120534781244e-06,
+      "loss": 0.6436,
+      "step": 31477
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3482018034641703,
+      "learning_rate": 3.3023454205118545e-06,
+      "loss": 0.6632,
+      "step": 31478
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6360497682592018,
+      "learning_rate": 3.301778826541454e-06,
+      "loss": 0.5529,
+      "step": 31479
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.350803516006737,
+      "learning_rate": 3.3012122715702145e-06,
+      "loss": 0.6539,
+      "step": 31480
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6167522881012923,
+      "learning_rate": 3.3006457556014415e-06,
+      "loss": 0.8595,
+      "step": 31481
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4106444960731497,
+      "learning_rate": 3.300079278638426e-06,
+      "loss": 0.7402,
+      "step": 31482
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.402211917546993,
+      "learning_rate": 3.2995128406844733e-06,
+      "loss": 0.6235,
+      "step": 31483
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.737534608560989,
+      "learning_rate": 3.2989464417428785e-06,
+      "loss": 0.5594,
+      "step": 31484
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3350374610975413,
+      "learning_rate": 3.2983800818169363e-06,
+      "loss": 0.7533,
+      "step": 31485
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3165378276196933,
+      "learning_rate": 3.297813760909946e-06,
+      "loss": 0.5986,
+      "step": 31486
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5152482649575136,
+      "learning_rate": 3.2972474790252097e-06,
+      "loss": 0.7769,
+      "step": 31487
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6585010843853847,
+      "learning_rate": 3.2966812361660204e-06,
+      "loss": 0.707,
+      "step": 31488
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3521623944771588,
+      "learning_rate": 3.2961150323356715e-06,
+      "loss": 0.6364,
+      "step": 31489
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5375079193686776,
+      "learning_rate": 3.2955488675374635e-06,
+      "loss": 0.604,
+      "step": 31490
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3848401501224792,
+      "learning_rate": 3.294982741774697e-06,
+      "loss": 0.6932,
+      "step": 31491
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6480559462815605,
+      "learning_rate": 3.294416655050663e-06,
+      "loss": 0.5109,
+      "step": 31492
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2518273653550747,
+      "learning_rate": 3.293850607368655e-06,
+      "loss": 0.6354,
+      "step": 31493
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4855128259862165,
+      "learning_rate": 3.2932845987319704e-06,
+      "loss": 0.6644,
+      "step": 31494
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3046569253031377,
+      "learning_rate": 3.2927186291439117e-06,
+      "loss": 0.7079,
+      "step": 31495
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.239828227297956,
+      "learning_rate": 3.292152698607768e-06,
+      "loss": 0.6057,
+      "step": 31496
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2962268451488166,
+      "learning_rate": 3.291586807126831e-06,
+      "loss": 0.7568,
+      "step": 31497
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4163733805184844,
+      "learning_rate": 3.2910209547044027e-06,
+      "loss": 0.6743,
+      "step": 31498
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5364681561516023,
+      "learning_rate": 3.290455141343771e-06,
+      "loss": 0.7078,
+      "step": 31499
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3951040381356192,
+      "learning_rate": 3.289889367048238e-06,
+      "loss": 0.6684,
+      "step": 31500
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2892187555301349,
+      "learning_rate": 3.28932363182109e-06,
+      "loss": 0.7058,
+      "step": 31501
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1590692685783568,
+      "learning_rate": 3.2887579356656274e-06,
+      "loss": 0.6937,
+      "step": 31502
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.178434588243278,
+      "learning_rate": 3.2881922785851374e-06,
+      "loss": 0.615,
+      "step": 31503
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3205001648723376,
+      "learning_rate": 3.28762666058292e-06,
+      "loss": 0.6644,
+      "step": 31504
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.2883402080977806,
+      "learning_rate": 3.2870610816622663e-06,
+      "loss": 0.6399,
+      "step": 31505
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4762017039815054,
+      "learning_rate": 3.286495541826463e-06,
+      "loss": 0.6624,
+      "step": 31506
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.155250254024494,
+      "learning_rate": 3.28593004107881e-06,
+      "loss": 0.7566,
+      "step": 31507
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6019780643128922,
+      "learning_rate": 3.2853645794226007e-06,
+      "loss": 0.6099,
+      "step": 31508
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.218487513521963,
+      "learning_rate": 3.284799156861125e-06,
+      "loss": 0.5363,
+      "step": 31509
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6189437812056877,
+      "learning_rate": 3.2842337733976704e-06,
+      "loss": 0.7156,
+      "step": 31510
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2352491042041696,
+      "learning_rate": 3.2836684290355336e-06,
+      "loss": 0.624,
+      "step": 31511
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4075277698536663,
+      "learning_rate": 3.2831031237780086e-06,
+      "loss": 0.7674,
+      "step": 31512
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4318821678662454,
+      "learning_rate": 3.2825378576283827e-06,
+      "loss": 0.6207,
+      "step": 31513
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.356367203792497,
+      "learning_rate": 3.2819726305899467e-06,
+      "loss": 0.6629,
+      "step": 31514
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4584085037506322,
+      "learning_rate": 3.2814074426659925e-06,
+      "loss": 0.6647,
+      "step": 31515
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3928033533972932,
+      "learning_rate": 3.280842293859814e-06,
+      "loss": 0.7408,
+      "step": 31516
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.325786745245334,
+      "learning_rate": 3.2802771841746984e-06,
+      "loss": 0.6681,
+      "step": 31517
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4955159499550883,
+      "learning_rate": 3.2797121136139345e-06,
+      "loss": 0.7198,
+      "step": 31518
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4331423702618904,
+      "learning_rate": 3.279147082180817e-06,
+      "loss": 0.7085,
+      "step": 31519
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.430416731664828,
+      "learning_rate": 3.278582089878629e-06,
+      "loss": 0.6516,
+      "step": 31520
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6499109404496017,
+      "learning_rate": 3.2780171367106684e-06,
+      "loss": 0.6319,
+      "step": 31521
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6554611954447092,
+      "learning_rate": 3.2774522226802162e-06,
+      "loss": 0.5366,
+      "step": 31522
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6165926526013527,
+      "learning_rate": 3.2768873477905696e-06,
+      "loss": 0.4933,
+      "step": 31523
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4382864013213903,
+      "learning_rate": 3.2763225120450094e-06,
+      "loss": 0.7853,
+      "step": 31524
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3580342067318314,
+      "learning_rate": 3.275757715446831e-06,
+      "loss": 0.547,
+      "step": 31525
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.499584018101048,
+      "learning_rate": 3.2751929579993203e-06,
+      "loss": 0.7069,
+      "step": 31526
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4000701152208428,
+      "learning_rate": 3.2746282397057627e-06,
+      "loss": 0.6213,
+      "step": 31527
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5179592893199076,
+      "learning_rate": 3.274063560569448e-06,
+      "loss": 0.7433,
+      "step": 31528
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.703285508642069,
+      "learning_rate": 3.2734989205936684e-06,
+      "loss": 0.7324,
+      "step": 31529
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4898138233750804,
+      "learning_rate": 3.2729343197817078e-06,
+      "loss": 0.6446,
+      "step": 31530
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3920101107312828,
+      "learning_rate": 3.2723697581368496e-06,
+      "loss": 0.7673,
+      "step": 31531
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3077005371317543,
+      "learning_rate": 3.271805235662384e-06,
+      "loss": 0.6827,
+      "step": 31532
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5202492010197601,
+      "learning_rate": 3.271240752361603e-06,
+      "loss": 0.5906,
+      "step": 31533
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6239387908431131,
+      "learning_rate": 3.270676308237788e-06,
+      "loss": 0.5358,
+      "step": 31534
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3355998070718342,
+      "learning_rate": 3.270111903294222e-06,
+      "loss": 0.5566,
+      "step": 31535
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2592834741502168,
+      "learning_rate": 3.2695475375341967e-06,
+      "loss": 0.6157,
+      "step": 31536
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3337025616562577,
+      "learning_rate": 3.2689832109609985e-06,
+      "loss": 0.6732,
+      "step": 31537
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4803870009850824,
+      "learning_rate": 3.268418923577912e-06,
+      "loss": 0.6469,
+      "step": 31538
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.712972267391539,
+      "learning_rate": 3.2678546753882177e-06,
+      "loss": 0.7914,
+      "step": 31539
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5122599998076844,
+      "learning_rate": 3.2672904663952078e-06,
+      "loss": 0.7106,
+      "step": 31540
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5804102201255987,
+      "learning_rate": 3.2667262966021617e-06,
+      "loss": 0.4844,
+      "step": 31541
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3270660104993768,
+      "learning_rate": 3.2661621660123666e-06,
+      "loss": 0.6673,
+      "step": 31542
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5959679268592278,
+      "learning_rate": 3.2655980746291114e-06,
+      "loss": 0.5373,
+      "step": 31543
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.429275957863397,
+      "learning_rate": 3.2650340224556754e-06,
+      "loss": 0.6492,
+      "step": 31544
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8065802694642434,
+      "learning_rate": 3.264470009495341e-06,
+      "loss": 0.5693,
+      "step": 31545
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.52118427398748,
+      "learning_rate": 3.263906035751393e-06,
+      "loss": 0.5895,
+      "step": 31546
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2826475816389926,
+      "learning_rate": 3.263342101227125e-06,
+      "loss": 0.6114,
+      "step": 31547
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4257607272484556,
+      "learning_rate": 3.2627782059258053e-06,
+      "loss": 0.5691,
+      "step": 31548
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.419299558248042,
+      "learning_rate": 3.262214349850723e-06,
+      "loss": 0.6068,
+      "step": 31549
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6650702229311534,
+      "learning_rate": 3.261650533005162e-06,
+      "loss": 0.596,
+      "step": 31550
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3033593874864906,
+      "learning_rate": 3.261086755392412e-06,
+      "loss": 0.6774,
+      "step": 31551
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.380649469298555,
+      "learning_rate": 3.2605230170157407e-06,
+      "loss": 0.7247,
+      "step": 31552
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.315915041462008,
+      "learning_rate": 3.2599593178784395e-06,
+      "loss": 0.6994,
+      "step": 31553
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.506025719491397,
+      "learning_rate": 3.2593956579837914e-06,
+      "loss": 0.7825,
+      "step": 31554
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3674143588615104,
+      "learning_rate": 3.2588320373350736e-06,
+      "loss": 0.7011,
+      "step": 31555
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6441325523028343,
+      "learning_rate": 3.2582684559355727e-06,
+      "loss": 0.8002,
+      "step": 31556
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.571585974721018,
+      "learning_rate": 3.257704913788564e-06,
+      "loss": 0.749,
+      "step": 31557
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5118812727530995,
+      "learning_rate": 3.257141410897334e-06,
+      "loss": 0.7207,
+      "step": 31558
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3508344112776949,
+      "learning_rate": 3.2565779472651595e-06,
+      "loss": 0.7295,
+      "step": 31559
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2639498275452603,
+      "learning_rate": 3.2560145228953267e-06,
+      "loss": 0.6009,
+      "step": 31560
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.66443596223875,
+      "learning_rate": 3.2554511377911115e-06,
+      "loss": 0.7683,
+      "step": 31561
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6151659223294975,
+      "learning_rate": 3.254887791955792e-06,
+      "loss": 0.5413,
+      "step": 31562
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2791930624171899,
+      "learning_rate": 3.2543244853926515e-06,
+      "loss": 0.738,
+      "step": 31563
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4920165941196488,
+      "learning_rate": 3.2537612181049726e-06,
+      "loss": 0.6189,
+      "step": 31564
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.549236543885307,
+      "learning_rate": 3.253197990096031e-06,
+      "loss": 0.7647,
+      "step": 31565
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4418623132070132,
+      "learning_rate": 3.2526348013691043e-06,
+      "loss": 0.7371,
+      "step": 31566
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2363686195253276,
+      "learning_rate": 3.2520716519274743e-06,
+      "loss": 0.6088,
+      "step": 31567
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3109670263485864,
+      "learning_rate": 3.251508541774422e-06,
+      "loss": 0.713,
+      "step": 31568
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6592621837502888,
+      "learning_rate": 3.250945470913224e-06,
+      "loss": 0.5367,
+      "step": 31569
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4618460921747014,
+      "learning_rate": 3.250382439347154e-06,
+      "loss": 0.6499,
+      "step": 31570
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3775865580763564,
+      "learning_rate": 3.2498194470794954e-06,
+      "loss": 0.6699,
+      "step": 31571
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3783901033919246,
+      "learning_rate": 3.2492564941135273e-06,
+      "loss": 0.6771,
+      "step": 31572
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4634341902987162,
+      "learning_rate": 3.2486935804525264e-06,
+      "loss": 0.5712,
+      "step": 31573
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4206648973151261,
+      "learning_rate": 3.2481307060997645e-06,
+      "loss": 0.7246,
+      "step": 31574
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5783789669248764,
+      "learning_rate": 3.2475678710585267e-06,
+      "loss": 0.5938,
+      "step": 31575
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3714644471465662,
+      "learning_rate": 3.247005075332084e-06,
+      "loss": 0.7943,
+      "step": 31576
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3964264121303793,
+      "learning_rate": 3.246442318923718e-06,
+      "loss": 0.8228,
+      "step": 31577
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3170751884485778,
+      "learning_rate": 3.245879601836701e-06,
+      "loss": 0.7028,
+      "step": 31578
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2162936483558098,
+      "learning_rate": 3.2453169240743132e-06,
+      "loss": 0.652,
+      "step": 31579
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3348620176161068,
+      "learning_rate": 3.244754285639826e-06,
+      "loss": 0.6938,
+      "step": 31580
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4541013737386026,
+      "learning_rate": 3.244191686536522e-06,
+      "loss": 0.7189,
+      "step": 31581
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.108582956737241,
+      "learning_rate": 3.2436291267676725e-06,
+      "loss": 0.612,
+      "step": 31582
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7980588188260025,
+      "learning_rate": 3.243066606336549e-06,
+      "loss": 0.7296,
+      "step": 31583
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3441133966707475,
+      "learning_rate": 3.242504125246432e-06,
+      "loss": 0.6067,
+      "step": 31584
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.719810994493711,
+      "learning_rate": 3.241941683500598e-06,
+      "loss": 0.5583,
+      "step": 31585
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3387721607487786,
+      "learning_rate": 3.241379281102319e-06,
+      "loss": 0.7739,
+      "step": 31586
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3252428418661442,
+      "learning_rate": 3.2408169180548675e-06,
+      "loss": 0.6808,
+      "step": 31587
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5846794967286836,
+      "learning_rate": 3.2402545943615195e-06,
+      "loss": 0.5159,
+      "step": 31588
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4199683842724589,
+      "learning_rate": 3.2396923100255515e-06,
+      "loss": 0.7224,
+      "step": 31589
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.399641706004984,
+      "learning_rate": 3.2391300650502356e-06,
+      "loss": 0.6636,
+      "step": 31590
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.48950743611697,
+      "learning_rate": 3.2385678594388416e-06,
+      "loss": 0.7068,
+      "step": 31591
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3317698838709129,
+      "learning_rate": 3.238005693194647e-06,
+      "loss": 0.6745,
+      "step": 31592
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2123832666928582,
+      "learning_rate": 3.2374435663209268e-06,
+      "loss": 0.6786,
+      "step": 31593
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4923991604301374,
+      "learning_rate": 3.23688147882095e-06,
+      "loss": 0.7746,
+      "step": 31594
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6443434804816466,
+      "learning_rate": 3.2363194306979884e-06,
+      "loss": 0.7465,
+      "step": 31595
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6100902566168498,
+      "learning_rate": 3.235757421955317e-06,
+      "loss": 0.5001,
+      "step": 31596
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3583575307005646,
+      "learning_rate": 3.235195452596209e-06,
+      "loss": 0.6569,
+      "step": 31597
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.184177645713754,
+      "learning_rate": 3.2346335226239357e-06,
+      "loss": 0.6554,
+      "step": 31598
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4556474367790415,
+      "learning_rate": 3.234071632041765e-06,
+      "loss": 0.7194,
+      "step": 31599
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3171327368962278,
+      "learning_rate": 3.2335097808529736e-06,
+      "loss": 0.6111,
+      "step": 31600
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3099512240399742,
+      "learning_rate": 3.232947969060828e-06,
+      "loss": 0.7668,
+      "step": 31601
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4652478033051983,
+      "learning_rate": 3.2323861966686055e-06,
+      "loss": 0.9095,
+      "step": 31602
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3940723552420773,
+      "learning_rate": 3.231824463679569e-06,
+      "loss": 0.6581,
+      "step": 31603
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3117398537404448,
+      "learning_rate": 3.231262770096997e-06,
+      "loss": 0.7184,
+      "step": 31604
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4433777053081045,
+      "learning_rate": 3.230701115924152e-06,
+      "loss": 0.6435,
+      "step": 31605
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2685609992789244,
+      "learning_rate": 3.2301395011643124e-06,
+      "loss": 0.5902,
+      "step": 31606
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5881083049739577,
+      "learning_rate": 3.2295779258207426e-06,
+      "loss": 0.6853,
+      "step": 31607
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2406925612629434,
+      "learning_rate": 3.229016389896712e-06,
+      "loss": 0.5923,
+      "step": 31608
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5980410437685841,
+      "learning_rate": 3.2284548933954905e-06,
+      "loss": 0.4347,
+      "step": 31609
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5400912719483204,
+      "learning_rate": 3.227893436320353e-06,
+      "loss": 0.7714,
+      "step": 31610
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.933126472814373,
+      "learning_rate": 3.2273320186745625e-06,
+      "loss": 0.5914,
+      "step": 31611
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3335844807877608,
+      "learning_rate": 3.2267706404613864e-06,
+      "loss": 0.5813,
+      "step": 31612
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3007572280995106,
+      "learning_rate": 3.2262093016840955e-06,
+      "loss": 0.6187,
+      "step": 31613
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6296995206085532,
+      "learning_rate": 3.2256480023459633e-06,
+      "loss": 0.4798,
+      "step": 31614
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3915294810604206,
+      "learning_rate": 3.225086742450252e-06,
+      "loss": 0.7538,
+      "step": 31615
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7394591327549995,
+      "learning_rate": 3.224525522000228e-06,
+      "loss": 0.5781,
+      "step": 31616
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4632718296584422,
+      "learning_rate": 3.2239643409991607e-06,
+      "loss": 0.594,
+      "step": 31617
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3568069274316563,
+      "learning_rate": 3.2234031994503223e-06,
+      "loss": 0.6307,
+      "step": 31618
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3311794303170215,
+      "learning_rate": 3.2228420973569753e-06,
+      "loss": 0.6242,
+      "step": 31619
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4564866282031688,
+      "learning_rate": 3.2222810347223845e-06,
+      "loss": 0.634,
+      "step": 31620
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4544178781863595,
+      "learning_rate": 3.2217200115498216e-06,
+      "loss": 0.6683,
+      "step": 31621
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6597520880703054,
+      "learning_rate": 3.2211590278425486e-06,
+      "loss": 0.6808,
+      "step": 31622
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3045853600412294,
+      "learning_rate": 3.2205980836038363e-06,
+      "loss": 0.7193,
+      "step": 31623
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2576086426232893,
+      "learning_rate": 3.2200371788369457e-06,
+      "loss": 0.7609,
+      "step": 31624
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6324804788627821,
+      "learning_rate": 3.219476313545148e-06,
+      "loss": 0.5226,
+      "step": 31625
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6249961291680288,
+      "learning_rate": 3.2189154877317032e-06,
+      "loss": 0.5605,
+      "step": 31626
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5461751043754082,
+      "learning_rate": 3.2183547013998816e-06,
+      "loss": 0.6776,
+      "step": 31627
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6749531273148888,
+      "learning_rate": 3.217793954552946e-06,
+      "loss": 0.5018,
+      "step": 31628
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3089549950601793,
+      "learning_rate": 3.217233247194158e-06,
+      "loss": 0.6583,
+      "step": 31629
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6303643145624219,
+      "learning_rate": 3.216672579326786e-06,
+      "loss": 0.5049,
+      "step": 31630
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6090260176554735,
+      "learning_rate": 3.216111950954094e-06,
+      "loss": 0.5245,
+      "step": 31631
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4856707153255186,
+      "learning_rate": 3.2155513620793523e-06,
+      "loss": 0.6847,
+      "step": 31632
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.603688813540923,
+      "learning_rate": 3.2149908127058116e-06,
+      "loss": 0.6989,
+      "step": 31633
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.415420888102764,
+      "learning_rate": 3.214430302836743e-06,
+      "loss": 0.6002,
+      "step": 31634
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5540442604085967,
+      "learning_rate": 3.2138698324754126e-06,
+      "loss": 0.8404,
+      "step": 31635
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4883553147698927,
+      "learning_rate": 3.2133094016250797e-06,
+      "loss": 0.6905,
+      "step": 31636
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2598532341106081,
+      "learning_rate": 3.212749010289006e-06,
+      "loss": 0.7161,
+      "step": 31637
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2543801563265904,
+      "learning_rate": 3.2121886584704553e-06,
+      "loss": 0.7575,
+      "step": 31638
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.500515899398601,
+      "learning_rate": 3.2116283461726948e-06,
+      "loss": 0.7638,
+      "step": 31639
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6213749804003932,
+      "learning_rate": 3.21106807339898e-06,
+      "loss": 0.6662,
+      "step": 31640
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6149864758675205,
+      "learning_rate": 3.210507840152579e-06,
+      "loss": 0.7164,
+      "step": 31641
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5816228087850122,
+      "learning_rate": 3.209947646436752e-06,
+      "loss": 0.6018,
+      "step": 31642
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5001208536380761,
+      "learning_rate": 3.2093874922547563e-06,
+      "loss": 0.6757,
+      "step": 31643
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3071260975924026,
+      "learning_rate": 3.2088273776098554e-06,
+      "loss": 0.5993,
+      "step": 31644
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2277950140562663,
+      "learning_rate": 3.2082673025053157e-06,
+      "loss": 0.662,
+      "step": 31645
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2387514174727678,
+      "learning_rate": 3.207707266944393e-06,
+      "loss": 0.7444,
+      "step": 31646
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4896707732306957,
+      "learning_rate": 3.2071472709303466e-06,
+      "loss": 0.7142,
+      "step": 31647
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.418286721004429,
+      "learning_rate": 3.2065873144664393e-06,
+      "loss": 0.6566,
+      "step": 31648
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2453260699653144,
+      "learning_rate": 3.2060273975559374e-06,
+      "loss": 0.5695,
+      "step": 31649
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4735593780935567,
+      "learning_rate": 3.205467520202089e-06,
+      "loss": 0.5686,
+      "step": 31650
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3272412185573794,
+      "learning_rate": 3.2049076824081595e-06,
+      "loss": 0.5673,
+      "step": 31651
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4630313554309031,
+      "learning_rate": 3.204347884177409e-06,
+      "loss": 0.6217,
+      "step": 31652
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4092695510285138,
+      "learning_rate": 3.203788125513099e-06,
+      "loss": 0.7061,
+      "step": 31653
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.31115234726315,
+      "learning_rate": 3.2032284064184872e-06,
+      "loss": 0.6621,
+      "step": 31654
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4884227255674942,
+      "learning_rate": 3.202668726896828e-06,
+      "loss": 0.6194,
+      "step": 31655
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.271920258866915,
+      "learning_rate": 3.202109086951386e-06,
+      "loss": 0.6923,
+      "step": 31656
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6721693838095475,
+      "learning_rate": 3.2015494865854144e-06,
+      "loss": 0.8601,
+      "step": 31657
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4920157187221903,
+      "learning_rate": 3.200989925802177e-06,
+      "loss": 0.714,
+      "step": 31658
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7907782960637357,
+      "learning_rate": 3.200430404604926e-06,
+      "loss": 0.6098,
+      "step": 31659
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2712761553329848,
+      "learning_rate": 3.199870922996925e-06,
+      "loss": 0.7578,
+      "step": 31660
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4617764613296114,
+      "learning_rate": 3.199311480981425e-06,
+      "loss": 0.6655,
+      "step": 31661
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.481796666079169,
+      "learning_rate": 3.1987520785616898e-06,
+      "loss": 0.6099,
+      "step": 31662
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.513248483083409,
+      "learning_rate": 3.198192715740973e-06,
+      "loss": 0.7773,
+      "step": 31663
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4290918696702972,
+      "learning_rate": 3.197633392522529e-06,
+      "loss": 0.6364,
+      "step": 31664
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.437313989191036,
+      "learning_rate": 3.1970741089096168e-06,
+      "loss": 0.6558,
+      "step": 31665
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3175421898879223,
+      "learning_rate": 3.196514864905497e-06,
+      "loss": 0.8315,
+      "step": 31666
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2422859494530472,
+      "learning_rate": 3.19595566051342e-06,
+      "loss": 0.6521,
+      "step": 31667
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6716714106294375,
+      "learning_rate": 3.1953964957366413e-06,
+      "loss": 0.6716,
+      "step": 31668
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3460146698926225,
+      "learning_rate": 3.1948373705784173e-06,
+      "loss": 0.58,
+      "step": 31669
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6377184076708786,
+      "learning_rate": 3.1942782850420086e-06,
+      "loss": 0.5791,
+      "step": 31670
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.770099435893598,
+      "learning_rate": 3.193719239130666e-06,
+      "loss": 0.7831,
+      "step": 31671
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4794461593691453,
+      "learning_rate": 3.193160232847642e-06,
+      "loss": 0.6886,
+      "step": 31672
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5987018385872163,
+      "learning_rate": 3.1926012661961935e-06,
+      "loss": 0.5111,
+      "step": 31673
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4835487774675415,
+      "learning_rate": 3.1920423391795795e-06,
+      "loss": 0.6165,
+      "step": 31674
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4674039241609858,
+      "learning_rate": 3.1914834518010506e-06,
+      "loss": 0.6363,
+      "step": 31675
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.611199478284884,
+      "learning_rate": 3.1909246040638562e-06,
+      "loss": 0.4826,
+      "step": 31676
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7187741831614345,
+      "learning_rate": 3.1903657959712585e-06,
+      "loss": 0.7062,
+      "step": 31677
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6165223933323511,
+      "learning_rate": 3.1898070275265035e-06,
+      "loss": 0.7493,
+      "step": 31678
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3169437474457992,
+      "learning_rate": 3.189248298732852e-06,
+      "loss": 0.6303,
+      "step": 31679
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.082241571643523,
+      "learning_rate": 3.1886896095935494e-06,
+      "loss": 0.6203,
+      "step": 31680
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3993262357730611,
+      "learning_rate": 3.188130960111856e-06,
+      "loss": 0.6196,
+      "step": 31681
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4221838419034385,
+      "learning_rate": 3.1875723502910173e-06,
+      "loss": 0.7335,
+      "step": 31682
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3029205497176242,
+      "learning_rate": 3.1870137801342917e-06,
+      "loss": 0.6675,
+      "step": 31683
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7221952080769356,
+      "learning_rate": 3.1864552496449297e-06,
+      "loss": 0.6973,
+      "step": 31684
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5950366536303626,
+      "learning_rate": 3.185896758826178e-06,
+      "loss": 0.7718,
+      "step": 31685
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7657019626553498,
+      "learning_rate": 3.185338307681294e-06,
+      "loss": 0.7597,
+      "step": 31686
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2974208316903335,
+      "learning_rate": 3.18477989621353e-06,
+      "loss": 0.7374,
+      "step": 31687
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.068311836921231,
+      "learning_rate": 3.1842215244261356e-06,
+      "loss": 0.7563,
+      "step": 31688
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3011121721830021,
+      "learning_rate": 3.1836631923223583e-06,
+      "loss": 0.6463,
+      "step": 31689
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5339764939699911,
+      "learning_rate": 3.1831048999054513e-06,
+      "loss": 0.6943,
+      "step": 31690
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4299205985496959,
+      "learning_rate": 3.1825466471786693e-06,
+      "loss": 0.7077,
+      "step": 31691
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3695123873559099,
+      "learning_rate": 3.18198843414526e-06,
+      "loss": 0.545,
+      "step": 31692
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7043997948606715,
+      "learning_rate": 3.1814302608084678e-06,
+      "loss": 0.6461,
+      "step": 31693
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5994460682880925,
+      "learning_rate": 3.180872127171548e-06,
+      "loss": 0.4655,
+      "step": 31694
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.395229717579485,
+      "learning_rate": 3.1803140332377535e-06,
+      "loss": 0.7688,
+      "step": 31695
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4344005073601762,
+      "learning_rate": 3.1797559790103293e-06,
+      "loss": 0.6929,
+      "step": 31696
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6736566157669799,
+      "learning_rate": 3.179197964492522e-06,
+      "loss": 0.6838,
+      "step": 31697
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3649595351324282,
+      "learning_rate": 3.1786399896875862e-06,
+      "loss": 0.726,
+      "step": 31698
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5074359879314352,
+      "learning_rate": 3.1780820545987647e-06,
+      "loss": 0.7317,
+      "step": 31699
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3389139229862563,
+      "learning_rate": 3.1775241592293127e-06,
+      "loss": 0.812,
+      "step": 31700
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4065769328782924,
+      "learning_rate": 3.176966303582473e-06,
+      "loss": 0.6424,
+      "step": 31701
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5641501556984825,
+      "learning_rate": 3.1764084876614965e-06,
+      "loss": 0.7554,
+      "step": 31702
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6159998875494079,
+      "learning_rate": 3.175850711469629e-06,
+      "loss": 0.5194,
+      "step": 31703
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.33753274774179,
+      "learning_rate": 3.1752929750101216e-06,
+      "loss": 0.6921,
+      "step": 31704
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3511584426146932,
+      "learning_rate": 3.1747352782862185e-06,
+      "loss": 0.6299,
+      "step": 31705
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5798261992670652,
+      "learning_rate": 3.1741776213011643e-06,
+      "loss": 0.7044,
+      "step": 31706
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2838949761268321,
+      "learning_rate": 3.17362000405821e-06,
+      "loss": 0.7228,
+      "step": 31707
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3214915984500677,
+      "learning_rate": 3.1730624265606037e-06,
+      "loss": 0.5755,
+      "step": 31708
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4170404252909947,
+      "learning_rate": 3.1725048888115893e-06,
+      "loss": 0.7126,
+      "step": 31709
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4540243686131569,
+      "learning_rate": 3.1719473908144106e-06,
+      "loss": 0.6494,
+      "step": 31710
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5498908035470194,
+      "learning_rate": 3.1713899325723153e-06,
+      "loss": 0.722,
+      "step": 31711
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2818825591151393,
+      "learning_rate": 3.1708325140885533e-06,
+      "loss": 0.5567,
+      "step": 31712
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.365031771624006,
+      "learning_rate": 3.1702751353663664e-06,
+      "loss": 0.7321,
+      "step": 31713
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6081162037441599,
+      "learning_rate": 3.1697177964089963e-06,
+      "loss": 0.6481,
+      "step": 31714
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.687326822943342,
+      "learning_rate": 3.1691604972196922e-06,
+      "loss": 0.7571,
+      "step": 31715
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1923305362133654,
+      "learning_rate": 3.168603237801701e-06,
+      "loss": 0.6005,
+      "step": 31716
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3187866740364222,
+      "learning_rate": 3.168046018158264e-06,
+      "loss": 0.7274,
+      "step": 31717
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4990561965246418,
+      "learning_rate": 3.1674888382926237e-06,
+      "loss": 0.6081,
+      "step": 31718
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.342279345066425,
+      "learning_rate": 3.1669316982080268e-06,
+      "loss": 0.6708,
+      "step": 31719
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2921614396612755,
+      "learning_rate": 3.1663745979077186e-06,
+      "loss": 0.7365,
+      "step": 31720
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7207768070513376,
+      "learning_rate": 3.165817537394943e-06,
+      "loss": 0.711,
+      "step": 31721
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4503727822324108,
+      "learning_rate": 3.1652605166729367e-06,
+      "loss": 0.705,
+      "step": 31722
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5933024152788502,
+      "learning_rate": 3.164703535744951e-06,
+      "loss": 0.7201,
+      "step": 31723
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.38349434965439,
+      "learning_rate": 3.1641465946142223e-06,
+      "loss": 0.6079,
+      "step": 31724
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5761291992783913,
+      "learning_rate": 3.1635896932839994e-06,
+      "loss": 0.4876,
+      "step": 31725
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7768005443502775,
+      "learning_rate": 3.1630328317575187e-06,
+      "loss": 0.6044,
+      "step": 31726
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5853008086517718,
+      "learning_rate": 3.162476010038028e-06,
+      "loss": 0.527,
+      "step": 31727
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3346681593618634,
+      "learning_rate": 3.1619192281287647e-06,
+      "loss": 0.6593,
+      "step": 31728
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4149350885376502,
+      "learning_rate": 3.1613624860329716e-06,
+      "loss": 0.5543,
+      "step": 31729
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6176829733718918,
+      "learning_rate": 3.1608057837538976e-06,
+      "loss": 0.5659,
+      "step": 31730
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2299437206775328,
+      "learning_rate": 3.1602491212947717e-06,
+      "loss": 0.7103,
+      "step": 31731
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6846267356787317,
+      "learning_rate": 3.1596924986588397e-06,
+      "loss": 0.6979,
+      "step": 31732
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4823792597953902,
+      "learning_rate": 3.1591359158493453e-06,
+      "loss": 0.8014,
+      "step": 31733
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.276849119665688,
+      "learning_rate": 3.1585793728695325e-06,
+      "loss": 0.7682,
+      "step": 31734
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6537201155296566,
+      "learning_rate": 3.15802286972263e-06,
+      "loss": 0.7105,
+      "step": 31735
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3134362991669166,
+      "learning_rate": 3.1574664064118856e-06,
+      "loss": 0.6195,
+      "step": 31736
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.384380646418769,
+      "learning_rate": 3.15690998294054e-06,
+      "loss": 0.6585,
+      "step": 31737
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5853339588696858,
+      "learning_rate": 3.1563535993118287e-06,
+      "loss": 0.5551,
+      "step": 31738
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4855252826283234,
+      "learning_rate": 3.155797255528996e-06,
+      "loss": 0.7834,
+      "step": 31739
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2979792041223102,
+      "learning_rate": 3.1552409515952764e-06,
+      "loss": 0.617,
+      "step": 31740
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3749588604746332,
+      "learning_rate": 3.1546846875139136e-06,
+      "loss": 0.6797,
+      "step": 31741
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5235708643025734,
+      "learning_rate": 3.15412846328814e-06,
+      "loss": 0.7247,
+      "step": 31742
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5694801083378203,
+      "learning_rate": 3.153572278921202e-06,
+      "loss": 0.6437,
+      "step": 31743
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4169200604600507,
+      "learning_rate": 3.1530161344163324e-06,
+      "loss": 0.6356,
+      "step": 31744
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6407259238389733,
+      "learning_rate": 3.1524600297767682e-06,
+      "loss": 0.4967,
+      "step": 31745
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.481200736093952,
+      "learning_rate": 3.1519039650057502e-06,
+      "loss": 0.6308,
+      "step": 31746
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3172047601454158,
+      "learning_rate": 3.1513479401065174e-06,
+      "loss": 0.6529,
+      "step": 31747
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4862780323752094,
+      "learning_rate": 3.1507919550823063e-06,
+      "loss": 0.5909,
+      "step": 31748
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4981997559687217,
+      "learning_rate": 3.150236009936349e-06,
+      "loss": 0.6996,
+      "step": 31749
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3196645056405583,
+      "learning_rate": 3.149680104671886e-06,
+      "loss": 0.6697,
+      "step": 31750
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5232894107840445,
+      "learning_rate": 3.149124239292158e-06,
+      "loss": 0.7111,
+      "step": 31751
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4914996280532629,
+      "learning_rate": 3.148568413800397e-06,
+      "loss": 0.6989,
+      "step": 31752
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5940107331822823,
+      "learning_rate": 3.148012628199837e-06,
+      "loss": 0.5364,
+      "step": 31753
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.660357199201985,
+      "learning_rate": 3.147456882493717e-06,
+      "loss": 0.6882,
+      "step": 31754
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3706262949632195,
+      "learning_rate": 3.1469011766852754e-06,
+      "loss": 0.7518,
+      "step": 31755
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.340418094901429,
+      "learning_rate": 3.146345510777745e-06,
+      "loss": 0.7868,
+      "step": 31756
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3934105916272974,
+      "learning_rate": 3.1457898847743573e-06,
+      "loss": 0.7619,
+      "step": 31757
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.365229441395684,
+      "learning_rate": 3.1452342986783545e-06,
+      "loss": 0.7142,
+      "step": 31758
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.437864171585828,
+      "learning_rate": 3.1446787524929644e-06,
+      "loss": 0.6605,
+      "step": 31759
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6251149024666034,
+      "learning_rate": 3.144123246221428e-06,
+      "loss": 0.514,
+      "step": 31760
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4673993630222741,
+      "learning_rate": 3.1435677798669738e-06,
+      "loss": 0.6341,
+      "step": 31761
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7107054188824404,
+      "learning_rate": 3.1430123534328415e-06,
+      "loss": 0.7078,
+      "step": 31762
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4615539257873906,
+      "learning_rate": 3.142456966922259e-06,
+      "loss": 0.7472,
+      "step": 31763
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5750902487821811,
+      "learning_rate": 3.1419016203384666e-06,
+      "loss": 0.4674,
+      "step": 31764
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4087444965322302,
+      "learning_rate": 3.1413463136846945e-06,
+      "loss": 0.6904,
+      "step": 31765
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4634951050681388,
+      "learning_rate": 3.140791046964172e-06,
+      "loss": 0.7272,
+      "step": 31766
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.704773741938361,
+      "learning_rate": 3.1402358201801355e-06,
+      "loss": 0.6361,
+      "step": 31767
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2339123895689725,
+      "learning_rate": 3.1396806333358208e-06,
+      "loss": 0.6724,
+      "step": 31768
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.22229996494404,
+      "learning_rate": 3.1391254864344577e-06,
+      "loss": 0.7185,
+      "step": 31769
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4962302055376955,
+      "learning_rate": 3.1385703794792742e-06,
+      "loss": 0.6038,
+      "step": 31770
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4421796728096563,
+      "learning_rate": 3.138015312473508e-06,
+      "loss": 0.6024,
+      "step": 31771
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5501414926384595,
+      "learning_rate": 3.1374602854203908e-06,
+      "loss": 0.468,
+      "step": 31772
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4957308322921141,
+      "learning_rate": 3.1369052983231527e-06,
+      "loss": 0.8464,
+      "step": 31773
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5605131075403642,
+      "learning_rate": 3.1363503511850222e-06,
+      "loss": 0.7332,
+      "step": 31774
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6380015115438671,
+      "learning_rate": 3.135795444009232e-06,
+      "loss": 0.5223,
+      "step": 31775
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4434042966944869,
+      "learning_rate": 3.135240576799018e-06,
+      "loss": 0.6165,
+      "step": 31776
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7582566020997286,
+      "learning_rate": 3.1346857495576066e-06,
+      "loss": 0.7976,
+      "step": 31777
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6068793870188203,
+      "learning_rate": 3.134130962288224e-06,
+      "loss": 0.5322,
+      "step": 31778
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.219724451818484,
+      "learning_rate": 3.1335762149941095e-06,
+      "loss": 0.5708,
+      "step": 31779
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3996208804299715,
+      "learning_rate": 3.133021507678484e-06,
+      "loss": 0.6827,
+      "step": 31780
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.290816028444897,
+      "learning_rate": 3.1324668403445845e-06,
+      "loss": 0.7228,
+      "step": 31781
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3129439703450871,
+      "learning_rate": 3.1319122129956337e-06,
+      "loss": 0.7354,
+      "step": 31782
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3330452151125196,
+      "learning_rate": 3.1313576256348687e-06,
+      "loss": 0.7204,
+      "step": 31783
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2679420367868086,
+      "learning_rate": 3.1308030782655106e-06,
+      "loss": 0.7329,
+      "step": 31784
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5142331802796603,
+      "learning_rate": 3.130248570890795e-06,
+      "loss": 0.6558,
+      "step": 31785
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3865102352610532,
+      "learning_rate": 3.129694103513946e-06,
+      "loss": 0.7708,
+      "step": 31786
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5693240814905012,
+      "learning_rate": 3.129139676138191e-06,
+      "loss": 0.6706,
+      "step": 31787
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3845998173138274,
+      "learning_rate": 3.12858528876676e-06,
+      "loss": 0.6762,
+      "step": 31788
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4202722406138681,
+      "learning_rate": 3.128030941402883e-06,
+      "loss": 0.6666,
+      "step": 31789
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4431042153288365,
+      "learning_rate": 3.1274766340497876e-06,
+      "loss": 0.7917,
+      "step": 31790
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1818222446335775,
+      "learning_rate": 3.126922366710694e-06,
+      "loss": 0.5646,
+      "step": 31791
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2911790059112884,
+      "learning_rate": 3.1263681393888347e-06,
+      "loss": 0.6123,
+      "step": 31792
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3352339855852497,
+      "learning_rate": 3.1258139520874396e-06,
+      "loss": 0.7434,
+      "step": 31793
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5636539281930864,
+      "learning_rate": 3.125259804809733e-06,
+      "loss": 0.7135,
+      "step": 31794
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.476716953919221,
+      "learning_rate": 3.1247056975589364e-06,
+      "loss": 0.7142,
+      "step": 31795
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.504854287345811,
+      "learning_rate": 3.124151630338279e-06,
+      "loss": 0.6335,
+      "step": 31796
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3575093164427379,
+      "learning_rate": 3.123597603150992e-06,
+      "loss": 0.7028,
+      "step": 31797
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4488055490900238,
+      "learning_rate": 3.1230436160002963e-06,
+      "loss": 0.6666,
+      "step": 31798
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4663165970361434,
+      "learning_rate": 3.1224896688894146e-06,
+      "loss": 0.7187,
+      "step": 31799
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4428299451824567,
+      "learning_rate": 3.1219357618215785e-06,
+      "loss": 0.6564,
+      "step": 31800
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4908498203814795,
+      "learning_rate": 3.1213818948000074e-06,
+      "loss": 0.6937,
+      "step": 31801
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3957356066561466,
+      "learning_rate": 3.120828067827931e-06,
+      "loss": 0.6794,
+      "step": 31802
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.477446029350301,
+      "learning_rate": 3.1202742809085694e-06,
+      "loss": 0.6456,
+      "step": 31803
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.9187557499543857,
+      "learning_rate": 3.119720534045152e-06,
+      "loss": 0.8117,
+      "step": 31804
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3947055783174442,
+      "learning_rate": 3.119166827240896e-06,
+      "loss": 0.634,
+      "step": 31805
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.37082251715618,
+      "learning_rate": 3.118613160499032e-06,
+      "loss": 0.6428,
+      "step": 31806
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3661566476459808,
+      "learning_rate": 3.1180595338227827e-06,
+      "loss": 0.6851,
+      "step": 31807
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3606285736734858,
+      "learning_rate": 3.1175059472153648e-06,
+      "loss": 0.7227,
+      "step": 31808
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3534252622882514,
+      "learning_rate": 3.1169524006800066e-06,
+      "loss": 0.8492,
+      "step": 31809
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6800925450606552,
+      "learning_rate": 3.116398894219934e-06,
+      "loss": 0.7327,
+      "step": 31810
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3947847701506335,
+      "learning_rate": 3.1158454278383666e-06,
+      "loss": 0.5871,
+      "step": 31811
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2674374796131649,
+      "learning_rate": 3.1152920015385225e-06,
+      "loss": 0.6011,
+      "step": 31812
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4227761268132426,
+      "learning_rate": 3.114738615323628e-06,
+      "loss": 0.8403,
+      "step": 31813
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.532077226953911,
+      "learning_rate": 3.1141852691969098e-06,
+      "loss": 0.6317,
+      "step": 31814
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2953087950240865,
+      "learning_rate": 3.113631963161584e-06,
+      "loss": 0.5781,
+      "step": 31815
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2308654858514902,
+      "learning_rate": 3.1130786972208694e-06,
+      "loss": 0.637,
+      "step": 31816
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2735532202795885,
+      "learning_rate": 3.112525471377992e-06,
+      "loss": 0.7299,
+      "step": 31817
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4480282062398306,
+      "learning_rate": 3.1119722856361744e-06,
+      "loss": 0.7574,
+      "step": 31818
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6636704980740566,
+      "learning_rate": 3.111419139998635e-06,
+      "loss": 0.7177,
+      "step": 31819
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8798169966577172,
+      "learning_rate": 3.1108660344685913e-06,
+      "loss": 0.6106,
+      "step": 31820
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2933655839337503,
+      "learning_rate": 3.1103129690492694e-06,
+      "loss": 0.7346,
+      "step": 31821
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2980899607983551,
+      "learning_rate": 3.109759943743884e-06,
+      "loss": 0.7168,
+      "step": 31822
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.599049707664344,
+      "learning_rate": 3.1092069585556573e-06,
+      "loss": 0.5022,
+      "step": 31823
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3280965220088292,
+      "learning_rate": 3.1086540134878117e-06,
+      "loss": 0.6321,
+      "step": 31824
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5007538494706176,
+      "learning_rate": 3.1081011085435653e-06,
+      "loss": 0.7318,
+      "step": 31825
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.346354949223593,
+      "learning_rate": 3.1075482437261326e-06,
+      "loss": 0.7908,
+      "step": 31826
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5968259410994541,
+      "learning_rate": 3.106995419038736e-06,
+      "loss": 0.5051,
+      "step": 31827
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.529839880277947,
+      "learning_rate": 3.106442634484601e-06,
+      "loss": 0.632,
+      "step": 31828
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3950549556431506,
+      "learning_rate": 3.105889890066933e-06,
+      "loss": 0.7428,
+      "step": 31829
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5117288846904806,
+      "learning_rate": 3.105337185788957e-06,
+      "loss": 0.6452,
+      "step": 31830
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3130525235550208,
+      "learning_rate": 3.1047845216538895e-06,
+      "loss": 0.6781,
+      "step": 31831
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.686450402730473,
+      "learning_rate": 3.1042318976649575e-06,
+      "loss": 0.5829,
+      "step": 31832
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7895917071688698,
+      "learning_rate": 3.1036793138253627e-06,
+      "loss": 0.6608,
+      "step": 31833
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3264747494086218,
+      "learning_rate": 3.103126770138332e-06,
+      "loss": 0.418,
+      "step": 31834
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4302658604960568,
+      "learning_rate": 3.1025742666070823e-06,
+      "loss": 0.762,
+      "step": 31835
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4882219497942928,
+      "learning_rate": 3.102021803234827e-06,
+      "loss": 0.7352,
+      "step": 31836
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2681959964558804,
+      "learning_rate": 3.1014693800247865e-06,
+      "loss": 0.6847,
+      "step": 31837
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4814122161797998,
+      "learning_rate": 3.1009169969801735e-06,
+      "loss": 0.7633,
+      "step": 31838
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3343130181840006,
+      "learning_rate": 3.100364654104208e-06,
+      "loss": 0.6619,
+      "step": 31839
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4308587600094596,
+      "learning_rate": 3.099812351400101e-06,
+      "loss": 0.7582,
+      "step": 31840
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.714403537694218,
+      "learning_rate": 3.0992600888710745e-06,
+      "loss": 0.7298,
+      "step": 31841
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1910526457171966,
+      "learning_rate": 3.098707866520341e-06,
+      "loss": 0.5121,
+      "step": 31842
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.383121772406285,
+      "learning_rate": 3.098155684351111e-06,
+      "loss": 0.6802,
+      "step": 31843
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4010165987297127,
+      "learning_rate": 3.097603542366604e-06,
+      "loss": 0.7102,
+      "step": 31844
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5721997841271724,
+      "learning_rate": 3.097051440570038e-06,
+      "loss": 0.6172,
+      "step": 31845
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5244557868222264,
+      "learning_rate": 3.0964993789646236e-06,
+      "loss": 0.6194,
+      "step": 31846
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6514952105298493,
+      "learning_rate": 3.0959473575535725e-06,
+      "loss": 0.7025,
+      "step": 31847
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4809364306763446,
+      "learning_rate": 3.0953953763401013e-06,
+      "loss": 0.6823,
+      "step": 31848
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.802942893067635,
+      "learning_rate": 3.094843435327427e-06,
+      "loss": 0.6519,
+      "step": 31849
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2518257772793837,
+      "learning_rate": 3.0942915345187617e-06,
+      "loss": 0.6247,
+      "step": 31850
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6533279603370084,
+      "learning_rate": 3.093739673917313e-06,
+      "loss": 0.4824,
+      "step": 31851
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4754297974853599,
+      "learning_rate": 3.0931878535262994e-06,
+      "loss": 0.5978,
+      "step": 31852
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6688990600934177,
+      "learning_rate": 3.0926360733489345e-06,
+      "loss": 0.8185,
+      "step": 31853
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6281811014016396,
+      "learning_rate": 3.092084333388431e-06,
+      "loss": 0.5134,
+      "step": 31854
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.252161654750359,
+      "learning_rate": 3.0915326336479947e-06,
+      "loss": 0.6117,
+      "step": 31855
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3071809187842796,
+      "learning_rate": 3.0909809741308437e-06,
+      "loss": 0.6952,
+      "step": 31856
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2696141332727335,
+      "learning_rate": 3.090429354840192e-06,
+      "loss": 0.7129,
+      "step": 31857
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5302194513523966,
+      "learning_rate": 3.0898777757792477e-06,
+      "loss": 0.7028,
+      "step": 31858
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.819075253859002,
+      "learning_rate": 3.089326236951219e-06,
+      "loss": 0.6489,
+      "step": 31859
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3393523514973318,
+      "learning_rate": 3.0887747383593245e-06,
+      "loss": 0.705,
+      "step": 31860
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0651223499971136,
+      "learning_rate": 3.088223280006769e-06,
+      "loss": 0.7911,
+      "step": 31861
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1592328286125726,
+      "learning_rate": 3.087671861896768e-06,
+      "loss": 0.6807,
+      "step": 31862
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5613404314398924,
+      "learning_rate": 3.0871204840325276e-06,
+      "loss": 0.7399,
+      "step": 31863
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.494871808581421,
+      "learning_rate": 3.086569146417263e-06,
+      "loss": 0.7472,
+      "step": 31864
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.367559591998052,
+      "learning_rate": 3.086017849054178e-06,
+      "loss": 0.7196,
+      "step": 31865
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.42519897876416,
+      "learning_rate": 3.08546659194649e-06,
+      "loss": 0.7208,
+      "step": 31866
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4345918993221638,
+      "learning_rate": 3.0849153750974036e-06,
+      "loss": 0.7897,
+      "step": 31867
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5181461789799078,
+      "learning_rate": 3.084364198510127e-06,
+      "loss": 0.7597,
+      "step": 31868
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0789958014956387,
+      "learning_rate": 3.0838130621878704e-06,
+      "loss": 0.5792,
+      "step": 31869
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4196250132363974,
+      "learning_rate": 3.0832619661338482e-06,
+      "loss": 0.7814,
+      "step": 31870
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6106890837840414,
+      "learning_rate": 3.0827109103512643e-06,
+      "loss": 0.5288,
+      "step": 31871
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4352302165832378,
+      "learning_rate": 3.082159894843324e-06,
+      "loss": 0.5962,
+      "step": 31872
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3022927341974333,
+      "learning_rate": 3.081608919613238e-06,
+      "loss": 0.6401,
+      "step": 31873
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3366155109588018,
+      "learning_rate": 3.0810579846642196e-06,
+      "loss": 0.5461,
+      "step": 31874
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.252421974996687,
+      "learning_rate": 3.0805070899994715e-06,
+      "loss": 0.6988,
+      "step": 31875
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.0953072852291275,
+      "learning_rate": 3.079956235622199e-06,
+      "loss": 0.6963,
+      "step": 31876
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5365878166013638,
+      "learning_rate": 3.079405421535613e-06,
+      "loss": 0.676,
+      "step": 31877
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4286291382029894,
+      "learning_rate": 3.0788546477429214e-06,
+      "loss": 0.8344,
+      "step": 31878
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4808236165642403,
+      "learning_rate": 3.0783039142473305e-06,
+      "loss": 0.7078,
+      "step": 31879
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3455428915249348,
+      "learning_rate": 3.0777532210520413e-06,
+      "loss": 0.6424,
+      "step": 31880
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.570326818943104,
+      "learning_rate": 3.077202568160268e-06,
+      "loss": 0.4644,
+      "step": 31881
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5652268691112927,
+      "learning_rate": 3.0766519555752095e-06,
+      "loss": 0.7051,
+      "step": 31882
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.436041182102783,
+      "learning_rate": 3.0761013833000785e-06,
+      "loss": 0.6919,
+      "step": 31883
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3958087450204253,
+      "learning_rate": 3.075550851338075e-06,
+      "loss": 0.716,
+      "step": 31884
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6467660608740975,
+      "learning_rate": 3.0750003596924095e-06,
+      "loss": 0.624,
+      "step": 31885
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3037567494612954,
+      "learning_rate": 3.0744499083662803e-06,
+      "loss": 0.6598,
+      "step": 31886
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4556982437541865,
+      "learning_rate": 3.0738994973629e-06,
+      "loss": 0.7145,
+      "step": 31887
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6629340149463214,
+      "learning_rate": 3.07334912668547e-06,
+      "loss": 0.6577,
+      "step": 31888
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3739034368519762,
+      "learning_rate": 3.0727987963371906e-06,
+      "loss": 0.6321,
+      "step": 31889
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2567667520399683,
+      "learning_rate": 3.0722485063212703e-06,
+      "loss": 0.5927,
+      "step": 31890
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.751661551693968,
+      "learning_rate": 3.071698256640915e-06,
+      "loss": 0.6913,
+      "step": 31891
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3487447259140568,
+      "learning_rate": 3.0711480472993262e-06,
+      "loss": 0.7309,
+      "step": 31892
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.29689156720268,
+      "learning_rate": 3.070597878299705e-06,
+      "loss": 0.5725,
+      "step": 31893
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.469580457540588,
+      "learning_rate": 3.070047749645255e-06,
+      "loss": 0.6503,
+      "step": 31894
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4736816665345847,
+      "learning_rate": 3.0694976613391845e-06,
+      "loss": 0.6998,
+      "step": 31895
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4572943504006504,
+      "learning_rate": 3.068947613384694e-06,
+      "loss": 0.7646,
+      "step": 31896
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.525967209425751,
+      "learning_rate": 3.068397605784981e-06,
+      "loss": 0.6587,
+      "step": 31897
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.232243954513304,
+      "learning_rate": 3.067847638543251e-06,
+      "loss": 0.5778,
+      "step": 31898
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.466787555544098,
+      "learning_rate": 3.0672977116627102e-06,
+      "loss": 0.7149,
+      "step": 31899
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3172297771581158,
+      "learning_rate": 3.066747825146558e-06,
+      "loss": 0.5692,
+      "step": 31900
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6149401844666819,
+      "learning_rate": 3.06619797899799e-06,
+      "loss": 0.5324,
+      "step": 31901
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.376272728939259,
+      "learning_rate": 3.065648173220217e-06,
+      "loss": 0.5942,
+      "step": 31902
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.282718041942477,
+      "learning_rate": 3.0650984078164325e-06,
+      "loss": 0.6109,
+      "step": 31903
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6356241342838431,
+      "learning_rate": 3.064548682789843e-06,
+      "loss": 0.7364,
+      "step": 31904
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1584011905067355,
+      "learning_rate": 3.0639989981436447e-06,
+      "loss": 0.6293,
+      "step": 31905
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4765620175058407,
+      "learning_rate": 3.063449353881043e-06,
+      "loss": 0.6639,
+      "step": 31906
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4272399291930837,
+      "learning_rate": 3.0628997500052314e-06,
+      "loss": 0.8768,
+      "step": 31907
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4799656933907792,
+      "learning_rate": 3.062350186519417e-06,
+      "loss": 0.748,
+      "step": 31908
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5542014322566984,
+      "learning_rate": 3.061800663426797e-06,
+      "loss": 0.7047,
+      "step": 31909
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5175208635900443,
+      "learning_rate": 3.0612511807305665e-06,
+      "loss": 0.6268,
+      "step": 31910
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4562602306132941,
+      "learning_rate": 3.060701738433929e-06,
+      "loss": 0.7208,
+      "step": 31911
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2164563168894267,
+      "learning_rate": 3.060152336540082e-06,
+      "loss": 0.5973,
+      "step": 31912
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6301245357116434,
+      "learning_rate": 3.0596029750522316e-06,
+      "loss": 0.674,
+      "step": 31913
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3617088051860344,
+      "learning_rate": 3.059053653973564e-06,
+      "loss": 0.5949,
+      "step": 31914
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4370405362131367,
+      "learning_rate": 3.0585043733072838e-06,
+      "loss": 0.843,
+      "step": 31915
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5689023933582815,
+      "learning_rate": 3.0579551330565915e-06,
+      "loss": 0.4881,
+      "step": 31916
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.373514488083275,
+      "learning_rate": 3.057405933224682e-06,
+      "loss": 0.5775,
+      "step": 31917
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3114741489138924,
+      "learning_rate": 3.0568567738147505e-06,
+      "loss": 0.7085,
+      "step": 31918
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3984140879778082,
+      "learning_rate": 3.0563076548299965e-06,
+      "loss": 0.695,
+      "step": 31919
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3709580081362975,
+      "learning_rate": 3.055758576273621e-06,
+      "loss": 0.711,
+      "step": 31920
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4031388066510977,
+      "learning_rate": 3.0552095381488144e-06,
+      "loss": 0.6793,
+      "step": 31921
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3674775379001098,
+      "learning_rate": 3.0546605404587803e-06,
+      "loss": 0.641,
+      "step": 31922
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.282537424732728,
+      "learning_rate": 3.0541115832067115e-06,
+      "loss": 0.6339,
+      "step": 31923
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5744092520284774,
+      "learning_rate": 3.0535626663957996e-06,
+      "loss": 0.6587,
+      "step": 31924
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3896881291028242,
+      "learning_rate": 3.0530137900292456e-06,
+      "loss": 0.7601,
+      "step": 31925
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5391354742107874,
+      "learning_rate": 3.052464954110249e-06,
+      "loss": 0.7009,
+      "step": 31926
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6097753373467825,
+      "learning_rate": 3.051916158642e-06,
+      "loss": 0.5144,
+      "step": 31927
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2459438981241004,
+      "learning_rate": 3.051367403627692e-06,
+      "loss": 0.6576,
+      "step": 31928
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.436132460636789,
+      "learning_rate": 3.0508186890705214e-06,
+      "loss": 0.6858,
+      "step": 31929
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5437474958570763,
+      "learning_rate": 3.050270014973693e-06,
+      "loss": 0.5945,
+      "step": 31930
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2456516879649802,
+      "learning_rate": 3.049721381340386e-06,
+      "loss": 0.6639,
+      "step": 31931
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6014454511431285,
+      "learning_rate": 3.049172788173801e-06,
+      "loss": 0.4967,
+      "step": 31932
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5845565791479763,
+      "learning_rate": 3.0486242354771322e-06,
+      "loss": 0.4973,
+      "step": 31933
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5794416933512097,
+      "learning_rate": 3.0480757232535773e-06,
+      "loss": 0.4991,
+      "step": 31934
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.31840910088177,
+      "learning_rate": 3.047527251506327e-06,
+      "loss": 0.767,
+      "step": 31935
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3270373953771677,
+      "learning_rate": 3.0469788202385707e-06,
+      "loss": 0.5957,
+      "step": 31936
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3345596044300512,
+      "learning_rate": 3.0464304294535086e-06,
+      "loss": 0.6767,
+      "step": 31937
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4041183126078025,
+      "learning_rate": 3.0458820791543263e-06,
+      "loss": 0.6054,
+      "step": 31938
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6145635255589113,
+      "learning_rate": 3.045333769344223e-06,
+      "loss": 0.7343,
+      "step": 31939
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1574057135302942,
+      "learning_rate": 3.044785500026386e-06,
+      "loss": 0.5457,
+      "step": 31940
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4648151532699818,
+      "learning_rate": 3.0442372712040123e-06,
+      "loss": 0.7306,
+      "step": 31941
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7425562285690692,
+      "learning_rate": 3.043689082880289e-06,
+      "loss": 0.7082,
+      "step": 31942
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.822055177355987,
+      "learning_rate": 3.043140935058412e-06,
+      "loss": 0.5721,
+      "step": 31943
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2844847356375215,
+      "learning_rate": 3.0425928277415716e-06,
+      "loss": 0.6988,
+      "step": 31944
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5595812878897308,
+      "learning_rate": 3.0420447609329564e-06,
+      "loss": 0.7271,
+      "step": 31945
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3648594290887375,
+      "learning_rate": 3.041496734635758e-06,
+      "loss": 0.7267,
+      "step": 31946
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3579957523912687,
+      "learning_rate": 3.040948748853172e-06,
+      "loss": 0.6689,
+      "step": 31947
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4141571536388557,
+      "learning_rate": 3.0404008035883857e-06,
+      "loss": 0.7586,
+      "step": 31948
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2798777267497305,
+      "learning_rate": 3.0398528988445875e-06,
+      "loss": 0.6822,
+      "step": 31949
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.338332768208719,
+      "learning_rate": 3.0393050346249677e-06,
+      "loss": 0.7311,
+      "step": 31950
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4361446988426312,
+      "learning_rate": 3.0387572109327214e-06,
+      "loss": 0.6828,
+      "step": 31951
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.487318674262916,
+      "learning_rate": 3.0382094277710337e-06,
+      "loss": 0.7212,
+      "step": 31952
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.343188854331327,
+      "learning_rate": 3.037661685143093e-06,
+      "loss": 0.6471,
+      "step": 31953
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3193491778101565,
+      "learning_rate": 3.037113983052089e-06,
+      "loss": 0.7673,
+      "step": 31954
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3225451216109763,
+      "learning_rate": 3.036566321501215e-06,
+      "loss": 0.6153,
+      "step": 31955
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2917225767063998,
+      "learning_rate": 3.036018700493658e-06,
+      "loss": 0.5561,
+      "step": 31956
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7180362524882056,
+      "learning_rate": 3.0354711200326004e-06,
+      "loss": 0.6855,
+      "step": 31957
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5041779189732245,
+      "learning_rate": 3.0349235801212383e-06,
+      "loss": 0.7783,
+      "step": 31958
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2341373491187346,
+      "learning_rate": 3.0343760807627533e-06,
+      "loss": 0.6218,
+      "step": 31959
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3556851198785878,
+      "learning_rate": 3.0338286219603385e-06,
+      "loss": 0.6744,
+      "step": 31960
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6040308496910014,
+      "learning_rate": 3.033281203717177e-06,
+      "loss": 0.7428,
+      "step": 31961
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3167150258825056,
+      "learning_rate": 3.0327338260364604e-06,
+      "loss": 0.6486,
+      "step": 31962
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2526686938474503,
+      "learning_rate": 3.032186488921369e-06,
+      "loss": 0.6956,
+      "step": 31963
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2797567788826305,
+      "learning_rate": 3.0316391923750986e-06,
+      "loss": 0.7578,
+      "step": 31964
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3916808492677382,
+      "learning_rate": 3.031091936400831e-06,
+      "loss": 0.6874,
+      "step": 31965
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3973954066129044,
+      "learning_rate": 3.030544721001748e-06,
+      "loss": 0.7466,
+      "step": 31966
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2378441113119527,
+      "learning_rate": 3.0299975461810406e-06,
+      "loss": 0.5671,
+      "step": 31967
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7949007280116962,
+      "learning_rate": 3.029450411941898e-06,
+      "loss": 0.6809,
+      "step": 31968
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3181942580058517,
+      "learning_rate": 3.0289033182875015e-06,
+      "loss": 0.7246,
+      "step": 31969
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7693765532639374,
+      "learning_rate": 3.028356265221033e-06,
+      "loss": 0.687,
+      "step": 31970
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4946899733492183,
+      "learning_rate": 3.0278092527456826e-06,
+      "loss": 0.587,
+      "step": 31971
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4981033175925107,
+      "learning_rate": 3.0272622808646366e-06,
+      "loss": 0.6928,
+      "step": 31972
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2713356972962564,
+      "learning_rate": 3.0267153495810776e-06,
+      "loss": 0.7095,
+      "step": 31973
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6621768036454885,
+      "learning_rate": 3.0261684588981866e-06,
+      "loss": 0.6511,
+      "step": 31974
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3386356131883752,
+      "learning_rate": 3.0256216088191502e-06,
+      "loss": 0.6465,
+      "step": 31975
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9387295827631026,
+      "learning_rate": 3.0250747993471563e-06,
+      "loss": 0.7446,
+      "step": 31976
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4666075262731144,
+      "learning_rate": 3.024528030485384e-06,
+      "loss": 0.8673,
+      "step": 31977
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5967504381159711,
+      "learning_rate": 3.023981302237016e-06,
+      "loss": 0.4975,
+      "step": 31978
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1810161457276414,
+      "learning_rate": 3.0234346146052374e-06,
+      "loss": 0.7047,
+      "step": 31979
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6061107371964816,
+      "learning_rate": 3.0228879675932343e-06,
+      "loss": 0.5969,
+      "step": 31980
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3219939731934538,
+      "learning_rate": 3.0223413612041865e-06,
+      "loss": 0.6593,
+      "step": 31981
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.427729520923201,
+      "learning_rate": 3.021794795441272e-06,
+      "loss": 0.7729,
+      "step": 31982
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.18027776449444,
+      "learning_rate": 3.0212482703076817e-06,
+      "loss": 0.7622,
+      "step": 31983
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6794280917045135,
+      "learning_rate": 3.0207017858065903e-06,
+      "loss": 0.5812,
+      "step": 31984
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4429945230739807,
+      "learning_rate": 3.0201553419411857e-06,
+      "loss": 0.6814,
+      "step": 31985
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3766253670679773,
+      "learning_rate": 3.0196089387146433e-06,
+      "loss": 0.7424,
+      "step": 31986
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.391918521906289,
+      "learning_rate": 3.0190625761301506e-06,
+      "loss": 0.7675,
+      "step": 31987
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.172233408692706,
+      "learning_rate": 3.018516254190883e-06,
+      "loss": 0.7174,
+      "step": 31988
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4214473155104241,
+      "learning_rate": 3.017969972900027e-06,
+      "loss": 0.5986,
+      "step": 31989
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.630435990241982,
+      "learning_rate": 3.0174237322607593e-06,
+      "loss": 0.6618,
+      "step": 31990
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3344197816515913,
+      "learning_rate": 3.016877532276259e-06,
+      "loss": 0.5925,
+      "step": 31991
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4188186545044712,
+      "learning_rate": 3.016331372949709e-06,
+      "loss": 0.5689,
+      "step": 31992
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3617382099986428,
+      "learning_rate": 3.0157852542842914e-06,
+      "loss": 0.7314,
+      "step": 31993
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0848391710658467,
+      "learning_rate": 3.015239176283183e-06,
+      "loss": 0.5965,
+      "step": 31994
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6168045390310166,
+      "learning_rate": 3.0146931389495604e-06,
+      "loss": 0.6827,
+      "step": 31995
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6986301710886773,
+      "learning_rate": 3.0141471422866065e-06,
+      "loss": 0.68,
+      "step": 31996
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3531805558916299,
+      "learning_rate": 3.0136011862975022e-06,
+      "loss": 0.7518,
+      "step": 31997
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.305600187071084,
+      "learning_rate": 3.0130552709854235e-06,
+      "loss": 0.7332,
+      "step": 31998
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5413656448966514,
+      "learning_rate": 3.0125093963535458e-06,
+      "loss": 0.4463,
+      "step": 31999
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9396590296749987,
+      "learning_rate": 3.0119635624050515e-06,
+      "loss": 0.6891,
+      "step": 32000
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7606243531332135,
+      "learning_rate": 3.0114177691431203e-06,
+      "loss": 0.5766,
+      "step": 32001
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5279385544091924,
+      "learning_rate": 3.0108720165709273e-06,
+      "loss": 0.746,
+      "step": 32002
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.28947370409235,
+      "learning_rate": 3.0103263046916464e-06,
+      "loss": 0.6818,
+      "step": 32003
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5516985316703593,
+      "learning_rate": 3.009780633508463e-06,
+      "loss": 0.5787,
+      "step": 32004
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.491311810811898,
+      "learning_rate": 3.0092350030245465e-06,
+      "loss": 0.6542,
+      "step": 32005
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5888279746783439,
+      "learning_rate": 3.00868941324308e-06,
+      "loss": 0.5117,
+      "step": 32006
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3642791540171997,
+      "learning_rate": 3.0081438641672334e-06,
+      "loss": 0.6511,
+      "step": 32007
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4035121213863326,
+      "learning_rate": 3.0075983558001898e-06,
+      "loss": 0.639,
+      "step": 32008
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0100914173996376,
+      "learning_rate": 3.00705288814512e-06,
+      "loss": 0.5429,
+      "step": 32009
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5953034830821425,
+      "learning_rate": 3.0065074612052024e-06,
+      "loss": 0.5005,
+      "step": 32010
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5780988111320314,
+      "learning_rate": 3.0059620749836182e-06,
+      "loss": 0.4713,
+      "step": 32011
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5826262699047527,
+      "learning_rate": 3.0054167294835314e-06,
+      "loss": 0.5256,
+      "step": 32012
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2662651165989012,
+      "learning_rate": 3.004871424708121e-06,
+      "loss": 0.6662,
+      "step": 32013
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.296699047259675,
+      "learning_rate": 3.0043261606605656e-06,
+      "loss": 0.6839,
+      "step": 32014
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6053745060875464,
+      "learning_rate": 3.003780937344044e-06,
+      "loss": 0.473,
+      "step": 32015
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3902500934246826,
+      "learning_rate": 3.0032357547617187e-06,
+      "loss": 0.6958,
+      "step": 32016
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3146922594562203,
+      "learning_rate": 3.00269061291677e-06,
+      "loss": 0.6797,
+      "step": 32017
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4031941923740114,
+      "learning_rate": 3.0021455118123745e-06,
+      "loss": 0.7168,
+      "step": 32018
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3915556170382741,
+      "learning_rate": 3.0016004514517005e-06,
+      "loss": 0.6873,
+      "step": 32019
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3493248904969468,
+      "learning_rate": 3.001055431837927e-06,
+      "loss": 0.6594,
+      "step": 32020
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7505170750777,
+      "learning_rate": 3.0005104529742214e-06,
+      "loss": 0.7315,
+      "step": 32021
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4746577489809956,
+      "learning_rate": 2.999965514863764e-06,
+      "loss": 0.6823,
+      "step": 32022
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3990922366239356,
+      "learning_rate": 2.9994206175097196e-06,
+      "loss": 0.6906,
+      "step": 32023
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6635849445773734,
+      "learning_rate": 2.9988757609152686e-06,
+      "loss": 0.7254,
+      "step": 32024
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4585776964517279,
+      "learning_rate": 2.9983309450835785e-06,
+      "loss": 0.603,
+      "step": 32025
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6196861628577353,
+      "learning_rate": 2.9977861700178202e-06,
+      "loss": 0.6404,
+      "step": 32026
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.359182574237502,
+      "learning_rate": 2.9972414357211667e-06,
+      "loss": 0.8227,
+      "step": 32027
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2980099806543544,
+      "learning_rate": 2.996696742196794e-06,
+      "loss": 0.666,
+      "step": 32028
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1360083748019338,
+      "learning_rate": 2.9961520894478703e-06,
+      "loss": 0.6943,
+      "step": 32029
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7079800920384993,
+      "learning_rate": 2.9956074774775636e-06,
+      "loss": 0.6132,
+      "step": 32030
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3972820744133254,
+      "learning_rate": 2.995062906289047e-06,
+      "loss": 0.4963,
+      "step": 32031
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3435467596698092,
+      "learning_rate": 2.9945183758854956e-06,
+      "loss": 0.6894,
+      "step": 32032
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2348645921036885,
+      "learning_rate": 2.9939738862700764e-06,
+      "loss": 0.7198,
+      "step": 32033
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1225909288948281,
+      "learning_rate": 2.9934294374459548e-06,
+      "loss": 0.4661,
+      "step": 32034
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6033005222099991,
+      "learning_rate": 2.992885029416307e-06,
+      "loss": 0.4923,
+      "step": 32035
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5396951666558234,
+      "learning_rate": 2.9923406621843034e-06,
+      "loss": 0.7764,
+      "step": 32036
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4846668549593827,
+      "learning_rate": 2.9917963357531112e-06,
+      "loss": 0.5894,
+      "step": 32037
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2338667708931075,
+      "learning_rate": 2.9912520501258958e-06,
+      "loss": 0.57,
+      "step": 32038
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5226948915739924,
+      "learning_rate": 2.9907078053058325e-06,
+      "loss": 0.6607,
+      "step": 32039
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.6601203998093146,
+      "learning_rate": 2.9901636012960856e-06,
+      "loss": 0.677,
+      "step": 32040
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5765166844250561,
+      "learning_rate": 2.9896194380998276e-06,
+      "loss": 0.5044,
+      "step": 32041
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2715987674420552,
+      "learning_rate": 2.989075315720222e-06,
+      "loss": 0.6431,
+      "step": 32042
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6198360719124216,
+      "learning_rate": 2.988531234160443e-06,
+      "loss": 0.5207,
+      "step": 32043
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3042394649268987,
+      "learning_rate": 2.9879871934236525e-06,
+      "loss": 0.7471,
+      "step": 32044
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3065990809896806,
+      "learning_rate": 2.9874431935130223e-06,
+      "loss": 0.7199,
+      "step": 32045
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4274819716100569,
+      "learning_rate": 2.9868992344317184e-06,
+      "loss": 0.8354,
+      "step": 32046
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6706350085063715,
+      "learning_rate": 2.986355316182905e-06,
+      "loss": 0.5211,
+      "step": 32047
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4638804614319842,
+      "learning_rate": 2.98581143876975e-06,
+      "loss": 0.6519,
+      "step": 32048
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.599016491908037,
+      "learning_rate": 2.985267602195426e-06,
+      "loss": 0.6752,
+      "step": 32049
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3879471801301158,
+      "learning_rate": 2.9847238064630947e-06,
+      "loss": 0.7611,
+      "step": 32050
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2223143405405414,
+      "learning_rate": 2.984180051575918e-06,
+      "loss": 0.7527,
+      "step": 32051
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3088293523235108,
+      "learning_rate": 2.983636337537067e-06,
+      "loss": 0.6545,
+      "step": 32052
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.128390461037248,
+      "learning_rate": 2.9830926643497106e-06,
+      "loss": 0.6888,
+      "step": 32053
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4996100581041625,
+      "learning_rate": 2.982549032017009e-06,
+      "loss": 0.6185,
+      "step": 32054
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.518912750039643,
+      "learning_rate": 2.9820054405421263e-06,
+      "loss": 0.594,
+      "step": 32055
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3762474453767277,
+      "learning_rate": 2.981461889928229e-06,
+      "loss": 0.7583,
+      "step": 32056
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.521648039528561,
+      "learning_rate": 2.980918380178486e-06,
+      "loss": 0.7085,
+      "step": 32057
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4191600879720865,
+      "learning_rate": 2.980374911296058e-06,
+      "loss": 0.6942,
+      "step": 32058
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6058515682897198,
+      "learning_rate": 2.979831483284107e-06,
+      "loss": 0.5219,
+      "step": 32059
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4094958361302332,
+      "learning_rate": 2.9792880961458028e-06,
+      "loss": 0.6806,
+      "step": 32060
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5323429614293136,
+      "learning_rate": 2.9787447498843024e-06,
+      "loss": 0.6136,
+      "step": 32061
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3532541141065342,
+      "learning_rate": 2.9782014445027774e-06,
+      "loss": 0.6521,
+      "step": 32062
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.321436828397587,
+      "learning_rate": 2.9776581800043823e-06,
+      "loss": 0.6321,
+      "step": 32063
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5559889452247462,
+      "learning_rate": 2.9771149563922884e-06,
+      "loss": 0.8033,
+      "step": 32064
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.45950830225385,
+      "learning_rate": 2.9765717736696507e-06,
+      "loss": 0.5686,
+      "step": 32065
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5122649589459227,
+      "learning_rate": 2.9760286318396394e-06,
+      "loss": 0.7319,
+      "step": 32066
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4047424020184736,
+      "learning_rate": 2.975485530905413e-06,
+      "loss": 0.7148,
+      "step": 32067
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9091207505846373,
+      "learning_rate": 2.9749424708701313e-06,
+      "loss": 0.6237,
+      "step": 32068
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8118277186631997,
+      "learning_rate": 2.9743994517369577e-06,
+      "loss": 0.7213,
+      "step": 32069
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6074459609199288,
+      "learning_rate": 2.973856473509058e-06,
+      "loss": 0.5328,
+      "step": 32070
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6121303884546186,
+      "learning_rate": 2.9733135361895905e-06,
+      "loss": 0.5215,
+      "step": 32071
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9227300625165649,
+      "learning_rate": 2.9727706397817136e-06,
+      "loss": 0.7331,
+      "step": 32072
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5036559006853438,
+      "learning_rate": 2.9722277842885893e-06,
+      "loss": 0.7153,
+      "step": 32073
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.511666066690793,
+      "learning_rate": 2.971684969713384e-06,
+      "loss": 0.7094,
+      "step": 32074
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6647499375220767,
+      "learning_rate": 2.9711421960592546e-06,
+      "loss": 0.7319,
+      "step": 32075
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3529543570231841,
+      "learning_rate": 2.9705994633293567e-06,
+      "loss": 0.7079,
+      "step": 32076
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.55746166017607,
+      "learning_rate": 2.9700567715268546e-06,
+      "loss": 0.5127,
+      "step": 32077
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9779823409220108,
+      "learning_rate": 2.9695141206549107e-06,
+      "loss": 0.6844,
+      "step": 32078
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2581754871820092,
+      "learning_rate": 2.9689715107166806e-06,
+      "loss": 0.6175,
+      "step": 32079
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4861756112899172,
+      "learning_rate": 2.968428941715321e-06,
+      "loss": 0.5559,
+      "step": 32080
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.427946474890135,
+      "learning_rate": 2.9678864136539985e-06,
+      "loss": 0.6947,
+      "step": 32081
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6833575219227042,
+      "learning_rate": 2.967343926535864e-06,
+      "loss": 0.7398,
+      "step": 32082
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7016607046297467,
+      "learning_rate": 2.966801480364083e-06,
+      "loss": 0.726,
+      "step": 32083
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.36400799020803,
+      "learning_rate": 2.9662590751418075e-06,
+      "loss": 0.7705,
+      "step": 32084
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2173732843011635,
+      "learning_rate": 2.9657167108722018e-06,
+      "loss": 0.6159,
+      "step": 32085
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5020123302932673,
+      "learning_rate": 2.965174387558416e-06,
+      "loss": 0.6199,
+      "step": 32086
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2798921891744839,
+      "learning_rate": 2.9646321052036164e-06,
+      "loss": 0.7829,
+      "step": 32087
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4316966015712178,
+      "learning_rate": 2.964089863810956e-06,
+      "loss": 0.7061,
+      "step": 32088
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5993021171925366,
+      "learning_rate": 2.9635476633835893e-06,
+      "loss": 0.5166,
+      "step": 32089
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5017748815376013,
+      "learning_rate": 2.963005503924674e-06,
+      "loss": 0.7147,
+      "step": 32090
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.708439643334882,
+      "learning_rate": 2.962463385437373e-06,
+      "loss": 0.6695,
+      "step": 32091
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2721421279818605,
+      "learning_rate": 2.961921307924838e-06,
+      "loss": 0.7172,
+      "step": 32092
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5745144211719238,
+      "learning_rate": 2.961379271390222e-06,
+      "loss": 0.614,
+      "step": 32093
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9108752323701275,
+      "learning_rate": 2.9608372758366842e-06,
+      "loss": 0.6616,
+      "step": 32094
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4109024382838986,
+      "learning_rate": 2.9602953212673813e-06,
+      "loss": 0.7014,
+      "step": 32095
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3136069160096933,
+      "learning_rate": 2.959753407685473e-06,
+      "loss": 0.7009,
+      "step": 32096
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5695547897182313,
+      "learning_rate": 2.9592115350941032e-06,
+      "loss": 0.49,
+      "step": 32097
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6353056910987194,
+      "learning_rate": 2.958669703496434e-06,
+      "loss": 0.6857,
+      "step": 32098
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.310749730937244,
+      "learning_rate": 2.958127912895621e-06,
+      "loss": 0.7055,
+      "step": 32099
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0510175352271927,
+      "learning_rate": 2.9575861632948168e-06,
+      "loss": 0.6651,
+      "step": 32100
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3064193087111173,
+      "learning_rate": 2.957044454697172e-06,
+      "loss": 0.7295,
+      "step": 32101
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3799337922482868,
+      "learning_rate": 2.956502787105845e-06,
+      "loss": 0.6464,
+      "step": 32102
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3851918753557142,
+      "learning_rate": 2.9559611605239914e-06,
+      "loss": 0.715,
+      "step": 32103
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3949260391080496,
+      "learning_rate": 2.955419574954762e-06,
+      "loss": 0.6571,
+      "step": 32104
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7385490410113065,
+      "learning_rate": 2.954878030401308e-06,
+      "loss": 0.6801,
+      "step": 32105
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6625237839730913,
+      "learning_rate": 2.9543365268667866e-06,
+      "loss": 0.6436,
+      "step": 32106
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.598758963074494,
+      "learning_rate": 2.9537950643543455e-06,
+      "loss": 0.6185,
+      "step": 32107
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3791342740954005,
+      "learning_rate": 2.953253642867141e-06,
+      "loss": 0.7525,
+      "step": 32108
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.338864214350935,
+      "learning_rate": 2.952712262408327e-06,
+      "loss": 0.7115,
+      "step": 32109
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.398568424016782,
+      "learning_rate": 2.9521709229810535e-06,
+      "loss": 0.6012,
+      "step": 32110
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8266885713683423,
+      "learning_rate": 2.951629624588469e-06,
+      "loss": 0.6369,
+      "step": 32111
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6000444925617727,
+      "learning_rate": 2.951088367233729e-06,
+      "loss": 0.6511,
+      "step": 32112
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5574057634171534,
+      "learning_rate": 2.95054715091999e-06,
+      "loss": 0.6597,
+      "step": 32113
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2461832311738443,
+      "learning_rate": 2.9500059756503907e-06,
+      "loss": 0.6963,
+      "step": 32114
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2402178382146523,
+      "learning_rate": 2.949464841428089e-06,
+      "loss": 0.6834,
+      "step": 32115
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6386520465896518,
+      "learning_rate": 2.948923748256236e-06,
+      "loss": 0.512,
+      "step": 32116
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4983250421676853,
+      "learning_rate": 2.9483826961379837e-06,
+      "loss": 0.6417,
+      "step": 32117
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2720798665075501,
+      "learning_rate": 2.947841685076479e-06,
+      "loss": 0.6253,
+      "step": 32118
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4505819246954048,
+      "learning_rate": 2.94730071507487e-06,
+      "loss": 0.7041,
+      "step": 32119
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4846670498615686,
+      "learning_rate": 2.946759786136313e-06,
+      "loss": 0.6018,
+      "step": 32120
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5854653296916483,
+      "learning_rate": 2.9462188982639494e-06,
+      "loss": 0.7464,
+      "step": 32121
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3093514230620238,
+      "learning_rate": 2.9456780514609364e-06,
+      "loss": 0.6344,
+      "step": 32122
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3915499623461207,
+      "learning_rate": 2.9451372457304163e-06,
+      "loss": 0.5934,
+      "step": 32123
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3718916438986857,
+      "learning_rate": 2.944596481075542e-06,
+      "loss": 0.6217,
+      "step": 32124
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3777615388528102,
+      "learning_rate": 2.94405575749946e-06,
+      "loss": 0.5921,
+      "step": 32125
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5936583137963082,
+      "learning_rate": 2.943515075005321e-06,
+      "loss": 0.4818,
+      "step": 32126
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4008008020436529,
+      "learning_rate": 2.9429744335962717e-06,
+      "loss": 0.7256,
+      "step": 32127
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5882119105790724,
+      "learning_rate": 2.9424338332754553e-06,
+      "loss": 0.6495,
+      "step": 32128
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4206008470765044,
+      "learning_rate": 2.941893274046025e-06,
+      "loss": 0.5209,
+      "step": 32129
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.543627880167219,
+      "learning_rate": 2.941352755911129e-06,
+      "loss": 0.472,
+      "step": 32130
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7725395937467,
+      "learning_rate": 2.9408122788739124e-06,
+      "loss": 0.6273,
+      "step": 32131
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.341975437728857,
+      "learning_rate": 2.940271842937518e-06,
+      "loss": 0.5899,
+      "step": 32132
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2656896106613864,
+      "learning_rate": 2.9397314481050953e-06,
+      "loss": 0.7981,
+      "step": 32133
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2970141293020923,
+      "learning_rate": 2.9391910943797963e-06,
+      "loss": 0.7166,
+      "step": 32134
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.317871474665364,
+      "learning_rate": 2.93865078176476e-06,
+      "loss": 0.6844,
+      "step": 32135
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3137444356059784,
+      "learning_rate": 2.9381105102631334e-06,
+      "loss": 0.6718,
+      "step": 32136
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3949853618273849,
+      "learning_rate": 2.9375702798780626e-06,
+      "loss": 0.6755,
+      "step": 32137
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5602668566948135,
+      "learning_rate": 2.937030090612696e-06,
+      "loss": 0.7158,
+      "step": 32138
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3202702292368502,
+      "learning_rate": 2.9364899424701765e-06,
+      "loss": 0.5237,
+      "step": 32139
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.551750549850219,
+      "learning_rate": 2.9359498354536465e-06,
+      "loss": 0.6493,
+      "step": 32140
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5279081796864702,
+      "learning_rate": 2.935409769566255e-06,
+      "loss": 0.6085,
+      "step": 32141
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6358063613967815,
+      "learning_rate": 2.934869744811142e-06,
+      "loss": 0.6845,
+      "step": 32142
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5880496432899933,
+      "learning_rate": 2.934329761191457e-06,
+      "loss": 0.4315,
+      "step": 32143
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3256413202888297,
+      "learning_rate": 2.9337898187103374e-06,
+      "loss": 0.598,
+      "step": 32144
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6950658384220174,
+      "learning_rate": 2.933249917370935e-06,
+      "loss": 0.7055,
+      "step": 32145
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.6934603343262076,
+      "learning_rate": 2.9327100571763843e-06,
+      "loss": 0.7256,
+      "step": 32146
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4947118456813206,
+      "learning_rate": 2.9321702381298378e-06,
+      "loss": 0.6816,
+      "step": 32147
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1899542493979718,
+      "learning_rate": 2.9316304602344324e-06,
+      "loss": 0.6475,
+      "step": 32148
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.733911749678374,
+      "learning_rate": 2.9310907234933104e-06,
+      "loss": 0.6981,
+      "step": 32149
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3033174857214225,
+      "learning_rate": 2.930551027909615e-06,
+      "loss": 0.6922,
+      "step": 32150
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7741724806980124,
+      "learning_rate": 2.930011373486494e-06,
+      "loss": 0.6918,
+      "step": 32151
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6104430636721685,
+      "learning_rate": 2.9294717602270837e-06,
+      "loss": 0.5373,
+      "step": 32152
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1954537426960383,
+      "learning_rate": 2.9289321881345257e-06,
+      "loss": 0.6617,
+      "step": 32153
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1677729748596426,
+      "learning_rate": 2.9283926572119625e-06,
+      "loss": 0.5397,
+      "step": 32154
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3836421068062972,
+      "learning_rate": 2.927853167462539e-06,
+      "loss": 0.701,
+      "step": 32155
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6065795424038667,
+      "learning_rate": 2.927313718889393e-06,
+      "loss": 0.7171,
+      "step": 32156
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4029863744018647,
+      "learning_rate": 2.9267743114956635e-06,
+      "loss": 0.6453,
+      "step": 32157
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5507658356030783,
+      "learning_rate": 2.926234945284493e-06,
+      "loss": 0.4799,
+      "step": 32158
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4389558726263432,
+      "learning_rate": 2.9256956202590248e-06,
+      "loss": 0.6609,
+      "step": 32159
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4895143277540228,
+      "learning_rate": 2.9251563364223967e-06,
+      "loss": 0.6242,
+      "step": 32160
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7775105839545735,
+      "learning_rate": 2.924617093777745e-06,
+      "loss": 0.6383,
+      "step": 32161
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.705328255484485,
+      "learning_rate": 2.9240778923282163e-06,
+      "loss": 0.7031,
+      "step": 32162
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3751787650501335,
+      "learning_rate": 2.9235387320769427e-06,
+      "loss": 0.7643,
+      "step": 32163
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.387257871737623,
+      "learning_rate": 2.92299961302707e-06,
+      "loss": 0.6754,
+      "step": 32164
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5962608914289547,
+      "learning_rate": 2.9224605351817314e-06,
+      "loss": 0.5042,
+      "step": 32165
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2382161038293313,
+      "learning_rate": 2.9219214985440715e-06,
+      "loss": 0.6746,
+      "step": 32166
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.698307377783601,
+      "learning_rate": 2.9213825031172216e-06,
+      "loss": 0.6462,
+      "step": 32167
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3224447598421716,
+      "learning_rate": 2.9208435489043276e-06,
+      "loss": 0.6517,
+      "step": 32168
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.60081118233207,
+      "learning_rate": 2.920304635908524e-06,
+      "loss": 0.7172,
+      "step": 32169
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.331070199835457,
+      "learning_rate": 2.9197657641329447e-06,
+      "loss": 0.7337,
+      "step": 32170
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5488174505455208,
+      "learning_rate": 2.91922693358073e-06,
+      "loss": 0.8686,
+      "step": 32171
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3164925808823884,
+      "learning_rate": 2.918688144255022e-06,
+      "loss": 0.5783,
+      "step": 32172
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.419222057598562,
+      "learning_rate": 2.9181493961589526e-06,
+      "loss": 0.7166,
+      "step": 32173
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5943057219886938,
+      "learning_rate": 2.917610689295657e-06,
+      "loss": 0.4528,
+      "step": 32174
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.394043833455504,
+      "learning_rate": 2.9170720236682737e-06,
+      "loss": 0.6361,
+      "step": 32175
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4691714606655135,
+      "learning_rate": 2.9165333992799428e-06,
+      "loss": 0.5874,
+      "step": 32176
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4348127972974791,
+      "learning_rate": 2.915994816133797e-06,
+      "loss": 0.7623,
+      "step": 32177
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5034646046927338,
+      "learning_rate": 2.9154562742329695e-06,
+      "loss": 0.7489,
+      "step": 32178
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6323693718443373,
+      "learning_rate": 2.9149177735805977e-06,
+      "loss": 0.7531,
+      "step": 32179
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6009271417044197,
+      "learning_rate": 2.9143793141798214e-06,
+      "loss": 0.5158,
+      "step": 32180
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2526317736653043,
+      "learning_rate": 2.9138408960337725e-06,
+      "loss": 0.5929,
+      "step": 32181
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3327731307568078,
+      "learning_rate": 2.913302519145582e-06,
+      "loss": 0.6858,
+      "step": 32182
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4108048552811572,
+      "learning_rate": 2.9127641835183906e-06,
+      "loss": 0.6125,
+      "step": 32183
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.519570782994211,
+      "learning_rate": 2.912225889155327e-06,
+      "loss": 0.6795,
+      "step": 32184
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.358090767798762,
+      "learning_rate": 2.9116876360595304e-06,
+      "loss": 0.7229,
+      "step": 32185
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.726269071539211,
+      "learning_rate": 2.9111494242341297e-06,
+      "loss": 0.6421,
+      "step": 32186
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.353089120134127,
+      "learning_rate": 2.910611253682265e-06,
+      "loss": 0.6772,
+      "step": 32187
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4972261209259476,
+      "learning_rate": 2.9100731244070635e-06,
+      "loss": 0.7133,
+      "step": 32188
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5757873832097148,
+      "learning_rate": 2.909535036411664e-06,
+      "loss": 0.4705,
+      "step": 32189
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.445331418491118,
+      "learning_rate": 2.9089969896991956e-06,
+      "loss": 0.6004,
+      "step": 32190
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.194010324607409,
+      "learning_rate": 2.9084589842727883e-06,
+      "loss": 0.7202,
+      "step": 32191
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4422470370495888,
+      "learning_rate": 2.907921020135578e-06,
+      "loss": 0.7441,
+      "step": 32192
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4111343555891616,
+      "learning_rate": 2.907383097290697e-06,
+      "loss": 0.6084,
+      "step": 32193
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9617677054754221,
+      "learning_rate": 2.906845215741283e-06,
+      "loss": 0.7403,
+      "step": 32194
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.467199709630997,
+      "learning_rate": 2.906307375490457e-06,
+      "loss": 0.6891,
+      "step": 32195
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4128963139989374,
+      "learning_rate": 2.905769576541354e-06,
+      "loss": 0.7282,
+      "step": 32196
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4757218514970187,
+      "learning_rate": 2.9052318188971097e-06,
+      "loss": 0.7511,
+      "step": 32197
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3702148475787839,
+      "learning_rate": 2.904694102560852e-06,
+      "loss": 0.6495,
+      "step": 32198
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.51690799697887,
+      "learning_rate": 2.904156427535709e-06,
+      "loss": 0.7569,
+      "step": 32199
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3091267565686278,
+      "learning_rate": 2.9036187938248126e-06,
+      "loss": 0.6836,
+      "step": 32200
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5141882613571271,
+      "learning_rate": 2.903081201431299e-06,
+      "loss": 0.6912,
+      "step": 32201
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4538228192153202,
+      "learning_rate": 2.9025436503582893e-06,
+      "loss": 0.6954,
+      "step": 32202
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8565468549126507,
+      "learning_rate": 2.9020061406089207e-06,
+      "loss": 0.653,
+      "step": 32203
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3812492132273146,
+      "learning_rate": 2.9014686721863192e-06,
+      "loss": 0.727,
+      "step": 32204
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3966721922987473,
+      "learning_rate": 2.900931245093612e-06,
+      "loss": 0.6972,
+      "step": 32205
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.647053072571321,
+      "learning_rate": 2.900393859333931e-06,
+      "loss": 0.7816,
+      "step": 32206
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4287617838489464,
+      "learning_rate": 2.8998565149104075e-06,
+      "loss": 0.6842,
+      "step": 32207
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8236112076857525,
+      "learning_rate": 2.899319211826167e-06,
+      "loss": 0.6563,
+      "step": 32208
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3879491665092618,
+      "learning_rate": 2.898781950084335e-06,
+      "loss": 0.6232,
+      "step": 32209
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5419429889601832,
+      "learning_rate": 2.8982447296880423e-06,
+      "loss": 0.6743,
+      "step": 32210
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6681987897184716,
+      "learning_rate": 2.897707550640424e-06,
+      "loss": 0.7122,
+      "step": 32211
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3400773236755832,
+      "learning_rate": 2.8971704129445944e-06,
+      "loss": 0.6673,
+      "step": 32212
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4225724452103417,
+      "learning_rate": 2.8966333166036864e-06,
+      "loss": 0.7234,
+      "step": 32213
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.215083512304457,
+      "learning_rate": 2.89609626162083e-06,
+      "loss": 0.5948,
+      "step": 32214
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9139879654949885,
+      "learning_rate": 2.8955592479991524e-06,
+      "loss": 0.7048,
+      "step": 32215
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5063709257718927,
+      "learning_rate": 2.895022275741778e-06,
+      "loss": 0.698,
+      "step": 32216
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2450072531751963,
+      "learning_rate": 2.8944853448518295e-06,
+      "loss": 0.6403,
+      "step": 32217
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1599135029197176,
+      "learning_rate": 2.893948455332438e-06,
+      "loss": 0.659,
+      "step": 32218
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.846122881903822,
+      "learning_rate": 2.8934116071867314e-06,
+      "loss": 0.7558,
+      "step": 32219
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3500723775972703,
+      "learning_rate": 2.892874800417832e-06,
+      "loss": 0.6878,
+      "step": 32220
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6502354927766456,
+      "learning_rate": 2.892338035028862e-06,
+      "loss": 0.6456,
+      "step": 32221
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4809850683079295,
+      "learning_rate": 2.8918013110229535e-06,
+      "loss": 0.7497,
+      "step": 32222
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4935187238687846,
+      "learning_rate": 2.8912646284032264e-06,
+      "loss": 0.6176,
+      "step": 32223
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.365877097536588,
+      "learning_rate": 2.890727987172809e-06,
+      "loss": 0.7453,
+      "step": 32224
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6249117173649072,
+      "learning_rate": 2.8901913873348208e-06,
+      "loss": 0.5991,
+      "step": 32225
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5019554327598572,
+      "learning_rate": 2.889654828892393e-06,
+      "loss": 0.6664,
+      "step": 32226
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5026824341632512,
+      "learning_rate": 2.889118311848643e-06,
+      "loss": 0.5119,
+      "step": 32227
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.703422280587533,
+      "learning_rate": 2.8885818362066997e-06,
+      "loss": 0.6173,
+      "step": 32228
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2194092955229228,
+      "learning_rate": 2.8880454019696846e-06,
+      "loss": 0.6751,
+      "step": 32229
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9063535005947794,
+      "learning_rate": 2.8875090091407187e-06,
+      "loss": 0.6674,
+      "step": 32230
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4354511496581797,
+      "learning_rate": 2.8869726577229264e-06,
+      "loss": 0.6654,
+      "step": 32231
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8666685383279152,
+      "learning_rate": 2.8864363477194336e-06,
+      "loss": 0.5305,
+      "step": 32232
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3320593026044503,
+      "learning_rate": 2.8859000791333613e-06,
+      "loss": 0.7149,
+      "step": 32233
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2359066293679106,
+      "learning_rate": 2.8853638519678285e-06,
+      "loss": 0.7759,
+      "step": 32234
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8587758429287826,
+      "learning_rate": 2.884827666225959e-06,
+      "loss": 0.7344,
+      "step": 32235
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2618450586644132,
+      "learning_rate": 2.884291521910878e-06,
+      "loss": 0.65,
+      "step": 32236
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5007846085527514,
+      "learning_rate": 2.8837554190257057e-06,
+      "loss": 0.621,
+      "step": 32237
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.6686865815003387,
+      "learning_rate": 2.883219357573559e-06,
+      "loss": 0.6754,
+      "step": 32238
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3105418160682178,
+      "learning_rate": 2.8826833375575615e-06,
+      "loss": 0.5867,
+      "step": 32239
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.250295799581945,
+      "learning_rate": 2.882147358980839e-06,
+      "loss": 0.5906,
+      "step": 32240
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.663327774705276,
+      "learning_rate": 2.8816114218465073e-06,
+      "loss": 0.76,
+      "step": 32241
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.698982753744762,
+      "learning_rate": 2.881075526157685e-06,
+      "loss": 0.7915,
+      "step": 32242
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6159339465564115,
+      "learning_rate": 2.880539671917497e-06,
+      "loss": 0.7233,
+      "step": 32243
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.299381697342824,
+      "learning_rate": 2.880003859129059e-06,
+      "loss": 0.6109,
+      "step": 32244
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4217036706795947,
+      "learning_rate": 2.879468087795495e-06,
+      "loss": 0.6748,
+      "step": 32245
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5517030418427329,
+      "learning_rate": 2.87893235791992e-06,
+      "loss": 0.4345,
+      "step": 32246
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.02380614101802,
+      "learning_rate": 2.8783966695054575e-06,
+      "loss": 0.8794,
+      "step": 32247
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5257284093644672,
+      "learning_rate": 2.877861022555222e-06,
+      "loss": 0.6098,
+      "step": 32248
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3655927724052936,
+      "learning_rate": 2.8773254170723363e-06,
+      "loss": 0.6834,
+      "step": 32249
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5095419370839542,
+      "learning_rate": 2.876789853059918e-06,
+      "loss": 0.7204,
+      "step": 32250
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4466063962699667,
+      "learning_rate": 2.876254330521081e-06,
+      "loss": 0.6645,
+      "step": 32251
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4436392924337267,
+      "learning_rate": 2.8757188494589474e-06,
+      "loss": 0.7781,
+      "step": 32252
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.520415211146554,
+      "learning_rate": 2.875183409876636e-06,
+      "loss": 0.7418,
+      "step": 32253
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2459673589476592,
+      "learning_rate": 2.8746480117772626e-06,
+      "loss": 0.7134,
+      "step": 32254
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5960149623275901,
+      "learning_rate": 2.874112655163942e-06,
+      "loss": 0.4708,
+      "step": 32255
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3963690742872488,
+      "learning_rate": 2.8735773400397926e-06,
+      "loss": 0.7394,
+      "step": 32256
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.384350340506316,
+      "learning_rate": 2.8730420664079363e-06,
+      "loss": 0.7143,
+      "step": 32257
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4790376396148175,
+      "learning_rate": 2.8725068342714847e-06,
+      "loss": 0.5941,
+      "step": 32258
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5485067085065354,
+      "learning_rate": 2.871971643633552e-06,
+      "loss": 0.6835,
+      "step": 32259
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.506485981517759,
+      "learning_rate": 2.871436494497256e-06,
+      "loss": 0.6599,
+      "step": 32260
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3542531327595992,
+      "learning_rate": 2.8709013868657186e-06,
+      "loss": 0.5956,
+      "step": 32261
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5049016608133767,
+      "learning_rate": 2.8703663207420494e-06,
+      "loss": 0.6673,
+      "step": 32262
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3195946236954739,
+      "learning_rate": 2.869831296129362e-06,
+      "loss": 0.9246,
+      "step": 32263
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.319709729568591,
+      "learning_rate": 2.869296313030776e-06,
+      "loss": 0.6472,
+      "step": 32264
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5241427815515254,
+      "learning_rate": 2.8687613714494023e-06,
+      "loss": 0.6393,
+      "step": 32265
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.788224351096233,
+      "learning_rate": 2.868226471388361e-06,
+      "loss": 0.5662,
+      "step": 32266
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4147848293892382,
+      "learning_rate": 2.867691612850759e-06,
+      "loss": 0.5911,
+      "step": 32267
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5791572474598286,
+      "learning_rate": 2.8671567958397183e-06,
+      "loss": 0.4966,
+      "step": 32268
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.54173842796729,
+      "learning_rate": 2.8666220203583462e-06,
+      "loss": 0.6099,
+      "step": 32269
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8930368862007108,
+      "learning_rate": 2.866087286409761e-06,
+      "loss": 0.6401,
+      "step": 32270
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4580194638839712,
+      "learning_rate": 2.865552593997075e-06,
+      "loss": 0.7244,
+      "step": 32271
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0499900387034353,
+      "learning_rate": 2.8650179431233984e-06,
+      "loss": 0.6271,
+      "step": 32272
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3917039423642839,
+      "learning_rate": 2.8644833337918444e-06,
+      "loss": 0.6904,
+      "step": 32273
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.478501423091738,
+      "learning_rate": 2.863948766005531e-06,
+      "loss": 0.5672,
+      "step": 32274
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7347087643669015,
+      "learning_rate": 2.8634142397675678e-06,
+      "loss": 0.6152,
+      "step": 32275
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5780148603094997,
+      "learning_rate": 2.862879755081063e-06,
+      "loss": 0.5346,
+      "step": 32276
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4000452748363545,
+      "learning_rate": 2.8623453119491316e-06,
+      "loss": 0.6606,
+      "step": 32277
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5694373901670797,
+      "learning_rate": 2.861810910374888e-06,
+      "loss": 0.4867,
+      "step": 32278
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.966155102887176,
+      "learning_rate": 2.8612765503614424e-06,
+      "loss": 0.5534,
+      "step": 32279
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2212855405438718,
+      "learning_rate": 2.8607422319119017e-06,
+      "loss": 0.7332,
+      "step": 32280
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.487715288577744,
+      "learning_rate": 2.8602079550293792e-06,
+      "loss": 0.71,
+      "step": 32281
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5952686811908855,
+      "learning_rate": 2.85967371971699e-06,
+      "loss": 0.544,
+      "step": 32282
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1783072091482263,
+      "learning_rate": 2.859139525977841e-06,
+      "loss": 0.6066,
+      "step": 32283
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3080153483174726,
+      "learning_rate": 2.8586053738150388e-06,
+      "loss": 0.621,
+      "step": 32284
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4016646412491993,
+      "learning_rate": 2.858071263231701e-06,
+      "loss": 0.5719,
+      "step": 32285
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.785049479129123,
+      "learning_rate": 2.8575371942309293e-06,
+      "loss": 0.574,
+      "step": 32286
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6842486391436284,
+      "learning_rate": 2.8570031668158403e-06,
+      "loss": 0.768,
+      "step": 32287
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3700694056633058,
+      "learning_rate": 2.8564691809895383e-06,
+      "loss": 0.6453,
+      "step": 32288
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2211483550652928,
+      "learning_rate": 2.8559352367551375e-06,
+      "loss": 0.7384,
+      "step": 32289
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4088040172669036,
+      "learning_rate": 2.855401334115739e-06,
+      "loss": 0.598,
+      "step": 32290
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.012309745959289,
+      "learning_rate": 2.8548674730744576e-06,
+      "loss": 0.7137,
+      "step": 32291
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6177446590586815,
+      "learning_rate": 2.8543336536344047e-06,
+      "loss": 0.7517,
+      "step": 32292
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5185438816377583,
+      "learning_rate": 2.8537998757986773e-06,
+      "loss": 0.6691,
+      "step": 32293
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2778526968810004,
+      "learning_rate": 2.853266139570391e-06,
+      "loss": 0.7461,
+      "step": 32294
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6686999093373955,
+      "learning_rate": 2.8527324449526505e-06,
+      "loss": 0.7071,
+      "step": 32295
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.679729981070627,
+      "learning_rate": 2.8521987919485705e-06,
+      "loss": 0.7009,
+      "step": 32296
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5143958487698608,
+      "learning_rate": 2.851665180561247e-06,
+      "loss": 0.7364,
+      "step": 32297
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5858696311521739,
+      "learning_rate": 2.8511316107937912e-06,
+      "loss": 0.4811,
+      "step": 32298
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5834089355173039,
+      "learning_rate": 2.850598082649314e-06,
+      "loss": 0.494,
+      "step": 32299
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5892492485650934,
+      "learning_rate": 2.850064596130915e-06,
+      "loss": 0.4949,
+      "step": 32300
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6096493155352486,
+      "learning_rate": 2.849531151241707e-06,
+      "loss": 0.5293,
+      "step": 32301
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4416975920877373,
+      "learning_rate": 2.8489977479847886e-06,
+      "loss": 0.6725,
+      "step": 32302
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5514612545430049,
+      "learning_rate": 2.8484643863632722e-06,
+      "loss": 0.6907,
+      "step": 32303
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6275927348858131,
+      "learning_rate": 2.8479310663802574e-06,
+      "loss": 0.4992,
+      "step": 32304
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3209189246896362,
+      "learning_rate": 2.847397788038856e-06,
+      "loss": 0.7777,
+      "step": 32305
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6268710436306721,
+      "learning_rate": 2.8468645513421688e-06,
+      "loss": 0.7158,
+      "step": 32306
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.346580039651113,
+      "learning_rate": 2.8463313562932983e-06,
+      "loss": 0.5261,
+      "step": 32307
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3594518981790442,
+      "learning_rate": 2.8457982028953502e-06,
+      "loss": 0.7573,
+      "step": 32308
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3755133618391284,
+      "learning_rate": 2.8452650911514336e-06,
+      "loss": 0.7203,
+      "step": 32309
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3064552801657012,
+      "learning_rate": 2.844732021064649e-06,
+      "loss": 0.6011,
+      "step": 32310
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.482926362805663,
+      "learning_rate": 2.8441989926380966e-06,
+      "loss": 0.6242,
+      "step": 32311
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5700843038470063,
+      "learning_rate": 2.843666005874883e-06,
+      "loss": 0.7065,
+      "step": 32312
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4414868343982228,
+      "learning_rate": 2.843133060778115e-06,
+      "loss": 0.6892,
+      "step": 32313
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4844656347419336,
+      "learning_rate": 2.842600157350892e-06,
+      "loss": 0.7539,
+      "step": 32314
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3956196725250467,
+      "learning_rate": 2.8420672955963136e-06,
+      "loss": 0.8278,
+      "step": 32315
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.708550903395981,
+      "learning_rate": 2.8415344755174855e-06,
+      "loss": 0.7055,
+      "step": 32316
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3977032326911005,
+      "learning_rate": 2.8410016971175136e-06,
+      "loss": 0.6278,
+      "step": 32317
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4838369285701418,
+      "learning_rate": 2.840468960399496e-06,
+      "loss": 0.6675,
+      "step": 32318
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.368536680589526,
+      "learning_rate": 2.8399362653665318e-06,
+      "loss": 0.6518,
+      "step": 32319
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3273503871393018,
+      "learning_rate": 2.8394036120217296e-06,
+      "loss": 0.6714,
+      "step": 32320
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2601014113917552,
+      "learning_rate": 2.838871000368182e-06,
+      "loss": 0.7454,
+      "step": 32321
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2627465242516998,
+      "learning_rate": 2.8383384304089978e-06,
+      "loss": 0.6517,
+      "step": 32322
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6849046153090286,
+      "learning_rate": 2.8378059021472725e-06,
+      "loss": 0.7354,
+      "step": 32323
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2731873116779342,
+      "learning_rate": 2.8372734155861105e-06,
+      "loss": 0.5976,
+      "step": 32324
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6165092914348054,
+      "learning_rate": 2.8367409707286085e-06,
+      "loss": 0.5948,
+      "step": 32325
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3997098118215907,
+      "learning_rate": 2.8362085675778708e-06,
+      "loss": 0.6926,
+      "step": 32326
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3433637804277072,
+      "learning_rate": 2.8356762061369945e-06,
+      "loss": 0.6863,
+      "step": 32327
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5435794542315238,
+      "learning_rate": 2.8351438864090774e-06,
+      "loss": 0.7289,
+      "step": 32328
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5489291981857423,
+      "learning_rate": 2.834611608397221e-06,
+      "loss": 0.6484,
+      "step": 32329
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6933574554732633,
+      "learning_rate": 2.8340793721045266e-06,
+      "loss": 0.6945,
+      "step": 32330
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4664614762291686,
+      "learning_rate": 2.8335471775340907e-06,
+      "loss": 0.6906,
+      "step": 32331
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3682450973159233,
+      "learning_rate": 2.8330150246890097e-06,
+      "loss": 0.704,
+      "step": 32332
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3928955667054774,
+      "learning_rate": 2.832482913572384e-06,
+      "loss": 0.746,
+      "step": 32333
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5460422821747966,
+      "learning_rate": 2.8319508441873143e-06,
+      "loss": 0.7408,
+      "step": 32334
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.37457013357548,
+      "learning_rate": 2.831418816536897e-06,
+      "loss": 0.6776,
+      "step": 32335
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7498103515760541,
+      "learning_rate": 2.8308868306242253e-06,
+      "loss": 0.6916,
+      "step": 32336
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.558509450404204,
+      "learning_rate": 2.8303548864524012e-06,
+      "loss": 0.7403,
+      "step": 32337
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7463763068417586,
+      "learning_rate": 2.829822984024523e-06,
+      "loss": 0.7782,
+      "step": 32338
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6010742740089746,
+      "learning_rate": 2.829291123343686e-06,
+      "loss": 0.4889,
+      "step": 32339
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3466711356002656,
+      "learning_rate": 2.8287593044129825e-06,
+      "loss": 0.8296,
+      "step": 32340
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8388170091215645,
+      "learning_rate": 2.828227527235513e-06,
+      "loss": 0.6552,
+      "step": 32341
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6784669354698818,
+      "learning_rate": 2.8276957918143766e-06,
+      "loss": 0.6871,
+      "step": 32342
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.245159368116096,
+      "learning_rate": 2.827164098152666e-06,
+      "loss": 0.6624,
+      "step": 32343
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3621158725475966,
+      "learning_rate": 2.8266324462534733e-06,
+      "loss": 0.7132,
+      "step": 32344
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5079714686343835,
+      "learning_rate": 2.8261008361199017e-06,
+      "loss": 0.7717,
+      "step": 32345
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.273322343321819,
+      "learning_rate": 2.8255692677550385e-06,
+      "loss": 0.5295,
+      "step": 32346
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5601686427066562,
+      "learning_rate": 2.8250377411619845e-06,
+      "loss": 0.4989,
+      "step": 32347
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6570065619052305,
+      "learning_rate": 2.8245062563438306e-06,
+      "loss": 0.7082,
+      "step": 32348
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.1618891411255134,
+      "learning_rate": 2.823974813303676e-06,
+      "loss": 0.6515,
+      "step": 32349
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.471233501067118,
+      "learning_rate": 2.8234434120446084e-06,
+      "loss": 0.6221,
+      "step": 32350
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2810899185686295,
+      "learning_rate": 2.8229120525697285e-06,
+      "loss": 0.6831,
+      "step": 32351
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3180091817542674,
+      "learning_rate": 2.8223807348821273e-06,
+      "loss": 0.6405,
+      "step": 32352
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.572731218217417,
+      "learning_rate": 2.821849458984894e-06,
+      "loss": 0.4851,
+      "step": 32353
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3414427815523193,
+      "learning_rate": 2.8213182248811265e-06,
+      "loss": 0.6301,
+      "step": 32354
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4235813587322663,
+      "learning_rate": 2.8207870325739196e-06,
+      "loss": 0.7513,
+      "step": 32355
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.393886022461757,
+      "learning_rate": 2.820255882066364e-06,
+      "loss": 0.706,
+      "step": 32356
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6245666820663786,
+      "learning_rate": 2.8197247733615474e-06,
+      "loss": 0.534,
+      "step": 32357
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0074969313715143,
+      "learning_rate": 2.8191937064625675e-06,
+      "loss": 0.7663,
+      "step": 32358
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.252765476529212,
+      "learning_rate": 2.818662681372518e-06,
+      "loss": 0.7006,
+      "step": 32359
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2416518199498725,
+      "learning_rate": 2.8181316980944884e-06,
+      "loss": 0.6965,
+      "step": 32360
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6224512024190529,
+      "learning_rate": 2.8176007566315665e-06,
+      "loss": 0.5106,
+      "step": 32361
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.366302514421844,
+      "learning_rate": 2.817069856986846e-06,
+      "loss": 0.7572,
+      "step": 32362
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.416290681531474,
+      "learning_rate": 2.816538999163423e-06,
+      "loss": 0.6422,
+      "step": 32363
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5830457351792586,
+      "learning_rate": 2.8160081831643836e-06,
+      "loss": 0.614,
+      "step": 32364
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4704180891020993,
+      "learning_rate": 2.8154774089928172e-06,
+      "loss": 0.7374,
+      "step": 32365
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4669224008600588,
+      "learning_rate": 2.8149466766518175e-06,
+      "loss": 0.6696,
+      "step": 32366
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2977661459641483,
+      "learning_rate": 2.814415986144471e-06,
+      "loss": 0.5508,
+      "step": 32367
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6299769881580317,
+      "learning_rate": 2.8138853374738717e-06,
+      "loss": 0.641,
+      "step": 32368
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6107979965088348,
+      "learning_rate": 2.813354730643104e-06,
+      "loss": 0.5157,
+      "step": 32369
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.516427016584345,
+      "learning_rate": 2.8128241656552637e-06,
+      "loss": 0.7128,
+      "step": 32370
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4906005247659806,
+      "learning_rate": 2.812293642513434e-06,
+      "loss": 0.5907,
+      "step": 32371
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.517465290814093,
+      "learning_rate": 2.8117631612207084e-06,
+      "loss": 0.6871,
+      "step": 32372
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4327049950563657,
+      "learning_rate": 2.811232721780174e-06,
+      "loss": 0.6087,
+      "step": 32373
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6237024812591503,
+      "learning_rate": 2.810702324194915e-06,
+      "loss": 0.5721,
+      "step": 32374
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6475383963957468,
+      "learning_rate": 2.810171968468024e-06,
+      "loss": 0.5046,
+      "step": 32375
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6396219581268707,
+      "learning_rate": 2.809641654602587e-06,
+      "loss": 0.7275,
+      "step": 32376
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.764941297469258,
+      "learning_rate": 2.8091113826017004e-06,
+      "loss": 0.6242,
+      "step": 32377
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4654196293755128,
+      "learning_rate": 2.808581152468436e-06,
+      "loss": 0.6134,
+      "step": 32378
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3282208772063937,
+      "learning_rate": 2.8080509642058894e-06,
+      "loss": 0.5957,
+      "step": 32379
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.35901003914795,
+      "learning_rate": 2.8075208178171507e-06,
+      "loss": 0.7145,
+      "step": 32380
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2589158870729684,
+      "learning_rate": 2.8069907133053033e-06,
+      "loss": 0.5936,
+      "step": 32381
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.206679498085865,
+      "learning_rate": 2.8064606506734292e-06,
+      "loss": 0.6117,
+      "step": 32382
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5976533282403065,
+      "learning_rate": 2.805930629924619e-06,
+      "loss": 0.6128,
+      "step": 32383
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5735056579483919,
+      "learning_rate": 2.8054006510619615e-06,
+      "loss": 0.4778,
+      "step": 32384
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4620328293170064,
+      "learning_rate": 2.804870714088539e-06,
+      "loss": 0.7237,
+      "step": 32385
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.446703255955039,
+      "learning_rate": 2.804340819007434e-06,
+      "loss": 0.7321,
+      "step": 32386
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.277356627773777,
+      "learning_rate": 2.803810965821737e-06,
+      "loss": 0.6586,
+      "step": 32387
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4827774250733552,
+      "learning_rate": 2.8032811545345294e-06,
+      "loss": 0.7104,
+      "step": 32388
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5724485098952977,
+      "learning_rate": 2.8027513851488963e-06,
+      "loss": 0.4519,
+      "step": 32389
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7127017019330129,
+      "learning_rate": 2.802221657667927e-06,
+      "loss": 0.5794,
+      "step": 32390
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5178401584369803,
+      "learning_rate": 2.8016919720947024e-06,
+      "loss": 0.8362,
+      "step": 32391
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6466560047174956,
+      "learning_rate": 2.801162328432302e-06,
+      "loss": 0.6905,
+      "step": 32392
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.680000404596524,
+      "learning_rate": 2.800632726683814e-06,
+      "loss": 0.806,
+      "step": 32393
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.482752852770281,
+      "learning_rate": 2.8001031668523273e-06,
+      "loss": 0.7642,
+      "step": 32394
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6001774965740643,
+      "learning_rate": 2.799573648940913e-06,
+      "loss": 0.5178,
+      "step": 32395
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5993468949824583,
+      "learning_rate": 2.7990441729526617e-06,
+      "loss": 0.7823,
+      "step": 32396
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5111628618104778,
+      "learning_rate": 2.798514738890654e-06,
+      "loss": 0.6992,
+      "step": 32397
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.3596263259635246,
+      "learning_rate": 2.7979853467579777e-06,
+      "loss": 0.6904,
+      "step": 32398
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.591572722222297,
+      "learning_rate": 2.7974559965577097e-06,
+      "loss": 0.6358,
+      "step": 32399
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3809202078192124,
+      "learning_rate": 2.7969266882929304e-06,
+      "loss": 0.6217,
+      "step": 32400
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5633238131379905,
+      "learning_rate": 2.796397421966728e-06,
+      "loss": 0.6471,
+      "step": 32401
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.405266610028165,
+      "learning_rate": 2.7958681975821767e-06,
+      "loss": 0.6779,
+      "step": 32402
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3944338637754197,
+      "learning_rate": 2.795339015142364e-06,
+      "loss": 0.5153,
+      "step": 32403
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.551982586126464,
+      "learning_rate": 2.794809874650367e-06,
+      "loss": 0.6878,
+      "step": 32404
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3809743130843748,
+      "learning_rate": 2.7942807761092704e-06,
+      "loss": 0.7362,
+      "step": 32405
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6442678856274622,
+      "learning_rate": 2.7937517195221498e-06,
+      "loss": 0.6267,
+      "step": 32406
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.33921786497821,
+      "learning_rate": 2.79322270489209e-06,
+      "loss": 0.6298,
+      "step": 32407
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4187431129336687,
+      "learning_rate": 2.7926937322221705e-06,
+      "loss": 0.689,
+      "step": 32408
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.528639673135613,
+      "learning_rate": 2.7921648015154655e-06,
+      "loss": 0.6495,
+      "step": 32409
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9862117175932394,
+      "learning_rate": 2.7916359127750605e-06,
+      "loss": 0.6121,
+      "step": 32410
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2636490029154792,
+      "learning_rate": 2.791107066004035e-06,
+      "loss": 0.7304,
+      "step": 32411
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4600924962969466,
+      "learning_rate": 2.7905782612054676e-06,
+      "loss": 0.5739,
+      "step": 32412
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3327285363190104,
+      "learning_rate": 2.790049498382432e-06,
+      "loss": 0.6635,
+      "step": 32413
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3664886690134985,
+      "learning_rate": 2.7895207775380116e-06,
+      "loss": 0.6518,
+      "step": 32414
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.19511276026283,
+      "learning_rate": 2.788992098675287e-06,
+      "loss": 0.5314,
+      "step": 32415
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5573416187171363,
+      "learning_rate": 2.788463461797333e-06,
+      "loss": 0.7767,
+      "step": 32416
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5124742062248595,
+      "learning_rate": 2.787934866907225e-06,
+      "loss": 0.6763,
+      "step": 32417
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5523563442950497,
+      "learning_rate": 2.787406314008043e-06,
+      "loss": 0.6797,
+      "step": 32418
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.559341924368896,
+      "learning_rate": 2.7868778031028688e-06,
+      "loss": 0.5882,
+      "step": 32419
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.551496485561731,
+      "learning_rate": 2.7863493341947756e-06,
+      "loss": 0.4603,
+      "step": 32420
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.619297507268102,
+      "learning_rate": 2.785820907286837e-06,
+      "loss": 0.6083,
+      "step": 32421
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7117015632918562,
+      "learning_rate": 2.785292522382137e-06,
+      "loss": 0.6849,
+      "step": 32422
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3968292862222396,
+      "learning_rate": 2.7847641794837432e-06,
+      "loss": 0.6738,
+      "step": 32423
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.43524935728622,
+      "learning_rate": 2.784235878594741e-06,
+      "loss": 0.741,
+      "step": 32424
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6969563823604294,
+      "learning_rate": 2.7837076197181987e-06,
+      "loss": 0.7007,
+      "step": 32425
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3065057100266788,
+      "learning_rate": 2.783179402857198e-06,
+      "loss": 0.8704,
+      "step": 32426
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5917182775247731,
+      "learning_rate": 2.782651228014809e-06,
+      "loss": 0.6182,
+      "step": 32427
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2768847961663663,
+      "learning_rate": 2.7821230951941125e-06,
+      "loss": 0.7635,
+      "step": 32428
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6392748607806915,
+      "learning_rate": 2.781595004398181e-06,
+      "loss": 0.7168,
+      "step": 32429
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4948143472182482,
+      "learning_rate": 2.781066955630085e-06,
+      "loss": 0.7385,
+      "step": 32430
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.53282710040187,
+      "learning_rate": 2.7805389488929025e-06,
+      "loss": 0.6051,
+      "step": 32431
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4449956017112366,
+      "learning_rate": 2.7800109841897105e-06,
+      "loss": 0.5215,
+      "step": 32432
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.4233212638443815,
+      "learning_rate": 2.7794830615235812e-06,
+      "loss": 0.7353,
+      "step": 32433
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5196606300009514,
+      "learning_rate": 2.7789551808975843e-06,
+      "loss": 0.7116,
+      "step": 32434
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3852121131305952,
+      "learning_rate": 2.7784273423147966e-06,
+      "loss": 0.6598,
+      "step": 32435
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5727834926366866,
+      "learning_rate": 2.777899545778294e-06,
+      "loss": 0.5939,
+      "step": 32436
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4430689350715447,
+      "learning_rate": 2.777371791291147e-06,
+      "loss": 0.638,
+      "step": 32437
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.485595230009996,
+      "learning_rate": 2.7768440788564243e-06,
+      "loss": 0.6103,
+      "step": 32438
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2771960202653578,
+      "learning_rate": 2.7763164084772033e-06,
+      "loss": 0.6492,
+      "step": 32439
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2685639425212474,
+      "learning_rate": 2.7757887801565577e-06,
+      "loss": 0.5829,
+      "step": 32440
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4227562838420682,
+      "learning_rate": 2.7752611938975573e-06,
+      "loss": 0.654,
+      "step": 32441
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4538448492542695,
+      "learning_rate": 2.7747336497032697e-06,
+      "loss": 0.5273,
+      "step": 32442
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.683330864011821,
+      "learning_rate": 2.7742061475767735e-06,
+      "loss": 0.6983,
+      "step": 32443
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3070963772838893,
+      "learning_rate": 2.773678687521134e-06,
+      "loss": 0.652,
+      "step": 32444
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3694051660208157,
+      "learning_rate": 2.773151269539428e-06,
+      "loss": 0.6229,
+      "step": 32445
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4446616571732775,
+      "learning_rate": 2.7726238936347204e-06,
+      "loss": 0.6024,
+      "step": 32446
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3261430477868514,
+      "learning_rate": 2.7720965598100867e-06,
+      "loss": 0.7229,
+      "step": 32447
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4730188908979838,
+      "learning_rate": 2.7715692680685933e-06,
+      "loss": 0.7057,
+      "step": 32448
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2244638004329833,
+      "learning_rate": 2.7710420184133148e-06,
+      "loss": 0.6076,
+      "step": 32449
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1481711526913203,
+      "learning_rate": 2.7705148108473177e-06,
+      "loss": 0.6942,
+      "step": 32450
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2420403150882748,
+      "learning_rate": 2.769987645373671e-06,
+      "loss": 0.7,
+      "step": 32451
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8597954647781485,
+      "learning_rate": 2.769460521995443e-06,
+      "loss": 0.6879,
+      "step": 32452
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2691123769423656,
+      "learning_rate": 2.7689334407157086e-06,
+      "loss": 0.7568,
+      "step": 32453
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.450790037139999,
+      "learning_rate": 2.7684064015375333e-06,
+      "loss": 0.6884,
+      "step": 32454
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4485842351347948,
+      "learning_rate": 2.7678794044639835e-06,
+      "loss": 0.7603,
+      "step": 32455
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5466741676164477,
+      "learning_rate": 2.7673524494981275e-06,
+      "loss": 0.873,
+      "step": 32456
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.689022791399376,
+      "learning_rate": 2.7668255366430397e-06,
+      "loss": 0.8061,
+      "step": 32457
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6610999713301235,
+      "learning_rate": 2.7662986659017833e-06,
+      "loss": 0.7381,
+      "step": 32458
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5079745521612797,
+      "learning_rate": 2.7657718372774233e-06,
+      "loss": 0.7081,
+      "step": 32459
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.543356906181106,
+      "learning_rate": 2.7652450507730297e-06,
+      "loss": 0.7166,
+      "step": 32460
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3689896923845941,
+      "learning_rate": 2.764718306391673e-06,
+      "loss": 0.7542,
+      "step": 32461
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4715194127397406,
+      "learning_rate": 2.764191604136417e-06,
+      "loss": 0.882,
+      "step": 32462
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.293724681649991,
+      "learning_rate": 2.7636649440103247e-06,
+      "loss": 0.6098,
+      "step": 32463
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3255289680994002,
+      "learning_rate": 2.763138326016467e-06,
+      "loss": 0.6686,
+      "step": 32464
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4945105680122162,
+      "learning_rate": 2.762611750157912e-06,
+      "loss": 0.748,
+      "step": 32465
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3225059243382862,
+      "learning_rate": 2.762085216437722e-06,
+      "loss": 0.6323,
+      "step": 32466
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6635190626679803,
+      "learning_rate": 2.7615587248589614e-06,
+      "loss": 0.7316,
+      "step": 32467
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2787083163999022,
+      "learning_rate": 2.7610322754247e-06,
+      "loss": 0.6875,
+      "step": 32468
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4087892666118067,
+      "learning_rate": 2.760505868137997e-06,
+      "loss": 0.734,
+      "step": 32469
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4911855780842502,
+      "learning_rate": 2.7599795030019237e-06,
+      "loss": 0.6267,
+      "step": 32470
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3676920187590424,
+      "learning_rate": 2.75945318001954e-06,
+      "loss": 0.6539,
+      "step": 32471
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3099722084205656,
+      "learning_rate": 2.7589268991939145e-06,
+      "loss": 0.6167,
+      "step": 32472
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6599695014663441,
+      "learning_rate": 2.758400660528107e-06,
+      "loss": 0.5239,
+      "step": 32473
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3288916281603933,
+      "learning_rate": 2.757874464025182e-06,
+      "loss": 0.643,
+      "step": 32474
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6745455329199426,
+      "learning_rate": 2.7573483096882114e-06,
+      "loss": 0.6961,
+      "step": 32475
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.562973752652725,
+      "learning_rate": 2.7568221975202468e-06,
+      "loss": 0.8169,
+      "step": 32476
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5899062168203195,
+      "learning_rate": 2.756296127524355e-06,
+      "loss": 0.6299,
+      "step": 32477
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2933825820510598,
+      "learning_rate": 2.755770099703602e-06,
+      "loss": 0.6725,
+      "step": 32478
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.272024619799217,
+      "learning_rate": 2.755244114061054e-06,
+      "loss": 0.7076,
+      "step": 32479
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3514799505909187,
+      "learning_rate": 2.7547181705997626e-06,
+      "loss": 0.6728,
+      "step": 32480
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.386032628537884,
+      "learning_rate": 2.754192269322795e-06,
+      "loss": 0.5072,
+      "step": 32481
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3826539708658032,
+      "learning_rate": 2.7536664102332177e-06,
+      "loss": 0.6667,
+      "step": 32482
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6238701457387552,
+      "learning_rate": 2.7531405933340894e-06,
+      "loss": 0.6551,
+      "step": 32483
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3729329264548036,
+      "learning_rate": 2.7526148186284674e-06,
+      "loss": 0.5955,
+      "step": 32484
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4801429306004996,
+      "learning_rate": 2.752089086119416e-06,
+      "loss": 0.6216,
+      "step": 32485
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3062490611721431,
+      "learning_rate": 2.7515633958099998e-06,
+      "loss": 0.6776,
+      "step": 32486
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2735410827291027,
+      "learning_rate": 2.7510377477032735e-06,
+      "loss": 0.6648,
+      "step": 32487
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3917079737827682,
+      "learning_rate": 2.750512141802304e-06,
+      "loss": 0.6977,
+      "step": 32488
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4772621318494212,
+      "learning_rate": 2.7499865781101466e-06,
+      "loss": 0.689,
+      "step": 32489
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3928664067045655,
+      "learning_rate": 2.7494610566298608e-06,
+      "loss": 0.6652,
+      "step": 32490
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.771667082556828,
+      "learning_rate": 2.7489355773645078e-06,
+      "loss": 0.6223,
+      "step": 32491
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5503857397965939,
+      "learning_rate": 2.7484101403171505e-06,
+      "loss": 0.6348,
+      "step": 32492
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2746378204111197,
+      "learning_rate": 2.7478847454908454e-06,
+      "loss": 0.6609,
+      "step": 32493
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.49424577656417,
+      "learning_rate": 2.747359392888648e-06,
+      "loss": 0.7109,
+      "step": 32494
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6002876524124897,
+      "learning_rate": 2.74683408251362e-06,
+      "loss": 0.5123,
+      "step": 32495
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1903229542482954,
+      "learning_rate": 2.7463088143688234e-06,
+      "loss": 0.6045,
+      "step": 32496
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2397748509438562,
+      "learning_rate": 2.7457835884573137e-06,
+      "loss": 0.6303,
+      "step": 32497
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4555168727276828,
+      "learning_rate": 2.7452584047821452e-06,
+      "loss": 0.7247,
+      "step": 32498
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3214023696948027,
+      "learning_rate": 2.7447332633463795e-06,
+      "loss": 0.6404,
+      "step": 32499
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5781446728034896,
+      "learning_rate": 2.744208164153077e-06,
+      "loss": 0.6685,
+      "step": 32500
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3797706377886505,
+      "learning_rate": 2.7436831072052915e-06,
+      "loss": 0.5665,
+      "step": 32501
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3390865334390447,
+      "learning_rate": 2.7431580925060762e-06,
+      "loss": 0.7299,
+      "step": 32502
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6242452963479468,
+      "learning_rate": 2.7426331200584966e-06,
+      "loss": 0.5292,
+      "step": 32503
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5838871132386075,
+      "learning_rate": 2.742108189865601e-06,
+      "loss": 0.4962,
+      "step": 32504
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3626439718225756,
+      "learning_rate": 2.7415833019304517e-06,
+      "loss": 0.7335,
+      "step": 32505
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.30505407296974,
+      "learning_rate": 2.7410584562561004e-06,
+      "loss": 0.6167,
+      "step": 32506
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.253221046607056,
+      "learning_rate": 2.7405336528456073e-06,
+      "loss": 0.6331,
+      "step": 32507
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3874622101120637,
+      "learning_rate": 2.740008891702023e-06,
+      "loss": 0.7263,
+      "step": 32508
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2831197574960924,
+      "learning_rate": 2.7394841728284083e-06,
+      "loss": 0.6198,
+      "step": 32509
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3897678452911473,
+      "learning_rate": 2.738959496227814e-06,
+      "loss": 0.6487,
+      "step": 32510
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3810117490499065,
+      "learning_rate": 2.738434861903295e-06,
+      "loss": 0.6771,
+      "step": 32511
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2165372781848056,
+      "learning_rate": 2.7379102698579075e-06,
+      "loss": 0.8215,
+      "step": 32512
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3570488403885135,
+      "learning_rate": 2.7373857200947073e-06,
+      "loss": 0.7333,
+      "step": 32513
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3589396091969255,
+      "learning_rate": 2.736861212616747e-06,
+      "loss": 0.7072,
+      "step": 32514
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.502277801178882,
+      "learning_rate": 2.7363367474270774e-06,
+      "loss": 0.7333,
+      "step": 32515
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6298578192452301,
+      "learning_rate": 2.7358123245287548e-06,
+      "loss": 0.5207,
+      "step": 32516
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6595673101817947,
+      "learning_rate": 2.7352879439248358e-06,
+      "loss": 0.8297,
+      "step": 32517
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3656352674189798,
+      "learning_rate": 2.73476360561837e-06,
+      "loss": 0.6028,
+      "step": 32518
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5843008998798025,
+      "learning_rate": 2.734239309612409e-06,
+      "loss": 0.4759,
+      "step": 32519
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2862761706074692,
+      "learning_rate": 2.7337150559100055e-06,
+      "loss": 0.6485,
+      "step": 32520
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2619786679992293,
+      "learning_rate": 2.7331908445142165e-06,
+      "loss": 0.6561,
+      "step": 32521
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3984123668430049,
+      "learning_rate": 2.7326666754280918e-06,
+      "loss": 0.6746,
+      "step": 32522
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4794512729094138,
+      "learning_rate": 2.7321425486546792e-06,
+      "loss": 0.6133,
+      "step": 32523
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3184627599158547,
+      "learning_rate": 2.7316184641970367e-06,
+      "loss": 0.6469,
+      "step": 32524
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5125677521140124,
+      "learning_rate": 2.7310944220582092e-06,
+      "loss": 0.6703,
+      "step": 32525
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3676601631266307,
+      "learning_rate": 2.7305704222412543e-06,
+      "loss": 0.7328,
+      "step": 32526
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5256235847580033,
+      "learning_rate": 2.7300464647492177e-06,
+      "loss": 0.6825,
+      "step": 32527
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.341566163180022,
+      "learning_rate": 2.729522549585154e-06,
+      "loss": 0.6157,
+      "step": 32528
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.754675938796151,
+      "learning_rate": 2.7289986767521105e-06,
+      "loss": 0.5795,
+      "step": 32529
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2608354936758583,
+      "learning_rate": 2.7284748462531394e-06,
+      "loss": 0.6707,
+      "step": 32530
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.42683220282815,
+      "learning_rate": 2.7279510580912916e-06,
+      "loss": 0.6178,
+      "step": 32531
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6461272432680139,
+      "learning_rate": 2.7274273122696118e-06,
+      "loss": 0.6202,
+      "step": 32532
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3948949868444078,
+      "learning_rate": 2.7269036087911525e-06,
+      "loss": 0.5847,
+      "step": 32533
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3007847907213974,
+      "learning_rate": 2.726379947658966e-06,
+      "loss": 0.6731,
+      "step": 32534
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5252779107991794,
+      "learning_rate": 2.7258563288760976e-06,
+      "loss": 0.7004,
+      "step": 32535
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6189890301520462,
+      "learning_rate": 2.725332752445594e-06,
+      "loss": 0.7507,
+      "step": 32536
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7032351957511578,
+      "learning_rate": 2.7248092183705065e-06,
+      "loss": 0.6674,
+      "step": 32537
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3524087790632244,
+      "learning_rate": 2.7242857266538857e-06,
+      "loss": 0.7573,
+      "step": 32538
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2707088931297528,
+      "learning_rate": 2.7237622772987772e-06,
+      "loss": 0.6618,
+      "step": 32539
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5830357775523726,
+      "learning_rate": 2.7232388703082258e-06,
+      "loss": 0.4943,
+      "step": 32540
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4719913018045436,
+      "learning_rate": 2.7227155056852806e-06,
+      "loss": 0.7546,
+      "step": 32541
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3328019352147737,
+      "learning_rate": 2.7221921834329923e-06,
+      "loss": 0.7676,
+      "step": 32542
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3584292114880085,
+      "learning_rate": 2.7216689035544064e-06,
+      "loss": 0.6882,
+      "step": 32543
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4595382932075873,
+      "learning_rate": 2.7211456660525647e-06,
+      "loss": 0.5991,
+      "step": 32544
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.398939840146683,
+      "learning_rate": 2.7206224709305206e-06,
+      "loss": 0.5684,
+      "step": 32545
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3804067542613003,
+      "learning_rate": 2.7200993181913137e-06,
+      "loss": 0.7552,
+      "step": 32546
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5151054445827623,
+      "learning_rate": 2.719576207837997e-06,
+      "loss": 0.655,
+      "step": 32547
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4617484086215147,
+      "learning_rate": 2.7190531398736098e-06,
+      "loss": 0.6206,
+      "step": 32548
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8789698174332392,
+      "learning_rate": 2.718530114301202e-06,
+      "loss": 0.5942,
+      "step": 32549
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.243484031704129,
+      "learning_rate": 2.7180071311238152e-06,
+      "loss": 0.7198,
+      "step": 32550
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5911107531875703,
+      "learning_rate": 2.717484190344499e-06,
+      "loss": 0.7623,
+      "step": 32551
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.374876029428935,
+      "learning_rate": 2.716961291966296e-06,
+      "loss": 0.7917,
+      "step": 32552
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4777616507436284,
+      "learning_rate": 2.7164384359922467e-06,
+      "loss": 0.6009,
+      "step": 32553
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9833049905790205,
+      "learning_rate": 2.715915622425398e-06,
+      "loss": 0.6577,
+      "step": 32554
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6147656794976022,
+      "learning_rate": 2.715392851268799e-06,
+      "loss": 0.7643,
+      "step": 32555
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.6721808048258966,
+      "learning_rate": 2.714870122525488e-06,
+      "loss": 0.739,
+      "step": 32556
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7519265329866291,
+      "learning_rate": 2.7143474361985065e-06,
+      "loss": 0.6344,
+      "step": 32557
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3575760487202793,
+      "learning_rate": 2.7138247922909024e-06,
+      "loss": 0.6572,
+      "step": 32558
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3606480154465357,
+      "learning_rate": 2.713302190805719e-06,
+      "loss": 0.6537,
+      "step": 32559
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.389624119209979,
+      "learning_rate": 2.7127796317459976e-06,
+      "loss": 0.5782,
+      "step": 32560
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.382788012092522,
+      "learning_rate": 2.7122571151147757e-06,
+      "loss": 0.7096,
+      "step": 32561
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.415176075725509,
+      "learning_rate": 2.7117346409151023e-06,
+      "loss": 0.7318,
+      "step": 32562
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2914647033308493,
+      "learning_rate": 2.7112122091500194e-06,
+      "loss": 0.6323,
+      "step": 32563
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5675165097082446,
+      "learning_rate": 2.710689819822566e-06,
+      "loss": 0.5094,
+      "step": 32564
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6017578836826639,
+      "learning_rate": 2.7101674729357817e-06,
+      "loss": 0.5025,
+      "step": 32565
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6632391203586343,
+      "learning_rate": 2.7096451684927138e-06,
+      "loss": 0.7897,
+      "step": 32566
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3999640798693784,
+      "learning_rate": 2.7091229064963964e-06,
+      "loss": 0.701,
+      "step": 32567
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.952910056917016,
+      "learning_rate": 2.708600686949876e-06,
+      "loss": 0.6987,
+      "step": 32568
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4157602431451615,
+      "learning_rate": 2.708078509856189e-06,
+      "loss": 0.7337,
+      "step": 32569
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4684762856505338,
+      "learning_rate": 2.70755637521838e-06,
+      "loss": 0.6642,
+      "step": 32570
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6675617751916085,
+      "learning_rate": 2.7070342830394836e-06,
+      "loss": 0.7299,
+      "step": 32571
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3340856436692285,
+      "learning_rate": 2.7065122333225426e-06,
+      "loss": 0.5911,
+      "step": 32572
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4165202899537321,
+      "learning_rate": 2.705990226070602e-06,
+      "loss": 0.5978,
+      "step": 32573
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4194740037809375,
+      "learning_rate": 2.70546826128669e-06,
+      "loss": 0.584,
+      "step": 32574
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.399218264660526,
+      "learning_rate": 2.7049463389738506e-06,
+      "loss": 0.7208,
+      "step": 32575
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4184231175149518,
+      "learning_rate": 2.704424459135123e-06,
+      "loss": 0.7433,
+      "step": 32576
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6072024802981932,
+      "learning_rate": 2.703902621773553e-06,
+      "loss": 0.7628,
+      "step": 32577
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.265528018561421,
+      "learning_rate": 2.7033808268921658e-06,
+      "loss": 0.7223,
+      "step": 32578
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5314989609447107,
+      "learning_rate": 2.702859074494004e-06,
+      "loss": 0.6519,
+      "step": 32579
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4943009898979487,
+      "learning_rate": 2.7023373645821115e-06,
+      "loss": 0.7819,
+      "step": 32580
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2904739420477893,
+      "learning_rate": 2.7018156971595177e-06,
+      "loss": 0.6531,
+      "step": 32581
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8361298017947087,
+      "learning_rate": 2.701294072229267e-06,
+      "loss": 0.7362,
+      "step": 32582
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8201398719187216,
+      "learning_rate": 2.7007724897943887e-06,
+      "loss": 0.6469,
+      "step": 32583
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4100457265166186,
+      "learning_rate": 2.7002509498579275e-06,
+      "loss": 0.6729,
+      "step": 32584
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.01533248339033,
+      "learning_rate": 2.6997294524229145e-06,
+      "loss": 0.764,
+      "step": 32585
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3705436783412432,
+      "learning_rate": 2.699207997492389e-06,
+      "loss": 0.6265,
+      "step": 32586
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.382275851982223,
+      "learning_rate": 2.6986865850693877e-06,
+      "loss": 0.8259,
+      "step": 32587
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7687245549659916,
+      "learning_rate": 2.6981652151569404e-06,
+      "loss": 0.7352,
+      "step": 32588
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.465219407183046,
+      "learning_rate": 2.6976438877580867e-06,
+      "loss": 0.7806,
+      "step": 32589
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4670067136783604,
+      "learning_rate": 2.6971226028758655e-06,
+      "loss": 0.6429,
+      "step": 32590
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5458753938073475,
+      "learning_rate": 2.696601360513309e-06,
+      "loss": 0.6916,
+      "step": 32591
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5890379459977626,
+      "learning_rate": 2.6960801606734487e-06,
+      "loss": 0.6825,
+      "step": 32592
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6002700598359727,
+      "learning_rate": 2.6955590033593217e-06,
+      "loss": 0.6968,
+      "step": 32593
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1444896388365113,
+      "learning_rate": 2.695037888573966e-06,
+      "loss": 0.7549,
+      "step": 32594
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.56988730259356,
+      "learning_rate": 2.6945168163204116e-06,
+      "loss": 0.6314,
+      "step": 32595
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3348221455352138,
+      "learning_rate": 2.693995786601691e-06,
+      "loss": 0.5653,
+      "step": 32596
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9104430640728718,
+      "learning_rate": 2.6934747994208397e-06,
+      "loss": 0.6655,
+      "step": 32597
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.569340280658573,
+      "learning_rate": 2.692953854780894e-06,
+      "loss": 0.7082,
+      "step": 32598
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7033028321955668,
+      "learning_rate": 2.692432952684886e-06,
+      "loss": 0.5957,
+      "step": 32599
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2972981027973554,
+      "learning_rate": 2.691912093135842e-06,
+      "loss": 0.6842,
+      "step": 32600
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3168392761927994,
+      "learning_rate": 2.6913912761368e-06,
+      "loss": 0.7142,
+      "step": 32601
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3295665815369384,
+      "learning_rate": 2.6908705016907956e-06,
+      "loss": 0.7276,
+      "step": 32602
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4972943738833948,
+      "learning_rate": 2.6903497698008564e-06,
+      "loss": 0.7023,
+      "step": 32603
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.523327311012009,
+      "learning_rate": 2.689829080470011e-06,
+      "loss": 0.6894,
+      "step": 32604
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.260543509010207,
+      "learning_rate": 2.6893084337013e-06,
+      "loss": 0.6386,
+      "step": 32605
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3605732741176535,
+      "learning_rate": 2.6887878294977456e-06,
+      "loss": 0.7012,
+      "step": 32606
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.798150674309205,
+      "learning_rate": 2.688267267862387e-06,
+      "loss": 0.8149,
+      "step": 32607
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4997060161298095,
+      "learning_rate": 2.687746748798248e-06,
+      "loss": 0.7862,
+      "step": 32608
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5362794000130264,
+      "learning_rate": 2.687226272308365e-06,
+      "loss": 0.66,
+      "step": 32609
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5655689766143293,
+      "learning_rate": 2.6867058383957633e-06,
+      "loss": 0.709,
+      "step": 32610
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3899181096913777,
+      "learning_rate": 2.6861854470634784e-06,
+      "loss": 0.8294,
+      "step": 32611
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4092454098364922,
+      "learning_rate": 2.685665098314537e-06,
+      "loss": 0.6726,
+      "step": 32612
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.29343557943021,
+      "learning_rate": 2.685144792151966e-06,
+      "loss": 0.6588,
+      "step": 32613
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6095994394742226,
+      "learning_rate": 2.6846245285787985e-06,
+      "loss": 0.678,
+      "step": 32614
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3494676000514332,
+      "learning_rate": 2.6841043075980653e-06,
+      "loss": 0.6691,
+      "step": 32615
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3182578244768346,
+      "learning_rate": 2.683584129212794e-06,
+      "loss": 0.6217,
+      "step": 32616
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4189807722261294,
+      "learning_rate": 2.6830639934260084e-06,
+      "loss": 0.6463,
+      "step": 32617
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9387233842137763,
+      "learning_rate": 2.6825439002407405e-06,
+      "loss": 0.6066,
+      "step": 32618
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5201480751405965,
+      "learning_rate": 2.682023849660023e-06,
+      "loss": 0.7157,
+      "step": 32619
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3194737316766592,
+      "learning_rate": 2.681503841686879e-06,
+      "loss": 0.726,
+      "step": 32620
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4490751440996892,
+      "learning_rate": 2.6809838763243325e-06,
+      "loss": 0.6706,
+      "step": 32621
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2813300075265726,
+      "learning_rate": 2.680463953575416e-06,
+      "loss": 0.7139,
+      "step": 32622
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.327052993029629,
+      "learning_rate": 2.679944073443158e-06,
+      "loss": 0.6521,
+      "step": 32623
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4340640965629536,
+      "learning_rate": 2.6794242359305843e-06,
+      "loss": 0.6816,
+      "step": 32624
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2673501404606229,
+      "learning_rate": 2.6789044410407162e-06,
+      "loss": 0.7768,
+      "step": 32625
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8206882029134739,
+      "learning_rate": 2.6783846887765886e-06,
+      "loss": 0.6239,
+      "step": 32626
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.633121530918781,
+      "learning_rate": 2.6778649791412194e-06,
+      "loss": 0.6515,
+      "step": 32627
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0723333661305126,
+      "learning_rate": 2.6773453121376414e-06,
+      "loss": 0.7045,
+      "step": 32628
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4412787412519181,
+      "learning_rate": 2.6768256877688735e-06,
+      "loss": 0.6349,
+      "step": 32629
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.429180224403281,
+      "learning_rate": 2.6763061060379492e-06,
+      "loss": 0.7889,
+      "step": 32630
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.521018678579512,
+      "learning_rate": 2.6757865669478854e-06,
+      "loss": 0.7,
+      "step": 32631
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.107504906755205,
+      "learning_rate": 2.6752670705017148e-06,
+      "loss": 0.4788,
+      "step": 32632
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.978135462767342,
+      "learning_rate": 2.6747476167024587e-06,
+      "loss": 0.7449,
+      "step": 32633
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3986816483447846,
+      "learning_rate": 2.674228205553138e-06,
+      "loss": 0.7472,
+      "step": 32634
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1329887851050766,
+      "learning_rate": 2.673708837056779e-06,
+      "loss": 0.6426,
+      "step": 32635
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3256198333147717,
+      "learning_rate": 2.6731895112164096e-06,
+      "loss": 0.6202,
+      "step": 32636
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2610041095596103,
+      "learning_rate": 2.6726702280350513e-06,
+      "loss": 0.787,
+      "step": 32637
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3891504968594803,
+      "learning_rate": 2.672150987515724e-06,
+      "loss": 0.6581,
+      "step": 32638
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3107733356957092,
+      "learning_rate": 2.671631789661453e-06,
+      "loss": 0.7171,
+      "step": 32639
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4114329751621606,
+      "learning_rate": 2.6711126344752647e-06,
+      "loss": 0.7525,
+      "step": 32640
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5882310715469705,
+      "learning_rate": 2.6705935219601788e-06,
+      "loss": 0.5179,
+      "step": 32641
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3176004240641612,
+      "learning_rate": 2.670074452119216e-06,
+      "loss": 0.6237,
+      "step": 32642
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.52003169518244,
+      "learning_rate": 2.669555424955399e-06,
+      "loss": 0.7084,
+      "step": 32643
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3133753286071765,
+      "learning_rate": 2.669036440471754e-06,
+      "loss": 0.8389,
+      "step": 32644
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6080529947229207,
+      "learning_rate": 2.6685174986713012e-06,
+      "loss": 0.5281,
+      "step": 32645
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3206438540073906,
+      "learning_rate": 2.667998599557057e-06,
+      "loss": 0.5669,
+      "step": 32646
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1134188363984086,
+      "learning_rate": 2.667479743132049e-06,
+      "loss": 0.8029,
+      "step": 32647
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4669409876260389,
+      "learning_rate": 2.6669609293992914e-06,
+      "loss": 0.6731,
+      "step": 32648
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.293819801669437,
+      "learning_rate": 2.6664421583618138e-06,
+      "loss": 0.5969,
+      "step": 32649
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2935209714016944,
+      "learning_rate": 2.665923430022628e-06,
+      "loss": 0.6267,
+      "step": 32650
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7518070889956787,
+      "learning_rate": 2.6654047443847607e-06,
+      "loss": 0.6082,
+      "step": 32651
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3485838012802405,
+      "learning_rate": 2.6648861014512263e-06,
+      "loss": 0.709,
+      "step": 32652
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4400240804506177,
+      "learning_rate": 2.664367501225049e-06,
+      "loss": 0.7457,
+      "step": 32653
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.316880716325912,
+      "learning_rate": 2.6638489437092486e-06,
+      "loss": 0.8984,
+      "step": 32654
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5111639413765012,
+      "learning_rate": 2.6633304289068384e-06,
+      "loss": 0.7098,
+      "step": 32655
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.967404367904294,
+      "learning_rate": 2.662811956820841e-06,
+      "loss": 0.6264,
+      "step": 32656
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4523165574796613,
+      "learning_rate": 2.662293527454275e-06,
+      "loss": 0.7015,
+      "step": 32657
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.575141691389324,
+      "learning_rate": 2.6617751408101655e-06,
+      "loss": 0.5364,
+      "step": 32658
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5731183656693262,
+      "learning_rate": 2.661256796891518e-06,
+      "loss": 0.7078,
+      "step": 32659
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5842913920407287,
+      "learning_rate": 2.6607384957013583e-06,
+      "loss": 0.4808,
+      "step": 32660
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4444791221590634,
+      "learning_rate": 2.6602202372427046e-06,
+      "loss": 0.7224,
+      "step": 32661
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4527470780172966,
+      "learning_rate": 2.659702021518572e-06,
+      "loss": 0.7241,
+      "step": 32662
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.379091028808453,
+      "learning_rate": 2.6591838485319754e-06,
+      "loss": 0.7073,
+      "step": 32663
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5077796554562917,
+      "learning_rate": 2.658665718285934e-06,
+      "loss": 0.7395,
+      "step": 32664
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.312652901953701,
+      "learning_rate": 2.6581476307834697e-06,
+      "loss": 0.5978,
+      "step": 32665
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1916412518949397,
+      "learning_rate": 2.6576295860275927e-06,
+      "loss": 0.7466,
+      "step": 32666
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7178162485024455,
+      "learning_rate": 2.657111584021318e-06,
+      "loss": 0.7773,
+      "step": 32667
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5896997550111648,
+      "learning_rate": 2.6565936247676684e-06,
+      "loss": 0.5165,
+      "step": 32668
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4127947982274223,
+      "learning_rate": 2.656075708269652e-06,
+      "loss": 0.6032,
+      "step": 32669
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3204552710742459,
+      "learning_rate": 2.655557834530288e-06,
+      "loss": 0.6984,
+      "step": 32670
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3488776618321312,
+      "learning_rate": 2.6550400035525936e-06,
+      "loss": 0.742,
+      "step": 32671
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7689710692560483,
+      "learning_rate": 2.6545222153395822e-06,
+      "loss": 0.7693,
+      "step": 32672
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4843181845299394,
+      "learning_rate": 2.6540044698942655e-06,
+      "loss": 0.4703,
+      "step": 32673
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4143183343021315,
+      "learning_rate": 2.6534867672196594e-06,
+      "loss": 0.7866,
+      "step": 32674
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9122689690075514,
+      "learning_rate": 2.652969107318787e-06,
+      "loss": 0.7746,
+      "step": 32675
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3604269322611326,
+      "learning_rate": 2.6524514901946473e-06,
+      "loss": 0.6412,
+      "step": 32676
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.27509697292915,
+      "learning_rate": 2.651933915850261e-06,
+      "loss": 0.6872,
+      "step": 32677
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2857409392065076,
+      "learning_rate": 2.651416384288642e-06,
+      "loss": 0.6141,
+      "step": 32678
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2633280926840373,
+      "learning_rate": 2.6508988955128067e-06,
+      "loss": 0.6002,
+      "step": 32679
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3417481142936016,
+      "learning_rate": 2.6503814495257652e-06,
+      "loss": 0.68,
+      "step": 32680
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4119105082365058,
+      "learning_rate": 2.649864046330526e-06,
+      "loss": 0.7005,
+      "step": 32681
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5481824644144349,
+      "learning_rate": 2.6493466859301074e-06,
+      "loss": 0.8478,
+      "step": 32682
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6270850311040355,
+      "learning_rate": 2.6488293683275177e-06,
+      "loss": 0.5056,
+      "step": 32683
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5626394270644517,
+      "learning_rate": 2.6483120935257744e-06,
+      "loss": 0.665,
+      "step": 32684
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5838916874927716,
+      "learning_rate": 2.6477948615278813e-06,
+      "loss": 0.5009,
+      "step": 32685
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5133693338307,
+      "learning_rate": 2.6472776723368577e-06,
+      "loss": 0.6246,
+      "step": 32686
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3335490302011452,
+      "learning_rate": 2.646760525955708e-06,
+      "loss": 0.672,
+      "step": 32687
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.263828346943404,
+      "learning_rate": 2.64624342238745e-06,
+      "loss": 0.5083,
+      "step": 32688
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6118504730545776,
+      "learning_rate": 2.6457263616350915e-06,
+      "loss": 0.5073,
+      "step": 32689
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6206084016222086,
+      "learning_rate": 2.645209343701638e-06,
+      "loss": 0.5582,
+      "step": 32690
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6088854541225486,
+      "learning_rate": 2.644692368590105e-06,
+      "loss": 0.7974,
+      "step": 32691
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5207788453087507,
+      "learning_rate": 2.6441754363035045e-06,
+      "loss": 0.771,
+      "step": 32692
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5733726796686842,
+      "learning_rate": 2.6436585468448438e-06,
+      "loss": 0.6768,
+      "step": 32693
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.81087024378781,
+      "learning_rate": 2.643141700217129e-06,
+      "loss": 0.6332,
+      "step": 32694
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4092853949647224,
+      "learning_rate": 2.6426248964233725e-06,
+      "loss": 0.6513,
+      "step": 32695
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0081682111955357,
+      "learning_rate": 2.6421081354665856e-06,
+      "loss": 0.7621,
+      "step": 32696
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2326831626842212,
+      "learning_rate": 2.6415914173497746e-06,
+      "loss": 0.7001,
+      "step": 32697
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5943123141773499,
+      "learning_rate": 2.6410747420759455e-06,
+      "loss": 0.7176,
+      "step": 32698
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3895282817608887,
+      "learning_rate": 2.6405581096481083e-06,
+      "loss": 0.6344,
+      "step": 32699
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6676269740347065,
+      "learning_rate": 2.6400415200692754e-06,
+      "loss": 0.7085,
+      "step": 32700
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5447933841682675,
+      "learning_rate": 2.63952497334245e-06,
+      "loss": 0.6374,
+      "step": 32701
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2709126459047158,
+      "learning_rate": 2.639008469470639e-06,
+      "loss": 0.6785,
+      "step": 32702
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2192709671325668,
+      "learning_rate": 2.638492008456853e-06,
+      "loss": 0.6653,
+      "step": 32703
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4395193047519517,
+      "learning_rate": 2.6379755903040947e-06,
+      "loss": 0.7536,
+      "step": 32704
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.541045586307352,
+      "learning_rate": 2.6374592150153755e-06,
+      "loss": 0.6648,
+      "step": 32705
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4829623402385033,
+      "learning_rate": 2.6369428825936973e-06,
+      "loss": 0.666,
+      "step": 32706
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3158281023793403,
+      "learning_rate": 2.636426593042072e-06,
+      "loss": 0.6458,
+      "step": 32707
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4106777854839248,
+      "learning_rate": 2.635910346363499e-06,
+      "loss": 0.7265,
+      "step": 32708
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4157806047196628,
+      "learning_rate": 2.6353941425609907e-06,
+      "loss": 0.7615,
+      "step": 32709
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.412670016784166,
+      "learning_rate": 2.634877981637548e-06,
+      "loss": 0.6274,
+      "step": 32710
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.550613953962705,
+      "learning_rate": 2.6343618635961765e-06,
+      "loss": 0.7546,
+      "step": 32711
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2882344808703192,
+      "learning_rate": 2.6338457884398805e-06,
+      "loss": 0.6405,
+      "step": 32712
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3106889195277827,
+      "learning_rate": 2.6333297561716696e-06,
+      "loss": 0.6079,
+      "step": 32713
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.515335510220888,
+      "learning_rate": 2.6328137667945463e-06,
+      "loss": 0.6913,
+      "step": 32714
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6035556528692634,
+      "learning_rate": 2.6322978203115102e-06,
+      "loss": 0.4768,
+      "step": 32715
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2289745302761066,
+      "learning_rate": 2.6317819167255676e-06,
+      "loss": 0.6259,
+      "step": 32716
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3619590219688495,
+      "learning_rate": 2.6312660560397273e-06,
+      "loss": 0.7396,
+      "step": 32717
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6575807356346388,
+      "learning_rate": 2.6307502382569894e-06,
+      "loss": 0.6613,
+      "step": 32718
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2545963945241525,
+      "learning_rate": 2.6302344633803534e-06,
+      "loss": 0.6856,
+      "step": 32719
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4745241059883647,
+      "learning_rate": 2.6297187314128247e-06,
+      "loss": 0.6159,
+      "step": 32720
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2923827550743419,
+      "learning_rate": 2.629203042357411e-06,
+      "loss": 0.633,
+      "step": 32721
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5425604247143059,
+      "learning_rate": 2.62868739621711e-06,
+      "loss": 0.6186,
+      "step": 32722
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1630670049952232,
+      "learning_rate": 2.6281717929949226e-06,
+      "loss": 0.6338,
+      "step": 32723
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6389068448823753,
+      "learning_rate": 2.6276562326938516e-06,
+      "loss": 0.4768,
+      "step": 32724
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2640561940384893,
+      "learning_rate": 2.6271407153169036e-06,
+      "loss": 0.6621,
+      "step": 32725
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5144336301092711,
+      "learning_rate": 2.6266252408670776e-06,
+      "loss": 0.8406,
+      "step": 32726
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4151431887957489,
+      "learning_rate": 2.626109809347369e-06,
+      "loss": 0.6802,
+      "step": 32727
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3165077465516328,
+      "learning_rate": 2.6255944207607885e-06,
+      "loss": 0.603,
+      "step": 32728
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5050606213426203,
+      "learning_rate": 2.6250790751103282e-06,
+      "loss": 0.6525,
+      "step": 32729
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2351278734389592,
+      "learning_rate": 2.6245637723989946e-06,
+      "loss": 0.6087,
+      "step": 32730
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6260126877537611,
+      "learning_rate": 2.6240485126297834e-06,
+      "loss": 0.5113,
+      "step": 32731
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.566060829780972,
+      "learning_rate": 2.6235332958057004e-06,
+      "loss": 0.5973,
+      "step": 32732
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.424308158429034,
+      "learning_rate": 2.6230181219297378e-06,
+      "loss": 0.7487,
+      "step": 32733
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.000467449710995,
+      "learning_rate": 2.6225029910049037e-06,
+      "loss": 0.8019,
+      "step": 32734
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5147355259329691,
+      "learning_rate": 2.621987903034192e-06,
+      "loss": 0.6808,
+      "step": 32735
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.709327898308779,
+      "learning_rate": 2.6214728580205997e-06,
+      "loss": 0.728,
+      "step": 32736
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5108011056291635,
+      "learning_rate": 2.620957855967128e-06,
+      "loss": 0.6068,
+      "step": 32737
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5853212423024376,
+      "learning_rate": 2.620442896876779e-06,
+      "loss": 0.6978,
+      "step": 32738
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3487080178476896,
+      "learning_rate": 2.619927980752548e-06,
+      "loss": 0.6222,
+      "step": 32739
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.658866118361751,
+      "learning_rate": 2.6194131075974294e-06,
+      "loss": 0.667,
+      "step": 32740
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5981317798133562,
+      "learning_rate": 2.6188982774144245e-06,
+      "loss": 0.4973,
+      "step": 32741
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.546961607235343,
+      "learning_rate": 2.618383490206534e-06,
+      "loss": 0.6101,
+      "step": 32742
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4285242554934199,
+      "learning_rate": 2.6178687459767514e-06,
+      "loss": 0.6941,
+      "step": 32743
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6475490169604066,
+      "learning_rate": 2.617354044728072e-06,
+      "loss": 0.788,
+      "step": 32744
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.587152883404943,
+      "learning_rate": 2.6168393864634935e-06,
+      "loss": 0.4842,
+      "step": 32745
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4940623657041436,
+      "learning_rate": 2.6163247711860173e-06,
+      "loss": 0.7028,
+      "step": 32746
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5664310885152177,
+      "learning_rate": 2.615810198898635e-06,
+      "loss": 0.6586,
+      "step": 32747
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9080204973674475,
+      "learning_rate": 2.615295669604342e-06,
+      "loss": 0.7557,
+      "step": 32748
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.274810768562397,
+      "learning_rate": 2.6147811833061376e-06,
+      "loss": 0.594,
+      "step": 32749
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.619384335968138,
+      "learning_rate": 2.614266740007012e-06,
+      "loss": 0.6812,
+      "step": 32750
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.263782834953271,
+      "learning_rate": 2.6137523397099672e-06,
+      "loss": 0.6846,
+      "step": 32751
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4280458536203406,
+      "learning_rate": 2.6132379824179923e-06,
+      "loss": 0.6167,
+      "step": 32752
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.270265921452884,
+      "learning_rate": 2.612723668134086e-06,
+      "loss": 0.6667,
+      "step": 32753
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4730456251939146,
+      "learning_rate": 2.6122093968612393e-06,
+      "loss": 0.6318,
+      "step": 32754
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4326896578555406,
+      "learning_rate": 2.611695168602448e-06,
+      "loss": 0.731,
+      "step": 32755
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.306845773417279,
+      "learning_rate": 2.6111809833607126e-06,
+      "loss": 0.8201,
+      "step": 32756
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6911100578520388,
+      "learning_rate": 2.6106668411390156e-06,
+      "loss": 0.6673,
+      "step": 32757
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2976245313904324,
+      "learning_rate": 2.610152741940355e-06,
+      "loss": 0.6623,
+      "step": 32758
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.600361131022815,
+      "learning_rate": 2.609638685767725e-06,
+      "loss": 0.6672,
+      "step": 32759
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3607683952655578,
+      "learning_rate": 2.609124672624124e-06,
+      "loss": 0.607,
+      "step": 32760
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.34444463105746,
+      "learning_rate": 2.608610702512533e-06,
+      "loss": 0.6499,
+      "step": 32761
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6194776382401221,
+      "learning_rate": 2.608096775435951e-06,
+      "loss": 0.4954,
+      "step": 32762
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4361446902197677,
+      "learning_rate": 2.607582891397372e-06,
+      "loss": 0.56,
+      "step": 32763
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.538843648000447,
+      "learning_rate": 2.6070690503997866e-06,
+      "loss": 0.5699,
+      "step": 32764
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5409803478912727,
+      "learning_rate": 2.6065552524461823e-06,
+      "loss": 0.7695,
+      "step": 32765
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5508697224921972,
+      "learning_rate": 2.6060414975395543e-06,
+      "loss": 0.5145,
+      "step": 32766
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4567098338242015,
+      "learning_rate": 2.6055277856828964e-06,
+      "loss": 0.5684,
+      "step": 32767
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4744762358477783,
+      "learning_rate": 2.605014116879193e-06,
+      "loss": 0.7868,
+      "step": 32768
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6529813614658071,
+      "learning_rate": 2.604500491131443e-06,
+      "loss": 0.5437,
+      "step": 32769
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.37201719900329,
+      "learning_rate": 2.603986908442632e-06,
+      "loss": 0.7005,
+      "step": 32770
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.333982378668342,
+      "learning_rate": 2.6034733688157475e-06,
+      "loss": 0.6687,
+      "step": 32771
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.573153750294744,
+      "learning_rate": 2.602959872253783e-06,
+      "loss": 0.6243,
+      "step": 32772
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7220850437475945,
+      "learning_rate": 2.602446418759731e-06,
+      "loss": 0.6361,
+      "step": 32773
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.504996346266082,
+      "learning_rate": 2.6019330083365778e-06,
+      "loss": 0.7003,
+      "step": 32774
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4171555799822422,
+      "learning_rate": 2.6014196409873104e-06,
+      "loss": 0.6543,
+      "step": 32775
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4348582440105604,
+      "learning_rate": 2.6009063167149195e-06,
+      "loss": 0.6855,
+      "step": 32776
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7230010887878526,
+      "learning_rate": 2.600393035522398e-06,
+      "loss": 0.766,
+      "step": 32777
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2714092899060978,
+      "learning_rate": 2.599879797412731e-06,
+      "loss": 0.5784,
+      "step": 32778
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.670269706609342,
+      "learning_rate": 2.599366602388903e-06,
+      "loss": 0.6463,
+      "step": 32779
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4006030291635023,
+      "learning_rate": 2.5988534504539065e-06,
+      "loss": 0.7097,
+      "step": 32780
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5290062068552677,
+      "learning_rate": 2.598340341610731e-06,
+      "loss": 0.6903,
+      "step": 32781
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4127274569439423,
+      "learning_rate": 2.597827275862361e-06,
+      "loss": 0.6497,
+      "step": 32782
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8118655966491901,
+      "learning_rate": 2.597314253211781e-06,
+      "loss": 0.6913,
+      "step": 32783
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6725140552269038,
+      "learning_rate": 2.596801273661985e-06,
+      "loss": 0.7203,
+      "step": 32784
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1113007909517734,
+      "learning_rate": 2.596288337215952e-06,
+      "loss": 0.5795,
+      "step": 32785
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2239239659058039,
+      "learning_rate": 2.595775443876676e-06,
+      "loss": 0.5318,
+      "step": 32786
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2576475559310398,
+      "learning_rate": 2.5952625936471354e-06,
+      "loss": 0.6053,
+      "step": 32787
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5459469307376386,
+      "learning_rate": 2.594749786530324e-06,
+      "loss": 0.6656,
+      "step": 32788
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4137572815848178,
+      "learning_rate": 2.5942370225292214e-06,
+      "loss": 0.6166,
+      "step": 32789
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5298043742800633,
+      "learning_rate": 2.593724301646817e-06,
+      "loss": 0.6722,
+      "step": 32790
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.305554823446775,
+      "learning_rate": 2.593211623886096e-06,
+      "loss": 0.6558,
+      "step": 32791
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4360809826340428,
+      "learning_rate": 2.5926989892500376e-06,
+      "loss": 0.7166,
+      "step": 32792
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.306849162571973,
+      "learning_rate": 2.5921863977416316e-06,
+      "loss": 0.7276,
+      "step": 32793
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3534595639073985,
+      "learning_rate": 2.5916738493638638e-06,
+      "loss": 0.6074,
+      "step": 32794
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.430664699567728,
+      "learning_rate": 2.591161344119717e-06,
+      "loss": 0.6806,
+      "step": 32795
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3178687077511633,
+      "learning_rate": 2.590648882012172e-06,
+      "loss": 0.5873,
+      "step": 32796
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3325942020162422,
+      "learning_rate": 2.5901364630442136e-06,
+      "loss": 0.7219,
+      "step": 32797
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4665645787434876,
+      "learning_rate": 2.5896240872188305e-06,
+      "loss": 0.6357,
+      "step": 32798
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6412618043472953,
+      "learning_rate": 2.589111754539002e-06,
+      "loss": 0.6858,
+      "step": 32799
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5856830660108595,
+      "learning_rate": 2.5885994650077086e-06,
+      "loss": 0.4929,
+      "step": 32800
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.420542799804251,
+      "learning_rate": 2.588087218627935e-06,
+      "loss": 0.7514,
+      "step": 32801
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.641454982691608,
+      "learning_rate": 2.587575015402667e-06,
+      "loss": 0.6593,
+      "step": 32802
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3150162497588018,
+      "learning_rate": 2.587062855334884e-06,
+      "loss": 0.8061,
+      "step": 32803
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3775851180177114,
+      "learning_rate": 2.5865507384275646e-06,
+      "loss": 0.7227,
+      "step": 32804
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6077548713921143,
+      "learning_rate": 2.586038664683698e-06,
+      "loss": 0.4692,
+      "step": 32805
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4124773697126425,
+      "learning_rate": 2.585526634106258e-06,
+      "loss": 0.7068,
+      "step": 32806
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4678108362811104,
+      "learning_rate": 2.585014646698232e-06,
+      "loss": 0.7885,
+      "step": 32807
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5980437740977647,
+      "learning_rate": 2.584502702462596e-06,
+      "loss": 0.6694,
+      "step": 32808
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.364386443477897,
+      "learning_rate": 2.5839908014023353e-06,
+      "loss": 0.6423,
+      "step": 32809
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5493742683649396,
+      "learning_rate": 2.5834789435204245e-06,
+      "loss": 0.7554,
+      "step": 32810
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3937182171780975,
+      "learning_rate": 2.582967128819851e-06,
+      "loss": 0.6905,
+      "step": 32811
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3454828442996734,
+      "learning_rate": 2.5824553573035914e-06,
+      "loss": 0.6435,
+      "step": 32812
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6917334057377533,
+      "learning_rate": 2.581943628974621e-06,
+      "loss": 0.6646,
+      "step": 32813
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.511346797546688,
+      "learning_rate": 2.581431943835925e-06,
+      "loss": 0.6948,
+      "step": 32814
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5509491277172858,
+      "learning_rate": 2.5809203018904828e-06,
+      "loss": 0.6422,
+      "step": 32815
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.318667424486194,
+      "learning_rate": 2.5804087031412705e-06,
+      "loss": 0.7186,
+      "step": 32816
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6318032940740032,
+      "learning_rate": 2.579897147591266e-06,
+      "loss": 0.5175,
+      "step": 32817
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5764945844138007,
+      "learning_rate": 2.57938563524345e-06,
+      "loss": 0.7881,
+      "step": 32818
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4432673156140823,
+      "learning_rate": 2.578874166100802e-06,
+      "loss": 0.8353,
+      "step": 32819
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.340016140533348,
+      "learning_rate": 2.578362740166298e-06,
+      "loss": 0.6181,
+      "step": 32820
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3337827528119182,
+      "learning_rate": 2.5778513574429143e-06,
+      "loss": 0.7478,
+      "step": 32821
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.065209516921004,
+      "learning_rate": 2.577340017933628e-06,
+      "loss": 0.6854,
+      "step": 32822
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.372081627663353,
+      "learning_rate": 2.5768287216414224e-06,
+      "loss": 0.7686,
+      "step": 32823
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5203784719827027,
+      "learning_rate": 2.5763174685692693e-06,
+      "loss": 0.7474,
+      "step": 32824
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4337761694937476,
+      "learning_rate": 2.5758062587201436e-06,
+      "loss": 0.5087,
+      "step": 32825
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0912440962226,
+      "learning_rate": 2.5752950920970278e-06,
+      "loss": 0.7434,
+      "step": 32826
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3323090589596025,
+      "learning_rate": 2.574783968702891e-06,
+      "loss": 0.6378,
+      "step": 32827
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0207199828259252,
+      "learning_rate": 2.5742728885407153e-06,
+      "loss": 0.8143,
+      "step": 32828
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3862756519665071,
+      "learning_rate": 2.573761851613471e-06,
+      "loss": 0.7392,
+      "step": 32829
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6092181110053494,
+      "learning_rate": 2.57325085792414e-06,
+      "loss": 0.5145,
+      "step": 32830
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.543198579333499,
+      "learning_rate": 2.57273990747569e-06,
+      "loss": 0.7284,
+      "step": 32831
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3379178970230374,
+      "learning_rate": 2.5722290002711025e-06,
+      "loss": 0.6684,
+      "step": 32832
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7691835055239309,
+      "learning_rate": 2.57171813631335e-06,
+      "loss": 0.7486,
+      "step": 32833
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9262421984812343,
+      "learning_rate": 2.571207315605404e-06,
+      "loss": 0.8047,
+      "step": 32834
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.561882433742939,
+      "learning_rate": 2.5706965381502392e-06,
+      "loss": 0.5021,
+      "step": 32835
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9688768945244977,
+      "learning_rate": 2.5701858039508354e-06,
+      "loss": 0.7441,
+      "step": 32836
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5040435830683925,
+      "learning_rate": 2.5696751130101614e-06,
+      "loss": 0.7318,
+      "step": 32837
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2809997714736625,
+      "learning_rate": 2.5691644653311887e-06,
+      "loss": 0.6655,
+      "step": 32838
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0047996462476307,
+      "learning_rate": 2.5686538609168932e-06,
+      "loss": 0.601,
+      "step": 32839
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5626712721966742,
+      "learning_rate": 2.568143299770247e-06,
+      "loss": 0.6989,
+      "step": 32840
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6787926015919596,
+      "learning_rate": 2.5676327818942305e-06,
+      "loss": 0.5377,
+      "step": 32841
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6540417160216168,
+      "learning_rate": 2.567122307291803e-06,
+      "loss": 0.5463,
+      "step": 32842
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.420682823869391,
+      "learning_rate": 2.566611875965941e-06,
+      "loss": 0.6589,
+      "step": 32843
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.126769135209906,
+      "learning_rate": 2.5661014879196223e-06,
+      "loss": 0.7185,
+      "step": 32844
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3615651533305477,
+      "learning_rate": 2.5655911431558145e-06,
+      "loss": 0.679,
+      "step": 32845
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3830548396676265,
+      "learning_rate": 2.565080841677484e-06,
+      "loss": 0.5497,
+      "step": 32846
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.109253812625313,
+      "learning_rate": 2.564570583487609e-06,
+      "loss": 0.6813,
+      "step": 32847
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4310361129622398,
+      "learning_rate": 2.5640603685891597e-06,
+      "loss": 0.7783,
+      "step": 32848
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4052072597917074,
+      "learning_rate": 2.563550196985105e-06,
+      "loss": 0.7454,
+      "step": 32849
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.482637819955324,
+      "learning_rate": 2.5630400686784128e-06,
+      "loss": 0.652,
+      "step": 32850
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6736847149147087,
+      "learning_rate": 2.5625299836720597e-06,
+      "loss": 0.5659,
+      "step": 32851
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6040586750869139,
+      "learning_rate": 2.5620199419690074e-06,
+      "loss": 0.5078,
+      "step": 32852
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3519801809768957,
+      "learning_rate": 2.56150994357223e-06,
+      "loss": 0.5925,
+      "step": 32853
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.331079699312303,
+      "learning_rate": 2.5609999884847003e-06,
+      "loss": 0.5715,
+      "step": 32854
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.402744675143166,
+      "learning_rate": 2.560490076709383e-06,
+      "loss": 0.7242,
+      "step": 32855
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3855530517116361,
+      "learning_rate": 2.559980208249245e-06,
+      "loss": 0.744,
+      "step": 32856
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5325498481490842,
+      "learning_rate": 2.5594703831072586e-06,
+      "loss": 0.6462,
+      "step": 32857
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3489166109805828,
+      "learning_rate": 2.5589606012863968e-06,
+      "loss": 0.6912,
+      "step": 32858
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4344314288923294,
+      "learning_rate": 2.5584508627896155e-06,
+      "loss": 0.5551,
+      "step": 32859
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.525514080856346,
+      "learning_rate": 2.5579411676198897e-06,
+      "loss": 0.6531,
+      "step": 32860
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5665343467006727,
+      "learning_rate": 2.5574315157801867e-06,
+      "loss": 0.6699,
+      "step": 32861
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5748800887395038,
+      "learning_rate": 2.55692190727348e-06,
+      "loss": 0.4913,
+      "step": 32862
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2865802459580693,
+      "learning_rate": 2.556412342102723e-06,
+      "loss": 0.6914,
+      "step": 32863
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4375193192459137,
+      "learning_rate": 2.5559028202708913e-06,
+      "loss": 0.8294,
+      "step": 32864
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3488062822483746,
+      "learning_rate": 2.555393341780953e-06,
+      "loss": 0.7293,
+      "step": 32865
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.449075145540746,
+      "learning_rate": 2.554883906635868e-06,
+      "loss": 0.7457,
+      "step": 32866
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2897701764528744,
+      "learning_rate": 2.554374514838609e-06,
+      "loss": 0.6471,
+      "step": 32867
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.295866541840082,
+      "learning_rate": 2.5538651663921367e-06,
+      "loss": 0.6496,
+      "step": 32868
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6583924333853632,
+      "learning_rate": 2.5533558612994213e-06,
+      "loss": 0.8079,
+      "step": 32869
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1830786670690714,
+      "learning_rate": 2.5528465995634223e-06,
+      "loss": 0.707,
+      "step": 32870
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2707211880757063,
+      "learning_rate": 2.552337381187111e-06,
+      "loss": 0.5373,
+      "step": 32871
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4248507170357856,
+      "learning_rate": 2.5518282061734502e-06,
+      "loss": 0.6104,
+      "step": 32872
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.443455866330965,
+      "learning_rate": 2.5513190745254e-06,
+      "loss": 0.5635,
+      "step": 32873
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.322697675672671,
+      "learning_rate": 2.550809986245929e-06,
+      "loss": 0.7619,
+      "step": 32874
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5348885347179153,
+      "learning_rate": 2.5503009413380033e-06,
+      "loss": 0.7505,
+      "step": 32875
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3482774244004034,
+      "learning_rate": 2.5497919398045833e-06,
+      "loss": 0.6713,
+      "step": 32876
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5607009017090936,
+      "learning_rate": 2.549282981648631e-06,
+      "loss": 0.7058,
+      "step": 32877
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5822463704290701,
+      "learning_rate": 2.548774066873112e-06,
+      "loss": 0.6758,
+      "step": 32878
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3244109001481073,
+      "learning_rate": 2.548265195480992e-06,
+      "loss": 0.6776,
+      "step": 32879
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8295516776722807,
+      "learning_rate": 2.5477563674752327e-06,
+      "loss": 0.6885,
+      "step": 32880
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3201188558952561,
+      "learning_rate": 2.5472475828587906e-06,
+      "loss": 0.8693,
+      "step": 32881
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2811613614369115,
+      "learning_rate": 2.546738841634633e-06,
+      "loss": 0.7575,
+      "step": 32882
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6058895562961301,
+      "learning_rate": 2.5462301438057245e-06,
+      "loss": 0.5556,
+      "step": 32883
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4056944545413053,
+      "learning_rate": 2.545721489375024e-06,
+      "loss": 0.6013,
+      "step": 32884
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5140655360664523,
+      "learning_rate": 2.54521287834549e-06,
+      "loss": 0.6123,
+      "step": 32885
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5137990506452688,
+      "learning_rate": 2.5447043107200895e-06,
+      "loss": 0.724,
+      "step": 32886
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.321880734025286,
+      "learning_rate": 2.5441957865017784e-06,
+      "loss": 0.6034,
+      "step": 32887
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2180953202378009,
+      "learning_rate": 2.543687305693523e-06,
+      "loss": 0.7121,
+      "step": 32888
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.386833837092174,
+      "learning_rate": 2.543178868298277e-06,
+      "loss": 0.7168,
+      "step": 32889
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9269841661348837,
+      "learning_rate": 2.542670474319008e-06,
+      "loss": 0.6889,
+      "step": 32890
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5916790832155132,
+      "learning_rate": 2.542162123758669e-06,
+      "loss": 0.6118,
+      "step": 32891
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4127747276259035,
+      "learning_rate": 2.5416538166202267e-06,
+      "loss": 0.6261,
+      "step": 32892
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4760425903292786,
+      "learning_rate": 2.5411455529066363e-06,
+      "loss": 0.6669,
+      "step": 32893
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5040317447142082,
+      "learning_rate": 2.5406373326208554e-06,
+      "loss": 0.6672,
+      "step": 32894
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3303179575225008,
+      "learning_rate": 2.5401291557658457e-06,
+      "loss": 0.8013,
+      "step": 32895
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3856462529397435,
+      "learning_rate": 2.5396210223445684e-06,
+      "loss": 0.6835,
+      "step": 32896
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4104934012202186,
+      "learning_rate": 2.5391129323599796e-06,
+      "loss": 0.7202,
+      "step": 32897
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5530809070884843,
+      "learning_rate": 2.5386048858150346e-06,
+      "loss": 0.7382,
+      "step": 32898
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5970269438930802,
+      "learning_rate": 2.5380968827126937e-06,
+      "loss": 0.6963,
+      "step": 32899
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6298535539546963,
+      "learning_rate": 2.5375889230559192e-06,
+      "loss": 0.6986,
+      "step": 32900
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6983630943112709,
+      "learning_rate": 2.5370810068476637e-06,
+      "loss": 0.6755,
+      "step": 32901
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3581746678176376,
+      "learning_rate": 2.536573134090883e-06,
+      "loss": 0.6169,
+      "step": 32902
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2819818764884114,
+      "learning_rate": 2.5360653047885352e-06,
+      "loss": 0.666,
+      "step": 32903
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3622573306370631,
+      "learning_rate": 2.535557518943582e-06,
+      "loss": 0.6723,
+      "step": 32904
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.00100680170917,
+      "learning_rate": 2.5350497765589765e-06,
+      "loss": 0.6567,
+      "step": 32905
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.391403811420742,
+      "learning_rate": 2.5345420776376706e-06,
+      "loss": 0.7697,
+      "step": 32906
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.40720175209608,
+      "learning_rate": 2.5340344221826287e-06,
+      "loss": 0.6522,
+      "step": 32907
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6030933031430875,
+      "learning_rate": 2.533526810196798e-06,
+      "loss": 0.4598,
+      "step": 32908
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5997472811828252,
+      "learning_rate": 2.5330192416831413e-06,
+      "loss": 0.4977,
+      "step": 32909
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3865362807162962,
+      "learning_rate": 2.5325117166446065e-06,
+      "loss": 0.6116,
+      "step": 32910
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.490276582234319,
+      "learning_rate": 2.5320042350841577e-06,
+      "loss": 0.7441,
+      "step": 32911
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4358477256944886,
+      "learning_rate": 2.53149679700474e-06,
+      "loss": 0.701,
+      "step": 32912
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.345813111972195,
+      "learning_rate": 2.5309894024093163e-06,
+      "loss": 0.7149,
+      "step": 32913
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.683215674199371,
+      "learning_rate": 2.5304820513008364e-06,
+      "loss": 0.7701,
+      "step": 32914
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4819062454431544,
+      "learning_rate": 2.529974743682252e-06,
+      "loss": 0.7349,
+      "step": 32915
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3801287925908037,
+      "learning_rate": 2.52946747955652e-06,
+      "loss": 0.4931,
+      "step": 32916
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.340831480857522,
+      "learning_rate": 2.528960258926596e-06,
+      "loss": 0.656,
+      "step": 32917
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3288010004404247,
+      "learning_rate": 2.52845308179543e-06,
+      "loss": 0.6758,
+      "step": 32918
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0468653270333856,
+      "learning_rate": 2.527945948165974e-06,
+      "loss": 0.6287,
+      "step": 32919
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.031307645011868,
+      "learning_rate": 2.5274388580411813e-06,
+      "loss": 0.7741,
+      "step": 32920
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4039922228419457,
+      "learning_rate": 2.5269318114240092e-06,
+      "loss": 0.6998,
+      "step": 32921
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6026915434167419,
+      "learning_rate": 2.5264248083174047e-06,
+      "loss": 0.4712,
+      "step": 32922
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3717954074502716,
+      "learning_rate": 2.5259178487243197e-06,
+      "loss": 0.6804,
+      "step": 32923
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2979601852107299,
+      "learning_rate": 2.525410932647706e-06,
+      "loss": 0.6955,
+      "step": 32924
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7790795720656811,
+      "learning_rate": 2.5249040600905195e-06,
+      "loss": 0.6198,
+      "step": 32925
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5186082922882225,
+      "learning_rate": 2.5243972310557087e-06,
+      "loss": 0.6743,
+      "step": 32926
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2726084673802442,
+      "learning_rate": 2.523890445546221e-06,
+      "loss": 0.6216,
+      "step": 32927
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.628836924455772,
+      "learning_rate": 2.523383703565012e-06,
+      "loss": 0.7645,
+      "step": 32928
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4045376875931777,
+      "learning_rate": 2.5228770051150274e-06,
+      "loss": 0.7497,
+      "step": 32929
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5701399319811986,
+      "learning_rate": 2.5223703501992234e-06,
+      "loss": 0.4778,
+      "step": 32930
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.309095029644567,
+      "learning_rate": 2.521863738820545e-06,
+      "loss": 0.6512,
+      "step": 32931
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4077929661427238,
+      "learning_rate": 2.521357170981944e-06,
+      "loss": 0.6359,
+      "step": 32932
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8745578080766008,
+      "learning_rate": 2.520850646686368e-06,
+      "loss": 0.7164,
+      "step": 32933
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3271049693169563,
+      "learning_rate": 2.52034416593677e-06,
+      "loss": 0.5927,
+      "step": 32934
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0419633174755103,
+      "learning_rate": 2.5198377287360954e-06,
+      "loss": 0.7841,
+      "step": 32935
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3149570409223863,
+      "learning_rate": 2.519331335087292e-06,
+      "loss": 0.624,
+      "step": 32936
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6450875627884703,
+      "learning_rate": 2.5188249849933087e-06,
+      "loss": 0.5136,
+      "step": 32937
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3390184387744437,
+      "learning_rate": 2.5183186784570957e-06,
+      "loss": 0.6259,
+      "step": 32938
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4493931125476454,
+      "learning_rate": 2.5178124154816052e-06,
+      "loss": 0.5801,
+      "step": 32939
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6130551290566493,
+      "learning_rate": 2.5173061960697744e-06,
+      "loss": 0.7236,
+      "step": 32940
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6017167067421925,
+      "learning_rate": 2.516800020224556e-06,
+      "loss": 0.6373,
+      "step": 32941
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.344143735302223,
+      "learning_rate": 2.5162938879488996e-06,
+      "loss": 0.6475,
+      "step": 32942
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6100428883722746,
+      "learning_rate": 2.5157877992457493e-06,
+      "loss": 0.5015,
+      "step": 32943
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3705403571980275,
+      "learning_rate": 2.5152817541180484e-06,
+      "loss": 0.5374,
+      "step": 32944
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3448205495287686,
+      "learning_rate": 2.5147757525687477e-06,
+      "loss": 0.7998,
+      "step": 32945
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3560206028032498,
+      "learning_rate": 2.5142697946007955e-06,
+      "loss": 0.6883,
+      "step": 32946
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.483690959005992,
+      "learning_rate": 2.5137638802171337e-06,
+      "loss": 0.6829,
+      "step": 32947
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2642462842033313,
+      "learning_rate": 2.513258009420707e-06,
+      "loss": 0.6756,
+      "step": 32948
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8192388510449102,
+      "learning_rate": 2.512752182214464e-06,
+      "loss": 0.7689,
+      "step": 32949
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.339792205772499,
+      "learning_rate": 2.5122463986013458e-06,
+      "loss": 0.6536,
+      "step": 32950
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6150213810593771,
+      "learning_rate": 2.5117406585842998e-06,
+      "loss": 0.5301,
+      "step": 32951
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4615942383838707,
+      "learning_rate": 2.5112349621662736e-06,
+      "loss": 0.6955,
+      "step": 32952
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3507137644184413,
+      "learning_rate": 2.510729309350208e-06,
+      "loss": 0.6857,
+      "step": 32953
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7599219320577737,
+      "learning_rate": 2.510223700139044e-06,
+      "loss": 0.6873,
+      "step": 32954
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4465755550606785,
+      "learning_rate": 2.509718134535729e-06,
+      "loss": 0.6519,
+      "step": 32955
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.343796299416845,
+      "learning_rate": 2.5092126125432137e-06,
+      "loss": 0.6772,
+      "step": 32956
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1878712317132085,
+      "learning_rate": 2.5087071341644265e-06,
+      "loss": 0.7285,
+      "step": 32957
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5399282572304376,
+      "learning_rate": 2.5082016994023185e-06,
+      "loss": 0.6595,
+      "step": 32958
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.31380388942347,
+      "learning_rate": 2.5076963082598327e-06,
+      "loss": 0.7882,
+      "step": 32959
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4191037561718247,
+      "learning_rate": 2.507190960739917e-06,
+      "loss": 0.7897,
+      "step": 32960
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4651246135190679,
+      "learning_rate": 2.5066856568455012e-06,
+      "loss": 0.7138,
+      "step": 32961
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3136564924882062,
+      "learning_rate": 2.506180396579534e-06,
+      "loss": 0.6721,
+      "step": 32962
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3844703840053076,
+      "learning_rate": 2.505675179944956e-06,
+      "loss": 0.6555,
+      "step": 32963
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2813488694671957,
+      "learning_rate": 2.5051700069447136e-06,
+      "loss": 0.6557,
+      "step": 32964
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3047390888523676,
+      "learning_rate": 2.504664877581745e-06,
+      "loss": 0.6169,
+      "step": 32965
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3421409438192473,
+      "learning_rate": 2.5041597918589867e-06,
+      "loss": 0.7175,
+      "step": 32966
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4511307355798408,
+      "learning_rate": 2.5036547497793863e-06,
+      "loss": 0.7728,
+      "step": 32967
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3010010218112038,
+      "learning_rate": 2.503149751345878e-06,
+      "loss": 0.6679,
+      "step": 32968
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3178595935589008,
+      "learning_rate": 2.5026447965614086e-06,
+      "loss": 0.5799,
+      "step": 32969
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4250266302204726,
+      "learning_rate": 2.5021398854289124e-06,
+      "loss": 0.5572,
+      "step": 32970
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.068473727866685,
+      "learning_rate": 2.501635017951334e-06,
+      "loss": 0.6069,
+      "step": 32971
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0110546776147844,
+      "learning_rate": 2.501130194131608e-06,
+      "loss": 0.6572,
+      "step": 32972
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7386113300451491,
+      "learning_rate": 2.500625413972678e-06,
+      "loss": 0.5493,
+      "step": 32973
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5917865415413965,
+      "learning_rate": 2.5001206774774813e-06,
+      "loss": 0.66,
+      "step": 32974
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.373203770960345,
+      "learning_rate": 2.4996159846489544e-06,
+      "loss": 0.7443,
+      "step": 32975
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5120270774785212,
+      "learning_rate": 2.4991113354900374e-06,
+      "loss": 0.5905,
+      "step": 32976
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3691209583763566,
+      "learning_rate": 2.4986067300036708e-06,
+      "loss": 0.6638,
+      "step": 32977
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4220725074989053,
+      "learning_rate": 2.498102168192792e-06,
+      "loss": 0.7248,
+      "step": 32978
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2724705265154843,
+      "learning_rate": 2.4975976500603348e-06,
+      "loss": 0.7267,
+      "step": 32979
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9379008329858063,
+      "learning_rate": 2.4970931756092387e-06,
+      "loss": 0.6603,
+      "step": 32980
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4634436520088216,
+      "learning_rate": 2.4965887448424443e-06,
+      "loss": 0.6171,
+      "step": 32981
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.492285168512619,
+      "learning_rate": 2.4960843577628858e-06,
+      "loss": 0.7889,
+      "step": 32982
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6088984084374248,
+      "learning_rate": 2.4955800143734966e-06,
+      "loss": 0.4993,
+      "step": 32983
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.328013257342334,
+      "learning_rate": 2.4950757146772175e-06,
+      "loss": 0.823,
+      "step": 32984
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6323649653173968,
+      "learning_rate": 2.4945714586769854e-06,
+      "loss": 0.7183,
+      "step": 32985
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3500089139706668,
+      "learning_rate": 2.4940672463757353e-06,
+      "loss": 0.5876,
+      "step": 32986
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3118676818205843,
+      "learning_rate": 2.493563077776399e-06,
+      "loss": 0.7381,
+      "step": 32987
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4039811264763193,
+      "learning_rate": 2.4930589528819172e-06,
+      "loss": 0.7942,
+      "step": 32988
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7309285794157367,
+      "learning_rate": 2.492554871695221e-06,
+      "loss": 0.7604,
+      "step": 32989
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.541957940344313,
+      "learning_rate": 2.4920508342192485e-06,
+      "loss": 0.5982,
+      "step": 32990
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3917703807329018,
+      "learning_rate": 2.491546840456931e-06,
+      "loss": 0.71,
+      "step": 32991
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3096657393597206,
+      "learning_rate": 2.491042890411208e-06,
+      "loss": 0.7476,
+      "step": 32992
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.674516880954435,
+      "learning_rate": 2.490538984085008e-06,
+      "loss": 0.6409,
+      "step": 32993
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.498356667163467,
+      "learning_rate": 2.4900351214812703e-06,
+      "loss": 0.7347,
+      "step": 32994
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3977146170063686,
+      "learning_rate": 2.4895313026029243e-06,
+      "loss": 0.7465,
+      "step": 32995
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2052658515328118,
+      "learning_rate": 2.489027527452903e-06,
+      "loss": 0.6685,
+      "step": 32996
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.57292473849375,
+      "learning_rate": 2.4885237960341413e-06,
+      "loss": 0.5724,
+      "step": 32997
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2961196741670598,
+      "learning_rate": 2.488020108349576e-06,
+      "loss": 0.6996,
+      "step": 32998
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4351482784459362,
+      "learning_rate": 2.4875164644021343e-06,
+      "loss": 0.6438,
+      "step": 32999
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2490654897097808,
+      "learning_rate": 2.487012864194748e-06,
+      "loss": 0.7524,
+      "step": 33000
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6782787347688406,
+      "learning_rate": 2.4865093077303514e-06,
+      "loss": 0.718,
+      "step": 33001
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5649325981279085,
+      "learning_rate": 2.48600579501188e-06,
+      "loss": 0.6582,
+      "step": 33002
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3709777367075686,
+      "learning_rate": 2.4855023260422605e-06,
+      "loss": 0.7612,
+      "step": 33003
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6599132530579048,
+      "learning_rate": 2.4849989008244224e-06,
+      "loss": 0.7364,
+      "step": 33004
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5247361183873676,
+      "learning_rate": 2.4844955193613007e-06,
+      "loss": 0.6449,
+      "step": 33005
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.263525905814764,
+      "learning_rate": 2.483992181655829e-06,
+      "loss": 0.5832,
+      "step": 33006
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4598374021555336,
+      "learning_rate": 2.483488887710933e-06,
+      "loss": 0.7051,
+      "step": 33007
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2888896433506827,
+      "learning_rate": 2.4829856375295426e-06,
+      "loss": 0.6899,
+      "step": 33008
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8373871999419888,
+      "learning_rate": 2.482482431114591e-06,
+      "loss": 0.5713,
+      "step": 33009
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4604909975210094,
+      "learning_rate": 2.4819792684690046e-06,
+      "loss": 0.6303,
+      "step": 33010
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8174063391033217,
+      "learning_rate": 2.4814761495957183e-06,
+      "loss": 0.6335,
+      "step": 33011
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7014931276454557,
+      "learning_rate": 2.4809730744976556e-06,
+      "loss": 0.7303,
+      "step": 33012
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3157370895848766,
+      "learning_rate": 2.480470043177751e-06,
+      "loss": 0.7081,
+      "step": 33013
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5844104027175154,
+      "learning_rate": 2.4799670556389267e-06,
+      "loss": 0.4752,
+      "step": 33014
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.598220593152543,
+      "learning_rate": 2.479464111884118e-06,
+      "loss": 0.653,
+      "step": 33015
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5686617939602243,
+      "learning_rate": 2.478961211916251e-06,
+      "loss": 0.6738,
+      "step": 33016
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5373292091282023,
+      "learning_rate": 2.4784583557382492e-06,
+      "loss": 0.6454,
+      "step": 33017
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8244684797582567,
+      "learning_rate": 2.4779555433530446e-06,
+      "loss": 0.8298,
+      "step": 33018
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3117007745932183,
+      "learning_rate": 2.4774527747635667e-06,
+      "loss": 0.736,
+      "step": 33019
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4975091995457692,
+      "learning_rate": 2.4769500499727417e-06,
+      "loss": 0.9228,
+      "step": 33020
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6052867067314016,
+      "learning_rate": 2.4764473689834912e-06,
+      "loss": 0.627,
+      "step": 33021
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.327829413290607,
+      "learning_rate": 2.475944731798746e-06,
+      "loss": 0.5746,
+      "step": 33022
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5389037691651448,
+      "learning_rate": 2.4754421384214355e-06,
+      "loss": 0.8149,
+      "step": 33023
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3898382565856071,
+      "learning_rate": 2.4749395888544837e-06,
+      "loss": 0.7615,
+      "step": 33024
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2868185448286806,
+      "learning_rate": 2.4744370831008125e-06,
+      "loss": 0.6928,
+      "step": 33025
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4133776581603927,
+      "learning_rate": 2.473934621163352e-06,
+      "loss": 0.707,
+      "step": 33026
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.480538173472024,
+      "learning_rate": 2.4734322030450286e-06,
+      "loss": 0.7478,
+      "step": 33027
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2932612570274855,
+      "learning_rate": 2.472929828748767e-06,
+      "loss": 0.6926,
+      "step": 33028
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7291061525709837,
+      "learning_rate": 2.4724274982774866e-06,
+      "loss": 0.7281,
+      "step": 33029
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9867229782843685,
+      "learning_rate": 2.4719252116341207e-06,
+      "loss": 0.5961,
+      "step": 33030
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3808114850147784,
+      "learning_rate": 2.4714229688215864e-06,
+      "loss": 0.8065,
+      "step": 33031
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2268457379563846,
+      "learning_rate": 2.470920769842814e-06,
+      "loss": 0.7392,
+      "step": 33032
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5768368741827645,
+      "learning_rate": 2.4704186147007216e-06,
+      "loss": 0.7558,
+      "step": 33033
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.43437632962042,
+      "learning_rate": 2.4699165033982377e-06,
+      "loss": 0.6391,
+      "step": 33034
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6170709013369386,
+      "learning_rate": 2.469414435938282e-06,
+      "loss": 0.7254,
+      "step": 33035
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7228825686595737,
+      "learning_rate": 2.4689124123237795e-06,
+      "loss": 0.6488,
+      "step": 33036
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.572740110443717,
+      "learning_rate": 2.4684104325576595e-06,
+      "loss": 0.6898,
+      "step": 33037
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5071959404756892,
+      "learning_rate": 2.467908496642831e-06,
+      "loss": 0.8033,
+      "step": 33038
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3885956728978293,
+      "learning_rate": 2.4674066045822256e-06,
+      "loss": 0.847,
+      "step": 33039
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.85028185854613,
+      "learning_rate": 2.466904756378763e-06,
+      "loss": 0.8521,
+      "step": 33040
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3211474805788814,
+      "learning_rate": 2.4664029520353716e-06,
+      "loss": 0.5626,
+      "step": 33041
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.395507287976942,
+      "learning_rate": 2.4659011915549626e-06,
+      "loss": 0.5848,
+      "step": 33042
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5591982439917114,
+      "learning_rate": 2.465399474940461e-06,
+      "loss": 0.5923,
+      "step": 33043
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5560578894781236,
+      "learning_rate": 2.4648978021947923e-06,
+      "loss": 0.6887,
+      "step": 33044
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2851989516138882,
+      "learning_rate": 2.464396173320873e-06,
+      "loss": 0.7076,
+      "step": 33045
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4359344007306494,
+      "learning_rate": 2.4638945883216236e-06,
+      "loss": 0.7131,
+      "step": 33046
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3404471291687787,
+      "learning_rate": 2.4633930471999657e-06,
+      "loss": 0.7743,
+      "step": 33047
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.91650049610199,
+      "learning_rate": 2.4628915499588213e-06,
+      "loss": 0.8414,
+      "step": 33048
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3758059951301784,
+      "learning_rate": 2.4623900966011062e-06,
+      "loss": 0.6672,
+      "step": 33049
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4004479434295931,
+      "learning_rate": 2.4618886871297454e-06,
+      "loss": 0.8182,
+      "step": 33050
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6478626535482335,
+      "learning_rate": 2.461387321547656e-06,
+      "loss": 0.5334,
+      "step": 33051
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5191088159229675,
+      "learning_rate": 2.4608859998577515e-06,
+      "loss": 0.8197,
+      "step": 33052
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.400398936217493,
+      "learning_rate": 2.4603847220629575e-06,
+      "loss": 0.66,
+      "step": 33053
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4251947690528937,
+      "learning_rate": 2.4598834881661914e-06,
+      "loss": 0.7671,
+      "step": 33054
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4304116366731303,
+      "learning_rate": 2.459382298170373e-06,
+      "loss": 0.6113,
+      "step": 33055
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6392099458863012,
+      "learning_rate": 2.4588811520784138e-06,
+      "loss": 0.5052,
+      "step": 33056
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6245598290764146,
+      "learning_rate": 2.4583800498932365e-06,
+      "loss": 0.5248,
+      "step": 33057
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5929908008638092,
+      "learning_rate": 2.4578789916177614e-06,
+      "loss": 0.4741,
+      "step": 33058
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4618016692420621,
+      "learning_rate": 2.4573779772549023e-06,
+      "loss": 0.6426,
+      "step": 33059
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3930227069875394,
+      "learning_rate": 2.456877006807573e-06,
+      "loss": 0.71,
+      "step": 33060
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.467513080425064,
+      "learning_rate": 2.456376080278695e-06,
+      "loss": 0.6968,
+      "step": 33061
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.260039215588193,
+      "learning_rate": 2.4558751976711857e-06,
+      "loss": 0.7113,
+      "step": 33062
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2910734659091503,
+      "learning_rate": 2.45537435898796e-06,
+      "loss": 0.6355,
+      "step": 33063
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3279333130345785,
+      "learning_rate": 2.4548735642319297e-06,
+      "loss": 0.7135,
+      "step": 33064
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2823280026761354,
+      "learning_rate": 2.4543728134060186e-06,
+      "loss": 0.6929,
+      "step": 33065
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6994836064907675,
+      "learning_rate": 2.4538721065131344e-06,
+      "loss": 0.6062,
+      "step": 33066
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3056861822226684,
+      "learning_rate": 2.453371443556198e-06,
+      "loss": 0.6602,
+      "step": 33067
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.34839071471662,
+      "learning_rate": 2.4528708245381203e-06,
+      "loss": 0.7256,
+      "step": 33068
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2772780736423375,
+      "learning_rate": 2.4523702494618206e-06,
+      "loss": 0.5892,
+      "step": 33069
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2292559732372885,
+      "learning_rate": 2.4518697183302076e-06,
+      "loss": 0.5711,
+      "step": 33070
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6031079326346416,
+      "learning_rate": 2.4513692311462024e-06,
+      "loss": 0.4936,
+      "step": 33071
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3037069460811308,
+      "learning_rate": 2.4508687879127145e-06,
+      "loss": 0.6319,
+      "step": 33072
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.433842182452394,
+      "learning_rate": 2.4503683886326557e-06,
+      "loss": 0.8181,
+      "step": 33073
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4187524315588969,
+      "learning_rate": 2.449868033308943e-06,
+      "loss": 0.6171,
+      "step": 33074
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6230445989939675,
+      "learning_rate": 2.449367721944492e-06,
+      "loss": 0.7453,
+      "step": 33075
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3739756128894391,
+      "learning_rate": 2.448867454542212e-06,
+      "loss": 0.7625,
+      "step": 33076
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.359710922900448,
+      "learning_rate": 2.4483672311050134e-06,
+      "loss": 0.664,
+      "step": 33077
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3136373951002163,
+      "learning_rate": 2.4478670516358116e-06,
+      "loss": 0.5916,
+      "step": 33078
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5110991979253117,
+      "learning_rate": 2.4473669161375225e-06,
+      "loss": 0.7574,
+      "step": 33079
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4088744140294116,
+      "learning_rate": 2.4468668246130534e-06,
+      "loss": 0.7061,
+      "step": 33080
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.378580922291801,
+      "learning_rate": 2.4463667770653144e-06,
+      "loss": 0.6256,
+      "step": 33081
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1479151897358966,
+      "learning_rate": 2.4458667734972196e-06,
+      "loss": 0.6158,
+      "step": 33082
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4219270269334787,
+      "learning_rate": 2.4453668139116824e-06,
+      "loss": 0.6783,
+      "step": 33083
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3449756090576215,
+      "learning_rate": 2.444866898311612e-06,
+      "loss": 0.6856,
+      "step": 33084
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.257701842036337,
+      "learning_rate": 2.4443670266999155e-06,
+      "loss": 0.5749,
+      "step": 33085
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3426985673405374,
+      "learning_rate": 2.443867199079506e-06,
+      "loss": 0.7506,
+      "step": 33086
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5293449206987517,
+      "learning_rate": 2.443367415453297e-06,
+      "loss": 0.6651,
+      "step": 33087
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5804325656996128,
+      "learning_rate": 2.4428676758241953e-06,
+      "loss": 0.4866,
+      "step": 33088
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6055982186670412,
+      "learning_rate": 2.4423679801951073e-06,
+      "loss": 0.7553,
+      "step": 33089
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6366441751941512,
+      "learning_rate": 2.441868328568948e-06,
+      "loss": 0.5265,
+      "step": 33090
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2613363647282254,
+      "learning_rate": 2.4413687209486223e-06,
+      "loss": 0.6525,
+      "step": 33091
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.585666211749502,
+      "learning_rate": 2.4408691573370438e-06,
+      "loss": 0.4761,
+      "step": 33092
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.447955992820726,
+      "learning_rate": 2.4403696377371146e-06,
+      "loss": 0.7038,
+      "step": 33093
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5377403205509441,
+      "learning_rate": 2.4398701621517507e-06,
+      "loss": 0.74,
+      "step": 33094
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2982268534097547,
+      "learning_rate": 2.439370730583852e-06,
+      "loss": 0.6808,
+      "step": 33095
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5648492689460456,
+      "learning_rate": 2.4388713430363343e-06,
+      "loss": 0.4813,
+      "step": 33096
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4004460469538869,
+      "learning_rate": 2.4383719995121013e-06,
+      "loss": 0.8201,
+      "step": 33097
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5204316965544857,
+      "learning_rate": 2.437872700014058e-06,
+      "loss": 0.5913,
+      "step": 33098
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3114637930533308,
+      "learning_rate": 2.4373734445451127e-06,
+      "loss": 0.6278,
+      "step": 33099
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2314892034249034,
+      "learning_rate": 2.4368742331081775e-06,
+      "loss": 0.6233,
+      "step": 33100
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3878652463151462,
+      "learning_rate": 2.4363750657061546e-06,
+      "loss": 0.6194,
+      "step": 33101
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2936160538445935,
+      "learning_rate": 2.4358759423419476e-06,
+      "loss": 0.5547,
+      "step": 33102
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5319668430888884,
+      "learning_rate": 2.4353768630184658e-06,
+      "loss": 0.6937,
+      "step": 33103
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3734918527823616,
+      "learning_rate": 2.4348778277386177e-06,
+      "loss": 0.7275,
+      "step": 33104
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.39793379328262,
+      "learning_rate": 2.4343788365053066e-06,
+      "loss": 0.8077,
+      "step": 33105
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2064090651416508,
+      "learning_rate": 2.4338798893214332e-06,
+      "loss": 0.692,
+      "step": 33106
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.524268329699728,
+      "learning_rate": 2.433380986189907e-06,
+      "loss": 0.6534,
+      "step": 33107
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.375561467480244,
+      "learning_rate": 2.432882127113635e-06,
+      "loss": 0.6038,
+      "step": 33108
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4611614993945232,
+      "learning_rate": 2.4323833120955197e-06,
+      "loss": 0.6523,
+      "step": 33109
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7093366874978246,
+      "learning_rate": 2.4318845411384608e-06,
+      "loss": 0.6747,
+      "step": 33110
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.449867977391724,
+      "learning_rate": 2.4313858142453695e-06,
+      "loss": 0.5864,
+      "step": 33111
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3651671277502344,
+      "learning_rate": 2.4308871314191427e-06,
+      "loss": 0.699,
+      "step": 33112
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4756884911042023,
+      "learning_rate": 2.430388492662691e-06,
+      "loss": 0.6619,
+      "step": 33113
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.450011473921783,
+      "learning_rate": 2.429889897978911e-06,
+      "loss": 0.6489,
+      "step": 33114
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5861762001305654,
+      "learning_rate": 2.4293913473707108e-06,
+      "loss": 0.5404,
+      "step": 33115
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.274680845428128,
+      "learning_rate": 2.4288928408409885e-06,
+      "loss": 0.6036,
+      "step": 33116
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.601999417040405,
+      "learning_rate": 2.4283943783926523e-06,
+      "loss": 0.4764,
+      "step": 33117
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4280946889222932,
+      "learning_rate": 2.427895960028601e-06,
+      "loss": 0.7083,
+      "step": 33118
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2954020713590881,
+      "learning_rate": 2.427397585751734e-06,
+      "loss": 0.609,
+      "step": 33119
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4839824323379915,
+      "learning_rate": 2.4268992555649553e-06,
+      "loss": 0.7045,
+      "step": 33120
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5993011679251394,
+      "learning_rate": 2.4264009694711665e-06,
+      "loss": 0.7035,
+      "step": 33121
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6337390061337442,
+      "learning_rate": 2.4259027274732748e-06,
+      "loss": 0.4934,
+      "step": 33122
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7851936980023517,
+      "learning_rate": 2.42540452957417e-06,
+      "loss": 0.6457,
+      "step": 33123
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.195486596849545,
+      "learning_rate": 2.4249063757767587e-06,
+      "loss": 0.7122,
+      "step": 33124
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5886290106625527,
+      "learning_rate": 2.424408266083943e-06,
+      "loss": 0.6996,
+      "step": 33125
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5391305602674596,
+      "learning_rate": 2.4239102004986203e-06,
+      "loss": 0.5652,
+      "step": 33126
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5925153122188767,
+      "learning_rate": 2.4234121790236887e-06,
+      "loss": 0.7248,
+      "step": 33127
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6047221561425102,
+      "learning_rate": 2.42291420166205e-06,
+      "loss": 0.4888,
+      "step": 33128
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7669055120175368,
+      "learning_rate": 2.4224162684166064e-06,
+      "loss": 0.7541,
+      "step": 33129
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6616300593030453,
+      "learning_rate": 2.4219183792902544e-06,
+      "loss": 0.6821,
+      "step": 33130
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3002879118971005,
+      "learning_rate": 2.4214205342858897e-06,
+      "loss": 0.6846,
+      "step": 33131
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4588799392118197,
+      "learning_rate": 2.4209227334064167e-06,
+      "loss": 0.7139,
+      "step": 33132
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5532665739090954,
+      "learning_rate": 2.4204249766547295e-06,
+      "loss": 0.6745,
+      "step": 33133
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2332641126169592,
+      "learning_rate": 2.4199272640337256e-06,
+      "loss": 0.7857,
+      "step": 33134
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3764216598986185,
+      "learning_rate": 2.41942959554631e-06,
+      "loss": 0.5454,
+      "step": 33135
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3819271782468099,
+      "learning_rate": 2.418931971195374e-06,
+      "loss": 0.6532,
+      "step": 33136
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.612646843337603,
+      "learning_rate": 2.4184343909838135e-06,
+      "loss": 0.6691,
+      "step": 33137
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.48964549879274,
+      "learning_rate": 2.4179368549145286e-06,
+      "loss": 0.6019,
+      "step": 33138
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4339349060353166,
+      "learning_rate": 2.4174393629904214e-06,
+      "loss": 0.81,
+      "step": 33139
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3895245685220574,
+      "learning_rate": 2.416941915214377e-06,
+      "loss": 0.7179,
+      "step": 33140
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2143690117209565,
+      "learning_rate": 2.4164445115892966e-06,
+      "loss": 0.632,
+      "step": 33141
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.345116464925688,
+      "learning_rate": 2.415947152118079e-06,
+      "loss": 0.6289,
+      "step": 33142
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2849647209316462,
+      "learning_rate": 2.4154498368036226e-06,
+      "loss": 0.7118,
+      "step": 33143
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2922124852701182,
+      "learning_rate": 2.4149525656488128e-06,
+      "loss": 0.7336,
+      "step": 33144
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.499607676498777,
+      "learning_rate": 2.4144553386565504e-06,
+      "loss": 0.755,
+      "step": 33145
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3765268994130822,
+      "learning_rate": 2.4139581558297344e-06,
+      "loss": 0.7026,
+      "step": 33146
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2635959860474941,
+      "learning_rate": 2.4134610171712525e-06,
+      "loss": 0.72,
+      "step": 33147
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4007018961136544,
+      "learning_rate": 2.412963922684004e-06,
+      "loss": 0.6692,
+      "step": 33148
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3021289537665834,
+      "learning_rate": 2.41246687237088e-06,
+      "loss": 0.6207,
+      "step": 33149
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2986609930407986,
+      "learning_rate": 2.411969866234779e-06,
+      "loss": 0.5274,
+      "step": 33150
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6806399289394165,
+      "learning_rate": 2.411472904278589e-06,
+      "loss": 0.6082,
+      "step": 33151
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7945094417092866,
+      "learning_rate": 2.4109759865052084e-06,
+      "loss": 0.6716,
+      "step": 33152
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4142323668115757,
+      "learning_rate": 2.4104791129175286e-06,
+      "loss": 0.7726,
+      "step": 33153
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5249851294915517,
+      "learning_rate": 2.4099822835184392e-06,
+      "loss": 0.5885,
+      "step": 33154
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5617142767201765,
+      "learning_rate": 2.409485498310835e-06,
+      "loss": 0.5068,
+      "step": 33155
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5763868672072614,
+      "learning_rate": 2.4089887572976134e-06,
+      "loss": 0.6804,
+      "step": 33156
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.39702317866876,
+      "learning_rate": 2.408492060481663e-06,
+      "loss": 0.6736,
+      "step": 33157
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.775102212483273,
+      "learning_rate": 2.407995407865872e-06,
+      "loss": 0.6309,
+      "step": 33158
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3780371578971191,
+      "learning_rate": 2.407498799453134e-06,
+      "loss": 0.6454,
+      "step": 33159
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4482138677421004,
+      "learning_rate": 2.407002235246346e-06,
+      "loss": 0.631,
+      "step": 33160
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4492932959613802,
+      "learning_rate": 2.406505715248395e-06,
+      "loss": 0.7119,
+      "step": 33161
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4819922152414302,
+      "learning_rate": 2.406009239462168e-06,
+      "loss": 0.6837,
+      "step": 33162
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3714739009367725,
+      "learning_rate": 2.40551280789056e-06,
+      "loss": 0.8612,
+      "step": 33163
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6167165671268646,
+      "learning_rate": 2.4050164205364633e-06,
+      "loss": 0.4877,
+      "step": 33164
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5481977251963681,
+      "learning_rate": 2.4045200774027653e-06,
+      "loss": 0.5611,
+      "step": 33165
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2910634507548708,
+      "learning_rate": 2.404023778492354e-06,
+      "loss": 0.6356,
+      "step": 33166
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4599580392262548,
+      "learning_rate": 2.4035275238081236e-06,
+      "loss": 0.5458,
+      "step": 33167
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6786591658504626,
+      "learning_rate": 2.4030313133529583e-06,
+      "loss": 0.69,
+      "step": 33168
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3588342601993109,
+      "learning_rate": 2.402535147129753e-06,
+      "loss": 0.7224,
+      "step": 33169
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2890009069606367,
+      "learning_rate": 2.4020390251413893e-06,
+      "loss": 0.7594,
+      "step": 33170
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5201795840509411,
+      "learning_rate": 2.4015429473907637e-06,
+      "loss": 0.6736,
+      "step": 33171
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4283203210714788,
+      "learning_rate": 2.4010469138807578e-06,
+      "loss": 0.6896,
+      "step": 33172
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4730384119372553,
+      "learning_rate": 2.4005509246142655e-06,
+      "loss": 0.7353,
+      "step": 33173
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.395164372624805,
+      "learning_rate": 2.400054979594172e-06,
+      "loss": 0.7452,
+      "step": 33174
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4589737979399076,
+      "learning_rate": 2.3995590788233604e-06,
+      "loss": 0.6117,
+      "step": 33175
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7480869544851523,
+      "learning_rate": 2.399063222304723e-06,
+      "loss": 0.6591,
+      "step": 33176
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1598640376593938,
+      "learning_rate": 2.3985674100411484e-06,
+      "loss": 0.7159,
+      "step": 33177
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5727029527593769,
+      "learning_rate": 2.398071642035521e-06,
+      "loss": 0.478,
+      "step": 33178
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2661138849623852,
+      "learning_rate": 2.3975759182907244e-06,
+      "loss": 0.6951,
+      "step": 33179
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6893702257305725,
+      "learning_rate": 2.3970802388096483e-06,
+      "loss": 0.5776,
+      "step": 33180
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4150258563793923,
+      "learning_rate": 2.39658460359518e-06,
+      "loss": 0.6618,
+      "step": 33181
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5973711845549106,
+      "learning_rate": 2.3960890126502034e-06,
+      "loss": 0.6526,
+      "step": 33182
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.38351253512761,
+      "learning_rate": 2.3955934659776013e-06,
+      "loss": 0.6926,
+      "step": 33183
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2664520813721774,
+      "learning_rate": 2.3950979635802617e-06,
+      "loss": 0.679,
+      "step": 33184
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3060704566786339,
+      "learning_rate": 2.3946025054610722e-06,
+      "loss": 0.7458,
+      "step": 33185
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4954541569623705,
+      "learning_rate": 2.394107091622915e-06,
+      "loss": 0.779,
+      "step": 33186
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5968939150709615,
+      "learning_rate": 2.3936117220686717e-06,
+      "loss": 0.5042,
+      "step": 33187
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4579374679403638,
+      "learning_rate": 2.3931163968012315e-06,
+      "loss": 0.7724,
+      "step": 33188
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.373270849376537,
+      "learning_rate": 2.3926211158234725e-06,
+      "loss": 0.6573,
+      "step": 33189
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.109385559491759,
+      "learning_rate": 2.392125879138286e-06,
+      "loss": 0.7894,
+      "step": 33190
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3843719925009195,
+      "learning_rate": 2.3916306867485474e-06,
+      "loss": 0.7654,
+      "step": 33191
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4380786079872814,
+      "learning_rate": 2.3911355386571466e-06,
+      "loss": 0.5796,
+      "step": 33192
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4749573097171698,
+      "learning_rate": 2.390640434866961e-06,
+      "loss": 0.6222,
+      "step": 33193
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.478248895400497,
+      "learning_rate": 2.390145375380879e-06,
+      "loss": 0.8166,
+      "step": 33194
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5822844606318912,
+      "learning_rate": 2.3896503602017794e-06,
+      "loss": 0.5378,
+      "step": 33195
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6076723383302144,
+      "learning_rate": 2.3891553893325424e-06,
+      "loss": 0.4811,
+      "step": 33196
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7642241118314044,
+      "learning_rate": 2.388660462776051e-06,
+      "loss": 0.7718,
+      "step": 33197
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6076805320828764,
+      "learning_rate": 2.388165580535192e-06,
+      "loss": 0.7787,
+      "step": 33198
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3459405562962559,
+      "learning_rate": 2.387670742612843e-06,
+      "loss": 0.6911,
+      "step": 33199
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.769551765350799,
+      "learning_rate": 2.387175949011882e-06,
+      "loss": 0.6715,
+      "step": 33200
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2759656802256707,
+      "learning_rate": 2.3866811997351925e-06,
+      "loss": 0.606,
+      "step": 33201
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2510370639972597,
+      "learning_rate": 2.3861864947856583e-06,
+      "loss": 0.727,
+      "step": 33202
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.465158435528068,
+      "learning_rate": 2.385691834166157e-06,
+      "loss": 0.6496,
+      "step": 33203
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5359571416179616,
+      "learning_rate": 2.385197217879566e-06,
+      "loss": 0.6378,
+      "step": 33204
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4375053923880852,
+      "learning_rate": 2.384702645928767e-06,
+      "loss": 0.5961,
+      "step": 33205
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4025855487985706,
+      "learning_rate": 2.3842081183166434e-06,
+      "loss": 0.8043,
+      "step": 33206
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4592162484548816,
+      "learning_rate": 2.383713635046071e-06,
+      "loss": 0.7679,
+      "step": 33207
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2882277746689312,
+      "learning_rate": 2.383219196119927e-06,
+      "loss": 0.7017,
+      "step": 33208
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5835965548032243,
+      "learning_rate": 2.3827248015410907e-06,
+      "loss": 0.5063,
+      "step": 33209
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2029290549103386,
+      "learning_rate": 2.382230451312446e-06,
+      "loss": 0.6461,
+      "step": 33210
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.400828023630815,
+      "learning_rate": 2.381736145436867e-06,
+      "loss": 0.7762,
+      "step": 33211
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4735228564225253,
+      "learning_rate": 2.38124188391723e-06,
+      "loss": 0.7176,
+      "step": 33212
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3387223855319466,
+      "learning_rate": 2.3807476667564168e-06,
+      "loss": 0.7462,
+      "step": 33213
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7025371876050195,
+      "learning_rate": 2.3802534939573006e-06,
+      "loss": 0.7148,
+      "step": 33214
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0569800628891306,
+      "learning_rate": 2.3797593655227637e-06,
+      "loss": 0.778,
+      "step": 33215
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3825822425353558,
+      "learning_rate": 2.379265281455677e-06,
+      "loss": 0.5918,
+      "step": 33216
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2486756899276301,
+      "learning_rate": 2.378771241758924e-06,
+      "loss": 0.6594,
+      "step": 33217
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4064124838487195,
+      "learning_rate": 2.378277246435374e-06,
+      "loss": 0.6237,
+      "step": 33218
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.564692672647457,
+      "learning_rate": 2.3777832954879075e-06,
+      "loss": 0.6975,
+      "step": 33219
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2317460215605551,
+      "learning_rate": 2.377289388919406e-06,
+      "loss": 0.7347,
+      "step": 33220
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3905187257338856,
+      "learning_rate": 2.3767955267327325e-06,
+      "loss": 0.5926,
+      "step": 33221
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5894563858597801,
+      "learning_rate": 2.3763017089307692e-06,
+      "loss": 0.4763,
+      "step": 33222
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0500052045709145,
+      "learning_rate": 2.375807935516391e-06,
+      "loss": 0.632,
+      "step": 33223
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6165312108005845,
+      "learning_rate": 2.3753142064924785e-06,
+      "loss": 0.6148,
+      "step": 33224
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.482080836066997,
+      "learning_rate": 2.374820521861895e-06,
+      "loss": 0.8255,
+      "step": 33225
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5481452033754102,
+      "learning_rate": 2.37432688162752e-06,
+      "loss": 0.8092,
+      "step": 33226
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6465113883725946,
+      "learning_rate": 2.3738332857922307e-06,
+      "loss": 0.5154,
+      "step": 33227
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7203799911064124,
+      "learning_rate": 2.3733397343588994e-06,
+      "loss": 0.659,
+      "step": 33228
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2736231376139062,
+      "learning_rate": 2.3728462273303954e-06,
+      "loss": 0.6835,
+      "step": 33229
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5197980525058357,
+      "learning_rate": 2.3723527647095956e-06,
+      "loss": 0.7057,
+      "step": 33230
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.455270468096922,
+      "learning_rate": 2.3718593464993757e-06,
+      "loss": 0.6861,
+      "step": 33231
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4039165478537246,
+      "learning_rate": 2.3713659727026016e-06,
+      "loss": 0.7412,
+      "step": 33232
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.490148126165046,
+      "learning_rate": 2.370872643322154e-06,
+      "loss": 0.6657,
+      "step": 33233
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4541488821137378,
+      "learning_rate": 2.3703793583609013e-06,
+      "loss": 0.6705,
+      "step": 33234
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3701028639574204,
+      "learning_rate": 2.3698861178217125e-06,
+      "loss": 0.8049,
+      "step": 33235
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3262946411259138,
+      "learning_rate": 2.369392921707463e-06,
+      "loss": 0.7153,
+      "step": 33236
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5171563814478317,
+      "learning_rate": 2.368899770021026e-06,
+      "loss": 0.6626,
+      "step": 33237
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4637084943849512,
+      "learning_rate": 2.36840666276527e-06,
+      "loss": 0.7738,
+      "step": 33238
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5656678945441025,
+      "learning_rate": 2.3679135999430646e-06,
+      "loss": 0.4871,
+      "step": 33239
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.43988202288251,
+      "learning_rate": 2.3674205815572816e-06,
+      "loss": 0.6016,
+      "step": 33240
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5621428715720781,
+      "learning_rate": 2.3669276076107985e-06,
+      "loss": 0.4623,
+      "step": 33241
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2118374990319989,
+      "learning_rate": 2.366434678106474e-06,
+      "loss": 0.6897,
+      "step": 33242
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1424496219977978,
+      "learning_rate": 2.365941793047184e-06,
+      "loss": 0.5369,
+      "step": 33243
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2797514708249969,
+      "learning_rate": 2.365448952435797e-06,
+      "loss": 0.6235,
+      "step": 33244
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4628590250851825,
+      "learning_rate": 2.3649561562751853e-06,
+      "loss": 0.6759,
+      "step": 33245
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.352532750116691,
+      "learning_rate": 2.364463404568218e-06,
+      "loss": 0.4959,
+      "step": 33246
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3740386013547838,
+      "learning_rate": 2.3639706973177577e-06,
+      "loss": 0.6736,
+      "step": 33247
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7228981634605767,
+      "learning_rate": 2.3634780345266805e-06,
+      "loss": 0.6889,
+      "step": 33248
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.184720051981299,
+      "learning_rate": 2.3629854161978505e-06,
+      "loss": 0.6035,
+      "step": 33249
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2940881428633515,
+      "learning_rate": 2.3624928423341387e-06,
+      "loss": 0.7381,
+      "step": 33250
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.272275906929871,
+      "learning_rate": 2.3620003129384095e-06,
+      "loss": 0.6195,
+      "step": 33251
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4800187946597902,
+      "learning_rate": 2.361507828013535e-06,
+      "loss": 0.6553,
+      "step": 33252
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5222050294100054,
+      "learning_rate": 2.3610153875623777e-06,
+      "loss": 0.7029,
+      "step": 33253
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6099636571062281,
+      "learning_rate": 2.3605229915878093e-06,
+      "loss": 0.6639,
+      "step": 33254
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5434397935508135,
+      "learning_rate": 2.3600306400926955e-06,
+      "loss": 0.6464,
+      "step": 33255
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.384615148303206,
+      "learning_rate": 2.359538333079898e-06,
+      "loss": 0.6847,
+      "step": 33256
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.746479460600456,
+      "learning_rate": 2.359046070552289e-06,
+      "loss": 0.6062,
+      "step": 33257
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3823053496245068,
+      "learning_rate": 2.358553852512735e-06,
+      "loss": 0.6079,
+      "step": 33258
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3106319916993243,
+      "learning_rate": 2.3580616789640997e-06,
+      "loss": 0.695,
+      "step": 33259
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.594855704061014,
+      "learning_rate": 2.3575695499092455e-06,
+      "loss": 0.4637,
+      "step": 33260
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.7589866495479196,
+      "learning_rate": 2.3570774653510407e-06,
+      "loss": 0.7723,
+      "step": 33261
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4957323123104176,
+      "learning_rate": 2.3565854252923547e-06,
+      "loss": 0.651,
+      "step": 33262
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.872085589341614,
+      "learning_rate": 2.356093429736047e-06,
+      "loss": 0.6771,
+      "step": 33263
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3213103988742225,
+      "learning_rate": 2.355601478684981e-06,
+      "loss": 0.7251,
+      "step": 33264
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5466441691807447,
+      "learning_rate": 2.355109572142024e-06,
+      "loss": 0.6415,
+      "step": 33265
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4081549677570597,
+      "learning_rate": 2.3546177101100422e-06,
+      "loss": 0.6797,
+      "step": 33266
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3236562216293688,
+      "learning_rate": 2.354125892591895e-06,
+      "loss": 0.7029,
+      "step": 33267
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3756031762796304,
+      "learning_rate": 2.3536341195904455e-06,
+      "loss": 0.6889,
+      "step": 33268
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2791520077639498,
+      "learning_rate": 2.3531423911085626e-06,
+      "loss": 0.7722,
+      "step": 33269
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.484911600602789,
+      "learning_rate": 2.3526507071491023e-06,
+      "loss": 0.7847,
+      "step": 33270
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5827766455167881,
+      "learning_rate": 2.352159067714933e-06,
+      "loss": 0.4785,
+      "step": 33271
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3527959528114653,
+      "learning_rate": 2.3516674728089127e-06,
+      "loss": 0.6556,
+      "step": 33272
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4201236077574586,
+      "learning_rate": 2.351175922433908e-06,
+      "loss": 0.7153,
+      "step": 33273
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3756279874636403,
+      "learning_rate": 2.3506844165927765e-06,
+      "loss": 0.6595,
+      "step": 33274
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.284597270677935,
+      "learning_rate": 2.350192955288385e-06,
+      "loss": 0.5833,
+      "step": 33275
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2778902431067307,
+      "learning_rate": 2.349701538523591e-06,
+      "loss": 0.578,
+      "step": 33276
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.396982904357114,
+      "learning_rate": 2.3492101663012546e-06,
+      "loss": 0.6028,
+      "step": 33277
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.341145533520972,
+      "learning_rate": 2.3487188386242387e-06,
+      "loss": 0.628,
+      "step": 33278
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2278425512407969,
+      "learning_rate": 2.348227555495407e-06,
+      "loss": 0.6304,
+      "step": 33279
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3704812838752214,
+      "learning_rate": 2.347736316917617e-06,
+      "loss": 0.7041,
+      "step": 33280
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5667935668984149,
+      "learning_rate": 2.3472451228937254e-06,
+      "loss": 0.4873,
+      "step": 33281
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2955782187073708,
+      "learning_rate": 2.3467539734265966e-06,
+      "loss": 0.7132,
+      "step": 33282
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3915474812765432,
+      "learning_rate": 2.3462628685190913e-06,
+      "loss": 0.6772,
+      "step": 33283
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2270935248036472,
+      "learning_rate": 2.3457718081740664e-06,
+      "loss": 0.6416,
+      "step": 33284
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2444683651429234,
+      "learning_rate": 2.345280792394379e-06,
+      "loss": 0.6305,
+      "step": 33285
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4905688784870077,
+      "learning_rate": 2.3447898211828903e-06,
+      "loss": 0.8085,
+      "step": 33286
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2894201831622303,
+      "learning_rate": 2.3442988945424617e-06,
+      "loss": 0.7743,
+      "step": 33287
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3901060136444567,
+      "learning_rate": 2.3438080124759476e-06,
+      "loss": 0.6532,
+      "step": 33288
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4907543130867273,
+      "learning_rate": 2.343317174986206e-06,
+      "loss": 0.6548,
+      "step": 33289
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6263310276809069,
+      "learning_rate": 2.342826382076098e-06,
+      "loss": 0.4752,
+      "step": 33290
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6031919428837904,
+      "learning_rate": 2.342335633748476e-06,
+      "loss": 0.533,
+      "step": 33291
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3538699335292719,
+      "learning_rate": 2.341844930006203e-06,
+      "loss": 0.7073,
+      "step": 33292
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3162858631210295,
+      "learning_rate": 2.341354270852131e-06,
+      "loss": 0.7079,
+      "step": 33293
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4577815304092552,
+      "learning_rate": 2.3408636562891218e-06,
+      "loss": 0.8872,
+      "step": 33294
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7468171657024194,
+      "learning_rate": 2.3403730863200256e-06,
+      "loss": 0.6189,
+      "step": 33295
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4068550164479443,
+      "learning_rate": 2.339882560947706e-06,
+      "loss": 0.6649,
+      "step": 33296
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.201360843791679,
+      "learning_rate": 2.3393920801750157e-06,
+      "loss": 0.7923,
+      "step": 33297
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.338617208748745,
+      "learning_rate": 2.3389016440048063e-06,
+      "loss": 0.7448,
+      "step": 33298
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.886860292795348,
+      "learning_rate": 2.338411252439937e-06,
+      "loss": 0.5744,
+      "step": 33299
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7934714666238583,
+      "learning_rate": 2.337920905483266e-06,
+      "loss": 0.6581,
+      "step": 33300
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4171797209305412,
+      "learning_rate": 2.337430603137646e-06,
+      "loss": 0.6697,
+      "step": 33301
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4575150994032984,
+      "learning_rate": 2.336940345405928e-06,
+      "loss": 0.7725,
+      "step": 33302
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5509425788145422,
+      "learning_rate": 2.3364501322909685e-06,
+      "loss": 0.4701,
+      "step": 33303
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4652187302179294,
+      "learning_rate": 2.3359599637956264e-06,
+      "loss": 0.6483,
+      "step": 33304
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.337918652191754,
+      "learning_rate": 2.3354698399227504e-06,
+      "loss": 0.7706,
+      "step": 33305
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4284843456541263,
+      "learning_rate": 2.3349797606751945e-06,
+      "loss": 0.776,
+      "step": 33306
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5917096724633839,
+      "learning_rate": 2.334489726055812e-06,
+      "loss": 0.4799,
+      "step": 33307
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2856711082178074,
+      "learning_rate": 2.3339997360674606e-06,
+      "loss": 0.8295,
+      "step": 33308
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3795370492060282,
+      "learning_rate": 2.333509790712989e-06,
+      "loss": 0.6889,
+      "step": 33309
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3901123434554852,
+      "learning_rate": 2.3330198899952484e-06,
+      "loss": 0.7576,
+      "step": 33310
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2021765010038257,
+      "learning_rate": 2.3325300339170954e-06,
+      "loss": 0.7127,
+      "step": 33311
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3609292282876764,
+      "learning_rate": 2.3320402224813776e-06,
+      "loss": 0.6212,
+      "step": 33312
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5552570692454237,
+      "learning_rate": 2.3315504556909517e-06,
+      "loss": 0.7394,
+      "step": 33313
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6349345935307609,
+      "learning_rate": 2.3310607335486645e-06,
+      "loss": 0.6169,
+      "step": 33314
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5676189647269623,
+      "learning_rate": 2.3305710560573723e-06,
+      "loss": 0.7088,
+      "step": 33315
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4142618657518249,
+      "learning_rate": 2.330081423219921e-06,
+      "loss": 0.7171,
+      "step": 33316
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4469604573447132,
+      "learning_rate": 2.3295918350391632e-06,
+      "loss": 0.6651,
+      "step": 33317
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5994922553361892,
+      "learning_rate": 2.3291022915179575e-06,
+      "loss": 0.6226,
+      "step": 33318
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5484762963028158,
+      "learning_rate": 2.3286127926591397e-06,
+      "loss": 0.6903,
+      "step": 33319
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2868324507356084,
+      "learning_rate": 2.3281233384655665e-06,
+      "loss": 0.6789,
+      "step": 33320
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6257396080422678,
+      "learning_rate": 2.3276339289400895e-06,
+      "loss": 0.5209,
+      "step": 33321
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0516904090009134,
+      "learning_rate": 2.3271445640855618e-06,
+      "loss": 0.5601,
+      "step": 33322
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.327217951186363,
+      "learning_rate": 2.326655243904822e-06,
+      "loss": 0.6457,
+      "step": 33323
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4836722112242204,
+      "learning_rate": 2.326165968400724e-06,
+      "loss": 0.6345,
+      "step": 33324
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6204054920772777,
+      "learning_rate": 2.325676737576121e-06,
+      "loss": 0.7775,
+      "step": 33325
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9230478973734526,
+      "learning_rate": 2.3251875514338575e-06,
+      "loss": 0.6388,
+      "step": 33326
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3525653079671998,
+      "learning_rate": 2.324698409976779e-06,
+      "loss": 0.6808,
+      "step": 33327
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4806901293378674,
+      "learning_rate": 2.324209313207736e-06,
+      "loss": 0.7171,
+      "step": 33328
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.548897176080617,
+      "learning_rate": 2.3237202611295795e-06,
+      "loss": 0.7351,
+      "step": 33329
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.619722581880879,
+      "learning_rate": 2.323231253745151e-06,
+      "loss": 0.631,
+      "step": 33330
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5139601830606086,
+      "learning_rate": 2.322742291057304e-06,
+      "loss": 0.6144,
+      "step": 33331
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6263103418860475,
+      "learning_rate": 2.322253373068879e-06,
+      "loss": 0.5347,
+      "step": 33332
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.290403508572702,
+      "learning_rate": 2.321764499782727e-06,
+      "loss": 0.7732,
+      "step": 33333
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5711856394439117,
+      "learning_rate": 2.3212756712016913e-06,
+      "loss": 0.6969,
+      "step": 33334
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3936993732820178,
+      "learning_rate": 2.3207868873286232e-06,
+      "loss": 0.8309,
+      "step": 33335
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.625791324079976,
+      "learning_rate": 2.3202981481663646e-06,
+      "loss": 0.7282,
+      "step": 33336
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3021864907583367,
+      "learning_rate": 2.319809453717757e-06,
+      "loss": 0.7164,
+      "step": 33337
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5260833558824052,
+      "learning_rate": 2.319320803985653e-06,
+      "loss": 0.7018,
+      "step": 33338
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.267288556806857,
+      "learning_rate": 2.3188321989728955e-06,
+      "loss": 0.6321,
+      "step": 33339
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.591269079056046,
+      "learning_rate": 2.31834363868233e-06,
+      "loss": 0.5427,
+      "step": 33340
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2395654752795502,
+      "learning_rate": 2.3178551231167968e-06,
+      "loss": 0.6589,
+      "step": 33341
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5852792400546873,
+      "learning_rate": 2.317366652279143e-06,
+      "loss": 0.7335,
+      "step": 33342
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3184371615350021,
+      "learning_rate": 2.3168782261722156e-06,
+      "loss": 0.7574,
+      "step": 33343
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.359227469545845,
+      "learning_rate": 2.3163898447988562e-06,
+      "loss": 0.6625,
+      "step": 33344
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3772136196751426,
+      "learning_rate": 2.315901508161904e-06,
+      "loss": 0.6933,
+      "step": 33345
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.480464357168514,
+      "learning_rate": 2.315413216264206e-06,
+      "loss": 0.604,
+      "step": 33346
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3307885928139478,
+      "learning_rate": 2.3149249691086094e-06,
+      "loss": 0.7562,
+      "step": 33347
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4276629732509118,
+      "learning_rate": 2.3144367666979517e-06,
+      "loss": 0.6805,
+      "step": 33348
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.360440772666464,
+      "learning_rate": 2.3139486090350727e-06,
+      "loss": 0.6507,
+      "step": 33349
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.149463106859622,
+      "learning_rate": 2.3134604961228225e-06,
+      "loss": 0.5191,
+      "step": 33350
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3593331506274917,
+      "learning_rate": 2.3129724279640355e-06,
+      "loss": 0.6056,
+      "step": 33351
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5191898158862618,
+      "learning_rate": 2.312484404561559e-06,
+      "loss": 0.7571,
+      "step": 33352
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3303351131477956,
+      "learning_rate": 2.311996425918229e-06,
+      "loss": 0.6427,
+      "step": 33353
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6752117183698534,
+      "learning_rate": 2.3115084920368935e-06,
+      "loss": 0.7681,
+      "step": 33354
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.507709119833559,
+      "learning_rate": 2.3110206029203872e-06,
+      "loss": 0.7086,
+      "step": 33355
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7431295948467667,
+      "learning_rate": 2.310532758571555e-06,
+      "loss": 0.7597,
+      "step": 33356
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2745003208043764,
+      "learning_rate": 2.310044958993236e-06,
+      "loss": 0.6395,
+      "step": 33357
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4667854312603972,
+      "learning_rate": 2.3095572041882675e-06,
+      "loss": 0.7372,
+      "step": 33358
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4963361173249858,
+      "learning_rate": 2.3090694941594914e-06,
+      "loss": 0.7182,
+      "step": 33359
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6098790084652675,
+      "learning_rate": 2.3085818289097506e-06,
+      "loss": 0.5387,
+      "step": 33360
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3570113301466538,
+      "learning_rate": 2.308094208441881e-06,
+      "loss": 0.6879,
+      "step": 33361
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4013879450272524,
+      "learning_rate": 2.307606632758719e-06,
+      "loss": 0.7503,
+      "step": 33362
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5243775328631994,
+      "learning_rate": 2.307119101863108e-06,
+      "loss": 0.6584,
+      "step": 33363
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4721345798361865,
+      "learning_rate": 2.306631615757886e-06,
+      "loss": 0.6184,
+      "step": 33364
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5307937488387593,
+      "learning_rate": 2.306144174445891e-06,
+      "loss": 0.8205,
+      "step": 33365
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5800750094698958,
+      "learning_rate": 2.3056567779299588e-06,
+      "loss": 0.4838,
+      "step": 33366
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4325511563733127,
+      "learning_rate": 2.305169426212928e-06,
+      "loss": 0.6862,
+      "step": 33367
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4481239999101505,
+      "learning_rate": 2.30468211929764e-06,
+      "loss": 0.7998,
+      "step": 33368
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4120323702533875,
+      "learning_rate": 2.304194857186929e-06,
+      "loss": 0.6245,
+      "step": 33369
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3583973254508452,
+      "learning_rate": 2.303707639883629e-06,
+      "loss": 0.723,
+      "step": 33370
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.307739631619752,
+      "learning_rate": 2.303220467390582e-06,
+      "loss": 0.6757,
+      "step": 33371
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2928390878084914,
+      "learning_rate": 2.30273333971062e-06,
+      "loss": 0.6888,
+      "step": 33372
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5909426150132385,
+      "learning_rate": 2.3022462568465843e-06,
+      "loss": 0.5154,
+      "step": 33373
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3676975730963057,
+      "learning_rate": 2.301759218801305e-06,
+      "loss": 0.7786,
+      "step": 33374
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.253628295987918,
+      "learning_rate": 2.3012722255776234e-06,
+      "loss": 0.7378,
+      "step": 33375
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.576696618644935,
+      "learning_rate": 2.300785277178369e-06,
+      "loss": 0.7126,
+      "step": 33376
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.334999441022027,
+      "learning_rate": 2.3002983736063843e-06,
+      "loss": 0.5989,
+      "step": 33377
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.367265759350603,
+      "learning_rate": 2.2998115148644997e-06,
+      "loss": 0.6347,
+      "step": 33378
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7067631833009576,
+      "learning_rate": 2.2993247009555463e-06,
+      "loss": 0.6834,
+      "step": 33379
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3831791101849342,
+      "learning_rate": 2.298837931882364e-06,
+      "loss": 0.6976,
+      "step": 33380
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3892152831307372,
+      "learning_rate": 2.2983512076477875e-06,
+      "loss": 0.7825,
+      "step": 33381
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.341343899566156,
+      "learning_rate": 2.2978645282546487e-06,
+      "loss": 0.66,
+      "step": 33382
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2697646128080269,
+      "learning_rate": 2.2973778937057787e-06,
+      "loss": 0.773,
+      "step": 33383
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3464713575502463,
+      "learning_rate": 2.296891304004013e-06,
+      "loss": 0.6663,
+      "step": 33384
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.301145067897444,
+      "learning_rate": 2.296404759152188e-06,
+      "loss": 0.6949,
+      "step": 33385
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6435874461147346,
+      "learning_rate": 2.2959182591531327e-06,
+      "loss": 0.8183,
+      "step": 33386
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4374080419727122,
+      "learning_rate": 2.295431804009678e-06,
+      "loss": 0.7657,
+      "step": 33387
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.556187722123878,
+      "learning_rate": 2.294945393724658e-06,
+      "loss": 0.7128,
+      "step": 33388
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7097512387694942,
+      "learning_rate": 2.2944590283009083e-06,
+      "loss": 0.7256,
+      "step": 33389
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4463593560185808,
+      "learning_rate": 2.2939727077412578e-06,
+      "loss": 0.6576,
+      "step": 33390
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.200865684569238,
+      "learning_rate": 2.293486432048535e-06,
+      "loss": 0.6766,
+      "step": 33391
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4157763512056245,
+      "learning_rate": 2.293000201225577e-06,
+      "loss": 0.638,
+      "step": 33392
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.338552517709629,
+      "learning_rate": 2.292514015275209e-06,
+      "loss": 0.6351,
+      "step": 33393
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6673692252485348,
+      "learning_rate": 2.2920278742002677e-06,
+      "loss": 0.6367,
+      "step": 33394
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3746813412501244,
+      "learning_rate": 2.2915417780035763e-06,
+      "loss": 0.6925,
+      "step": 33395
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.299826057794982,
+      "learning_rate": 2.2910557266879742e-06,
+      "loss": 0.5574,
+      "step": 33396
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4980963380148682,
+      "learning_rate": 2.2905697202562817e-06,
+      "loss": 0.7248,
+      "step": 33397
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4683437432232986,
+      "learning_rate": 2.2900837587113366e-06,
+      "loss": 0.637,
+      "step": 33398
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1622833560109114,
+      "learning_rate": 2.289597842055964e-06,
+      "loss": 0.6532,
+      "step": 33399
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4896552849665887,
+      "learning_rate": 2.289111970292992e-06,
+      "loss": 0.6139,
+      "step": 33400
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3385988812619518,
+      "learning_rate": 2.28862614342525e-06,
+      "loss": 0.6126,
+      "step": 33401
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3314553368640158,
+      "learning_rate": 2.288140361455572e-06,
+      "loss": 0.6631,
+      "step": 33402
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3367567343913347,
+      "learning_rate": 2.2876546243867826e-06,
+      "loss": 0.6101,
+      "step": 33403
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8126427141787649,
+      "learning_rate": 2.287168932221706e-06,
+      "loss": 0.692,
+      "step": 33404
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2954809117486088,
+      "learning_rate": 2.286683284963174e-06,
+      "loss": 0.6484,
+      "step": 33405
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.788051691609868,
+      "learning_rate": 2.2861976826140164e-06,
+      "loss": 0.6155,
+      "step": 33406
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3543258346768055,
+      "learning_rate": 2.2857121251770585e-06,
+      "loss": 0.7999,
+      "step": 33407
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4302018119019633,
+      "learning_rate": 2.285226612655124e-06,
+      "loss": 0.8112,
+      "step": 33408
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4359258532128796,
+      "learning_rate": 2.2847411450510426e-06,
+      "loss": 0.7369,
+      "step": 33409
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2356803119716147,
+      "learning_rate": 2.284255722367643e-06,
+      "loss": 0.668,
+      "step": 33410
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5081811990180924,
+      "learning_rate": 2.2837703446077496e-06,
+      "loss": 0.6439,
+      "step": 33411
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.33686302556032,
+      "learning_rate": 2.2832850117741866e-06,
+      "loss": 0.6387,
+      "step": 33412
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5752188024396074,
+      "learning_rate": 2.282799723869783e-06,
+      "loss": 0.6516,
+      "step": 33413
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3884021570809755,
+      "learning_rate": 2.2823144808973596e-06,
+      "loss": 0.7291,
+      "step": 33414
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2484604798049361,
+      "learning_rate": 2.2818292828597457e-06,
+      "loss": 0.5565,
+      "step": 33415
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5480500211730053,
+      "learning_rate": 2.2813441297597673e-06,
+      "loss": 0.6549,
+      "step": 33416
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.633389350256683,
+      "learning_rate": 2.280859021600247e-06,
+      "loss": 0.7498,
+      "step": 33417
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.146183300585046,
+      "learning_rate": 2.2803739583840066e-06,
+      "loss": 0.5453,
+      "step": 33418
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5722386946874456,
+      "learning_rate": 2.2798889401138735e-06,
+      "loss": 0.5463,
+      "step": 33419
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4486563047417245,
+      "learning_rate": 2.279403966792676e-06,
+      "loss": 0.7347,
+      "step": 33420
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2802996918151732,
+      "learning_rate": 2.2789190384232286e-06,
+      "loss": 0.5636,
+      "step": 33421
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3639648198035872,
+      "learning_rate": 2.2784341550083577e-06,
+      "loss": 0.7587,
+      "step": 33422
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3513407121218566,
+      "learning_rate": 2.277949316550888e-06,
+      "loss": 0.8531,
+      "step": 33423
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.282634717051328,
+      "learning_rate": 2.2774645230536485e-06,
+      "loss": 0.5368,
+      "step": 33424
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5342297369997726,
+      "learning_rate": 2.276979774519449e-06,
+      "loss": 0.8024,
+      "step": 33425
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5972231532335222,
+      "learning_rate": 2.276495070951118e-06,
+      "loss": 0.7036,
+      "step": 33426
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.423399440961679,
+      "learning_rate": 2.2760104123514814e-06,
+      "loss": 0.6549,
+      "step": 33427
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5703472327399246,
+      "learning_rate": 2.275525798723355e-06,
+      "loss": 0.7002,
+      "step": 33428
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3202684456351932,
+      "learning_rate": 2.2750412300695656e-06,
+      "loss": 0.5831,
+      "step": 33429
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7507179276959304,
+      "learning_rate": 2.2745567063929285e-06,
+      "loss": 0.8078,
+      "step": 33430
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8835478184035572,
+      "learning_rate": 2.2740722276962714e-06,
+      "loss": 0.8066,
+      "step": 33431
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4046794608701478,
+      "learning_rate": 2.273587793982409e-06,
+      "loss": 0.7066,
+      "step": 33432
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4492882438853811,
+      "learning_rate": 2.273103405254167e-06,
+      "loss": 0.6658,
+      "step": 33433
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2406599590358691,
+      "learning_rate": 2.2726190615143638e-06,
+      "loss": 0.5432,
+      "step": 33434
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2772861813189533,
+      "learning_rate": 2.272134762765815e-06,
+      "loss": 0.6182,
+      "step": 33435
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.47151858605527,
+      "learning_rate": 2.271650509011346e-06,
+      "loss": 0.7206,
+      "step": 33436
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8628197867880347,
+      "learning_rate": 2.2711663002537754e-06,
+      "loss": 0.8113,
+      "step": 33437
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.518242007685187,
+      "learning_rate": 2.270682136495923e-06,
+      "loss": 0.705,
+      "step": 33438
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3952936459101792,
+      "learning_rate": 2.2701980177406026e-06,
+      "loss": 0.6326,
+      "step": 33439
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4715066097288416,
+      "learning_rate": 2.2697139439906367e-06,
+      "loss": 0.689,
+      "step": 33440
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.53087082480488,
+      "learning_rate": 2.269229915248846e-06,
+      "loss": 0.6677,
+      "step": 33441
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3683711538969487,
+      "learning_rate": 2.2687459315180473e-06,
+      "loss": 0.7368,
+      "step": 33442
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4653512715963666,
+      "learning_rate": 2.268261992801054e-06,
+      "loss": 0.7086,
+      "step": 33443
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5630794972702028,
+      "learning_rate": 2.267778099100687e-06,
+      "loss": 0.4661,
+      "step": 33444
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.541350990236992,
+      "learning_rate": 2.267294250419767e-06,
+      "loss": 0.7471,
+      "step": 33445
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.487513648256257,
+      "learning_rate": 2.2668104467611087e-06,
+      "loss": 0.7177,
+      "step": 33446
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3319852909671273,
+      "learning_rate": 2.2663266881275247e-06,
+      "loss": 0.6639,
+      "step": 33447
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4232075992630044,
+      "learning_rate": 2.265842974521838e-06,
+      "loss": 0.7648,
+      "step": 33448
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3524501397754842,
+      "learning_rate": 2.265359305946859e-06,
+      "loss": 0.6228,
+      "step": 33449
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2766468623945022,
+      "learning_rate": 2.2648756824054097e-06,
+      "loss": 0.6207,
+      "step": 33450
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6672427292884688,
+      "learning_rate": 2.2643921039003013e-06,
+      "loss": 0.6735,
+      "step": 33451
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.257109713427319,
+      "learning_rate": 2.263908570434352e-06,
+      "loss": 0.6014,
+      "step": 33452
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2937375698802258,
+      "learning_rate": 2.2634250820103752e-06,
+      "loss": 0.5935,
+      "step": 33453
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3195160706128097,
+      "learning_rate": 2.26294163863119e-06,
+      "loss": 0.7274,
+      "step": 33454
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.2065094395234066,
+      "learning_rate": 2.2624582402996066e-06,
+      "loss": 0.6356,
+      "step": 33455
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2686575218088698,
+      "learning_rate": 2.261974887018439e-06,
+      "loss": 0.6737,
+      "step": 33456
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4132487027264085,
+      "learning_rate": 2.2614915787905024e-06,
+      "loss": 0.669,
+      "step": 33457
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5435370384087248,
+      "learning_rate": 2.261008315618616e-06,
+      "loss": 0.6495,
+      "step": 33458
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.487659841766597,
+      "learning_rate": 2.260525097505588e-06,
+      "loss": 0.5525,
+      "step": 33459
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3679610594450504,
+      "learning_rate": 2.2600419244542316e-06,
+      "loss": 0.6436,
+      "step": 33460
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2082527675127788,
+      "learning_rate": 2.2595587964673614e-06,
+      "loss": 0.7587,
+      "step": 33461
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2852577649522638,
+      "learning_rate": 2.2590757135477927e-06,
+      "loss": 0.642,
+      "step": 33462
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3384240078992478,
+      "learning_rate": 2.2585926756983355e-06,
+      "loss": 0.638,
+      "step": 33463
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8119907213302278,
+      "learning_rate": 2.2581096829218008e-06,
+      "loss": 0.6926,
+      "step": 33464
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6559960395795164,
+      "learning_rate": 2.2576267352210022e-06,
+      "loss": 0.7044,
+      "step": 33465
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5910262831313025,
+      "learning_rate": 2.257143832598755e-06,
+      "loss": 0.4813,
+      "step": 33466
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4740289619517164,
+      "learning_rate": 2.256660975057867e-06,
+      "loss": 0.6665,
+      "step": 33467
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2728449130929083,
+      "learning_rate": 2.256178162601148e-06,
+      "loss": 0.6348,
+      "step": 33468
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6365592886136124,
+      "learning_rate": 2.255695395231412e-06,
+      "loss": 0.6095,
+      "step": 33469
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6935363252850104,
+      "learning_rate": 2.255212672951471e-06,
+      "loss": 0.6619,
+      "step": 33470
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9277224849388113,
+      "learning_rate": 2.2547299957641345e-06,
+      "loss": 0.6689,
+      "step": 33471
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3279266087955484,
+      "learning_rate": 2.2542473636722095e-06,
+      "loss": 0.5956,
+      "step": 33472
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4252023288734665,
+      "learning_rate": 2.2537647766785107e-06,
+      "loss": 0.6647,
+      "step": 33473
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3559654429052568,
+      "learning_rate": 2.253282234785843e-06,
+      "loss": 0.7041,
+      "step": 33474
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.969239400769156,
+      "learning_rate": 2.2527997379970214e-06,
+      "loss": 0.6018,
+      "step": 33475
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7827896701135393,
+      "learning_rate": 2.25231728631485e-06,
+      "loss": 0.772,
+      "step": 33476
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3232173475480085,
+      "learning_rate": 2.251834879742143e-06,
+      "loss": 0.5404,
+      "step": 33477
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9262107455855932,
+      "learning_rate": 2.251352518281704e-06,
+      "loss": 0.6164,
+      "step": 33478
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5794852217521187,
+      "learning_rate": 2.250870201936346e-06,
+      "loss": 0.6472,
+      "step": 33479
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3486248322895331,
+      "learning_rate": 2.250387930708875e-06,
+      "loss": 0.7227,
+      "step": 33480
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6261579851854288,
+      "learning_rate": 2.2499057046020966e-06,
+      "loss": 0.5018,
+      "step": 33481
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.316435937951733,
+      "learning_rate": 2.24942352361882e-06,
+      "loss": 0.6071,
+      "step": 33482
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3707512586076729,
+      "learning_rate": 2.248941387761857e-06,
+      "loss": 0.6998,
+      "step": 33483
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.813424493215513,
+      "learning_rate": 2.2484592970340115e-06,
+      "loss": 0.6523,
+      "step": 33484
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3548113934175006,
+      "learning_rate": 2.247977251438086e-06,
+      "loss": 0.6456,
+      "step": 33485
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4650226983399688,
+      "learning_rate": 2.247495250976892e-06,
+      "loss": 0.6214,
+      "step": 33486
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2859255174854625,
+      "learning_rate": 2.2470132956532376e-06,
+      "loss": 0.6312,
+      "step": 33487
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.793229820698695,
+      "learning_rate": 2.2465313854699265e-06,
+      "loss": 0.659,
+      "step": 33488
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7727782572535378,
+      "learning_rate": 2.2460495204297615e-06,
+      "loss": 0.6657,
+      "step": 33489
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6027766058078587,
+      "learning_rate": 2.2455677005355513e-06,
+      "loss": 0.6334,
+      "step": 33490
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3503974053117582,
+      "learning_rate": 2.245085925790104e-06,
+      "loss": 0.7022,
+      "step": 33491
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3948705237699428,
+      "learning_rate": 2.2446041961962207e-06,
+      "loss": 0.6151,
+      "step": 33492
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5314970193755588,
+      "learning_rate": 2.2441225117567056e-06,
+      "loss": 0.6538,
+      "step": 33493
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4177484880961126,
+      "learning_rate": 2.2436408724743664e-06,
+      "loss": 0.7632,
+      "step": 33494
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6315217931463064,
+      "learning_rate": 2.243159278352003e-06,
+      "loss": 0.5045,
+      "step": 33495
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2813450769771235,
+      "learning_rate": 2.242677729392425e-06,
+      "loss": 0.5997,
+      "step": 33496
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1756901461299012,
+      "learning_rate": 2.242196225598431e-06,
+      "loss": 0.7101,
+      "step": 33497
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.316808870909684,
+      "learning_rate": 2.241714766972828e-06,
+      "loss": 0.5769,
+      "step": 33498
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.585478782653522,
+      "learning_rate": 2.241233353518416e-06,
+      "loss": 0.6438,
+      "step": 33499
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1903750644207576,
+      "learning_rate": 2.240751985238e-06,
+      "loss": 0.6557,
+      "step": 33500
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.3229056714590257,
+      "learning_rate": 2.2402706621343873e-06,
+      "loss": 0.7224,
+      "step": 33501
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8024310992356418,
+      "learning_rate": 2.2397893842103714e-06,
+      "loss": 0.5553,
+      "step": 33502
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3525649393400168,
+      "learning_rate": 2.2393081514687564e-06,
+      "loss": 0.6904,
+      "step": 33503
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.375248973216778,
+      "learning_rate": 2.2388269639123482e-06,
+      "loss": 0.6785,
+      "step": 33504
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3340837843710625,
+      "learning_rate": 2.2383458215439513e-06,
+      "loss": 0.8255,
+      "step": 33505
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.455528262936497,
+      "learning_rate": 2.2378647243663563e-06,
+      "loss": 0.6948,
+      "step": 33506
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.298698530765625,
+      "learning_rate": 2.2373836723823714e-06,
+      "loss": 0.6776,
+      "step": 33507
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3062731957461415,
+      "learning_rate": 2.2369026655947988e-06,
+      "loss": 0.6263,
+      "step": 33508
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4071746144076822,
+      "learning_rate": 2.236421704006436e-06,
+      "loss": 0.6712,
+      "step": 33509
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4061194696315946,
+      "learning_rate": 2.2359407876200822e-06,
+      "loss": 0.6635,
+      "step": 33510
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2993017140363834,
+      "learning_rate": 2.235459916438538e-06,
+      "loss": 0.6071,
+      "step": 33511
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4628726485385173,
+      "learning_rate": 2.2349790904646085e-06,
+      "loss": 0.6418,
+      "step": 33512
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3342613757572668,
+      "learning_rate": 2.2344983097010854e-06,
+      "loss": 0.7472,
+      "step": 33513
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7230375604106412,
+      "learning_rate": 2.234017574150774e-06,
+      "loss": 0.6997,
+      "step": 33514
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2146679193909873,
+      "learning_rate": 2.2335368838164715e-06,
+      "loss": 0.6912,
+      "step": 33515
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6559552365688857,
+      "learning_rate": 2.2330562387009745e-06,
+      "loss": 0.803,
+      "step": 33516
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3356483158942247,
+      "learning_rate": 2.2325756388070817e-06,
+      "loss": 0.5895,
+      "step": 33517
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6348262006713786,
+      "learning_rate": 2.2320950841375956e-06,
+      "loss": 0.5008,
+      "step": 33518
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5312653489934047,
+      "learning_rate": 2.2316145746953112e-06,
+      "loss": 0.5851,
+      "step": 33519
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8632537246107137,
+      "learning_rate": 2.231134110483023e-06,
+      "loss": 0.8343,
+      "step": 33520
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4065383737044306,
+      "learning_rate": 2.2306536915035326e-06,
+      "loss": 0.8057,
+      "step": 33521
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3968486279785,
+      "learning_rate": 2.2301733177596407e-06,
+      "loss": 0.7066,
+      "step": 33522
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.546506171414735,
+      "learning_rate": 2.229692989254135e-06,
+      "loss": 0.6757,
+      "step": 33523
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8953039534145069,
+      "learning_rate": 2.229212705989816e-06,
+      "loss": 0.6347,
+      "step": 33524
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4282943418946132,
+      "learning_rate": 2.228732467969481e-06,
+      "loss": 0.7126,
+      "step": 33525
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4433797007464906,
+      "learning_rate": 2.228252275195929e-06,
+      "loss": 0.6999,
+      "step": 33526
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.417318145700942,
+      "learning_rate": 2.2277721276719532e-06,
+      "loss": 0.6495,
+      "step": 33527
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3269171991615876,
+      "learning_rate": 2.227292025400345e-06,
+      "loss": 0.6284,
+      "step": 33528
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.577832353385968,
+      "learning_rate": 2.226811968383906e-06,
+      "loss": 0.5016,
+      "step": 33529
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2646654869370042,
+      "learning_rate": 2.226331956625427e-06,
+      "loss": 0.6635,
+      "step": 33530
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1821379866479267,
+      "learning_rate": 2.2258519901277066e-06,
+      "loss": 0.6004,
+      "step": 33531
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3854359900865167,
+      "learning_rate": 2.2253720688935344e-06,
+      "loss": 0.7169,
+      "step": 33532
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.397682977943109,
+      "learning_rate": 2.2248921929257106e-06,
+      "loss": 0.6358,
+      "step": 33533
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3689905906981163,
+      "learning_rate": 2.2244123622270218e-06,
+      "loss": 0.7388,
+      "step": 33534
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.610247222246306,
+      "learning_rate": 2.223932576800271e-06,
+      "loss": 0.7054,
+      "step": 33535
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1680289470841827,
+      "learning_rate": 2.2234528366482445e-06,
+      "loss": 0.659,
+      "step": 33536
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3838557604355588,
+      "learning_rate": 2.2229731417737354e-06,
+      "loss": 0.6844,
+      "step": 33537
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2013471820373527,
+      "learning_rate": 2.2224934921795392e-06,
+      "loss": 0.689,
+      "step": 33538
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3697940356813008,
+      "learning_rate": 2.2220138878684516e-06,
+      "loss": 0.6868,
+      "step": 33539
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.22883458884834,
+      "learning_rate": 2.221534328843261e-06,
+      "loss": 0.6197,
+      "step": 33540
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6166477933238897,
+      "learning_rate": 2.221054815106757e-06,
+      "loss": 0.4813,
+      "step": 33541
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3144702125353709,
+      "learning_rate": 2.2205753466617353e-06,
+      "loss": 0.7182,
+      "step": 33542
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6547265211688547,
+      "learning_rate": 2.2200959235109886e-06,
+      "loss": 0.6517,
+      "step": 33543
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.302208333498052,
+      "learning_rate": 2.219616545657307e-06,
+      "loss": 0.6325,
+      "step": 33544
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3575162933208225,
+      "learning_rate": 2.2191372131034793e-06,
+      "loss": 0.6612,
+      "step": 33545
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9434652250467945,
+      "learning_rate": 2.2186579258522966e-06,
+      "loss": 0.6458,
+      "step": 33546
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4207262789939599,
+      "learning_rate": 2.2181786839065544e-06,
+      "loss": 0.6611,
+      "step": 33547
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5188504663601727,
+      "learning_rate": 2.2176994872690404e-06,
+      "loss": 0.7038,
+      "step": 33548
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.595078660480396,
+      "learning_rate": 2.2172203359425405e-06,
+      "loss": 0.6252,
+      "step": 33549
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3566636726752344,
+      "learning_rate": 2.2167412299298497e-06,
+      "loss": 0.5226,
+      "step": 33550
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5968598606535567,
+      "learning_rate": 2.2162621692337537e-06,
+      "loss": 0.7996,
+      "step": 33551
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8978755940312497,
+      "learning_rate": 2.2157831538570464e-06,
+      "loss": 0.6793,
+      "step": 33552
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.595594458941051,
+      "learning_rate": 2.2153041838025114e-06,
+      "loss": 0.6226,
+      "step": 33553
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5520619643720772,
+      "learning_rate": 2.2148252590729424e-06,
+      "loss": 0.6947,
+      "step": 33554
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.599096322923512,
+      "learning_rate": 2.2143463796711228e-06,
+      "loss": 0.5101,
+      "step": 33555
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5946642117454919,
+      "learning_rate": 2.213867545599846e-06,
+      "loss": 0.5242,
+      "step": 33556
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.360353901081735,
+      "learning_rate": 2.2133887568618972e-06,
+      "loss": 0.8155,
+      "step": 33557
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2470701921471543,
+      "learning_rate": 2.2129100134600624e-06,
+      "loss": 0.6689,
+      "step": 33558
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6334099955993988,
+      "learning_rate": 2.21243131539713e-06,
+      "loss": 0.676,
+      "step": 33559
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4464199367876283,
+      "learning_rate": 2.2119526626758914e-06,
+      "loss": 0.6631,
+      "step": 33560
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6362780994428432,
+      "learning_rate": 2.21147405529913e-06,
+      "loss": 0.5974,
+      "step": 33561
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4762822731465037,
+      "learning_rate": 2.2109954932696285e-06,
+      "loss": 0.6659,
+      "step": 33562
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6863268559041775,
+      "learning_rate": 2.210516976590179e-06,
+      "loss": 0.6176,
+      "step": 33563
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.469467242704862,
+      "learning_rate": 2.2100385052635677e-06,
+      "loss": 0.7078,
+      "step": 33564
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3573467641067003,
+      "learning_rate": 2.209560079292579e-06,
+      "loss": 0.6633,
+      "step": 33565
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4064970012570563,
+      "learning_rate": 2.2090816986799934e-06,
+      "loss": 0.587,
+      "step": 33566
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.175747958113008,
+      "learning_rate": 2.208603363428602e-06,
+      "loss": 0.5535,
+      "step": 33567
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6687971712409695,
+      "learning_rate": 2.2081250735411917e-06,
+      "loss": 0.7216,
+      "step": 33568
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.387641502257323,
+      "learning_rate": 2.2076468290205433e-06,
+      "loss": 0.6047,
+      "step": 33569
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.873410024220682,
+      "learning_rate": 2.2071686298694393e-06,
+      "loss": 0.6371,
+      "step": 33570
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6038244293213375,
+      "learning_rate": 2.2066904760906704e-06,
+      "loss": 0.524,
+      "step": 33571
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3560164437381095,
+      "learning_rate": 2.206212367687014e-06,
+      "loss": 0.6623,
+      "step": 33572
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6227638512476337,
+      "learning_rate": 2.205734304661259e-06,
+      "loss": 0.4634,
+      "step": 33573
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3106874579392092,
+      "learning_rate": 2.2052562870161832e-06,
+      "loss": 0.6111,
+      "step": 33574
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.954507591294822,
+      "learning_rate": 2.2047783147545763e-06,
+      "loss": 0.7165,
+      "step": 33575
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.625558881051511,
+      "learning_rate": 2.2043003878792156e-06,
+      "loss": 0.7367,
+      "step": 33576
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1905400446385306,
+      "learning_rate": 2.2038225063928887e-06,
+      "loss": 0.5705,
+      "step": 33577
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5457202809340247,
+      "learning_rate": 2.203344670298375e-06,
+      "loss": 0.7428,
+      "step": 33578
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4443776090288356,
+      "learning_rate": 2.202866879598454e-06,
+      "loss": 0.6695,
+      "step": 33579
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6262764284147802,
+      "learning_rate": 2.2023891342959114e-06,
+      "loss": 0.7368,
+      "step": 33580
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.414402760914303,
+      "learning_rate": 2.20191143439353e-06,
+      "loss": 0.6459,
+      "step": 33581
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5145552911761377,
+      "learning_rate": 2.20143377989409e-06,
+      "loss": 0.7909,
+      "step": 33582
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3965314497153434,
+      "learning_rate": 2.200956170800368e-06,
+      "loss": 0.5948,
+      "step": 33583
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.449870283165137,
+      "learning_rate": 2.2004786071151486e-06,
+      "loss": 0.6916,
+      "step": 33584
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5356469714375804,
+      "learning_rate": 2.2000010888412137e-06,
+      "loss": 0.585,
+      "step": 33585
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.385522269759824,
+      "learning_rate": 2.1995236159813437e-06,
+      "loss": 0.7789,
+      "step": 33586
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6111508611506773,
+      "learning_rate": 2.199046188538313e-06,
+      "loss": 0.7801,
+      "step": 33587
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2779717421040633,
+      "learning_rate": 2.198568806514906e-06,
+      "loss": 0.6182,
+      "step": 33588
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4072977277570897,
+      "learning_rate": 2.198091469913902e-06,
+      "loss": 0.6032,
+      "step": 33589
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3691324938628402,
+      "learning_rate": 2.197614178738081e-06,
+      "loss": 0.7827,
+      "step": 33590
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7491033940636767,
+      "learning_rate": 2.197136932990218e-06,
+      "loss": 0.7039,
+      "step": 33591
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.681079340272606,
+      "learning_rate": 2.1966597326730944e-06,
+      "loss": 0.8065,
+      "step": 33592
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.283071989532434,
+      "learning_rate": 2.19618257778949e-06,
+      "loss": 0.5798,
+      "step": 33593
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3333894267210096,
+      "learning_rate": 2.1957054683421818e-06,
+      "loss": 0.6908,
+      "step": 33594
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3712935126624903,
+      "learning_rate": 2.1952284043339443e-06,
+      "loss": 0.7728,
+      "step": 33595
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.346535584956228,
+      "learning_rate": 2.194751385767561e-06,
+      "loss": 0.6277,
+      "step": 33596
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3989442654644262,
+      "learning_rate": 2.1942744126458027e-06,
+      "loss": 0.663,
+      "step": 33597
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.685649398829474,
+      "learning_rate": 2.1937974849714505e-06,
+      "loss": 0.6591,
+      "step": 33598
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2550395381998416,
+      "learning_rate": 2.193320602747283e-06,
+      "loss": 0.6349,
+      "step": 33599
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3694334695217245,
+      "learning_rate": 2.192843765976075e-06,
+      "loss": 0.6697,
+      "step": 33600
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.439531301900179,
+      "learning_rate": 2.1923669746606004e-06,
+      "loss": 0.6842,
+      "step": 33601
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.224736539410822,
+      "learning_rate": 2.191890228803636e-06,
+      "loss": 0.7032,
+      "step": 33602
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5049357273626733,
+      "learning_rate": 2.191413528407964e-06,
+      "loss": 0.755,
+      "step": 33603
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4200525564539126,
+      "learning_rate": 2.1909368734763505e-06,
+      "loss": 0.8109,
+      "step": 33604
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2919639557308775,
+      "learning_rate": 2.1904602640115733e-06,
+      "loss": 0.7884,
+      "step": 33605
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4311877110945284,
+      "learning_rate": 2.189983700016409e-06,
+      "loss": 0.7402,
+      "step": 33606
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.316083501857172,
+      "learning_rate": 2.1895071814936386e-06,
+      "loss": 0.7521,
+      "step": 33607
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5439466763373655,
+      "learning_rate": 2.189030708446024e-06,
+      "loss": 0.648,
+      "step": 33608
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4242086805120833,
+      "learning_rate": 2.188554280876346e-06,
+      "loss": 0.7261,
+      "step": 33609
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2731688085508701,
+      "learning_rate": 2.1880778987873806e-06,
+      "loss": 0.718,
+      "step": 33610
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1886829463065187,
+      "learning_rate": 2.1876015621818956e-06,
+      "loss": 0.7255,
+      "step": 33611
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3870814354727612,
+      "learning_rate": 2.187125271062671e-06,
+      "loss": 0.7949,
+      "step": 33612
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5691479078033805,
+      "learning_rate": 2.1866490254324745e-06,
+      "loss": 0.4539,
+      "step": 33613
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6320635193357096,
+      "learning_rate": 2.186172825294083e-06,
+      "loss": 0.6713,
+      "step": 33614
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3701098466564017,
+      "learning_rate": 2.1856966706502647e-06,
+      "loss": 0.6587,
+      "step": 33615
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3058070639215362,
+      "learning_rate": 2.1852205615037968e-06,
+      "loss": 0.7041,
+      "step": 33616
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2844401381396873,
+      "learning_rate": 2.1847444978574485e-06,
+      "loss": 0.7685,
+      "step": 33617
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5968468205022123,
+      "learning_rate": 2.18426847971399e-06,
+      "loss": 0.5383,
+      "step": 33618
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6071592420339612,
+      "learning_rate": 2.1837925070761946e-06,
+      "loss": 0.5314,
+      "step": 33619
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5569231481256497,
+      "learning_rate": 2.183316579946836e-06,
+      "loss": 0.4636,
+      "step": 33620
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3861383347530964,
+      "learning_rate": 2.1828406983286832e-06,
+      "loss": 0.6463,
+      "step": 33621
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9375156255770203,
+      "learning_rate": 2.1823648622245043e-06,
+      "loss": 0.6552,
+      "step": 33622
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6237916859367933,
+      "learning_rate": 2.181889071637072e-06,
+      "loss": 0.7207,
+      "step": 33623
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3782942291685825,
+      "learning_rate": 2.1814133265691594e-06,
+      "loss": 0.6702,
+      "step": 33624
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.663803488443826,
+      "learning_rate": 2.1809376270235348e-06,
+      "loss": 0.7051,
+      "step": 33625
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3165186169500627,
+      "learning_rate": 2.180461973002963e-06,
+      "loss": 0.6984,
+      "step": 33626
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2975325927865173,
+      "learning_rate": 2.179986364510217e-06,
+      "loss": 0.6014,
+      "step": 33627
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5771462782584522,
+      "learning_rate": 2.1795108015480694e-06,
+      "loss": 0.4762,
+      "step": 33628
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4056141235557573,
+      "learning_rate": 2.1790352841192863e-06,
+      "loss": 0.7232,
+      "step": 33629
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1922456850911436,
+      "learning_rate": 2.1785598122266327e-06,
+      "loss": 0.5806,
+      "step": 33630
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2454648738152425,
+      "learning_rate": 2.178084385872882e-06,
+      "loss": 0.6053,
+      "step": 33631
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.494434758722581,
+      "learning_rate": 2.177609005060799e-06,
+      "loss": 0.6365,
+      "step": 33632
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3215491135383137,
+      "learning_rate": 2.1771336697931544e-06,
+      "loss": 0.6694,
+      "step": 33633
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.080604974699878,
+      "learning_rate": 2.1766583800727128e-06,
+      "loss": 0.6072,
+      "step": 33634
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5461453029480405,
+      "learning_rate": 2.1761831359022446e-06,
+      "loss": 0.6777,
+      "step": 33635
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5911482371732119,
+      "learning_rate": 2.1757079372845125e-06,
+      "loss": 0.49,
+      "step": 33636
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2790373097854206,
+      "learning_rate": 2.175232784222289e-06,
+      "loss": 0.6173,
+      "step": 33637
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.504014087956348,
+      "learning_rate": 2.1747576767183374e-06,
+      "loss": 0.6272,
+      "step": 33638
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4704260549586272,
+      "learning_rate": 2.1742826147754213e-06,
+      "loss": 0.5291,
+      "step": 33639
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.006079633975556,
+      "learning_rate": 2.173807598396309e-06,
+      "loss": 0.6034,
+      "step": 33640
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4698368208777841,
+      "learning_rate": 2.1733326275837708e-06,
+      "loss": 0.5971,
+      "step": 33641
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.53820242238217,
+      "learning_rate": 2.172857702340566e-06,
+      "loss": 0.7134,
+      "step": 33642
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3289191718487874,
+      "learning_rate": 2.17238282266946e-06,
+      "loss": 0.7313,
+      "step": 33643
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3125523682769142,
+      "learning_rate": 2.17190798857322e-06,
+      "loss": 0.6451,
+      "step": 33644
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3981449725029034,
+      "learning_rate": 2.1714332000546114e-06,
+      "loss": 0.5653,
+      "step": 33645
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4474550166225801,
+      "learning_rate": 2.1709584571163976e-06,
+      "loss": 0.5981,
+      "step": 33646
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.605757843552407,
+      "learning_rate": 2.1704837597613406e-06,
+      "loss": 0.7783,
+      "step": 33647
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6838852359501189,
+      "learning_rate": 2.1700091079922047e-06,
+      "loss": 0.8011,
+      "step": 33648
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3993679211619605,
+      "learning_rate": 2.1695345018117577e-06,
+      "loss": 0.7031,
+      "step": 33649
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.311143903122554,
+      "learning_rate": 2.16905994122276e-06,
+      "loss": 0.7323,
+      "step": 33650
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.694120443094086,
+      "learning_rate": 2.1685854262279714e-06,
+      "loss": 0.7224,
+      "step": 33651
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4632073796953904,
+      "learning_rate": 2.1681109568301617e-06,
+      "loss": 0.4896,
+      "step": 33652
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.517547439133167,
+      "learning_rate": 2.1676365330320857e-06,
+      "loss": 0.6508,
+      "step": 33653
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5929978128316028,
+      "learning_rate": 2.167162154836513e-06,
+      "loss": 0.492,
+      "step": 33654
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5284416823994489,
+      "learning_rate": 2.1666878222461986e-06,
+      "loss": 0.639,
+      "step": 33655
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4715565041648995,
+      "learning_rate": 2.1662135352639115e-06,
+      "loss": 0.7466,
+      "step": 33656
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7479097905131227,
+      "learning_rate": 2.1657392938924047e-06,
+      "loss": 0.703,
+      "step": 33657
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5162269069222103,
+      "learning_rate": 2.165265098134448e-06,
+      "loss": 0.5765,
+      "step": 33658
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.651300287033603,
+      "learning_rate": 2.1647909479927987e-06,
+      "loss": 0.7207,
+      "step": 33659
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4017791159546127,
+      "learning_rate": 2.1643168434702142e-06,
+      "loss": 0.7262,
+      "step": 33660
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6031179074138061,
+      "learning_rate": 2.1638427845694574e-06,
+      "loss": 0.4749,
+      "step": 33661
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1942103170167437,
+      "learning_rate": 2.1633687712932917e-06,
+      "loss": 0.5915,
+      "step": 33662
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6067662694864747,
+      "learning_rate": 2.1628948036444743e-06,
+      "loss": 0.495,
+      "step": 33663
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3544537299308754,
+      "learning_rate": 2.162420881625761e-06,
+      "loss": 0.6305,
+      "step": 33664
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3455072630872178,
+      "learning_rate": 2.161947005239915e-06,
+      "loss": 0.6575,
+      "step": 33665
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2530270915825166,
+      "learning_rate": 2.1614731744896976e-06,
+      "loss": 0.5976,
+      "step": 33666
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4127934634894772,
+      "learning_rate": 2.160999389377866e-06,
+      "loss": 0.6704,
+      "step": 33667
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6783790296495957,
+      "learning_rate": 2.160525649907175e-06,
+      "loss": 0.6868,
+      "step": 33668
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5360558901736854,
+      "learning_rate": 2.1600519560803844e-06,
+      "loss": 0.8965,
+      "step": 33669
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6790215809438485,
+      "learning_rate": 2.1595783079002565e-06,
+      "loss": 0.6741,
+      "step": 33670
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4588540390344826,
+      "learning_rate": 2.159104705369546e-06,
+      "loss": 0.7384,
+      "step": 33671
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4451004060978785,
+      "learning_rate": 2.158631148491007e-06,
+      "loss": 0.7668,
+      "step": 33672
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.305895239401425,
+      "learning_rate": 2.158157637267404e-06,
+      "loss": 0.6447,
+      "step": 33673
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2510940497541756,
+      "learning_rate": 2.1576841717014864e-06,
+      "loss": 0.6477,
+      "step": 33674
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4664204736323925,
+      "learning_rate": 2.1572107517960174e-06,
+      "loss": 0.6964,
+      "step": 33675
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2277170854169803,
+      "learning_rate": 2.156737377553748e-06,
+      "loss": 0.5808,
+      "step": 33676
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.547741173708881,
+      "learning_rate": 2.1562640489774388e-06,
+      "loss": 0.7142,
+      "step": 33677
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.287420751028368,
+      "learning_rate": 2.1557907660698408e-06,
+      "loss": 0.7524,
+      "step": 33678
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.897664327718468,
+      "learning_rate": 2.155317528833716e-06,
+      "loss": 0.7933,
+      "step": 33679
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2695281083097925,
+      "learning_rate": 2.1548443372718153e-06,
+      "loss": 0.6532,
+      "step": 33680
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3324807613427765,
+      "learning_rate": 2.1543711913868915e-06,
+      "loss": 0.6876,
+      "step": 33681
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2000092914235054,
+      "learning_rate": 2.153898091181703e-06,
+      "loss": 0.6794,
+      "step": 33682
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3301582408793804,
+      "learning_rate": 2.153425036659007e-06,
+      "loss": 0.6043,
+      "step": 33683
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4000593938552064,
+      "learning_rate": 2.1529520278215543e-06,
+      "loss": 0.574,
+      "step": 33684
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3748558912663305,
+      "learning_rate": 2.152479064672096e-06,
+      "loss": 0.6888,
+      "step": 33685
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.655129574488674,
+      "learning_rate": 2.1520061472133903e-06,
+      "loss": 0.6367,
+      "step": 33686
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2113487033494763,
+      "learning_rate": 2.151533275448191e-06,
+      "loss": 0.7068,
+      "step": 33687
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6125065829669027,
+      "learning_rate": 2.151060449379251e-06,
+      "loss": 0.7121,
+      "step": 33688
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5508489394121174,
+      "learning_rate": 2.1505876690093176e-06,
+      "loss": 0.7204,
+      "step": 33689
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5313386307011536,
+      "learning_rate": 2.1501149343411486e-06,
+      "loss": 0.6677,
+      "step": 33690
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4746995232088356,
+      "learning_rate": 2.149642245377499e-06,
+      "loss": 0.7175,
+      "step": 33691
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.199742684808051,
+      "learning_rate": 2.1491696021211174e-06,
+      "loss": 0.7021,
+      "step": 33692
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4754131127647376,
+      "learning_rate": 2.1486970045747525e-06,
+      "loss": 0.632,
+      "step": 33693
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4147968495773828,
+      "learning_rate": 2.148224452741162e-06,
+      "loss": 0.6458,
+      "step": 33694
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3246131633843337,
+      "learning_rate": 2.1477519466230935e-06,
+      "loss": 0.6625,
+      "step": 33695
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4577443634258858,
+      "learning_rate": 2.147279486223297e-06,
+      "loss": 0.8817,
+      "step": 33696
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5640572788246278,
+      "learning_rate": 2.14680707154453e-06,
+      "loss": 0.4923,
+      "step": 33697
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3307839234412555,
+      "learning_rate": 2.1463347025895375e-06,
+      "loss": 0.8283,
+      "step": 33698
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.42005519354007,
+      "learning_rate": 2.1458623793610678e-06,
+      "loss": 0.803,
+      "step": 33699
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0943527210377,
+      "learning_rate": 2.145390101861875e-06,
+      "loss": 0.6453,
+      "step": 33700
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3627372993680622,
+      "learning_rate": 2.1449178700947125e-06,
+      "loss": 0.6897,
+      "step": 33701
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.457950416488855,
+      "learning_rate": 2.1444456840623195e-06,
+      "loss": 0.6391,
+      "step": 33702
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3160036601077256,
+      "learning_rate": 2.1439735437674504e-06,
+      "loss": 0.6356,
+      "step": 33703
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.292551107161986,
+      "learning_rate": 2.1435014492128547e-06,
+      "loss": 0.6762,
+      "step": 33704
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1386953478569621,
+      "learning_rate": 2.1430294004012874e-06,
+      "loss": 0.5305,
+      "step": 33705
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3648743063603654,
+      "learning_rate": 2.142557397335484e-06,
+      "loss": 0.7295,
+      "step": 33706
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6172308945815569,
+      "learning_rate": 2.142085440018199e-06,
+      "loss": 0.5073,
+      "step": 33707
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6711503577625992,
+      "learning_rate": 2.1416135284521812e-06,
+      "loss": 0.6469,
+      "step": 33708
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4014363381474086,
+      "learning_rate": 2.1411416626401793e-06,
+      "loss": 0.7463,
+      "step": 33709
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3977335533006732,
+      "learning_rate": 2.1406698425849383e-06,
+      "loss": 0.7673,
+      "step": 33710
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4730139472337935,
+      "learning_rate": 2.140198068289204e-06,
+      "loss": 0.5994,
+      "step": 33711
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6073035239913018,
+      "learning_rate": 2.1397263397557266e-06,
+      "loss": 0.7131,
+      "step": 33712
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3353900990048946,
+      "learning_rate": 2.1392546569872498e-06,
+      "loss": 0.763,
+      "step": 33713
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4141458427228584,
+      "learning_rate": 2.1387830199865224e-06,
+      "loss": 0.6632,
+      "step": 33714
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.022516716677057,
+      "learning_rate": 2.138311428756287e-06,
+      "loss": 0.8606,
+      "step": 33715
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5271074667525681,
+      "learning_rate": 2.137839883299294e-06,
+      "loss": 0.7628,
+      "step": 33716
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5054451196517173,
+      "learning_rate": 2.1373683836182835e-06,
+      "loss": 0.7824,
+      "step": 33717
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7216417844283105,
+      "learning_rate": 2.136896929716006e-06,
+      "loss": 0.5584,
+      "step": 33718
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2626341298174195,
+      "learning_rate": 2.1364255215952047e-06,
+      "loss": 0.6817,
+      "step": 33719
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2509098009805302,
+      "learning_rate": 2.1359541592586197e-06,
+      "loss": 0.7286,
+      "step": 33720
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5098048597970153,
+      "learning_rate": 2.1354828427090004e-06,
+      "loss": 0.7211,
+      "step": 33721
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.385984555303032,
+      "learning_rate": 2.135011571949093e-06,
+      "loss": 0.7983,
+      "step": 33722
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5037492847136988,
+      "learning_rate": 2.134540346981637e-06,
+      "loss": 0.7849,
+      "step": 33723
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4668159892159691,
+      "learning_rate": 2.134069167809375e-06,
+      "loss": 0.7255,
+      "step": 33724
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1838109670274073,
+      "learning_rate": 2.1335980344350527e-06,
+      "loss": 0.7344,
+      "step": 33725
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3935612260139465,
+      "learning_rate": 2.1331269468614156e-06,
+      "loss": 0.614,
+      "step": 33726
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6142354959953353,
+      "learning_rate": 2.1326559050912042e-06,
+      "loss": 0.5915,
+      "step": 33727
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2763648361356417,
+      "learning_rate": 2.1321849091271573e-06,
+      "loss": 0.6682,
+      "step": 33728
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3268097393732656,
+      "learning_rate": 2.1317139589720227e-06,
+      "loss": 0.665,
+      "step": 33729
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9324682019732078,
+      "learning_rate": 2.1312430546285414e-06,
+      "loss": 0.6563,
+      "step": 33730
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.465334163479984,
+      "learning_rate": 2.1307721960994556e-06,
+      "loss": 0.7323,
+      "step": 33731
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7273972515521263,
+      "learning_rate": 2.130301383387502e-06,
+      "loss": 0.6661,
+      "step": 33732
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.121794201206721,
+      "learning_rate": 2.1298306164954275e-06,
+      "loss": 0.5647,
+      "step": 33733
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5337919802398414,
+      "learning_rate": 2.129359895425969e-06,
+      "loss": 0.6264,
+      "step": 33734
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2465553609329314,
+      "learning_rate": 2.128889220181871e-06,
+      "loss": 0.6806,
+      "step": 33735
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.372151833111478,
+      "learning_rate": 2.1284185907658695e-06,
+      "loss": 0.6487,
+      "step": 33736
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3821161386120597,
+      "learning_rate": 2.12794800718071e-06,
+      "loss": 0.7035,
+      "step": 33737
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3362169707702662,
+      "learning_rate": 2.1274774694291267e-06,
+      "loss": 0.717,
+      "step": 33738
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5284831408496273,
+      "learning_rate": 2.1270069775138647e-06,
+      "loss": 0.7157,
+      "step": 33739
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4457898714250743,
+      "learning_rate": 2.1265365314376598e-06,
+      "loss": 0.7128,
+      "step": 33740
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.73223864748674,
+      "learning_rate": 2.1260661312032506e-06,
+      "loss": 0.7709,
+      "step": 33741
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3115344456949551,
+      "learning_rate": 2.125595776813376e-06,
+      "loss": 0.6717,
+      "step": 33742
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4449475137946761,
+      "learning_rate": 2.125125468270779e-06,
+      "loss": 0.6691,
+      "step": 33743
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2708553222352585,
+      "learning_rate": 2.124655205578194e-06,
+      "loss": 0.6943,
+      "step": 33744
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2894320015746574,
+      "learning_rate": 2.1241849887383583e-06,
+      "loss": 0.6798,
+      "step": 33745
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5773451390518323,
+      "learning_rate": 2.12371481775401e-06,
+      "loss": 0.7226,
+      "step": 33746
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7684742887406175,
+      "learning_rate": 2.1232446926278905e-06,
+      "loss": 0.5633,
+      "step": 33747
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4217606311988102,
+      "learning_rate": 2.1227746133627346e-06,
+      "loss": 0.739,
+      "step": 33748
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4889862596820693,
+      "learning_rate": 2.1223045799612763e-06,
+      "loss": 0.7174,
+      "step": 33749
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2151323096985744,
+      "learning_rate": 2.121834592426254e-06,
+      "loss": 0.5921,
+      "step": 33750
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5419941774596129,
+      "learning_rate": 2.121364650760408e-06,
+      "loss": 0.6978,
+      "step": 33751
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6115352755531998,
+      "learning_rate": 2.1208947549664706e-06,
+      "loss": 0.5196,
+      "step": 33752
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5346354350787332,
+      "learning_rate": 2.120424905047176e-06,
+      "loss": 0.6287,
+      "step": 33753
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2668748527001936,
+      "learning_rate": 2.1199551010052653e-06,
+      "loss": 0.5644,
+      "step": 33754
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4034229054637735,
+      "learning_rate": 2.119485342843468e-06,
+      "loss": 0.7014,
+      "step": 33755
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3134695688180256,
+      "learning_rate": 2.1190156305645247e-06,
+      "loss": 0.6894,
+      "step": 33756
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.394484142449385,
+      "learning_rate": 2.1185459641711646e-06,
+      "loss": 0.6921,
+      "step": 33757
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.264945874535942,
+      "learning_rate": 2.118076343666129e-06,
+      "loss": 0.7675,
+      "step": 33758
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3954773078941982,
+      "learning_rate": 2.117606769052144e-06,
+      "loss": 0.6229,
+      "step": 33759
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3825381961170569,
+      "learning_rate": 2.1171372403319514e-06,
+      "loss": 0.7297,
+      "step": 33760
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4054596456165536,
+      "learning_rate": 2.116667757508282e-06,
+      "loss": 0.8018,
+      "step": 33761
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4804821485897448,
+      "learning_rate": 2.1161983205838645e-06,
+      "loss": 0.7574,
+      "step": 33762
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.850247340740656,
+      "learning_rate": 2.115728929561437e-06,
+      "loss": 0.7287,
+      "step": 33763
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5002094688669183,
+      "learning_rate": 2.115259584443735e-06,
+      "loss": 0.6898,
+      "step": 33764
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0330272162465612,
+      "learning_rate": 2.114790285233488e-06,
+      "loss": 0.5125,
+      "step": 33765
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4804502069909375,
+      "learning_rate": 2.114321031933425e-06,
+      "loss": 0.6182,
+      "step": 33766
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3688785291313672,
+      "learning_rate": 2.1138518245462816e-06,
+      "loss": 0.6368,
+      "step": 33767
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5111591093666785,
+      "learning_rate": 2.113382663074792e-06,
+      "loss": 0.7004,
+      "step": 33768
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1322644096388457,
+      "learning_rate": 2.1129135475216866e-06,
+      "loss": 0.6093,
+      "step": 33769
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4577755710038385,
+      "learning_rate": 2.1124444778896914e-06,
+      "loss": 0.5737,
+      "step": 33770
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3317433872135158,
+      "learning_rate": 2.111975454181543e-06,
+      "loss": 0.7381,
+      "step": 33771
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6354844303537178,
+      "learning_rate": 2.1115064763999726e-06,
+      "loss": 0.5011,
+      "step": 33772
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4694838565845396,
+      "learning_rate": 2.1110375445477093e-06,
+      "loss": 0.7542,
+      "step": 33773
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.308287035814595,
+      "learning_rate": 2.1105686586274797e-06,
+      "loss": 0.7427,
+      "step": 33774
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4353959865282229,
+      "learning_rate": 2.110099818642021e-06,
+      "loss": 0.645,
+      "step": 33775
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2061487735758494,
+      "learning_rate": 2.1096310245940555e-06,
+      "loss": 0.7072,
+      "step": 33776
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6851362835391541,
+      "learning_rate": 2.1091622764863185e-06,
+      "loss": 0.7206,
+      "step": 33777
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.361905517578501,
+      "learning_rate": 2.1086935743215354e-06,
+      "loss": 0.7021,
+      "step": 33778
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6064398100126087,
+      "learning_rate": 2.1082249181024384e-06,
+      "loss": 0.6759,
+      "step": 33779
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.472703096756662,
+      "learning_rate": 2.107756307831752e-06,
+      "loss": 0.6267,
+      "step": 33780
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.616886443548224,
+      "learning_rate": 2.1072877435122095e-06,
+      "loss": 0.5057,
+      "step": 33781
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3022112964644639,
+      "learning_rate": 2.106819225146538e-06,
+      "loss": 0.6113,
+      "step": 33782
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6081896344943725,
+      "learning_rate": 2.1063507527374594e-06,
+      "loss": 0.7378,
+      "step": 33783
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.468245937934095,
+      "learning_rate": 2.105882326287707e-06,
+      "loss": 0.733,
+      "step": 33784
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3888864851961482,
+      "learning_rate": 2.105413945800007e-06,
+      "loss": 0.6806,
+      "step": 33785
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6132954455764055,
+      "learning_rate": 2.1049456112770907e-06,
+      "loss": 0.481,
+      "step": 33786
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5037829123661286,
+      "learning_rate": 2.104477322721675e-06,
+      "loss": 0.6268,
+      "step": 33787
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4472550000912061,
+      "learning_rate": 2.104009080136493e-06,
+      "loss": 0.6366,
+      "step": 33788
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6214944799544001,
+      "learning_rate": 2.103540883524272e-06,
+      "loss": 0.4821,
+      "step": 33789
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2477593549908508,
+      "learning_rate": 2.1030727328877367e-06,
+      "loss": 0.6335,
+      "step": 33790
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4586715211206909,
+      "learning_rate": 2.1026046282296085e-06,
+      "loss": 0.688,
+      "step": 33791
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.622133713260741,
+      "learning_rate": 2.1021365695526176e-06,
+      "loss": 0.6887,
+      "step": 33792
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.557134644789213,
+      "learning_rate": 2.10166855685949e-06,
+      "loss": 0.7681,
+      "step": 33793
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6999108030436,
+      "learning_rate": 2.1012005901529463e-06,
+      "loss": 0.6904,
+      "step": 33794
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4761037096920133,
+      "learning_rate": 2.1007326694357153e-06,
+      "loss": 0.5735,
+      "step": 33795
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6216334715703407,
+      "learning_rate": 2.1002647947105205e-06,
+      "loss": 0.5072,
+      "step": 33796
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4771435314901276,
+      "learning_rate": 2.099796965980082e-06,
+      "loss": 0.6409,
+      "step": 33797
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5593381709614713,
+      "learning_rate": 2.099329183247126e-06,
+      "loss": 0.63,
+      "step": 33798
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2270625876684298,
+      "learning_rate": 2.0988614465143807e-06,
+      "loss": 0.6225,
+      "step": 33799
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.321314494759793,
+      "learning_rate": 2.0983937557845647e-06,
+      "loss": 0.7524,
+      "step": 33800
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6037899925456819,
+      "learning_rate": 2.0979261110604e-06,
+      "loss": 0.5269,
+      "step": 33801
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3138845975775366,
+      "learning_rate": 2.09745851234461e-06,
+      "loss": 0.7267,
+      "step": 33802
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7182133760268379,
+      "learning_rate": 2.0969909596399253e-06,
+      "loss": 0.6893,
+      "step": 33803
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5961778960891015,
+      "learning_rate": 2.0965234529490552e-06,
+      "loss": 0.4976,
+      "step": 33804
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4079336956448407,
+      "learning_rate": 2.096055992274727e-06,
+      "loss": 0.7383,
+      "step": 33805
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4236391293893462,
+      "learning_rate": 2.0955885776196637e-06,
+      "loss": 0.7214,
+      "step": 33806
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4011805877620376,
+      "learning_rate": 2.09512120898659e-06,
+      "loss": 0.699,
+      "step": 33807
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.626733694181567,
+      "learning_rate": 2.0946538863782216e-06,
+      "loss": 0.6762,
+      "step": 33808
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2870505736747657,
+      "learning_rate": 2.0941866097972797e-06,
+      "loss": 0.5921,
+      "step": 33809
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.28941789994244,
+      "learning_rate": 2.0937193792464883e-06,
+      "loss": 0.6341,
+      "step": 33810
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7110408087919207,
+      "learning_rate": 2.0932521947285635e-06,
+      "loss": 0.8544,
+      "step": 33811
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6287402137021307,
+      "learning_rate": 2.092785056246229e-06,
+      "loss": 0.6615,
+      "step": 33812
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3778026923574656,
+      "learning_rate": 2.092317963802202e-06,
+      "loss": 0.6145,
+      "step": 33813
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.438885386865841,
+      "learning_rate": 2.091850917399206e-06,
+      "loss": 0.7759,
+      "step": 33814
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2942453213057654,
+      "learning_rate": 2.091383917039954e-06,
+      "loss": 0.6442,
+      "step": 33815
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5819935612178965,
+      "learning_rate": 2.090916962727172e-06,
+      "loss": 0.4839,
+      "step": 33816
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5138372461963325,
+      "learning_rate": 2.0904500544635752e-06,
+      "loss": 0.7754,
+      "step": 33817
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2732281283411422,
+      "learning_rate": 2.08998319225188e-06,
+      "loss": 0.6141,
+      "step": 33818
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2799954565414497,
+      "learning_rate": 2.089516376094807e-06,
+      "loss": 0.6832,
+      "step": 33819
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3921107090452052,
+      "learning_rate": 2.089049605995076e-06,
+      "loss": 0.687,
+      "step": 33820
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3503496268723356,
+      "learning_rate": 2.0885828819554035e-06,
+      "loss": 0.7045,
+      "step": 33821
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7024869383301717,
+      "learning_rate": 2.088116203978503e-06,
+      "loss": 0.7819,
+      "step": 33822
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5985092466586976,
+      "learning_rate": 2.087649572067095e-06,
+      "loss": 0.623,
+      "step": 33823
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6836639532396576,
+      "learning_rate": 2.087182986223899e-06,
+      "loss": 0.642,
+      "step": 33824
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1746707147882276,
+      "learning_rate": 2.0867164464516287e-06,
+      "loss": 0.649,
+      "step": 33825
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6477002305266006,
+      "learning_rate": 2.086249952752998e-06,
+      "loss": 0.6621,
+      "step": 33826
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2682482882438304,
+      "learning_rate": 2.0857835051307253e-06,
+      "loss": 0.6058,
+      "step": 33827
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4220424942924044,
+      "learning_rate": 2.0853171035875296e-06,
+      "loss": 0.6882,
+      "step": 33828
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4570021664382855,
+      "learning_rate": 2.084850748126124e-06,
+      "loss": 0.5337,
+      "step": 33829
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5308548054414535,
+      "learning_rate": 2.0843844387492197e-06,
+      "loss": 0.7187,
+      "step": 33830
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7819376914860563,
+      "learning_rate": 2.083918175459535e-06,
+      "loss": 0.7655,
+      "step": 33831
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4051825058291874,
+      "learning_rate": 2.083451958259788e-06,
+      "loss": 0.6003,
+      "step": 33832
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2583939553106156,
+      "learning_rate": 2.0829857871526904e-06,
+      "loss": 0.6068,
+      "step": 33833
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5736239966810421,
+      "learning_rate": 2.0825196621409526e-06,
+      "loss": 0.4774,
+      "step": 33834
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2645999327550694,
+      "learning_rate": 2.0820535832272947e-06,
+      "loss": 0.6071,
+      "step": 33835
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.559784891585974,
+      "learning_rate": 2.0815875504144257e-06,
+      "loss": 0.5016,
+      "step": 33836
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6459904781995884,
+      "learning_rate": 2.081121563705063e-06,
+      "loss": 0.6619,
+      "step": 33837
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5833307466756974,
+      "learning_rate": 2.080655623101915e-06,
+      "loss": 0.7645,
+      "step": 33838
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3177458190724407,
+      "learning_rate": 2.0801897286077e-06,
+      "loss": 0.7525,
+      "step": 33839
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3346838676554706,
+      "learning_rate": 2.0797238802251252e-06,
+      "loss": 0.6754,
+      "step": 33840
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4328538860966087,
+      "learning_rate": 2.079258077956907e-06,
+      "loss": 0.8088,
+      "step": 33841
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3777643838403353,
+      "learning_rate": 2.078792321805757e-06,
+      "loss": 0.6992,
+      "step": 33842
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5006583751092348,
+      "learning_rate": 2.078326611774384e-06,
+      "loss": 0.6846,
+      "step": 33843
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.352203528681812,
+      "learning_rate": 2.0778609478655e-06,
+      "loss": 0.5937,
+      "step": 33844
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6555687312216945,
+      "learning_rate": 2.0773953300818204e-06,
+      "loss": 0.5203,
+      "step": 33845
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9117727084952418,
+      "learning_rate": 2.076929758426055e-06,
+      "loss": 0.7556,
+      "step": 33846
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5883680386894962,
+      "learning_rate": 2.0764642329009087e-06,
+      "loss": 0.5358,
+      "step": 33847
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5363891973838044,
+      "learning_rate": 2.0759987535090966e-06,
+      "loss": 0.6573,
+      "step": 33848
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6208999317387018,
+      "learning_rate": 2.075533320253331e-06,
+      "loss": 0.7736,
+      "step": 33849
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2357563954190427,
+      "learning_rate": 2.07506793313632e-06,
+      "loss": 0.6769,
+      "step": 33850
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7046368887686199,
+      "learning_rate": 2.0746025921607695e-06,
+      "loss": 0.7018,
+      "step": 33851
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6089587597969628,
+      "learning_rate": 2.0741372973293915e-06,
+      "loss": 0.8007,
+      "step": 33852
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2907353945601803,
+      "learning_rate": 2.073672048644898e-06,
+      "loss": 0.6671,
+      "step": 33853
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.348266306059422,
+      "learning_rate": 2.0732068461099952e-06,
+      "loss": 0.6445,
+      "step": 33854
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4007573257291681,
+      "learning_rate": 2.07274168972739e-06,
+      "loss": 0.628,
+      "step": 33855
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4084011078541605,
+      "learning_rate": 2.0722765794997947e-06,
+      "loss": 0.6263,
+      "step": 33856
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5425260004510415,
+      "learning_rate": 2.0718115154299113e-06,
+      "loss": 0.6601,
+      "step": 33857
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.396029322766777,
+      "learning_rate": 2.0713464975204556e-06,
+      "loss": 0.6905,
+      "step": 33858
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3733085918361396,
+      "learning_rate": 2.0708815257741264e-06,
+      "loss": 0.7087,
+      "step": 33859
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4734834566458737,
+      "learning_rate": 2.0704166001936387e-06,
+      "loss": 0.7447,
+      "step": 33860
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.712327050907902,
+      "learning_rate": 2.0699517207816944e-06,
+      "loss": 0.6688,
+      "step": 33861
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4215796136267196,
+      "learning_rate": 2.0694868875410036e-06,
+      "loss": 0.8162,
+      "step": 33862
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5362975937089727,
+      "learning_rate": 2.0690221004742704e-06,
+      "loss": 0.6438,
+      "step": 33863
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.39536151090311,
+      "learning_rate": 2.068557359584199e-06,
+      "loss": 0.7389,
+      "step": 33864
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.536643912083368,
+      "learning_rate": 2.068092664873498e-06,
+      "loss": 0.8034,
+      "step": 33865
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5884955887527094,
+      "learning_rate": 2.067628016344875e-06,
+      "loss": 0.4964,
+      "step": 33866
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.187439693714429,
+      "learning_rate": 2.0671634140010333e-06,
+      "loss": 0.6008,
+      "step": 33867
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5598559639564817,
+      "learning_rate": 2.0666988578446757e-06,
+      "loss": 0.7414,
+      "step": 33868
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3860684805454937,
+      "learning_rate": 2.0662343478785074e-06,
+      "loss": 0.6995,
+      "step": 33869
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5742165673666157,
+      "learning_rate": 2.0657698841052386e-06,
+      "loss": 0.7435,
+      "step": 33870
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3982852259212448,
+      "learning_rate": 2.0653054665275686e-06,
+      "loss": 0.6975,
+      "step": 33871
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4412132306542087,
+      "learning_rate": 2.0648410951482e-06,
+      "loss": 0.6148,
+      "step": 33872
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3321910496876712,
+      "learning_rate": 2.064376769969839e-06,
+      "loss": 0.7141,
+      "step": 33873
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5799671589729352,
+      "learning_rate": 2.0639124909951913e-06,
+      "loss": 0.5332,
+      "step": 33874
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4307823279552903,
+      "learning_rate": 2.0634482582269578e-06,
+      "loss": 0.8041,
+      "step": 33875
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5940855761378536,
+      "learning_rate": 2.0629840716678384e-06,
+      "loss": 0.6719,
+      "step": 33876
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3560729199849864,
+      "learning_rate": 2.0625199313205403e-06,
+      "loss": 0.565,
+      "step": 33877
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4179213376942768,
+      "learning_rate": 2.062055837187763e-06,
+      "loss": 0.7018,
+      "step": 33878
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3896471971531772,
+      "learning_rate": 2.061591789272208e-06,
+      "loss": 0.6642,
+      "step": 33879
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4533526436217725,
+      "learning_rate": 2.0611277875765835e-06,
+      "loss": 0.555,
+      "step": 33880
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4520206254073502,
+      "learning_rate": 2.060663832103584e-06,
+      "loss": 0.8219,
+      "step": 33881
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3991392863642564,
+      "learning_rate": 2.0601999228559134e-06,
+      "loss": 0.6826,
+      "step": 33882
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3066322650208984,
+      "learning_rate": 2.0597360598362705e-06,
+      "loss": 0.7495,
+      "step": 33883
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.404518344064755,
+      "learning_rate": 2.0592722430473645e-06,
+      "loss": 0.5684,
+      "step": 33884
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2642378383763895,
+      "learning_rate": 2.0588084724918834e-06,
+      "loss": 0.7137,
+      "step": 33885
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6221824501225948,
+      "learning_rate": 2.058344748172534e-06,
+      "loss": 0.5325,
+      "step": 33886
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.57358477062937,
+      "learning_rate": 2.057881070092015e-06,
+      "loss": 0.6473,
+      "step": 33887
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5768537533379476,
+      "learning_rate": 2.0574174382530332e-06,
+      "loss": 0.7298,
+      "step": 33888
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1506293288288005,
+      "learning_rate": 2.0569538526582753e-06,
+      "loss": 0.6105,
+      "step": 33889
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3509605645203433,
+      "learning_rate": 2.0564903133104474e-06,
+      "loss": 0.6453,
+      "step": 33890
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6026807377736261,
+      "learning_rate": 2.0560268202122514e-06,
+      "loss": 0.473,
+      "step": 33891
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6142516346091784,
+      "learning_rate": 2.055563373366378e-06,
+      "loss": 0.6663,
+      "step": 33892
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5705788786794888,
+      "learning_rate": 2.0550999727755327e-06,
+      "loss": 0.4678,
+      "step": 33893
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8858102889755337,
+      "learning_rate": 2.0546366184424094e-06,
+      "loss": 0.7721,
+      "step": 33894
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.465133801235009,
+      "learning_rate": 2.054173310369708e-06,
+      "loss": 0.6726,
+      "step": 33895
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3574892901456805,
+      "learning_rate": 2.0537100485601238e-06,
+      "loss": 0.7499,
+      "step": 33896
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.483593708171208,
+      "learning_rate": 2.053246833016358e-06,
+      "loss": 0.708,
+      "step": 33897
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.516158108136848,
+      "learning_rate": 2.052783663741105e-06,
+      "loss": 0.6373,
+      "step": 33898
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.551805323163928,
+      "learning_rate": 2.052320540737058e-06,
+      "loss": 0.7761,
+      "step": 33899
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.257134351711237,
+      "learning_rate": 2.0518574640069178e-06,
+      "loss": 0.6375,
+      "step": 33900
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6399390204766413,
+      "learning_rate": 2.051394433553382e-06,
+      "loss": 0.5257,
+      "step": 33901
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.525121142620485,
+      "learning_rate": 2.0509314493791444e-06,
+      "loss": 0.7714,
+      "step": 33902
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6515581145404978,
+      "learning_rate": 2.050468511486898e-06,
+      "loss": 0.7467,
+      "step": 33903
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4101638568571626,
+      "learning_rate": 2.0500056198793405e-06,
+      "loss": 0.597,
+      "step": 33904
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5921066360632299,
+      "learning_rate": 2.04954277455917e-06,
+      "loss": 0.526,
+      "step": 33905
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5253406566767764,
+      "learning_rate": 2.049079975529079e-06,
+      "loss": 0.7166,
+      "step": 33906
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6639228620931805,
+      "learning_rate": 2.0486172227917587e-06,
+      "loss": 0.6105,
+      "step": 33907
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3366656018866663,
+      "learning_rate": 2.0481545163499062e-06,
+      "loss": 0.7046,
+      "step": 33908
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.480752829768215,
+      "learning_rate": 2.0476918562062186e-06,
+      "loss": 0.5945,
+      "step": 33909
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.709448942715548,
+      "learning_rate": 2.0472292423633857e-06,
+      "loss": 0.5408,
+      "step": 33910
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2987202159097144,
+      "learning_rate": 2.0467666748241e-06,
+      "loss": 0.5901,
+      "step": 33911
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6448037636952402,
+      "learning_rate": 2.0463041535910587e-06,
+      "loss": 0.7401,
+      "step": 33912
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2329129274084563,
+      "learning_rate": 2.045841678666951e-06,
+      "loss": 0.69,
+      "step": 33913
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.458817490682668,
+      "learning_rate": 2.0453792500544734e-06,
+      "loss": 0.6884,
+      "step": 33914
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.621629167465992,
+      "learning_rate": 2.0449168677563136e-06,
+      "loss": 0.528,
+      "step": 33915
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3959053852405416,
+      "learning_rate": 2.04445453177517e-06,
+      "loss": 0.8071,
+      "step": 33916
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2741478993910325,
+      "learning_rate": 2.043992242113727e-06,
+      "loss": 0.6308,
+      "step": 33917
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4911460571060702,
+      "learning_rate": 2.0435299987746825e-06,
+      "loss": 0.6899,
+      "step": 33918
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2936940133699772,
+      "learning_rate": 2.0430678017607266e-06,
+      "loss": 0.6252,
+      "step": 33919
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7539513238985118,
+      "learning_rate": 2.0426056510745453e-06,
+      "loss": 0.6739,
+      "step": 33920
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4267333440429442,
+      "learning_rate": 2.0421435467188333e-06,
+      "loss": 0.7299,
+      "step": 33921
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4410472495049313,
+      "learning_rate": 2.041681488696283e-06,
+      "loss": 0.8618,
+      "step": 33922
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4983280196196034,
+      "learning_rate": 2.041219477009584e-06,
+      "loss": 0.7418,
+      "step": 33923
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.649140878503866,
+      "learning_rate": 2.040757511661421e-06,
+      "loss": 0.7051,
+      "step": 33924
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4656041949448015,
+      "learning_rate": 2.0402955926544887e-06,
+      "loss": 0.7869,
+      "step": 33925
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.0281921150775473,
+      "learning_rate": 2.0398337199914774e-06,
+      "loss": 0.7122,
+      "step": 33926
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.476071930870371,
+      "learning_rate": 2.039371893675075e-06,
+      "loss": 0.6442,
+      "step": 33927
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.172659344234367,
+      "learning_rate": 2.038910113707966e-06,
+      "loss": 0.754,
+      "step": 33928
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3759991455003757,
+      "learning_rate": 2.0384483800928433e-06,
+      "loss": 0.6615,
+      "step": 33929
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2373292682442711,
+      "learning_rate": 2.0379866928323977e-06,
+      "loss": 0.693,
+      "step": 33930
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4291154644390507,
+      "learning_rate": 2.0375250519293145e-06,
+      "loss": 0.74,
+      "step": 33931
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5303186766711019,
+      "learning_rate": 2.0370634573862778e-06,
+      "loss": 0.507,
+      "step": 33932
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7797565031426978,
+      "learning_rate": 2.036601909205982e-06,
+      "loss": 0.6633,
+      "step": 33933
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.424373057080336,
+      "learning_rate": 2.0361404073911086e-06,
+      "loss": 0.657,
+      "step": 33934
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.35679081487232,
+      "learning_rate": 2.035678951944349e-06,
+      "loss": 0.5602,
+      "step": 33935
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3702304014310538,
+      "learning_rate": 2.0352175428683853e-06,
+      "loss": 0.7308,
+      "step": 33936
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5554558665875384,
+      "learning_rate": 2.03475618016591e-06,
+      "loss": 0.6195,
+      "step": 33937
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.403573781734114,
+      "learning_rate": 2.034294863839603e-06,
+      "loss": 0.7179,
+      "step": 33938
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4539059393212064,
+      "learning_rate": 2.033833593892156e-06,
+      "loss": 0.7461,
+      "step": 33939
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4707939993976191,
+      "learning_rate": 2.033372370326252e-06,
+      "loss": 0.7374,
+      "step": 33940
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.522877421046177,
+      "learning_rate": 2.032911193144573e-06,
+      "loss": 0.6407,
+      "step": 33941
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6023342523567687,
+      "learning_rate": 2.0324500623498076e-06,
+      "loss": 0.6809,
+      "step": 33942
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3801660396801252,
+      "learning_rate": 2.0319889779446432e-06,
+      "loss": 0.6556,
+      "step": 33943
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.352817601312353,
+      "learning_rate": 2.031527939931761e-06,
+      "loss": 0.7507,
+      "step": 33944
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2971819674923253,
+      "learning_rate": 2.0310669483138446e-06,
+      "loss": 0.599,
+      "step": 33945
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.521650822597136,
+      "learning_rate": 2.030606003093578e-06,
+      "loss": 0.723,
+      "step": 33946
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2542205565913491,
+      "learning_rate": 2.0301451042736496e-06,
+      "loss": 0.5887,
+      "step": 33947
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4024995226337724,
+      "learning_rate": 2.0296842518567395e-06,
+      "loss": 0.6802,
+      "step": 33948
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.593535367930124,
+      "learning_rate": 2.0292234458455284e-06,
+      "loss": 0.4767,
+      "step": 33949
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6372640287237517,
+      "learning_rate": 2.0287626862427024e-06,
+      "loss": 0.6361,
+      "step": 33950
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.373857050125437,
+      "learning_rate": 2.0283019730509456e-06,
+      "loss": 0.5822,
+      "step": 33951
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3672880618543575,
+      "learning_rate": 2.0278413062729394e-06,
+      "loss": 0.6856,
+      "step": 33952
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5904119349646778,
+      "learning_rate": 2.027380685911361e-06,
+      "loss": 0.5002,
+      "step": 33953
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9207493770980149,
+      "learning_rate": 2.0269201119688975e-06,
+      "loss": 0.8776,
+      "step": 33954
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5880518638245328,
+      "learning_rate": 2.0264595844482315e-06,
+      "loss": 0.4969,
+      "step": 33955
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4859022989283914,
+      "learning_rate": 2.025999103352042e-06,
+      "loss": 0.6598,
+      "step": 33956
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.753039321386466,
+      "learning_rate": 2.025538668683008e-06,
+      "loss": 0.6885,
+      "step": 33957
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.579686743085785,
+      "learning_rate": 2.025078280443815e-06,
+      "loss": 0.5005,
+      "step": 33958
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3433584476907563,
+      "learning_rate": 2.024617938637139e-06,
+      "loss": 0.6589,
+      "step": 33959
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.368048266515319,
+      "learning_rate": 2.024157643265665e-06,
+      "loss": 0.7217,
+      "step": 33960
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4429552849219829,
+      "learning_rate": 2.023697394332067e-06,
+      "loss": 0.7701,
+      "step": 33961
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.371528141629872,
+      "learning_rate": 2.0232371918390305e-06,
+      "loss": 0.5814,
+      "step": 33962
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.463375740378783,
+      "learning_rate": 2.0227770357892307e-06,
+      "loss": 0.5857,
+      "step": 33963
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5978388951040816,
+      "learning_rate": 2.02231692618535e-06,
+      "loss": 0.4832,
+      "step": 33964
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6120924715576672,
+      "learning_rate": 2.021856863030067e-06,
+      "loss": 0.7783,
+      "step": 33965
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4358622837006054,
+      "learning_rate": 2.0213968463260556e-06,
+      "loss": 0.5462,
+      "step": 33966
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2978657260717352,
+      "learning_rate": 2.0209368760759983e-06,
+      "loss": 0.6873,
+      "step": 33967
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.605512096599238,
+      "learning_rate": 2.020476952282573e-06,
+      "loss": 0.5454,
+      "step": 33968
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5619698265663513,
+      "learning_rate": 2.0200170749484614e-06,
+      "loss": 0.6402,
+      "step": 33969
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7593982977642917,
+      "learning_rate": 2.019557244076332e-06,
+      "loss": 0.6796,
+      "step": 33970
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.0435263738994083,
+      "learning_rate": 2.0190974596688673e-06,
+      "loss": 0.6717,
+      "step": 33971
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.578345870320721,
+      "learning_rate": 2.018637721728747e-06,
+      "loss": 0.6785,
+      "step": 33972
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4234157212825138,
+      "learning_rate": 2.018178030258643e-06,
+      "loss": 0.7867,
+      "step": 33973
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5142973520652863,
+      "learning_rate": 2.0177183852612316e-06,
+      "loss": 0.8151,
+      "step": 33974
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5591258308924958,
+      "learning_rate": 2.017258786739191e-06,
+      "loss": 0.4918,
+      "step": 33975
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.525603327709199,
+      "learning_rate": 2.0167992346952005e-06,
+      "loss": 0.6906,
+      "step": 33976
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6536113922209812,
+      "learning_rate": 2.016339729131929e-06,
+      "loss": 0.7278,
+      "step": 33977
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4008640238426606,
+      "learning_rate": 2.0158802700520576e-06,
+      "loss": 0.5239,
+      "step": 33978
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2999565360672574,
+      "learning_rate": 2.015420857458259e-06,
+      "loss": 0.6108,
+      "step": 33979
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2198316583346287,
+      "learning_rate": 2.0149614913532055e-06,
+      "loss": 0.5743,
+      "step": 33980
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3413543706921671,
+      "learning_rate": 2.014502171739574e-06,
+      "loss": 0.6516,
+      "step": 33981
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4007241529244239,
+      "learning_rate": 2.014042898620042e-06,
+      "loss": 0.6797,
+      "step": 33982
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4113753243531644,
+      "learning_rate": 2.013583671997281e-06,
+      "loss": 0.7956,
+      "step": 33983
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6836451343399766,
+      "learning_rate": 2.0131244918739612e-06,
+      "loss": 0.5489,
+      "step": 33984
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3328975564428014,
+      "learning_rate": 2.012665358252759e-06,
+      "loss": 0.739,
+      "step": 33985
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3543260300488857,
+      "learning_rate": 2.012206271136353e-06,
+      "loss": 0.7017,
+      "step": 33986
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.527339392536447,
+      "learning_rate": 2.0117472305274065e-06,
+      "loss": 0.7361,
+      "step": 33987
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2366842363419301,
+      "learning_rate": 2.0112882364285968e-06,
+      "loss": 0.6584,
+      "step": 33988
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.441598383589644,
+      "learning_rate": 2.010829288842595e-06,
+      "loss": 0.5322,
+      "step": 33989
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.368460270944275,
+      "learning_rate": 2.0103703877720783e-06,
+      "loss": 0.5435,
+      "step": 33990
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6196404402402798,
+      "learning_rate": 2.009911533219714e-06,
+      "loss": 0.5134,
+      "step": 33991
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2178115380737808,
+      "learning_rate": 2.0094527251881713e-06,
+      "loss": 0.754,
+      "step": 33992
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2948154673020595,
+      "learning_rate": 2.008993963680127e-06,
+      "loss": 0.7019,
+      "step": 33993
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.412903181486578,
+      "learning_rate": 2.008535248698248e-06,
+      "loss": 0.6648,
+      "step": 33994
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.5782702964568536,
+      "learning_rate": 2.0080765802452097e-06,
+      "loss": 0.7105,
+      "step": 33995
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6029401315648792,
+      "learning_rate": 2.0076179583236765e-06,
+      "loss": 0.5219,
+      "step": 33996
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5925263293814411,
+      "learning_rate": 2.0071593829363245e-06,
+      "loss": 0.5096,
+      "step": 33997
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2930868141291776,
+      "learning_rate": 2.006700854085819e-06,
+      "loss": 0.7098,
+      "step": 33998
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.516578876094402,
+      "learning_rate": 2.0062423717748336e-06,
+      "loss": 0.6346,
+      "step": 33999
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4517778527387113,
+      "learning_rate": 2.005783936006036e-06,
+      "loss": 0.7356,
+      "step": 34000
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2068629551742074,
+      "learning_rate": 2.0053255467820933e-06,
+      "loss": 0.5699,
+      "step": 34001
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.535975834879882,
+      "learning_rate": 2.004867204105676e-06,
+      "loss": 0.7815,
+      "step": 34002
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.272097109910987,
+      "learning_rate": 2.004408907979456e-06,
+      "loss": 0.6734,
+      "step": 34003
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3224717281929252,
+      "learning_rate": 2.0039506584060987e-06,
+      "loss": 0.6535,
+      "step": 34004
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6156919197390902,
+      "learning_rate": 2.0034924553882683e-06,
+      "loss": 0.6777,
+      "step": 34005
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3731123825029572,
+      "learning_rate": 2.0030342989286376e-06,
+      "loss": 0.6597,
+      "step": 34006
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5494550625684909,
+      "learning_rate": 2.0025761890298757e-06,
+      "loss": 0.6716,
+      "step": 34007
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.401261294933937,
+      "learning_rate": 2.0021181256946465e-06,
+      "loss": 0.7558,
+      "step": 34008
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.143753797748778,
+      "learning_rate": 2.001660108925616e-06,
+      "loss": 0.7616,
+      "step": 34009
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5656326977030828,
+      "learning_rate": 2.001202138725451e-06,
+      "loss": 0.5971,
+      "step": 34010
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3112427710927452,
+      "learning_rate": 2.0007442150968236e-06,
+      "loss": 0.5891,
+      "step": 34011
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3982415891292677,
+      "learning_rate": 2.000286338042395e-06,
+      "loss": 0.6391,
+      "step": 34012
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.337958160948597,
+      "learning_rate": 1.9998285075648304e-06,
+      "loss": 0.617,
+      "step": 34013
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.103906475146408,
+      "learning_rate": 1.9993707236667992e-06,
+      "loss": 0.5948,
+      "step": 34014
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4718668829366548,
+      "learning_rate": 1.9989129863509617e-06,
+      "loss": 0.6785,
+      "step": 34015
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3453440932743506,
+      "learning_rate": 1.998455295619989e-06,
+      "loss": 0.6272,
+      "step": 34016
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6320724176385234,
+      "learning_rate": 1.9979976514765397e-06,
+      "loss": 0.5621,
+      "step": 34017
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5436419862786868,
+      "learning_rate": 1.9975400539232836e-06,
+      "loss": 0.7009,
+      "step": 34018
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3892475387610035,
+      "learning_rate": 1.9970825029628814e-06,
+      "loss": 0.8394,
+      "step": 34019
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3685960979619871,
+      "learning_rate": 1.996624998598e-06,
+      "loss": 0.7751,
+      "step": 34020
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6454511173184971,
+      "learning_rate": 1.996167540831303e-06,
+      "loss": 0.7308,
+      "step": 34021
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5922551090038412,
+      "learning_rate": 1.995710129665449e-06,
+      "loss": 0.4494,
+      "step": 34022
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.405366145635465,
+      "learning_rate": 1.995252765103105e-06,
+      "loss": 0.6118,
+      "step": 34023
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2228999377771699,
+      "learning_rate": 1.9947954471469365e-06,
+      "loss": 0.7605,
+      "step": 34024
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6709941168061753,
+      "learning_rate": 1.9943381757996027e-06,
+      "loss": 0.737,
+      "step": 34025
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1487307788719672,
+      "learning_rate": 1.9938809510637648e-06,
+      "loss": 0.6799,
+      "step": 34026
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4680145347164149,
+      "learning_rate": 1.993423772942086e-06,
+      "loss": 0.6642,
+      "step": 34027
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5030914415315424,
+      "learning_rate": 1.992966641437232e-06,
+      "loss": 0.7489,
+      "step": 34028
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2204078836673429,
+      "learning_rate": 1.9925095565518605e-06,
+      "loss": 0.7146,
+      "step": 34029
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3442957028200766,
+      "learning_rate": 1.992052518288632e-06,
+      "loss": 0.7025,
+      "step": 34030
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.615588348614717,
+      "learning_rate": 1.9915955266502073e-06,
+      "loss": 0.5075,
+      "step": 34031
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4332608764902546,
+      "learning_rate": 1.9911385816392536e-06,
+      "loss": 0.6656,
+      "step": 34032
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2689925820675076,
+      "learning_rate": 1.990681683258425e-06,
+      "loss": 0.7868,
+      "step": 34033
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3511812185721646,
+      "learning_rate": 1.990224831510381e-06,
+      "loss": 0.6745,
+      "step": 34034
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.838341013833228,
+      "learning_rate": 1.989768026397787e-06,
+      "loss": 0.8049,
+      "step": 34035
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6516064567222706,
+      "learning_rate": 1.989311267923296e-06,
+      "loss": 0.693,
+      "step": 34036
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.539471568628045,
+      "learning_rate": 1.9888545560895746e-06,
+      "loss": 0.6206,
+      "step": 34037
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4234220533365103,
+      "learning_rate": 1.988397890899275e-06,
+      "loss": 0.6927,
+      "step": 34038
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0850326327263047,
+      "learning_rate": 1.987941272355063e-06,
+      "loss": 0.6992,
+      "step": 34039
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3672441957737047,
+      "learning_rate": 1.98748470045959e-06,
+      "loss": 0.547,
+      "step": 34040
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4155435918952497,
+      "learning_rate": 1.9870281752155208e-06,
+      "loss": 0.8362,
+      "step": 34041
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.633596818354297,
+      "learning_rate": 1.98657169662551e-06,
+      "loss": 0.7186,
+      "step": 34042
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5148482240793373,
+      "learning_rate": 1.9861152646922133e-06,
+      "loss": 0.5607,
+      "step": 34043
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8026247302776848,
+      "learning_rate": 1.9856588794182906e-06,
+      "loss": 0.6677,
+      "step": 34044
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5976492388286182,
+      "learning_rate": 1.985202540806401e-06,
+      "loss": 0.5008,
+      "step": 34045
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6217149771550743,
+      "learning_rate": 1.9847462488592017e-06,
+      "loss": 0.694,
+      "step": 34046
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0572851692911454,
+      "learning_rate": 1.9842900035793435e-06,
+      "loss": 0.6517,
+      "step": 34047
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.044650590450024,
+      "learning_rate": 1.983833804969486e-06,
+      "loss": 0.658,
+      "step": 34048
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2964489961494006,
+      "learning_rate": 1.983377653032289e-06,
+      "loss": 0.7542,
+      "step": 34049
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6453637713520597,
+      "learning_rate": 1.9829215477704066e-06,
+      "loss": 0.5221,
+      "step": 34050
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3361708257143141,
+      "learning_rate": 1.9824654891864893e-06,
+      "loss": 0.7881,
+      "step": 34051
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2292751690153405,
+      "learning_rate": 1.982009477283197e-06,
+      "loss": 0.6977,
+      "step": 34052
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.576715772133298,
+      "learning_rate": 1.9815535120631846e-06,
+      "loss": 0.5123,
+      "step": 34053
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3571483043510093,
+      "learning_rate": 1.9810975935291076e-06,
+      "loss": 0.532,
+      "step": 34054
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5097832063208587,
+      "learning_rate": 1.980641721683617e-06,
+      "loss": 0.6309,
+      "step": 34055
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.412621679756516,
+      "learning_rate": 1.980185896529372e-06,
+      "loss": 0.693,
+      "step": 34056
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4278773899263364,
+      "learning_rate": 1.9797301180690197e-06,
+      "loss": 0.7198,
+      "step": 34057
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7074722525474555,
+      "learning_rate": 1.9792743863052223e-06,
+      "loss": 0.5682,
+      "step": 34058
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3945885555475905,
+      "learning_rate": 1.9788187012406247e-06,
+      "loss": 0.7566,
+      "step": 34059
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5460290133930905,
+      "learning_rate": 1.978363062877887e-06,
+      "loss": 0.6984,
+      "step": 34060
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.232476678127019,
+      "learning_rate": 1.9779074712196567e-06,
+      "loss": 0.6596,
+      "step": 34061
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.214494330705356,
+      "learning_rate": 1.9774519262685908e-06,
+      "loss": 0.7198,
+      "step": 34062
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3409810138684481,
+      "learning_rate": 1.97699642802734e-06,
+      "loss": 0.7236,
+      "step": 34063
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6083294841497516,
+      "learning_rate": 1.9765409764985543e-06,
+      "loss": 0.4645,
+      "step": 34064
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5742625149432793,
+      "learning_rate": 1.976085571684887e-06,
+      "loss": 0.4739,
+      "step": 34065
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2641160972609335,
+      "learning_rate": 1.9756302135889894e-06,
+      "loss": 0.6206,
+      "step": 34066
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4759670873829598,
+      "learning_rate": 1.975174902213518e-06,
+      "loss": 0.779,
+      "step": 34067
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3306275733287847,
+      "learning_rate": 1.9747196375611143e-06,
+      "loss": 0.7126,
+      "step": 34068
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5090576324034584,
+      "learning_rate": 1.974264419634433e-06,
+      "loss": 0.6856,
+      "step": 34069
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4984733413751594,
+      "learning_rate": 1.973809248436128e-06,
+      "loss": 0.847,
+      "step": 34070
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.215128337577533,
+      "learning_rate": 1.973354123968847e-06,
+      "loss": 0.5685,
+      "step": 34071
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5862817809448116,
+      "learning_rate": 1.972899046235237e-06,
+      "loss": 0.4816,
+      "step": 34072
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4220803384132916,
+      "learning_rate": 1.97244401523795e-06,
+      "loss": 0.6046,
+      "step": 34073
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7509535539919003,
+      "learning_rate": 1.9719890309796374e-06,
+      "loss": 0.836,
+      "step": 34074
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4282265252892345,
+      "learning_rate": 1.9715340934629446e-06,
+      "loss": 0.6489,
+      "step": 34075
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2663240215518294,
+      "learning_rate": 1.9710792026905243e-06,
+      "loss": 0.581,
+      "step": 34076
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.638886645588218,
+      "learning_rate": 1.97062435866502e-06,
+      "loss": 0.7495,
+      "step": 34077
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3947011259867257,
+      "learning_rate": 1.970169561389086e-06,
+      "loss": 0.4993,
+      "step": 34078
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3142154076649752,
+      "learning_rate": 1.969714810865363e-06,
+      "loss": 0.6224,
+      "step": 34079
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.477322113793642,
+      "learning_rate": 1.969260107096507e-06,
+      "loss": 0.6279,
+      "step": 34080
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5977608459397515,
+      "learning_rate": 1.9688054500851596e-06,
+      "loss": 0.5056,
+      "step": 34081
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5610774019385547,
+      "learning_rate": 1.968350839833968e-06,
+      "loss": 0.5344,
+      "step": 34082
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7687244206915276,
+      "learning_rate": 1.96789627634558e-06,
+      "loss": 0.6995,
+      "step": 34083
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5765472972488537,
+      "learning_rate": 1.967441759622646e-06,
+      "loss": 0.7363,
+      "step": 34084
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6760647088517824,
+      "learning_rate": 1.9669872896678087e-06,
+      "loss": 0.5935,
+      "step": 34085
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4527807994421564,
+      "learning_rate": 1.9665328664837114e-06,
+      "loss": 0.6383,
+      "step": 34086
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4193662888222967,
+      "learning_rate": 1.9660784900730047e-06,
+      "loss": 0.5486,
+      "step": 34087
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.507800333596747,
+      "learning_rate": 1.9656241604383343e-06,
+      "loss": 0.5618,
+      "step": 34088
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.885404962805884,
+      "learning_rate": 1.9651698775823425e-06,
+      "loss": 0.6232,
+      "step": 34089
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2038947856083686,
+      "learning_rate": 1.964715641507674e-06,
+      "loss": 0.6349,
+      "step": 34090
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3957877177056375,
+      "learning_rate": 1.9642614522169754e-06,
+      "loss": 0.6409,
+      "step": 34091
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5760515992278528,
+      "learning_rate": 1.963807309712893e-06,
+      "loss": 0.4966,
+      "step": 34092
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4655008802887832,
+      "learning_rate": 1.9633532139980684e-06,
+      "loss": 0.6159,
+      "step": 34093
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2524185152313576,
+      "learning_rate": 1.962899165075143e-06,
+      "loss": 0.6416,
+      "step": 34094
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5425139744701644,
+      "learning_rate": 1.9624451629467658e-06,
+      "loss": 0.7258,
+      "step": 34095
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.319603804408982,
+      "learning_rate": 1.9619912076155757e-06,
+      "loss": 0.7459,
+      "step": 34096
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.604011664224134,
+      "learning_rate": 1.9615372990842184e-06,
+      "loss": 0.5429,
+      "step": 34097
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3561057117035957,
+      "learning_rate": 1.9610834373553347e-06,
+      "loss": 0.6547,
+      "step": 34098
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3533388746139308,
+      "learning_rate": 1.9606296224315713e-06,
+      "loss": 0.6761,
+      "step": 34099
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3713088988824855,
+      "learning_rate": 1.9601758543155636e-06,
+      "loss": 0.6309,
+      "step": 34100
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.99447241358783,
+      "learning_rate": 1.9597221330099613e-06,
+      "loss": 0.7784,
+      "step": 34101
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4896251583919533,
+      "learning_rate": 1.9592684585174014e-06,
+      "loss": 0.7182,
+      "step": 34102
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.542625753314522,
+      "learning_rate": 1.958814830840523e-06,
+      "loss": 0.6324,
+      "step": 34103
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4907124271363612,
+      "learning_rate": 1.9583612499819713e-06,
+      "loss": 0.6945,
+      "step": 34104
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6090718861931183,
+      "learning_rate": 1.9579077159443893e-06,
+      "loss": 0.4716,
+      "step": 34105
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5042067658396117,
+      "learning_rate": 1.957454228730414e-06,
+      "loss": 0.6263,
+      "step": 34106
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3860565372899867,
+      "learning_rate": 1.9570007883426845e-06,
+      "loss": 0.7772,
+      "step": 34107
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.325874019103304,
+      "learning_rate": 1.956547394783842e-06,
+      "loss": 0.6015,
+      "step": 34108
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2789734170194589,
+      "learning_rate": 1.9560940480565304e-06,
+      "loss": 0.6472,
+      "step": 34109
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4147592143131011,
+      "learning_rate": 1.9556407481633854e-06,
+      "loss": 0.6249,
+      "step": 34110
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5445255512838418,
+      "learning_rate": 1.9551874951070438e-06,
+      "loss": 0.5359,
+      "step": 34111
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5962560930634397,
+      "learning_rate": 1.954734288890149e-06,
+      "loss": 0.7467,
+      "step": 34112
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4510819078343027,
+      "learning_rate": 1.9542811295153396e-06,
+      "loss": 0.6935,
+      "step": 34113
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2300460054165867,
+      "learning_rate": 1.953828016985253e-06,
+      "loss": 0.7478,
+      "step": 34114
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5164016195576588,
+      "learning_rate": 1.9533749513025234e-06,
+      "loss": 0.5723,
+      "step": 34115
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2830242215483305,
+      "learning_rate": 1.9529219324697966e-06,
+      "loss": 0.8069,
+      "step": 34116
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3293224344174341,
+      "learning_rate": 1.9524689604897028e-06,
+      "loss": 0.6749,
+      "step": 34117
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4039789106669405,
+      "learning_rate": 1.9520160353648852e-06,
+      "loss": 0.5835,
+      "step": 34118
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4472313719127847,
+      "learning_rate": 1.951563157097975e-06,
+      "loss": 0.6947,
+      "step": 34119
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2993899654495515,
+      "learning_rate": 1.9511103256916154e-06,
+      "loss": 0.7384,
+      "step": 34120
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5588647095438888,
+      "learning_rate": 1.9506575411484373e-06,
+      "loss": 0.7926,
+      "step": 34121
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4247012878383816,
+      "learning_rate": 1.9502048034710818e-06,
+      "loss": 0.6986,
+      "step": 34122
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3438964648228227,
+      "learning_rate": 1.9497521126621823e-06,
+      "loss": 0.6488,
+      "step": 34123
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3758605228676386,
+      "learning_rate": 1.9492994687243715e-06,
+      "loss": 0.6543,
+      "step": 34124
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5880766292949945,
+      "learning_rate": 1.9488468716602895e-06,
+      "loss": 0.5355,
+      "step": 34125
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5775682494677992,
+      "learning_rate": 1.948394321472571e-06,
+      "loss": 0.7034,
+      "step": 34126
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.412820100700112,
+      "learning_rate": 1.9479418181638508e-06,
+      "loss": 0.5857,
+      "step": 34127
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2459824654245097,
+      "learning_rate": 1.9474893617367597e-06,
+      "loss": 0.6373,
+      "step": 34128
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3209947489741014,
+      "learning_rate": 1.947036952193935e-06,
+      "loss": 0.7525,
+      "step": 34129
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6728424582745833,
+      "learning_rate": 1.946584589538013e-06,
+      "loss": 0.7156,
+      "step": 34130
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3443704821165856,
+      "learning_rate": 1.946132273771625e-06,
+      "loss": 0.5964,
+      "step": 34131
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4875925573646045,
+      "learning_rate": 1.9456800048974023e-06,
+      "loss": 0.5748,
+      "step": 34132
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6094009436827756,
+      "learning_rate": 1.945227782917981e-06,
+      "loss": 0.5466,
+      "step": 34133
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3200557075202504,
+      "learning_rate": 1.944775607835996e-06,
+      "loss": 0.5928,
+      "step": 34134
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5157404545857127,
+      "learning_rate": 1.944323479654078e-06,
+      "loss": 0.738,
+      "step": 34135
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4576546562140704,
+      "learning_rate": 1.9438713983748556e-06,
+      "loss": 0.6324,
+      "step": 34136
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4486828077846874,
+      "learning_rate": 1.943419364000968e-06,
+      "loss": 0.6593,
+      "step": 34137
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4095924270364089,
+      "learning_rate": 1.94296737653504e-06,
+      "loss": 0.5932,
+      "step": 34138
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3445968501955652,
+      "learning_rate": 1.9425154359797103e-06,
+      "loss": 0.5931,
+      "step": 34139
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4968389854674444,
+      "learning_rate": 1.9420635423376033e-06,
+      "loss": 0.8223,
+      "step": 34140
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6986927397387648,
+      "learning_rate": 1.941611695611356e-06,
+      "loss": 0.5983,
+      "step": 34141
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6798163021977188,
+      "learning_rate": 1.941159895803595e-06,
+      "loss": 0.7363,
+      "step": 34142
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4251172375123689,
+      "learning_rate": 1.940708142916955e-06,
+      "loss": 0.5803,
+      "step": 34143
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4163400965368507,
+      "learning_rate": 1.9402564369540634e-06,
+      "loss": 0.8101,
+      "step": 34144
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3556320890279108,
+      "learning_rate": 1.939804777917548e-06,
+      "loss": 0.743,
+      "step": 34145
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2528399292020036,
+      "learning_rate": 1.9393531658100395e-06,
+      "loss": 0.7322,
+      "step": 34146
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6576717774214265,
+      "learning_rate": 1.938901600634173e-06,
+      "loss": 0.5814,
+      "step": 34147
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4338199989253428,
+      "learning_rate": 1.9384500823925732e-06,
+      "loss": 0.6998,
+      "step": 34148
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4066989928407727,
+      "learning_rate": 1.9379986110878666e-06,
+      "loss": 0.8626,
+      "step": 34149
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6078127750350559,
+      "learning_rate": 1.9375471867226846e-06,
+      "loss": 0.4529,
+      "step": 34150
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4706179906881598,
+      "learning_rate": 1.9370958092996582e-06,
+      "loss": 0.6938,
+      "step": 34151
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3472020190770844,
+      "learning_rate": 1.9366444788214123e-06,
+      "loss": 0.6948,
+      "step": 34152
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6294031914123709,
+      "learning_rate": 1.9361931952905723e-06,
+      "loss": 0.5099,
+      "step": 34153
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6076057481257956,
+      "learning_rate": 1.935741958709768e-06,
+      "loss": 0.5096,
+      "step": 34154
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.6227423972356396,
+      "learning_rate": 1.9352907690816313e-06,
+      "loss": 0.6352,
+      "step": 34155
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3967741667858988,
+      "learning_rate": 1.9348396264087845e-06,
+      "loss": 0.6298,
+      "step": 34156
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.6748896124502495,
+      "learning_rate": 1.9343885306938514e-06,
+      "loss": 0.694,
+      "step": 34157
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.437277119456453,
+      "learning_rate": 1.933937481939464e-06,
+      "loss": 0.7259,
+      "step": 34158
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.414312247823353,
+      "learning_rate": 1.9334864801482446e-06,
+      "loss": 0.7605,
+      "step": 34159
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3432593730690148,
+      "learning_rate": 1.9330355253228237e-06,
+      "loss": 0.6344,
+      "step": 34160
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3959305944538578,
+      "learning_rate": 1.93258461746582e-06,
+      "loss": 0.6623,
+      "step": 34161
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5218949178372274,
+      "learning_rate": 1.932133756579867e-06,
+      "loss": 0.5738,
+      "step": 34162
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7328054084429292,
+      "learning_rate": 1.931682942667582e-06,
+      "loss": 0.6017,
+      "step": 34163
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2658009043484155,
+      "learning_rate": 1.931232175731593e-06,
+      "loss": 0.7177,
+      "step": 34164
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5562515738756335,
+      "learning_rate": 1.9307814557745307e-06,
+      "loss": 0.6868,
+      "step": 34165
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1595134061657737,
+      "learning_rate": 1.930330782799008e-06,
+      "loss": 0.4457,
+      "step": 34166
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8899320471525027,
+      "learning_rate": 1.9298801568076543e-06,
+      "loss": 0.7698,
+      "step": 34167
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6390111148301676,
+      "learning_rate": 1.9294295778030924e-06,
+      "loss": 0.5236,
+      "step": 34168
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3751202572824588,
+      "learning_rate": 1.928979045787953e-06,
+      "loss": 0.658,
+      "step": 34169
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3778579887239415,
+      "learning_rate": 1.9285285607648476e-06,
+      "loss": 0.7525,
+      "step": 34170
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4291727151879727,
+      "learning_rate": 1.928078122736403e-06,
+      "loss": 0.664,
+      "step": 34171
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.391387038677917,
+      "learning_rate": 1.927627731705247e-06,
+      "loss": 0.6049,
+      "step": 34172
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.138243295102522,
+      "learning_rate": 1.9271773876739953e-06,
+      "loss": 0.7,
+      "step": 34173
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4380230958610927,
+      "learning_rate": 1.926727090645275e-06,
+      "loss": 0.6344,
+      "step": 34174
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3670403450509727,
+      "learning_rate": 1.9262768406217026e-06,
+      "loss": 0.5772,
+      "step": 34175
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.342645433187616,
+      "learning_rate": 1.9258266376059053e-06,
+      "loss": 0.7021,
+      "step": 34176
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4820273670947894,
+      "learning_rate": 1.9253764816004993e-06,
+      "loss": 0.5732,
+      "step": 34177
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4602457699908258,
+      "learning_rate": 1.9249263726081103e-06,
+      "loss": 0.7506,
+      "step": 34178
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6374406747586747,
+      "learning_rate": 1.924476310631357e-06,
+      "loss": 0.6126,
+      "step": 34179
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.803957049538264,
+      "learning_rate": 1.924026295672856e-06,
+      "loss": 0.784,
+      "step": 34180
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5563002802530235,
+      "learning_rate": 1.9235763277352304e-06,
+      "loss": 0.6772,
+      "step": 34181
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2779790016000936,
+      "learning_rate": 1.9231264068211033e-06,
+      "loss": 0.5831,
+      "step": 34182
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.58889259638689,
+      "learning_rate": 1.922676532933092e-06,
+      "loss": 0.6701,
+      "step": 34183
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3650318687837595,
+      "learning_rate": 1.922226706073812e-06,
+      "loss": 0.634,
+      "step": 34184
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2517036795561867,
+      "learning_rate": 1.921776926245885e-06,
+      "loss": 0.6139,
+      "step": 34185
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.336527147441069,
+      "learning_rate": 1.921327193451934e-06,
+      "loss": 0.5975,
+      "step": 34186
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3409028266442886,
+      "learning_rate": 1.920877507694573e-06,
+      "loss": 0.6249,
+      "step": 34187
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4086503141517397,
+      "learning_rate": 1.920427868976418e-06,
+      "loss": 0.6941,
+      "step": 34188
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2838980167248908,
+      "learning_rate": 1.9199782773000908e-06,
+      "loss": 0.6889,
+      "step": 34189
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3399185280331194,
+      "learning_rate": 1.91952873266821e-06,
+      "loss": 0.8347,
+      "step": 34190
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1963386074214222,
+      "learning_rate": 1.919079235083391e-06,
+      "loss": 0.6619,
+      "step": 34191
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.880850602791981,
+      "learning_rate": 1.9186297845482495e-06,
+      "loss": 0.6968,
+      "step": 34192
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.636408491131373,
+      "learning_rate": 1.9181803810654053e-06,
+      "loss": 0.58,
+      "step": 34193
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5874773491523565,
+      "learning_rate": 1.9177310246374725e-06,
+      "loss": 0.481,
+      "step": 34194
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4499797551266211,
+      "learning_rate": 1.9172817152670696e-06,
+      "loss": 0.6447,
+      "step": 34195
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6267410590345905,
+      "learning_rate": 1.9168324529568093e-06,
+      "loss": 0.6335,
+      "step": 34196
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.131001813892047,
+      "learning_rate": 1.9163832377093126e-06,
+      "loss": 0.7543,
+      "step": 34197
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4863117192205095,
+      "learning_rate": 1.915934069527189e-06,
+      "loss": 0.6814,
+      "step": 34198
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5549291400354757,
+      "learning_rate": 1.91548494841306e-06,
+      "loss": 0.6106,
+      "step": 34199
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2850313548420937,
+      "learning_rate": 1.915035874369533e-06,
+      "loss": 0.6315,
+      "step": 34200
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6909844931010987,
+      "learning_rate": 1.91458684739923e-06,
+      "loss": 0.6178,
+      "step": 34201
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3182620094359787,
+      "learning_rate": 1.9141378675047608e-06,
+      "loss": 0.4768,
+      "step": 34202
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5692106297180748,
+      "learning_rate": 1.9136889346887423e-06,
+      "loss": 0.4851,
+      "step": 34203
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5534281940743282,
+      "learning_rate": 1.9132400489537872e-06,
+      "loss": 0.855,
+      "step": 34204
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.212936441186098,
+      "learning_rate": 1.9127912103025058e-06,
+      "loss": 0.6539,
+      "step": 34205
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5092732137653166,
+      "learning_rate": 1.912342418737515e-06,
+      "loss": 0.6018,
+      "step": 34206
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.478438963585249,
+      "learning_rate": 1.911893674261429e-06,
+      "loss": 0.794,
+      "step": 34207
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4543708592693905,
+      "learning_rate": 1.9114449768768584e-06,
+      "loss": 0.7163,
+      "step": 34208
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6614707454151094,
+      "learning_rate": 1.910996326586414e-06,
+      "loss": 0.5113,
+      "step": 34209
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2858758433003357,
+      "learning_rate": 1.9105477233927105e-06,
+      "loss": 0.695,
+      "step": 34210
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5818675948280067,
+      "learning_rate": 1.9100991672983616e-06,
+      "loss": 0.4473,
+      "step": 34211
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1077815966396614,
+      "learning_rate": 1.909650658305976e-06,
+      "loss": 0.5615,
+      "step": 34212
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3613140060284488,
+      "learning_rate": 1.909202196418163e-06,
+      "loss": 0.7669,
+      "step": 34213
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5166562587829395,
+      "learning_rate": 1.908753781637537e-06,
+      "loss": 0.7633,
+      "step": 34214
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2337662652573975,
+      "learning_rate": 1.9083054139667103e-06,
+      "loss": 0.5717,
+      "step": 34215
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1492458034523605,
+      "learning_rate": 1.9078570934082907e-06,
+      "loss": 0.6328,
+      "step": 34216
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.399207477632584,
+      "learning_rate": 1.9074088199648866e-06,
+      "loss": 0.7228,
+      "step": 34217
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7366453094175827,
+      "learning_rate": 1.9069605936391133e-06,
+      "loss": 0.5848,
+      "step": 34218
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4824800108091625,
+      "learning_rate": 1.9065124144335756e-06,
+      "loss": 0.6841,
+      "step": 34219
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6516752849877379,
+      "learning_rate": 1.9060642823508869e-06,
+      "loss": 0.6581,
+      "step": 34220
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5979472463148648,
+      "learning_rate": 1.9056161973936516e-06,
+      "loss": 0.5021,
+      "step": 34221
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3807243721474527,
+      "learning_rate": 1.9051681595644844e-06,
+      "loss": 0.6848,
+      "step": 34222
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5132262790941202,
+      "learning_rate": 1.904720168865989e-06,
+      "loss": 0.6916,
+      "step": 34223
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.504654651579738,
+      "learning_rate": 1.9042722253007772e-06,
+      "loss": 0.6705,
+      "step": 34224
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5093933462830043,
+      "learning_rate": 1.9038243288714563e-06,
+      "loss": 0.6962,
+      "step": 34225
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6269116575035675,
+      "learning_rate": 1.903376479580631e-06,
+      "loss": 0.4828,
+      "step": 34226
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3792767892439255,
+      "learning_rate": 1.9029286774309108e-06,
+      "loss": 0.6841,
+      "step": 34227
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5675782428819663,
+      "learning_rate": 1.9024809224249052e-06,
+      "loss": 0.5012,
+      "step": 34228
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7946809361575475,
+      "learning_rate": 1.9020332145652197e-06,
+      "loss": 0.6747,
+      "step": 34229
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2414724264719132,
+      "learning_rate": 1.9015855538544592e-06,
+      "loss": 0.6761,
+      "step": 34230
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3364123918389956,
+      "learning_rate": 1.90113794029523e-06,
+      "loss": 0.6714,
+      "step": 34231
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.9829103691934153,
+      "learning_rate": 1.9006903738901427e-06,
+      "loss": 0.8864,
+      "step": 34232
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5890422113609463,
+      "learning_rate": 1.9002428546418005e-06,
+      "loss": 0.5295,
+      "step": 34233
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3760273480409035,
+      "learning_rate": 1.8997953825528071e-06,
+      "loss": 0.6177,
+      "step": 34234
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3461431614357444,
+      "learning_rate": 1.8993479576257678e-06,
+      "loss": 0.5217,
+      "step": 34235
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.463843872663284,
+      "learning_rate": 1.898900579863293e-06,
+      "loss": 0.5906,
+      "step": 34236
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4386300250453106,
+      "learning_rate": 1.898453249267983e-06,
+      "loss": 0.7265,
+      "step": 34237
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2765885782348614,
+      "learning_rate": 1.898005965842441e-06,
+      "loss": 0.6568,
+      "step": 34238
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3289641573191064,
+      "learning_rate": 1.8975587295892761e-06,
+      "loss": 0.6156,
+      "step": 34239
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4619477253784388,
+      "learning_rate": 1.8971115405110863e-06,
+      "loss": 0.657,
+      "step": 34240
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9612883808760553,
+      "learning_rate": 1.8966643986104816e-06,
+      "loss": 0.7965,
+      "step": 34241
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5878741423148349,
+      "learning_rate": 1.896217303890059e-06,
+      "loss": 0.4514,
+      "step": 34242
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6235222984690441,
+      "learning_rate": 1.8957702563524282e-06,
+      "loss": 0.4809,
+      "step": 34243
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5182393590501404,
+      "learning_rate": 1.8953232560001855e-06,
+      "loss": 0.7625,
+      "step": 34244
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6265583201806152,
+      "learning_rate": 1.8948763028359396e-06,
+      "loss": 0.5049,
+      "step": 34245
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4387361215998082,
+      "learning_rate": 1.8944293968622896e-06,
+      "loss": 0.5127,
+      "step": 34246
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.844361978643529,
+      "learning_rate": 1.8939825380818354e-06,
+      "loss": 0.6204,
+      "step": 34247
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4941906510031622,
+      "learning_rate": 1.893535726497182e-06,
+      "loss": 0.5332,
+      "step": 34248
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4960181466496287,
+      "learning_rate": 1.8930889621109295e-06,
+      "loss": 0.6575,
+      "step": 34249
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5404455608232153,
+      "learning_rate": 1.8926422449256842e-06,
+      "loss": 0.8243,
+      "step": 34250
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3146340720477194,
+      "learning_rate": 1.892195574944038e-06,
+      "loss": 0.6004,
+      "step": 34251
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2172401662367824,
+      "learning_rate": 1.8917489521685961e-06,
+      "loss": 0.737,
+      "step": 34252
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1330572735516,
+      "learning_rate": 1.891302376601961e-06,
+      "loss": 0.774,
+      "step": 34253
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6432554218549369,
+      "learning_rate": 1.890855848246731e-06,
+      "loss": 0.5761,
+      "step": 34254
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5000119741508726,
+      "learning_rate": 1.8904093671055023e-06,
+      "loss": 0.7608,
+      "step": 34255
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3567723150482691,
+      "learning_rate": 1.8899629331808788e-06,
+      "loss": 0.6202,
+      "step": 34256
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2652195774315076,
+      "learning_rate": 1.8895165464754605e-06,
+      "loss": 0.5159,
+      "step": 34257
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2069762266662802,
+      "learning_rate": 1.8890702069918442e-06,
+      "loss": 0.7552,
+      "step": 34258
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.775453518707973,
+      "learning_rate": 1.888623914732628e-06,
+      "loss": 0.7158,
+      "step": 34259
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2978347198290985,
+      "learning_rate": 1.8881776697004128e-06,
+      "loss": 0.5928,
+      "step": 34260
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5806238151418208,
+      "learning_rate": 1.8877314718977923e-06,
+      "loss": 0.5048,
+      "step": 34261
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3444908385505163,
+      "learning_rate": 1.887285321327369e-06,
+      "loss": 0.6933,
+      "step": 34262
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5824369310176192,
+      "learning_rate": 1.8868392179917404e-06,
+      "loss": 0.4939,
+      "step": 34263
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4444430084698154,
+      "learning_rate": 1.8863931618935028e-06,
+      "loss": 0.8388,
+      "step": 34264
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.140650533898347,
+      "learning_rate": 1.8859471530352502e-06,
+      "loss": 0.7366,
+      "step": 34265
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4584856244469107,
+      "learning_rate": 1.885501191419582e-06,
+      "loss": 0.7737,
+      "step": 34266
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3189214769198525,
+      "learning_rate": 1.8850552770491004e-06,
+      "loss": 0.7058,
+      "step": 34267
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.789968959813544,
+      "learning_rate": 1.8846094099263911e-06,
+      "loss": 0.6287,
+      "step": 34268
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4178900324925958,
+      "learning_rate": 1.8841635900540544e-06,
+      "loss": 0.7286,
+      "step": 34269
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0288804102875386,
+      "learning_rate": 1.8837178174346882e-06,
+      "loss": 0.762,
+      "step": 34270
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7224091845544331,
+      "learning_rate": 1.883272092070888e-06,
+      "loss": 0.7473,
+      "step": 34271
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3771746969512437,
+      "learning_rate": 1.8828264139652474e-06,
+      "loss": 0.6345,
+      "step": 34272
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3142240883465786,
+      "learning_rate": 1.882380783120359e-06,
+      "loss": 0.6994,
+      "step": 34273
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4029388430949834,
+      "learning_rate": 1.8819351995388214e-06,
+      "loss": 0.6164,
+      "step": 34274
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3400846456590756,
+      "learning_rate": 1.881489663223225e-06,
+      "loss": 0.6767,
+      "step": 34275
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2473559938373546,
+      "learning_rate": 1.8810441741761687e-06,
+      "loss": 0.5647,
+      "step": 34276
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.580105404710796,
+      "learning_rate": 1.8805987324002418e-06,
+      "loss": 0.5157,
+      "step": 34277
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.695704709176342,
+      "learning_rate": 1.8801533378980407e-06,
+      "loss": 0.6719,
+      "step": 34278
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5582499880299641,
+      "learning_rate": 1.879707990672156e-06,
+      "loss": 0.5005,
+      "step": 34279
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3056989398681564,
+      "learning_rate": 1.8792626907251855e-06,
+      "loss": 0.6654,
+      "step": 34280
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2741455292714967,
+      "learning_rate": 1.878817438059718e-06,
+      "loss": 0.6077,
+      "step": 34281
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6032969247007982,
+      "learning_rate": 1.8783722326783427e-06,
+      "loss": 0.7631,
+      "step": 34282
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6351929032465744,
+      "learning_rate": 1.8779270745836576e-06,
+      "loss": 0.5518,
+      "step": 34283
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.428918696485712,
+      "learning_rate": 1.8774819637782538e-06,
+      "loss": 0.6687,
+      "step": 34284
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4650222317918618,
+      "learning_rate": 1.8770369002647215e-06,
+      "loss": 0.7069,
+      "step": 34285
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4833603865486156,
+      "learning_rate": 1.87659188404565e-06,
+      "loss": 0.6933,
+      "step": 34286
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4581040156941438,
+      "learning_rate": 1.8761469151236322e-06,
+      "loss": 0.6187,
+      "step": 34287
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2930592354069201,
+      "learning_rate": 1.875701993501261e-06,
+      "loss": 0.7696,
+      "step": 34288
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5178908004837766,
+      "learning_rate": 1.8752571191811253e-06,
+      "loss": 0.6348,
+      "step": 34289
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5252551678977955,
+      "learning_rate": 1.874812292165813e-06,
+      "loss": 0.725,
+      "step": 34290
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.505568579017704,
+      "learning_rate": 1.8743675124579152e-06,
+      "loss": 0.6719,
+      "step": 34291
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3063275597155009,
+      "learning_rate": 1.8739227800600246e-06,
+      "loss": 0.5217,
+      "step": 34292
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3550295251909903,
+      "learning_rate": 1.873478094974729e-06,
+      "loss": 0.8203,
+      "step": 34293
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6151739490664865,
+      "learning_rate": 1.8730334572046138e-06,
+      "loss": 0.7404,
+      "step": 34294
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4212743115753939,
+      "learning_rate": 1.872588866752273e-06,
+      "loss": 0.7773,
+      "step": 34295
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4089017541477835,
+      "learning_rate": 1.8721443236202907e-06,
+      "loss": 0.6936,
+      "step": 34296
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4073794972056821,
+      "learning_rate": 1.8716998278112597e-06,
+      "loss": 0.7264,
+      "step": 34297
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.535271586825893,
+      "learning_rate": 1.8712553793277643e-06,
+      "loss": 0.6713,
+      "step": 34298
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2178365675328973,
+      "learning_rate": 1.8708109781723949e-06,
+      "loss": 0.639,
+      "step": 34299
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3359779314887623,
+      "learning_rate": 1.8703666243477347e-06,
+      "loss": 0.6706,
+      "step": 34300
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.568784985834378,
+      "learning_rate": 1.8699223178563775e-06,
+      "loss": 0.7973,
+      "step": 34301
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4225074158820548,
+      "learning_rate": 1.8694780587009065e-06,
+      "loss": 0.699,
+      "step": 34302
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3240943198536013,
+      "learning_rate": 1.8690338468839053e-06,
+      "loss": 0.7153,
+      "step": 34303
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.490864136466903,
+      "learning_rate": 1.868589682407963e-06,
+      "loss": 0.7339,
+      "step": 34304
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4496637806595096,
+      "learning_rate": 1.8681455652756685e-06,
+      "loss": 0.6777,
+      "step": 34305
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.630179107940581,
+      "learning_rate": 1.8677014954896055e-06,
+      "loss": 0.4986,
+      "step": 34306
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3813532694622983,
+      "learning_rate": 1.8672574730523563e-06,
+      "loss": 0.6888,
+      "step": 34307
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3598342406794943,
+      "learning_rate": 1.8668134979665086e-06,
+      "loss": 0.6239,
+      "step": 34308
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.479847571543341,
+      "learning_rate": 1.8663695702346508e-06,
+      "loss": 0.7665,
+      "step": 34309
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.854106431064786,
+      "learning_rate": 1.8659256898593648e-06,
+      "loss": 0.6544,
+      "step": 34310
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.352839425680996,
+      "learning_rate": 1.8654818568432308e-06,
+      "loss": 0.5369,
+      "step": 34311
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3692875781312415,
+      "learning_rate": 1.8650380711888372e-06,
+      "loss": 0.6962,
+      "step": 34312
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5923417975357584,
+      "learning_rate": 1.8645943328987704e-06,
+      "loss": 0.5127,
+      "step": 34313
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3990317677123445,
+      "learning_rate": 1.8641506419756105e-06,
+      "loss": 0.6702,
+      "step": 34314
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4992449851155385,
+      "learning_rate": 1.8637069984219391e-06,
+      "loss": 0.685,
+      "step": 34315
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9021235837099986,
+      "learning_rate": 1.863263402240344e-06,
+      "loss": 0.7322,
+      "step": 34316
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5219333971453328,
+      "learning_rate": 1.862819853433403e-06,
+      "loss": 0.5883,
+      "step": 34317
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3239881203590025,
+      "learning_rate": 1.862376352003703e-06,
+      "loss": 0.6333,
+      "step": 34318
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5248584322060863,
+      "learning_rate": 1.8619328979538221e-06,
+      "loss": 0.7035,
+      "step": 34319
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6515409561751009,
+      "learning_rate": 1.8614894912863469e-06,
+      "loss": 0.7054,
+      "step": 34320
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.471674152256989,
+      "learning_rate": 1.8610461320038541e-06,
+      "loss": 0.8304,
+      "step": 34321
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5607770461842099,
+      "learning_rate": 1.8606028201089298e-06,
+      "loss": 0.8369,
+      "step": 34322
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7558237793785094,
+      "learning_rate": 1.860159555604153e-06,
+      "loss": 0.7716,
+      "step": 34323
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6941766644620548,
+      "learning_rate": 1.8597163384921013e-06,
+      "loss": 0.6722,
+      "step": 34324
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5471256425967574,
+      "learning_rate": 1.8592731687753585e-06,
+      "loss": 0.7312,
+      "step": 34325
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3110736188627614,
+      "learning_rate": 1.8588300464565068e-06,
+      "loss": 0.6921,
+      "step": 34326
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2044352794367807,
+      "learning_rate": 1.8583869715381242e-06,
+      "loss": 0.6721,
+      "step": 34327
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3966351475949212,
+      "learning_rate": 1.8579439440227886e-06,
+      "loss": 0.7687,
+      "step": 34328
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2332787385332693,
+      "learning_rate": 1.8575009639130803e-06,
+      "loss": 0.751,
+      "step": 34329
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6243782319752632,
+      "learning_rate": 1.8570580312115815e-06,
+      "loss": 0.5232,
+      "step": 34330
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7514988463908838,
+      "learning_rate": 1.8566151459208692e-06,
+      "loss": 0.655,
+      "step": 34331
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3544396851148655,
+      "learning_rate": 1.856172308043519e-06,
+      "loss": 0.6278,
+      "step": 34332
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3677861450325028,
+      "learning_rate": 1.855729517582112e-06,
+      "loss": 0.6803,
+      "step": 34333
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4226474923257602,
+      "learning_rate": 1.855286774539229e-06,
+      "loss": 0.674,
+      "step": 34334
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6807186350085377,
+      "learning_rate": 1.854844078917445e-06,
+      "loss": 0.7417,
+      "step": 34335
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4267595511513338,
+      "learning_rate": 1.8544014307193348e-06,
+      "loss": 0.6093,
+      "step": 34336
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.8699207037672743,
+      "learning_rate": 1.8539588299474775e-06,
+      "loss": 0.6553,
+      "step": 34337
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3034233284331178,
+      "learning_rate": 1.8535162766044544e-06,
+      "loss": 0.6955,
+      "step": 34338
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9008175028666574,
+      "learning_rate": 1.853073770692838e-06,
+      "loss": 0.8022,
+      "step": 34339
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6107705993977253,
+      "learning_rate": 1.852631312215204e-06,
+      "loss": 0.8324,
+      "step": 34340
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3304423547146018,
+      "learning_rate": 1.8521889011741312e-06,
+      "loss": 0.6222,
+      "step": 34341
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4256335208078772,
+      "learning_rate": 1.851746537572192e-06,
+      "loss": 0.8129,
+      "step": 34342
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2990471994568962,
+      "learning_rate": 1.8513042214119669e-06,
+      "loss": 0.734,
+      "step": 34343
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4033590442520352,
+      "learning_rate": 1.8508619526960258e-06,
+      "loss": 0.7542,
+      "step": 34344
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2009114473165772,
+      "learning_rate": 1.850419731426949e-06,
+      "loss": 0.6026,
+      "step": 34345
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3503977309200688,
+      "learning_rate": 1.8499775576073064e-06,
+      "loss": 0.7241,
+      "step": 34346
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4370802727245586,
+      "learning_rate": 1.849535431239674e-06,
+      "loss": 0.7341,
+      "step": 34347
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6644917771135144,
+      "learning_rate": 1.8490933523266318e-06,
+      "loss": 0.6472,
+      "step": 34348
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3577502034822622,
+      "learning_rate": 1.8486513208707446e-06,
+      "loss": 0.6512,
+      "step": 34349
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5329918571754646,
+      "learning_rate": 1.848209336874589e-06,
+      "loss": 0.8139,
+      "step": 34350
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5319396610295495,
+      "learning_rate": 1.8477674003407398e-06,
+      "loss": 0.7661,
+      "step": 34351
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5297804945908349,
+      "learning_rate": 1.8473255112717747e-06,
+      "loss": 0.6081,
+      "step": 34352
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2994097384741639,
+      "learning_rate": 1.8468836696702574e-06,
+      "loss": 0.717,
+      "step": 34353
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0132430530878467,
+      "learning_rate": 1.8464418755387635e-06,
+      "loss": 0.5927,
+      "step": 34354
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3401203130746253,
+      "learning_rate": 1.846000128879869e-06,
+      "loss": 0.6823,
+      "step": 34355
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.334913125499615,
+      "learning_rate": 1.8455584296961415e-06,
+      "loss": 0.6912,
+      "step": 34356
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1924909895777205,
+      "learning_rate": 1.8451167779901568e-06,
+      "loss": 0.5615,
+      "step": 34357
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1676618506397074,
+      "learning_rate": 1.8446751737644808e-06,
+      "loss": 0.6119,
+      "step": 34358
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.264782125147952,
+      "learning_rate": 1.8442336170216912e-06,
+      "loss": 0.6918,
+      "step": 34359
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7813536283773832,
+      "learning_rate": 1.8437921077643516e-06,
+      "loss": 0.7181,
+      "step": 34360
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.303889696551636,
+      "learning_rate": 1.8433506459950412e-06,
+      "loss": 0.6629,
+      "step": 34361
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3817371599744996,
+      "learning_rate": 1.8429092317163244e-06,
+      "loss": 0.4501,
+      "step": 34362
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4417028590989827,
+      "learning_rate": 1.8424678649307702e-06,
+      "loss": 0.7861,
+      "step": 34363
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1844531424249132,
+      "learning_rate": 1.8420265456409502e-06,
+      "loss": 0.5935,
+      "step": 34364
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8737973404149164,
+      "learning_rate": 1.8415852738494378e-06,
+      "loss": 0.7574,
+      "step": 34365
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3660702562760263,
+      "learning_rate": 1.8411440495587985e-06,
+      "loss": 0.8419,
+      "step": 34366
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5179911840339637,
+      "learning_rate": 1.8407028727715992e-06,
+      "loss": 0.64,
+      "step": 34367
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6599522377372761,
+      "learning_rate": 1.8402617434904103e-06,
+      "loss": 0.5339,
+      "step": 34368
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5912822983006505,
+      "learning_rate": 1.839820661717804e-06,
+      "loss": 0.5355,
+      "step": 34369
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5754839596487168,
+      "learning_rate": 1.8393796274563458e-06,
+      "loss": 0.4978,
+      "step": 34370
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2536748396553068,
+      "learning_rate": 1.8389386407085997e-06,
+      "loss": 0.7601,
+      "step": 34371
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2487286289974684,
+      "learning_rate": 1.8384977014771365e-06,
+      "loss": 0.5899,
+      "step": 34372
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5370475918967552,
+      "learning_rate": 1.8380568097645268e-06,
+      "loss": 0.6141,
+      "step": 34373
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.407922149535299,
+      "learning_rate": 1.837615965573334e-06,
+      "loss": 0.6727,
+      "step": 34374
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3187487150734802,
+      "learning_rate": 1.8371751689061223e-06,
+      "loss": 0.581,
+      "step": 34375
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2499475760102505,
+      "learning_rate": 1.8367344197654647e-06,
+      "loss": 0.7904,
+      "step": 34376
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2654467439878254,
+      "learning_rate": 1.8362937181539198e-06,
+      "loss": 0.5361,
+      "step": 34377
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3948470523676315,
+      "learning_rate": 1.8358530640740612e-06,
+      "loss": 0.6755,
+      "step": 34378
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2572935825499603,
+      "learning_rate": 1.8354124575284471e-06,
+      "loss": 0.6505,
+      "step": 34379
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.45217601943774,
+      "learning_rate": 1.8349718985196496e-06,
+      "loss": 0.61,
+      "step": 34380
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4742320353985094,
+      "learning_rate": 1.834531387050229e-06,
+      "loss": 0.5485,
+      "step": 34381
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.605503873589816,
+      "learning_rate": 1.8340909231227532e-06,
+      "loss": 0.503,
+      "step": 34382
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4255690488485038,
+      "learning_rate": 1.8336505067397858e-06,
+      "loss": 0.7476,
+      "step": 34383
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4403385338995989,
+      "learning_rate": 1.833210137903888e-06,
+      "loss": 0.805,
+      "step": 34384
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3600519847022814,
+      "learning_rate": 1.8327698166176266e-06,
+      "loss": 0.7185,
+      "step": 34385
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3462430179970208,
+      "learning_rate": 1.8323295428835674e-06,
+      "loss": 0.6398,
+      "step": 34386
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3523171121556847,
+      "learning_rate": 1.8318893167042717e-06,
+      "loss": 0.7861,
+      "step": 34387
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4097771169805353,
+      "learning_rate": 1.8314491380822985e-06,
+      "loss": 0.7662,
+      "step": 34388
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5093556916074151,
+      "learning_rate": 1.8310090070202158e-06,
+      "loss": 0.7011,
+      "step": 34389
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5553595564574736,
+      "learning_rate": 1.830568923520587e-06,
+      "loss": 0.6513,
+      "step": 34390
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3570643449416284,
+      "learning_rate": 1.8301288875859724e-06,
+      "loss": 0.6459,
+      "step": 34391
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3807552711948354,
+      "learning_rate": 1.829688899218932e-06,
+      "loss": 0.6896,
+      "step": 34392
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.632357143690301,
+      "learning_rate": 1.8292489584220286e-06,
+      "loss": 0.6555,
+      "step": 34393
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.336203531403163,
+      "learning_rate": 1.8288090651978274e-06,
+      "loss": 0.6286,
+      "step": 34394
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7191007941606486,
+      "learning_rate": 1.8283692195488877e-06,
+      "loss": 0.6727,
+      "step": 34395
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5767355709784286,
+      "learning_rate": 1.8279294214777653e-06,
+      "loss": 0.4983,
+      "step": 34396
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.198977406202269,
+      "learning_rate": 1.8274896709870294e-06,
+      "loss": 0.5594,
+      "step": 34397
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3157066597985765,
+      "learning_rate": 1.827049968079232e-06,
+      "loss": 0.6707,
+      "step": 34398
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2967804313930542,
+      "learning_rate": 1.8266103127569401e-06,
+      "loss": 0.7001,
+      "step": 34399
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4829395013556146,
+      "learning_rate": 1.8261707050227084e-06,
+      "loss": 0.6267,
+      "step": 34400
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3005164741548831,
+      "learning_rate": 1.8257311448791015e-06,
+      "loss": 0.6054,
+      "step": 34401
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2451854186027518,
+      "learning_rate": 1.825291632328673e-06,
+      "loss": 0.7165,
+      "step": 34402
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7185431945009413,
+      "learning_rate": 1.8248521673739872e-06,
+      "loss": 0.6843,
+      "step": 34403
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9478340372041787,
+      "learning_rate": 1.8244127500176e-06,
+      "loss": 0.577,
+      "step": 34404
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5462605567593253,
+      "learning_rate": 1.8239733802620674e-06,
+      "loss": 0.8295,
+      "step": 34405
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.441830201355755,
+      "learning_rate": 1.8235340581099503e-06,
+      "loss": 0.7292,
+      "step": 34406
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.35518195184664,
+      "learning_rate": 1.823094783563809e-06,
+      "loss": 0.7915,
+      "step": 34407
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4368052657742048,
+      "learning_rate": 1.8226555566261995e-06,
+      "loss": 0.7111,
+      "step": 34408
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.253536421776821,
+      "learning_rate": 1.8222163772996749e-06,
+      "loss": 0.607,
+      "step": 34409
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3426997852499951,
+      "learning_rate": 1.8217772455867965e-06,
+      "loss": 0.6827,
+      "step": 34410
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4991813928117637,
+      "learning_rate": 1.8213381614901215e-06,
+      "loss": 0.6551,
+      "step": 34411
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5988791369798085,
+      "learning_rate": 1.8208991250122054e-06,
+      "loss": 0.649,
+      "step": 34412
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2953149721616222,
+      "learning_rate": 1.8204601361556019e-06,
+      "loss": 0.6917,
+      "step": 34413
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.482393376166932,
+      "learning_rate": 1.8200211949228686e-06,
+      "loss": 0.7105,
+      "step": 34414
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.166777623898841,
+      "learning_rate": 1.8195823013165648e-06,
+      "loss": 0.7784,
+      "step": 34415
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6717535775348067,
+      "learning_rate": 1.819143455339243e-06,
+      "loss": 0.6672,
+      "step": 34416
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2474297870153968,
+      "learning_rate": 1.818704656993454e-06,
+      "loss": 0.6336,
+      "step": 34417
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2951548924404404,
+      "learning_rate": 1.8182659062817609e-06,
+      "loss": 0.7147,
+      "step": 34418
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4090841964980756,
+      "learning_rate": 1.8178272032067102e-06,
+      "loss": 0.7141,
+      "step": 34419
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3826608521224668,
+      "learning_rate": 1.8173885477708631e-06,
+      "loss": 0.8045,
+      "step": 34420
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5832683364752538,
+      "learning_rate": 1.8169499399767676e-06,
+      "loss": 0.5086,
+      "step": 34421
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2784228684739363,
+      "learning_rate": 1.8165113798269828e-06,
+      "loss": 0.6953,
+      "step": 34422
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.383973477277799,
+      "learning_rate": 1.8160728673240568e-06,
+      "loss": 0.6128,
+      "step": 34423
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4225425992166398,
+      "learning_rate": 1.8156344024705485e-06,
+      "loss": 0.7692,
+      "step": 34424
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6369297234555322,
+      "learning_rate": 1.8151959852690083e-06,
+      "loss": 0.756,
+      "step": 34425
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3455300477286443,
+      "learning_rate": 1.814757615721985e-06,
+      "loss": 0.6298,
+      "step": 34426
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.345228540600471,
+      "learning_rate": 1.8143192938320353e-06,
+      "loss": 0.6747,
+      "step": 34427
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2190272280059191,
+      "learning_rate": 1.813881019601712e-06,
+      "loss": 0.6094,
+      "step": 34428
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3781464819606724,
+      "learning_rate": 1.8134427930335651e-06,
+      "loss": 0.7132,
+      "step": 34429
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.166205205548499,
+      "learning_rate": 1.8130046141301426e-06,
+      "loss": 0.5192,
+      "step": 34430
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.312660098866419,
+      "learning_rate": 1.8125664828940005e-06,
+      "loss": 0.665,
+      "step": 34431
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7455356784020626,
+      "learning_rate": 1.8121283993276906e-06,
+      "loss": 0.6976,
+      "step": 34432
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.670167972278536,
+      "learning_rate": 1.8116903634337612e-06,
+      "loss": 0.6344,
+      "step": 34433
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5756944986747856,
+      "learning_rate": 1.8112523752147615e-06,
+      "loss": 0.6545,
+      "step": 34434
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5032670142670335,
+      "learning_rate": 1.8108144346732414e-06,
+      "loss": 0.6757,
+      "step": 34435
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9209156068639155,
+      "learning_rate": 1.810376541811756e-06,
+      "loss": 0.7112,
+      "step": 34436
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6073801862429842,
+      "learning_rate": 1.8099386966328503e-06,
+      "loss": 0.5457,
+      "step": 34437
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6440041940816705,
+      "learning_rate": 1.8095008991390728e-06,
+      "loss": 0.6937,
+      "step": 34438
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4737296051017077,
+      "learning_rate": 1.8090631493329759e-06,
+      "loss": 0.5993,
+      "step": 34439
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5815254130529177,
+      "learning_rate": 1.8086254472171039e-06,
+      "loss": 0.5163,
+      "step": 34440
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4153366882708973,
+      "learning_rate": 1.8081877927940107e-06,
+      "loss": 0.8227,
+      "step": 34441
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2714057000427963,
+      "learning_rate": 1.8077501860662395e-06,
+      "loss": 0.6151,
+      "step": 34442
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6047260223100837,
+      "learning_rate": 1.8073126270363428e-06,
+      "loss": 0.6717,
+      "step": 34443
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6813636001354233,
+      "learning_rate": 1.8068751157068632e-06,
+      "loss": 0.539,
+      "step": 34444
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6093498993725144,
+      "learning_rate": 1.8064376520803495e-06,
+      "loss": 0.5016,
+      "step": 34445
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3482499403612695,
+      "learning_rate": 1.8060002361593565e-06,
+      "loss": 0.6745,
+      "step": 34446
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6101050790072584,
+      "learning_rate": 1.8055628679464177e-06,
+      "loss": 0.4746,
+      "step": 34447
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5942615437234946,
+      "learning_rate": 1.8051255474440876e-06,
+      "loss": 0.4645,
+      "step": 34448
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3779470881762321,
+      "learning_rate": 1.8046882746549099e-06,
+      "loss": 0.6956,
+      "step": 34449
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6745194020202362,
+      "learning_rate": 1.8042510495814369e-06,
+      "loss": 0.5142,
+      "step": 34450
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6168892856194745,
+      "learning_rate": 1.8038138722262045e-06,
+      "loss": 0.715,
+      "step": 34451
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3022918485176882,
+      "learning_rate": 1.8033767425917625e-06,
+      "loss": 0.5592,
+      "step": 34452
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5367105115535338,
+      "learning_rate": 1.802939660680656e-06,
+      "loss": 0.6901,
+      "step": 34453
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5874701977618919,
+      "learning_rate": 1.8025026264954326e-06,
+      "loss": 0.4817,
+      "step": 34454
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5951625353421384,
+      "learning_rate": 1.8020656400386338e-06,
+      "loss": 0.7016,
+      "step": 34455
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4569284260624802,
+      "learning_rate": 1.801628701312802e-06,
+      "loss": 0.8023,
+      "step": 34456
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3747624328169397,
+      "learning_rate": 1.801191810320486e-06,
+      "loss": 0.8179,
+      "step": 34457
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4753150760670508,
+      "learning_rate": 1.8007549670642244e-06,
+      "loss": 0.7456,
+      "step": 34458
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6495006601380195,
+      "learning_rate": 1.8003181715465657e-06,
+      "loss": 0.7192,
+      "step": 34459
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3021322760758978,
+      "learning_rate": 1.799881423770049e-06,
+      "loss": 0.7101,
+      "step": 34460
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2400478295372728,
+      "learning_rate": 1.7994447237372204e-06,
+      "loss": 0.6587,
+      "step": 34461
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5843074164851547,
+      "learning_rate": 1.7990080714506175e-06,
+      "loss": 0.5002,
+      "step": 34462
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4112673177683543,
+      "learning_rate": 1.7985714669127897e-06,
+      "loss": 0.8294,
+      "step": 34463
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.607493809597049,
+      "learning_rate": 1.7981349101262746e-06,
+      "loss": 0.6015,
+      "step": 34464
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6506899281804766,
+      "learning_rate": 1.7976984010936128e-06,
+      "loss": 0.6504,
+      "step": 34465
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.647466504606814,
+      "learning_rate": 1.7972619398173474e-06,
+      "loss": 0.7947,
+      "step": 34466
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8041446384674005,
+      "learning_rate": 1.7968255263000235e-06,
+      "loss": 0.746,
+      "step": 34467
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.505609313699745,
+      "learning_rate": 1.7963891605441775e-06,
+      "loss": 0.7371,
+      "step": 34468
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.31688164473789,
+      "learning_rate": 1.7959528425523488e-06,
+      "loss": 0.6374,
+      "step": 34469
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.8094942718502183,
+      "learning_rate": 1.7955165723270796e-06,
+      "loss": 0.6669,
+      "step": 34470
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5884066727030195,
+      "learning_rate": 1.7950803498709124e-06,
+      "loss": 0.4888,
+      "step": 34471
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.421073678339006,
+      "learning_rate": 1.7946441751863864e-06,
+      "loss": 0.5935,
+      "step": 34472
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3897539870658568,
+      "learning_rate": 1.794208048276036e-06,
+      "loss": 0.5449,
+      "step": 34473
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5987972927433877,
+      "learning_rate": 1.793771969142405e-06,
+      "loss": 0.5229,
+      "step": 34474
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.453695078804152,
+      "learning_rate": 1.7933359377880344e-06,
+      "loss": 0.7663,
+      "step": 34475
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.364244818294893,
+      "learning_rate": 1.79289995421546e-06,
+      "loss": 0.704,
+      "step": 34476
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6962551599777662,
+      "learning_rate": 1.7924640184272169e-06,
+      "loss": 0.5797,
+      "step": 34477
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.574442597113664,
+      "learning_rate": 1.7920281304258502e-06,
+      "loss": 0.6465,
+      "step": 34478
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3653784989915112,
+      "learning_rate": 1.7915922902138916e-06,
+      "loss": 0.7172,
+      "step": 34479
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7409247622877226,
+      "learning_rate": 1.7911564977938845e-06,
+      "loss": 0.6848,
+      "step": 34480
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.602042659228479,
+      "learning_rate": 1.7907207531683601e-06,
+      "loss": 0.5057,
+      "step": 34481
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3190203080249148,
+      "learning_rate": 1.790285056339861e-06,
+      "loss": 0.5407,
+      "step": 34482
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3094618211520272,
+      "learning_rate": 1.7898494073109184e-06,
+      "loss": 0.6026,
+      "step": 34483
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6445964309419817,
+      "learning_rate": 1.7894138060840749e-06,
+      "loss": 0.5285,
+      "step": 34484
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.318020349892859,
+      "learning_rate": 1.7889782526618637e-06,
+      "loss": 0.5791,
+      "step": 34485
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4066752550018589,
+      "learning_rate": 1.7885427470468185e-06,
+      "loss": 0.7252,
+      "step": 34486
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2190426692641982,
+      "learning_rate": 1.7881072892414763e-06,
+      "loss": 0.5782,
+      "step": 34487
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5100950417169332,
+      "learning_rate": 1.7876718792483772e-06,
+      "loss": 0.5228,
+      "step": 34488
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4697755634735092,
+      "learning_rate": 1.7872365170700511e-06,
+      "loss": 0.7455,
+      "step": 34489
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4995536415790038,
+      "learning_rate": 1.786801202709032e-06,
+      "loss": 0.7273,
+      "step": 34490
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4617424832483388,
+      "learning_rate": 1.7863659361678566e-06,
+      "loss": 0.7014,
+      "step": 34491
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5491018783036307,
+      "learning_rate": 1.7859307174490614e-06,
+      "loss": 0.4794,
+      "step": 34492
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.851311299665735,
+      "learning_rate": 1.7854955465551783e-06,
+      "loss": 0.7426,
+      "step": 34493
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3654382894690618,
+      "learning_rate": 1.7850604234887382e-06,
+      "loss": 0.7957,
+      "step": 34494
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3868925893773103,
+      "learning_rate": 1.7846253482522768e-06,
+      "loss": 0.726,
+      "step": 34495
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.572704720074053,
+      "learning_rate": 1.7841903208483302e-06,
+      "loss": 0.653,
+      "step": 34496
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6319099464115514,
+      "learning_rate": 1.7837553412794295e-06,
+      "loss": 0.811,
+      "step": 34497
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7428656785916457,
+      "learning_rate": 1.7833204095481027e-06,
+      "loss": 0.6287,
+      "step": 34498
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3506653220783258,
+      "learning_rate": 1.7828855256568889e-06,
+      "loss": 0.6444,
+      "step": 34499
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.547612582638124,
+      "learning_rate": 1.782450689608315e-06,
+      "loss": 0.5627,
+      "step": 34500
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3187304039128853,
+      "learning_rate": 1.7820159014049165e-06,
+      "loss": 0.6803,
+      "step": 34501
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.355614414486381,
+      "learning_rate": 1.7815811610492217e-06,
+      "loss": 0.7199,
+      "step": 34502
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5895074954427406,
+      "learning_rate": 1.781146468543765e-06,
+      "loss": 0.5112,
+      "step": 34503
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4329917748280585,
+      "learning_rate": 1.7807118238910725e-06,
+      "loss": 0.6222,
+      "step": 34504
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4795969637785458,
+      "learning_rate": 1.7802772270936818e-06,
+      "loss": 0.766,
+      "step": 34505
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1971192009555864,
+      "learning_rate": 1.779842678154119e-06,
+      "loss": 0.5168,
+      "step": 34506
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.795013966653755,
+      "learning_rate": 1.7794081770749116e-06,
+      "loss": 0.7727,
+      "step": 34507
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4168906110131159,
+      "learning_rate": 1.7789737238585925e-06,
+      "loss": 0.6455,
+      "step": 34508
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4344397280158516,
+      "learning_rate": 1.7785393185076939e-06,
+      "loss": 0.6494,
+      "step": 34509
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2793507528471104,
+      "learning_rate": 1.7781049610247425e-06,
+      "loss": 0.7128,
+      "step": 34510
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.433601765188362,
+      "learning_rate": 1.7776706514122644e-06,
+      "loss": 0.7456,
+      "step": 34511
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.31678123195804,
+      "learning_rate": 1.7772363896727907e-06,
+      "loss": 0.7436,
+      "step": 34512
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2809127384380954,
+      "learning_rate": 1.7768021758088517e-06,
+      "loss": 0.534,
+      "step": 34513
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4370400003693171,
+      "learning_rate": 1.7763680098229752e-06,
+      "loss": 0.835,
+      "step": 34514
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6995147643353539,
+      "learning_rate": 1.775933891717685e-06,
+      "loss": 0.6654,
+      "step": 34515
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3287063553523955,
+      "learning_rate": 1.775499821495511e-06,
+      "loss": 0.6524,
+      "step": 34516
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5217282725397776,
+      "learning_rate": 1.7750657991589827e-06,
+      "loss": 0.7284,
+      "step": 34517
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4386607023596885,
+      "learning_rate": 1.7746318247106265e-06,
+      "loss": 0.6815,
+      "step": 34518
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5897904872061399,
+      "learning_rate": 1.7741978981529651e-06,
+      "loss": 0.4806,
+      "step": 34519
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3899114368941003,
+      "learning_rate": 1.7737640194885297e-06,
+      "loss": 0.6853,
+      "step": 34520
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4526539699297436,
+      "learning_rate": 1.773330188719843e-06,
+      "loss": 0.8239,
+      "step": 34521
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.360993933253971,
+      "learning_rate": 1.7728964058494336e-06,
+      "loss": 0.6158,
+      "step": 34522
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2619967798903784,
+      "learning_rate": 1.7724626708798253e-06,
+      "loss": 0.6394,
+      "step": 34523
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4110638534796645,
+      "learning_rate": 1.7720289838135452e-06,
+      "loss": 0.7702,
+      "step": 34524
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.337425861177116,
+      "learning_rate": 1.7715953446531153e-06,
+      "loss": 0.6955,
+      "step": 34525
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.435177157971891,
+      "learning_rate": 1.771161753401065e-06,
+      "loss": 0.635,
+      "step": 34526
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2311349626429293,
+      "learning_rate": 1.7707282100599154e-06,
+      "loss": 0.5424,
+      "step": 34527
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5963353592799283,
+      "learning_rate": 1.7702947146321904e-06,
+      "loss": 0.6542,
+      "step": 34528
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5030495519501048,
+      "learning_rate": 1.769861267120414e-06,
+      "loss": 0.588,
+      "step": 34529
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3530625272303407,
+      "learning_rate": 1.7694278675271115e-06,
+      "loss": 0.6762,
+      "step": 34530
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4796104369831344,
+      "learning_rate": 1.7689945158548094e-06,
+      "loss": 0.6064,
+      "step": 34531
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7528858739938098,
+      "learning_rate": 1.7685612121060235e-06,
+      "loss": 0.7327,
+      "step": 34532
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.606810737149687,
+      "learning_rate": 1.7681279562832798e-06,
+      "loss": 0.5004,
+      "step": 34533
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6876251587744626,
+      "learning_rate": 1.7676947483891039e-06,
+      "loss": 0.6507,
+      "step": 34534
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5958589833007427,
+      "learning_rate": 1.7672615884260159e-06,
+      "loss": 0.5316,
+      "step": 34535
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.771233274206595,
+      "learning_rate": 1.7668284763965337e-06,
+      "loss": 0.8406,
+      "step": 34536
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4875324999626702,
+      "learning_rate": 1.7663954123031834e-06,
+      "loss": 0.6908,
+      "step": 34537
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.330590958029724,
+      "learning_rate": 1.7659623961484883e-06,
+      "loss": 0.6518,
+      "step": 34538
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5680766938053305,
+      "learning_rate": 1.7655294279349666e-06,
+      "loss": 0.7858,
+      "step": 34539
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.459466999949687,
+      "learning_rate": 1.7650965076651383e-06,
+      "loss": 0.6926,
+      "step": 34540
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2968414134586106,
+      "learning_rate": 1.764663635341527e-06,
+      "loss": 0.6898,
+      "step": 34541
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4442957923977422,
+      "learning_rate": 1.7642308109666483e-06,
+      "loss": 0.6372,
+      "step": 34542
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1913566273914067,
+      "learning_rate": 1.763798034543025e-06,
+      "loss": 0.7185,
+      "step": 34543
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.499593080521984,
+      "learning_rate": 1.7633653060731803e-06,
+      "loss": 0.6663,
+      "step": 34544
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7291370282087597,
+      "learning_rate": 1.7629326255596302e-06,
+      "loss": 0.7188,
+      "step": 34545
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5782347162022679,
+      "learning_rate": 1.7624999930048913e-06,
+      "loss": 0.5316,
+      "step": 34546
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4610293540257075,
+      "learning_rate": 1.7620674084114865e-06,
+      "loss": 0.7813,
+      "step": 34547
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2885973288039148,
+      "learning_rate": 1.7616348717819376e-06,
+      "loss": 0.6157,
+      "step": 34548
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6136603193162385,
+      "learning_rate": 1.761202383118754e-06,
+      "loss": 0.5207,
+      "step": 34549
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4975822457532073,
+      "learning_rate": 1.7607699424244583e-06,
+      "loss": 0.7015,
+      "step": 34550
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2475117331795613,
+      "learning_rate": 1.7603375497015684e-06,
+      "loss": 0.6595,
+      "step": 34551
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3793908375504926,
+      "learning_rate": 1.7599052049526045e-06,
+      "loss": 0.6563,
+      "step": 34552
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2829534350174465,
+      "learning_rate": 1.759472908180081e-06,
+      "loss": 0.6418,
+      "step": 34553
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3052857318691173,
+      "learning_rate": 1.7590406593865128e-06,
+      "loss": 0.6636,
+      "step": 34554
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3546159166758351,
+      "learning_rate": 1.7586084585744213e-06,
+      "loss": 0.6421,
+      "step": 34555
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8331049591614936,
+      "learning_rate": 1.7581763057463187e-06,
+      "loss": 0.632,
+      "step": 34556
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5197749269247305,
+      "learning_rate": 1.757744200904724e-06,
+      "loss": 0.8132,
+      "step": 34557
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7745345396698755,
+      "learning_rate": 1.7573121440521513e-06,
+      "loss": 0.6423,
+      "step": 34558
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9329153830074493,
+      "learning_rate": 1.756880135191118e-06,
+      "loss": 0.7155,
+      "step": 34559
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2622547627039353,
+      "learning_rate": 1.756448174324137e-06,
+      "loss": 0.6775,
+      "step": 34560
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3851310656626845,
+      "learning_rate": 1.7560162614537257e-06,
+      "loss": 0.718,
+      "step": 34561
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4042275290036625,
+      "learning_rate": 1.7555843965823992e-06,
+      "loss": 0.6043,
+      "step": 34562
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3887501282656287,
+      "learning_rate": 1.7551525797126668e-06,
+      "loss": 0.63,
+      "step": 34563
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4105587921287701,
+      "learning_rate": 1.7547208108470471e-06,
+      "loss": 0.6781,
+      "step": 34564
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3339810684980484,
+      "learning_rate": 1.7542890899880548e-06,
+      "loss": 0.7171,
+      "step": 34565
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2840964354133395,
+      "learning_rate": 1.7538574171382028e-06,
+      "loss": 0.7034,
+      "step": 34566
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2446623984185914,
+      "learning_rate": 1.7534257923000008e-06,
+      "loss": 0.7396,
+      "step": 34567
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.114407485088748,
+      "learning_rate": 1.752994215475965e-06,
+      "loss": 0.7624,
+      "step": 34568
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6752578332241297,
+      "learning_rate": 1.7525626866686107e-06,
+      "loss": 0.6399,
+      "step": 34569
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3095777681153233,
+      "learning_rate": 1.7521312058804462e-06,
+      "loss": 0.571,
+      "step": 34570
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5728377327274254,
+      "learning_rate": 1.751699773113984e-06,
+      "loss": 0.503,
+      "step": 34571
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5740044350964443,
+      "learning_rate": 1.7512683883717363e-06,
+      "loss": 0.49,
+      "step": 34572
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6121203209297956,
+      "learning_rate": 1.7508370516562179e-06,
+      "loss": 0.7341,
+      "step": 34573
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5023143518825028,
+      "learning_rate": 1.7504057629699389e-06,
+      "loss": 0.6295,
+      "step": 34574
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.264035590702138,
+      "learning_rate": 1.7499745223154053e-06,
+      "loss": 0.6942,
+      "step": 34575
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3847366024934038,
+      "learning_rate": 1.7495433296951325e-06,
+      "loss": 0.828,
+      "step": 34576
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6558991782555639,
+      "learning_rate": 1.7491121851116332e-06,
+      "loss": 0.5311,
+      "step": 34577
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6141410150848471,
+      "learning_rate": 1.7486810885674154e-06,
+      "loss": 0.5333,
+      "step": 34578
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5645216095913368,
+      "learning_rate": 1.7482500400649849e-06,
+      "loss": 0.6891,
+      "step": 34579
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.230169966635423,
+      "learning_rate": 1.7478190396068584e-06,
+      "loss": 0.5536,
+      "step": 34580
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4840700615292985,
+      "learning_rate": 1.7473880871955395e-06,
+      "loss": 0.739,
+      "step": 34581
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2080776294576399,
+      "learning_rate": 1.746957182833543e-06,
+      "loss": 0.5708,
+      "step": 34582
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4241116827408054,
+      "learning_rate": 1.7465263265233712e-06,
+      "loss": 0.7614,
+      "step": 34583
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5188227452273815,
+      "learning_rate": 1.7460955182675387e-06,
+      "loss": 0.792,
+      "step": 34584
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2781250314415862,
+      "learning_rate": 1.7456647580685492e-06,
+      "loss": 0.7729,
+      "step": 34585
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4471267541045294,
+      "learning_rate": 1.7452340459289152e-06,
+      "loss": 0.609,
+      "step": 34586
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.166381078403878,
+      "learning_rate": 1.7448033818511422e-06,
+      "loss": 0.7163,
+      "step": 34587
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.42666693405313,
+      "learning_rate": 1.744372765837734e-06,
+      "loss": 0.6547,
+      "step": 34588
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4623400902294237,
+      "learning_rate": 1.743942197891202e-06,
+      "loss": 0.6689,
+      "step": 34589
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2523672223777786,
+      "learning_rate": 1.743511678014055e-06,
+      "loss": 0.5588,
+      "step": 34590
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.253348928363974,
+      "learning_rate": 1.7430812062087965e-06,
+      "loss": 0.6978,
+      "step": 34591
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3950152878184363,
+      "learning_rate": 1.7426507824779303e-06,
+      "loss": 0.6589,
+      "step": 34592
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4905403863346005,
+      "learning_rate": 1.7422204068239667e-06,
+      "loss": 0.5871,
+      "step": 34593
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3277114480856518,
+      "learning_rate": 1.7417900792494126e-06,
+      "loss": 0.7494,
+      "step": 34594
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.01246076023147,
+      "learning_rate": 1.7413597997567699e-06,
+      "loss": 0.6315,
+      "step": 34595
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3105130928438073,
+      "learning_rate": 1.740929568348544e-06,
+      "loss": 0.5977,
+      "step": 34596
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.48994536324854,
+      "learning_rate": 1.74049938502724e-06,
+      "loss": 0.7208,
+      "step": 34597
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3165345647975057,
+      "learning_rate": 1.7400692497953665e-06,
+      "loss": 0.7258,
+      "step": 34598
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.578588735966015,
+      "learning_rate": 1.7396391626554254e-06,
+      "loss": 0.6992,
+      "step": 34599
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6199816080272289,
+      "learning_rate": 1.7392091236099173e-06,
+      "loss": 0.5069,
+      "step": 34600
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.310943041549865,
+      "learning_rate": 1.7387791326613511e-06,
+      "loss": 0.7571,
+      "step": 34601
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4623868493857792,
+      "learning_rate": 1.7383491898122262e-06,
+      "loss": 0.6639,
+      "step": 34602
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3072627768485765,
+      "learning_rate": 1.7379192950650504e-06,
+      "loss": 0.7102,
+      "step": 34603
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.276406663394645,
+      "learning_rate": 1.7374894484223215e-06,
+      "loss": 0.6566,
+      "step": 34604
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3554486534880108,
+      "learning_rate": 1.7370596498865467e-06,
+      "loss": 0.73,
+      "step": 34605
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3705727923701287,
+      "learning_rate": 1.736629899460225e-06,
+      "loss": 0.693,
+      "step": 34606
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3996108420408007,
+      "learning_rate": 1.736200197145863e-06,
+      "loss": 0.6764,
+      "step": 34607
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.320443154620807,
+      "learning_rate": 1.7357705429459592e-06,
+      "loss": 0.6535,
+      "step": 34608
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6215860968774837,
+      "learning_rate": 1.7353409368630126e-06,
+      "loss": 0.5048,
+      "step": 34609
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4567803514991755,
+      "learning_rate": 1.7349113788995288e-06,
+      "loss": 0.678,
+      "step": 34610
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5962993164904556,
+      "learning_rate": 1.734481869058009e-06,
+      "loss": 0.5167,
+      "step": 34611
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7032965616457723,
+      "learning_rate": 1.7340524073409537e-06,
+      "loss": 0.6574,
+      "step": 34612
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6200680772181221,
+      "learning_rate": 1.7336229937508586e-06,
+      "loss": 0.5072,
+      "step": 34613
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6204384190460991,
+      "learning_rate": 1.7331936282902284e-06,
+      "loss": 0.5342,
+      "step": 34614
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3204114199234092,
+      "learning_rate": 1.7327643109615644e-06,
+      "loss": 0.6281,
+      "step": 34615
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3721922972736547,
+      "learning_rate": 1.7323350417673634e-06,
+      "loss": 0.7083,
+      "step": 34616
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6123578419268767,
+      "learning_rate": 1.7319058207101237e-06,
+      "loss": 0.7583,
+      "step": 34617
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6013912950426794,
+      "learning_rate": 1.7314766477923461e-06,
+      "loss": 0.6036,
+      "step": 34618
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2715019308043207,
+      "learning_rate": 1.7310475230165302e-06,
+      "loss": 0.6716,
+      "step": 34619
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3528002226665419,
+      "learning_rate": 1.7306184463851749e-06,
+      "loss": 0.6546,
+      "step": 34620
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2252647103128789,
+      "learning_rate": 1.7301894179007738e-06,
+      "loss": 0.6259,
+      "step": 34621
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.321452124545696,
+      "learning_rate": 1.7297604375658317e-06,
+      "loss": 0.6455,
+      "step": 34622
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.238399831946486,
+      "learning_rate": 1.7293315053828385e-06,
+      "loss": 0.717,
+      "step": 34623
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3943375835135634,
+      "learning_rate": 1.7289026213542992e-06,
+      "loss": 0.7126,
+      "step": 34624
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2643209804401163,
+      "learning_rate": 1.728473785482705e-06,
+      "loss": 0.5845,
+      "step": 34625
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.452208020249181,
+      "learning_rate": 1.7280449977705572e-06,
+      "loss": 0.6391,
+      "step": 34626
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3163502745158129,
+      "learning_rate": 1.727616258220348e-06,
+      "loss": 0.7148,
+      "step": 34627
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5661956777565402,
+      "learning_rate": 1.727187566834576e-06,
+      "loss": 0.653,
+      "step": 34628
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.429717144820006,
+      "learning_rate": 1.7267589236157422e-06,
+      "loss": 0.6987,
+      "step": 34629
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3605461346443966,
+      "learning_rate": 1.7263303285663324e-06,
+      "loss": 0.6681,
+      "step": 34630
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6000928244229401,
+      "learning_rate": 1.7259017816888469e-06,
+      "loss": 0.5024,
+      "step": 34631
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.318352143270032,
+      "learning_rate": 1.72547328298578e-06,
+      "loss": 0.6764,
+      "step": 34632
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3948349516185639,
+      "learning_rate": 1.7250448324596325e-06,
+      "loss": 0.6294,
+      "step": 34633
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3388243515683291,
+      "learning_rate": 1.7246164301128888e-06,
+      "loss": 0.8183,
+      "step": 34634
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.195639099978328,
+      "learning_rate": 1.724188075948049e-06,
+      "loss": 0.8674,
+      "step": 34635
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3334991582785858,
+      "learning_rate": 1.7237597699676079e-06,
+      "loss": 0.6719,
+      "step": 34636
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4799860423834066,
+      "learning_rate": 1.723331512174058e-06,
+      "loss": 0.6168,
+      "step": 34637
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5840710409863796,
+      "learning_rate": 1.7229033025698894e-06,
+      "loss": 0.4617,
+      "step": 34638
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4172482398758122,
+      "learning_rate": 1.7224751411575991e-06,
+      "loss": 0.6661,
+      "step": 34639
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3452407252116139,
+      "learning_rate": 1.7220470279396806e-06,
+      "loss": 0.8106,
+      "step": 34640
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0044022568973148,
+      "learning_rate": 1.721618962918623e-06,
+      "loss": 0.6759,
+      "step": 34641
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4135800903615086,
+      "learning_rate": 1.7211909460969233e-06,
+      "loss": 0.7892,
+      "step": 34642
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4374058910553749,
+      "learning_rate": 1.7207629774770718e-06,
+      "loss": 0.6325,
+      "step": 34643
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3710855053983402,
+      "learning_rate": 1.7203350570615563e-06,
+      "loss": 0.6643,
+      "step": 34644
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6016282117534542,
+      "learning_rate": 1.7199071848528703e-06,
+      "loss": 0.5018,
+      "step": 34645
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3788430711750574,
+      "learning_rate": 1.7194793608535099e-06,
+      "loss": 0.7944,
+      "step": 34646
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.937284170862255,
+      "learning_rate": 1.7190515850659616e-06,
+      "loss": 0.7681,
+      "step": 34647
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3212278674324662,
+      "learning_rate": 1.718623857492715e-06,
+      "loss": 0.7712,
+      "step": 34648
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6168938839055205,
+      "learning_rate": 1.718196178136261e-06,
+      "loss": 0.7446,
+      "step": 34649
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.961565997248221,
+      "learning_rate": 1.7177685469990946e-06,
+      "loss": 0.6804,
+      "step": 34650
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6024289295208065,
+      "learning_rate": 1.7173409640837002e-06,
+      "loss": 0.4973,
+      "step": 34651
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3888899827878753,
+      "learning_rate": 1.7169134293925682e-06,
+      "loss": 0.5837,
+      "step": 34652
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3918131286172541,
+      "learning_rate": 1.7164859429281876e-06,
+      "loss": 0.6917,
+      "step": 34653
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1778939906737713,
+      "learning_rate": 1.716058504693051e-06,
+      "loss": 0.6647,
+      "step": 34654
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4975424604814758,
+      "learning_rate": 1.7156311146896442e-06,
+      "loss": 0.733,
+      "step": 34655
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.395596559544697,
+      "learning_rate": 1.715203772920453e-06,
+      "loss": 0.6291,
+      "step": 34656
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6410279759368693,
+      "learning_rate": 1.7147764793879718e-06,
+      "loss": 0.6015,
+      "step": 34657
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3006649499357499,
+      "learning_rate": 1.7143492340946821e-06,
+      "loss": 0.5933,
+      "step": 34658
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.20561063749265,
+      "learning_rate": 1.7139220370430765e-06,
+      "loss": 0.7313,
+      "step": 34659
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3548449444654875,
+      "learning_rate": 1.7134948882356383e-06,
+      "loss": 0.7641,
+      "step": 34660
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3215413081756393,
+      "learning_rate": 1.7130677876748592e-06,
+      "loss": 0.6743,
+      "step": 34661
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5686663250278416,
+      "learning_rate": 1.7126407353632191e-06,
+      "loss": 0.6141,
+      "step": 34662
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4699508814622837,
+      "learning_rate": 1.7122137313032116e-06,
+      "loss": 0.7103,
+      "step": 34663
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1233519001051437,
+      "learning_rate": 1.7117867754973206e-06,
+      "loss": 0.6323,
+      "step": 34664
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4451762496533065,
+      "learning_rate": 1.711359867948027e-06,
+      "loss": 0.7615,
+      "step": 34665
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4183821569147044,
+      "learning_rate": 1.7109330086578213e-06,
+      "loss": 0.6949,
+      "step": 34666
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.602770289229601,
+      "learning_rate": 1.7105061976291893e-06,
+      "loss": 0.6253,
+      "step": 34667
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2887043337956146,
+      "learning_rate": 1.7100794348646155e-06,
+      "loss": 0.596,
+      "step": 34668
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5850340100650883,
+      "learning_rate": 1.7096527203665814e-06,
+      "loss": 0.5262,
+      "step": 34669
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3900937989328543,
+      "learning_rate": 1.7092260541375738e-06,
+      "loss": 0.682,
+      "step": 34670
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6391144686708203,
+      "learning_rate": 1.7087994361800787e-06,
+      "loss": 0.5777,
+      "step": 34671
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4294084810396368,
+      "learning_rate": 1.7083728664965793e-06,
+      "loss": 0.8192,
+      "step": 34672
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4660996275032376,
+      "learning_rate": 1.707946345089555e-06,
+      "loss": 0.6326,
+      "step": 34673
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.594281427802162,
+      "learning_rate": 1.7075198719614927e-06,
+      "loss": 0.7082,
+      "step": 34674
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6821363334394903,
+      "learning_rate": 1.707093447114877e-06,
+      "loss": 0.7099,
+      "step": 34675
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.712990979928105,
+      "learning_rate": 1.7066670705521893e-06,
+      "loss": 0.7407,
+      "step": 34676
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4623688986540844,
+      "learning_rate": 1.7062407422759085e-06,
+      "loss": 0.7441,
+      "step": 34677
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.569982823496224,
+      "learning_rate": 1.705814462288522e-06,
+      "loss": 0.4937,
+      "step": 34678
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3092293679082327,
+      "learning_rate": 1.7053882305925085e-06,
+      "loss": 0.5663,
+      "step": 34679
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.195791103493806,
+      "learning_rate": 1.7049620471903517e-06,
+      "loss": 0.5386,
+      "step": 34680
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5185267611830522,
+      "learning_rate": 1.7045359120845307e-06,
+      "loss": 0.6525,
+      "step": 34681
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5495385689267667,
+      "learning_rate": 1.70410982527753e-06,
+      "loss": 0.6423,
+      "step": 34682
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5958582010717186,
+      "learning_rate": 1.7036837867718259e-06,
+      "loss": 0.8154,
+      "step": 34683
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3877118946144422,
+      "learning_rate": 1.7032577965699037e-06,
+      "loss": 0.8095,
+      "step": 34684
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3335387042737115,
+      "learning_rate": 1.7028318546742418e-06,
+      "loss": 0.6593,
+      "step": 34685
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3228339825454305,
+      "learning_rate": 1.7024059610873167e-06,
+      "loss": 0.7173,
+      "step": 34686
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3408237582452964,
+      "learning_rate": 1.7019801158116112e-06,
+      "loss": 0.7096,
+      "step": 34687
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.273705826879398,
+      "learning_rate": 1.7015543188496065e-06,
+      "loss": 0.7163,
+      "step": 34688
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3037869493777845,
+      "learning_rate": 1.7011285702037804e-06,
+      "loss": 0.6156,
+      "step": 34689
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3160946172201047,
+      "learning_rate": 1.7007028698766082e-06,
+      "loss": 0.6098,
+      "step": 34690
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5635197446869857,
+      "learning_rate": 1.7002772178705717e-06,
+      "loss": 0.675,
+      "step": 34691
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.765455341559907,
+      "learning_rate": 1.6998516141881517e-06,
+      "loss": 0.7525,
+      "step": 34692
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2970772036985931,
+      "learning_rate": 1.6994260588318235e-06,
+      "loss": 0.6247,
+      "step": 34693
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4654986677419697,
+      "learning_rate": 1.699000551804062e-06,
+      "loss": 0.61,
+      "step": 34694
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3017173404511972,
+      "learning_rate": 1.6985750931073476e-06,
+      "loss": 0.6568,
+      "step": 34695
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.509717345271998,
+      "learning_rate": 1.698149682744159e-06,
+      "loss": 0.699,
+      "step": 34696
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5446208191243636,
+      "learning_rate": 1.6977243207169714e-06,
+      "loss": 0.729,
+      "step": 34697
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.173577746261782,
+      "learning_rate": 1.697299007028259e-06,
+      "loss": 0.6086,
+      "step": 34698
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3682257430996592,
+      "learning_rate": 1.6968737416805003e-06,
+      "loss": 0.5181,
+      "step": 34699
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.600430502104417,
+      "learning_rate": 1.696448524676173e-06,
+      "loss": 0.7163,
+      "step": 34700
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.421546302712352,
+      "learning_rate": 1.6960233560177531e-06,
+      "loss": 0.7056,
+      "step": 34701
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7002511373075053,
+      "learning_rate": 1.6955982357077106e-06,
+      "loss": 0.7159,
+      "step": 34702
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3166369720126139,
+      "learning_rate": 1.6951731637485269e-06,
+      "loss": 0.7037,
+      "step": 34703
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3076989093688833,
+      "learning_rate": 1.6947481401426712e-06,
+      "loss": 0.7253,
+      "step": 34704
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5807772753559369,
+      "learning_rate": 1.6943231648926251e-06,
+      "loss": 0.5058,
+      "step": 34705
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4878799203092075,
+      "learning_rate": 1.693898238000856e-06,
+      "loss": 0.5605,
+      "step": 34706
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7648019770115202,
+      "learning_rate": 1.6934733594698426e-06,
+      "loss": 0.7125,
+      "step": 34707
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7556972447787955,
+      "learning_rate": 1.693048529302056e-06,
+      "loss": 0.7372,
+      "step": 34708
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2537776064577406,
+      "learning_rate": 1.692623747499973e-06,
+      "loss": 0.6189,
+      "step": 34709
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.526430107219106,
+      "learning_rate": 1.6921990140660638e-06,
+      "loss": 0.7018,
+      "step": 34710
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.347589722933641,
+      "learning_rate": 1.6917743290027999e-06,
+      "loss": 0.77,
+      "step": 34711
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3432532777004509,
+      "learning_rate": 1.691349692312657e-06,
+      "loss": 0.6696,
+      "step": 34712
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2638343059304886,
+      "learning_rate": 1.6909251039981056e-06,
+      "loss": 0.6394,
+      "step": 34713
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4930546202681898,
+      "learning_rate": 1.6905005640616245e-06,
+      "loss": 0.6216,
+      "step": 34714
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3917439897331747,
+      "learning_rate": 1.6900760725056742e-06,
+      "loss": 0.6817,
+      "step": 34715
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4990171975796516,
+      "learning_rate": 1.6896516293327326e-06,
+      "loss": 0.5873,
+      "step": 34716
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1490702765575465,
+      "learning_rate": 1.6892272345452732e-06,
+      "loss": 0.6286,
+      "step": 34717
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6030473311978775,
+      "learning_rate": 1.688802888145763e-06,
+      "loss": 0.6181,
+      "step": 34718
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3733384494113057,
+      "learning_rate": 1.6883785901366712e-06,
+      "loss": 0.8345,
+      "step": 34719
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3569948357329844,
+      "learning_rate": 1.6879543405204712e-06,
+      "loss": 0.5683,
+      "step": 34720
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4769120375635365,
+      "learning_rate": 1.6875301392996346e-06,
+      "loss": 0.6892,
+      "step": 34721
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5199405907367198,
+      "learning_rate": 1.6871059864766293e-06,
+      "loss": 0.6127,
+      "step": 34722
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6490609067590982,
+      "learning_rate": 1.6866818820539232e-06,
+      "loss": 0.7043,
+      "step": 34723
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6067757497107649,
+      "learning_rate": 1.686257826033988e-06,
+      "loss": 0.7409,
+      "step": 34724
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5548983291516572,
+      "learning_rate": 1.68583381841929e-06,
+      "loss": 0.5805,
+      "step": 34725
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.382114200410036,
+      "learning_rate": 1.6854098592123002e-06,
+      "loss": 0.6753,
+      "step": 34726
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.579529460584915,
+      "learning_rate": 1.6849859484154885e-06,
+      "loss": 0.5939,
+      "step": 34727
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.308542220047198,
+      "learning_rate": 1.6845620860313216e-06,
+      "loss": 0.5884,
+      "step": 34728
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3403529838559807,
+      "learning_rate": 1.6841382720622646e-06,
+      "loss": 0.6113,
+      "step": 34729
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.707256303880262,
+      "learning_rate": 1.6837145065107862e-06,
+      "loss": 0.7388,
+      "step": 34730
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7057096234069091,
+      "learning_rate": 1.6832907893793604e-06,
+      "loss": 0.7722,
+      "step": 34731
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3839241148293586,
+      "learning_rate": 1.682867120670444e-06,
+      "loss": 0.6882,
+      "step": 34732
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5837396060080877,
+      "learning_rate": 1.6824435003865081e-06,
+      "loss": 0.4708,
+      "step": 34733
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3313021486614065,
+      "learning_rate": 1.6820199285300199e-06,
+      "loss": 0.5094,
+      "step": 34734
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2787945485745686,
+      "learning_rate": 1.6815964051034472e-06,
+      "loss": 0.6377,
+      "step": 34735
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5869079503515292,
+      "learning_rate": 1.6811729301092528e-06,
+      "loss": 0.4844,
+      "step": 34736
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3134429384912454,
+      "learning_rate": 1.6807495035499022e-06,
+      "loss": 0.7024,
+      "step": 34737
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4024332231811871,
+      "learning_rate": 1.6803261254278635e-06,
+      "loss": 0.8624,
+      "step": 34738
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.433208019867949,
+      "learning_rate": 1.6799027957455983e-06,
+      "loss": 0.6804,
+      "step": 34739
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5149731877226065,
+      "learning_rate": 1.6794795145055753e-06,
+      "loss": 0.6657,
+      "step": 34740
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3868141551384294,
+      "learning_rate": 1.6790562817102541e-06,
+      "loss": 0.6373,
+      "step": 34741
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3258456575996975,
+      "learning_rate": 1.6786330973621045e-06,
+      "loss": 0.6972,
+      "step": 34742
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.348576847400531,
+      "learning_rate": 1.6782099614635848e-06,
+      "loss": 0.5526,
+      "step": 34743
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3027200246782613,
+      "learning_rate": 1.6777868740171633e-06,
+      "loss": 0.6939,
+      "step": 34744
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4508469216153894,
+      "learning_rate": 1.677363835025302e-06,
+      "loss": 0.577,
+      "step": 34745
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.400048378605554,
+      "learning_rate": 1.6769408444904612e-06,
+      "loss": 0.5592,
+      "step": 34746
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3860935628538833,
+      "learning_rate": 1.6765179024151047e-06,
+      "loss": 0.7233,
+      "step": 34747
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.461399327461541,
+      "learning_rate": 1.6760950088016991e-06,
+      "loss": 0.813,
+      "step": 34748
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3041775694313837,
+      "learning_rate": 1.675672163652704e-06,
+      "loss": 0.841,
+      "step": 34749
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4867440678333488,
+      "learning_rate": 1.6752493669705782e-06,
+      "loss": 0.683,
+      "step": 34750
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3964094020047557,
+      "learning_rate": 1.6748266187577866e-06,
+      "loss": 0.6591,
+      "step": 34751
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5021991497525042,
+      "learning_rate": 1.6744039190167926e-06,
+      "loss": 0.7097,
+      "step": 34752
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4794410702145415,
+      "learning_rate": 1.6739812677500543e-06,
+      "loss": 0.7179,
+      "step": 34753
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.898932307993366,
+      "learning_rate": 1.673558664960031e-06,
+      "loss": 0.5191,
+      "step": 34754
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4851313568713245,
+      "learning_rate": 1.673136110649185e-06,
+      "loss": 0.7123,
+      "step": 34755
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2839840674201852,
+      "learning_rate": 1.6727136048199788e-06,
+      "loss": 0.5692,
+      "step": 34756
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3526667456746684,
+      "learning_rate": 1.6722911474748715e-06,
+      "loss": 0.7414,
+      "step": 34757
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3337009903417363,
+      "learning_rate": 1.671868738616319e-06,
+      "loss": 0.636,
+      "step": 34758
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1503656667771183,
+      "learning_rate": 1.6714463782467848e-06,
+      "loss": 0.5658,
+      "step": 34759
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.283882943922044,
+      "learning_rate": 1.671024066368725e-06,
+      "loss": 0.684,
+      "step": 34760
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8696622826384626,
+      "learning_rate": 1.6706018029846016e-06,
+      "loss": 0.6226,
+      "step": 34761
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3549032686338858,
+      "learning_rate": 1.6701795880968696e-06,
+      "loss": 0.5763,
+      "step": 34762
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.340644265060116,
+      "learning_rate": 1.6697574217079926e-06,
+      "loss": 0.6659,
+      "step": 34763
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4547952832015438,
+      "learning_rate": 1.6693353038204218e-06,
+      "loss": 0.5181,
+      "step": 34764
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4895588560287376,
+      "learning_rate": 1.668913234436621e-06,
+      "loss": 0.6749,
+      "step": 34765
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9778789598074464,
+      "learning_rate": 1.6684912135590447e-06,
+      "loss": 0.6254,
+      "step": 34766
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4874034732628727,
+      "learning_rate": 1.6680692411901478e-06,
+      "loss": 0.7419,
+      "step": 34767
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4150587056317254,
+      "learning_rate": 1.6676473173323892e-06,
+      "loss": 0.6785,
+      "step": 34768
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.544293299083796,
+      "learning_rate": 1.6672254419882295e-06,
+      "loss": 0.7663,
+      "step": 34769
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.340145351463173,
+      "learning_rate": 1.6668036151601197e-06,
+      "loss": 0.7414,
+      "step": 34770
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.488840351573794,
+      "learning_rate": 1.6663818368505168e-06,
+      "loss": 0.6543,
+      "step": 34771
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7431216450647364,
+      "learning_rate": 1.6659601070618758e-06,
+      "loss": 0.6864,
+      "step": 34772
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5462584059688569,
+      "learning_rate": 1.6655384257966566e-06,
+      "loss": 0.6523,
+      "step": 34773
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.369632594699449,
+      "learning_rate": 1.665116793057312e-06,
+      "loss": 0.6611,
+      "step": 34774
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.355540723295772,
+      "learning_rate": 1.6646952088462932e-06,
+      "loss": 0.6861,
+      "step": 34775
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7184344500704622,
+      "learning_rate": 1.664273673166058e-06,
+      "loss": 0.6899,
+      "step": 34776
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2792500591993912,
+      "learning_rate": 1.6638521860190625e-06,
+      "loss": 0.699,
+      "step": 34777
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4165912099099138,
+      "learning_rate": 1.6634307474077594e-06,
+      "loss": 0.5824,
+      "step": 34778
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3318156293185037,
+      "learning_rate": 1.6630093573345985e-06,
+      "loss": 0.6023,
+      "step": 34779
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.57214923559829,
+      "learning_rate": 1.6625880158020403e-06,
+      "loss": 0.7157,
+      "step": 34780
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.280723842937669,
+      "learning_rate": 1.6621667228125305e-06,
+      "loss": 0.7036,
+      "step": 34781
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4884581556537941,
+      "learning_rate": 1.6617454783685283e-06,
+      "loss": 0.6849,
+      "step": 34782
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1643620352441715,
+      "learning_rate": 1.6613242824724818e-06,
+      "loss": 0.5425,
+      "step": 34783
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.296599237182202,
+      "learning_rate": 1.660903135126848e-06,
+      "loss": 0.6331,
+      "step": 34784
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.367602329969854,
+      "learning_rate": 1.6604820363340735e-06,
+      "loss": 0.7774,
+      "step": 34785
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.05173585412782,
+      "learning_rate": 1.6600609860966155e-06,
+      "loss": 0.6958,
+      "step": 34786
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3589008614043683,
+      "learning_rate": 1.659639984416922e-06,
+      "loss": 0.6639,
+      "step": 34787
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.23717328276498,
+      "learning_rate": 1.6592190312974433e-06,
+      "loss": 0.593,
+      "step": 34788
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5051924076663576,
+      "learning_rate": 1.6587981267406328e-06,
+      "loss": 0.7268,
+      "step": 34789
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2852215261342277,
+      "learning_rate": 1.658377270748941e-06,
+      "loss": 0.6341,
+      "step": 34790
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.313812351468262,
+      "learning_rate": 1.6579564633248191e-06,
+      "loss": 0.7001,
+      "step": 34791
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8676114694346617,
+      "learning_rate": 1.657535704470713e-06,
+      "loss": 0.6212,
+      "step": 34792
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3397735489262854,
+      "learning_rate": 1.6571149941890753e-06,
+      "loss": 0.6412,
+      "step": 34793
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4010213385805954,
+      "learning_rate": 1.656694332482357e-06,
+      "loss": 0.8045,
+      "step": 34794
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5995247836350824,
+      "learning_rate": 1.6562737193530065e-06,
+      "loss": 0.5314,
+      "step": 34795
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3584415830788332,
+      "learning_rate": 1.6558531548034696e-06,
+      "loss": 0.6111,
+      "step": 34796
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.340170672887574,
+      "learning_rate": 1.6554326388361963e-06,
+      "loss": 0.5849,
+      "step": 34797
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2410025400332914,
+      "learning_rate": 1.6550121714536393e-06,
+      "loss": 0.6085,
+      "step": 34798
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6005260008344964,
+      "learning_rate": 1.6545917526582434e-06,
+      "loss": 0.6836,
+      "step": 34799
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.9076027175492323,
+      "learning_rate": 1.6541713824524542e-06,
+      "loss": 0.5911,
+      "step": 34800
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5540382460721653,
+      "learning_rate": 1.653751060838723e-06,
+      "loss": 0.4785,
+      "step": 34801
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.600944842595669,
+      "learning_rate": 1.653330787819495e-06,
+      "loss": 0.6245,
+      "step": 34802
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8790671358208277,
+      "learning_rate": 1.6529105633972176e-06,
+      "loss": 0.5193,
+      "step": 34803
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.254278734408814,
+      "learning_rate": 1.652490387574337e-06,
+      "loss": 0.7155,
+      "step": 34804
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.295275599650744,
+      "learning_rate": 1.6520702603533012e-06,
+      "loss": 0.7868,
+      "step": 34805
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4222056769065707,
+      "learning_rate": 1.6516501817365538e-06,
+      "loss": 0.6345,
+      "step": 34806
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.006846370794621,
+      "learning_rate": 1.651230151726545e-06,
+      "loss": 0.6419,
+      "step": 34807
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.29060649446432,
+      "learning_rate": 1.6508101703257161e-06,
+      "loss": 0.6638,
+      "step": 34808
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5276709850226253,
+      "learning_rate": 1.6503902375365121e-06,
+      "loss": 0.6028,
+      "step": 34809
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7604712854259637,
+      "learning_rate": 1.6499703533613788e-06,
+      "loss": 0.586,
+      "step": 34810
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.311399763403793,
+      "learning_rate": 1.6495505178027626e-06,
+      "loss": 0.6703,
+      "step": 34811
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4993232254969628,
+      "learning_rate": 1.649130730863111e-06,
+      "loss": 0.5557,
+      "step": 34812
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5144274799351882,
+      "learning_rate": 1.6487109925448597e-06,
+      "loss": 0.5859,
+      "step": 34813
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.516576853784286,
+      "learning_rate": 1.6482913028504577e-06,
+      "loss": 0.8959,
+      "step": 34814
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2781886179052282,
+      "learning_rate": 1.6478716617823488e-06,
+      "loss": 0.7044,
+      "step": 34815
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2766191211662923,
+      "learning_rate": 1.6474520693429763e-06,
+      "loss": 0.6494,
+      "step": 34816
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6699852272484468,
+      "learning_rate": 1.6470325255347798e-06,
+      "loss": 0.7004,
+      "step": 34817
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.58847643366435,
+      "learning_rate": 1.6466130303602047e-06,
+      "loss": 0.7137,
+      "step": 34818
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.661480152099539,
+      "learning_rate": 1.6461935838216947e-06,
+      "loss": 0.6131,
+      "step": 34819
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4213754779165777,
+      "learning_rate": 1.6457741859216903e-06,
+      "loss": 0.8619,
+      "step": 34820
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5862794222555587,
+      "learning_rate": 1.6453548366626316e-06,
+      "loss": 0.6337,
+      "step": 34821
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3727932066513013,
+      "learning_rate": 1.6449355360469621e-06,
+      "loss": 0.6198,
+      "step": 34822
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5814405231639774,
+      "learning_rate": 1.6445162840771256e-06,
+      "loss": 0.5199,
+      "step": 34823
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.663627320794926,
+      "learning_rate": 1.644097080755558e-06,
+      "loss": 0.68,
+      "step": 34824
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5187214735638563,
+      "learning_rate": 1.6436779260847058e-06,
+      "loss": 0.6512,
+      "step": 34825
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5802413551978682,
+      "learning_rate": 1.6432588200670053e-06,
+      "loss": 0.7576,
+      "step": 34826
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3745785563761328,
+      "learning_rate": 1.6428397627048953e-06,
+      "loss": 0.6984,
+      "step": 34827
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4652095620337453,
+      "learning_rate": 1.6424207540008185e-06,
+      "loss": 0.6706,
+      "step": 34828
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4372410934811002,
+      "learning_rate": 1.6420017939572164e-06,
+      "loss": 0.6591,
+      "step": 34829
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4477144191822486,
+      "learning_rate": 1.6415828825765256e-06,
+      "loss": 0.648,
+      "step": 34830
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3245955009478787,
+      "learning_rate": 1.6411640198611834e-06,
+      "loss": 0.6013,
+      "step": 34831
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4893738042255045,
+      "learning_rate": 1.6407452058136298e-06,
+      "loss": 0.6285,
+      "step": 34832
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4177192801944745,
+      "learning_rate": 1.6403264404363073e-06,
+      "loss": 0.7032,
+      "step": 34833
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3760212201958353,
+      "learning_rate": 1.6399077237316507e-06,
+      "loss": 0.6196,
+      "step": 34834
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.7208212450906215,
+      "learning_rate": 1.639489055702096e-06,
+      "loss": 0.7669,
+      "step": 34835
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4819196925069018,
+      "learning_rate": 1.6390704363500832e-06,
+      "loss": 0.742,
+      "step": 34836
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.6358441795968925,
+      "learning_rate": 1.6386518656780504e-06,
+      "loss": 0.5398,
+      "step": 34837
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.509502056177718,
+      "learning_rate": 1.6382333436884345e-06,
+      "loss": 0.7455,
+      "step": 34838
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2992375745319222,
+      "learning_rate": 1.6378148703836695e-06,
+      "loss": 0.4647,
+      "step": 34839
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4137685410360468,
+      "learning_rate": 1.637396445766195e-06,
+      "loss": 0.6565,
+      "step": 34840
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3318107961922359,
+      "learning_rate": 1.6369780698384442e-06,
+      "loss": 0.6808,
+      "step": 34841
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4171327197393462,
+      "learning_rate": 1.6365597426028569e-06,
+      "loss": 0.5849,
+      "step": 34842
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4763285847643002,
+      "learning_rate": 1.6361414640618645e-06,
+      "loss": 0.7103,
+      "step": 34843
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4052951365504454,
+      "learning_rate": 1.6357232342179063e-06,
+      "loss": 0.6429,
+      "step": 34844
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.280855117998612,
+      "learning_rate": 1.6353050530734139e-06,
+      "loss": 0.7091,
+      "step": 34845
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6616483094142775,
+      "learning_rate": 1.634886920630826e-06,
+      "loss": 0.7207,
+      "step": 34846
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2632344240499003,
+      "learning_rate": 1.6344688368925743e-06,
+      "loss": 0.6504,
+      "step": 34847
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.418848784130413,
+      "learning_rate": 1.6340508018610913e-06,
+      "loss": 0.5024,
+      "step": 34848
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6477922670981473,
+      "learning_rate": 1.6336328155388138e-06,
+      "loss": 0.7066,
+      "step": 34849
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5759918759436463,
+      "learning_rate": 1.6332148779281765e-06,
+      "loss": 0.4475,
+      "step": 34850
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4685442723270694,
+      "learning_rate": 1.6327969890316109e-06,
+      "loss": 0.6369,
+      "step": 34851
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5329819311846142,
+      "learning_rate": 1.6323791488515472e-06,
+      "loss": 0.781,
+      "step": 34852
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5686883361644129,
+      "learning_rate": 1.6319613573904226e-06,
+      "loss": 0.4884,
+      "step": 34853
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5940282264972697,
+      "learning_rate": 1.6315436146506702e-06,
+      "loss": 0.5473,
+      "step": 34854
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2975509398058096,
+      "learning_rate": 1.6311259206347207e-06,
+      "loss": 0.5822,
+      "step": 34855
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.626075280940148,
+      "learning_rate": 1.6307082753450022e-06,
+      "loss": 0.6416,
+      "step": 34856
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5056158429569537,
+      "learning_rate": 1.6302906787839501e-06,
+      "loss": 0.6948,
+      "step": 34857
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.439125044644187,
+      "learning_rate": 1.6298731309539984e-06,
+      "loss": 0.5457,
+      "step": 34858
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4596461721560288,
+      "learning_rate": 1.629455631857575e-06,
+      "loss": 0.8211,
+      "step": 34859
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1975672898829923,
+      "learning_rate": 1.629038181497109e-06,
+      "loss": 0.5647,
+      "step": 34860
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4331654873193667,
+      "learning_rate": 1.6286207798750342e-06,
+      "loss": 0.6892,
+      "step": 34861
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3671272565988484,
+      "learning_rate": 1.6282034269937785e-06,
+      "loss": 0.7277,
+      "step": 34862
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.545977425294997,
+      "learning_rate": 1.6277861228557745e-06,
+      "loss": 0.8275,
+      "step": 34863
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6353585655151947,
+      "learning_rate": 1.627368867463448e-06,
+      "loss": 0.5153,
+      "step": 34864
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5982593842976688,
+      "learning_rate": 1.6269516608192338e-06,
+      "loss": 0.4931,
+      "step": 34865
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5699692958418923,
+      "learning_rate": 1.6265345029255542e-06,
+      "loss": 0.656,
+      "step": 34866
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.257153599149417,
+      "learning_rate": 1.6261173937848451e-06,
+      "loss": 0.7017,
+      "step": 34867
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3677785391488906,
+      "learning_rate": 1.6257003333995315e-06,
+      "loss": 0.5997,
+      "step": 34868
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6912608843713057,
+      "learning_rate": 1.6252833217720386e-06,
+      "loss": 0.7575,
+      "step": 34869
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.279856614368262,
+      "learning_rate": 1.6248663589047985e-06,
+      "loss": 0.7779,
+      "step": 34870
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6660777220998806,
+      "learning_rate": 1.6244494448002402e-06,
+      "loss": 0.5712,
+      "step": 34871
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4451520983680395,
+      "learning_rate": 1.624032579460788e-06,
+      "loss": 0.5499,
+      "step": 34872
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6344365802144764,
+      "learning_rate": 1.6236157628888694e-06,
+      "loss": 0.4885,
+      "step": 34873
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3258796815917515,
+      "learning_rate": 1.6231989950869099e-06,
+      "loss": 0.7165,
+      "step": 34874
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7305782760308224,
+      "learning_rate": 1.6227822760573408e-06,
+      "loss": 0.6124,
+      "step": 34875
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.298780947716676,
+      "learning_rate": 1.6223656058025861e-06,
+      "loss": 0.6098,
+      "step": 34876
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4587719394717868,
+      "learning_rate": 1.621948984325068e-06,
+      "loss": 0.7534,
+      "step": 34877
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.384496760155998,
+      "learning_rate": 1.6215324116272158e-06,
+      "loss": 0.777,
+      "step": 34878
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.299665217698744,
+      "learning_rate": 1.6211158877114574e-06,
+      "loss": 0.6265,
+      "step": 34879
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4978316207515916,
+      "learning_rate": 1.6206994125802144e-06,
+      "loss": 0.6012,
+      "step": 34880
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.174043393506379,
+      "learning_rate": 1.6202829862359093e-06,
+      "loss": 0.6144,
+      "step": 34881
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1951355497230851,
+      "learning_rate": 1.6198666086809733e-06,
+      "loss": 0.6756,
+      "step": 34882
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7967997185746072,
+      "learning_rate": 1.6194502799178236e-06,
+      "loss": 0.6334,
+      "step": 34883
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.058027712300853,
+      "learning_rate": 1.6190339999488902e-06,
+      "loss": 0.6262,
+      "step": 34884
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.486401576315075,
+      "learning_rate": 1.6186177687765926e-06,
+      "loss": 0.7586,
+      "step": 34885
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6236893519875848,
+      "learning_rate": 1.6182015864033574e-06,
+      "loss": 0.6302,
+      "step": 34886
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4212371757804645,
+      "learning_rate": 1.617785452831604e-06,
+      "loss": 0.663,
+      "step": 34887
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3235056086083001,
+      "learning_rate": 1.6173693680637592e-06,
+      "loss": 0.6616,
+      "step": 34888
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.39531812129822,
+      "learning_rate": 1.6169533321022446e-06,
+      "loss": 0.6687,
+      "step": 34889
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4312253571198228,
+      "learning_rate": 1.6165373449494793e-06,
+      "loss": 0.7239,
+      "step": 34890
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3117921372274208,
+      "learning_rate": 1.6161214066078878e-06,
+      "loss": 0.6424,
+      "step": 34891
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4744814257204937,
+      "learning_rate": 1.615705517079894e-06,
+      "loss": 0.6729,
+      "step": 34892
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7827651575229682,
+      "learning_rate": 1.615289676367917e-06,
+      "loss": 0.6159,
+      "step": 34893
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.043237452305832,
+      "learning_rate": 1.614873884474376e-06,
+      "loss": 0.6303,
+      "step": 34894
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4949602873451457,
+      "learning_rate": 1.6144581414016936e-06,
+      "loss": 0.6637,
+      "step": 34895
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5846098353426967,
+      "learning_rate": 1.614042447152293e-06,
+      "loss": 0.7762,
+      "step": 34896
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.322170255763362,
+      "learning_rate": 1.6136268017285928e-06,
+      "loss": 0.6771,
+      "step": 34897
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5930067718688534,
+      "learning_rate": 1.6132112051330095e-06,
+      "loss": 0.5099,
+      "step": 34898
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7281050107473073,
+      "learning_rate": 1.612795657367966e-06,
+      "loss": 0.5872,
+      "step": 34899
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4382889893343227,
+      "learning_rate": 1.6123801584358845e-06,
+      "loss": 0.5717,
+      "step": 34900
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4648553563890006,
+      "learning_rate": 1.6119647083391798e-06,
+      "loss": 0.7089,
+      "step": 34901
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6312757262462243,
+      "learning_rate": 1.6115493070802712e-06,
+      "loss": 0.7761,
+      "step": 34902
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.820863391392325,
+      "learning_rate": 1.61113395466158e-06,
+      "loss": 0.7474,
+      "step": 34903
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.657548697492633,
+      "learning_rate": 1.61071865108552e-06,
+      "loss": 0.7302,
+      "step": 34904
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3868455488136768,
+      "learning_rate": 1.6103033963545156e-06,
+      "loss": 0.7577,
+      "step": 34905
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2987282668107485,
+      "learning_rate": 1.6098881904709774e-06,
+      "loss": 0.565,
+      "step": 34906
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5478677485267276,
+      "learning_rate": 1.6094730334373287e-06,
+      "loss": 0.6922,
+      "step": 34907
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3254597363349032,
+      "learning_rate": 1.6090579252559834e-06,
+      "loss": 0.6708,
+      "step": 34908
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.386367180703213,
+      "learning_rate": 1.6086428659293585e-06,
+      "loss": 0.7216,
+      "step": 34909
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6004809912271099,
+      "learning_rate": 1.6082278554598763e-06,
+      "loss": 0.6083,
+      "step": 34910
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6108150973383841,
+      "learning_rate": 1.6078128938499427e-06,
+      "loss": 0.7447,
+      "step": 34911
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3336239429126076,
+      "learning_rate": 1.6073979811019803e-06,
+      "loss": 0.6115,
+      "step": 34912
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5815405525241766,
+      "learning_rate": 1.6069831172184035e-06,
+      "loss": 0.6942,
+      "step": 34913
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3272155371909176,
+      "learning_rate": 1.6065683022016343e-06,
+      "loss": 0.6093,
+      "step": 34914
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5193030247805976,
+      "learning_rate": 1.6061535360540758e-06,
+      "loss": 0.6076,
+      "step": 34915
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.307993348965375,
+      "learning_rate": 1.6057388187781486e-06,
+      "loss": 0.637,
+      "step": 34916
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.687351050163145,
+      "learning_rate": 1.6053241503762707e-06,
+      "loss": 0.6396,
+      "step": 34917
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4487541000411153,
+      "learning_rate": 1.6049095308508521e-06,
+      "loss": 0.5462,
+      "step": 34918
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5884854786563238,
+      "learning_rate": 1.6044949602043069e-06,
+      "loss": 0.7256,
+      "step": 34919
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5883302005013256,
+      "learning_rate": 1.6040804384390496e-06,
+      "loss": 0.6867,
+      "step": 34920
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.224625071193218,
+      "learning_rate": 1.603665965557497e-06,
+      "loss": 0.626,
+      "step": 34921
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4027525463184765,
+      "learning_rate": 1.6032515415620564e-06,
+      "loss": 0.7859,
+      "step": 34922
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.449441068916166,
+      "learning_rate": 1.6028371664551456e-06,
+      "loss": 0.6604,
+      "step": 34923
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2982857751814076,
+      "learning_rate": 1.6024228402391762e-06,
+      "loss": 0.7241,
+      "step": 34924
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3798248767396124,
+      "learning_rate": 1.6020085629165572e-06,
+      "loss": 0.6997,
+      "step": 34925
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3968579334316482,
+      "learning_rate": 1.6015943344897022e-06,
+      "loss": 0.6362,
+      "step": 34926
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5384325703602775,
+      "learning_rate": 1.6011801549610273e-06,
+      "loss": 0.7537,
+      "step": 34927
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5255007682223996,
+      "learning_rate": 1.6007660243329405e-06,
+      "loss": 0.6399,
+      "step": 34928
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5520090632761856,
+      "learning_rate": 1.6003519426078507e-06,
+      "loss": 0.6625,
+      "step": 34929
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3243210821441556,
+      "learning_rate": 1.5999379097881706e-06,
+      "loss": 0.6497,
+      "step": 34930
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5757785309327498,
+      "learning_rate": 1.599523925876314e-06,
+      "loss": 0.4664,
+      "step": 34931
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.390001410196989,
+      "learning_rate": 1.5991099908746887e-06,
+      "loss": 0.63,
+      "step": 34932
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3828080669605798,
+      "learning_rate": 1.598696104785702e-06,
+      "loss": 0.644,
+      "step": 34933
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3402450587481913,
+      "learning_rate": 1.5982822676117672e-06,
+      "loss": 0.602,
+      "step": 34934
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.305345324324419,
+      "learning_rate": 1.5978684793552946e-06,
+      "loss": 0.7502,
+      "step": 34935
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3107306408464148,
+      "learning_rate": 1.5974547400186914e-06,
+      "loss": 0.7579,
+      "step": 34936
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.518495251833634,
+      "learning_rate": 1.5970410496043653e-06,
+      "loss": 0.7204,
+      "step": 34937
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5982538740605063,
+      "learning_rate": 1.5966274081147293e-06,
+      "loss": 0.6912,
+      "step": 34938
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1087795385533368,
+      "learning_rate": 1.5962138155521855e-06,
+      "loss": 0.6643,
+      "step": 34939
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4174690819226512,
+      "learning_rate": 1.595800271919149e-06,
+      "loss": 0.604,
+      "step": 34940
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6169325507569658,
+      "learning_rate": 1.595386777218021e-06,
+      "loss": 0.661,
+      "step": 34941
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6676533729467157,
+      "learning_rate": 1.594973331451214e-06,
+      "loss": 0.5229,
+      "step": 34942
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3936562231536906,
+      "learning_rate": 1.5945599346211315e-06,
+      "loss": 0.645,
+      "step": 34943
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3641109897834176,
+      "learning_rate": 1.5941465867301842e-06,
+      "loss": 0.566,
+      "step": 34944
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4329133748513,
+      "learning_rate": 1.5937332877807753e-06,
+      "loss": 0.6424,
+      "step": 34945
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5783434404257377,
+      "learning_rate": 1.5933200377753145e-06,
+      "loss": 0.4981,
+      "step": 34946
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3077545155110768,
+      "learning_rate": 1.5929068367162026e-06,
+      "loss": 0.7112,
+      "step": 34947
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5569917482488478,
+      "learning_rate": 1.5924936846058515e-06,
+      "loss": 0.7709,
+      "step": 34948
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1973336080438621,
+      "learning_rate": 1.5920805814466643e-06,
+      "loss": 0.6776,
+      "step": 34949
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5139100982688478,
+      "learning_rate": 1.5916675272410442e-06,
+      "loss": 0.5915,
+      "step": 34950
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.394005109515911,
+      "learning_rate": 1.5912545219913966e-06,
+      "loss": 0.6787,
+      "step": 34951
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3525215287440595,
+      "learning_rate": 1.5908415657001297e-06,
+      "loss": 0.6767,
+      "step": 34952
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.869516092673539,
+      "learning_rate": 1.590428658369646e-06,
+      "loss": 0.6186,
+      "step": 34953
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7501873264744086,
+      "learning_rate": 1.5900158000023469e-06,
+      "loss": 0.6362,
+      "step": 34954
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3003437753564695,
+      "learning_rate": 1.5896029906006383e-06,
+      "loss": 0.8197,
+      "step": 34955
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.244523577498006,
+      "learning_rate": 1.589190230166926e-06,
+      "loss": 0.5739,
+      "step": 34956
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4034094430131647,
+      "learning_rate": 1.5887775187036113e-06,
+      "loss": 0.6013,
+      "step": 34957
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6680550567933057,
+      "learning_rate": 1.5883648562130937e-06,
+      "loss": 0.7762,
+      "step": 34958
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3279972585062945,
+      "learning_rate": 1.587952242697779e-06,
+      "loss": 0.585,
+      "step": 34959
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3439324198595097,
+      "learning_rate": 1.5875396781600727e-06,
+      "loss": 0.5444,
+      "step": 34960
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 12.524727367861274,
+      "learning_rate": 1.5871271626023733e-06,
+      "loss": 0.6575,
+      "step": 34961
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.386045611787066,
+      "learning_rate": 1.5867146960270806e-06,
+      "loss": 0.6005,
+      "step": 34962
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5292659878923018,
+      "learning_rate": 1.5863022784366012e-06,
+      "loss": 0.7538,
+      "step": 34963
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8257791229099665,
+      "learning_rate": 1.5858899098333314e-06,
+      "loss": 0.6915,
+      "step": 34964
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6249151115050022,
+      "learning_rate": 1.5854775902196762e-06,
+      "loss": 0.4837,
+      "step": 34965
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4327662349142778,
+      "learning_rate": 1.5850653195980325e-06,
+      "loss": 0.5882,
+      "step": 34966
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3169103323310631,
+      "learning_rate": 1.5846530979708053e-06,
+      "loss": 0.6314,
+      "step": 34967
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2519883856488845,
+      "learning_rate": 1.584240925340389e-06,
+      "loss": 0.5788,
+      "step": 34968
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3453617366234554,
+      "learning_rate": 1.5838288017091885e-06,
+      "loss": 0.7887,
+      "step": 34969
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6190932977530177,
+      "learning_rate": 1.583416727079602e-06,
+      "loss": 0.649,
+      "step": 34970
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6654328192846062,
+      "learning_rate": 1.583004701454025e-06,
+      "loss": 0.6186,
+      "step": 34971
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.349185357195649,
+      "learning_rate": 1.5825927248348604e-06,
+      "loss": 0.6684,
+      "step": 34972
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6254003570792769,
+      "learning_rate": 1.5821807972245073e-06,
+      "loss": 0.5266,
+      "step": 34973
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4252586239316032,
+      "learning_rate": 1.5817689186253627e-06,
+      "loss": 0.6892,
+      "step": 34974
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.581746529127585,
+      "learning_rate": 1.5813570890398234e-06,
+      "loss": 0.4948,
+      "step": 34975
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7262951917092455,
+      "learning_rate": 1.5809453084702875e-06,
+      "loss": 0.83,
+      "step": 34976
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6094260713192243,
+      "learning_rate": 1.5805335769191555e-06,
+      "loss": 0.6901,
+      "step": 34977
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5919289558904846,
+      "learning_rate": 1.580121894388823e-06,
+      "loss": 0.809,
+      "step": 34978
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5963126282364444,
+      "learning_rate": 1.5797102608816838e-06,
+      "loss": 0.639,
+      "step": 34979
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3768637091585854,
+      "learning_rate": 1.5792986764001372e-06,
+      "loss": 0.6761,
+      "step": 34980
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6147917513191469,
+      "learning_rate": 1.5788871409465822e-06,
+      "loss": 0.5264,
+      "step": 34981
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4816209438834755,
+      "learning_rate": 1.5784756545234126e-06,
+      "loss": 0.5177,
+      "step": 34982
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3920750697896371,
+      "learning_rate": 1.5780642171330217e-06,
+      "loss": 0.6617,
+      "step": 34983
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3343026205617803,
+      "learning_rate": 1.5776528287778093e-06,
+      "loss": 0.8392,
+      "step": 34984
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3845142407258078,
+      "learning_rate": 1.5772414894601663e-06,
+      "loss": 0.7677,
+      "step": 34985
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4363581318389216,
+      "learning_rate": 1.576830199182493e-06,
+      "loss": 0.6318,
+      "step": 34986
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6292999180188001,
+      "learning_rate": 1.5764189579471778e-06,
+      "loss": 0.5163,
+      "step": 34987
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4988961332798931,
+      "learning_rate": 1.5760077657566208e-06,
+      "loss": 0.572,
+      "step": 34988
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4424658270628439,
+      "learning_rate": 1.5755966226132114e-06,
+      "loss": 0.6889,
+      "step": 34989
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5113560350554973,
+      "learning_rate": 1.5751855285193474e-06,
+      "loss": 0.6735,
+      "step": 34990
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7137700046186182,
+      "learning_rate": 1.5747744834774215e-06,
+      "loss": 0.76,
+      "step": 34991
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3335694685846988,
+      "learning_rate": 1.5743634874898228e-06,
+      "loss": 0.6699,
+      "step": 34992
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4000562160768015,
+      "learning_rate": 1.5739525405589474e-06,
+      "loss": 0.5502,
+      "step": 34993
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3928684408704382,
+      "learning_rate": 1.5735416426871886e-06,
+      "loss": 0.6035,
+      "step": 34994
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3977528244821222,
+      "learning_rate": 1.5731307938769426e-06,
+      "loss": 0.6147,
+      "step": 34995
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6342512358700554,
+      "learning_rate": 1.5727199941305914e-06,
+      "loss": 0.6727,
+      "step": 34996
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9215143839315605,
+      "learning_rate": 1.5723092434505327e-06,
+      "loss": 0.575,
+      "step": 34997
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6633422001445181,
+      "learning_rate": 1.5718985418391609e-06,
+      "loss": 0.5038,
+      "step": 34998
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7184719263672077,
+      "learning_rate": 1.5714878892988638e-06,
+      "loss": 0.6154,
+      "step": 34999
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.483766859422462,
+      "learning_rate": 1.57107728583203e-06,
+      "loss": 0.6411,
+      "step": 35000
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2369712489184759,
+      "learning_rate": 1.5706667314410528e-06,
+      "loss": 0.5867,
+      "step": 35001
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3807517416147597,
+      "learning_rate": 1.570256226128325e-06,
+      "loss": 0.6941,
+      "step": 35002
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.595401921282247,
+      "learning_rate": 1.5698457698962344e-06,
+      "loss": 0.5423,
+      "step": 35003
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.626843003364521,
+      "learning_rate": 1.569435362747168e-06,
+      "loss": 0.7676,
+      "step": 35004
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.6372319299768305,
+      "learning_rate": 1.5690250046835209e-06,
+      "loss": 0.8205,
+      "step": 35005
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9046032810910125,
+      "learning_rate": 1.5686146957076776e-06,
+      "loss": 0.612,
+      "step": 35006
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5781474685700144,
+      "learning_rate": 1.5682044358220272e-06,
+      "loss": 0.5537,
+      "step": 35007
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6334958455542021,
+      "learning_rate": 1.5677942250289634e-06,
+      "loss": 0.6963,
+      "step": 35008
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.463822308172718,
+      "learning_rate": 1.5673840633308713e-06,
+      "loss": 0.6698,
+      "step": 35009
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5642983159479571,
+      "learning_rate": 1.5669739507301362e-06,
+      "loss": 0.4249,
+      "step": 35010
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5101065329539582,
+      "learning_rate": 1.5665638872291488e-06,
+      "loss": 0.7737,
+      "step": 35011
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5480238214689845,
+      "learning_rate": 1.5661538728303016e-06,
+      "loss": 0.4505,
+      "step": 35012
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9839329966251655,
+      "learning_rate": 1.5657439075359716e-06,
+      "loss": 0.6233,
+      "step": 35013
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4393772324114869,
+      "learning_rate": 1.5653339913485498e-06,
+      "loss": 0.6576,
+      "step": 35014
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.370277769210239,
+      "learning_rate": 1.564924124270425e-06,
+      "loss": 0.6624,
+      "step": 35015
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3741402223417953,
+      "learning_rate": 1.564514306303987e-06,
+      "loss": 0.6475,
+      "step": 35016
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3618352108085698,
+      "learning_rate": 1.5641045374516117e-06,
+      "loss": 0.7065,
+      "step": 35017
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8753673842597836,
+      "learning_rate": 1.563694817715692e-06,
+      "loss": 0.7588,
+      "step": 35018
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6190763279220499,
+      "learning_rate": 1.5632851470986132e-06,
+      "loss": 0.4917,
+      "step": 35019
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4321458615677347,
+      "learning_rate": 1.5628755256027573e-06,
+      "loss": 0.5084,
+      "step": 35020
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5310898071028416,
+      "learning_rate": 1.5624659532305142e-06,
+      "loss": 0.7316,
+      "step": 35021
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2032040472953258,
+      "learning_rate": 1.5620564299842623e-06,
+      "loss": 0.6781,
+      "step": 35022
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2495225142772937,
+      "learning_rate": 1.5616469558663927e-06,
+      "loss": 0.7006,
+      "step": 35023
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2947429345273025,
+      "learning_rate": 1.5612375308792826e-06,
+      "loss": 0.6622,
+      "step": 35024
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2464173232564002,
+      "learning_rate": 1.5608281550253224e-06,
+      "loss": 0.7097,
+      "step": 35025
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5359632331924093,
+      "learning_rate": 1.5604188283068921e-06,
+      "loss": 0.6412,
+      "step": 35026
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6972183108216703,
+      "learning_rate": 1.5600095507263725e-06,
+      "loss": 0.5831,
+      "step": 35027
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.414982596482457,
+      "learning_rate": 1.55960032228615e-06,
+      "loss": 0.6616,
+      "step": 35028
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.657393185332504,
+      "learning_rate": 1.5591911429886087e-06,
+      "loss": 0.7123,
+      "step": 35029
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4409948504999823,
+      "learning_rate": 1.5587820128361297e-06,
+      "loss": 0.6831,
+      "step": 35030
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.867911072196384,
+      "learning_rate": 1.5583729318310902e-06,
+      "loss": 0.669,
+      "step": 35031
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.55395382744768,
+      "learning_rate": 1.5579638999758772e-06,
+      "loss": 0.7444,
+      "step": 35032
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3519549808576257,
+      "learning_rate": 1.5575549172728733e-06,
+      "loss": 0.7545,
+      "step": 35033
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3719848459833714,
+      "learning_rate": 1.5571459837244585e-06,
+      "loss": 0.6735,
+      "step": 35034
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5370978751515043,
+      "learning_rate": 1.556737099333009e-06,
+      "loss": 0.7015,
+      "step": 35035
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4166405595894151,
+      "learning_rate": 1.5563282641009092e-06,
+      "loss": 0.5005,
+      "step": 35036
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3239738047285377,
+      "learning_rate": 1.555919478030542e-06,
+      "loss": 0.6853,
+      "step": 35037
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2667243296875057,
+      "learning_rate": 1.5555107411242854e-06,
+      "loss": 0.7569,
+      "step": 35038
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6293615033757651,
+      "learning_rate": 1.5551020533845162e-06,
+      "loss": 0.471,
+      "step": 35039
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4279634154130851,
+      "learning_rate": 1.5546934148136196e-06,
+      "loss": 0.6633,
+      "step": 35040
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3235903454895739,
+      "learning_rate": 1.5542848254139674e-06,
+      "loss": 0.7355,
+      "step": 35041
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5608852596477167,
+      "learning_rate": 1.5538762851879464e-06,
+      "loss": 0.6018,
+      "step": 35042
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6450521165866452,
+      "learning_rate": 1.5534677941379295e-06,
+      "loss": 0.7594,
+      "step": 35043
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.260811423861669,
+      "learning_rate": 1.5530593522662995e-06,
+      "loss": 0.6756,
+      "step": 35044
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2775034180217135,
+      "learning_rate": 1.5526509595754291e-06,
+      "loss": 0.7143,
+      "step": 35045
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7437312473312407,
+      "learning_rate": 1.5522426160677029e-06,
+      "loss": 0.7804,
+      "step": 35046
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.326404137950868,
+      "learning_rate": 1.5518343217454934e-06,
+      "loss": 0.6469,
+      "step": 35047
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4530659590480033,
+      "learning_rate": 1.5514260766111777e-06,
+      "loss": 0.7447,
+      "step": 35048
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.291515619313859,
+      "learning_rate": 1.5510178806671328e-06,
+      "loss": 0.7049,
+      "step": 35049
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5085028982498285,
+      "learning_rate": 1.5506097339157399e-06,
+      "loss": 0.6639,
+      "step": 35050
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7417680769037958,
+      "learning_rate": 1.550201636359373e-06,
+      "loss": 0.7025,
+      "step": 35051
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7507959963015607,
+      "learning_rate": 1.5497935880004034e-06,
+      "loss": 0.7579,
+      "step": 35052
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6383719614990573,
+      "learning_rate": 1.5493855888412123e-06,
+      "loss": 0.6186,
+      "step": 35053
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.575529842236628,
+      "learning_rate": 1.5489776388841748e-06,
+      "loss": 0.6968,
+      "step": 35054
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5279983837290394,
+      "learning_rate": 1.5485697381316656e-06,
+      "loss": 0.7032,
+      "step": 35055
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6353592093878753,
+      "learning_rate": 1.548161886586057e-06,
+      "loss": 0.6328,
+      "step": 35056
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.443424113639185,
+      "learning_rate": 1.5477540842497251e-06,
+      "loss": 0.7102,
+      "step": 35057
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.316798982134785,
+      "learning_rate": 1.5473463311250481e-06,
+      "loss": 0.7015,
+      "step": 35058
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4210599984002759,
+      "learning_rate": 1.5469386272143961e-06,
+      "loss": 0.5985,
+      "step": 35059
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4823739456905458,
+      "learning_rate": 1.5465309725201416e-06,
+      "loss": 0.6712,
+      "step": 35060
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3973502372830802,
+      "learning_rate": 1.5461233670446618e-06,
+      "loss": 0.6919,
+      "step": 35061
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2849364426926884,
+      "learning_rate": 1.5457158107903269e-06,
+      "loss": 0.64,
+      "step": 35062
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5602538144505622,
+      "learning_rate": 1.5453083037595128e-06,
+      "loss": 0.7507,
+      "step": 35063
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3569117450829706,
+      "learning_rate": 1.5449008459545877e-06,
+      "loss": 0.6761,
+      "step": 35064
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5897289058334961,
+      "learning_rate": 1.5444934373779296e-06,
+      "loss": 0.4806,
+      "step": 35065
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4038952585088849,
+      "learning_rate": 1.5440860780319055e-06,
+      "loss": 0.6826,
+      "step": 35066
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3158763432819633,
+      "learning_rate": 1.5436787679188913e-06,
+      "loss": 0.8004,
+      "step": 35067
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5069302010452423,
+      "learning_rate": 1.5432715070412552e-06,
+      "loss": 0.6344,
+      "step": 35068
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0296844743746076,
+      "learning_rate": 1.5428642954013706e-06,
+      "loss": 0.5199,
+      "step": 35069
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6108343668876255,
+      "learning_rate": 1.5424571330016059e-06,
+      "loss": 0.5332,
+      "step": 35070
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2737922908705603,
+      "learning_rate": 1.5420500198443367e-06,
+      "loss": 0.6893,
+      "step": 35071
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.372705093258003,
+      "learning_rate": 1.541642955931929e-06,
+      "loss": 0.6808,
+      "step": 35072
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3661221275804831,
+      "learning_rate": 1.541235941266751e-06,
+      "loss": 0.7236,
+      "step": 35073
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2829856066353056,
+      "learning_rate": 1.5408289758511763e-06,
+      "loss": 0.585,
+      "step": 35074
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5724066564448185,
+      "learning_rate": 1.5404220596875752e-06,
+      "loss": 0.5214,
+      "step": 35075
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.423537171695588,
+      "learning_rate": 1.5400151927783158e-06,
+      "loss": 0.6831,
+      "step": 35076
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2263040902271902,
+      "learning_rate": 1.5396083751257628e-06,
+      "loss": 0.7128,
+      "step": 35077
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8573302837690608,
+      "learning_rate": 1.5392016067322891e-06,
+      "loss": 0.6931,
+      "step": 35078
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3508978634843467,
+      "learning_rate": 1.5387948876002645e-06,
+      "loss": 0.6074,
+      "step": 35079
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6342852211784584,
+      "learning_rate": 1.5383882177320553e-06,
+      "loss": 0.5714,
+      "step": 35080
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5358732600982077,
+      "learning_rate": 1.537981597130026e-06,
+      "loss": 0.7181,
+      "step": 35081
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4881999124582963,
+      "learning_rate": 1.537575025796547e-06,
+      "loss": 0.6141,
+      "step": 35082
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.308890988458593,
+      "learning_rate": 1.5371685037339878e-06,
+      "loss": 0.665,
+      "step": 35083
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6060199514925065,
+      "learning_rate": 1.536762030944713e-06,
+      "loss": 0.5239,
+      "step": 35084
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4731678163998385,
+      "learning_rate": 1.536355607431086e-06,
+      "loss": 0.7154,
+      "step": 35085
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4732290415543676,
+      "learning_rate": 1.5359492331954795e-06,
+      "loss": 0.6742,
+      "step": 35086
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.414857051795953,
+      "learning_rate": 1.5355429082402528e-06,
+      "loss": 0.6987,
+      "step": 35087
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3528427224857453,
+      "learning_rate": 1.5351366325677785e-06,
+      "loss": 0.6272,
+      "step": 35088
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.330903330063197,
+      "learning_rate": 1.5347304061804158e-06,
+      "loss": 0.5614,
+      "step": 35089
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.540301007748593,
+      "learning_rate": 1.5343242290805348e-06,
+      "loss": 0.7205,
+      "step": 35090
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5577306976249017,
+      "learning_rate": 1.5339181012704974e-06,
+      "loss": 0.448,
+      "step": 35091
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7312717167502183,
+      "learning_rate": 1.5335120227526678e-06,
+      "loss": 0.6363,
+      "step": 35092
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5053300268462582,
+      "learning_rate": 1.5331059935294168e-06,
+      "loss": 0.6597,
+      "step": 35093
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.295258389014167,
+      "learning_rate": 1.532700013603098e-06,
+      "loss": 0.7102,
+      "step": 35094
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3750651756158079,
+      "learning_rate": 1.5322940829760802e-06,
+      "loss": 0.8286,
+      "step": 35095
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3633073266012339,
+      "learning_rate": 1.5318882016507286e-06,
+      "loss": 0.786,
+      "step": 35096
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3750321312482217,
+      "learning_rate": 1.531482369629408e-06,
+      "loss": 0.6345,
+      "step": 35097
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5495412061317302,
+      "learning_rate": 1.531076586914474e-06,
+      "loss": 0.7515,
+      "step": 35098
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4779065635990918,
+      "learning_rate": 1.5306708535082937e-06,
+      "loss": 0.6131,
+      "step": 35099
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5704751802104346,
+      "learning_rate": 1.530265169413231e-06,
+      "loss": 0.493,
+      "step": 35100
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5529209683685001,
+      "learning_rate": 1.5298595346316457e-06,
+      "loss": 0.6597,
+      "step": 35101
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3923824504165765,
+      "learning_rate": 1.5294539491658978e-06,
+      "loss": 0.6296,
+      "step": 35102
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.379582774432841,
+      "learning_rate": 1.5290484130183515e-06,
+      "loss": 0.5982,
+      "step": 35103
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3404670308988713,
+      "learning_rate": 1.5286429261913682e-06,
+      "loss": 0.6791,
+      "step": 35104
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3113311022091212,
+      "learning_rate": 1.5282374886873063e-06,
+      "loss": 0.5925,
+      "step": 35105
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4116675869576472,
+      "learning_rate": 1.5278321005085296e-06,
+      "loss": 0.6841,
+      "step": 35106
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5036408978784892,
+      "learning_rate": 1.5274267616573968e-06,
+      "loss": 0.7426,
+      "step": 35107
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3592912856104424,
+      "learning_rate": 1.5270214721362653e-06,
+      "loss": 0.6548,
+      "step": 35108
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5557214606017828,
+      "learning_rate": 1.5266162319474974e-06,
+      "loss": 0.618,
+      "step": 35109
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3440508779387144,
+      "learning_rate": 1.5262110410934561e-06,
+      "loss": 0.619,
+      "step": 35110
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.425171404201328,
+      "learning_rate": 1.5258058995764946e-06,
+      "loss": 0.5893,
+      "step": 35111
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3603489599824303,
+      "learning_rate": 1.5254008073989734e-06,
+      "loss": 0.7775,
+      "step": 35112
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.2197160065308394,
+      "learning_rate": 1.5249957645632508e-06,
+      "loss": 0.7155,
+      "step": 35113
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.370129025640788,
+      "learning_rate": 1.5245907710716912e-06,
+      "loss": 0.6376,
+      "step": 35114
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4508474360110717,
+      "learning_rate": 1.524185826926643e-06,
+      "loss": 0.6529,
+      "step": 35115
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8869519670801775,
+      "learning_rate": 1.5237809321304674e-06,
+      "loss": 0.8025,
+      "step": 35116
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.63613481449762,
+      "learning_rate": 1.523376086685523e-06,
+      "loss": 0.7557,
+      "step": 35117
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5151612888320027,
+      "learning_rate": 1.5229712905941696e-06,
+      "loss": 0.5249,
+      "step": 35118
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2741330812292806,
+      "learning_rate": 1.52256654385876e-06,
+      "loss": 0.689,
+      "step": 35119
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.571296108075504,
+      "learning_rate": 1.5221618464816501e-06,
+      "loss": 0.6389,
+      "step": 35120
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6007157797702194,
+      "learning_rate": 1.5217571984652012e-06,
+      "loss": 0.6301,
+      "step": 35121
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4045943649234875,
+      "learning_rate": 1.5213525998117628e-06,
+      "loss": 0.8389,
+      "step": 35122
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.30883145625387,
+      "learning_rate": 1.5209480505236963e-06,
+      "loss": 0.6924,
+      "step": 35123
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3567530012465698,
+      "learning_rate": 1.5205435506033529e-06,
+      "loss": 0.6285,
+      "step": 35124
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.201196637300597,
+      "learning_rate": 1.5201391000530908e-06,
+      "loss": 0.7183,
+      "step": 35125
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4924105516057917,
+      "learning_rate": 1.5197346988752615e-06,
+      "loss": 0.6174,
+      "step": 35126
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3122692507570655,
+      "learning_rate": 1.5193303470722243e-06,
+      "loss": 0.788,
+      "step": 35127
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.598169491651926,
+      "learning_rate": 1.5189260446463306e-06,
+      "loss": 0.628,
+      "step": 35128
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.690334071439071,
+      "learning_rate": 1.5185217915999328e-06,
+      "loss": 0.664,
+      "step": 35129
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4856537333431394,
+      "learning_rate": 1.518117587935385e-06,
+      "loss": 0.6458,
+      "step": 35130
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2509958042877716,
+      "learning_rate": 1.517713433655046e-06,
+      "loss": 0.6247,
+      "step": 35131
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3119276546137295,
+      "learning_rate": 1.5173093287612628e-06,
+      "loss": 0.6979,
+      "step": 35132
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4740547526034515,
+      "learning_rate": 1.5169052732563895e-06,
+      "loss": 0.6393,
+      "step": 35133
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5349963258205341,
+      "learning_rate": 1.5165012671427793e-06,
+      "loss": 0.8177,
+      "step": 35134
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2038874742439878,
+      "learning_rate": 1.516097310422786e-06,
+      "loss": 0.6693,
+      "step": 35135
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4569705996333677,
+      "learning_rate": 1.5156934030987603e-06,
+      "loss": 0.6715,
+      "step": 35136
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7665893680938851,
+      "learning_rate": 1.515289545173052e-06,
+      "loss": 0.6344,
+      "step": 35137
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.474730976418606,
+      "learning_rate": 1.5148857366480141e-06,
+      "loss": 0.7129,
+      "step": 35138
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4180351471371635,
+      "learning_rate": 1.514481977525999e-06,
+      "loss": 0.7041,
+      "step": 35139
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2924529871700376,
+      "learning_rate": 1.514078267809358e-06,
+      "loss": 0.6782,
+      "step": 35140
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2675994786318525,
+      "learning_rate": 1.513674607500436e-06,
+      "loss": 0.5863,
+      "step": 35141
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5858538671998355,
+      "learning_rate": 1.5132709966015901e-06,
+      "loss": 0.483,
+      "step": 35142
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9809154204681378,
+      "learning_rate": 1.512867435115165e-06,
+      "loss": 0.6705,
+      "step": 35143
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5158229036589965,
+      "learning_rate": 1.5124639230435157e-06,
+      "loss": 0.6876,
+      "step": 35144
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.423348874714502,
+      "learning_rate": 1.5120604603889855e-06,
+      "loss": 0.7405,
+      "step": 35145
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4286095084074968,
+      "learning_rate": 1.5116570471539294e-06,
+      "loss": 0.7313,
+      "step": 35146
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5908248571396452,
+      "learning_rate": 1.5112536833406898e-06,
+      "loss": 0.5025,
+      "step": 35147
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2743202250972876,
+      "learning_rate": 1.5108503689516218e-06,
+      "loss": 0.6977,
+      "step": 35148
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.781738630808584,
+      "learning_rate": 1.5104471039890711e-06,
+      "loss": 0.5986,
+      "step": 35149
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2357311963211801,
+      "learning_rate": 1.5100438884553826e-06,
+      "loss": 0.6521,
+      "step": 35150
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5297978387074678,
+      "learning_rate": 1.5096407223529053e-06,
+      "loss": 0.6992,
+      "step": 35151
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7843959194283734,
+      "learning_rate": 1.5092376056839908e-06,
+      "loss": 0.7617,
+      "step": 35152
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6192221324529763,
+      "learning_rate": 1.5088345384509838e-06,
+      "loss": 0.6528,
+      "step": 35153
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4632967002757258,
+      "learning_rate": 1.5084315206562272e-06,
+      "loss": 0.5385,
+      "step": 35154
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1700778769643303,
+      "learning_rate": 1.508028552302071e-06,
+      "loss": 0.7668,
+      "step": 35155
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.294001096319232,
+      "learning_rate": 1.5076256333908635e-06,
+      "loss": 0.5936,
+      "step": 35156
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7203527586576164,
+      "learning_rate": 1.5072227639249482e-06,
+      "loss": 0.71,
+      "step": 35157
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3431638832518247,
+      "learning_rate": 1.5068199439066678e-06,
+      "loss": 0.6077,
+      "step": 35158
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4540094748512138,
+      "learning_rate": 1.5064171733383715e-06,
+      "loss": 0.5929,
+      "step": 35159
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4559048411780717,
+      "learning_rate": 1.5060144522224052e-06,
+      "loss": 0.8136,
+      "step": 35160
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6207003921810383,
+      "learning_rate": 1.5056117805611115e-06,
+      "loss": 0.4904,
+      "step": 35161
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.353831799098169,
+      "learning_rate": 1.505209158356834e-06,
+      "loss": 0.579,
+      "step": 35162
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6007053445058463,
+      "learning_rate": 1.50480658561192e-06,
+      "loss": 0.6131,
+      "step": 35163
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.299180980440967,
+      "learning_rate": 1.5044040623287092e-06,
+      "loss": 0.673,
+      "step": 35164
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3884393821029521,
+      "learning_rate": 1.5040015885095494e-06,
+      "loss": 0.5425,
+      "step": 35165
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3526898630434847,
+      "learning_rate": 1.5035991641567803e-06,
+      "loss": 0.7259,
+      "step": 35166
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2930414404241368,
+      "learning_rate": 1.5031967892727494e-06,
+      "loss": 0.5281,
+      "step": 35167
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2377638264264244,
+      "learning_rate": 1.5027944638597947e-06,
+      "loss": 0.648,
+      "step": 35168
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6781179135560766,
+      "learning_rate": 1.502392187920263e-06,
+      "loss": 0.7675,
+      "step": 35169
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3979655271144167,
+      "learning_rate": 1.5019899614564937e-06,
+      "loss": 0.691,
+      "step": 35170
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7337997421372253,
+      "learning_rate": 1.5015877844708272e-06,
+      "loss": 0.684,
+      "step": 35171
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4376383509984998,
+      "learning_rate": 1.501185656965607e-06,
+      "loss": 0.7624,
+      "step": 35172
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5882191425039169,
+      "learning_rate": 1.500783578943177e-06,
+      "loss": 0.4926,
+      "step": 35173
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4856670618648675,
+      "learning_rate": 1.500381550405876e-06,
+      "loss": 0.6743,
+      "step": 35174
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5873198400138329,
+      "learning_rate": 1.4999795713560428e-06,
+      "loss": 0.7682,
+      "step": 35175
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6957280665123355,
+      "learning_rate": 1.4995776417960184e-06,
+      "loss": 0.5246,
+      "step": 35176
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5975719471888483,
+      "learning_rate": 1.4991757617281466e-06,
+      "loss": 0.7518,
+      "step": 35177
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5535215532348852,
+      "learning_rate": 1.4987739311547656e-06,
+      "loss": 0.7877,
+      "step": 35178
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5750055611305185,
+      "learning_rate": 1.4983721500782122e-06,
+      "loss": 0.6372,
+      "step": 35179
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3046658260252475,
+      "learning_rate": 1.497970418500827e-06,
+      "loss": 0.6642,
+      "step": 35180
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.569945644571215,
+      "learning_rate": 1.4975687364249513e-06,
+      "loss": 0.7666,
+      "step": 35181
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8234678866946143,
+      "learning_rate": 1.4971671038529235e-06,
+      "loss": 0.7322,
+      "step": 35182
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.337667702356066,
+      "learning_rate": 1.496765520787078e-06,
+      "loss": 0.5818,
+      "step": 35183
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4867996357866928,
+      "learning_rate": 1.4963639872297586e-06,
+      "loss": 0.7547,
+      "step": 35184
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.396747983954094,
+      "learning_rate": 1.4959625031832969e-06,
+      "loss": 0.7535,
+      "step": 35185
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.245233125010581,
+      "learning_rate": 1.4955610686500365e-06,
+      "loss": 0.633,
+      "step": 35186
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1776137643114821,
+      "learning_rate": 1.4951596836323102e-06,
+      "loss": 0.604,
+      "step": 35187
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5321556314923739,
+      "learning_rate": 1.494758348132459e-06,
+      "loss": 0.6701,
+      "step": 35188
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4533585584741622,
+      "learning_rate": 1.4943570621528148e-06,
+      "loss": 0.7953,
+      "step": 35189
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5617071878924107,
+      "learning_rate": 1.4939558256957166e-06,
+      "loss": 0.5718,
+      "step": 35190
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4682185762068791,
+      "learning_rate": 1.4935546387635047e-06,
+      "loss": 0.8032,
+      "step": 35191
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5346766591034635,
+      "learning_rate": 1.4931535013585075e-06,
+      "loss": 0.5642,
+      "step": 35192
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.304864814568714,
+      "learning_rate": 1.4927524134830629e-06,
+      "loss": 0.6423,
+      "step": 35193
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8824006726356841,
+      "learning_rate": 1.492351375139507e-06,
+      "loss": 0.6168,
+      "step": 35194
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1806268386345313,
+      "learning_rate": 1.4919503863301799e-06,
+      "loss": 0.6593,
+      "step": 35195
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.497264525856047,
+      "learning_rate": 1.4915494470574065e-06,
+      "loss": 0.6766,
+      "step": 35196
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6185230705604587,
+      "learning_rate": 1.4911485573235262e-06,
+      "loss": 0.8041,
+      "step": 35197
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5403683902142267,
+      "learning_rate": 1.4907477171308727e-06,
+      "loss": 0.7617,
+      "step": 35198
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6419450744262807,
+      "learning_rate": 1.4903469264817838e-06,
+      "loss": 0.7887,
+      "step": 35199
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.372775838095924,
+      "learning_rate": 1.4899461853785857e-06,
+      "loss": 0.5985,
+      "step": 35200
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.517944941973057,
+      "learning_rate": 1.4895454938236142e-06,
+      "loss": 0.6714,
+      "step": 35201
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.272178737725724,
+      "learning_rate": 1.4891448518192065e-06,
+      "loss": 0.7322,
+      "step": 35202
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2912956720276287,
+      "learning_rate": 1.4887442593676893e-06,
+      "loss": 0.6344,
+      "step": 35203
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.331847448193885,
+      "learning_rate": 1.4883437164713999e-06,
+      "loss": 0.6588,
+      "step": 35204
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3541384712245619,
+      "learning_rate": 1.4879432231326663e-06,
+      "loss": 0.7144,
+      "step": 35205
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2763220534105546,
+      "learning_rate": 1.4875427793538233e-06,
+      "loss": 0.6876,
+      "step": 35206
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3802918031582994,
+      "learning_rate": 1.487142385137199e-06,
+      "loss": 0.6666,
+      "step": 35207
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3051618522971307,
+      "learning_rate": 1.4867420404851306e-06,
+      "loss": 0.7397,
+      "step": 35208
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4234590508489697,
+      "learning_rate": 1.486341745399944e-06,
+      "loss": 0.652,
+      "step": 35209
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.662579228028427,
+      "learning_rate": 1.4859414998839694e-06,
+      "loss": 0.7296,
+      "step": 35210
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.0758406739241435,
+      "learning_rate": 1.4855413039395383e-06,
+      "loss": 0.6064,
+      "step": 35211
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5430475184417236,
+      "learning_rate": 1.4851411575689845e-06,
+      "loss": 0.676,
+      "step": 35212
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4152546056852655,
+      "learning_rate": 1.4847410607746337e-06,
+      "loss": 0.663,
+      "step": 35213
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.19067068055073,
+      "learning_rate": 1.4843410135588155e-06,
+      "loss": 0.5824,
+      "step": 35214
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3663445020936171,
+      "learning_rate": 1.4839410159238587e-06,
+      "loss": 0.6552,
+      "step": 35215
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8763423700600856,
+      "learning_rate": 1.4835410678720963e-06,
+      "loss": 0.7313,
+      "step": 35216
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3161994473685428,
+      "learning_rate": 1.483141169405854e-06,
+      "loss": 0.6593,
+      "step": 35217
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.241961364077069,
+      "learning_rate": 1.4827413205274565e-06,
+      "loss": 0.7818,
+      "step": 35218
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5903792289946894,
+      "learning_rate": 1.4823415212392378e-06,
+      "loss": 0.4884,
+      "step": 35219
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4275871737414072,
+      "learning_rate": 1.4819417715435247e-06,
+      "loss": 0.6279,
+      "step": 35220
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.32043053843812,
+      "learning_rate": 1.481542071442643e-06,
+      "loss": 0.6225,
+      "step": 35221
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4106768649859622,
+      "learning_rate": 1.481142420938918e-06,
+      "loss": 0.7288,
+      "step": 35222
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8032012065606056,
+      "learning_rate": 1.4807428200346817e-06,
+      "loss": 0.7065,
+      "step": 35223
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4079261151768863,
+      "learning_rate": 1.480343268732255e-06,
+      "loss": 0.7139,
+      "step": 35224
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3490344664470053,
+      "learning_rate": 1.4799437670339701e-06,
+      "loss": 0.5738,
+      "step": 35225
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4981427621468875,
+      "learning_rate": 1.4795443149421473e-06,
+      "loss": 0.7172,
+      "step": 35226
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1623369092172795,
+      "learning_rate": 1.4791449124591173e-06,
+      "loss": 0.6481,
+      "step": 35227
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8388812625622872,
+      "learning_rate": 1.4787455595872002e-06,
+      "loss": 0.7047,
+      "step": 35228
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1469798020545334,
+      "learning_rate": 1.4783462563287277e-06,
+      "loss": 0.5111,
+      "step": 35229
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3315016119648397,
+      "learning_rate": 1.477947002686021e-06,
+      "loss": 0.7713,
+      "step": 35230
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3704389042680698,
+      "learning_rate": 1.477547798661403e-06,
+      "loss": 0.6671,
+      "step": 35231
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2083103471636456,
+      "learning_rate": 1.477148644257198e-06,
+      "loss": 0.624,
+      "step": 35232
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2883457556889872,
+      "learning_rate": 1.4767495394757359e-06,
+      "loss": 0.652,
+      "step": 35233
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4293243910051876,
+      "learning_rate": 1.4763504843193366e-06,
+      "loss": 0.6355,
+      "step": 35234
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3607505016779058,
+      "learning_rate": 1.4759514787903195e-06,
+      "loss": 0.5593,
+      "step": 35235
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1989114047691471,
+      "learning_rate": 1.4755525228910128e-06,
+      "loss": 0.5339,
+      "step": 35236
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3135019391262537,
+      "learning_rate": 1.4751536166237402e-06,
+      "loss": 0.6688,
+      "step": 35237
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3984372915814387,
+      "learning_rate": 1.4747547599908218e-06,
+      "loss": 0.7094,
+      "step": 35238
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2770455202219968,
+      "learning_rate": 1.4743559529945784e-06,
+      "loss": 0.586,
+      "step": 35239
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6048973575498353,
+      "learning_rate": 1.4739571956373333e-06,
+      "loss": 0.7621,
+      "step": 35240
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4077912858822417,
+      "learning_rate": 1.4735584879214116e-06,
+      "loss": 0.6757,
+      "step": 35241
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3078638611919953,
+      "learning_rate": 1.4731598298491313e-06,
+      "loss": 0.7234,
+      "step": 35242
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3721002910709932,
+      "learning_rate": 1.4727612214228115e-06,
+      "loss": 0.7224,
+      "step": 35243
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3813456786659173,
+      "learning_rate": 1.4723626626447784e-06,
+      "loss": 0.7556,
+      "step": 35244
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8052086093180297,
+      "learning_rate": 1.471964153517348e-06,
+      "loss": 0.673,
+      "step": 35245
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2656415477348073,
+      "learning_rate": 1.4715656940428437e-06,
+      "loss": 0.6403,
+      "step": 35246
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2064493613761744,
+      "learning_rate": 1.4711672842235814e-06,
+      "loss": 0.6275,
+      "step": 35247
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.462311005705278,
+      "learning_rate": 1.470768924061886e-06,
+      "loss": 0.7344,
+      "step": 35248
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4443607120548447,
+      "learning_rate": 1.4703706135600715e-06,
+      "loss": 0.5528,
+      "step": 35249
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5370141850537056,
+      "learning_rate": 1.4699723527204623e-06,
+      "loss": 0.7481,
+      "step": 35250
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3984903244512064,
+      "learning_rate": 1.4695741415453757e-06,
+      "loss": 0.7777,
+      "step": 35251
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5213148899530404,
+      "learning_rate": 1.4691759800371252e-06,
+      "loss": 0.8136,
+      "step": 35252
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.274838290672348,
+      "learning_rate": 1.4687778681980335e-06,
+      "loss": 0.5184,
+      "step": 35253
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4769824335347677,
+      "learning_rate": 1.468379806030421e-06,
+      "loss": 0.7132,
+      "step": 35254
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5253347846905867,
+      "learning_rate": 1.4679817935366014e-06,
+      "loss": 0.6586,
+      "step": 35255
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4309025186855515,
+      "learning_rate": 1.4675838307188916e-06,
+      "loss": 0.7132,
+      "step": 35256
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5878262054877255,
+      "learning_rate": 1.4671859175796087e-06,
+      "loss": 0.6788,
+      "step": 35257
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6706733672735588,
+      "learning_rate": 1.4667880541210744e-06,
+      "loss": 0.8022,
+      "step": 35258
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6353512357895612,
+      "learning_rate": 1.4663902403456009e-06,
+      "loss": 0.6224,
+      "step": 35259
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6498678123487813,
+      "learning_rate": 1.4659924762555033e-06,
+      "loss": 0.7919,
+      "step": 35260
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.602709279966796,
+      "learning_rate": 1.4655947618530998e-06,
+      "loss": 0.6692,
+      "step": 35261
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2456072870758643,
+      "learning_rate": 1.465197097140707e-06,
+      "loss": 0.584,
+      "step": 35262
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4425910221826332,
+      "learning_rate": 1.464799482120639e-06,
+      "loss": 0.6696,
+      "step": 35263
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2928778456131744,
+      "learning_rate": 1.4644019167952084e-06,
+      "loss": 0.6854,
+      "step": 35264
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4134788292816838,
+      "learning_rate": 1.4640044011667343e-06,
+      "loss": 0.6113,
+      "step": 35265
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5755359565033548,
+      "learning_rate": 1.463606935237526e-06,
+      "loss": 0.4678,
+      "step": 35266
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2250923669915357,
+      "learning_rate": 1.4632095190099039e-06,
+      "loss": 0.6742,
+      "step": 35267
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2855552926141756,
+      "learning_rate": 1.462812152486176e-06,
+      "loss": 0.7025,
+      "step": 35268
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4162650662876843,
+      "learning_rate": 1.4624148356686619e-06,
+      "loss": 0.6859,
+      "step": 35269
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4236562321416584,
+      "learning_rate": 1.4620175685596683e-06,
+      "loss": 0.5978,
+      "step": 35270
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4628657780278038,
+      "learning_rate": 1.4616203511615134e-06,
+      "loss": 0.7789,
+      "step": 35271
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6562341852588616,
+      "learning_rate": 1.461223183476509e-06,
+      "loss": 0.7596,
+      "step": 35272
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5802577918541512,
+      "learning_rate": 1.4608260655069639e-06,
+      "loss": 0.6893,
+      "step": 35273
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4091181925046967,
+      "learning_rate": 1.4604289972551933e-06,
+      "loss": 0.761,
+      "step": 35274
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.404201586923984,
+      "learning_rate": 1.4600319787235073e-06,
+      "loss": 0.5966,
+      "step": 35275
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.556202742336064,
+      "learning_rate": 1.4596350099142243e-06,
+      "loss": 0.8549,
+      "step": 35276
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4294806053996176,
+      "learning_rate": 1.4592380908296455e-06,
+      "loss": 0.6526,
+      "step": 35277
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2479386837410547,
+      "learning_rate": 1.458841221472087e-06,
+      "loss": 0.5587,
+      "step": 35278
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3969918586262213,
+      "learning_rate": 1.4584444018438616e-06,
+      "loss": 0.6132,
+      "step": 35279
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2569839405725542,
+      "learning_rate": 1.4580476319472758e-06,
+      "loss": 0.6307,
+      "step": 35280
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4280420518304096,
+      "learning_rate": 1.4576509117846394e-06,
+      "loss": 0.7449,
+      "step": 35281
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.598890874776287,
+      "learning_rate": 1.4572542413582635e-06,
+      "loss": 0.6983,
+      "step": 35282
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4151048949569391,
+      "learning_rate": 1.456857620670461e-06,
+      "loss": 0.741,
+      "step": 35283
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3093097040961539,
+      "learning_rate": 1.4564610497235377e-06,
+      "loss": 0.6398,
+      "step": 35284
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3396099301939655,
+      "learning_rate": 1.4560645285198006e-06,
+      "loss": 0.689,
+      "step": 35285
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.668351774657109,
+      "learning_rate": 1.4556680570615623e-06,
+      "loss": 0.6626,
+      "step": 35286
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4255117824433237,
+      "learning_rate": 1.4552716353511286e-06,
+      "loss": 0.6494,
+      "step": 35287
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5780569954573741,
+      "learning_rate": 1.454875263390808e-06,
+      "loss": 0.4942,
+      "step": 35288
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0543964769643233,
+      "learning_rate": 1.4544789411829118e-06,
+      "loss": 0.6941,
+      "step": 35289
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3829086289745227,
+      "learning_rate": 1.454082668729745e-06,
+      "loss": 0.707,
+      "step": 35290
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.422489298771386,
+      "learning_rate": 1.4536864460336108e-06,
+      "loss": 0.6402,
+      "step": 35291
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.448699395563213,
+      "learning_rate": 1.4532902730968213e-06,
+      "loss": 0.659,
+      "step": 35292
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.603615996918114,
+      "learning_rate": 1.4528941499216853e-06,
+      "loss": 0.7023,
+      "step": 35293
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.067153009040792,
+      "learning_rate": 1.4524980765105013e-06,
+      "loss": 0.5911,
+      "step": 35294
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6373859648637696,
+      "learning_rate": 1.4521020528655794e-06,
+      "loss": 0.6991,
+      "step": 35295
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5853220004456878,
+      "learning_rate": 1.4517060789892257e-06,
+      "loss": 0.612,
+      "step": 35296
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.647878814164669,
+      "learning_rate": 1.4513101548837505e-06,
+      "loss": 0.6477,
+      "step": 35297
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2905513803338762,
+      "learning_rate": 1.4509142805514486e-06,
+      "loss": 0.5063,
+      "step": 35298
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.616842253479958,
+      "learning_rate": 1.4505184559946306e-06,
+      "loss": 0.795,
+      "step": 35299
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5818967332368128,
+      "learning_rate": 1.4501226812156032e-06,
+      "loss": 0.5186,
+      "step": 35300
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5168732993064495,
+      "learning_rate": 1.4497269562166672e-06,
+      "loss": 0.5953,
+      "step": 35301
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4223356551291704,
+      "learning_rate": 1.4493312810001293e-06,
+      "loss": 0.7617,
+      "step": 35302
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.265045351612049,
+      "learning_rate": 1.4489356555682887e-06,
+      "loss": 0.6733,
+      "step": 35303
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3187663562857124,
+      "learning_rate": 1.448540079923455e-06,
+      "loss": 0.717,
+      "step": 35304
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6615752145741749,
+      "learning_rate": 1.4481445540679261e-06,
+      "loss": 0.4773,
+      "step": 35305
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0097586903076508,
+      "learning_rate": 1.4477490780040093e-06,
+      "loss": 0.6789,
+      "step": 35306
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6697487919666854,
+      "learning_rate": 1.4473536517340047e-06,
+      "loss": 0.5403,
+      "step": 35307
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4924338826667036,
+      "learning_rate": 1.446958275260213e-06,
+      "loss": 0.5918,
+      "step": 35308
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7122862118294204,
+      "learning_rate": 1.4465629485849376e-06,
+      "loss": 0.6634,
+      "step": 35309
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5142917472850645,
+      "learning_rate": 1.4461676717104834e-06,
+      "loss": 0.7104,
+      "step": 35310
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.510352703180472,
+      "learning_rate": 1.4457724446391486e-06,
+      "loss": 0.6364,
+      "step": 35311
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2530883121431002,
+      "learning_rate": 1.4453772673732336e-06,
+      "loss": 0.5852,
+      "step": 35312
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4342887295137572,
+      "learning_rate": 1.4449821399150398e-06,
+      "loss": 0.6583,
+      "step": 35313
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6158279832216875,
+      "learning_rate": 1.44458706226687e-06,
+      "loss": 0.7504,
+      "step": 35314
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4263139048199873,
+      "learning_rate": 1.4441920344310235e-06,
+      "loss": 0.6905,
+      "step": 35315
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.635631265322193,
+      "learning_rate": 1.443797056409798e-06,
+      "loss": 0.6779,
+      "step": 35316
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4206792243553528,
+      "learning_rate": 1.4434021282054943e-06,
+      "loss": 0.8157,
+      "step": 35317
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3802997155427168,
+      "learning_rate": 1.443007249820415e-06,
+      "loss": 0.7058,
+      "step": 35318
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6307762368041854,
+      "learning_rate": 1.442612421256856e-06,
+      "loss": 0.5805,
+      "step": 35319
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.589465415630185,
+      "learning_rate": 1.4422176425171153e-06,
+      "loss": 0.6793,
+      "step": 35320
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3039239577679638,
+      "learning_rate": 1.4418229136034922e-06,
+      "loss": 0.5566,
+      "step": 35321
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.848197892738636,
+      "learning_rate": 1.4414282345182885e-06,
+      "loss": 0.8236,
+      "step": 35322
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4287103895690214,
+      "learning_rate": 1.4410336052637985e-06,
+      "loss": 0.5083,
+      "step": 35323
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.273910846966268,
+      "learning_rate": 1.4406390258423186e-06,
+      "loss": 0.5263,
+      "step": 35324
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.967472637610098,
+      "learning_rate": 1.4402444962561502e-06,
+      "loss": 0.666,
+      "step": 35325
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.379542976489628,
+      "learning_rate": 1.4398500165075857e-06,
+      "loss": 0.6741,
+      "step": 35326
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5135039095016163,
+      "learning_rate": 1.4394555865989279e-06,
+      "loss": 0.6921,
+      "step": 35327
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5598303448221381,
+      "learning_rate": 1.439061206532466e-06,
+      "loss": 0.4745,
+      "step": 35328
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4196294460731373,
+      "learning_rate": 1.4386668763105039e-06,
+      "loss": 0.6208,
+      "step": 35329
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3066432166857882,
+      "learning_rate": 1.4382725959353305e-06,
+      "loss": 0.6092,
+      "step": 35330
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3398146089961616,
+      "learning_rate": 1.4378783654092476e-06,
+      "loss": 0.7509,
+      "step": 35331
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5942778944063938,
+      "learning_rate": 1.4374841847345478e-06,
+      "loss": 0.6433,
+      "step": 35332
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3666768349295757,
+      "learning_rate": 1.4370900539135225e-06,
+      "loss": 0.6928,
+      "step": 35333
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2584565196997042,
+      "learning_rate": 1.4366959729484708e-06,
+      "loss": 0.665,
+      "step": 35334
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6014195753701265,
+      "learning_rate": 1.4363019418416878e-06,
+      "loss": 0.518,
+      "step": 35335
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5538768213240686,
+      "learning_rate": 1.4359079605954673e-06,
+      "loss": 0.6414,
+      "step": 35336
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0478406246489644,
+      "learning_rate": 1.4355140292120983e-06,
+      "loss": 0.7462,
+      "step": 35337
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5792475387875111,
+      "learning_rate": 1.4351201476938792e-06,
+      "loss": 0.5002,
+      "step": 35338
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6663713427798632,
+      "learning_rate": 1.4347263160431036e-06,
+      "loss": 0.7481,
+      "step": 35339
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5805611265904326,
+      "learning_rate": 1.434332534262064e-06,
+      "loss": 0.4556,
+      "step": 35340
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.474646906363261,
+      "learning_rate": 1.4339388023530498e-06,
+      "loss": 0.8218,
+      "step": 35341
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3904988490840071,
+      "learning_rate": 1.433545120318355e-06,
+      "loss": 0.6955,
+      "step": 35342
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6205821424866733,
+      "learning_rate": 1.4331514881602759e-06,
+      "loss": 0.6136,
+      "step": 35343
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0700287448098686,
+      "learning_rate": 1.4327579058811003e-06,
+      "loss": 0.7537,
+      "step": 35344
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3957853971125742,
+      "learning_rate": 1.4323643734831182e-06,
+      "loss": 0.7896,
+      "step": 35345
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5142432514942314,
+      "learning_rate": 1.4319708909686259e-06,
+      "loss": 0.7354,
+      "step": 35346
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3601564361990262,
+      "learning_rate": 1.43157745833991e-06,
+      "loss": 0.6444,
+      "step": 35347
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7246177897141368,
+      "learning_rate": 1.4311840755992634e-06,
+      "loss": 0.6246,
+      "step": 35348
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2407229016822459,
+      "learning_rate": 1.4307907427489753e-06,
+      "loss": 0.8349,
+      "step": 35349
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3652240003521963,
+      "learning_rate": 1.4303974597913383e-06,
+      "loss": 0.5867,
+      "step": 35350
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5107033097257163,
+      "learning_rate": 1.4300042267286374e-06,
+      "loss": 0.619,
+      "step": 35351
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.338941723729003,
+      "learning_rate": 1.4296110435631683e-06,
+      "loss": 0.641,
+      "step": 35352
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6450412641428722,
+      "learning_rate": 1.429217910297217e-06,
+      "loss": 0.5552,
+      "step": 35353
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2467423691706065,
+      "learning_rate": 1.4288248269330695e-06,
+      "loss": 0.6286,
+      "step": 35354
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.381114369875806,
+      "learning_rate": 1.4284317934730175e-06,
+      "loss": 0.5958,
+      "step": 35355
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.562468348933186,
+      "learning_rate": 1.428038809919352e-06,
+      "loss": 0.7492,
+      "step": 35356
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.193258131759366,
+      "learning_rate": 1.4276458762743583e-06,
+      "loss": 0.7006,
+      "step": 35357
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.695919623581575,
+      "learning_rate": 1.4272529925403221e-06,
+      "loss": 0.7344,
+      "step": 35358
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9937406873358219,
+      "learning_rate": 1.426860158719533e-06,
+      "loss": 0.6621,
+      "step": 35359
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3078276555512083,
+      "learning_rate": 1.426467374814281e-06,
+      "loss": 0.602,
+      "step": 35360
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.807445322322607,
+      "learning_rate": 1.426074640826849e-06,
+      "loss": 0.6548,
+      "step": 35361
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2455755073707813,
+      "learning_rate": 1.4256819567595238e-06,
+      "loss": 0.6159,
+      "step": 35362
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9680964589696242,
+      "learning_rate": 1.4252893226145924e-06,
+      "loss": 0.5581,
+      "step": 35363
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.27843699465053,
+      "learning_rate": 1.4248967383943435e-06,
+      "loss": 0.7595,
+      "step": 35364
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.718301670784334,
+      "learning_rate": 1.4245042041010604e-06,
+      "loss": 0.72,
+      "step": 35365
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3120389300400475,
+      "learning_rate": 1.424111719737027e-06,
+      "loss": 0.7182,
+      "step": 35366
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.61278109052153,
+      "learning_rate": 1.4237192853045323e-06,
+      "loss": 0.4987,
+      "step": 35367
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.624474129412666,
+      "learning_rate": 1.423326900805857e-06,
+      "loss": 0.8578,
+      "step": 35368
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3440336798591215,
+      "learning_rate": 1.4229345662432903e-06,
+      "loss": 0.5808,
+      "step": 35369
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5787730372865674,
+      "learning_rate": 1.4225422816191115e-06,
+      "loss": 0.7329,
+      "step": 35370
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4984404465455698,
+      "learning_rate": 1.4221500469356086e-06,
+      "loss": 0.5997,
+      "step": 35371
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.269092664324896,
+      "learning_rate": 1.4217578621950623e-06,
+      "loss": 0.7259,
+      "step": 35372
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5515202324332367,
+      "learning_rate": 1.4213657273997571e-06,
+      "loss": 0.7185,
+      "step": 35373
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.853197520036338,
+      "learning_rate": 1.4209736425519827e-06,
+      "loss": 0.6981,
+      "step": 35374
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4744191684363195,
+      "learning_rate": 1.420581607654009e-06,
+      "loss": 0.6412,
+      "step": 35375
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3694315621758406,
+      "learning_rate": 1.420189622708127e-06,
+      "loss": 0.812,
+      "step": 35376
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4336784547623442,
+      "learning_rate": 1.4197976877166165e-06,
+      "loss": 0.6608,
+      "step": 35377
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.390672778245741,
+      "learning_rate": 1.4194058026817658e-06,
+      "loss": 0.6713,
+      "step": 35378
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4604521622220785,
+      "learning_rate": 1.4190139676058457e-06,
+      "loss": 0.6246,
+      "step": 35379
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.479154474165393,
+      "learning_rate": 1.4186221824911428e-06,
+      "loss": 0.6689,
+      "step": 35380
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7392901941035475,
+      "learning_rate": 1.418230447339941e-06,
+      "loss": 0.6698,
+      "step": 35381
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4167895405086888,
+      "learning_rate": 1.4178387621545187e-06,
+      "loss": 0.5594,
+      "step": 35382
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.309808569416757,
+      "learning_rate": 1.4174471269371537e-06,
+      "loss": 0.6747,
+      "step": 35383
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3425801700958893,
+      "learning_rate": 1.417055541690129e-06,
+      "loss": 0.5707,
+      "step": 35384
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.266477784913992,
+      "learning_rate": 1.4166640064157267e-06,
+      "loss": 0.7526,
+      "step": 35385
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.112924280136889,
+      "learning_rate": 1.416272521116221e-06,
+      "loss": 0.5876,
+      "step": 35386
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.112111573947319,
+      "learning_rate": 1.4158810857938964e-06,
+      "loss": 0.6664,
+      "step": 35387
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7075044484044355,
+      "learning_rate": 1.4154897004510304e-06,
+      "loss": 0.8018,
+      "step": 35388
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5788200283557247,
+      "learning_rate": 1.4150983650898987e-06,
+      "loss": 0.515,
+      "step": 35389
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.383220703658127,
+      "learning_rate": 1.4147070797127826e-06,
+      "loss": 0.7457,
+      "step": 35390
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6767118337162186,
+      "learning_rate": 1.4143158443219617e-06,
+      "loss": 0.7028,
+      "step": 35391
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.818776664260503,
+      "learning_rate": 1.413924658919712e-06,
+      "loss": 0.6123,
+      "step": 35392
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5029498555005742,
+      "learning_rate": 1.4135335235083092e-06,
+      "loss": 0.682,
+      "step": 35393
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3008343468996928,
+      "learning_rate": 1.4131424380900327e-06,
+      "loss": 0.7475,
+      "step": 35394
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3918502911381831,
+      "learning_rate": 1.412751402667164e-06,
+      "loss": 0.7461,
+      "step": 35395
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1465534913438273,
+      "learning_rate": 1.4123604172419714e-06,
+      "loss": 0.7401,
+      "step": 35396
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4709284181099653,
+      "learning_rate": 1.4119694818167351e-06,
+      "loss": 0.6089,
+      "step": 35397
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7005246860331333,
+      "learning_rate": 1.4115785963937302e-06,
+      "loss": 0.659,
+      "step": 35398
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.572273917020825,
+      "learning_rate": 1.4111877609752367e-06,
+      "loss": 0.6893,
+      "step": 35399
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3911557615158487,
+      "learning_rate": 1.4107969755635277e-06,
+      "loss": 0.6884,
+      "step": 35400
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.610162612653663,
+      "learning_rate": 1.4104062401608753e-06,
+      "loss": 0.4897,
+      "step": 35401
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4597136021761907,
+      "learning_rate": 1.4100155547695592e-06,
+      "loss": 0.6655,
+      "step": 35402
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4603724532835245,
+      "learning_rate": 1.4096249193918509e-06,
+      "loss": 0.6739,
+      "step": 35403
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3199819609102825,
+      "learning_rate": 1.4092343340300273e-06,
+      "loss": 0.6482,
+      "step": 35404
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2449295905383286,
+      "learning_rate": 1.408843798686359e-06,
+      "loss": 0.6941,
+      "step": 35405
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3292906097769046,
+      "learning_rate": 1.408453313363124e-06,
+      "loss": 0.6144,
+      "step": 35406
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.314363536641934,
+      "learning_rate": 1.4080628780625927e-06,
+      "loss": 0.6486,
+      "step": 35407
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4614435695933503,
+      "learning_rate": 1.407672492787041e-06,
+      "loss": 0.5637,
+      "step": 35408
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3297481718183062,
+      "learning_rate": 1.4072821575387407e-06,
+      "loss": 0.628,
+      "step": 35409
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3513100496112427,
+      "learning_rate": 1.406891872319961e-06,
+      "loss": 0.6689,
+      "step": 35410
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5963033926936947,
+      "learning_rate": 1.4065016371329775e-06,
+      "loss": 0.6663,
+      "step": 35411
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4343692898863898,
+      "learning_rate": 1.4061114519800655e-06,
+      "loss": 0.6718,
+      "step": 35412
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2551203012475667,
+      "learning_rate": 1.405721316863492e-06,
+      "loss": 0.6646,
+      "step": 35413
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3828530808649342,
+      "learning_rate": 1.4053312317855284e-06,
+      "loss": 0.6554,
+      "step": 35414
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.503616015667983,
+      "learning_rate": 1.4049411967484472e-06,
+      "loss": 0.6815,
+      "step": 35415
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.581208935940438,
+      "learning_rate": 1.4045512117545224e-06,
+      "loss": 0.6497,
+      "step": 35416
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4340443683622497,
+      "learning_rate": 1.4041612768060208e-06,
+      "loss": 0.7927,
+      "step": 35417
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.398702590153426,
+      "learning_rate": 1.4037713919052121e-06,
+      "loss": 0.7471,
+      "step": 35418
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4078614114979067,
+      "learning_rate": 1.4033815570543673e-06,
+      "loss": 0.7025,
+      "step": 35419
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6307981091532587,
+      "learning_rate": 1.4029917722557597e-06,
+      "loss": 0.5258,
+      "step": 35420
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.687812775879044,
+      "learning_rate": 1.4026020375116545e-06,
+      "loss": 0.6051,
+      "step": 35421
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2809424578323605,
+      "learning_rate": 1.4022123528243204e-06,
+      "loss": 0.7286,
+      "step": 35422
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2661165030688903,
+      "learning_rate": 1.4018227181960308e-06,
+      "loss": 0.5216,
+      "step": 35423
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5991205520737218,
+      "learning_rate": 1.4014331336290487e-06,
+      "loss": 0.5278,
+      "step": 35424
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.654890193430886,
+      "learning_rate": 1.4010435991256477e-06,
+      "loss": 0.6266,
+      "step": 35425
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3855123672606529,
+      "learning_rate": 1.4006541146880914e-06,
+      "loss": 0.7187,
+      "step": 35426
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.67939883565749,
+      "learning_rate": 1.4002646803186514e-06,
+      "loss": 0.7252,
+      "step": 35427
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6171236178201558,
+      "learning_rate": 1.3998752960195905e-06,
+      "loss": 0.6919,
+      "step": 35428
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4530963340189877,
+      "learning_rate": 1.39948596179318e-06,
+      "loss": 0.7593,
+      "step": 35429
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2621769732958188,
+      "learning_rate": 1.3990966776416859e-06,
+      "loss": 0.6201,
+      "step": 35430
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3306250820652556,
+      "learning_rate": 1.398707443567372e-06,
+      "loss": 0.7743,
+      "step": 35431
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6152717653830772,
+      "learning_rate": 1.3983182595725054e-06,
+      "loss": 0.4919,
+      "step": 35432
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6143514729571151,
+      "learning_rate": 1.3979291256593564e-06,
+      "loss": 0.5443,
+      "step": 35433
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3765341825035615,
+      "learning_rate": 1.3975400418301866e-06,
+      "loss": 0.6096,
+      "step": 35434
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3067553036697632,
+      "learning_rate": 1.3971510080872608e-06,
+      "loss": 0.6196,
+      "step": 35435
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.394694524736466,
+      "learning_rate": 1.396762024432845e-06,
+      "loss": 0.7927,
+      "step": 35436
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.411853203853346,
+      "learning_rate": 1.3963730908692065e-06,
+      "loss": 0.6805,
+      "step": 35437
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.292478802746188,
+      "learning_rate": 1.3959842073986085e-06,
+      "loss": 0.7906,
+      "step": 35438
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5297944512645716,
+      "learning_rate": 1.3955953740233107e-06,
+      "loss": 0.7774,
+      "step": 35439
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4435044454955994,
+      "learning_rate": 1.3952065907455825e-06,
+      "loss": 0.8886,
+      "step": 35440
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6623230973780987,
+      "learning_rate": 1.3948178575676862e-06,
+      "loss": 0.6775,
+      "step": 35441
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5711896969976338,
+      "learning_rate": 1.394429174491886e-06,
+      "loss": 0.478,
+      "step": 35442
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.488535561565158,
+      "learning_rate": 1.3940405415204417e-06,
+      "loss": 0.8168,
+      "step": 35443
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5674037387944934,
+      "learning_rate": 1.3936519586556163e-06,
+      "loss": 0.5237,
+      "step": 35444
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3796407584578945,
+      "learning_rate": 1.393263425899678e-06,
+      "loss": 0.7222,
+      "step": 35445
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2899204042852004,
+      "learning_rate": 1.3928749432548837e-06,
+      "loss": 0.7095,
+      "step": 35446
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2920891712494171,
+      "learning_rate": 1.3924865107234942e-06,
+      "loss": 0.7819,
+      "step": 35447
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3154792019098822,
+      "learning_rate": 1.3920981283077762e-06,
+      "loss": 0.6233,
+      "step": 35448
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5248197256380205,
+      "learning_rate": 1.3917097960099856e-06,
+      "loss": 0.6559,
+      "step": 35449
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4570320246204793,
+      "learning_rate": 1.3913215138323877e-06,
+      "loss": 0.7492,
+      "step": 35450
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2428611516639165,
+      "learning_rate": 1.3909332817772402e-06,
+      "loss": 0.68,
+      "step": 35451
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4896483012865391,
+      "learning_rate": 1.3905450998468062e-06,
+      "loss": 0.7484,
+      "step": 35452
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2685793736774444,
+      "learning_rate": 1.3901569680433425e-06,
+      "loss": 0.7399,
+      "step": 35453
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4256663241020813,
+      "learning_rate": 1.3897688863691128e-06,
+      "loss": 0.6834,
+      "step": 35454
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4437768552018888,
+      "learning_rate": 1.3893808548263742e-06,
+      "loss": 0.6768,
+      "step": 35455
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5072474927601054,
+      "learning_rate": 1.388992873417384e-06,
+      "loss": 0.6689,
+      "step": 35456
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3074004072800025,
+      "learning_rate": 1.3886049421444048e-06,
+      "loss": 0.7118,
+      "step": 35457
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3554690005284333,
+      "learning_rate": 1.3882170610096946e-06,
+      "loss": 0.5322,
+      "step": 35458
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.585280452552238,
+      "learning_rate": 1.3878292300155115e-06,
+      "loss": 0.6839,
+      "step": 35459
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6004466473416828,
+      "learning_rate": 1.387441449164112e-06,
+      "loss": 0.4767,
+      "step": 35460
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2550491456783528,
+      "learning_rate": 1.3870537184577538e-06,
+      "loss": 0.7512,
+      "step": 35461
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3648364047645802,
+      "learning_rate": 1.3866660378986985e-06,
+      "loss": 0.5419,
+      "step": 35462
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5979655585463038,
+      "learning_rate": 1.386278407489201e-06,
+      "loss": 0.5106,
+      "step": 35463
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2927961920424158,
+      "learning_rate": 1.3858908272315153e-06,
+      "loss": 0.581,
+      "step": 35464
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3306754419618176,
+      "learning_rate": 1.3855032971278991e-06,
+      "loss": 0.7423,
+      "step": 35465
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6978315230773253,
+      "learning_rate": 1.3851158171806134e-06,
+      "loss": 0.6726,
+      "step": 35466
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3999000815679123,
+      "learning_rate": 1.3847283873919103e-06,
+      "loss": 0.729,
+      "step": 35467
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5503449623332644,
+      "learning_rate": 1.384341007764044e-06,
+      "loss": 0.6419,
+      "step": 35468
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4060389805480429,
+      "learning_rate": 1.3839536782992746e-06,
+      "loss": 0.7338,
+      "step": 35469
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4395844105765456,
+      "learning_rate": 1.3835663989998528e-06,
+      "loss": 0.6182,
+      "step": 35470
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5264302431925492,
+      "learning_rate": 1.3831791698680342e-06,
+      "loss": 0.6243,
+      "step": 35471
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5651818327274489,
+      "learning_rate": 1.3827919909060772e-06,
+      "loss": 0.5145,
+      "step": 35472
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.674671951211041,
+      "learning_rate": 1.3824048621162334e-06,
+      "loss": 0.6415,
+      "step": 35473
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4905067875923836,
+      "learning_rate": 1.3820177835007541e-06,
+      "loss": 0.6325,
+      "step": 35474
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3713286799626974,
+      "learning_rate": 1.3816307550618968e-06,
+      "loss": 0.6355,
+      "step": 35475
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3839005471301074,
+      "learning_rate": 1.3812437768019172e-06,
+      "loss": 0.7041,
+      "step": 35476
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4250043853837506,
+      "learning_rate": 1.3808568487230611e-06,
+      "loss": 0.8251,
+      "step": 35477
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4418304951230467,
+      "learning_rate": 1.3804699708275847e-06,
+      "loss": 0.6151,
+      "step": 35478
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2806781637991276,
+      "learning_rate": 1.3800831431177419e-06,
+      "loss": 0.5075,
+      "step": 35479
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4086669607785096,
+      "learning_rate": 1.3796963655957872e-06,
+      "loss": 0.5737,
+      "step": 35480
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4496728855772896,
+      "learning_rate": 1.3793096382639658e-06,
+      "loss": 0.6952,
+      "step": 35481
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5165938594148722,
+      "learning_rate": 1.3789229611245335e-06,
+      "loss": 0.6196,
+      "step": 35482
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3078257500658859,
+      "learning_rate": 1.378536334179743e-06,
+      "loss": 0.7096,
+      "step": 35483
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.629417099462748,
+      "learning_rate": 1.3781497574318402e-06,
+      "loss": 0.6196,
+      "step": 35484
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3655364780202193,
+      "learning_rate": 1.3777632308830824e-06,
+      "loss": 0.6918,
+      "step": 35485
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4628687921789292,
+      "learning_rate": 1.3773767545357141e-06,
+      "loss": 0.6855,
+      "step": 35486
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1504790037190245,
+      "learning_rate": 1.3769903283919905e-06,
+      "loss": 0.5852,
+      "step": 35487
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.515573699051999,
+      "learning_rate": 1.3766039524541575e-06,
+      "loss": 0.6316,
+      "step": 35488
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5068408821858288,
+      "learning_rate": 1.3762176267244686e-06,
+      "loss": 0.631,
+      "step": 35489
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9997123288017573,
+      "learning_rate": 1.3758313512051702e-06,
+      "loss": 0.6172,
+      "step": 35490
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5977439030253107,
+      "learning_rate": 1.3754451258985102e-06,
+      "loss": 0.5126,
+      "step": 35491
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.144250954818651,
+      "learning_rate": 1.375058950806739e-06,
+      "loss": 0.622,
+      "step": 35492
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3469600586403023,
+      "learning_rate": 1.3746728259321085e-06,
+      "loss": 0.6619,
+      "step": 35493
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6341365021368065,
+      "learning_rate": 1.374286751276862e-06,
+      "loss": 0.5614,
+      "step": 35494
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3720222055326783,
+      "learning_rate": 1.373900726843247e-06,
+      "loss": 0.6015,
+      "step": 35495
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6818963581650026,
+      "learning_rate": 1.3735147526335135e-06,
+      "loss": 0.7801,
+      "step": 35496
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4051258159803042,
+      "learning_rate": 1.37312882864991e-06,
+      "loss": 0.7348,
+      "step": 35497
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5452654453670855,
+      "learning_rate": 1.3727429548946825e-06,
+      "loss": 0.7515,
+      "step": 35498
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3381908756587393,
+      "learning_rate": 1.3723571313700734e-06,
+      "loss": 0.5771,
+      "step": 35499
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5929536615695338,
+      "learning_rate": 1.3719713580783334e-06,
+      "loss": 0.5266,
+      "step": 35500
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.528773384456363,
+      "learning_rate": 1.3715856350217095e-06,
+      "loss": 0.6253,
+      "step": 35501
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7552304495387108,
+      "learning_rate": 1.3711999622024463e-06,
+      "loss": 0.6435,
+      "step": 35502
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9650365155037834,
+      "learning_rate": 1.3708143396227857e-06,
+      "loss": 0.6886,
+      "step": 35503
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.636125960804871,
+      "learning_rate": 1.3704287672849793e-06,
+      "loss": 0.5095,
+      "step": 35504
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4527353797494118,
+      "learning_rate": 1.370043245191266e-06,
+      "loss": 0.6064,
+      "step": 35505
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3252435861670662,
+      "learning_rate": 1.3696577733438943e-06,
+      "loss": 0.6622,
+      "step": 35506
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.293122465720914,
+      "learning_rate": 1.3692723517451067e-06,
+      "loss": 0.6752,
+      "step": 35507
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3758619184164989,
+      "learning_rate": 1.3688869803971493e-06,
+      "loss": 0.6194,
+      "step": 35508
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.584657289477601,
+      "learning_rate": 1.3685016593022627e-06,
+      "loss": 0.6701,
+      "step": 35509
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5740041821637467,
+      "learning_rate": 1.3681163884626935e-06,
+      "loss": 0.4799,
+      "step": 35510
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.295969570625682,
+      "learning_rate": 1.3677311678806838e-06,
+      "loss": 0.6457,
+      "step": 35511
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.489392552592741,
+      "learning_rate": 1.3673459975584746e-06,
+      "loss": 0.6403,
+      "step": 35512
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7090358897942903,
+      "learning_rate": 1.3669608774983102e-06,
+      "loss": 0.7901,
+      "step": 35513
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5223822156409963,
+      "learning_rate": 1.3665758077024338e-06,
+      "loss": 0.6662,
+      "step": 35514
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3601121491966475,
+      "learning_rate": 1.3661907881730874e-06,
+      "loss": 0.7095,
+      "step": 35515
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9369455432718459,
+      "learning_rate": 1.365805818912509e-06,
+      "loss": 0.6165,
+      "step": 35516
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.355462000841483,
+      "learning_rate": 1.3654208999229423e-06,
+      "loss": 0.633,
+      "step": 35517
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4188835524985743,
+      "learning_rate": 1.3650360312066314e-06,
+      "loss": 0.6076,
+      "step": 35518
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5139517817765469,
+      "learning_rate": 1.364651212765814e-06,
+      "loss": 0.5975,
+      "step": 35519
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5333271530445751,
+      "learning_rate": 1.3642664446027298e-06,
+      "loss": 0.6512,
+      "step": 35520
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.532099774698893,
+      "learning_rate": 1.3638817267196203e-06,
+      "loss": 0.7985,
+      "step": 35521
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3485218134876484,
+      "learning_rate": 1.3634970591187269e-06,
+      "loss": 0.6057,
+      "step": 35522
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5882468913274322,
+      "learning_rate": 1.363112441802289e-06,
+      "loss": 0.5373,
+      "step": 35523
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.359163612498187,
+      "learning_rate": 1.3627278747725415e-06,
+      "loss": 0.6962,
+      "step": 35524
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3025834246518053,
+      "learning_rate": 1.3623433580317303e-06,
+      "loss": 0.5991,
+      "step": 35525
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5359921303716437,
+      "learning_rate": 1.3619588915820869e-06,
+      "loss": 0.7807,
+      "step": 35526
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2577580191104138,
+      "learning_rate": 1.3615744754258575e-06,
+      "loss": 0.7037,
+      "step": 35527
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3428614985027003,
+      "learning_rate": 1.3611901095652736e-06,
+      "loss": 0.6085,
+      "step": 35528
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.269607185811644,
+      "learning_rate": 1.3608057940025775e-06,
+      "loss": 0.5902,
+      "step": 35529
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4254666625287737,
+      "learning_rate": 1.3604215287400036e-06,
+      "loss": 0.6459,
+      "step": 35530
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3761647313960823,
+      "learning_rate": 1.3600373137797917e-06,
+      "loss": 0.6969,
+      "step": 35531
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.613193183059565,
+      "learning_rate": 1.3596531491241793e-06,
+      "loss": 0.573,
+      "step": 35532
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5753912364412075,
+      "learning_rate": 1.3592690347754e-06,
+      "loss": 0.5152,
+      "step": 35533
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4423854239236131,
+      "learning_rate": 1.358884970735691e-06,
+      "loss": 0.6452,
+      "step": 35534
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7103702157756644,
+      "learning_rate": 1.3585009570072917e-06,
+      "loss": 0.8576,
+      "step": 35535
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4327816080108016,
+      "learning_rate": 1.3581169935924355e-06,
+      "loss": 0.7688,
+      "step": 35536
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.381648104032231,
+      "learning_rate": 1.3577330804933563e-06,
+      "loss": 0.5935,
+      "step": 35537
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7533118548557232,
+      "learning_rate": 1.3573492177122916e-06,
+      "loss": 0.6462,
+      "step": 35538
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0053541685209932,
+      "learning_rate": 1.356965405251478e-06,
+      "loss": 0.7077,
+      "step": 35539
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4133274889705802,
+      "learning_rate": 1.3565816431131474e-06,
+      "loss": 0.6479,
+      "step": 35540
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3125152627249634,
+      "learning_rate": 1.3561979312995333e-06,
+      "loss": 0.6472,
+      "step": 35541
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4014273093947325,
+      "learning_rate": 1.3558142698128719e-06,
+      "loss": 0.7121,
+      "step": 35542
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3396169953189614,
+      "learning_rate": 1.355430658655398e-06,
+      "loss": 0.5003,
+      "step": 35543
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4352223783932843,
+      "learning_rate": 1.355047097829344e-06,
+      "loss": 0.5915,
+      "step": 35544
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6152061004258114,
+      "learning_rate": 1.3546635873369396e-06,
+      "loss": 0.5148,
+      "step": 35545
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3381351878598435,
+      "learning_rate": 1.354280127180424e-06,
+      "loss": 0.7098,
+      "step": 35546
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7813326317456821,
+      "learning_rate": 1.3538967173620243e-06,
+      "loss": 0.587,
+      "step": 35547
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2479256259462974,
+      "learning_rate": 1.3535133578839777e-06,
+      "loss": 0.6884,
+      "step": 35548
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4821141827778779,
+      "learning_rate": 1.3531300487485111e-06,
+      "loss": 0.7252,
+      "step": 35549
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6505341606388979,
+      "learning_rate": 1.3527467899578617e-06,
+      "loss": 0.545,
+      "step": 35550
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.513074411662009,
+      "learning_rate": 1.3523635815142556e-06,
+      "loss": 0.6422,
+      "step": 35551
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1759389309528672,
+      "learning_rate": 1.3519804234199286e-06,
+      "loss": 0.744,
+      "step": 35552
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.431496828555912,
+      "learning_rate": 1.351597315677109e-06,
+      "loss": 0.623,
+      "step": 35553
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3871343482791505,
+      "learning_rate": 1.351214258288026e-06,
+      "loss": 0.6188,
+      "step": 35554
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6584402347256679,
+      "learning_rate": 1.3508312512549127e-06,
+      "loss": 0.5235,
+      "step": 35555
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3860655329067406,
+      "learning_rate": 1.3504482945799991e-06,
+      "loss": 0.7081,
+      "step": 35556
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.391357159495308,
+      "learning_rate": 1.3500653882655146e-06,
+      "loss": 0.608,
+      "step": 35557
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0012408648017224,
+      "learning_rate": 1.3496825323136852e-06,
+      "loss": 0.7346,
+      "step": 35558
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.595278129125539,
+      "learning_rate": 1.349299726726744e-06,
+      "loss": 0.5021,
+      "step": 35559
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.301094111735131,
+      "learning_rate": 1.3489169715069195e-06,
+      "loss": 0.7138,
+      "step": 35560
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3199033678139447,
+      "learning_rate": 1.3485342666564394e-06,
+      "loss": 0.7329,
+      "step": 35561
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3548368643793554,
+      "learning_rate": 1.3481516121775295e-06,
+      "loss": 0.6285,
+      "step": 35562
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3490282191820318,
+      "learning_rate": 1.3477690080724203e-06,
+      "loss": 0.8207,
+      "step": 35563
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5787384522089364,
+      "learning_rate": 1.3473864543433423e-06,
+      "loss": 0.516,
+      "step": 35564
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4838354306606782,
+      "learning_rate": 1.3470039509925192e-06,
+      "loss": 0.5604,
+      "step": 35565
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.461027120683553,
+      "learning_rate": 1.346621498022176e-06,
+      "loss": 0.7499,
+      "step": 35566
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8745119654284408,
+      "learning_rate": 1.3462390954345416e-06,
+      "loss": 0.6595,
+      "step": 35567
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4095788709159973,
+      "learning_rate": 1.3458567432318458e-06,
+      "loss": 0.6803,
+      "step": 35568
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5923141865820485,
+      "learning_rate": 1.3454744414163101e-06,
+      "loss": 0.7172,
+      "step": 35569
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.487443272957201,
+      "learning_rate": 1.3450921899901637e-06,
+      "loss": 0.6759,
+      "step": 35570
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5525219700043127,
+      "learning_rate": 1.3447099889556302e-06,
+      "loss": 0.7077,
+      "step": 35571
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3742093756305762,
+      "learning_rate": 1.3443278383149338e-06,
+      "loss": 0.6756,
+      "step": 35572
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.565696112453018,
+      "learning_rate": 1.3439457380703015e-06,
+      "loss": 0.4693,
+      "step": 35573
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3743221180140552,
+      "learning_rate": 1.3435636882239588e-06,
+      "loss": 0.6633,
+      "step": 35574
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2844418427179825,
+      "learning_rate": 1.3431816887781302e-06,
+      "loss": 0.5766,
+      "step": 35575
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.581088382831862,
+      "learning_rate": 1.3427997397350356e-06,
+      "loss": 0.5404,
+      "step": 35576
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4755418329638401,
+      "learning_rate": 1.3424178410969013e-06,
+      "loss": 0.6418,
+      "step": 35577
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5725023400136762,
+      "learning_rate": 1.3420359928659564e-06,
+      "loss": 0.4948,
+      "step": 35578
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2960965154960555,
+      "learning_rate": 1.341654195044414e-06,
+      "loss": 0.7345,
+      "step": 35579
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4386610322875968,
+      "learning_rate": 1.341272447634503e-06,
+      "loss": 0.5689,
+      "step": 35580
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2493117293602078,
+      "learning_rate": 1.340890750638444e-06,
+      "loss": 0.711,
+      "step": 35581
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5868241648871618,
+      "learning_rate": 1.3405091040584628e-06,
+      "loss": 0.4979,
+      "step": 35582
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2959972716678987,
+      "learning_rate": 1.34012750789678e-06,
+      "loss": 0.8112,
+      "step": 35583
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4667383342728217,
+      "learning_rate": 1.339745962155613e-06,
+      "loss": 0.7358,
+      "step": 35584
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5849059654537797,
+      "learning_rate": 1.3393644668371897e-06,
+      "loss": 0.4825,
+      "step": 35585
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5097440292461206,
+      "learning_rate": 1.3389830219437272e-06,
+      "loss": 0.7567,
+      "step": 35586
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5718251623348225,
+      "learning_rate": 1.3386016274774482e-06,
+      "loss": 0.7041,
+      "step": 35587
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.312477198940932,
+      "learning_rate": 1.338220283440571e-06,
+      "loss": 0.5971,
+      "step": 35588
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3354503267203415,
+      "learning_rate": 1.3378389898353206e-06,
+      "loss": 0.749,
+      "step": 35589
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.306508001944085,
+      "learning_rate": 1.3374577466639105e-06,
+      "loss": 0.5935,
+      "step": 35590
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5667309263009058,
+      "learning_rate": 1.337076553928567e-06,
+      "loss": 0.489,
+      "step": 35591
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6074933040202672,
+      "learning_rate": 1.3366954116315057e-06,
+      "loss": 0.5276,
+      "step": 35592
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3171306688361228,
+      "learning_rate": 1.3363143197749451e-06,
+      "loss": 0.6164,
+      "step": 35593
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4763515504501004,
+      "learning_rate": 1.3359332783611045e-06,
+      "loss": 0.689,
+      "step": 35594
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4061525244217299,
+      "learning_rate": 1.3355522873922068e-06,
+      "loss": 0.7464,
+      "step": 35595
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.438873957064413,
+      "learning_rate": 1.3351713468704663e-06,
+      "loss": 0.668,
+      "step": 35596
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.22060689841323,
+      "learning_rate": 1.3347904567980984e-06,
+      "loss": 0.6781,
+      "step": 35597
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4468123314093284,
+      "learning_rate": 1.3344096171773246e-06,
+      "loss": 0.6409,
+      "step": 35598
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4627721867412387,
+      "learning_rate": 1.3340288280103641e-06,
+      "loss": 0.6579,
+      "step": 35599
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.593379421592723,
+      "learning_rate": 1.333648089299431e-06,
+      "loss": 0.563,
+      "step": 35600
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3639051905109298,
+      "learning_rate": 1.3332674010467395e-06,
+      "loss": 0.5843,
+      "step": 35601
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.568549698943629,
+      "learning_rate": 1.3328867632545094e-06,
+      "loss": 0.6757,
+      "step": 35602
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3648780736438948,
+      "learning_rate": 1.3325061759249591e-06,
+      "loss": 0.6634,
+      "step": 35603
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3375151913366168,
+      "learning_rate": 1.332125639060302e-06,
+      "loss": 0.7984,
+      "step": 35604
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4470598749790955,
+      "learning_rate": 1.3317451526627512e-06,
+      "loss": 0.6894,
+      "step": 35605
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.118417229599652,
+      "learning_rate": 1.3313647167345266e-06,
+      "loss": 0.6373,
+      "step": 35606
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8383454276784728,
+      "learning_rate": 1.330984331277838e-06,
+      "loss": 0.6982,
+      "step": 35607
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.358567874967713,
+      "learning_rate": 1.330603996294907e-06,
+      "loss": 0.6386,
+      "step": 35608
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2728959215135227,
+      "learning_rate": 1.3302237117879413e-06,
+      "loss": 0.5273,
+      "step": 35609
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4653117153595556,
+      "learning_rate": 1.3298434777591596e-06,
+      "loss": 0.7358,
+      "step": 35610
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.722534363076,
+      "learning_rate": 1.3294632942107733e-06,
+      "loss": 0.5685,
+      "step": 35611
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0348300648272932,
+      "learning_rate": 1.3290831611449973e-06,
+      "loss": 0.6229,
+      "step": 35612
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4663824102199976,
+      "learning_rate": 1.3287030785640454e-06,
+      "loss": 0.5609,
+      "step": 35613
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3921858808724987,
+      "learning_rate": 1.3283230464701268e-06,
+      "loss": 0.5937,
+      "step": 35614
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.32111539718364,
+      "learning_rate": 1.3279430648654567e-06,
+      "loss": 0.7152,
+      "step": 35615
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2806258131537447,
+      "learning_rate": 1.327563133752251e-06,
+      "loss": 0.6291,
+      "step": 35616
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4046473695545922,
+      "learning_rate": 1.3271832531327166e-06,
+      "loss": 0.6702,
+      "step": 35617
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4409963484025679,
+      "learning_rate": 1.3268034230090664e-06,
+      "loss": 0.6754,
+      "step": 35618
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3055022337960165,
+      "learning_rate": 1.326423643383512e-06,
+      "loss": 0.6576,
+      "step": 35619
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5517343921335147,
+      "learning_rate": 1.326043914258267e-06,
+      "loss": 0.6325,
+      "step": 35620
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2766489443254383,
+      "learning_rate": 1.325664235635541e-06,
+      "loss": 0.5791,
+      "step": 35621
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3448628396130182,
+      "learning_rate": 1.3252846075175407e-06,
+      "loss": 0.7676,
+      "step": 35622
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5444836629694207,
+      "learning_rate": 1.3249050299064815e-06,
+      "loss": 0.7538,
+      "step": 35623
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3675564615270834,
+      "learning_rate": 1.3245255028045722e-06,
+      "loss": 0.61,
+      "step": 35624
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6323829315855087,
+      "learning_rate": 1.3241460262140237e-06,
+      "loss": 0.653,
+      "step": 35625
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3481285446217186,
+      "learning_rate": 1.323766600137041e-06,
+      "loss": 0.772,
+      "step": 35626
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3753540218262092,
+      "learning_rate": 1.3233872245758373e-06,
+      "loss": 0.6792,
+      "step": 35627
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5675957360487849,
+      "learning_rate": 1.323007899532619e-06,
+      "loss": 0.7438,
+      "step": 35628
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0428230441774535,
+      "learning_rate": 1.3226286250095978e-06,
+      "loss": 0.6939,
+      "step": 35629
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7006339778495538,
+      "learning_rate": 1.322249401008977e-06,
+      "loss": 0.5977,
+      "step": 35630
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6590006281688054,
+      "learning_rate": 1.3218702275329708e-06,
+      "loss": 0.7268,
+      "step": 35631
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3403809115416025,
+      "learning_rate": 1.321491104583781e-06,
+      "loss": 0.7341,
+      "step": 35632
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8305951563540517,
+      "learning_rate": 1.3211120321636194e-06,
+      "loss": 0.7592,
+      "step": 35633
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4622102613767074,
+      "learning_rate": 1.3207330102746907e-06,
+      "loss": 0.8402,
+      "step": 35634
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.515054020782299,
+      "learning_rate": 1.3203540389192014e-06,
+      "loss": 0.7047,
+      "step": 35635
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7699278847382784,
+      "learning_rate": 1.319975118099358e-06,
+      "loss": 0.7102,
+      "step": 35636
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4357849198538564,
+      "learning_rate": 1.3195962478173695e-06,
+      "loss": 0.8039,
+      "step": 35637
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5391374057282883,
+      "learning_rate": 1.3192174280754411e-06,
+      "loss": 0.7053,
+      "step": 35638
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7398240589297241,
+      "learning_rate": 1.318838658875774e-06,
+      "loss": 0.5554,
+      "step": 35639
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5516042618937707,
+      "learning_rate": 1.3184599402205766e-06,
+      "loss": 0.761,
+      "step": 35640
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6708103672958328,
+      "learning_rate": 1.318081272112056e-06,
+      "loss": 0.7182,
+      "step": 35641
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5935038130515822,
+      "learning_rate": 1.3177026545524152e-06,
+      "loss": 0.4977,
+      "step": 35642
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5570818307375756,
+      "learning_rate": 1.3173240875438565e-06,
+      "loss": 0.5631,
+      "step": 35643
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4533316199529631,
+      "learning_rate": 1.3169455710885847e-06,
+      "loss": 0.5709,
+      "step": 35644
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1808238973738505,
+      "learning_rate": 1.3165671051888073e-06,
+      "loss": 0.5708,
+      "step": 35645
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.371735809392446,
+      "learning_rate": 1.316188689846727e-06,
+      "loss": 0.6162,
+      "step": 35646
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6816450835262249,
+      "learning_rate": 1.3158103250645416e-06,
+      "loss": 0.7163,
+      "step": 35647
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.096916964426221,
+      "learning_rate": 1.3154320108444597e-06,
+      "loss": 0.6766,
+      "step": 35648
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5871104634152979,
+      "learning_rate": 1.3150537471886814e-06,
+      "loss": 0.5278,
+      "step": 35649
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.7878421872360692,
+      "learning_rate": 1.314675534099411e-06,
+      "loss": 0.6296,
+      "step": 35650
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.381365307760963,
+      "learning_rate": 1.3142973715788477e-06,
+      "loss": 0.6584,
+      "step": 35651
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4536794946960059,
+      "learning_rate": 1.313919259629196e-06,
+      "loss": 0.635,
+      "step": 35652
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2841334090646679,
+      "learning_rate": 1.3135411982526546e-06,
+      "loss": 0.7217,
+      "step": 35653
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9926951596321032,
+      "learning_rate": 1.3131631874514272e-06,
+      "loss": 0.5964,
+      "step": 35654
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3751342263681294,
+      "learning_rate": 1.3127852272277187e-06,
+      "loss": 0.7193,
+      "step": 35655
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6211597192556507,
+      "learning_rate": 1.3124073175837193e-06,
+      "loss": 0.5001,
+      "step": 35656
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5786426808839498,
+      "learning_rate": 1.3120294585216353e-06,
+      "loss": 0.4932,
+      "step": 35657
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.345460283206198,
+      "learning_rate": 1.311651650043666e-06,
+      "loss": 0.5761,
+      "step": 35658
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2742537661431075,
+      "learning_rate": 1.3112738921520164e-06,
+      "loss": 0.4666,
+      "step": 35659
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2668965422823784,
+      "learning_rate": 1.3108961848488767e-06,
+      "loss": 0.6689,
+      "step": 35660
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4347328756477054,
+      "learning_rate": 1.3105185281364507e-06,
+      "loss": 0.7495,
+      "step": 35661
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4094902033393513,
+      "learning_rate": 1.310140922016938e-06,
+      "loss": 0.6262,
+      "step": 35662
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3477418329332855,
+      "learning_rate": 1.309763366492537e-06,
+      "loss": 0.5764,
+      "step": 35663
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8376834354861011,
+      "learning_rate": 1.3093858615654421e-06,
+      "loss": 0.6998,
+      "step": 35664
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4382369059969697,
+      "learning_rate": 1.309008407237854e-06,
+      "loss": 0.6896,
+      "step": 35665
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4874505346074298,
+      "learning_rate": 1.3086310035119732e-06,
+      "loss": 0.6859,
+      "step": 35666
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3874079180094852,
+      "learning_rate": 1.3082536503899922e-06,
+      "loss": 0.6606,
+      "step": 35667
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3072000065678882,
+      "learning_rate": 1.3078763478741118e-06,
+      "loss": 0.6286,
+      "step": 35668
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2009732367336057,
+      "learning_rate": 1.3074990959665279e-06,
+      "loss": 0.6385,
+      "step": 35669
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4310841562422087,
+      "learning_rate": 1.307121894669433e-06,
+      "loss": 0.6285,
+      "step": 35670
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.400683261272983,
+      "learning_rate": 1.3067447439850279e-06,
+      "loss": 0.7842,
+      "step": 35671
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.296019557034698,
+      "learning_rate": 1.306367643915508e-06,
+      "loss": 0.6151,
+      "step": 35672
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0546271874297597,
+      "learning_rate": 1.3059905944630691e-06,
+      "loss": 0.6397,
+      "step": 35673
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4319512115343933,
+      "learning_rate": 1.3056135956299032e-06,
+      "loss": 0.7323,
+      "step": 35674
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.29469225253738,
+      "learning_rate": 1.3052366474182065e-06,
+      "loss": 0.6509,
+      "step": 35675
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3253837489162097,
+      "learning_rate": 1.3048597498301808e-06,
+      "loss": 0.7607,
+      "step": 35676
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.284931932653307,
+      "learning_rate": 1.3044829028680084e-06,
+      "loss": 0.6398,
+      "step": 35677
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3261285154179314,
+      "learning_rate": 1.30410610653389e-06,
+      "loss": 0.708,
+      "step": 35678
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.594477789334314,
+      "learning_rate": 1.3037293608300195e-06,
+      "loss": 0.6826,
+      "step": 35679
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3490082368893348,
+      "learning_rate": 1.3033526657585915e-06,
+      "loss": 0.7983,
+      "step": 35680
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3163545791431566,
+      "learning_rate": 1.302976021321798e-06,
+      "loss": 0.662,
+      "step": 35681
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.422516528495996,
+      "learning_rate": 1.302599427521829e-06,
+      "loss": 0.7098,
+      "step": 35682
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.354028751077943,
+      "learning_rate": 1.302222884360882e-06,
+      "loss": 0.7146,
+      "step": 35683
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.624390082148913,
+      "learning_rate": 1.301846391841145e-06,
+      "loss": 0.7632,
+      "step": 35684
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7865801761426081,
+      "learning_rate": 1.3014699499648142e-06,
+      "loss": 0.5127,
+      "step": 35685
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4857620027263367,
+      "learning_rate": 1.3010935587340767e-06,
+      "loss": 0.7248,
+      "step": 35686
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4165527332417083,
+      "learning_rate": 1.3007172181511296e-06,
+      "loss": 0.6133,
+      "step": 35687
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.411268529132244,
+      "learning_rate": 1.3003409282181579e-06,
+      "loss": 0.8034,
+      "step": 35688
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.591896988786394,
+      "learning_rate": 1.2999646889373585e-06,
+      "loss": 0.509,
+      "step": 35689
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.434245283199726,
+      "learning_rate": 1.2995885003109166e-06,
+      "loss": 0.6316,
+      "step": 35690
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.278716934799549,
+      "learning_rate": 1.2992123623410268e-06,
+      "loss": 0.6945,
+      "step": 35691
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.586375142349574,
+      "learning_rate": 1.2988362750298754e-06,
+      "loss": 0.512,
+      "step": 35692
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2860378950499598,
+      "learning_rate": 1.2984602383796551e-06,
+      "loss": 0.7322,
+      "step": 35693
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3703203564180464,
+      "learning_rate": 1.2980842523925553e-06,
+      "loss": 0.5403,
+      "step": 35694
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4100946591528112,
+      "learning_rate": 1.2977083170707616e-06,
+      "loss": 0.6532,
+      "step": 35695
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9373500627401892,
+      "learning_rate": 1.2973324324164649e-06,
+      "loss": 0.6905,
+      "step": 35696
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5765670416680189,
+      "learning_rate": 1.2969565984318567e-06,
+      "loss": 0.6774,
+      "step": 35697
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5271671826234203,
+      "learning_rate": 1.2965808151191229e-06,
+      "loss": 0.713,
+      "step": 35698
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4558244879178397,
+      "learning_rate": 1.2962050824804472e-06,
+      "loss": 0.8022,
+      "step": 35699
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4428868974870672,
+      "learning_rate": 1.2958294005180228e-06,
+      "loss": 0.7615,
+      "step": 35700
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.387674063826748,
+      "learning_rate": 1.2954537692340363e-06,
+      "loss": 0.6392,
+      "step": 35701
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.614693816714957,
+      "learning_rate": 1.295078188630674e-06,
+      "loss": 0.5931,
+      "step": 35702
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5241841757297425,
+      "learning_rate": 1.2947026587101208e-06,
+      "loss": 0.8063,
+      "step": 35703
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8994194074968629,
+      "learning_rate": 1.2943271794745649e-06,
+      "loss": 0.589,
+      "step": 35704
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.230121546406475,
+      "learning_rate": 1.2939517509261934e-06,
+      "loss": 0.6253,
+      "step": 35705
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4243747932688005,
+      "learning_rate": 1.2935763730671923e-06,
+      "loss": 0.6931,
+      "step": 35706
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3553715723620285,
+      "learning_rate": 1.2932010458997434e-06,
+      "loss": 0.6501,
+      "step": 35707
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4057131804978713,
+      "learning_rate": 1.2928257694260359e-06,
+      "loss": 0.6685,
+      "step": 35708
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2969505107376453,
+      "learning_rate": 1.2924505436482527e-06,
+      "loss": 0.7033,
+      "step": 35709
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4775577884824969,
+      "learning_rate": 1.2920753685685805e-06,
+      "loss": 0.832,
+      "step": 35710
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3495598683480095,
+      "learning_rate": 1.2917002441892002e-06,
+      "loss": 0.7379,
+      "step": 35711
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.587588710917254,
+      "learning_rate": 1.2913251705122998e-06,
+      "loss": 0.7779,
+      "step": 35712
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5479697142145983,
+      "learning_rate": 1.2909501475400598e-06,
+      "loss": 0.6889,
+      "step": 35713
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4709022106694134,
+      "learning_rate": 1.2905751752746676e-06,
+      "loss": 0.7231,
+      "step": 35714
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6102082965622129,
+      "learning_rate": 1.2902002537183044e-06,
+      "loss": 0.6858,
+      "step": 35715
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.404434789324103,
+      "learning_rate": 1.28982538287315e-06,
+      "loss": 0.5763,
+      "step": 35716
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3469652136399173,
+      "learning_rate": 1.28945056274139e-06,
+      "loss": 0.641,
+      "step": 35717
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.328334224836587,
+      "learning_rate": 1.2890757933252084e-06,
+      "loss": 0.627,
+      "step": 35718
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.655481777453116,
+      "learning_rate": 1.2887010746267859e-06,
+      "loss": 0.7627,
+      "step": 35719
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5536091024894253,
+      "learning_rate": 1.2883264066483015e-06,
+      "loss": 0.5018,
+      "step": 35720
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4311482039360488,
+      "learning_rate": 1.287951789391938e-06,
+      "loss": 0.549,
+      "step": 35721
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3468653671241606,
+      "learning_rate": 1.2875772228598805e-06,
+      "loss": 0.6386,
+      "step": 35722
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.405539516128179,
+      "learning_rate": 1.2872027070543057e-06,
+      "loss": 0.794,
+      "step": 35723
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.358776450791326,
+      "learning_rate": 1.2868282419773936e-06,
+      "loss": 0.7046,
+      "step": 35724
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.609984031180561,
+      "learning_rate": 1.2864538276313265e-06,
+      "loss": 0.7444,
+      "step": 35725
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.49002148553505,
+      "learning_rate": 1.286079464018285e-06,
+      "loss": 0.6849,
+      "step": 35726
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.762029885038148,
+      "learning_rate": 1.2857051511404484e-06,
+      "loss": 0.6563,
+      "step": 35727
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5951527156264138,
+      "learning_rate": 1.2853308889999927e-06,
+      "loss": 0.499,
+      "step": 35728
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4772341650034349,
+      "learning_rate": 1.2849566775991029e-06,
+      "loss": 0.6605,
+      "step": 35729
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4330138878260055,
+      "learning_rate": 1.2845825169399506e-06,
+      "loss": 0.6935,
+      "step": 35730
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2800813267503428,
+      "learning_rate": 1.2842084070247217e-06,
+      "loss": 0.5843,
+      "step": 35731
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6511103144964472,
+      "learning_rate": 1.283834347855588e-06,
+      "loss": 0.7781,
+      "step": 35732
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4549755857067939,
+      "learning_rate": 1.2834603394347323e-06,
+      "loss": 0.6764,
+      "step": 35733
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6111574517443443,
+      "learning_rate": 1.2830863817643279e-06,
+      "loss": 0.5268,
+      "step": 35734
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5098245577302736,
+      "learning_rate": 1.2827124748465569e-06,
+      "loss": 0.6523,
+      "step": 35735
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.409160340248863,
+      "learning_rate": 1.282338618683594e-06,
+      "loss": 0.6942,
+      "step": 35736
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8046133607169341,
+      "learning_rate": 1.281964813277612e-06,
+      "loss": 0.7189,
+      "step": 35737
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5528351990467768,
+      "learning_rate": 1.2815910586307923e-06,
+      "loss": 0.671,
+      "step": 35738
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4086034956430697,
+      "learning_rate": 1.281217354745311e-06,
+      "loss": 0.6642,
+      "step": 35739
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3397203291895923,
+      "learning_rate": 1.2808437016233433e-06,
+      "loss": 0.7135,
+      "step": 35740
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6296012901276316,
+      "learning_rate": 1.2804700992670616e-06,
+      "loss": 0.7215,
+      "step": 35741
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6574337133374746,
+      "learning_rate": 1.2800965476786431e-06,
+      "loss": 0.6522,
+      "step": 35742
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6978218457343905,
+      "learning_rate": 1.279723046860266e-06,
+      "loss": 0.6113,
+      "step": 35743
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5015124672216866,
+      "learning_rate": 1.2793495968141023e-06,
+      "loss": 0.6358,
+      "step": 35744
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.2374703668398976,
+      "learning_rate": 1.2789761975423231e-06,
+      "loss": 0.7165,
+      "step": 35745
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4742299297188368,
+      "learning_rate": 1.2786028490471058e-06,
+      "loss": 0.6057,
+      "step": 35746
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6422129705880506,
+      "learning_rate": 1.2782295513306264e-06,
+      "loss": 0.6115,
+      "step": 35747
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2920020096593652,
+      "learning_rate": 1.2778563043950553e-06,
+      "loss": 0.6845,
+      "step": 35748
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.759911835438691,
+      "learning_rate": 1.2774831082425644e-06,
+      "loss": 0.7295,
+      "step": 35749
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2578438389728692,
+      "learning_rate": 1.2771099628753304e-06,
+      "loss": 0.5983,
+      "step": 35750
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6705056842748363,
+      "learning_rate": 1.2767368682955216e-06,
+      "loss": 0.5746,
+      "step": 35751
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3130772141091727,
+      "learning_rate": 1.2763638245053134e-06,
+      "loss": 0.5369,
+      "step": 35752
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.42069411277801,
+      "learning_rate": 1.275990831506878e-06,
+      "loss": 0.6719,
+      "step": 35753
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4640431752220526,
+      "learning_rate": 1.2756178893023862e-06,
+      "loss": 0.6658,
+      "step": 35754
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6015292147304638,
+      "learning_rate": 1.2752449978940062e-06,
+      "loss": 0.7166,
+      "step": 35755
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4206694759428078,
+      "learning_rate": 1.274872157283913e-06,
+      "loss": 0.7348,
+      "step": 35756
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3280770829507156,
+      "learning_rate": 1.2744993674742812e-06,
+      "loss": 0.6581,
+      "step": 35757
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.613947456671666,
+      "learning_rate": 1.2741266284672704e-06,
+      "loss": 0.6652,
+      "step": 35758
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6001047335811,
+      "learning_rate": 1.273753940265059e-06,
+      "loss": 0.6536,
+      "step": 35759
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5929776721464954,
+      "learning_rate": 1.2733813028698138e-06,
+      "loss": 0.8957,
+      "step": 35760
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4007229009581745,
+      "learning_rate": 1.273008716283709e-06,
+      "loss": 0.5245,
+      "step": 35761
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3457499320098012,
+      "learning_rate": 1.2726361805089072e-06,
+      "loss": 0.5932,
+      "step": 35762
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.371226611814872,
+      "learning_rate": 1.272263695547581e-06,
+      "loss": 0.6161,
+      "step": 35763
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4163594134306687,
+      "learning_rate": 1.2718912614018996e-06,
+      "loss": 0.7744,
+      "step": 35764
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2431223725659333,
+      "learning_rate": 1.2715188780740295e-06,
+      "loss": 0.7822,
+      "step": 35765
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2482511796419986,
+      "learning_rate": 1.271146545566141e-06,
+      "loss": 0.6015,
+      "step": 35766
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2268564322020319,
+      "learning_rate": 1.2707742638803987e-06,
+      "loss": 0.63,
+      "step": 35767
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4719589310952819,
+      "learning_rate": 1.2704020330189759e-06,
+      "loss": 0.7164,
+      "step": 35768
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5738291106059128,
+      "learning_rate": 1.2700298529840327e-06,
+      "loss": 0.8011,
+      "step": 35769
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5034627779210996,
+      "learning_rate": 1.2696577237777418e-06,
+      "loss": 0.7912,
+      "step": 35770
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.510524076221643,
+      "learning_rate": 1.2692856454022684e-06,
+      "loss": 0.7084,
+      "step": 35771
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3236975548582346,
+      "learning_rate": 1.268913617859775e-06,
+      "loss": 0.6424,
+      "step": 35772
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1430669523474415,
+      "learning_rate": 1.268541641152431e-06,
+      "loss": 0.6677,
+      "step": 35773
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3132017844975035,
+      "learning_rate": 1.2681697152824045e-06,
+      "loss": 0.5722,
+      "step": 35774
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.628329571246082,
+      "learning_rate": 1.2677978402518576e-06,
+      "loss": 0.5914,
+      "step": 35775
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4155878097566919,
+      "learning_rate": 1.2674260160629547e-06,
+      "loss": 0.7443,
+      "step": 35776
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2425475868511153,
+      "learning_rate": 1.2670542427178623e-06,
+      "loss": 0.6547,
+      "step": 35777
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3025816020854175,
+      "learning_rate": 1.266682520218747e-06,
+      "loss": 0.5891,
+      "step": 35778
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7942208258191583,
+      "learning_rate": 1.266310848567771e-06,
+      "loss": 0.6669,
+      "step": 35779
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2921724965222157,
+      "learning_rate": 1.2659392277670956e-06,
+      "loss": 0.6692,
+      "step": 35780
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4995492213449635,
+      "learning_rate": 1.265567657818888e-06,
+      "loss": 0.677,
+      "step": 35781
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2258811564963483,
+      "learning_rate": 1.265196138725312e-06,
+      "loss": 0.7099,
+      "step": 35782
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.201485828513419,
+      "learning_rate": 1.2648246704885303e-06,
+      "loss": 0.6262,
+      "step": 35783
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4731940778100296,
+      "learning_rate": 1.2644532531107024e-06,
+      "loss": 0.6559,
+      "step": 35784
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5940237159307011,
+      "learning_rate": 1.2640818865939952e-06,
+      "loss": 0.6531,
+      "step": 35785
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3744329941426598,
+      "learning_rate": 1.2637105709405662e-06,
+      "loss": 0.5482,
+      "step": 35786
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9881545997677428,
+      "learning_rate": 1.2633393061525834e-06,
+      "loss": 0.5556,
+      "step": 35787
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.411480926004902,
+      "learning_rate": 1.262968092232203e-06,
+      "loss": 0.7626,
+      "step": 35788
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2552817206950064,
+      "learning_rate": 1.2625969291815899e-06,
+      "loss": 0.6697,
+      "step": 35789
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5873568294967788,
+      "learning_rate": 1.2622258170029023e-06,
+      "loss": 0.7952,
+      "step": 35790
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2645879442082255,
+      "learning_rate": 1.2618547556983042e-06,
+      "loss": 0.7203,
+      "step": 35791
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.744848215521724,
+      "learning_rate": 1.2614837452699536e-06,
+      "loss": 0.6546,
+      "step": 35792
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6893639676218877,
+      "learning_rate": 1.26111278572001e-06,
+      "loss": 0.6552,
+      "step": 35793
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6479579446630137,
+      "learning_rate": 1.2607418770506342e-06,
+      "loss": 0.509,
+      "step": 35794
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5248654682789484,
+      "learning_rate": 1.2603710192639874e-06,
+      "loss": 0.7768,
+      "step": 35795
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.503160675606056,
+      "learning_rate": 1.2600002123622279e-06,
+      "loss": 0.8253,
+      "step": 35796
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0273635749936303,
+      "learning_rate": 1.259629456347512e-06,
+      "loss": 0.8054,
+      "step": 35797
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5111483681220974,
+      "learning_rate": 1.2592587512220011e-06,
+      "loss": 0.5944,
+      "step": 35798
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2000459172544182,
+      "learning_rate": 1.2588880969878547e-06,
+      "loss": 0.603,
+      "step": 35799
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.403473451984593,
+      "learning_rate": 1.2585174936472288e-06,
+      "loss": 0.7117,
+      "step": 35800
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2247946322696843,
+      "learning_rate": 1.2581469412022806e-06,
+      "loss": 0.7402,
+      "step": 35801
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4040636616848594,
+      "learning_rate": 1.2577764396551673e-06,
+      "loss": 0.7508,
+      "step": 35802
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6094679034133461,
+      "learning_rate": 1.2574059890080502e-06,
+      "loss": 0.4992,
+      "step": 35803
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5327987335296518,
+      "learning_rate": 1.2570355892630825e-06,
+      "loss": 0.8164,
+      "step": 35804
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3499368609035214,
+      "learning_rate": 1.25666524042242e-06,
+      "loss": 0.6904,
+      "step": 35805
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1624938208162692,
+      "learning_rate": 1.2562949424882232e-06,
+      "loss": 0.6323,
+      "step": 35806
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5326030346738837,
+      "learning_rate": 1.2559246954626424e-06,
+      "loss": 0.7171,
+      "step": 35807
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5731583212981957,
+      "learning_rate": 1.2555544993478386e-06,
+      "loss": 0.6219,
+      "step": 35808
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2250830099423664,
+      "learning_rate": 1.255184354145963e-06,
+      "loss": 0.5887,
+      "step": 35809
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4371835692878259,
+      "learning_rate": 1.254814259859175e-06,
+      "loss": 0.7617,
+      "step": 35810
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4358443237637968,
+      "learning_rate": 1.2544442164896254e-06,
+      "loss": 0.7278,
+      "step": 35811
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.710474555513353,
+      "learning_rate": 1.254074224039471e-06,
+      "loss": 0.7098,
+      "step": 35812
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.350978686683094,
+      "learning_rate": 1.2537042825108647e-06,
+      "loss": 0.6236,
+      "step": 35813
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2910058745878614,
+      "learning_rate": 1.2533343919059626e-06,
+      "loss": 0.622,
+      "step": 35814
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5795788263779826,
+      "learning_rate": 1.2529645522269151e-06,
+      "loss": 0.7202,
+      "step": 35815
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4417604131690203,
+      "learning_rate": 1.2525947634758784e-06,
+      "loss": 0.5878,
+      "step": 35816
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.720687708673307,
+      "learning_rate": 1.252225025655005e-06,
+      "loss": 0.6332,
+      "step": 35817
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2661306557238325,
+      "learning_rate": 1.2518553387664456e-06,
+      "loss": 0.6223,
+      "step": 35818
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3324115468467597,
+      "learning_rate": 1.251485702812353e-06,
+      "loss": 0.6486,
+      "step": 35819
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.591096019722577,
+      "learning_rate": 1.2511161177948827e-06,
+      "loss": 0.6039,
+      "step": 35820
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7359112130102246,
+      "learning_rate": 1.2507465837161848e-06,
+      "loss": 0.7378,
+      "step": 35821
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.757516950440842,
+      "learning_rate": 1.2503771005784071e-06,
+      "loss": 0.6792,
+      "step": 35822
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.371055367234298,
+      "learning_rate": 1.2500076683837048e-06,
+      "loss": 0.7383,
+      "step": 35823
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4875879524788815,
+      "learning_rate": 1.2496382871342295e-06,
+      "loss": 0.7374,
+      "step": 35824
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.370577782802991,
+      "learning_rate": 1.2492689568321314e-06,
+      "loss": 0.6293,
+      "step": 35825
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5803478265089693,
+      "learning_rate": 1.248899677479557e-06,
+      "loss": 0.4869,
+      "step": 35826
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.37564069997275,
+      "learning_rate": 1.2485304490786587e-06,
+      "loss": 0.5992,
+      "step": 35827
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3406969189561353,
+      "learning_rate": 1.2481612716315905e-06,
+      "loss": 0.7316,
+      "step": 35828
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3919167725883632,
+      "learning_rate": 1.2477921451404973e-06,
+      "loss": 0.7524,
+      "step": 35829
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.555641430941916,
+      "learning_rate": 1.2474230696075262e-06,
+      "loss": 0.7708,
+      "step": 35830
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2469565700868155,
+      "learning_rate": 1.2470540450348323e-06,
+      "loss": 0.5706,
+      "step": 35831
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3771161266577743,
+      "learning_rate": 1.2466850714245581e-06,
+      "loss": 0.6234,
+      "step": 35832
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.411875444707523,
+      "learning_rate": 1.2463161487788578e-06,
+      "loss": 0.6623,
+      "step": 35833
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5856214272802929,
+      "learning_rate": 1.2459472770998727e-06,
+      "loss": 0.6431,
+      "step": 35834
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5328558229866056,
+      "learning_rate": 1.2455784563897577e-06,
+      "loss": 0.66,
+      "step": 35835
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2454483298551098,
+      "learning_rate": 1.2452096866506536e-06,
+      "loss": 0.7861,
+      "step": 35836
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3632341086943702,
+      "learning_rate": 1.2448409678847117e-06,
+      "loss": 0.7118,
+      "step": 35837
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3882598330310238,
+      "learning_rate": 1.2444723000940785e-06,
+      "loss": 0.5385,
+      "step": 35838
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4846623709376525,
+      "learning_rate": 1.2441036832808973e-06,
+      "loss": 0.6008,
+      "step": 35839
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6084121648941498,
+      "learning_rate": 1.2437351174473167e-06,
+      "loss": 0.7574,
+      "step": 35840
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3693255952759047,
+      "learning_rate": 1.2433666025954815e-06,
+      "loss": 0.5971,
+      "step": 35841
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.828953670993052,
+      "learning_rate": 1.2429981387275435e-06,
+      "loss": 0.6675,
+      "step": 35842
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3457907364665875,
+      "learning_rate": 1.2426297258456388e-06,
+      "loss": 0.6529,
+      "step": 35843
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3600779443786042,
+      "learning_rate": 1.2422613639519154e-06,
+      "loss": 0.6861,
+      "step": 35844
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3810138797582325,
+      "learning_rate": 1.2418930530485219e-06,
+      "loss": 0.6401,
+      "step": 35845
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2847213534436779,
+      "learning_rate": 1.2415247931375996e-06,
+      "loss": 0.6708,
+      "step": 35846
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6250766576521235,
+      "learning_rate": 1.2411565842212902e-06,
+      "loss": 0.5341,
+      "step": 35847
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.758605338456856,
+      "learning_rate": 1.24078842630174e-06,
+      "loss": 0.7846,
+      "step": 35848
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.430754839609702,
+      "learning_rate": 1.240420319381096e-06,
+      "loss": 0.8442,
+      "step": 35849
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.244287693166373,
+      "learning_rate": 1.2400522634614953e-06,
+      "loss": 0.6072,
+      "step": 35850
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5870295522071703,
+      "learning_rate": 1.2396842585450852e-06,
+      "loss": 0.5306,
+      "step": 35851
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2976864798861392,
+      "learning_rate": 1.2393163046340073e-06,
+      "loss": 0.5795,
+      "step": 35852
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.386835149187309,
+      "learning_rate": 1.238948401730402e-06,
+      "loss": 0.6567,
+      "step": 35853
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3116130286092254,
+      "learning_rate": 1.2385805498364124e-06,
+      "loss": 0.7432,
+      "step": 35854
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4109491657290778,
+      "learning_rate": 1.238212748954183e-06,
+      "loss": 0.7209,
+      "step": 35855
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5930678043831096,
+      "learning_rate": 1.2378449990858522e-06,
+      "loss": 0.5167,
+      "step": 35856
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4264966668224952,
+      "learning_rate": 1.2374773002335605e-06,
+      "loss": 0.6378,
+      "step": 35857
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4567974715462062,
+      "learning_rate": 1.2371096523994498e-06,
+      "loss": 0.6915,
+      "step": 35858
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7774882128763383,
+      "learning_rate": 1.2367420555856646e-06,
+      "loss": 0.6738,
+      "step": 35859
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3065288789252216,
+      "learning_rate": 1.236374509794338e-06,
+      "loss": 0.6176,
+      "step": 35860
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2580968142775644,
+      "learning_rate": 1.2360070150276137e-06,
+      "loss": 0.6109,
+      "step": 35861
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3882584111193494,
+      "learning_rate": 1.2356395712876312e-06,
+      "loss": 0.6077,
+      "step": 35862
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7323496382972803,
+      "learning_rate": 1.235272178576531e-06,
+      "loss": 0.481,
+      "step": 35863
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3588854996920419,
+      "learning_rate": 1.2349048368964522e-06,
+      "loss": 0.7512,
+      "step": 35864
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2881437834013587,
+      "learning_rate": 1.234537546249529e-06,
+      "loss": 0.7949,
+      "step": 35865
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.339973332225132,
+      "learning_rate": 1.2341703066379073e-06,
+      "loss": 0.5796,
+      "step": 35866
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.319076548334569,
+      "learning_rate": 1.2338031180637179e-06,
+      "loss": 0.6375,
+      "step": 35867
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3542537927849518,
+      "learning_rate": 1.233435980529104e-06,
+      "loss": 0.7584,
+      "step": 35868
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4946388572522373,
+      "learning_rate": 1.2330688940362002e-06,
+      "loss": 0.6999,
+      "step": 35869
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4151077107569618,
+      "learning_rate": 1.2327018585871463e-06,
+      "loss": 0.683,
+      "step": 35870
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4503368846428821,
+      "learning_rate": 1.2323348741840756e-06,
+      "loss": 0.6762,
+      "step": 35871
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.375506191049042,
+      "learning_rate": 1.2319679408291296e-06,
+      "loss": 0.5356,
+      "step": 35872
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2895549320827961,
+      "learning_rate": 1.231601058524442e-06,
+      "loss": 0.5949,
+      "step": 35873
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4303719456896324,
+      "learning_rate": 1.2312342272721467e-06,
+      "loss": 0.6024,
+      "step": 35874
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.309026228302189,
+      "learning_rate": 1.2308674470743819e-06,
+      "loss": 0.7052,
+      "step": 35875
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.087130183315622,
+      "learning_rate": 1.2305007179332851e-06,
+      "loss": 0.6941,
+      "step": 35876
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4307883109417863,
+      "learning_rate": 1.2301340398509899e-06,
+      "loss": 0.766,
+      "step": 35877
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.474433153048519,
+      "learning_rate": 1.2297674128296288e-06,
+      "loss": 0.6772,
+      "step": 35878
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6092724130678397,
+      "learning_rate": 1.2294008368713385e-06,
+      "loss": 0.5278,
+      "step": 35879
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5908401281628798,
+      "learning_rate": 1.2290343119782545e-06,
+      "loss": 0.4828,
+      "step": 35880
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.329853723258088,
+      "learning_rate": 1.2286678381525108e-06,
+      "loss": 0.604,
+      "step": 35881
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5114310800759168,
+      "learning_rate": 1.2283014153962368e-06,
+      "loss": 0.5783,
+      "step": 35882
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3423652873703615,
+      "learning_rate": 1.227935043711569e-06,
+      "loss": 0.6538,
+      "step": 35883
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2971889079430976,
+      "learning_rate": 1.2275687231006428e-06,
+      "loss": 0.6844,
+      "step": 35884
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2543448146488803,
+      "learning_rate": 1.227202453565589e-06,
+      "loss": 0.7513,
+      "step": 35885
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2763477517572162,
+      "learning_rate": 1.2268362351085383e-06,
+      "loss": 0.6982,
+      "step": 35886
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2593831938974889,
+      "learning_rate": 1.2264700677316266e-06,
+      "loss": 0.6438,
+      "step": 35887
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3700085704256444,
+      "learning_rate": 1.226103951436981e-06,
+      "loss": 0.819,
+      "step": 35888
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4728144429322443,
+      "learning_rate": 1.2257378862267388e-06,
+      "loss": 0.6559,
+      "step": 35889
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.582281095241822,
+      "learning_rate": 1.225371872103026e-06,
+      "loss": 0.7036,
+      "step": 35890
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5628490809538194,
+      "learning_rate": 1.2250059090679779e-06,
+      "loss": 0.6734,
+      "step": 35891
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.566878567739044,
+      "learning_rate": 1.2246399971237221e-06,
+      "loss": 0.7305,
+      "step": 35892
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3479447123895008,
+      "learning_rate": 1.224274136272392e-06,
+      "loss": 0.6434,
+      "step": 35893
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2647486984511949,
+      "learning_rate": 1.2239083265161166e-06,
+      "loss": 0.7121,
+      "step": 35894
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2440129012931194,
+      "learning_rate": 1.2235425678570234e-06,
+      "loss": 0.5917,
+      "step": 35895
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5550187274508263,
+      "learning_rate": 1.2231768602972439e-06,
+      "loss": 0.7019,
+      "step": 35896
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2424310256920787,
+      "learning_rate": 1.2228112038389085e-06,
+      "loss": 0.6557,
+      "step": 35897
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5966696660503086,
+      "learning_rate": 1.2224455984841466e-06,
+      "loss": 0.6987,
+      "step": 35898
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.504982987699897,
+      "learning_rate": 1.2220800442350823e-06,
+      "loss": 0.6669,
+      "step": 35899
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7557531403638542,
+      "learning_rate": 1.2217145410938468e-06,
+      "loss": 0.6715,
+      "step": 35900
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4597174767107257,
+      "learning_rate": 1.2213490890625711e-06,
+      "loss": 0.6493,
+      "step": 35901
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.41130159490225,
+      "learning_rate": 1.2209836881433802e-06,
+      "loss": 0.6509,
+      "step": 35902
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6317582273687378,
+      "learning_rate": 1.2206183383383997e-06,
+      "loss": 0.5212,
+      "step": 35903
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2889824497747318,
+      "learning_rate": 1.2202530396497592e-06,
+      "loss": 0.6221,
+      "step": 35904
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7520091993400881,
+      "learning_rate": 1.2198877920795859e-06,
+      "loss": 0.6686,
+      "step": 35905
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1324196395600095,
+      "learning_rate": 1.2195225956300073e-06,
+      "loss": 0.6626,
+      "step": 35906
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2914091463065165,
+      "learning_rate": 1.2191574503031445e-06,
+      "loss": 0.5738,
+      "step": 35907
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4456067928406262,
+      "learning_rate": 1.2187923561011295e-06,
+      "loss": 0.5738,
+      "step": 35908
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8652988912858162,
+      "learning_rate": 1.2184273130260837e-06,
+      "loss": 0.5346,
+      "step": 35909
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4543564808495637,
+      "learning_rate": 1.2180623210801357e-06,
+      "loss": 0.8617,
+      "step": 35910
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3307944754915317,
+      "learning_rate": 1.2176973802654068e-06,
+      "loss": 0.699,
+      "step": 35911
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3267253982914051,
+      "learning_rate": 1.2173324905840278e-06,
+      "loss": 0.6391,
+      "step": 35912
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4842729309381333,
+      "learning_rate": 1.2169676520381168e-06,
+      "loss": 0.6195,
+      "step": 35913
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2921749191973746,
+      "learning_rate": 1.2166028646298022e-06,
+      "loss": 0.6279,
+      "step": 35914
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3891717416476301,
+      "learning_rate": 1.2162381283612067e-06,
+      "loss": 0.7396,
+      "step": 35915
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.812488971027743,
+      "learning_rate": 1.2158734432344521e-06,
+      "loss": 0.6952,
+      "step": 35916
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5543660338652777,
+      "learning_rate": 1.2155088092516632e-06,
+      "loss": 0.5088,
+      "step": 35917
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6099734089690658,
+      "learning_rate": 1.215144226414965e-06,
+      "loss": 0.6493,
+      "step": 35918
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9910210476330517,
+      "learning_rate": 1.214779694726479e-06,
+      "loss": 0.7631,
+      "step": 35919
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.249495163921912,
+      "learning_rate": 1.2144152141883247e-06,
+      "loss": 0.627,
+      "step": 35920
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6870327562981604,
+      "learning_rate": 1.2140507848026261e-06,
+      "loss": 0.652,
+      "step": 35921
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3346486144629635,
+      "learning_rate": 1.213686406571507e-06,
+      "loss": 0.6961,
+      "step": 35922
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6520900056356773,
+      "learning_rate": 1.2133220794970879e-06,
+      "loss": 0.6617,
+      "step": 35923
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4421957173866011,
+      "learning_rate": 1.2129578035814881e-06,
+      "loss": 0.6067,
+      "step": 35924
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4735888036875484,
+      "learning_rate": 1.2125935788268296e-06,
+      "loss": 0.6963,
+      "step": 35925
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.572526712379022,
+      "learning_rate": 1.2122294052352347e-06,
+      "loss": 0.511,
+      "step": 35926
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3487547826028954,
+      "learning_rate": 1.211865282808823e-06,
+      "loss": 0.7437,
+      "step": 35927
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5244751597594348,
+      "learning_rate": 1.2115012115497115e-06,
+      "loss": 0.8047,
+      "step": 35928
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.35286895561356,
+      "learning_rate": 1.2111371914600245e-06,
+      "loss": 0.7335,
+      "step": 35929
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4796144837701948,
+      "learning_rate": 1.2107732225418766e-06,
+      "loss": 0.6496,
+      "step": 35930
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2523280270070665,
+      "learning_rate": 1.210409304797392e-06,
+      "loss": 0.6728,
+      "step": 35931
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6139541892225381,
+      "learning_rate": 1.2100454382286853e-06,
+      "loss": 0.5163,
+      "step": 35932
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4910207092530865,
+      "learning_rate": 1.2096816228378794e-06,
+      "loss": 0.6433,
+      "step": 35933
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3491909115796028,
+      "learning_rate": 1.209317858627087e-06,
+      "loss": 0.678,
+      "step": 35934
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.544119196620394,
+      "learning_rate": 1.2089541455984321e-06,
+      "loss": 0.6168,
+      "step": 35935
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7165320298125328,
+      "learning_rate": 1.2085904837540275e-06,
+      "loss": 0.5785,
+      "step": 35936
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3368561433053834,
+      "learning_rate": 1.2082268730959935e-06,
+      "loss": 0.6923,
+      "step": 35937
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3009298143045198,
+      "learning_rate": 1.207863313626445e-06,
+      "loss": 0.7409,
+      "step": 35938
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3788505415167458,
+      "learning_rate": 1.2074998053474984e-06,
+      "loss": 0.6888,
+      "step": 35939
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4044193788280754,
+      "learning_rate": 1.2071363482612775e-06,
+      "loss": 0.6453,
+      "step": 35940
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6011771317397246,
+      "learning_rate": 1.2067729423698871e-06,
+      "loss": 0.5401,
+      "step": 35941
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4758180766883717,
+      "learning_rate": 1.206409587675449e-06,
+      "loss": 0.7338,
+      "step": 35942
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4651956682420613,
+      "learning_rate": 1.2060462841800779e-06,
+      "loss": 0.6323,
+      "step": 35943
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3451657201664593,
+      "learning_rate": 1.2056830318858947e-06,
+      "loss": 0.5476,
+      "step": 35944
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8918743335060415,
+      "learning_rate": 1.2053198307950042e-06,
+      "loss": 0.6804,
+      "step": 35945
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4478849245081475,
+      "learning_rate": 1.2049566809095259e-06,
+      "loss": 0.5998,
+      "step": 35946
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4007058769817613,
+      "learning_rate": 1.2045935822315768e-06,
+      "loss": 0.6108,
+      "step": 35947
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4242734613244432,
+      "learning_rate": 1.2042305347632655e-06,
+      "loss": 0.5557,
+      "step": 35948
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.304305924194333,
+      "learning_rate": 1.2038675385067112e-06,
+      "loss": 0.7031,
+      "step": 35949
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5775063625765112,
+      "learning_rate": 1.2035045934640233e-06,
+      "loss": 0.5192,
+      "step": 35950
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.790219770938241,
+      "learning_rate": 1.2031416996373179e-06,
+      "loss": 0.6538,
+      "step": 35951
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3925934943343996,
+      "learning_rate": 1.2027788570287047e-06,
+      "loss": 0.8254,
+      "step": 35952
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.773770679689139,
+      "learning_rate": 1.2024160656403006e-06,
+      "loss": 0.7001,
+      "step": 35953
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5081281838970908,
+      "learning_rate": 1.202053325474215e-06,
+      "loss": 0.73,
+      "step": 35954
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6914719999581653,
+      "learning_rate": 1.2016906365325586e-06,
+      "loss": 0.5946,
+      "step": 35955
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5951530703763128,
+      "learning_rate": 1.2013279988174454e-06,
+      "loss": 0.665,
+      "step": 35956
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.274944758631932,
+      "learning_rate": 1.2009654123309878e-06,
+      "loss": 0.7331,
+      "step": 35957
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3979892855297706,
+      "learning_rate": 1.2006028770752942e-06,
+      "loss": 0.655,
+      "step": 35958
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4964945830416385,
+      "learning_rate": 1.2002403930524752e-06,
+      "loss": 0.6694,
+      "step": 35959
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4107590988562397,
+      "learning_rate": 1.1998779602646438e-06,
+      "loss": 0.7036,
+      "step": 35960
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5950998987908371,
+      "learning_rate": 1.199515578713909e-06,
+      "loss": 0.4804,
+      "step": 35961
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3287580424764471,
+      "learning_rate": 1.1991532484023816e-06,
+      "loss": 0.6052,
+      "step": 35962
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3775989941031133,
+      "learning_rate": 1.1987909693321687e-06,
+      "loss": 0.7132,
+      "step": 35963
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6123363737903864,
+      "learning_rate": 1.1984287415053807e-06,
+      "loss": 0.6777,
+      "step": 35964
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5566646720071806,
+      "learning_rate": 1.1980665649241297e-06,
+      "loss": 0.7921,
+      "step": 35965
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.315056483020775,
+      "learning_rate": 1.1977044395905214e-06,
+      "loss": 0.8386,
+      "step": 35966
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3589343334493065,
+      "learning_rate": 1.1973423655066619e-06,
+      "loss": 0.6278,
+      "step": 35967
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5550061529141069,
+      "learning_rate": 1.1969803426746651e-06,
+      "loss": 0.5119,
+      "step": 35968
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.543297399870762,
+      "learning_rate": 1.1966183710966329e-06,
+      "loss": 0.7637,
+      "step": 35969
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.557606154771973,
+      "learning_rate": 1.1962564507746776e-06,
+      "loss": 0.7667,
+      "step": 35970
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3925173675232527,
+      "learning_rate": 1.1958945817109035e-06,
+      "loss": 0.5973,
+      "step": 35971
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.26095436208927,
+      "learning_rate": 1.1955327639074198e-06,
+      "loss": 0.6357,
+      "step": 35972
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.410235630895128,
+      "learning_rate": 1.1951709973663294e-06,
+      "loss": 0.6697,
+      "step": 35973
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4478259626628966,
+      "learning_rate": 1.194809282089744e-06,
+      "loss": 0.7353,
+      "step": 35974
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3746084502545437,
+      "learning_rate": 1.194447618079766e-06,
+      "loss": 0.7507,
+      "step": 35975
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6426831542043454,
+      "learning_rate": 1.1940860053384994e-06,
+      "loss": 0.6794,
+      "step": 35976
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5229974561263735,
+      "learning_rate": 1.1937244438680529e-06,
+      "loss": 0.6737,
+      "step": 35977
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4462719884454776,
+      "learning_rate": 1.1933629336705322e-06,
+      "loss": 0.5795,
+      "step": 35978
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.419315567145768,
+      "learning_rate": 1.19300147474804e-06,
+      "loss": 0.6108,
+      "step": 35979
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.379486904701441,
+      "learning_rate": 1.1926400671026805e-06,
+      "loss": 0.5738,
+      "step": 35980
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3196314982872968,
+      "learning_rate": 1.1922787107365585e-06,
+      "loss": 0.5261,
+      "step": 35981
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.599291126897465,
+      "learning_rate": 1.19191740565178e-06,
+      "loss": 0.706,
+      "step": 35982
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2874622946038932,
+      "learning_rate": 1.1915561518504482e-06,
+      "loss": 0.6743,
+      "step": 35983
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3542713223586118,
+      "learning_rate": 1.191194949334662e-06,
+      "loss": 0.6967,
+      "step": 35984
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1353093829647576,
+      "learning_rate": 1.1908337981065276e-06,
+      "loss": 0.7594,
+      "step": 35985
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2954033908243154,
+      "learning_rate": 1.1904726981681502e-06,
+      "loss": 0.7383,
+      "step": 35986
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9064059388157926,
+      "learning_rate": 1.1901116495216303e-06,
+      "loss": 0.6759,
+      "step": 35987
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3270768105207422,
+      "learning_rate": 1.189750652169066e-06,
+      "loss": 0.6369,
+      "step": 35988
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.941679150910282,
+      "learning_rate": 1.189389706112566e-06,
+      "loss": 0.7305,
+      "step": 35989
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5271420394978044,
+      "learning_rate": 1.189028811354226e-06,
+      "loss": 0.7197,
+      "step": 35990
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3796976430029935,
+      "learning_rate": 1.188667967896152e-06,
+      "loss": 0.644,
+      "step": 35991
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6929350056052888,
+      "learning_rate": 1.1883071757404408e-06,
+      "loss": 0.7828,
+      "step": 35992
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2829768146318035,
+      "learning_rate": 1.1879464348891967e-06,
+      "loss": 0.5437,
+      "step": 35993
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2848002960946399,
+      "learning_rate": 1.1875857453445162e-06,
+      "loss": 0.7225,
+      "step": 35994
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2341626720332532,
+      "learning_rate": 1.1872251071085038e-06,
+      "loss": 0.5846,
+      "step": 35995
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2436355066577855,
+      "learning_rate": 1.1868645201832563e-06,
+      "loss": 0.556,
+      "step": 35996
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.536815527491998,
+      "learning_rate": 1.1865039845708725e-06,
+      "loss": 0.639,
+      "step": 35997
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5004866312691654,
+      "learning_rate": 1.1861435002734523e-06,
+      "loss": 0.7768,
+      "step": 35998
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3428043246834949,
+      "learning_rate": 1.1857830672930982e-06,
+      "loss": 0.7351,
+      "step": 35999
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3620837194552269,
+      "learning_rate": 1.185422685631904e-06,
+      "loss": 0.7012,
+      "step": 36000
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5873797292108803,
+      "learning_rate": 1.1850623552919683e-06,
+      "loss": 0.5435,
+      "step": 36001
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.357085276240555,
+      "learning_rate": 1.1847020762753903e-06,
+      "loss": 0.715,
+      "step": 36002
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4319786301077082,
+      "learning_rate": 1.1843418485842696e-06,
+      "loss": 0.7415,
+      "step": 36003
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3004681233874549,
+      "learning_rate": 1.183981672220702e-06,
+      "loss": 0.6831,
+      "step": 36004
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.920893841611247,
+      "learning_rate": 1.1836215471867819e-06,
+      "loss": 0.6198,
+      "step": 36005
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9572943438734893,
+      "learning_rate": 1.1832614734846082e-06,
+      "loss": 0.7284,
+      "step": 36006
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4599899259511289,
+      "learning_rate": 1.1829014511162794e-06,
+      "loss": 0.7455,
+      "step": 36007
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4541329004751786,
+      "learning_rate": 1.1825414800838908e-06,
+      "loss": 0.5379,
+      "step": 36008
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5911326164737495,
+      "learning_rate": 1.1821815603895336e-06,
+      "loss": 0.4724,
+      "step": 36009
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4139573537546781,
+      "learning_rate": 1.1818216920353109e-06,
+      "loss": 0.7444,
+      "step": 36010
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3452906962564735,
+      "learning_rate": 1.181461875023311e-06,
+      "loss": 0.5973,
+      "step": 36011
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2530502994517718,
+      "learning_rate": 1.1811021093556341e-06,
+      "loss": 0.6586,
+      "step": 36012
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.441673169795898,
+      "learning_rate": 1.180742395034371e-06,
+      "loss": 0.7143,
+      "step": 36013
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5600203602505487,
+      "learning_rate": 1.180382732061619e-06,
+      "loss": 0.7322,
+      "step": 36014
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6197444585718895,
+      "learning_rate": 1.1800231204394696e-06,
+      "loss": 0.4903,
+      "step": 36015
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4533654377730907,
+      "learning_rate": 1.1796635601700201e-06,
+      "loss": 0.7668,
+      "step": 36016
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.41494926794793,
+      "learning_rate": 1.1793040512553623e-06,
+      "loss": 0.7576,
+      "step": 36017
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.475294561237924,
+      "learning_rate": 1.1789445936975862e-06,
+      "loss": 0.7769,
+      "step": 36018
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9135473317484384,
+      "learning_rate": 1.1785851874987874e-06,
+      "loss": 0.7485,
+      "step": 36019
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3837466802827265,
+      "learning_rate": 1.1782258326610606e-06,
+      "loss": 0.7071,
+      "step": 36020
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4343260902445345,
+      "learning_rate": 1.1778665291864955e-06,
+      "loss": 0.6882,
+      "step": 36021
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4695575944017132,
+      "learning_rate": 1.1775072770771833e-06,
+      "loss": 0.7377,
+      "step": 36022
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2047476184791874,
+      "learning_rate": 1.177148076335216e-06,
+      "loss": 0.6127,
+      "step": 36023
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6586906918012367,
+      "learning_rate": 1.1767889269626887e-06,
+      "loss": 0.6939,
+      "step": 36024
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1649000913155887,
+      "learning_rate": 1.1764298289616893e-06,
+      "loss": 0.589,
+      "step": 36025
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6367983441713736,
+      "learning_rate": 1.1760707823343065e-06,
+      "loss": 0.4912,
+      "step": 36026
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3104785187150625,
+      "learning_rate": 1.1757117870826328e-06,
+      "loss": 0.7666,
+      "step": 36027
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3868858572359049,
+      "learning_rate": 1.1753528432087624e-06,
+      "loss": 0.7115,
+      "step": 36028
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.77939709062437,
+      "learning_rate": 1.174993950714781e-06,
+      "loss": 0.6916,
+      "step": 36029
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2805787529733255,
+      "learning_rate": 1.1746351096027763e-06,
+      "loss": 0.597,
+      "step": 36030
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5545165314863905,
+      "learning_rate": 1.174276319874842e-06,
+      "loss": 0.6715,
+      "step": 36031
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2761609056283838,
+      "learning_rate": 1.173917581533064e-06,
+      "loss": 0.6672,
+      "step": 36032
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2279392546898724,
+      "learning_rate": 1.1735588945795307e-06,
+      "loss": 0.6374,
+      "step": 36033
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.591759231231945,
+      "learning_rate": 1.1732002590163349e-06,
+      "loss": 0.4999,
+      "step": 36034
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3828165363777176,
+      "learning_rate": 1.1728416748455619e-06,
+      "loss": 0.7721,
+      "step": 36035
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5881715429904474,
+      "learning_rate": 1.1724831420692973e-06,
+      "loss": 0.5219,
+      "step": 36036
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8584143284829504,
+      "learning_rate": 1.1721246606896297e-06,
+      "loss": 0.5577,
+      "step": 36037
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.32476175124819,
+      "learning_rate": 1.171766230708652e-06,
+      "loss": 0.7221,
+      "step": 36038
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4382778140770824,
+      "learning_rate": 1.1714078521284422e-06,
+      "loss": 0.7299,
+      "step": 36039
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4058694695908298,
+      "learning_rate": 1.1710495249510911e-06,
+      "loss": 0.7073,
+      "step": 36040
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3791875108304184,
+      "learning_rate": 1.1706912491786837e-06,
+      "loss": 0.7376,
+      "step": 36041
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4369879415050486,
+      "learning_rate": 1.1703330248133128e-06,
+      "loss": 0.7977,
+      "step": 36042
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8062224935095974,
+      "learning_rate": 1.169974851857053e-06,
+      "loss": 0.6021,
+      "step": 36043
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7302553540724896,
+      "learning_rate": 1.1696167303119964e-06,
+      "loss": 0.656,
+      "step": 36044
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5660292506914955,
+      "learning_rate": 1.169258660180228e-06,
+      "loss": 0.7772,
+      "step": 36045
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3163509035471752,
+      "learning_rate": 1.1689006414638294e-06,
+      "loss": 0.573,
+      "step": 36046
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4868680067594153,
+      "learning_rate": 1.1685426741648897e-06,
+      "loss": 0.6403,
+      "step": 36047
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1634064417962922,
+      "learning_rate": 1.168184758285489e-06,
+      "loss": 0.5486,
+      "step": 36048
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4433940917877341,
+      "learning_rate": 1.1678268938277137e-06,
+      "loss": 0.5519,
+      "step": 36049
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4743560314049162,
+      "learning_rate": 1.167469080793645e-06,
+      "loss": 0.7252,
+      "step": 36050
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4681559689628356,
+      "learning_rate": 1.1671113191853688e-06,
+      "loss": 0.7135,
+      "step": 36051
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3450579924848545,
+      "learning_rate": 1.166753609004968e-06,
+      "loss": 0.6663,
+      "step": 36052
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6530044039656115,
+      "learning_rate": 1.1663959502545218e-06,
+      "loss": 0.7148,
+      "step": 36053
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6085010465728268,
+      "learning_rate": 1.1660383429361155e-06,
+      "loss": 0.534,
+      "step": 36054
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2523012262317417,
+      "learning_rate": 1.1656807870518316e-06,
+      "loss": 0.6289,
+      "step": 36055
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3563431410449465,
+      "learning_rate": 1.165323282603752e-06,
+      "loss": 0.7083,
+      "step": 36056
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5128275780629368,
+      "learning_rate": 1.1649658295939548e-06,
+      "loss": 0.7325,
+      "step": 36057
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2733965111938998,
+      "learning_rate": 1.164608428024524e-06,
+      "loss": 0.5104,
+      "step": 36058
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.46613416425616,
+      "learning_rate": 1.1642510778975425e-06,
+      "loss": 0.7821,
+      "step": 36059
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5544223578496834,
+      "learning_rate": 1.1638937792150883e-06,
+      "loss": 0.7945,
+      "step": 36060
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7944488349011032,
+      "learning_rate": 1.1635365319792402e-06,
+      "loss": 0.6338,
+      "step": 36061
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4432673629418977,
+      "learning_rate": 1.1631793361920795e-06,
+      "loss": 0.6958,
+      "step": 36062
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5567731523054158,
+      "learning_rate": 1.1628221918556882e-06,
+      "loss": 0.6324,
+      "step": 36063
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3607954284166772,
+      "learning_rate": 1.1624650989721443e-06,
+      "loss": 0.6813,
+      "step": 36064
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3984791483334673,
+      "learning_rate": 1.162108057543524e-06,
+      "loss": 0.7289,
+      "step": 36065
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.413046461774798,
+      "learning_rate": 1.1617510675719091e-06,
+      "loss": 0.649,
+      "step": 36066
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3686552608126665,
+      "learning_rate": 1.16139412905938e-06,
+      "loss": 0.5924,
+      "step": 36067
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5494371645084206,
+      "learning_rate": 1.1610372420080118e-06,
+      "loss": 0.768,
+      "step": 36068
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5526250561770059,
+      "learning_rate": 1.1606804064198817e-06,
+      "loss": 0.654,
+      "step": 36069
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.382232661720269,
+      "learning_rate": 1.1603236222970704e-06,
+      "loss": 0.6629,
+      "step": 36070
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5758889830469521,
+      "learning_rate": 1.1599668896416516e-06,
+      "loss": 0.441,
+      "step": 36071
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6018522951564912,
+      "learning_rate": 1.159610208455707e-06,
+      "loss": 0.5339,
+      "step": 36072
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5588643882910826,
+      "learning_rate": 1.1592535787413074e-06,
+      "loss": 0.5649,
+      "step": 36073
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9290952670404478,
+      "learning_rate": 1.158897000500535e-06,
+      "loss": 0.6714,
+      "step": 36074
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7224481822605364,
+      "learning_rate": 1.158540473735461e-06,
+      "loss": 0.7078,
+      "step": 36075
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3661810595143011,
+      "learning_rate": 1.1581839984481657e-06,
+      "loss": 0.8527,
+      "step": 36076
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.41670558143497,
+      "learning_rate": 1.157827574640723e-06,
+      "loss": 0.728,
+      "step": 36077
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1484241029163134,
+      "learning_rate": 1.1574712023152058e-06,
+      "loss": 0.6708,
+      "step": 36078
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4348894838876245,
+      "learning_rate": 1.1571148814736898e-06,
+      "loss": 0.6485,
+      "step": 36079
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5482600111855462,
+      "learning_rate": 1.1567586121182529e-06,
+      "loss": 0.6236,
+      "step": 36080
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.324680567731741,
+      "learning_rate": 1.1564023942509662e-06,
+      "loss": 0.571,
+      "step": 36081
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4188806069317954,
+      "learning_rate": 1.1560462278739037e-06,
+      "loss": 0.7245,
+      "step": 36082
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2002108039939632,
+      "learning_rate": 1.1556901129891396e-06,
+      "loss": 0.5808,
+      "step": 36083
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6344531491726022,
+      "learning_rate": 1.1553340495987497e-06,
+      "loss": 0.748,
+      "step": 36084
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4157183482678923,
+      "learning_rate": 1.154978037704806e-06,
+      "loss": 0.7662,
+      "step": 36085
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.42546801115625,
+      "learning_rate": 1.1546220773093774e-06,
+      "loss": 0.7012,
+      "step": 36086
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4918394908960473,
+      "learning_rate": 1.1542661684145396e-06,
+      "loss": 0.7101,
+      "step": 36087
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2522146805191698,
+      "learning_rate": 1.153910311022367e-06,
+      "loss": 0.5755,
+      "step": 36088
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4809347889701012,
+      "learning_rate": 1.1535545051349284e-06,
+      "loss": 0.6359,
+      "step": 36089
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4096183024235167,
+      "learning_rate": 1.1531987507542953e-06,
+      "loss": 0.7138,
+      "step": 36090
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6599097895087528,
+      "learning_rate": 1.1528430478825414e-06,
+      "loss": 0.4981,
+      "step": 36091
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.566789528745906,
+      "learning_rate": 1.1524873965217342e-06,
+      "loss": 0.4681,
+      "step": 36092
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5943394945899586,
+      "learning_rate": 1.1521317966739486e-06,
+      "loss": 0.4807,
+      "step": 36093
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2697602172156577,
+      "learning_rate": 1.1517762483412509e-06,
+      "loss": 0.5851,
+      "step": 36094
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3176625514823181,
+      "learning_rate": 1.1514207515257147e-06,
+      "loss": 0.6315,
+      "step": 36095
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6688000011496475,
+      "learning_rate": 1.1510653062294075e-06,
+      "loss": 0.6468,
+      "step": 36096
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4476993911170468,
+      "learning_rate": 1.1507099124544008e-06,
+      "loss": 0.7285,
+      "step": 36097
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.623657369889684,
+      "learning_rate": 1.1503545702027619e-06,
+      "loss": 0.6947,
+      "step": 36098
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6217515228150116,
+      "learning_rate": 1.1499992794765592e-06,
+      "loss": 0.531,
+      "step": 36099
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5197697915697945,
+      "learning_rate": 1.149644040277863e-06,
+      "loss": 0.6761,
+      "step": 36100
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4348335943134733,
+      "learning_rate": 1.149288852608743e-06,
+      "loss": 0.6657,
+      "step": 36101
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.292414003065786,
+      "learning_rate": 1.1489337164712644e-06,
+      "loss": 0.7381,
+      "step": 36102
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3594962480924815,
+      "learning_rate": 1.1485786318674951e-06,
+      "loss": 0.6636,
+      "step": 36103
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0831517672877147,
+      "learning_rate": 1.1482235987995027e-06,
+      "loss": 0.6274,
+      "step": 36104
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3856832870300946,
+      "learning_rate": 1.1478686172693576e-06,
+      "loss": 0.8277,
+      "step": 36105
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3896175138641096,
+      "learning_rate": 1.1475136872791226e-06,
+      "loss": 0.7028,
+      "step": 36106
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.547828347162886,
+      "learning_rate": 1.1471588088308649e-06,
+      "loss": 0.5516,
+      "step": 36107
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.543728191141386,
+      "learning_rate": 1.1468039819266508e-06,
+      "loss": 0.7535,
+      "step": 36108
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.1318047481812465,
+      "learning_rate": 1.1464492065685483e-06,
+      "loss": 0.6837,
+      "step": 36109
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6012418517188007,
+      "learning_rate": 1.1460944827586228e-06,
+      "loss": 0.6962,
+      "step": 36110
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4366873112047722,
+      "learning_rate": 1.1457398104989347e-06,
+      "loss": 0.7323,
+      "step": 36111
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3737548399510513,
+      "learning_rate": 1.1453851897915557e-06,
+      "loss": 0.7177,
+      "step": 36112
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8393518611781061,
+      "learning_rate": 1.1450306206385453e-06,
+      "loss": 0.672,
+      "step": 36113
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5843677943556465,
+      "learning_rate": 1.1446761030419718e-06,
+      "loss": 0.5391,
+      "step": 36114
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3473594739248178,
+      "learning_rate": 1.1443216370038968e-06,
+      "loss": 0.5979,
+      "step": 36115
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3066080955940715,
+      "learning_rate": 1.1439672225263853e-06,
+      "loss": 0.5983,
+      "step": 36116
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3229302238611407,
+      "learning_rate": 1.1436128596114992e-06,
+      "loss": 0.735,
+      "step": 36117
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9310047607779028,
+      "learning_rate": 1.1432585482613045e-06,
+      "loss": 0.7742,
+      "step": 36118
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2773194577547646,
+      "learning_rate": 1.142904288477863e-06,
+      "loss": 0.6171,
+      "step": 36119
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6164104825238537,
+      "learning_rate": 1.142550080263234e-06,
+      "loss": 0.5081,
+      "step": 36120
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.353470017507589,
+      "learning_rate": 1.1421959236194835e-06,
+      "loss": 0.6305,
+      "step": 36121
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6130759461103912,
+      "learning_rate": 1.1418418185486725e-06,
+      "loss": 0.7715,
+      "step": 36122
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5868969159942237,
+      "learning_rate": 1.1414877650528667e-06,
+      "loss": 0.8256,
+      "step": 36123
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3744195141110915,
+      "learning_rate": 1.1411337631341202e-06,
+      "loss": 0.565,
+      "step": 36124
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3918590681242846,
+      "learning_rate": 1.140779812794497e-06,
+      "loss": 0.7513,
+      "step": 36125
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4647188463306053,
+      "learning_rate": 1.1404259140360606e-06,
+      "loss": 0.6685,
+      "step": 36126
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6419800185105076,
+      "learning_rate": 1.140072066860869e-06,
+      "loss": 0.7068,
+      "step": 36127
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6913417879485424,
+      "learning_rate": 1.139718271270982e-06,
+      "loss": 0.7146,
+      "step": 36128
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.269569444998801,
+      "learning_rate": 1.1393645272684595e-06,
+      "loss": 0.7378,
+      "step": 36129
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.559717600574844,
+      "learning_rate": 1.139010834855364e-06,
+      "loss": 0.7467,
+      "step": 36130
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2978167239023992,
+      "learning_rate": 1.138657194033751e-06,
+      "loss": 0.6492,
+      "step": 36131
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3540262715206595,
+      "learning_rate": 1.1383036048056828e-06,
+      "loss": 0.6412,
+      "step": 36132
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.326382027640302,
+      "learning_rate": 1.137950067173217e-06,
+      "loss": 0.6897,
+      "step": 36133
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6961884170371968,
+      "learning_rate": 1.1375965811384093e-06,
+      "loss": 0.8138,
+      "step": 36134
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6930368186394518,
+      "learning_rate": 1.1372431467033208e-06,
+      "loss": 0.8039,
+      "step": 36135
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5556848781838122,
+      "learning_rate": 1.1368897638700105e-06,
+      "loss": 0.6006,
+      "step": 36136
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3426654715373363,
+      "learning_rate": 1.1365364326405325e-06,
+      "loss": 0.5986,
+      "step": 36137
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3164914787878257,
+      "learning_rate": 1.1361831530169453e-06,
+      "loss": 0.6753,
+      "step": 36138
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6463340498473394,
+      "learning_rate": 1.1358299250013049e-06,
+      "loss": 0.4804,
+      "step": 36139
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.396564744173891,
+      "learning_rate": 1.1354767485956741e-06,
+      "loss": 0.7068,
+      "step": 36140
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6016215752727407,
+      "learning_rate": 1.135123623802099e-06,
+      "loss": 0.5977,
+      "step": 36141
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4875070357798834,
+      "learning_rate": 1.1347705506226415e-06,
+      "loss": 0.5499,
+      "step": 36142
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6000317067431158,
+      "learning_rate": 1.1344175290593563e-06,
+      "loss": 0.4855,
+      "step": 36143
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6785496072739986,
+      "learning_rate": 1.1340645591143007e-06,
+      "loss": 0.5462,
+      "step": 36144
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3587188070711596,
+      "learning_rate": 1.1337116407895289e-06,
+      "loss": 0.5287,
+      "step": 36145
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3652258603028486,
+      "learning_rate": 1.1333587740870921e-06,
+      "loss": 0.7814,
+      "step": 36146
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0380448569764673,
+      "learning_rate": 1.1330059590090502e-06,
+      "loss": 0.7804,
+      "step": 36147
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5702022691002453,
+      "learning_rate": 1.1326531955574526e-06,
+      "loss": 0.668,
+      "step": 36148
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3738601490491134,
+      "learning_rate": 1.1323004837343576e-06,
+      "loss": 0.7377,
+      "step": 36149
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5675631296487802,
+      "learning_rate": 1.1319478235418146e-06,
+      "loss": 0.6859,
+      "step": 36150
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4439874571751594,
+      "learning_rate": 1.131595214981881e-06,
+      "loss": 0.5904,
+      "step": 36151
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6200903863611035,
+      "learning_rate": 1.131242658056606e-06,
+      "loss": 0.5105,
+      "step": 36152
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.335190374742151,
+      "learning_rate": 1.130890152768046e-06,
+      "loss": 0.5576,
+      "step": 36153
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8241038071812814,
+      "learning_rate": 1.1305376991182516e-06,
+      "loss": 0.7406,
+      "step": 36154
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4002352750718656,
+      "learning_rate": 1.130185297109272e-06,
+      "loss": 0.7477,
+      "step": 36155
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6260337036948369,
+      "learning_rate": 1.1298329467431624e-06,
+      "loss": 0.5254,
+      "step": 36156
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3274561608463842,
+      "learning_rate": 1.1294806480219744e-06,
+      "loss": 0.6947,
+      "step": 36157
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2948981164425588,
+      "learning_rate": 1.1291284009477598e-06,
+      "loss": 0.5681,
+      "step": 36158
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2837826081695765,
+      "learning_rate": 1.128776205522566e-06,
+      "loss": 0.6673,
+      "step": 36159
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4119696898823861,
+      "learning_rate": 1.1284240617484455e-06,
+      "loss": 0.6933,
+      "step": 36160
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3940005632730113,
+      "learning_rate": 1.12807196962745e-06,
+      "loss": 0.6254,
+      "step": 36161
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3688158775680166,
+      "learning_rate": 1.1277199291616291e-06,
+      "loss": 0.6334,
+      "step": 36162
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3640381894230482,
+      "learning_rate": 1.127367940353029e-06,
+      "loss": 0.706,
+      "step": 36163
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.815425346134878,
+      "learning_rate": 1.1270160032037024e-06,
+      "loss": 0.8859,
+      "step": 36164
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4780307191372914,
+      "learning_rate": 1.1266641177157e-06,
+      "loss": 0.6954,
+      "step": 36165
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.602440057807414,
+      "learning_rate": 1.126312283891069e-06,
+      "loss": 0.5078,
+      "step": 36166
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3369969487989601,
+      "learning_rate": 1.1259605017318543e-06,
+      "loss": 0.7692,
+      "step": 36167
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3994387226230467,
+      "learning_rate": 1.1256087712401087e-06,
+      "loss": 0.6799,
+      "step": 36168
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.299118171032082,
+      "learning_rate": 1.1252570924178763e-06,
+      "loss": 0.5846,
+      "step": 36169
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3401648117025915,
+      "learning_rate": 1.1249054652672097e-06,
+      "loss": 0.5873,
+      "step": 36170
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4076011884861408,
+      "learning_rate": 1.124553889790151e-06,
+      "loss": 0.5647,
+      "step": 36171
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3223588949786895,
+      "learning_rate": 1.1242023659887513e-06,
+      "loss": 0.6149,
+      "step": 36172
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3025762215544507,
+      "learning_rate": 1.1238508938650527e-06,
+      "loss": 0.573,
+      "step": 36173
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.361488041058433,
+      "learning_rate": 1.123499473421108e-06,
+      "loss": 0.6769,
+      "step": 36174
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5013712462176303,
+      "learning_rate": 1.1231481046589587e-06,
+      "loss": 0.808,
+      "step": 36175
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3146959006391856,
+      "learning_rate": 1.1227967875806489e-06,
+      "loss": 0.6519,
+      "step": 36176
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.217449371210125,
+      "learning_rate": 1.122445522188228e-06,
+      "loss": 0.621,
+      "step": 36177
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.430370459083492,
+      "learning_rate": 1.122094308483741e-06,
+      "loss": 0.7157,
+      "step": 36178
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3460447813349938,
+      "learning_rate": 1.121743146469232e-06,
+      "loss": 0.6404,
+      "step": 36179
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3197907600424748,
+      "learning_rate": 1.1213920361467422e-06,
+      "loss": 0.7581,
+      "step": 36180
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5991069300038854,
+      "learning_rate": 1.1210409775183207e-06,
+      "loss": 0.647,
+      "step": 36181
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.531187115998596,
+      "learning_rate": 1.1206899705860097e-06,
+      "loss": 0.6201,
+      "step": 36182
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.33174990525809,
+      "learning_rate": 1.1203390153518546e-06,
+      "loss": 0.6439,
+      "step": 36183
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3615621174715773,
+      "learning_rate": 1.1199881118178934e-06,
+      "loss": 0.7731,
+      "step": 36184
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.534647251482612,
+      "learning_rate": 1.1196372599861738e-06,
+      "loss": 0.7419,
+      "step": 36185
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3098309972479034,
+      "learning_rate": 1.1192864598587395e-06,
+      "loss": 0.7013,
+      "step": 36186
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3696721785810284,
+      "learning_rate": 1.1189357114376308e-06,
+      "loss": 0.5344,
+      "step": 36187
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.654196259062336,
+      "learning_rate": 1.1185850147248879e-06,
+      "loss": 0.6777,
+      "step": 36188
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.6550721264541637,
+      "learning_rate": 1.118234369722555e-06,
+      "loss": 0.7054,
+      "step": 36189
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5391230766610962,
+      "learning_rate": 1.1178837764326766e-06,
+      "loss": 0.6825,
+      "step": 36190
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3479967001176663,
+      "learning_rate": 1.1175332348572898e-06,
+      "loss": 0.68,
+      "step": 36191
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2677671135549236,
+      "learning_rate": 1.1171827449984341e-06,
+      "loss": 0.6897,
+      "step": 36192
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3942638087738697,
+      "learning_rate": 1.1168323068581555e-06,
+      "loss": 0.7192,
+      "step": 36193
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.318337429192902,
+      "learning_rate": 1.1164819204384891e-06,
+      "loss": 0.7447,
+      "step": 36194
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3837250549378204,
+      "learning_rate": 1.11613158574148e-06,
+      "loss": 0.6991,
+      "step": 36195
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4734667221147357,
+      "learning_rate": 1.1157813027691632e-06,
+      "loss": 0.8001,
+      "step": 36196
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3454027804770912,
+      "learning_rate": 1.1154310715235827e-06,
+      "loss": 0.6833,
+      "step": 36197
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4871939851387754,
+      "learning_rate": 1.1150808920067723e-06,
+      "loss": 0.6691,
+      "step": 36198
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3658196056239318,
+      "learning_rate": 1.1147307642207771e-06,
+      "loss": 0.653,
+      "step": 36199
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4503060150519718,
+      "learning_rate": 1.1143806881676311e-06,
+      "loss": 0.7109,
+      "step": 36200
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2102572460471108,
+      "learning_rate": 1.1140306638493725e-06,
+      "loss": 0.68,
+      "step": 36201
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4113969749884097,
+      "learning_rate": 1.113680691268041e-06,
+      "loss": 0.7351,
+      "step": 36202
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4551502225122628,
+      "learning_rate": 1.1133307704256758e-06,
+      "loss": 0.8096,
+      "step": 36203
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5875951269100137,
+      "learning_rate": 1.112980901324312e-06,
+      "loss": 0.487,
+      "step": 36204
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.945520148133698,
+      "learning_rate": 1.112631083965985e-06,
+      "loss": 0.7104,
+      "step": 36205
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3323062668822219,
+      "learning_rate": 1.1122813183527337e-06,
+      "loss": 0.6704,
+      "step": 36206
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5848531199928468,
+      "learning_rate": 1.111931604486597e-06,
+      "loss": 0.6066,
+      "step": 36207
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.425896439988011,
+      "learning_rate": 1.1115819423696073e-06,
+      "loss": 0.5905,
+      "step": 36208
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3759798152747569,
+      "learning_rate": 1.1112323320037998e-06,
+      "loss": 0.6821,
+      "step": 36209
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4071385180392715,
+      "learning_rate": 1.110882773391212e-06,
+      "loss": 0.7877,
+      "step": 36210
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2665212189912838,
+      "learning_rate": 1.1105332665338808e-06,
+      "loss": 0.5111,
+      "step": 36211
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2428782979292563,
+      "learning_rate": 1.1101838114338393e-06,
+      "loss": 0.7895,
+      "step": 36212
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5901900847898964,
+      "learning_rate": 1.10983440809312e-06,
+      "loss": 0.4662,
+      "step": 36213
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4808656166878449,
+      "learning_rate": 1.1094850565137616e-06,
+      "loss": 0.7179,
+      "step": 36214
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3162184430831554,
+      "learning_rate": 1.1091357566977934e-06,
+      "loss": 0.5863,
+      "step": 36215
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.645367529659537,
+      "learning_rate": 1.108786508647255e-06,
+      "loss": 0.6987,
+      "step": 36216
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.479430581925296,
+      "learning_rate": 1.1084373123641734e-06,
+      "loss": 0.7384,
+      "step": 36217
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.176401681513944,
+      "learning_rate": 1.1080881678505862e-06,
+      "loss": 0.7167,
+      "step": 36218
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4000110247445134,
+      "learning_rate": 1.1077390751085237e-06,
+      "loss": 0.6966,
+      "step": 36219
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4890378357087932,
+      "learning_rate": 1.107390034140019e-06,
+      "loss": 0.7355,
+      "step": 36220
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4444781324227776,
+      "learning_rate": 1.1070410449471103e-06,
+      "loss": 0.5473,
+      "step": 36221
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7489748634745794,
+      "learning_rate": 1.1066921075318183e-06,
+      "loss": 0.632,
+      "step": 36222
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5083964812955912,
+      "learning_rate": 1.106343221896181e-06,
+      "loss": 0.7832,
+      "step": 36223
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3456191631288432,
+      "learning_rate": 1.1059943880422297e-06,
+      "loss": 0.7196,
+      "step": 36224
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5723928281150606,
+      "learning_rate": 1.1056456059719989e-06,
+      "loss": 0.5176,
+      "step": 36225
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5377326842936532,
+      "learning_rate": 1.1052968756875104e-06,
+      "loss": 0.7543,
+      "step": 36226
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6664901469071576,
+      "learning_rate": 1.1049481971907993e-06,
+      "loss": 0.7499,
+      "step": 36227
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5159769771528806,
+      "learning_rate": 1.1045995704838997e-06,
+      "loss": 0.6978,
+      "step": 36228
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.29943320883386,
+      "learning_rate": 1.104250995568834e-06,
+      "loss": 0.6416,
+      "step": 36229
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.615020739270908,
+      "learning_rate": 1.1039024724476376e-06,
+      "loss": 0.5192,
+      "step": 36230
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.385567630467797,
+      "learning_rate": 1.1035540011223355e-06,
+      "loss": 0.6055,
+      "step": 36231
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.286381343906277,
+      "learning_rate": 1.1032055815949616e-06,
+      "loss": 0.7091,
+      "step": 36232
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2363102121384573,
+      "learning_rate": 1.1028572138675387e-06,
+      "loss": 0.7782,
+      "step": 36233
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6333832859834938,
+      "learning_rate": 1.1025088979420995e-06,
+      "loss": 0.52,
+      "step": 36234
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.343271095328063,
+      "learning_rate": 1.1021606338206714e-06,
+      "loss": 0.7275,
+      "step": 36235
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1962802683983864,
+      "learning_rate": 1.1018124215052783e-06,
+      "loss": 0.6745,
+      "step": 36236
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2767792306061736,
+      "learning_rate": 1.1014642609979498e-06,
+      "loss": 0.5995,
+      "step": 36237
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5958279478884585,
+      "learning_rate": 1.1011161523007163e-06,
+      "loss": 0.6328,
+      "step": 36238
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1325145572307243,
+      "learning_rate": 1.100768095415603e-06,
+      "loss": 0.6848,
+      "step": 36239
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4013883556078097,
+      "learning_rate": 1.1004200903446316e-06,
+      "loss": 0.7772,
+      "step": 36240
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3477429536989156,
+      "learning_rate": 1.1000721370898326e-06,
+      "loss": 0.6206,
+      "step": 36241
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2783756025808577,
+      "learning_rate": 1.0997242356532335e-06,
+      "loss": 0.733,
+      "step": 36242
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4204852085936455,
+      "learning_rate": 1.099376386036858e-06,
+      "loss": 0.6855,
+      "step": 36243
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3295254032194896,
+      "learning_rate": 1.0990285882427286e-06,
+      "loss": 0.6554,
+      "step": 36244
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3865045729432366,
+      "learning_rate": 1.0986808422728722e-06,
+      "loss": 0.6999,
+      "step": 36245
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4560035993667633,
+      "learning_rate": 1.0983331481293168e-06,
+      "loss": 0.7049,
+      "step": 36246
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3548989193235124,
+      "learning_rate": 1.0979855058140842e-06,
+      "loss": 0.6621,
+      "step": 36247
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.561037011161477,
+      "learning_rate": 1.097637915329196e-06,
+      "loss": 0.6587,
+      "step": 36248
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6232108911728294,
+      "learning_rate": 1.0972903766766797e-06,
+      "loss": 0.6666,
+      "step": 36249
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3867177298682283,
+      "learning_rate": 1.0969428898585565e-06,
+      "loss": 0.7172,
+      "step": 36250
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3103418695761841,
+      "learning_rate": 1.0965954548768509e-06,
+      "loss": 0.7247,
+      "step": 36251
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3771835848434866,
+      "learning_rate": 1.0962480717335843e-06,
+      "loss": 0.5465,
+      "step": 36252
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2894342549193207,
+      "learning_rate": 1.095900740430783e-06,
+      "loss": 0.6979,
+      "step": 36253
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5909395208776544,
+      "learning_rate": 1.0955534609704644e-06,
+      "loss": 0.5472,
+      "step": 36254
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.087717934110769,
+      "learning_rate": 1.095206233354653e-06,
+      "loss": 0.6762,
+      "step": 36255
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6070875777611944,
+      "learning_rate": 1.0948590575853723e-06,
+      "loss": 0.7578,
+      "step": 36256
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8543431857057302,
+      "learning_rate": 1.0945119336646382e-06,
+      "loss": 0.5764,
+      "step": 36257
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4213103395701774,
+      "learning_rate": 1.0941648615944756e-06,
+      "loss": 0.6406,
+      "step": 36258
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.566565482626902,
+      "learning_rate": 1.0938178413769062e-06,
+      "loss": 0.6744,
+      "step": 36259
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5945108663913372,
+      "learning_rate": 1.0934708730139486e-06,
+      "loss": 0.5262,
+      "step": 36260
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.437969117085485,
+      "learning_rate": 1.0931239565076224e-06,
+      "loss": 0.6837,
+      "step": 36261
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.10655173652996,
+      "learning_rate": 1.0927770918599478e-06,
+      "loss": 0.7902,
+      "step": 36262
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3270740373377334,
+      "learning_rate": 1.0924302790729468e-06,
+      "loss": 0.6542,
+      "step": 36263
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.357185310148168,
+      "learning_rate": 1.0920835181486378e-06,
+      "loss": 0.802,
+      "step": 36264
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3449961404676558,
+      "learning_rate": 1.0917368090890357e-06,
+      "loss": 0.7208,
+      "step": 36265
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5962483626505453,
+      "learning_rate": 1.0913901518961623e-06,
+      "loss": 0.4971,
+      "step": 36266
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2416014790366172,
+      "learning_rate": 1.0910435465720382e-06,
+      "loss": 0.5581,
+      "step": 36267
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3687454825243681,
+      "learning_rate": 1.0906969931186795e-06,
+      "loss": 0.5947,
+      "step": 36268
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8863800037941845,
+      "learning_rate": 1.0903504915381002e-06,
+      "loss": 0.6189,
+      "step": 36269
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5373345454707594,
+      "learning_rate": 1.0900040418323243e-06,
+      "loss": 0.7309,
+      "step": 36270
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7303743230202746,
+      "learning_rate": 1.0896576440033636e-06,
+      "loss": 0.7189,
+      "step": 36271
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.378400454501123,
+      "learning_rate": 1.0893112980532384e-06,
+      "loss": 0.78,
+      "step": 36272
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8585801379028835,
+      "learning_rate": 1.0889650039839627e-06,
+      "loss": 0.6632,
+      "step": 36273
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4724699028849126,
+      "learning_rate": 1.0886187617975552e-06,
+      "loss": 0.6887,
+      "step": 36274
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2498259521922195,
+      "learning_rate": 1.0882725714960296e-06,
+      "loss": 0.6491,
+      "step": 36275
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8286663022586194,
+      "learning_rate": 1.0879264330814033e-06,
+      "loss": 0.6692,
+      "step": 36276
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8178103456278265,
+      "learning_rate": 1.0875803465556911e-06,
+      "loss": 0.7372,
+      "step": 36277
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.362726063477986,
+      "learning_rate": 1.087234311920906e-06,
+      "loss": 0.5403,
+      "step": 36278
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4165578679102901,
+      "learning_rate": 1.0868883291790645e-06,
+      "loss": 0.6009,
+      "step": 36279
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5010952054227589,
+      "learning_rate": 1.0865423983321832e-06,
+      "loss": 0.6991,
+      "step": 36280
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.909783674594031,
+      "learning_rate": 1.0861965193822743e-06,
+      "loss": 0.6898,
+      "step": 36281
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4654505379799123,
+      "learning_rate": 1.0858506923313484e-06,
+      "loss": 0.5917,
+      "step": 36282
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3350892777900414,
+      "learning_rate": 1.0855049171814226e-06,
+      "loss": 0.6838,
+      "step": 36283
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4704128315258638,
+      "learning_rate": 1.0851591939345108e-06,
+      "loss": 0.6156,
+      "step": 36284
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7442772602925765,
+      "learning_rate": 1.0848135225926248e-06,
+      "loss": 0.6539,
+      "step": 36285
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3163428340275158,
+      "learning_rate": 1.0844679031577754e-06,
+      "loss": 0.5363,
+      "step": 36286
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4583203087101166,
+      "learning_rate": 1.0841223356319763e-06,
+      "loss": 0.6219,
+      "step": 36287
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.362914761293616,
+      "learning_rate": 1.0837768200172417e-06,
+      "loss": 0.6452,
+      "step": 36288
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.459974686572232,
+      "learning_rate": 1.0834313563155807e-06,
+      "loss": 0.7066,
+      "step": 36289
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7047036682083812,
+      "learning_rate": 1.0830859445290044e-06,
+      "loss": 0.747,
+      "step": 36290
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6227008454966778,
+      "learning_rate": 1.0827405846595251e-06,
+      "loss": 0.6613,
+      "step": 36291
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.636127400179402,
+      "learning_rate": 1.0823952767091527e-06,
+      "loss": 0.7525,
+      "step": 36292
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5870029853606208,
+      "learning_rate": 1.0820500206799002e-06,
+      "loss": 0.4787,
+      "step": 36293
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2826722909093595,
+      "learning_rate": 1.0817048165737732e-06,
+      "loss": 0.6916,
+      "step": 36294
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6594764058878406,
+      "learning_rate": 1.0813596643927871e-06,
+      "loss": 0.5601,
+      "step": 36295
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4171656042428507,
+      "learning_rate": 1.0810145641389458e-06,
+      "loss": 0.5778,
+      "step": 36296
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3091064903910894,
+      "learning_rate": 1.0806695158142633e-06,
+      "loss": 0.6807,
+      "step": 36297
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3062426155415943,
+      "learning_rate": 1.0803245194207468e-06,
+      "loss": 0.7696,
+      "step": 36298
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.403443216022663,
+      "learning_rate": 1.0799795749604037e-06,
+      "loss": 0.6677,
+      "step": 36299
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3867771890083402,
+      "learning_rate": 1.0796346824352433e-06,
+      "loss": 0.6138,
+      "step": 36300
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3478109067782011,
+      "learning_rate": 1.0792898418472753e-06,
+      "loss": 0.6826,
+      "step": 36301
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2804010465375588,
+      "learning_rate": 1.0789450531985068e-06,
+      "loss": 0.7001,
+      "step": 36302
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4885512473587141,
+      "learning_rate": 1.078600316490942e-06,
+      "loss": 0.6182,
+      "step": 36303
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7494706878154571,
+      "learning_rate": 1.0782556317265902e-06,
+      "loss": 0.7267,
+      "step": 36304
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4678649923689298,
+      "learning_rate": 1.0779109989074621e-06,
+      "loss": 0.736,
+      "step": 36305
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6841453763336067,
+      "learning_rate": 1.0775664180355593e-06,
+      "loss": 0.6331,
+      "step": 36306
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5654112204368515,
+      "learning_rate": 1.0772218891128882e-06,
+      "loss": 0.77,
+      "step": 36307
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3372480255049073,
+      "learning_rate": 1.076877412141457e-06,
+      "loss": 0.6669,
+      "step": 36308
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.322892745156284,
+      "learning_rate": 1.076532987123271e-06,
+      "loss": 0.6473,
+      "step": 36309
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2873872366270003,
+      "learning_rate": 1.0761886140603362e-06,
+      "loss": 0.7158,
+      "step": 36310
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2886400597260974,
+      "learning_rate": 1.0758442929546543e-06,
+      "loss": 0.7198,
+      "step": 36311
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2371435125114993,
+      "learning_rate": 1.0755000238082326e-06,
+      "loss": 0.7294,
+      "step": 36312
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3203981545518226,
+      "learning_rate": 1.075155806623076e-06,
+      "loss": 0.8194,
+      "step": 36313
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5865971034114504,
+      "learning_rate": 1.074811641401189e-06,
+      "loss": 0.4735,
+      "step": 36314
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2808653575127684,
+      "learning_rate": 1.0744675281445726e-06,
+      "loss": 0.6874,
+      "step": 36315
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6226268373980237,
+      "learning_rate": 1.0741234668552324e-06,
+      "loss": 0.4937,
+      "step": 36316
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3791147967914934,
+      "learning_rate": 1.0737794575351712e-06,
+      "loss": 0.8191,
+      "step": 36317
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3510960571811272,
+      "learning_rate": 1.0734355001863904e-06,
+      "loss": 0.7092,
+      "step": 36318
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4928398007211108,
+      "learning_rate": 1.0730915948108966e-06,
+      "loss": 0.6985,
+      "step": 36319
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2796249439571588,
+      "learning_rate": 1.0727477414106903e-06,
+      "loss": 0.7792,
+      "step": 36320
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6569712297939247,
+      "learning_rate": 1.0724039399877705e-06,
+      "loss": 0.6866,
+      "step": 36321
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.515018660726222,
+      "learning_rate": 1.0720601905441419e-06,
+      "loss": 0.6748,
+      "step": 36322
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1827397623202796,
+      "learning_rate": 1.071716493081809e-06,
+      "loss": 0.5539,
+      "step": 36323
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7499208931027859,
+      "learning_rate": 1.0713728476027653e-06,
+      "loss": 0.6688,
+      "step": 36324
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2702689648664045,
+      "learning_rate": 1.0710292541090161e-06,
+      "loss": 0.8478,
+      "step": 36325
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5314960550797652,
+      "learning_rate": 1.0706857126025604e-06,
+      "loss": 0.6621,
+      "step": 36326
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.392356140150378,
+      "learning_rate": 1.0703422230854022e-06,
+      "loss": 0.5305,
+      "step": 36327
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5199869100801324,
+      "learning_rate": 1.0699987855595384e-06,
+      "loss": 0.6129,
+      "step": 36328
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4684268253340689,
+      "learning_rate": 1.0696554000269677e-06,
+      "loss": 0.5863,
+      "step": 36329
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.256308611091381,
+      "learning_rate": 1.0693120664896916e-06,
+      "loss": 0.6447,
+      "step": 36330
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.23065321427535,
+      "learning_rate": 1.0689687849497066e-06,
+      "loss": 0.6501,
+      "step": 36331
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.715335520646235,
+      "learning_rate": 1.068625555409014e-06,
+      "loss": 0.6923,
+      "step": 36332
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5672976202949117,
+      "learning_rate": 1.0682823778696105e-06,
+      "loss": 0.4794,
+      "step": 36333
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6634033857141884,
+      "learning_rate": 1.0679392523334954e-06,
+      "loss": 0.5264,
+      "step": 36334
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6449656098178596,
+      "learning_rate": 1.0675961788026646e-06,
+      "loss": 0.8742,
+      "step": 36335
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.331246858112689,
+      "learning_rate": 1.0672531572791178e-06,
+      "loss": 0.6829,
+      "step": 36336
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7325100340659363,
+      "learning_rate": 1.0669101877648524e-06,
+      "loss": 0.7495,
+      "step": 36337
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.284045807106611,
+      "learning_rate": 1.0665672702618612e-06,
+      "loss": 0.7432,
+      "step": 36338
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2051951704729011,
+      "learning_rate": 1.0662244047721437e-06,
+      "loss": 0.5133,
+      "step": 36339
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6268956502060796,
+      "learning_rate": 1.0658815912976983e-06,
+      "loss": 0.6155,
+      "step": 36340
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.349542861790805,
+      "learning_rate": 1.065538829840519e-06,
+      "loss": 0.6502,
+      "step": 36341
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.420379946321607,
+      "learning_rate": 1.0651961204025995e-06,
+      "loss": 0.6934,
+      "step": 36342
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6874472010399035,
+      "learning_rate": 1.0648534629859363e-06,
+      "loss": 0.7231,
+      "step": 36343
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.359490390529802,
+      "learning_rate": 1.0645108575925267e-06,
+      "loss": 0.6453,
+      "step": 36344
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3969605033704897,
+      "learning_rate": 1.0641683042243645e-06,
+      "loss": 0.5794,
+      "step": 36345
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4954618377188449,
+      "learning_rate": 1.0638258028834415e-06,
+      "loss": 0.687,
+      "step": 36346
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.358012790980188,
+      "learning_rate": 1.0634833535717537e-06,
+      "loss": 0.6769,
+      "step": 36347
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4383812794076418,
+      "learning_rate": 1.0631409562912975e-06,
+      "loss": 0.7518,
+      "step": 36348
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3320771457034657,
+      "learning_rate": 1.0627986110440637e-06,
+      "loss": 0.773,
+      "step": 36349
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3978587032518723,
+      "learning_rate": 1.0624563178320435e-06,
+      "loss": 0.681,
+      "step": 36350
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.471671016837804,
+      "learning_rate": 1.0621140766572336e-06,
+      "loss": 0.7428,
+      "step": 36351
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1880375782048427,
+      "learning_rate": 1.0617718875216242e-06,
+      "loss": 0.8077,
+      "step": 36352
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1235230725491268,
+      "learning_rate": 1.0614297504272108e-06,
+      "loss": 0.5415,
+      "step": 36353
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2243885357779758,
+      "learning_rate": 1.0610876653759805e-06,
+      "loss": 0.7666,
+      "step": 36354
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4680378352477632,
+      "learning_rate": 1.0607456323699306e-06,
+      "loss": 0.714,
+      "step": 36355
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3273512991782415,
+      "learning_rate": 1.0604036514110471e-06,
+      "loss": 0.6787,
+      "step": 36356
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4655833621752652,
+      "learning_rate": 1.0600617225013265e-06,
+      "loss": 0.6963,
+      "step": 36357
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.196341376352874,
+      "learning_rate": 1.059719845642756e-06,
+      "loss": 0.6466,
+      "step": 36358
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5624910880184828,
+      "learning_rate": 1.059378020837326e-06,
+      "loss": 0.7387,
+      "step": 36359
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4554060521450207,
+      "learning_rate": 1.0590362480870275e-06,
+      "loss": 0.671,
+      "step": 36360
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2818621482374268,
+      "learning_rate": 1.058694527393852e-06,
+      "loss": 0.6329,
+      "step": 36361
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5251514593367348,
+      "learning_rate": 1.058352858759788e-06,
+      "loss": 0.652,
+      "step": 36362
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4284429069787612,
+      "learning_rate": 1.0580112421868226e-06,
+      "loss": 0.6844,
+      "step": 36363
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3611524242462967,
+      "learning_rate": 1.0576696776769469e-06,
+      "loss": 0.6619,
+      "step": 36364
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3605503094895677,
+      "learning_rate": 1.057328165232151e-06,
+      "loss": 0.634,
+      "step": 36365
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.702334052486785,
+      "learning_rate": 1.0569867048544225e-06,
+      "loss": 0.6951,
+      "step": 36366
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4465367209918871,
+      "learning_rate": 1.0566452965457452e-06,
+      "loss": 0.6768,
+      "step": 36367
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6268057670548728,
+      "learning_rate": 1.056303940308112e-06,
+      "loss": 0.4879,
+      "step": 36368
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2725938295915098,
+      "learning_rate": 1.0559626361435104e-06,
+      "loss": 0.6147,
+      "step": 36369
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.506580559807311,
+      "learning_rate": 1.0556213840539253e-06,
+      "loss": 0.5762,
+      "step": 36370
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2613940610639331,
+      "learning_rate": 1.0552801840413428e-06,
+      "loss": 0.667,
+      "step": 36371
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.32998069960933,
+      "learning_rate": 1.0549390361077527e-06,
+      "loss": 0.6666,
+      "step": 36372
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4460017902320093,
+      "learning_rate": 1.0545979402551375e-06,
+      "loss": 0.7454,
+      "step": 36373
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3514593050669435,
+      "learning_rate": 1.054256896485487e-06,
+      "loss": 0.6899,
+      "step": 36374
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.388652897843349,
+      "learning_rate": 1.0539159048007842e-06,
+      "loss": 0.6756,
+      "step": 36375
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9380889147350033,
+      "learning_rate": 1.0535749652030159e-06,
+      "loss": 0.6986,
+      "step": 36376
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3066492466185966,
+      "learning_rate": 1.0532340776941653e-06,
+      "loss": 0.6986,
+      "step": 36377
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.63089695531339,
+      "learning_rate": 1.0528932422762195e-06,
+      "loss": 0.6034,
+      "step": 36378
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2716957399369704,
+      "learning_rate": 1.0525524589511627e-06,
+      "loss": 0.7918,
+      "step": 36379
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.309200152733545,
+      "learning_rate": 1.0522117277209764e-06,
+      "loss": 0.5813,
+      "step": 36380
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3933288840757774,
+      "learning_rate": 1.051871048587646e-06,
+      "loss": 0.6036,
+      "step": 36381
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3010763289048235,
+      "learning_rate": 1.0515304215531563e-06,
+      "loss": 0.6583,
+      "step": 36382
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3395625518498397,
+      "learning_rate": 1.0511898466194903e-06,
+      "loss": 0.6613,
+      "step": 36383
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2548482654516477,
+      "learning_rate": 1.0508493237886286e-06,
+      "loss": 0.6849,
+      "step": 36384
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.415311122273927,
+      "learning_rate": 1.050508853062555e-06,
+      "loss": 0.7537,
+      "step": 36385
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3104924330541734,
+      "learning_rate": 1.0501684344432538e-06,
+      "loss": 0.586,
+      "step": 36386
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6112609863333625,
+      "learning_rate": 1.0498280679327067e-06,
+      "loss": 0.5072,
+      "step": 36387
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4217788664016042,
+      "learning_rate": 1.0494877535328907e-06,
+      "loss": 0.6879,
+      "step": 36388
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.541141173415461,
+      "learning_rate": 1.049147491245791e-06,
+      "loss": 0.8151,
+      "step": 36389
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3325525188592673,
+      "learning_rate": 1.0488072810733906e-06,
+      "loss": 0.7037,
+      "step": 36390
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5858019214847512,
+      "learning_rate": 1.048467123017669e-06,
+      "loss": 0.4715,
+      "step": 36391
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4717873718099352,
+      "learning_rate": 1.0481270170806024e-06,
+      "loss": 0.6712,
+      "step": 36392
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5766660884792536,
+      "learning_rate": 1.0477869632641769e-06,
+      "loss": 0.6365,
+      "step": 36393
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4730397182160881,
+      "learning_rate": 1.0474469615703686e-06,
+      "loss": 0.7077,
+      "step": 36394
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5849141308384558,
+      "learning_rate": 1.0471070120011595e-06,
+      "loss": 0.617,
+      "step": 36395
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6150304651349425,
+      "learning_rate": 1.0467671145585268e-06,
+      "loss": 0.7136,
+      "step": 36396
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6721901763628265,
+      "learning_rate": 1.0464272692444511e-06,
+      "loss": 0.5436,
+      "step": 36397
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2821949064705265,
+      "learning_rate": 1.0460874760609096e-06,
+      "loss": 0.6773,
+      "step": 36398
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6102383237219896,
+      "learning_rate": 1.0457477350098832e-06,
+      "loss": 0.4682,
+      "step": 36399
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4047565567511682,
+      "learning_rate": 1.045408046093348e-06,
+      "loss": 0.7028,
+      "step": 36400
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4045665744816216,
+      "learning_rate": 1.0450684093132813e-06,
+      "loss": 0.7043,
+      "step": 36401
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5182233482771053,
+      "learning_rate": 1.0447288246716602e-06,
+      "loss": 0.6815,
+      "step": 36402
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5770012175710845,
+      "learning_rate": 1.0443892921704635e-06,
+      "loss": 0.5356,
+      "step": 36403
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3353223335025168,
+      "learning_rate": 1.0440498118116715e-06,
+      "loss": 0.5933,
+      "step": 36404
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.276485200620425,
+      "learning_rate": 1.0437103835972528e-06,
+      "loss": 0.5487,
+      "step": 36405
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2973055599675536,
+      "learning_rate": 1.043371007529188e-06,
+      "loss": 0.6508,
+      "step": 36406
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2826258002823485,
+      "learning_rate": 1.0430316836094555e-06,
+      "loss": 0.7995,
+      "step": 36407
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.375033832109358,
+      "learning_rate": 1.0426924118400272e-06,
+      "loss": 0.7336,
+      "step": 36408
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3255693001988424,
+      "learning_rate": 1.042353192222878e-06,
+      "loss": 0.6373,
+      "step": 36409
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.0528815838498615,
+      "learning_rate": 1.0420140247599842e-06,
+      "loss": 0.6461,
+      "step": 36410
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4567559325357657,
+      "learning_rate": 1.041674909453323e-06,
+      "loss": 0.669,
+      "step": 36411
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4912009371088522,
+      "learning_rate": 1.0413358463048674e-06,
+      "loss": 0.7075,
+      "step": 36412
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.64451419362125,
+      "learning_rate": 1.0409968353165878e-06,
+      "loss": 0.6799,
+      "step": 36413
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7226783111630426,
+      "learning_rate": 1.040657876490463e-06,
+      "loss": 0.7832,
+      "step": 36414
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.003544081061722,
+      "learning_rate": 1.040318969828462e-06,
+      "loss": 0.5691,
+      "step": 36415
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.534919472991187,
+      "learning_rate": 1.0399801153325616e-06,
+      "loss": 0.7266,
+      "step": 36416
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2117800483953685,
+      "learning_rate": 1.0396413130047356e-06,
+      "loss": 0.6565,
+      "step": 36417
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2605517961886914,
+      "learning_rate": 1.0393025628469545e-06,
+      "loss": 0.6465,
+      "step": 36418
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4271096028541623,
+      "learning_rate": 1.038963864861189e-06,
+      "loss": 0.7061,
+      "step": 36419
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4533015271146845,
+      "learning_rate": 1.0386252190494117e-06,
+      "loss": 0.7943,
+      "step": 36420
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6083414301388068,
+      "learning_rate": 1.0382866254136004e-06,
+      "loss": 0.557,
+      "step": 36421
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5813088314957073,
+      "learning_rate": 1.0379480839557177e-06,
+      "loss": 0.7075,
+      "step": 36422
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.619993729648624,
+      "learning_rate": 1.0376095946777387e-06,
+      "loss": 0.7318,
+      "step": 36423
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4077342040862262,
+      "learning_rate": 1.0372711575816341e-06,
+      "loss": 0.7053,
+      "step": 36424
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5944859158963524,
+      "learning_rate": 1.0369327726693757e-06,
+      "loss": 0.4908,
+      "step": 36425
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2634831388967307,
+      "learning_rate": 1.0365944399429327e-06,
+      "loss": 0.6059,
+      "step": 36426
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4524053895102813,
+      "learning_rate": 1.036256159404273e-06,
+      "loss": 0.5696,
+      "step": 36427
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.546497857835943,
+      "learning_rate": 1.035917931055369e-06,
+      "loss": 0.6164,
+      "step": 36428
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2725206211232853,
+      "learning_rate": 1.035579754898187e-06,
+      "loss": 0.6447,
+      "step": 36429
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3812125257843038,
+      "learning_rate": 1.0352416309347003e-06,
+      "loss": 0.554,
+      "step": 36430
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.457658038924386,
+      "learning_rate": 1.0349035591668732e-06,
+      "loss": 0.6813,
+      "step": 36431
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.663036599303436,
+      "learning_rate": 1.034565539596677e-06,
+      "loss": 0.605,
+      "step": 36432
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.408797116771725,
+      "learning_rate": 1.0342275722260776e-06,
+      "loss": 0.5677,
+      "step": 36433
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.430808050667401,
+      "learning_rate": 1.0338896570570456e-06,
+      "loss": 0.6783,
+      "step": 36434
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4040157224291154,
+      "learning_rate": 1.0335517940915441e-06,
+      "loss": 0.6296,
+      "step": 36435
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5660442047381902,
+      "learning_rate": 1.0332139833315447e-06,
+      "loss": 0.7892,
+      "step": 36436
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3739548738098049,
+      "learning_rate": 1.0328762247790103e-06,
+      "loss": 0.6773,
+      "step": 36437
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.982753832860237,
+      "learning_rate": 1.0325385184359126e-06,
+      "loss": 0.6616,
+      "step": 36438
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7162747914535115,
+      "learning_rate": 1.0322008643042136e-06,
+      "loss": 0.8863,
+      "step": 36439
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.338511204702208,
+      "learning_rate": 1.031863262385878e-06,
+      "loss": 0.6695,
+      "step": 36440
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7188599581793225,
+      "learning_rate": 1.0315257126828749e-06,
+      "loss": 0.8154,
+      "step": 36441
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3164500897574847,
+      "learning_rate": 1.0311882151971697e-06,
+      "loss": 0.5847,
+      "step": 36442
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2798546512640157,
+      "learning_rate": 1.0308507699307268e-06,
+      "loss": 0.6081,
+      "step": 36443
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3862406770103393,
+      "learning_rate": 1.030513376885508e-06,
+      "loss": 0.5575,
+      "step": 36444
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5891251676891573,
+      "learning_rate": 1.0301760360634794e-06,
+      "loss": 0.4525,
+      "step": 36445
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2402179385700272,
+      "learning_rate": 1.0298387474666083e-06,
+      "loss": 0.7007,
+      "step": 36446
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.408023521031712,
+      "learning_rate": 1.0295015110968564e-06,
+      "loss": 0.6042,
+      "step": 36447
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.233541514032827,
+      "learning_rate": 1.0291643269561835e-06,
+      "loss": 0.6848,
+      "step": 36448
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6429772264601656,
+      "learning_rate": 1.0288271950465566e-06,
+      "loss": 0.5115,
+      "step": 36449
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3226969044883377,
+      "learning_rate": 1.0284901153699412e-06,
+      "loss": 0.6595,
+      "step": 36450
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4899622615539874,
+      "learning_rate": 1.0281530879282953e-06,
+      "loss": 0.7072,
+      "step": 36451
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8109373082433233,
+      "learning_rate": 1.0278161127235808e-06,
+      "loss": 0.6691,
+      "step": 36452
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.48976910047683,
+      "learning_rate": 1.027479189757763e-06,
+      "loss": 0.74,
+      "step": 36453
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3780937179882593,
+      "learning_rate": 1.027142319032801e-06,
+      "loss": 0.6543,
+      "step": 36454
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2886825625596996,
+      "learning_rate": 1.0268055005506583e-06,
+      "loss": 0.7187,
+      "step": 36455
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5927146016780214,
+      "learning_rate": 1.0264687343132928e-06,
+      "loss": 0.5188,
+      "step": 36456
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2925465132648717,
+      "learning_rate": 1.02613202032267e-06,
+      "loss": 0.6027,
+      "step": 36457
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3324370193004353,
+      "learning_rate": 1.0257953585807456e-06,
+      "loss": 0.6557,
+      "step": 36458
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.443932904360013,
+      "learning_rate": 1.025458749089484e-06,
+      "loss": 0.5071,
+      "step": 36459
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5915903217320259,
+      "learning_rate": 1.0251221918508425e-06,
+      "loss": 0.5053,
+      "step": 36460
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5982330992947353,
+      "learning_rate": 1.0247856868667782e-06,
+      "loss": 0.6861,
+      "step": 36461
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.4488782894977876,
+      "learning_rate": 1.0244492341392543e-06,
+      "loss": 0.7374,
+      "step": 36462
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5432218167978564,
+      "learning_rate": 1.02411283367023e-06,
+      "loss": 0.5487,
+      "step": 36463
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.411490607172647,
+      "learning_rate": 1.0237764854616628e-06,
+      "loss": 0.6648,
+      "step": 36464
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.437876190559297,
+      "learning_rate": 1.0234401895155076e-06,
+      "loss": 0.6856,
+      "step": 36465
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.330532587807148,
+      "learning_rate": 1.0231039458337267e-06,
+      "loss": 0.6617,
+      "step": 36466
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5814596930218903,
+      "learning_rate": 1.022767754418279e-06,
+      "loss": 0.4614,
+      "step": 36467
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3940293865911753,
+      "learning_rate": 1.022431615271119e-06,
+      "loss": 0.769,
+      "step": 36468
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2753126080065402,
+      "learning_rate": 1.0220955283942024e-06,
+      "loss": 0.6789,
+      "step": 36469
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2923909577134203,
+      "learning_rate": 1.021759493789488e-06,
+      "loss": 0.7144,
+      "step": 36470
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3820974242733715,
+      "learning_rate": 1.0214235114589344e-06,
+      "loss": 0.7148,
+      "step": 36471
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.45307246498653,
+      "learning_rate": 1.021087581404495e-06,
+      "loss": 0.802,
+      "step": 36472
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6729375103551278,
+      "learning_rate": 1.0207517036281256e-06,
+      "loss": 0.6331,
+      "step": 36473
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6190750884808508,
+      "learning_rate": 1.0204158781317841e-06,
+      "loss": 0.4905,
+      "step": 36474
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6800361058081141,
+      "learning_rate": 1.0200801049174214e-06,
+      "loss": 0.6829,
+      "step": 36475
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.971231003061916,
+      "learning_rate": 1.019744383986998e-06,
+      "loss": 0.6265,
+      "step": 36476
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6471827583439362,
+      "learning_rate": 1.0194087153424648e-06,
+      "loss": 0.5678,
+      "step": 36477
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3948594395099352,
+      "learning_rate": 1.0190730989857777e-06,
+      "loss": 0.6256,
+      "step": 36478
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4156950304680525,
+      "learning_rate": 1.0187375349188887e-06,
+      "loss": 0.6973,
+      "step": 36479
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.478328984118405,
+      "learning_rate": 1.0184020231437564e-06,
+      "loss": 0.6874,
+      "step": 36480
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5326058928028048,
+      "learning_rate": 1.01806656366233e-06,
+      "loss": 0.7682,
+      "step": 36481
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.361375227375646,
+      "learning_rate": 1.0177311564765612e-06,
+      "loss": 0.6708,
+      "step": 36482
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3868098233158357,
+      "learning_rate": 1.0173958015884066e-06,
+      "loss": 0.8399,
+      "step": 36483
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4555991797379038,
+      "learning_rate": 1.0170604989998189e-06,
+      "loss": 0.6829,
+      "step": 36484
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6091309751930332,
+      "learning_rate": 1.0167252487127488e-06,
+      "loss": 0.6079,
+      "step": 36485
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.31801347434467,
+      "learning_rate": 1.0163900507291457e-06,
+      "loss": 0.7166,
+      "step": 36486
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5565097187588126,
+      "learning_rate": 1.016054905050965e-06,
+      "loss": 0.6833,
+      "step": 36487
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9323130265958635,
+      "learning_rate": 1.0157198116801592e-06,
+      "loss": 0.6226,
+      "step": 36488
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.334436866942988,
+      "learning_rate": 1.015384770618676e-06,
+      "loss": 0.7873,
+      "step": 36489
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.861122671045174,
+      "learning_rate": 1.0150497818684657e-06,
+      "loss": 0.6358,
+      "step": 36490
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3395534412896335,
+      "learning_rate": 1.0147148454314792e-06,
+      "loss": 0.6665,
+      "step": 36491
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.321075352520441,
+      "learning_rate": 1.0143799613096706e-06,
+      "loss": 0.617,
+      "step": 36492
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4292504237818797,
+      "learning_rate": 1.014045129504987e-06,
+      "loss": 0.7355,
+      "step": 36493
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4727324751545219,
+      "learning_rate": 1.0137103500193746e-06,
+      "loss": 0.7817,
+      "step": 36494
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3829810814406498,
+      "learning_rate": 1.0133756228547875e-06,
+      "loss": 0.617,
+      "step": 36495
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4579662470296681,
+      "learning_rate": 1.013040948013172e-06,
+      "loss": 0.6566,
+      "step": 36496
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2343826967794274,
+      "learning_rate": 1.0127063254964776e-06,
+      "loss": 0.5977,
+      "step": 36497
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.360572594989811,
+      "learning_rate": 1.0123717553066515e-06,
+      "loss": 0.6984,
+      "step": 36498
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4766363834147764,
+      "learning_rate": 1.0120372374456445e-06,
+      "loss": 0.5777,
+      "step": 36499
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6825886108170685,
+      "learning_rate": 1.0117027719153993e-06,
+      "loss": 0.7198,
+      "step": 36500
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6333698022910255,
+      "learning_rate": 1.011368358717867e-06,
+      "loss": 0.6273,
+      "step": 36501
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7372540232655036,
+      "learning_rate": 1.0110339978549977e-06,
+      "loss": 0.7262,
+      "step": 36502
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9802097486636299,
+      "learning_rate": 1.0106996893287303e-06,
+      "loss": 0.6092,
+      "step": 36503
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3386233338045326,
+      "learning_rate": 1.010365433141015e-06,
+      "loss": 0.7355,
+      "step": 36504
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.567535937303319,
+      "learning_rate": 1.0100312292937986e-06,
+      "loss": 0.5018,
+      "step": 36505
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3656458091992587,
+      "learning_rate": 1.0096970777890302e-06,
+      "loss": 0.7162,
+      "step": 36506
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7430668641821838,
+      "learning_rate": 1.0093629786286475e-06,
+      "loss": 0.6199,
+      "step": 36507
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3798323184377774,
+      "learning_rate": 1.0090289318146007e-06,
+      "loss": 0.6981,
+      "step": 36508
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4211432206283154,
+      "learning_rate": 1.0086949373488352e-06,
+      "loss": 0.6583,
+      "step": 36509
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4480372713956737,
+      "learning_rate": 1.0083609952332928e-06,
+      "loss": 0.6184,
+      "step": 36510
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5368321668204936,
+      "learning_rate": 1.0080271054699208e-06,
+      "loss": 0.8007,
+      "step": 36511
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.50098674870339,
+      "learning_rate": 1.007693268060661e-06,
+      "loss": 0.7068,
+      "step": 36512
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.463929854378776,
+      "learning_rate": 1.0073594830074596e-06,
+      "loss": 0.7435,
+      "step": 36513
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6119164150646517,
+      "learning_rate": 1.007025750312256e-06,
+      "loss": 0.5769,
+      "step": 36514
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6405249813507217,
+      "learning_rate": 1.0066920699769966e-06,
+      "loss": 0.5416,
+      "step": 36515
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3615909144942338,
+      "learning_rate": 1.0063584420036243e-06,
+      "loss": 0.6284,
+      "step": 36516
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3364743851773722,
+      "learning_rate": 1.0060248663940787e-06,
+      "loss": 0.7833,
+      "step": 36517
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4627649010868344,
+      "learning_rate": 1.0056913431503035e-06,
+      "loss": 0.7495,
+      "step": 36518
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.374734565352827,
+      "learning_rate": 1.005357872274243e-06,
+      "loss": 0.7002,
+      "step": 36519
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4383136787101893,
+      "learning_rate": 1.0050244537678356e-06,
+      "loss": 0.8073,
+      "step": 36520
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5146350762778948,
+      "learning_rate": 1.004691087633023e-06,
+      "loss": 0.6357,
+      "step": 36521
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.353900847099567,
+      "learning_rate": 1.0043577738717459e-06,
+      "loss": 0.6214,
+      "step": 36522
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4200207452101201,
+      "learning_rate": 1.0040245124859483e-06,
+      "loss": 0.6604,
+      "step": 36523
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5129750123264982,
+      "learning_rate": 1.0036913034775675e-06,
+      "loss": 0.6006,
+      "step": 36524
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3653433263550867,
+      "learning_rate": 1.0033581468485422e-06,
+      "loss": 0.7436,
+      "step": 36525
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.318210478026159,
+      "learning_rate": 1.0030250426008148e-06,
+      "loss": 0.6587,
+      "step": 36526
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2674302923257472,
+      "learning_rate": 1.0026919907363252e-06,
+      "loss": 0.7461,
+      "step": 36527
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5688414442289562,
+      "learning_rate": 1.002358991257012e-06,
+      "loss": 0.485,
+      "step": 36528
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5850123930200495,
+      "learning_rate": 1.0020260441648111e-06,
+      "loss": 0.5115,
+      "step": 36529
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6064321974286987,
+      "learning_rate": 1.0016931494616644e-06,
+      "loss": 0.5272,
+      "step": 36530
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3273762264432043,
+      "learning_rate": 1.001360307149508e-06,
+      "loss": 0.6851,
+      "step": 36531
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5026107222411154,
+      "learning_rate": 1.001027517230283e-06,
+      "loss": 0.6006,
+      "step": 36532
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2946847793258665,
+      "learning_rate": 1.000694779705922e-06,
+      "loss": 0.5976,
+      "step": 36533
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3776458164633172,
+      "learning_rate": 1.0003620945783676e-06,
+      "loss": 0.644,
+      "step": 36534
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9766032042082802,
+      "learning_rate": 1.000029461849552e-06,
+      "loss": 0.6669,
+      "step": 36535
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2810997688952226,
+      "learning_rate": 9.99696881521417e-07,
+      "loss": 0.6549,
+      "step": 36536
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3269842295053131,
+      "learning_rate": 9.993643535958952e-07,
+      "loss": 0.6233,
+      "step": 36537
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5461113761103864,
+      "learning_rate": 9.990318780749219e-07,
+      "loss": 0.6252,
+      "step": 36538
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3656097757525216,
+      "learning_rate": 9.986994549604346e-07,
+      "loss": 0.6467,
+      "step": 36539
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5745779326876019,
+      "learning_rate": 9.983670842543713e-07,
+      "loss": 0.4675,
+      "step": 36540
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2780375221246214,
+      "learning_rate": 9.980347659586643e-07,
+      "loss": 0.7474,
+      "step": 36541
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.486688252874393,
+      "learning_rate": 9.977025000752472e-07,
+      "loss": 0.6499,
+      "step": 36542
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3083269747741526,
+      "learning_rate": 9.973702866060565e-07,
+      "loss": 0.7779,
+      "step": 36543
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.20134659422373,
+      "learning_rate": 9.97038125553028e-07,
+      "loss": 0.745,
+      "step": 36544
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8086061183962123,
+      "learning_rate": 9.96706016918093e-07,
+      "loss": 0.7068,
+      "step": 36545
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2678922368957137,
+      "learning_rate": 9.963739607031852e-07,
+      "loss": 0.7432,
+      "step": 36546
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2206676048713805,
+      "learning_rate": 9.960419569102375e-07,
+      "loss": 0.6097,
+      "step": 36547
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5488503675339538,
+      "learning_rate": 9.957100055411872e-07,
+      "loss": 0.7693,
+      "step": 36548
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3965638114562493,
+      "learning_rate": 9.95378106597964e-07,
+      "loss": 0.6883,
+      "step": 36549
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6376094557452446,
+      "learning_rate": 9.950462600824973e-07,
+      "loss": 0.5433,
+      "step": 36550
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4159932521787335,
+      "learning_rate": 9.947144659967245e-07,
+      "loss": 0.6048,
+      "step": 36551
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4121289291349486,
+      "learning_rate": 9.94382724342573e-07,
+      "loss": 0.6216,
+      "step": 36552
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1547727731863398,
+      "learning_rate": 9.94051035121979e-07,
+      "loss": 0.6357,
+      "step": 36553
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3837037239417673,
+      "learning_rate": 9.937193983368688e-07,
+      "loss": 0.6098,
+      "step": 36554
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.5006461899640153,
+      "learning_rate": 9.933878139891772e-07,
+      "loss": 0.7863,
+      "step": 36555
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5523602135315109,
+      "learning_rate": 9.93056282080832e-07,
+      "loss": 0.7589,
+      "step": 36556
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3905025134020264,
+      "learning_rate": 9.92724802613766e-07,
+      "loss": 0.6108,
+      "step": 36557
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.840800259799347,
+      "learning_rate": 9.923933755899062e-07,
+      "loss": 0.7416,
+      "step": 36558
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3725942461837262,
+      "learning_rate": 9.920620010111858e-07,
+      "loss": 0.5777,
+      "step": 36559
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3913338770357715,
+      "learning_rate": 9.9173067887953e-07,
+      "loss": 0.7136,
+      "step": 36560
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5710932732423977,
+      "learning_rate": 9.913994091968725e-07,
+      "loss": 0.4923,
+      "step": 36561
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.286740722107477,
+      "learning_rate": 9.910681919651399e-07,
+      "loss": 0.6988,
+      "step": 36562
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.596918158857558,
+      "learning_rate": 9.907370271862582e-07,
+      "loss": 0.7397,
+      "step": 36563
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5659227816736294,
+      "learning_rate": 9.904059148621581e-07,
+      "loss": 0.6707,
+      "step": 36564
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9454481446479623,
+      "learning_rate": 9.900748549947692e-07,
+      "loss": 0.6879,
+      "step": 36565
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.286096200131315,
+      "learning_rate": 9.897438475860165e-07,
+      "loss": 0.6413,
+      "step": 36566
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6008409965019879,
+      "learning_rate": 9.894128926378255e-07,
+      "loss": 0.4853,
+      "step": 36567
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4089690293321344,
+      "learning_rate": 9.890819901521266e-07,
+      "loss": 0.5823,
+      "step": 36568
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3195553153111548,
+      "learning_rate": 9.887511401308458e-07,
+      "loss": 0.6612,
+      "step": 36569
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4263513459608357,
+      "learning_rate": 9.884203425759098e-07,
+      "loss": 0.6653,
+      "step": 36570
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3547782171147367,
+      "learning_rate": 9.880895974892413e-07,
+      "loss": 0.6547,
+      "step": 36571
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.496852136409142,
+      "learning_rate": 9.877589048727676e-07,
+      "loss": 0.6876,
+      "step": 36572
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5868673743477528,
+      "learning_rate": 9.874282647284173e-07,
+      "loss": 0.7045,
+      "step": 36573
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.272557824786728,
+      "learning_rate": 9.87097677058113e-07,
+      "loss": 0.6959,
+      "step": 36574
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3470712986674314,
+      "learning_rate": 9.86767141863777e-07,
+      "loss": 0.7123,
+      "step": 36575
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4533764722281286,
+      "learning_rate": 9.864366591473385e-07,
+      "loss": 0.8048,
+      "step": 36576
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3268821983813432,
+      "learning_rate": 9.86106228910718e-07,
+      "loss": 0.6425,
+      "step": 36577
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.525583607352079,
+      "learning_rate": 9.857758511558423e-07,
+      "loss": 0.6857,
+      "step": 36578
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5241439548802813,
+      "learning_rate": 9.854455258846319e-07,
+      "loss": 0.6669,
+      "step": 36579
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.090752202825085,
+      "learning_rate": 9.851152530990127e-07,
+      "loss": 0.6986,
+      "step": 36580
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.8594365753219404,
+      "learning_rate": 9.847850328009056e-07,
+      "loss": 0.6929,
+      "step": 36581
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2204109546414834,
+      "learning_rate": 9.844548649922358e-07,
+      "loss": 0.6215,
+      "step": 36582
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3785550690731556,
+      "learning_rate": 9.84124749674923e-07,
+      "loss": 0.6925,
+      "step": 36583
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3682511887404836,
+      "learning_rate": 9.837946868508897e-07,
+      "loss": 0.5832,
+      "step": 36584
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.62616477983474,
+      "learning_rate": 9.83464676522058e-07,
+      "loss": 0.7447,
+      "step": 36585
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4771830136934154,
+      "learning_rate": 9.831347186903495e-07,
+      "loss": 0.7473,
+      "step": 36586
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5913365421799442,
+      "learning_rate": 9.828048133576884e-07,
+      "loss": 0.4891,
+      "step": 36587
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3280591997507913,
+      "learning_rate": 9.824749605259898e-07,
+      "loss": 0.7004,
+      "step": 36588
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5997130861336392,
+      "learning_rate": 9.821451601971755e-07,
+      "loss": 0.4908,
+      "step": 36589
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.522485953925902,
+      "learning_rate": 9.818154123731704e-07,
+      "loss": 0.7146,
+      "step": 36590
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5396312397609708,
+      "learning_rate": 9.8148571705589e-07,
+      "loss": 0.6841,
+      "step": 36591
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7067989050857673,
+      "learning_rate": 9.811560742472537e-07,
+      "loss": 0.6034,
+      "step": 36592
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5999688236887747,
+      "learning_rate": 9.808264839491822e-07,
+      "loss": 0.5186,
+      "step": 36593
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3688727875498037,
+      "learning_rate": 9.80496946163596e-07,
+      "loss": 0.7071,
+      "step": 36594
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4376030642055677,
+      "learning_rate": 9.801674608924118e-07,
+      "loss": 0.7095,
+      "step": 36595
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6175424532506089,
+      "learning_rate": 9.798380281375463e-07,
+      "loss": 0.5564,
+      "step": 36596
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6161694921753897,
+      "learning_rate": 9.795086479009218e-07,
+      "loss": 0.6782,
+      "step": 36597
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4486918320701228,
+      "learning_rate": 9.791793201844525e-07,
+      "loss": 0.5969,
+      "step": 36598
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2755387355168126,
+      "learning_rate": 9.788500449900563e-07,
+      "loss": 0.6113,
+      "step": 36599
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4565485085146397,
+      "learning_rate": 9.78520822319653e-07,
+      "loss": 0.6943,
+      "step": 36600
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.62190767674267,
+      "learning_rate": 9.781916521751577e-07,
+      "loss": 0.5899,
+      "step": 36601
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2692054277069804,
+      "learning_rate": 9.778625345584857e-07,
+      "loss": 0.5923,
+      "step": 36602
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6589633723441841,
+      "learning_rate": 9.775334694715533e-07,
+      "loss": 0.6626,
+      "step": 36603
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3981412040778838,
+      "learning_rate": 9.77204456916282e-07,
+      "loss": 0.8063,
+      "step": 36604
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6279411224496492,
+      "learning_rate": 9.768754968945792e-07,
+      "loss": 0.4384,
+      "step": 36605
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6671415858184195,
+      "learning_rate": 9.765465894083637e-07,
+      "loss": 0.5928,
+      "step": 36606
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3930088909782297,
+      "learning_rate": 9.762177344595514e-07,
+      "loss": 0.8278,
+      "step": 36607
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.7182354234668487,
+      "learning_rate": 9.758889320500575e-07,
+      "loss": 0.6602,
+      "step": 36608
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4331859758531929,
+      "learning_rate": 9.75560182181795e-07,
+      "loss": 0.6352,
+      "step": 36609
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4829134871783323,
+      "learning_rate": 9.75231484856677e-07,
+      "loss": 0.6663,
+      "step": 36610
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4853029689890838,
+      "learning_rate": 9.749028400766191e-07,
+      "loss": 0.6974,
+      "step": 36611
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4182706327548285,
+      "learning_rate": 9.745742478435339e-07,
+      "loss": 0.5511,
+      "step": 36612
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6825551468029083,
+      "learning_rate": 9.74245708159336e-07,
+      "loss": 0.6656,
+      "step": 36613
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7283790348609538,
+      "learning_rate": 9.73917221025935e-07,
+      "loss": 0.7164,
+      "step": 36614
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.230417049299603,
+      "learning_rate": 9.735887864452476e-07,
+      "loss": 0.66,
+      "step": 36615
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.368790129845457,
+      "learning_rate": 9.732604044191818e-07,
+      "loss": 0.715,
+      "step": 36616
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3156140558621214,
+      "learning_rate": 9.72932074949654e-07,
+      "loss": 0.6953,
+      "step": 36617
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.775475754306208,
+      "learning_rate": 9.72603798038574e-07,
+      "loss": 0.7506,
+      "step": 36618
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.291462805162487,
+      "learning_rate": 9.722755736878498e-07,
+      "loss": 0.6209,
+      "step": 36619
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.340589850327627,
+      "learning_rate": 9.719474018993957e-07,
+      "loss": 0.7345,
+      "step": 36620
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1942528690445886,
+      "learning_rate": 9.716192826751246e-07,
+      "loss": 0.6034,
+      "step": 36621
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.299247231213573,
+      "learning_rate": 9.71291216016944e-07,
+      "loss": 0.7529,
+      "step": 36622
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3366952068188729,
+      "learning_rate": 9.709632019267624e-07,
+      "loss": 0.6731,
+      "step": 36623
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3750733511472797,
+      "learning_rate": 9.706352404064911e-07,
+      "loss": 0.6868,
+      "step": 36624
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.653893358885528,
+      "learning_rate": 9.703073314580425e-07,
+      "loss": 0.6992,
+      "step": 36625
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3415401066892145,
+      "learning_rate": 9.699794750833235e-07,
+      "loss": 0.6958,
+      "step": 36626
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3795126013207923,
+      "learning_rate": 9.696516712842407e-07,
+      "loss": 0.6586,
+      "step": 36627
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.813894031263704,
+      "learning_rate": 9.693239200627047e-07,
+      "loss": 0.5449,
+      "step": 36628
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5201259177204587,
+      "learning_rate": 9.68996221420626e-07,
+      "loss": 0.5928,
+      "step": 36629
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4743091456828346,
+      "learning_rate": 9.686685753599113e-07,
+      "loss": 0.6722,
+      "step": 36630
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2355827997002053,
+      "learning_rate": 9.683409818824641e-07,
+      "loss": 0.6424,
+      "step": 36631
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5265807572800558,
+      "learning_rate": 9.680134409901975e-07,
+      "loss": 0.6934,
+      "step": 36632
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5979182598336645,
+      "learning_rate": 9.676859526850145e-07,
+      "loss": 0.6467,
+      "step": 36633
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6678887516775485,
+      "learning_rate": 9.67358516968826e-07,
+      "loss": 0.7136,
+      "step": 36634
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7790826113469855,
+      "learning_rate": 9.670311338435324e-07,
+      "loss": 0.5985,
+      "step": 36635
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3428606284996376,
+      "learning_rate": 9.667038033110465e-07,
+      "loss": 0.6621,
+      "step": 36636
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3611297756562113,
+      "learning_rate": 9.663765253732682e-07,
+      "loss": 0.7304,
+      "step": 36637
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2732561812464398,
+      "learning_rate": 9.660493000321081e-07,
+      "loss": 0.6165,
+      "step": 36638
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6478313914049467,
+      "learning_rate": 9.65722127289469e-07,
+      "loss": 0.6207,
+      "step": 36639
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2439721006527868,
+      "learning_rate": 9.65395007147254e-07,
+      "loss": 0.591,
+      "step": 36640
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5601531659837589,
+      "learning_rate": 9.650679396073691e-07,
+      "loss": 0.6011,
+      "step": 36641
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3655562598007083,
+      "learning_rate": 9.647409246717222e-07,
+      "loss": 0.6534,
+      "step": 36642
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4094428778802657,
+      "learning_rate": 9.644139623422132e-07,
+      "loss": 0.7676,
+      "step": 36643
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.798570924180655,
+      "learning_rate": 9.640870526207445e-07,
+      "loss": 0.6091,
+      "step": 36644
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3482417224017826,
+      "learning_rate": 9.637601955092224e-07,
+      "loss": 0.5726,
+      "step": 36645
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4387351420021235,
+      "learning_rate": 9.634333910095516e-07,
+      "loss": 0.743,
+      "step": 36646
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3586916263463342,
+      "learning_rate": 9.63106639123632e-07,
+      "loss": 0.7325,
+      "step": 36647
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7981971668393486,
+      "learning_rate": 9.627799398533644e-07,
+      "loss": 0.7178,
+      "step": 36648
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4427671871111882,
+      "learning_rate": 9.624532932006536e-07,
+      "loss": 0.7109,
+      "step": 36649
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8286595602889408,
+      "learning_rate": 9.621266991674017e-07,
+      "loss": 0.6117,
+      "step": 36650
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.32511751611843,
+      "learning_rate": 9.618001577555113e-07,
+      "loss": 0.6205,
+      "step": 36651
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6597206679062488,
+      "learning_rate": 9.61473668966879e-07,
+      "loss": 0.7069,
+      "step": 36652
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5420069538288819,
+      "learning_rate": 9.6114723280341e-07,
+      "loss": 0.5914,
+      "step": 36653
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2727561768769244,
+      "learning_rate": 9.608208492670023e-07,
+      "loss": 0.639,
+      "step": 36654
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.606094111642574,
+      "learning_rate": 9.604945183595593e-07,
+      "loss": 0.653,
+      "step": 36655
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3178766814141154,
+      "learning_rate": 9.60168240082976e-07,
+      "loss": 0.5983,
+      "step": 36656
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.202025935021969,
+      "learning_rate": 9.598420144391585e-07,
+      "loss": 0.567,
+      "step": 36657
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4170124235534587,
+      "learning_rate": 9.5951584143e-07,
+      "loss": 0.7021,
+      "step": 36658
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3901365737266786,
+      "learning_rate": 9.591897210574053e-07,
+      "loss": 0.6923,
+      "step": 36659
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5089776978751888,
+      "learning_rate": 9.588636533232697e-07,
+      "loss": 0.7109,
+      "step": 36660
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2782907000606398,
+      "learning_rate": 9.585376382294898e-07,
+      "loss": 0.7356,
+      "step": 36661
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4885440172287923,
+      "learning_rate": 9.582116757779668e-07,
+      "loss": 0.66,
+      "step": 36662
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7484244537376803,
+      "learning_rate": 9.57885765970601e-07,
+      "loss": 0.6429,
+      "step": 36663
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5838964937294829,
+      "learning_rate": 9.575599088092858e-07,
+      "loss": 0.4767,
+      "step": 36664
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2847393533761995,
+      "learning_rate": 9.572341042959177e-07,
+      "loss": 0.5495,
+      "step": 36665
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.678737236915968,
+      "learning_rate": 9.569083524323952e-07,
+      "loss": 0.5535,
+      "step": 36666
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4956847462945568,
+      "learning_rate": 9.56582653220618e-07,
+      "loss": 0.729,
+      "step": 36667
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4112958236855497,
+      "learning_rate": 9.562570066624788e-07,
+      "loss": 0.673,
+      "step": 36668
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4124903966379228,
+      "learning_rate": 9.55931412759874e-07,
+      "loss": 0.6506,
+      "step": 36669
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4725381220529041,
+      "learning_rate": 9.556058715146987e-07,
+      "loss": 0.6177,
+      "step": 36670
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2866364137264052,
+      "learning_rate": 9.552803829288514e-07,
+      "loss": 0.5161,
+      "step": 36671
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4639841618525151,
+      "learning_rate": 9.549549470042251e-07,
+      "loss": 0.7796,
+      "step": 36672
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.616956605510332,
+      "learning_rate": 9.546295637427138e-07,
+      "loss": 0.7884,
+      "step": 36673
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.562497614062919,
+      "learning_rate": 9.543042331462139e-07,
+      "loss": 0.4677,
+      "step": 36674
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.060441136354631,
+      "learning_rate": 9.539789552166167e-07,
+      "loss": 0.6844,
+      "step": 36675
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4824871197218688,
+      "learning_rate": 9.536537299558202e-07,
+      "loss": 0.5573,
+      "step": 36676
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.958027955139511,
+      "learning_rate": 9.533285573657137e-07,
+      "loss": 0.6741,
+      "step": 36677
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6203937492489644,
+      "learning_rate": 9.530034374481945e-07,
+      "loss": 0.5073,
+      "step": 36678
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.251270213101272,
+      "learning_rate": 9.526783702051512e-07,
+      "loss": 0.5672,
+      "step": 36679
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3953363500816582,
+      "learning_rate": 9.523533556384812e-07,
+      "loss": 0.7268,
+      "step": 36680
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.290825671029866,
+      "learning_rate": 9.520283937500718e-07,
+      "loss": 0.6191,
+      "step": 36681
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.23361624018741,
+      "learning_rate": 9.517034845418194e-07,
+      "loss": 0.7194,
+      "step": 36682
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.658991944611922,
+      "learning_rate": 9.513786280156135e-07,
+      "loss": 0.7353,
+      "step": 36683
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5060879532829652,
+      "learning_rate": 9.510538241733446e-07,
+      "loss": 0.7429,
+      "step": 36684
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3216695105085425,
+      "learning_rate": 9.507290730169094e-07,
+      "loss": 0.655,
+      "step": 36685
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3262263583582317,
+      "learning_rate": 9.504043745481905e-07,
+      "loss": 0.6124,
+      "step": 36686
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7720678233582268,
+      "learning_rate": 9.500797287690833e-07,
+      "loss": 0.6885,
+      "step": 36687
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1845598066498435,
+      "learning_rate": 9.49755135681476e-07,
+      "loss": 0.6217,
+      "step": 36688
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5110383043496458,
+      "learning_rate": 9.49430595287264e-07,
+      "loss": 0.751,
+      "step": 36689
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.462678492763659,
+      "learning_rate": 9.491061075883279e-07,
+      "loss": 0.6544,
+      "step": 36690
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.507174878595141,
+      "learning_rate": 9.487816725865618e-07,
+      "loss": 0.7751,
+      "step": 36691
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3106911117132938,
+      "learning_rate": 9.484572902838563e-07,
+      "loss": 0.7486,
+      "step": 36692
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5601628563632228,
+      "learning_rate": 9.481329606820988e-07,
+      "loss": 0.6834,
+      "step": 36693
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6727926328759035,
+      "learning_rate": 9.478086837831746e-07,
+      "loss": 0.642,
+      "step": 36694
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4642650553420473,
+      "learning_rate": 9.474844595889743e-07,
+      "loss": 0.6689,
+      "step": 36695
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.401090656813196,
+      "learning_rate": 9.471602881013864e-07,
+      "loss": 0.5481,
+      "step": 36696
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.558633514222008,
+      "learning_rate": 9.468361693222972e-07,
+      "loss": 0.4897,
+      "step": 36697
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4005079579690907,
+      "learning_rate": 9.46512103253594e-07,
+      "loss": 0.7201,
+      "step": 36698
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.263329861726924,
+      "learning_rate": 9.461880898971654e-07,
+      "loss": 0.6565,
+      "step": 36699
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2668992833225865,
+      "learning_rate": 9.458641292548931e-07,
+      "loss": 0.6644,
+      "step": 36700
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.310849527917942,
+      "learning_rate": 9.455402213286669e-07,
+      "loss": 0.7322,
+      "step": 36701
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.325269139251445,
+      "learning_rate": 9.452163661203728e-07,
+      "loss": 0.6122,
+      "step": 36702
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3276097215611444,
+      "learning_rate": 9.448925636318973e-07,
+      "loss": 0.7848,
+      "step": 36703
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5008434278278893,
+      "learning_rate": 9.44568813865121e-07,
+      "loss": 0.5702,
+      "step": 36704
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6191190336165628,
+      "learning_rate": 9.442451168219325e-07,
+      "loss": 0.5523,
+      "step": 36705
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3057018468009751,
+      "learning_rate": 9.439214725042178e-07,
+      "loss": 0.7512,
+      "step": 36706
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.269002984388687,
+      "learning_rate": 9.435978809138601e-07,
+      "loss": 0.7235,
+      "step": 36707
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.514835140309765,
+      "learning_rate": 9.432743420527401e-07,
+      "loss": 0.6052,
+      "step": 36708
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.445775614454287,
+      "learning_rate": 9.429508559227441e-07,
+      "loss": 0.7959,
+      "step": 36709
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4424341777342928,
+      "learning_rate": 9.426274225257582e-07,
+      "loss": 0.5571,
+      "step": 36710
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4619515204834739,
+      "learning_rate": 9.42304041863662e-07,
+      "loss": 0.6416,
+      "step": 36711
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6540439265270026,
+      "learning_rate": 9.419807139383374e-07,
+      "loss": 0.7915,
+      "step": 36712
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.609054018498717,
+      "learning_rate": 9.41657438751672e-07,
+      "loss": 0.6878,
+      "step": 36713
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6011367626420104,
+      "learning_rate": 9.413342163055416e-07,
+      "loss": 0.4753,
+      "step": 36714
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.293164354542021,
+      "learning_rate": 9.410110466018341e-07,
+      "loss": 0.7179,
+      "step": 36715
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.9464248718029586,
+      "learning_rate": 9.406879296424254e-07,
+      "loss": 0.6635,
+      "step": 36716
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.36568937548347,
+      "learning_rate": 9.403648654292019e-07,
+      "loss": 0.6098,
+      "step": 36717
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.165293513908681,
+      "learning_rate": 9.40041853964041e-07,
+      "loss": 0.6592,
+      "step": 36718
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.307718018445939,
+      "learning_rate": 9.397188952488267e-07,
+      "loss": 0.5574,
+      "step": 36719
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3567578074512552,
+      "learning_rate": 9.393959892854366e-07,
+      "loss": 0.5789,
+      "step": 36720
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2732841472359662,
+      "learning_rate": 9.39073136075751e-07,
+      "loss": 0.6443,
+      "step": 36721
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5905697102858983,
+      "learning_rate": 9.387503356216488e-07,
+      "loss": 0.5055,
+      "step": 36722
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3541725191722411,
+      "learning_rate": 9.384275879250138e-07,
+      "loss": 0.5355,
+      "step": 36723
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5402152557890638,
+      "learning_rate": 9.381048929877223e-07,
+      "loss": 0.6555,
+      "step": 36724
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2461894343723259,
+      "learning_rate": 9.377822508116507e-07,
+      "loss": 0.5314,
+      "step": 36725
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5682637501984482,
+      "learning_rate": 9.374596613986797e-07,
+      "loss": 0.4739,
+      "step": 36726
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.248970154314014,
+      "learning_rate": 9.371371247506899e-07,
+      "loss": 0.5716,
+      "step": 36727
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3922075861889083,
+      "learning_rate": 9.368146408695566e-07,
+      "loss": 0.6899,
+      "step": 36728
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.418038355279259,
+      "learning_rate": 9.364922097571572e-07,
+      "loss": 0.6776,
+      "step": 36729
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.650621203656842,
+      "learning_rate": 9.361698314153678e-07,
+      "loss": 0.639,
+      "step": 36730
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.648793202164291,
+      "learning_rate": 9.358475058460704e-07,
+      "loss": 0.741,
+      "step": 36731
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1551012142165151,
+      "learning_rate": 9.355252330511377e-07,
+      "loss": 0.6587,
+      "step": 36732
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5784933693423155,
+      "learning_rate": 9.352030130324463e-07,
+      "loss": 0.8549,
+      "step": 36733
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6109409914378474,
+      "learning_rate": 9.348808457918734e-07,
+      "loss": 0.5252,
+      "step": 36734
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5961415284393651,
+      "learning_rate": 9.345587313312921e-07,
+      "loss": 0.4966,
+      "step": 36735
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5528040129737238,
+      "learning_rate": 9.342366696525829e-07,
+      "loss": 0.4365,
+      "step": 36736
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.658987910628116,
+      "learning_rate": 9.339146607576155e-07,
+      "loss": 0.7204,
+      "step": 36737
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3720119451813126,
+      "learning_rate": 9.335927046482695e-07,
+      "loss": 0.6988,
+      "step": 36738
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5721407972563843,
+      "learning_rate": 9.332708013264147e-07,
+      "loss": 0.5789,
+      "step": 36739
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5424637006590578,
+      "learning_rate": 9.329489507939304e-07,
+      "loss": 0.644,
+      "step": 36740
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.862269848095154,
+      "learning_rate": 9.326271530526865e-07,
+      "loss": 0.7695,
+      "step": 36741
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0195874992699414,
+      "learning_rate": 9.323054081045579e-07,
+      "loss": 0.6506,
+      "step": 36742
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3659894465907236,
+      "learning_rate": 9.319837159514167e-07,
+      "loss": 0.7572,
+      "step": 36743
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2857842699034097,
+      "learning_rate": 9.31662076595139e-07,
+      "loss": 0.6455,
+      "step": 36744
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3594561761693031,
+      "learning_rate": 9.313404900375966e-07,
+      "loss": 0.6187,
+      "step": 36745
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4053396779066476,
+      "learning_rate": 9.310189562806582e-07,
+      "loss": 0.5973,
+      "step": 36746
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3704331648862222,
+      "learning_rate": 9.306974753261999e-07,
+      "loss": 0.6933,
+      "step": 36747
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8737044296312029,
+      "learning_rate": 9.303760471760925e-07,
+      "loss": 0.7212,
+      "step": 36748
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7295570448080237,
+      "learning_rate": 9.300546718322079e-07,
+      "loss": 0.6476,
+      "step": 36749
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3683243121788906,
+      "learning_rate": 9.297333492964144e-07,
+      "loss": 0.6984,
+      "step": 36750
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3208113951860336,
+      "learning_rate": 9.294120795705852e-07,
+      "loss": 0.6425,
+      "step": 36751
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4513081498986633,
+      "learning_rate": 9.290908626565931e-07,
+      "loss": 0.6006,
+      "step": 36752
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5208634010384579,
+      "learning_rate": 9.287696985563044e-07,
+      "loss": 0.6404,
+      "step": 36753
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4689972188643767,
+      "learning_rate": 9.284485872715898e-07,
+      "loss": 0.5886,
+      "step": 36754
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5780285484844927,
+      "learning_rate": 9.281275288043212e-07,
+      "loss": 0.6668,
+      "step": 36755
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.582580305182357,
+      "learning_rate": 9.278065231563637e-07,
+      "loss": 0.5055,
+      "step": 36756
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2061214679318935,
+      "learning_rate": 9.274855703295904e-07,
+      "loss": 0.5951,
+      "step": 36757
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.595199463241866,
+      "learning_rate": 9.271646703258674e-07,
+      "loss": 0.7838,
+      "step": 36758
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4419826226496548,
+      "learning_rate": 9.268438231470656e-07,
+      "loss": 0.684,
+      "step": 36759
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.380479794352851,
+      "learning_rate": 9.265230287950499e-07,
+      "loss": 0.6555,
+      "step": 36760
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2744929013630755,
+      "learning_rate": 9.262022872716914e-07,
+      "loss": 0.6706,
+      "step": 36761
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2713534602316794,
+      "learning_rate": 9.25881598578856e-07,
+      "loss": 0.6407,
+      "step": 36762
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2625540991200666,
+      "learning_rate": 9.255609627184092e-07,
+      "loss": 0.746,
+      "step": 36763
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5830544115968712,
+      "learning_rate": 9.252403796922193e-07,
+      "loss": 0.6115,
+      "step": 36764
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6169005045292885,
+      "learning_rate": 9.249198495021549e-07,
+      "loss": 0.5296,
+      "step": 36765
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.297132881150514,
+      "learning_rate": 9.245993721500801e-07,
+      "loss": 0.5609,
+      "step": 36766
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5588020594480962,
+      "learning_rate": 9.242789476378589e-07,
+      "loss": 0.6774,
+      "step": 36767
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2649420538271732,
+      "learning_rate": 9.239585759673586e-07,
+      "loss": 0.7464,
+      "step": 36768
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.561385816145933,
+      "learning_rate": 9.236382571404478e-07,
+      "loss": 0.6273,
+      "step": 36769
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5538644923254654,
+      "learning_rate": 9.233179911589874e-07,
+      "loss": 0.7378,
+      "step": 36770
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3168322320934254,
+      "learning_rate": 9.229977780248423e-07,
+      "loss": 0.5511,
+      "step": 36771
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.35035479398727,
+      "learning_rate": 9.226776177398778e-07,
+      "loss": 0.621,
+      "step": 36772
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5374655694377823,
+      "learning_rate": 9.223575103059601e-07,
+      "loss": 0.774,
+      "step": 36773
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.061924594354119,
+      "learning_rate": 9.2203745572495e-07,
+      "loss": 0.77,
+      "step": 36774
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3816416402784697,
+      "learning_rate": 9.217174539987106e-07,
+      "loss": 0.6498,
+      "step": 36775
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.344348586085616,
+      "learning_rate": 9.21397505129108e-07,
+      "loss": 0.6496,
+      "step": 36776
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4768101966206222,
+      "learning_rate": 9.210776091180018e-07,
+      "loss": 0.674,
+      "step": 36777
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3411253762495778,
+      "learning_rate": 9.207577659672573e-07,
+      "loss": 0.5582,
+      "step": 36778
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8156250203134114,
+      "learning_rate": 9.204379756787341e-07,
+      "loss": 0.7254,
+      "step": 36779
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2446107740295511,
+      "learning_rate": 9.201182382542983e-07,
+      "loss": 0.6342,
+      "step": 36780
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3707996166731746,
+      "learning_rate": 9.197985536958054e-07,
+      "loss": 0.7379,
+      "step": 36781
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6563318172328618,
+      "learning_rate": 9.194789220051214e-07,
+      "loss": 0.7392,
+      "step": 36782
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.295469246336035,
+      "learning_rate": 9.191593431841084e-07,
+      "loss": 0.591,
+      "step": 36783
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5061097059058508,
+      "learning_rate": 9.188398172346214e-07,
+      "loss": 0.6993,
+      "step": 36784
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.359357489615131,
+      "learning_rate": 9.185203441585244e-07,
+      "loss": 0.6425,
+      "step": 36785
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.465823466916569,
+      "learning_rate": 9.182009239576773e-07,
+      "loss": 0.8249,
+      "step": 36786
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4506223642635854,
+      "learning_rate": 9.178815566339427e-07,
+      "loss": 0.6303,
+      "step": 36787
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5918809631144445,
+      "learning_rate": 9.175622421891738e-07,
+      "loss": 0.5349,
+      "step": 36788
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2151390182567459,
+      "learning_rate": 9.172429806252325e-07,
+      "loss": 0.6962,
+      "step": 36789
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3385981139100966,
+      "learning_rate": 9.169237719439805e-07,
+      "loss": 0.6544,
+      "step": 36790
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4739723130979037,
+      "learning_rate": 9.16604616147273e-07,
+      "loss": 0.5726,
+      "step": 36791
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1471123744696103,
+      "learning_rate": 9.162855132369686e-07,
+      "loss": 0.6409,
+      "step": 36792
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3123979802241488,
+      "learning_rate": 9.159664632149245e-07,
+      "loss": 0.7464,
+      "step": 36793
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5476876823421777,
+      "learning_rate": 9.156474660830028e-07,
+      "loss": 0.6033,
+      "step": 36794
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4951843710146668,
+      "learning_rate": 9.153285218430552e-07,
+      "loss": 0.7248,
+      "step": 36795
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6066562951564207,
+      "learning_rate": 9.150096304969425e-07,
+      "loss": 0.7032,
+      "step": 36796
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4617604448382588,
+      "learning_rate": 9.146907920465197e-07,
+      "loss": 0.5873,
+      "step": 36797
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6325622415257044,
+      "learning_rate": 9.143720064936423e-07,
+      "loss": 0.5692,
+      "step": 36798
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.266366138843311,
+      "learning_rate": 9.140532738401675e-07,
+      "loss": 0.6559,
+      "step": 36799
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5467392779668232,
+      "learning_rate": 9.137345940879538e-07,
+      "loss": 0.7475,
+      "step": 36800
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6091995903673139,
+      "learning_rate": 9.134159672388531e-07,
+      "loss": 0.516,
+      "step": 36801
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.306245626846325,
+      "learning_rate": 9.130973932947196e-07,
+      "loss": 0.7933,
+      "step": 36802
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4066500099393295,
+      "learning_rate": 9.127788722574094e-07,
+      "loss": 0.8115,
+      "step": 36803
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5813172232518813,
+      "learning_rate": 9.124604041287799e-07,
+      "loss": 0.7058,
+      "step": 36804
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4733004011420292,
+      "learning_rate": 9.121419889106831e-07,
+      "loss": 0.6416,
+      "step": 36805
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3034240083326063,
+      "learning_rate": 9.118236266049707e-07,
+      "loss": 0.5308,
+      "step": 36806
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5000705236172174,
+      "learning_rate": 9.115053172134991e-07,
+      "loss": 0.7309,
+      "step": 36807
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4087619429552538,
+      "learning_rate": 9.111870607381223e-07,
+      "loss": 0.5226,
+      "step": 36808
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.385607259023361,
+      "learning_rate": 9.108688571806912e-07,
+      "loss": 0.7701,
+      "step": 36809
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2199600956588106,
+      "learning_rate": 9.105507065430586e-07,
+      "loss": 0.5696,
+      "step": 36810
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4051022454840392,
+      "learning_rate": 9.102326088270763e-07,
+      "loss": 0.7125,
+      "step": 36811
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5977211139662368,
+      "learning_rate": 9.099145640346007e-07,
+      "loss": 0.7064,
+      "step": 36812
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3495947457542306,
+      "learning_rate": 9.095965721674793e-07,
+      "loss": 0.7045,
+      "step": 36813
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4860955517535506,
+      "learning_rate": 9.092786332275627e-07,
+      "loss": 0.6643,
+      "step": 36814
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5783877707143064,
+      "learning_rate": 9.089607472167061e-07,
+      "loss": 0.8025,
+      "step": 36815
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5949548846258786,
+      "learning_rate": 9.08642914136757e-07,
+      "loss": 0.5207,
+      "step": 36816
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4595336563408816,
+      "learning_rate": 9.083251339895682e-07,
+      "loss": 0.6635,
+      "step": 36817
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4282742314573336,
+      "learning_rate": 9.080074067769872e-07,
+      "loss": 0.6248,
+      "step": 36818
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.091662739802969,
+      "learning_rate": 9.076897325008671e-07,
+      "loss": 0.7575,
+      "step": 36819
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5921168382582003,
+      "learning_rate": 9.073721111630552e-07,
+      "loss": 0.5812,
+      "step": 36820
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4554616701427545,
+      "learning_rate": 9.070545427654021e-07,
+      "loss": 0.6257,
+      "step": 36821
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5924524049763016,
+      "learning_rate": 9.067370273097575e-07,
+      "loss": 0.5063,
+      "step": 36822
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4329376904459636,
+      "learning_rate": 9.064195647979656e-07,
+      "loss": 0.7199,
+      "step": 36823
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5617732055405993,
+      "learning_rate": 9.061021552318783e-07,
+      "loss": 0.6997,
+      "step": 36824
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3367335992195106,
+      "learning_rate": 9.057847986133461e-07,
+      "loss": 0.5815,
+      "step": 36825
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3473619446364544,
+      "learning_rate": 9.054674949442132e-07,
+      "loss": 0.6816,
+      "step": 36826
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4194686194532333,
+      "learning_rate": 9.05150244226326e-07,
+      "loss": 0.6618,
+      "step": 36827
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.373028191529903,
+      "learning_rate": 9.04833046461534e-07,
+      "loss": 0.6043,
+      "step": 36828
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4404338619406067,
+      "learning_rate": 9.045159016516847e-07,
+      "loss": 0.6551,
+      "step": 36829
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3714610787880708,
+      "learning_rate": 9.041988097986232e-07,
+      "loss": 0.6423,
+      "step": 36830
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8197253703900809,
+      "learning_rate": 9.038817709041947e-07,
+      "loss": 0.7055,
+      "step": 36831
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.394167943524247,
+      "learning_rate": 9.035647849702467e-07,
+      "loss": 0.6987,
+      "step": 36832
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4780913959905093,
+      "learning_rate": 9.032478519986254e-07,
+      "loss": 0.5528,
+      "step": 36833
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4284638270203154,
+      "learning_rate": 9.02930971991175e-07,
+      "loss": 0.744,
+      "step": 36834
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5452015780741999,
+      "learning_rate": 9.026141449497394e-07,
+      "loss": 0.6935,
+      "step": 36835
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3624866152832056,
+      "learning_rate": 9.022973708761662e-07,
+      "loss": 0.6834,
+      "step": 36836
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3010657250076039,
+      "learning_rate": 9.019806497722961e-07,
+      "loss": 0.6329,
+      "step": 36837
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3857304639014572,
+      "learning_rate": 9.016639816399764e-07,
+      "loss": 0.8046,
+      "step": 36838
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.459440728779489,
+      "learning_rate": 9.013473664810479e-07,
+      "loss": 0.635,
+      "step": 36839
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2842004095278068,
+      "learning_rate": 9.01030804297357e-07,
+      "loss": 0.5213,
+      "step": 36840
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3446576573432172,
+      "learning_rate": 9.007142950907444e-07,
+      "loss": 0.6732,
+      "step": 36841
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.02006576484097,
+      "learning_rate": 9.003978388630552e-07,
+      "loss": 0.8182,
+      "step": 36842
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6322724204797712,
+      "learning_rate": 9.000814356161314e-07,
+      "loss": 0.5033,
+      "step": 36843
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3464615132989475,
+      "learning_rate": 8.997650853518125e-07,
+      "loss": 0.7073,
+      "step": 36844
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5949133377616576,
+      "learning_rate": 8.994487880719415e-07,
+      "loss": 0.4624,
+      "step": 36845
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4353435032919093,
+      "learning_rate": 8.991325437783627e-07,
+      "loss": 0.6214,
+      "step": 36846
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4970402092855803,
+      "learning_rate": 8.988163524729165e-07,
+      "loss": 0.6233,
+      "step": 36847
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3798080177167789,
+      "learning_rate": 8.985002141574406e-07,
+      "loss": 0.5076,
+      "step": 36848
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.529083413814136,
+      "learning_rate": 8.981841288337778e-07,
+      "loss": 0.6671,
+      "step": 36849
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4461123515215413,
+      "learning_rate": 8.978680965037712e-07,
+      "loss": 0.6155,
+      "step": 36850
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3296491672600286,
+      "learning_rate": 8.975521171692569e-07,
+      "loss": 0.6175,
+      "step": 36851
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6071479249817312,
+      "learning_rate": 8.972361908320748e-07,
+      "loss": 0.509,
+      "step": 36852
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5942141436213326,
+      "learning_rate": 8.969203174940655e-07,
+      "loss": 0.5091,
+      "step": 36853
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0497919611305204,
+      "learning_rate": 8.966044971570698e-07,
+      "loss": 0.6527,
+      "step": 36854
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5980414388453035,
+      "learning_rate": 8.96288729822925e-07,
+      "loss": 0.5331,
+      "step": 36855
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7129128842762251,
+      "learning_rate": 8.959730154934676e-07,
+      "loss": 0.7699,
+      "step": 36856
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6812259473744597,
+      "learning_rate": 8.956573541705393e-07,
+      "loss": 0.6557,
+      "step": 36857
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.3039720319362837,
+      "learning_rate": 8.953417458559743e-07,
+      "loss": 0.6255,
+      "step": 36858
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3363543635640374,
+      "learning_rate": 8.950261905516145e-07,
+      "loss": 0.6839,
+      "step": 36859
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.410761034713906,
+      "learning_rate": 8.947106882592927e-07,
+      "loss": 0.7129,
+      "step": 36860
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4110389745920535,
+      "learning_rate": 8.943952389808497e-07,
+      "loss": 0.7136,
+      "step": 36861
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0024600071366936,
+      "learning_rate": 8.940798427181185e-07,
+      "loss": 0.6333,
+      "step": 36862
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5654704740775598,
+      "learning_rate": 8.937644994729388e-07,
+      "loss": 0.7241,
+      "step": 36863
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.416908702502012,
+      "learning_rate": 8.934492092471459e-07,
+      "loss": 0.6993,
+      "step": 36864
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4008366808127697,
+      "learning_rate": 8.931339720425725e-07,
+      "loss": 0.6291,
+      "step": 36865
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2953351344427573,
+      "learning_rate": 8.928187878610561e-07,
+      "loss": 0.7236,
+      "step": 36866
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3410284640346828,
+      "learning_rate": 8.925036567044332e-07,
+      "loss": 0.689,
+      "step": 36867
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5747367134573285,
+      "learning_rate": 8.921885785745399e-07,
+      "loss": 0.4848,
+      "step": 36868
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3188607622333623,
+      "learning_rate": 8.918735534732048e-07,
+      "loss": 0.537,
+      "step": 36869
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3591892216046448,
+      "learning_rate": 8.915585814022653e-07,
+      "loss": 0.5725,
+      "step": 36870
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4815479152094573,
+      "learning_rate": 8.912436623635579e-07,
+      "loss": 0.6903,
+      "step": 36871
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5022532221556744,
+      "learning_rate": 8.90928796358913e-07,
+      "loss": 0.9031,
+      "step": 36872
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.253644115015669,
+      "learning_rate": 8.906139833901629e-07,
+      "loss": 0.5681,
+      "step": 36873
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6347666185261642,
+      "learning_rate": 8.902992234591424e-07,
+      "loss": 0.7853,
+      "step": 36874
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.439038513508514,
+      "learning_rate": 8.899845165676857e-07,
+      "loss": 0.5484,
+      "step": 36875
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3735844429084068,
+      "learning_rate": 8.896698627176226e-07,
+      "loss": 0.7212,
+      "step": 36876
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3486607827685253,
+      "learning_rate": 8.893552619107848e-07,
+      "loss": 0.6689,
+      "step": 36877
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8442569459784213,
+      "learning_rate": 8.890407141490065e-07,
+      "loss": 0.5149,
+      "step": 36878
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4415552912351357,
+      "learning_rate": 8.887262194341162e-07,
+      "loss": 0.8181,
+      "step": 36879
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5741195676303021,
+      "learning_rate": 8.884117777679458e-07,
+      "loss": 0.4618,
+      "step": 36880
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4269463374914721,
+      "learning_rate": 8.880973891523292e-07,
+      "loss": 0.7202,
+      "step": 36881
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3817850388821915,
+      "learning_rate": 8.877830535890941e-07,
+      "loss": 0.7927,
+      "step": 36882
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.421107141078405,
+      "learning_rate": 8.874687710800689e-07,
+      "loss": 0.7253,
+      "step": 36883
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2463206480203786,
+      "learning_rate": 8.871545416270855e-07,
+      "loss": 0.6456,
+      "step": 36884
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8914051574992938,
+      "learning_rate": 8.868403652319779e-07,
+      "loss": 0.613,
+      "step": 36885
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4133532917246723,
+      "learning_rate": 8.865262418965659e-07,
+      "loss": 0.6112,
+      "step": 36886
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6659534572961539,
+      "learning_rate": 8.862121716226846e-07,
+      "loss": 0.6016,
+      "step": 36887
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3278555425128393,
+      "learning_rate": 8.858981544121614e-07,
+      "loss": 0.8148,
+      "step": 36888
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.934744742443329,
+      "learning_rate": 8.85584190266825e-07,
+      "loss": 0.7539,
+      "step": 36889
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4585970886725779,
+      "learning_rate": 8.852702791885048e-07,
+      "loss": 0.6264,
+      "step": 36890
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6835664301933904,
+      "learning_rate": 8.84956421179024e-07,
+      "loss": 0.7526,
+      "step": 36891
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4316004539043596,
+      "learning_rate": 8.846426162402144e-07,
+      "loss": 0.531,
+      "step": 36892
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.236709342161516,
+      "learning_rate": 8.843288643739001e-07,
+      "loss": 0.538,
+      "step": 36893
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4913042978175908,
+      "learning_rate": 8.840151655819107e-07,
+      "loss": 0.5222,
+      "step": 36894
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1829519580856853,
+      "learning_rate": 8.837015198660681e-07,
+      "loss": 0.6517,
+      "step": 36895
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4532477747736428,
+      "learning_rate": 8.833879272282053e-07,
+      "loss": 0.6712,
+      "step": 36896
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3190297497494838,
+      "learning_rate": 8.830743876701408e-07,
+      "loss": 0.5371,
+      "step": 36897
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6006084485959198,
+      "learning_rate": 8.827609011937066e-07,
+      "loss": 0.6116,
+      "step": 36898
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.523246787357514,
+      "learning_rate": 8.824474678007244e-07,
+      "loss": 0.6891,
+      "step": 36899
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2148695708865394,
+      "learning_rate": 8.821340874930173e-07,
+      "loss": 0.7175,
+      "step": 36900
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.297726419631985,
+      "learning_rate": 8.818207602724127e-07,
+      "loss": 0.7007,
+      "step": 36901
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4559899822621414,
+      "learning_rate": 8.815074861407358e-07,
+      "loss": 0.5839,
+      "step": 36902
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3873363120170257,
+      "learning_rate": 8.811942650998096e-07,
+      "loss": 0.5723,
+      "step": 36903
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5803398156426134,
+      "learning_rate": 8.808810971514559e-07,
+      "loss": 0.5015,
+      "step": 36904
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3489955632677404,
+      "learning_rate": 8.805679822974988e-07,
+      "loss": 0.5835,
+      "step": 36905
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1520452805696533,
+      "learning_rate": 8.802549205397637e-07,
+      "loss": 0.4816,
+      "step": 36906
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5728189827583879,
+      "learning_rate": 8.799419118800722e-07,
+      "loss": 0.484,
+      "step": 36907
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.400271880135334,
+      "learning_rate": 8.796289563202454e-07,
+      "loss": 0.7855,
+      "step": 36908
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.556305627237205,
+      "learning_rate": 8.793160538621048e-07,
+      "loss": 0.7016,
+      "step": 36909
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4813250417997057,
+      "learning_rate": 8.790032045074759e-07,
+      "loss": 0.6986,
+      "step": 36910
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8629106537864968,
+      "learning_rate": 8.786904082581782e-07,
+      "loss": 0.7617,
+      "step": 36911
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4511022065885593,
+      "learning_rate": 8.783776651160303e-07,
+      "loss": 0.7326,
+      "step": 36912
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5131890510175712,
+      "learning_rate": 8.780649750828585e-07,
+      "loss": 0.6852,
+      "step": 36913
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.375268594183949,
+      "learning_rate": 8.77752338160478e-07,
+      "loss": 0.7319,
+      "step": 36914
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5817894879288064,
+      "learning_rate": 8.774397543507129e-07,
+      "loss": 0.6891,
+      "step": 36915
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3081244588155128,
+      "learning_rate": 8.771272236553796e-07,
+      "loss": 0.5499,
+      "step": 36916
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4148422783149714,
+      "learning_rate": 8.768147460763021e-07,
+      "loss": 0.7205,
+      "step": 36917
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2871726543798205,
+      "learning_rate": 8.765023216152969e-07,
+      "loss": 0.7255,
+      "step": 36918
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2550944462020541,
+      "learning_rate": 8.761899502741833e-07,
+      "loss": 0.6685,
+      "step": 36919
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4521978090077836,
+      "learning_rate": 8.758776320547812e-07,
+      "loss": 0.6714,
+      "step": 36920
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8627011999830758,
+      "learning_rate": 8.75565366958907e-07,
+      "loss": 0.7568,
+      "step": 36921
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.540500850032157,
+      "learning_rate": 8.752531549883791e-07,
+      "loss": 0.6592,
+      "step": 36922
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6224317094583407,
+      "learning_rate": 8.749409961450183e-07,
+      "loss": 0.4987,
+      "step": 36923
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.391639602166319,
+      "learning_rate": 8.74628890430641e-07,
+      "loss": 0.7759,
+      "step": 36924
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3167606655415462,
+      "learning_rate": 8.743168378470601e-07,
+      "loss": 0.5853,
+      "step": 36925
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.625957771248531,
+      "learning_rate": 8.740048383960964e-07,
+      "loss": 0.8189,
+      "step": 36926
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3874663113197332,
+      "learning_rate": 8.736928920795674e-07,
+      "loss": 0.609,
+      "step": 36927
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.459468221879041,
+      "learning_rate": 8.733809988992881e-07,
+      "loss": 0.7213,
+      "step": 36928
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2456502505081346,
+      "learning_rate": 8.730691588570716e-07,
+      "loss": 0.6851,
+      "step": 36929
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6365628353276842,
+      "learning_rate": 8.727573719547367e-07,
+      "loss": 0.5195,
+      "step": 36930
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3376003543167778,
+      "learning_rate": 8.724456381940994e-07,
+      "loss": 0.6952,
+      "step": 36931
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2990057214341317,
+      "learning_rate": 8.721339575769727e-07,
+      "loss": 0.6975,
+      "step": 36932
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3262049615952158,
+      "learning_rate": 8.71822330105171e-07,
+      "loss": 0.6229,
+      "step": 36933
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5758738307521134,
+      "learning_rate": 8.715107557805092e-07,
+      "loss": 0.604,
+      "step": 36934
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4474285594227194,
+      "learning_rate": 8.711992346048038e-07,
+      "loss": 0.5442,
+      "step": 36935
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3666114473148963,
+      "learning_rate": 8.708877665798666e-07,
+      "loss": 0.5966,
+      "step": 36936
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.528417092344394,
+      "learning_rate": 8.705763517075094e-07,
+      "loss": 0.7802,
+      "step": 36937
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.272642099989058,
+      "learning_rate": 8.702649899895488e-07,
+      "loss": 0.6146,
+      "step": 36938
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3079183821778009,
+      "learning_rate": 8.699536814277931e-07,
+      "loss": 0.6473,
+      "step": 36939
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2927312711871843,
+      "learning_rate": 8.69642426024061e-07,
+      "loss": 0.6692,
+      "step": 36940
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2814828606603457,
+      "learning_rate": 8.693312237801588e-07,
+      "loss": 0.7204,
+      "step": 36941
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2880257860162698,
+      "learning_rate": 8.690200746979016e-07,
+      "loss": 0.6689,
+      "step": 36942
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.556566630403639,
+      "learning_rate": 8.687089787791003e-07,
+      "loss": 0.7747,
+      "step": 36943
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6182533126794676,
+      "learning_rate": 8.683979360255668e-07,
+      "loss": 0.6856,
+      "step": 36944
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.333677198164037,
+      "learning_rate": 8.680869464391129e-07,
+      "loss": 0.6624,
+      "step": 36945
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6109814035298002,
+      "learning_rate": 8.677760100215449e-07,
+      "loss": 0.7164,
+      "step": 36946
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.564958006111322,
+      "learning_rate": 8.674651267746769e-07,
+      "loss": 0.7156,
+      "step": 36947
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5914181782031437,
+      "learning_rate": 8.671542967003188e-07,
+      "loss": 0.5466,
+      "step": 36948
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3905695635512059,
+      "learning_rate": 8.668435198002812e-07,
+      "loss": 0.7526,
+      "step": 36949
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2757047962913073,
+      "learning_rate": 8.665327960763692e-07,
+      "loss": 0.6734,
+      "step": 36950
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6046156560067202,
+      "learning_rate": 8.66222125530396e-07,
+      "loss": 0.6913,
+      "step": 36951
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5992513362610787,
+      "learning_rate": 8.659115081641712e-07,
+      "loss": 0.5761,
+      "step": 36952
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4285593843214917,
+      "learning_rate": 8.656009439794998e-07,
+      "loss": 0.7895,
+      "step": 36953
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4390025681212923,
+      "learning_rate": 8.652904329781908e-07,
+      "loss": 0.5907,
+      "step": 36954
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.237266338208718,
+      "learning_rate": 8.649799751620536e-07,
+      "loss": 0.6335,
+      "step": 36955
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6367847699168636,
+      "learning_rate": 8.646695705328955e-07,
+      "loss": 0.6588,
+      "step": 36956
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.33030536880532,
+      "learning_rate": 8.643592190925243e-07,
+      "loss": 0.648,
+      "step": 36957
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5614736893893314,
+      "learning_rate": 8.640489208427438e-07,
+      "loss": 0.4725,
+      "step": 36958
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3309030488442004,
+      "learning_rate": 8.637386757853649e-07,
+      "loss": 0.6522,
+      "step": 36959
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6141188395776438,
+      "learning_rate": 8.634284839221907e-07,
+      "loss": 0.6204,
+      "step": 36960
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3694366844307526,
+      "learning_rate": 8.631183452550295e-07,
+      "loss": 0.7449,
+      "step": 36961
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4736335623274255,
+      "learning_rate": 8.628082597856835e-07,
+      "loss": 0.6595,
+      "step": 36962
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.412812966946118,
+      "learning_rate": 8.624982275159633e-07,
+      "loss": 0.7416,
+      "step": 36963
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3337319568865618,
+      "learning_rate": 8.621882484476695e-07,
+      "loss": 0.7357,
+      "step": 36964
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.236062310020732,
+      "learning_rate": 8.618783225826088e-07,
+      "loss": 0.7007,
+      "step": 36965
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2889137765224254,
+      "learning_rate": 8.615684499225884e-07,
+      "loss": 0.7191,
+      "step": 36966
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.309237157397676,
+      "learning_rate": 8.61258630469407e-07,
+      "loss": 0.7366,
+      "step": 36967
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7994050279979061,
+      "learning_rate": 8.609488642248709e-07,
+      "loss": 0.7468,
+      "step": 36968
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6037921117205587,
+      "learning_rate": 8.606391511907841e-07,
+      "loss": 0.5792,
+      "step": 36969
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5948097913084385,
+      "learning_rate": 8.60329491368953e-07,
+      "loss": 0.518,
+      "step": 36970
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4551995447462318,
+      "learning_rate": 8.60019884761173e-07,
+      "loss": 0.6678,
+      "step": 36971
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4953934300052023,
+      "learning_rate": 8.597103313692512e-07,
+      "loss": 0.7256,
+      "step": 36972
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4572708514481263,
+      "learning_rate": 8.59400831194992e-07,
+      "loss": 0.7196,
+      "step": 36973
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6119593501145527,
+      "learning_rate": 8.590913842401949e-07,
+      "loss": 0.5118,
+      "step": 36974
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3211806463773847,
+      "learning_rate": 8.587819905066597e-07,
+      "loss": 0.6895,
+      "step": 36975
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4479996512048299,
+      "learning_rate": 8.584726499961893e-07,
+      "loss": 0.7576,
+      "step": 36976
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.410969264356177,
+      "learning_rate": 8.581633627105879e-07,
+      "loss": 0.684,
+      "step": 36977
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3684616130126241,
+      "learning_rate": 8.578541286516506e-07,
+      "loss": 0.6571,
+      "step": 36978
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3376717688745512,
+      "learning_rate": 8.575449478211839e-07,
+      "loss": 0.6223,
+      "step": 36979
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.428415991125849,
+      "learning_rate": 8.572358202209851e-07,
+      "loss": 0.4781,
+      "step": 36980
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4131418542007639,
+      "learning_rate": 8.569267458528519e-07,
+      "loss": 0.5853,
+      "step": 36981
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6414385977766046,
+      "learning_rate": 8.566177247185847e-07,
+      "loss": 0.6117,
+      "step": 36982
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.437164729792659,
+      "learning_rate": 8.563087568199868e-07,
+      "loss": 0.6234,
+      "step": 36983
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3906556760041198,
+      "learning_rate": 8.559998421588545e-07,
+      "loss": 0.7445,
+      "step": 36984
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4723615758461503,
+      "learning_rate": 8.55690980736984e-07,
+      "loss": 0.7864,
+      "step": 36985
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5254524682779704,
+      "learning_rate": 8.553821725561751e-07,
+      "loss": 0.7599,
+      "step": 36986
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4717950969474767,
+      "learning_rate": 8.550734176182285e-07,
+      "loss": 0.647,
+      "step": 36987
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.439621355951344,
+      "learning_rate": 8.547647159249395e-07,
+      "loss": 0.6699,
+      "step": 36988
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.5500672354622735,
+      "learning_rate": 8.544560674781033e-07,
+      "loss": 0.7133,
+      "step": 36989
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9864392908082185,
+      "learning_rate": 8.541474722795207e-07,
+      "loss": 0.615,
+      "step": 36990
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3738946426338037,
+      "learning_rate": 8.538389303309879e-07,
+      "loss": 0.6049,
+      "step": 36991
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.82504294812952,
+      "learning_rate": 8.535304416343005e-07,
+      "loss": 0.6771,
+      "step": 36992
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4229884351606918,
+      "learning_rate": 8.532220061912533e-07,
+      "loss": 0.6506,
+      "step": 36993
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5186963762424786,
+      "learning_rate": 8.529136240036439e-07,
+      "loss": 0.5395,
+      "step": 36994
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4109207023296646,
+      "learning_rate": 8.526052950732666e-07,
+      "loss": 0.7252,
+      "step": 36995
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3912277600950766,
+      "learning_rate": 8.522970194019187e-07,
+      "loss": 0.7319,
+      "step": 36996
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2685272946026604,
+      "learning_rate": 8.51988796991392e-07,
+      "loss": 0.5724,
+      "step": 36997
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3324334239234665,
+      "learning_rate": 8.516806278434852e-07,
+      "loss": 0.5964,
+      "step": 36998
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4247236162380061,
+      "learning_rate": 8.51372511959988e-07,
+      "loss": 0.656,
+      "step": 36999
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.366246460256134,
+      "learning_rate": 8.510644493426979e-07,
+      "loss": 0.6883,
+      "step": 37000
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2556077742572864,
+      "learning_rate": 8.507564399934065e-07,
+      "loss": 0.6405,
+      "step": 37001
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6119928464125199,
+      "learning_rate": 8.504484839139071e-07,
+      "loss": 0.5096,
+      "step": 37002
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4420394745891048,
+      "learning_rate": 8.501405811059926e-07,
+      "loss": 0.5947,
+      "step": 37003
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.506945062873742,
+      "learning_rate": 8.498327315714583e-07,
+      "loss": 0.6179,
+      "step": 37004
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2230625253229526,
+      "learning_rate": 8.495249353120949e-07,
+      "loss": 0.7196,
+      "step": 37005
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6131370798971101,
+      "learning_rate": 8.492171923296922e-07,
+      "loss": 0.5197,
+      "step": 37006
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2991260529258375,
+      "learning_rate": 8.489095026260452e-07,
+      "loss": 0.5856,
+      "step": 37007
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.437187670134913,
+      "learning_rate": 8.48601866202945e-07,
+      "loss": 0.6968,
+      "step": 37008
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.338613792574455,
+      "learning_rate": 8.482942830621821e-07,
+      "loss": 0.6899,
+      "step": 37009
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3806564214450192,
+      "learning_rate": 8.479867532055452e-07,
+      "loss": 0.8114,
+      "step": 37010
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3745873466052232,
+      "learning_rate": 8.476792766348274e-07,
+      "loss": 0.6504,
+      "step": 37011
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6240889250752711,
+      "learning_rate": 8.473718533518205e-07,
+      "loss": 0.4995,
+      "step": 37012
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.240119435367769,
+      "learning_rate": 8.470644833583119e-07,
+      "loss": 0.7204,
+      "step": 37013
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6015696774304576,
+      "learning_rate": 8.467571666560903e-07,
+      "loss": 0.5042,
+      "step": 37014
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4285059044843376,
+      "learning_rate": 8.464499032469475e-07,
+      "loss": 0.6974,
+      "step": 37015
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2658681697873924,
+      "learning_rate": 8.461426931326699e-07,
+      "loss": 0.7109,
+      "step": 37016
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7425254594513573,
+      "learning_rate": 8.458355363150494e-07,
+      "loss": 0.6454,
+      "step": 37017
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3306382154742706,
+      "learning_rate": 8.455284327958702e-07,
+      "loss": 0.7569,
+      "step": 37018
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1961647679753689,
+      "learning_rate": 8.452213825769251e-07,
+      "loss": 0.5516,
+      "step": 37019
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3810287987497107,
+      "learning_rate": 8.449143856599973e-07,
+      "loss": 0.6521,
+      "step": 37020
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4838653845860135,
+      "learning_rate": 8.446074420468786e-07,
+      "loss": 0.6957,
+      "step": 37021
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6099894425959387,
+      "learning_rate": 8.443005517393543e-07,
+      "loss": 0.519,
+      "step": 37022
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.413588355706898,
+      "learning_rate": 8.439937147392086e-07,
+      "loss": 0.6779,
+      "step": 37023
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4305751571009875,
+      "learning_rate": 8.436869310482298e-07,
+      "loss": 0.5491,
+      "step": 37024
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2840173025318453,
+      "learning_rate": 8.433802006682068e-07,
+      "loss": 0.6634,
+      "step": 37025
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5823433440709225,
+      "learning_rate": 8.430735236009225e-07,
+      "loss": 0.6271,
+      "step": 37026
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3532823068813833,
+      "learning_rate": 8.427668998481619e-07,
+      "loss": 0.7561,
+      "step": 37027
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5983005543648855,
+      "learning_rate": 8.424603294117118e-07,
+      "loss": 0.4973,
+      "step": 37028
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2757506852092488,
+      "learning_rate": 8.421538122933593e-07,
+      "loss": 0.6951,
+      "step": 37029
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3437471881176188,
+      "learning_rate": 8.418473484948853e-07,
+      "loss": 0.6961,
+      "step": 37030
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.549187194184649,
+      "learning_rate": 8.41540938018074e-07,
+      "loss": 0.6796,
+      "step": 37031
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6006087866992946,
+      "learning_rate": 8.412345808647115e-07,
+      "loss": 0.6613,
+      "step": 37032
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5798366416843368,
+      "learning_rate": 8.409282770365823e-07,
+      "loss": 0.7422,
+      "step": 37033
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.301783072137743,
+      "learning_rate": 8.406220265354681e-07,
+      "loss": 0.5207,
+      "step": 37034
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.278877429959302,
+      "learning_rate": 8.403158293631508e-07,
+      "loss": 0.7415,
+      "step": 37035
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3698062904576507,
+      "learning_rate": 8.400096855214157e-07,
+      "loss": 0.7195,
+      "step": 37036
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2694834833833244,
+      "learning_rate": 8.397035950120435e-07,
+      "loss": 0.5734,
+      "step": 37037
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4068689442142548,
+      "learning_rate": 8.393975578368186e-07,
+      "loss": 0.6854,
+      "step": 37038
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4603100075796358,
+      "learning_rate": 8.390915739975192e-07,
+      "loss": 0.725,
+      "step": 37039
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3459902194544557,
+      "learning_rate": 8.387856434959307e-07,
+      "loss": 0.6515,
+      "step": 37040
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3652819937826526,
+      "learning_rate": 8.384797663338307e-07,
+      "loss": 0.7252,
+      "step": 37041
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3233782730384351,
+      "learning_rate": 8.381739425130042e-07,
+      "loss": 0.6768,
+      "step": 37042
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5881450495878786,
+      "learning_rate": 8.3786817203523e-07,
+      "loss": 0.4572,
+      "step": 37043
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.453197877779098,
+      "learning_rate": 8.375624549022854e-07,
+      "loss": 0.7331,
+      "step": 37044
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.524212472132825,
+      "learning_rate": 8.372567911159535e-07,
+      "loss": 0.6254,
+      "step": 37045
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3666671251157112,
+      "learning_rate": 8.369511806780151e-07,
+      "loss": 0.7589,
+      "step": 37046
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2469344213838245,
+      "learning_rate": 8.366456235902476e-07,
+      "loss": 0.7548,
+      "step": 37047
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2619344448438738,
+      "learning_rate": 8.363401198544285e-07,
+      "loss": 0.6388,
+      "step": 37048
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4385257713622883,
+      "learning_rate": 8.360346694723387e-07,
+      "loss": 0.5569,
+      "step": 37049
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5980577287409392,
+      "learning_rate": 8.357292724457577e-07,
+      "loss": 0.7659,
+      "step": 37050
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.486021247884261,
+      "learning_rate": 8.35423928776462e-07,
+      "loss": 0.7604,
+      "step": 37051
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3123897412336034,
+      "learning_rate": 8.351186384662291e-07,
+      "loss": 0.7469,
+      "step": 37052
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3372265342246274,
+      "learning_rate": 8.348134015168364e-07,
+      "loss": 0.709,
+      "step": 37053
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3206171958366404,
+      "learning_rate": 8.345082179300634e-07,
+      "loss": 0.6371,
+      "step": 37054
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.590871088635903,
+      "learning_rate": 8.342030877076856e-07,
+      "loss": 0.7016,
+      "step": 37055
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6525634412693421,
+      "learning_rate": 8.338980108514772e-07,
+      "loss": 0.5352,
+      "step": 37056
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3408126683453123,
+      "learning_rate": 8.335929873632164e-07,
+      "loss": 0.7327,
+      "step": 37057
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.570900133029277,
+      "learning_rate": 8.332880172446811e-07,
+      "loss": 0.7388,
+      "step": 37058
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1495308845346437,
+      "learning_rate": 8.329831004976452e-07,
+      "loss": 0.6341,
+      "step": 37059
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7246563186046877,
+      "learning_rate": 8.326782371238818e-07,
+      "loss": 0.6631,
+      "step": 37060
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3016917540993085,
+      "learning_rate": 8.323734271251704e-07,
+      "loss": 0.5769,
+      "step": 37061
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.101610710796885,
+      "learning_rate": 8.32068670503281e-07,
+      "loss": 0.6511,
+      "step": 37062
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.616307968680503,
+      "learning_rate": 8.317639672599909e-07,
+      "loss": 0.6533,
+      "step": 37063
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3900106361550177,
+      "learning_rate": 8.314593173970753e-07,
+      "loss": 0.797,
+      "step": 37064
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2507521676221784,
+      "learning_rate": 8.311547209163052e-07,
+      "loss": 0.6163,
+      "step": 37065
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5986915785191045,
+      "learning_rate": 8.308501778194545e-07,
+      "loss": 0.5152,
+      "step": 37066
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4747186494457174,
+      "learning_rate": 8.305456881082963e-07,
+      "loss": 0.6097,
+      "step": 37067
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5619998491451264,
+      "learning_rate": 8.302412517846081e-07,
+      "loss": 0.6398,
+      "step": 37068
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6196287898706285,
+      "learning_rate": 8.299368688501552e-07,
+      "loss": 0.5341,
+      "step": 37069
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6641101315883904,
+      "learning_rate": 8.296325393067128e-07,
+      "loss": 0.6476,
+      "step": 37070
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.677293867068703,
+      "learning_rate": 8.293282631560529e-07,
+      "loss": 0.6877,
+      "step": 37071
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2675556090536204,
+      "learning_rate": 8.290240403999505e-07,
+      "loss": 0.6515,
+      "step": 37072
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4588398921913688,
+      "learning_rate": 8.287198710401701e-07,
+      "loss": 0.7088,
+      "step": 37073
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6052685359713363,
+      "learning_rate": 8.284157550784866e-07,
+      "loss": 0.4924,
+      "step": 37074
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.379336331229455,
+      "learning_rate": 8.281116925166721e-07,
+      "loss": 0.6182,
+      "step": 37075
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6351388173685832,
+      "learning_rate": 8.278076833564929e-07,
+      "loss": 0.7023,
+      "step": 37076
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.479695668721637,
+      "learning_rate": 8.275037275997233e-07,
+      "loss": 0.8044,
+      "step": 37077
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5078633883535744,
+      "learning_rate": 8.271998252481283e-07,
+      "loss": 0.6973,
+      "step": 37078
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3127316399849582,
+      "learning_rate": 8.268959763034834e-07,
+      "loss": 0.6879,
+      "step": 37079
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.492967155064939,
+      "learning_rate": 8.265921807675515e-07,
+      "loss": 0.7803,
+      "step": 37080
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4166211334213112,
+      "learning_rate": 8.262884386421055e-07,
+      "loss": 0.6952,
+      "step": 37081
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.605632115490803,
+      "learning_rate": 8.25984749928913e-07,
+      "loss": 0.7389,
+      "step": 37082
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4240784042464143,
+      "learning_rate": 8.256811146297405e-07,
+      "loss": 0.7081,
+      "step": 37083
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3488239675359592,
+      "learning_rate": 8.253775327463565e-07,
+      "loss": 0.6889,
+      "step": 37084
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2986453565914524,
+      "learning_rate": 8.250740042805317e-07,
+      "loss": 0.5914,
+      "step": 37085
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2888982018681354,
+      "learning_rate": 8.247705292340302e-07,
+      "loss": 0.664,
+      "step": 37086
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4002854944422785,
+      "learning_rate": 8.244671076086174e-07,
+      "loss": 0.7065,
+      "step": 37087
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3169100634393,
+      "learning_rate": 8.241637394060619e-07,
+      "loss": 0.7049,
+      "step": 37088
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6112966157421352,
+      "learning_rate": 8.238604246281323e-07,
+      "loss": 0.4914,
+      "step": 37089
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3221277096296988,
+      "learning_rate": 8.235571632765927e-07,
+      "loss": 0.6915,
+      "step": 37090
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3300336659970398,
+      "learning_rate": 8.232539553532071e-07,
+      "loss": 0.6905,
+      "step": 37091
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4807339237901802,
+      "learning_rate": 8.229508008597409e-07,
+      "loss": 0.7774,
+      "step": 37092
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3558544621791404,
+      "learning_rate": 8.226476997979638e-07,
+      "loss": 0.6501,
+      "step": 37093
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5364844872889551,
+      "learning_rate": 8.223446521696366e-07,
+      "loss": 0.7314,
+      "step": 37094
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.36605764382884,
+      "learning_rate": 8.220416579765222e-07,
+      "loss": 0.6511,
+      "step": 37095
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4082870828419323,
+      "learning_rate": 8.217387172203895e-07,
+      "loss": 0.7142,
+      "step": 37096
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4092420892487025,
+      "learning_rate": 8.214358299029968e-07,
+      "loss": 0.8468,
+      "step": 37097
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5931578982427476,
+      "learning_rate": 8.211329960261138e-07,
+      "loss": 0.5335,
+      "step": 37098
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4592348461490883,
+      "learning_rate": 8.208302155914982e-07,
+      "loss": 0.6919,
+      "step": 37099
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3468834329318322,
+      "learning_rate": 8.205274886009162e-07,
+      "loss": 0.7309,
+      "step": 37100
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9312994330226119,
+      "learning_rate": 8.202248150561288e-07,
+      "loss": 0.6904,
+      "step": 37101
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5879415178431495,
+      "learning_rate": 8.199221949588998e-07,
+      "loss": 0.4896,
+      "step": 37102
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4253792224331188,
+      "learning_rate": 8.196196283109903e-07,
+      "loss": 0.8062,
+      "step": 37103
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.61817836989739,
+      "learning_rate": 8.193171151141599e-07,
+      "loss": 0.5396,
+      "step": 37104
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2417545592106451,
+      "learning_rate": 8.190146553701716e-07,
+      "loss": 0.4825,
+      "step": 37105
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3307105835070447,
+      "learning_rate": 8.187122490807897e-07,
+      "loss": 0.6805,
+      "step": 37106
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6099974445535016,
+      "learning_rate": 8.184098962477705e-07,
+      "loss": 0.4813,
+      "step": 37107
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4478356623361879,
+      "learning_rate": 8.181075968728746e-07,
+      "loss": 0.7169,
+      "step": 37108
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.560353990827042,
+      "learning_rate": 8.178053509578631e-07,
+      "loss": 0.7858,
+      "step": 37109
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4792502111742398,
+      "learning_rate": 8.175031585044968e-07,
+      "loss": 0.6098,
+      "step": 37110
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4371726991040117,
+      "learning_rate": 8.172010195145352e-07,
+      "loss": 0.7173,
+      "step": 37111
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5662203628693863,
+      "learning_rate": 8.168989339897348e-07,
+      "loss": 0.4954,
+      "step": 37112
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7617164217581056,
+      "learning_rate": 8.165969019318554e-07,
+      "loss": 0.6643,
+      "step": 37113
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4627520004893388,
+      "learning_rate": 8.162949233426587e-07,
+      "loss": 0.7946,
+      "step": 37114
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.328903917480909,
+      "learning_rate": 8.159929982239001e-07,
+      "loss": 0.6136,
+      "step": 37115
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5197859505206084,
+      "learning_rate": 8.156911265773359e-07,
+      "loss": 0.673,
+      "step": 37116
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6433139262361596,
+      "learning_rate": 8.153893084047281e-07,
+      "loss": 0.5723,
+      "step": 37117
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6522867703942004,
+      "learning_rate": 8.150875437078298e-07,
+      "loss": 0.6851,
+      "step": 37118
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3887532419263764,
+      "learning_rate": 8.147858324884005e-07,
+      "loss": 0.6737,
+      "step": 37119
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.477484209430907,
+      "learning_rate": 8.144841747481947e-07,
+      "loss": 0.8447,
+      "step": 37120
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3916781876198026,
+      "learning_rate": 8.141825704889728e-07,
+      "loss": 0.5879,
+      "step": 37121
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5179777230869316,
+      "learning_rate": 8.138810197124846e-07,
+      "loss": 0.565,
+      "step": 37122
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4747064891531358,
+      "learning_rate": 8.135795224204923e-07,
+      "loss": 0.7441,
+      "step": 37123
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3213022078799765,
+      "learning_rate": 8.132780786147487e-07,
+      "loss": 0.6317,
+      "step": 37124
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5815904015531624,
+      "learning_rate": 8.129766882970069e-07,
+      "loss": 0.4556,
+      "step": 37125
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9690681976228204,
+      "learning_rate": 8.126753514690222e-07,
+      "loss": 0.7044,
+      "step": 37126
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6032758017162221,
+      "learning_rate": 8.123740681325531e-07,
+      "loss": 0.5139,
+      "step": 37127
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3763771138802412,
+      "learning_rate": 8.120728382893505e-07,
+      "loss": 0.6198,
+      "step": 37128
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6840202579733097,
+      "learning_rate": 8.117716619411675e-07,
+      "loss": 0.6757,
+      "step": 37129
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.758749749686448,
+      "learning_rate": 8.114705390897581e-07,
+      "loss": 0.6609,
+      "step": 37130
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3724764904709863,
+      "learning_rate": 8.111694697368778e-07,
+      "loss": 0.6854,
+      "step": 37131
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3786133632288673,
+      "learning_rate": 8.108684538842781e-07,
+      "loss": 0.6185,
+      "step": 37132
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.697789907708007,
+      "learning_rate": 8.105674915337103e-07,
+      "loss": 0.6182,
+      "step": 37133
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3975148347000168,
+      "learning_rate": 8.102665826869283e-07,
+      "loss": 0.6501,
+      "step": 37134
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4907253295344682,
+      "learning_rate": 8.09965727345684e-07,
+      "loss": 0.669,
+      "step": 37135
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.500071968650777,
+      "learning_rate": 8.096649255117294e-07,
+      "loss": 0.7669,
+      "step": 37136
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4659607305052755,
+      "learning_rate": 8.093641771868144e-07,
+      "loss": 0.6591,
+      "step": 37137
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2863001137913097,
+      "learning_rate": 8.090634823726917e-07,
+      "loss": 0.6634,
+      "step": 37138
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.0436161460744136,
+      "learning_rate": 8.08762841071109e-07,
+      "loss": 0.5177,
+      "step": 37139
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2393902266496801,
+      "learning_rate": 8.084622532838215e-07,
+      "loss": 0.6548,
+      "step": 37140
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3261863976382684,
+      "learning_rate": 8.081617190125745e-07,
+      "loss": 0.6858,
+      "step": 37141
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6098945807355803,
+      "learning_rate": 8.078612382591222e-07,
+      "loss": 0.5117,
+      "step": 37142
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3326901811706018,
+      "learning_rate": 8.075608110252098e-07,
+      "loss": 0.6035,
+      "step": 37143
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3413866189186254,
+      "learning_rate": 8.072604373125914e-07,
+      "loss": 0.6688,
+      "step": 37144
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3188369224933165,
+      "learning_rate": 8.069601171230124e-07,
+      "loss": 0.6913,
+      "step": 37145
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6320440267869465,
+      "learning_rate": 8.066598504582201e-07,
+      "loss": 0.659,
+      "step": 37146
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5777162088594064,
+      "learning_rate": 8.063596373199645e-07,
+      "loss": 0.4925,
+      "step": 37147
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3472079833310893,
+      "learning_rate": 8.060594777099939e-07,
+      "loss": 0.804,
+      "step": 37148
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3765958244550036,
+      "learning_rate": 8.057593716300605e-07,
+      "loss": 0.7162,
+      "step": 37149
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.432190406883777,
+      "learning_rate": 8.054593190819016e-07,
+      "loss": 0.5835,
+      "step": 37150
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.505077807505067,
+      "learning_rate": 8.051593200672703e-07,
+      "loss": 0.7176,
+      "step": 37151
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.538997564210625,
+      "learning_rate": 8.048593745879141e-07,
+      "loss": 0.6598,
+      "step": 37152
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.271657690370306,
+      "learning_rate": 8.045594826455783e-07,
+      "loss": 0.745,
+      "step": 37153
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.67948386382555,
+      "learning_rate": 8.042596442420059e-07,
+      "loss": 0.6912,
+      "step": 37154
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.457004765771705,
+      "learning_rate": 8.039598593789455e-07,
+      "loss": 0.6874,
+      "step": 37155
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.511199788416123,
+      "learning_rate": 8.036601280581446e-07,
+      "loss": 0.7677,
+      "step": 37156
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5755545251060916,
+      "learning_rate": 8.033604502813453e-07,
+      "loss": 0.7113,
+      "step": 37157
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3477968343241034,
+      "learning_rate": 8.030608260502915e-07,
+      "loss": 0.6365,
+      "step": 37158
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.032052914692509,
+      "learning_rate": 8.027612553667308e-07,
+      "loss": 0.5697,
+      "step": 37159
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.510540193148479,
+      "learning_rate": 8.02461738232404e-07,
+      "loss": 0.755,
+      "step": 37160
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.26877172971415,
+      "learning_rate": 8.021622746490565e-07,
+      "loss": 0.6074,
+      "step": 37161
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4265437924270983,
+      "learning_rate": 8.018628646184345e-07,
+      "loss": 0.5984,
+      "step": 37162
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5879050861445165,
+      "learning_rate": 8.01563508142279e-07,
+      "loss": 0.6678,
+      "step": 37163
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5995229148707255,
+      "learning_rate": 8.012642052223319e-07,
+      "loss": 0.5651,
+      "step": 37164
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5865337490456534,
+      "learning_rate": 8.009649558603361e-07,
+      "loss": 0.4903,
+      "step": 37165
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.450576977766651,
+      "learning_rate": 8.006657600580381e-07,
+      "loss": 0.7097,
+      "step": 37166
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3481835199063996,
+      "learning_rate": 8.003666178171732e-07,
+      "loss": 0.6691,
+      "step": 37167
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3497369346050332,
+      "learning_rate": 8.000675291394855e-07,
+      "loss": 0.6791,
+      "step": 37168
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6024714527438404,
+      "learning_rate": 7.997684940267181e-07,
+      "loss": 0.5388,
+      "step": 37169
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4710771348712646,
+      "learning_rate": 7.994695124806152e-07,
+      "loss": 0.6483,
+      "step": 37170
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5625140924931868,
+      "learning_rate": 7.991705845029096e-07,
+      "loss": 0.5072,
+      "step": 37171
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7928603902293045,
+      "learning_rate": 7.988717100953458e-07,
+      "loss": 0.6488,
+      "step": 37172
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.527935067557539,
+      "learning_rate": 7.985728892596656e-07,
+      "loss": 0.8034,
+      "step": 37173
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3779898860736406,
+      "learning_rate": 7.982741219976065e-07,
+      "loss": 0.7507,
+      "step": 37174
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.505148784151568,
+      "learning_rate": 7.979754083109092e-07,
+      "loss": 0.6433,
+      "step": 37175
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4684745624707995,
+      "learning_rate": 7.976767482013115e-07,
+      "loss": 0.6994,
+      "step": 37176
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4651307693664306,
+      "learning_rate": 7.973781416705539e-07,
+      "loss": 0.6673,
+      "step": 37177
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8262394466431697,
+      "learning_rate": 7.970795887203731e-07,
+      "loss": 0.7324,
+      "step": 37178
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6344830440640548,
+      "learning_rate": 7.967810893525096e-07,
+      "loss": 0.6414,
+      "step": 37179
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4592662855435155,
+      "learning_rate": 7.964826435687001e-07,
+      "loss": 0.5469,
+      "step": 37180
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8854227815389704,
+      "learning_rate": 7.96184251370683e-07,
+      "loss": 0.588,
+      "step": 37181
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5438347309247151,
+      "learning_rate": 7.958859127601937e-07,
+      "loss": 0.6696,
+      "step": 37182
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.586810513831869,
+      "learning_rate": 7.955876277389718e-07,
+      "loss": 0.6614,
+      "step": 37183
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5751995351674636,
+      "learning_rate": 7.952893963087527e-07,
+      "loss": 0.4836,
+      "step": 37184
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.423677241868111,
+      "learning_rate": 7.949912184712716e-07,
+      "loss": 0.6309,
+      "step": 37185
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.425539469784352,
+      "learning_rate": 7.94693094228266e-07,
+      "loss": 0.6108,
+      "step": 37186
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7362812296497823,
+      "learning_rate": 7.943950235814734e-07,
+      "loss": 0.798,
+      "step": 37187
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5986949176586264,
+      "learning_rate": 7.94097006532627e-07,
+      "loss": 0.4981,
+      "step": 37188
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4011070510323742,
+      "learning_rate": 7.937990430834607e-07,
+      "loss": 0.7385,
+      "step": 37189
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2616106243982321,
+      "learning_rate": 7.935011332357113e-07,
+      "loss": 0.716,
+      "step": 37190
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.517699740988216,
+      "learning_rate": 7.932032769911146e-07,
+      "loss": 0.6779,
+      "step": 37191
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5099072729766396,
+      "learning_rate": 7.929054743514042e-07,
+      "loss": 0.5044,
+      "step": 37192
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4633579089959314,
+      "learning_rate": 7.926077253183096e-07,
+      "loss": 0.5926,
+      "step": 37193
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.435833838364555,
+      "learning_rate": 7.923100298935693e-07,
+      "loss": 0.7297,
+      "step": 37194
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7101444588205992,
+      "learning_rate": 7.920123880789155e-07,
+      "loss": 0.724,
+      "step": 37195
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4005549738962886,
+      "learning_rate": 7.917147998760821e-07,
+      "loss": 0.578,
+      "step": 37196
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6916730813649539,
+      "learning_rate": 7.91417265286798e-07,
+      "loss": 0.577,
+      "step": 37197
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4475578981339126,
+      "learning_rate": 7.911197843127993e-07,
+      "loss": 0.5509,
+      "step": 37198
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5183564881443492,
+      "learning_rate": 7.90822356955816e-07,
+      "loss": 0.7204,
+      "step": 37199
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4059612018909764,
+      "learning_rate": 7.905249832175809e-07,
+      "loss": 0.622,
+      "step": 37200
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3290337931159093,
+      "learning_rate": 7.90227663099824e-07,
+      "loss": 0.6456,
+      "step": 37201
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5122407806707139,
+      "learning_rate": 7.899303966042793e-07,
+      "loss": 0.702,
+      "step": 37202
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5873367627561674,
+      "learning_rate": 7.896331837326743e-07,
+      "loss": 0.6952,
+      "step": 37203
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3053432586646925,
+      "learning_rate": 7.893360244867421e-07,
+      "loss": 0.66,
+      "step": 37204
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.374220386658196,
+      "learning_rate": 7.890389188682112e-07,
+      "loss": 0.5784,
+      "step": 37205
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2437785554113618,
+      "learning_rate": 7.887418668788104e-07,
+      "loss": 0.5331,
+      "step": 37206
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4215460210244701,
+      "learning_rate": 7.884448685202717e-07,
+      "loss": 0.6489,
+      "step": 37207
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4225252356102307,
+      "learning_rate": 7.881479237943235e-07,
+      "loss": 0.8211,
+      "step": 37208
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5325716834172474,
+      "learning_rate": 7.878510327026956e-07,
+      "loss": 0.6255,
+      "step": 37209
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5703610671930015,
+      "learning_rate": 7.875541952471133e-07,
+      "loss": 0.6681,
+      "step": 37210
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6031544455372633,
+      "learning_rate": 7.872574114293075e-07,
+      "loss": 0.4785,
+      "step": 37211
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.42330081683353,
+      "learning_rate": 7.869606812510067e-07,
+      "loss": 0.6604,
+      "step": 37212
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5605934875232412,
+      "learning_rate": 7.866640047139384e-07,
+      "loss": 0.7928,
+      "step": 37213
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3223113139662057,
+      "learning_rate": 7.863673818198269e-07,
+      "loss": 0.6908,
+      "step": 37214
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3689058291012701,
+      "learning_rate": 7.860708125704019e-07,
+      "loss": 0.4832,
+      "step": 37215
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8884106176157895,
+      "learning_rate": 7.857742969673909e-07,
+      "loss": 0.6797,
+      "step": 37216
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5978659703978537,
+      "learning_rate": 7.854778350125191e-07,
+      "loss": 0.5966,
+      "step": 37217
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2776574148576632,
+      "learning_rate": 7.85181426707512e-07,
+      "loss": 0.8291,
+      "step": 37218
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3101094528196682,
+      "learning_rate": 7.848850720540968e-07,
+      "loss": 0.6739,
+      "step": 37219
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3493955197407312,
+      "learning_rate": 7.845887710539957e-07,
+      "loss": 0.6067,
+      "step": 37220
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4898290410720507,
+      "learning_rate": 7.842925237089394e-07,
+      "loss": 0.6404,
+      "step": 37221
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3675608202745981,
+      "learning_rate": 7.839963300206477e-07,
+      "loss": 0.5533,
+      "step": 37222
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1175756112142485,
+      "learning_rate": 7.837001899908481e-07,
+      "loss": 0.5059,
+      "step": 37223
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0017291776419848,
+      "learning_rate": 7.834041036212625e-07,
+      "loss": 0.8058,
+      "step": 37224
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.456205933536647,
+      "learning_rate": 7.831080709136185e-07,
+      "loss": 0.5226,
+      "step": 37225
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3077347675103252,
+      "learning_rate": 7.828120918696369e-07,
+      "loss": 0.5813,
+      "step": 37226
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3118873415521406,
+      "learning_rate": 7.825161664910397e-07,
+      "loss": 0.5654,
+      "step": 37227
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4050519487463191,
+      "learning_rate": 7.822202947795509e-07,
+      "loss": 0.6375,
+      "step": 37228
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6145268790474722,
+      "learning_rate": 7.81924476736896e-07,
+      "loss": 0.4674,
+      "step": 37229
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5831516030758723,
+      "learning_rate": 7.816287123647958e-07,
+      "loss": 0.7163,
+      "step": 37230
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4211526937170602,
+      "learning_rate": 7.813330016649689e-07,
+      "loss": 0.7966,
+      "step": 37231
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7454133399521776,
+      "learning_rate": 7.810373446391406e-07,
+      "loss": 0.669,
+      "step": 37232
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1738832232793686,
+      "learning_rate": 7.807417412890339e-07,
+      "loss": 0.6428,
+      "step": 37233
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5880170521645018,
+      "learning_rate": 7.804461916163663e-07,
+      "loss": 0.4905,
+      "step": 37234
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5766300677080411,
+      "learning_rate": 7.801506956228588e-07,
+      "loss": 0.7899,
+      "step": 37235
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3603568409667057,
+      "learning_rate": 7.798552533102322e-07,
+      "loss": 0.6181,
+      "step": 37236
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4908678191937539,
+      "learning_rate": 7.795598646802105e-07,
+      "loss": 0.6352,
+      "step": 37237
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2791051291863673,
+      "learning_rate": 7.792645297345091e-07,
+      "loss": 0.6365,
+      "step": 37238
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3136349843409998,
+      "learning_rate": 7.789692484748479e-07,
+      "loss": 0.7341,
+      "step": 37239
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5083637357744422,
+      "learning_rate": 7.786740209029475e-07,
+      "loss": 0.6366,
+      "step": 37240
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4261405666689637,
+      "learning_rate": 7.783788470205256e-07,
+      "loss": 0.696,
+      "step": 37241
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.210024821511387,
+      "learning_rate": 7.78083726829304e-07,
+      "loss": 0.5639,
+      "step": 37242
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3274602265438573,
+      "learning_rate": 7.777886603309958e-07,
+      "loss": 0.607,
+      "step": 37243
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4232774950876395,
+      "learning_rate": 7.774936475273243e-07,
+      "loss": 0.5677,
+      "step": 37244
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3331008535608284,
+      "learning_rate": 7.771986884200022e-07,
+      "loss": 0.7032,
+      "step": 37245
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4429314226158707,
+      "learning_rate": 7.769037830107484e-07,
+      "loss": 0.6527,
+      "step": 37246
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4341560346968667,
+      "learning_rate": 7.766089313012859e-07,
+      "loss": 0.7297,
+      "step": 37247
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5915566273334594,
+      "learning_rate": 7.763141332933222e-07,
+      "loss": 0.7484,
+      "step": 37248
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.295080904395492,
+      "learning_rate": 7.76019388988577e-07,
+      "loss": 0.6132,
+      "step": 37249
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5806674734013928,
+      "learning_rate": 7.757246983887679e-07,
+      "loss": 0.6132,
+      "step": 37250
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2701784118495,
+      "learning_rate": 7.754300614956134e-07,
+      "loss": 0.6556,
+      "step": 37251
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5096579243653991,
+      "learning_rate": 7.751354783108223e-07,
+      "loss": 0.7043,
+      "step": 37252
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3706427891689226,
+      "learning_rate": 7.748409488361131e-07,
+      "loss": 0.7218,
+      "step": 37253
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.335246444584445,
+      "learning_rate": 7.745464730732011e-07,
+      "loss": 0.6974,
+      "step": 37254
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.431362483287919,
+      "learning_rate": 7.742520510238016e-07,
+      "loss": 0.6922,
+      "step": 37255
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3592786565323534,
+      "learning_rate": 7.739576826896244e-07,
+      "loss": 0.6372,
+      "step": 37256
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3915245021535814,
+      "learning_rate": 7.736633680723882e-07,
+      "loss": 0.7217,
+      "step": 37257
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.459154013711084,
+      "learning_rate": 7.733691071738047e-07,
+      "loss": 0.7579,
+      "step": 37258
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5910636184493956,
+      "learning_rate": 7.730748999955872e-07,
+      "loss": 0.6544,
+      "step": 37259
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6014946510295767,
+      "learning_rate": 7.7278074653945e-07,
+      "loss": 0.5388,
+      "step": 37260
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3473019999880174,
+      "learning_rate": 7.724866468071046e-07,
+      "loss": 0.6606,
+      "step": 37261
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2298102072178951,
+      "learning_rate": 7.721926008002611e-07,
+      "loss": 0.6165,
+      "step": 37262
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5508606398627272,
+      "learning_rate": 7.718986085206348e-07,
+      "loss": 0.7154,
+      "step": 37263
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4289575974144684,
+      "learning_rate": 7.716046699699376e-07,
+      "loss": 0.711,
+      "step": 37264
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.608979521039086,
+      "learning_rate": 7.713107851498791e-07,
+      "loss": 0.6966,
+      "step": 37265
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5739655647116702,
+      "learning_rate": 7.710169540621692e-07,
+      "loss": 0.6137,
+      "step": 37266
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4695462863738977,
+      "learning_rate": 7.70723176708521e-07,
+      "loss": 0.6823,
+      "step": 37267
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4503091699003097,
+      "learning_rate": 7.704294530906487e-07,
+      "loss": 0.7076,
+      "step": 37268
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.0968351648266115,
+      "learning_rate": 7.70135783210253e-07,
+      "loss": 0.6874,
+      "step": 37269
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.459411378407527,
+      "learning_rate": 7.698421670690503e-07,
+      "loss": 0.6302,
+      "step": 37270
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3769639336575203,
+      "learning_rate": 7.695486046687473e-07,
+      "loss": 0.5905,
+      "step": 37271
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3680587370172699,
+      "learning_rate": 7.692550960110578e-07,
+      "loss": 0.7598,
+      "step": 37272
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2529913750408068,
+      "learning_rate": 7.689616410976863e-07,
+      "loss": 0.7823,
+      "step": 37273
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.607874841471006,
+      "learning_rate": 7.686682399303425e-07,
+      "loss": 0.4992,
+      "step": 37274
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4520523182416791,
+      "learning_rate": 7.683748925107348e-07,
+      "loss": 0.6597,
+      "step": 37275
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1531544806453076,
+      "learning_rate": 7.680815988405698e-07,
+      "loss": 0.5614,
+      "step": 37276
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.416801788887464,
+      "learning_rate": 7.677883589215596e-07,
+      "loss": 0.702,
+      "step": 37277
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.549853703836405,
+      "learning_rate": 7.674951727554058e-07,
+      "loss": 0.6952,
+      "step": 37278
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.230944648789723,
+      "learning_rate": 7.672020403438196e-07,
+      "loss": 0.6001,
+      "step": 37279
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4284101399626519,
+      "learning_rate": 7.66908961688504e-07,
+      "loss": 0.6455,
+      "step": 37280
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3687109074097161,
+      "learning_rate": 7.666159367911696e-07,
+      "loss": 0.5464,
+      "step": 37281
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4310709121034957,
+      "learning_rate": 7.663229656535209e-07,
+      "loss": 0.7185,
+      "step": 37282
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.47100563283724,
+      "learning_rate": 7.66030048277262e-07,
+      "loss": 0.652,
+      "step": 37283
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.513031010538166,
+      "learning_rate": 7.657371846640982e-07,
+      "loss": 0.62,
+      "step": 37284
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5925077007755731,
+      "learning_rate": 7.654443748157391e-07,
+      "loss": 0.4783,
+      "step": 37285
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5610860517463145,
+      "learning_rate": 7.651516187338859e-07,
+      "loss": 0.7339,
+      "step": 37286
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3818948366279988,
+      "learning_rate": 7.648589164202413e-07,
+      "loss": 0.7338,
+      "step": 37287
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6095033579776954,
+      "learning_rate": 7.645662678765131e-07,
+      "loss": 0.5884,
+      "step": 37288
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3610491095873265,
+      "learning_rate": 7.642736731044043e-07,
+      "loss": 0.7079,
+      "step": 37289
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2250315122306357,
+      "learning_rate": 7.639811321056189e-07,
+      "loss": 0.6192,
+      "step": 37290
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5075806214146648,
+      "learning_rate": 7.636886448818582e-07,
+      "loss": 0.7998,
+      "step": 37291
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3387471067648244,
+      "learning_rate": 7.633962114348259e-07,
+      "loss": 0.7197,
+      "step": 37292
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4280886977698948,
+      "learning_rate": 7.631038317662265e-07,
+      "loss": 0.6723,
+      "step": 37293
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1717921827030269,
+      "learning_rate": 7.628115058777619e-07,
+      "loss": 0.6048,
+      "step": 37294
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4894189281500871,
+      "learning_rate": 7.625192337711306e-07,
+      "loss": 0.7426,
+      "step": 37295
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7304456537355757,
+      "learning_rate": 7.622270154480382e-07,
+      "loss": 0.6921,
+      "step": 37296
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3181475095251658,
+      "learning_rate": 7.619348509101843e-07,
+      "loss": 0.6835,
+      "step": 37297
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7333000264472946,
+      "learning_rate": 7.616427401592708e-07,
+      "loss": 0.6512,
+      "step": 37298
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3429888453233234,
+      "learning_rate": 7.613506831969975e-07,
+      "loss": 0.6031,
+      "step": 37299
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.318752571685001,
+      "learning_rate": 7.610586800250663e-07,
+      "loss": 0.626,
+      "step": 37300
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.918910223175052,
+      "learning_rate": 7.607667306451749e-07,
+      "loss": 0.5604,
+      "step": 37301
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.734584019660764,
+      "learning_rate": 7.604748350590274e-07,
+      "loss": 0.7388,
+      "step": 37302
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5898345750960914,
+      "learning_rate": 7.60182993268318e-07,
+      "loss": 0.6902,
+      "step": 37303
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7462770181976932,
+      "learning_rate": 7.598912052747498e-07,
+      "loss": 0.745,
+      "step": 37304
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2772279027550073,
+      "learning_rate": 7.595994710800192e-07,
+      "loss": 0.6198,
+      "step": 37305
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3211025489244146,
+      "learning_rate": 7.593077906858281e-07,
+      "loss": 0.708,
+      "step": 37306
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5951765719858599,
+      "learning_rate": 7.59016164093872e-07,
+      "loss": 0.493,
+      "step": 37307
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4452749862828596,
+      "learning_rate": 7.587245913058472e-07,
+      "loss": 0.6666,
+      "step": 37308
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5454876119572842,
+      "learning_rate": 7.584330723234545e-07,
+      "loss": 0.7743,
+      "step": 37309
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3470427360873487,
+      "learning_rate": 7.581416071483915e-07,
+      "loss": 0.7244,
+      "step": 37310
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2934687228541146,
+      "learning_rate": 7.578501957823547e-07,
+      "loss": 0.6474,
+      "step": 37311
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2860736066175116,
+      "learning_rate": 7.575588382270372e-07,
+      "loss": 0.6388,
+      "step": 37312
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3642593624545476,
+      "learning_rate": 7.572675344841396e-07,
+      "loss": 0.7391,
+      "step": 37313
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.618322367989649,
+      "learning_rate": 7.569762845553574e-07,
+      "loss": 0.7369,
+      "step": 37314
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5924486460930685,
+      "learning_rate": 7.566850884423859e-07,
+      "loss": 0.5059,
+      "step": 37315
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.223317945690146,
+      "learning_rate": 7.563939461469183e-07,
+      "loss": 0.6046,
+      "step": 37316
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5152595445858772,
+      "learning_rate": 7.561028576706519e-07,
+      "loss": 0.6809,
+      "step": 37317
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4668974633575156,
+      "learning_rate": 7.558118230152822e-07,
+      "loss": 0.7632,
+      "step": 37318
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.865693269440373,
+      "learning_rate": 7.555208421825033e-07,
+      "loss": 0.628,
+      "step": 37319
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6114994100615604,
+      "learning_rate": 7.552299151740072e-07,
+      "loss": 0.512,
+      "step": 37320
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8969032797928742,
+      "learning_rate": 7.549390419914915e-07,
+      "loss": 0.6996,
+      "step": 37321
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3079312195741293,
+      "learning_rate": 7.546482226366447e-07,
+      "loss": 0.7191,
+      "step": 37322
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7949345932014151,
+      "learning_rate": 7.543574571111656e-07,
+      "loss": 0.7189,
+      "step": 37323
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.61149663032739,
+      "learning_rate": 7.540667454167427e-07,
+      "loss": 0.4698,
+      "step": 37324
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.451988980768847,
+      "learning_rate": 7.537760875550714e-07,
+      "loss": 0.7358,
+      "step": 37325
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5283304220617802,
+      "learning_rate": 7.534854835278427e-07,
+      "loss": 0.6278,
+      "step": 37326
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9501106285636525,
+      "learning_rate": 7.531949333367495e-07,
+      "loss": 0.6746,
+      "step": 37327
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.716103044336939,
+      "learning_rate": 7.529044369834826e-07,
+      "loss": 0.7233,
+      "step": 37328
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.463554686446423,
+      "learning_rate": 7.526139944697331e-07,
+      "loss": 0.6384,
+      "step": 37329
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5462526063889843,
+      "learning_rate": 7.523236057971917e-07,
+      "loss": 0.6563,
+      "step": 37330
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4204556899342986,
+      "learning_rate": 7.520332709675504e-07,
+      "loss": 0.5926,
+      "step": 37331
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3857524840486186,
+      "learning_rate": 7.517429899825035e-07,
+      "loss": 0.5913,
+      "step": 37332
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.497183356314258,
+      "learning_rate": 7.514527628437318e-07,
+      "loss": 0.7434,
+      "step": 37333
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3129765539283522,
+      "learning_rate": 7.511625895529317e-07,
+      "loss": 0.7461,
+      "step": 37334
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2641258502149486,
+      "learning_rate": 7.50872470111792e-07,
+      "loss": 0.7877,
+      "step": 37335
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4322803805177073,
+      "learning_rate": 7.505824045220012e-07,
+      "loss": 0.7964,
+      "step": 37336
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3186797971832835,
+      "learning_rate": 7.502923927852457e-07,
+      "loss": 0.4804,
+      "step": 37337
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3377745952638647,
+      "learning_rate": 7.500024349032165e-07,
+      "loss": 0.7775,
+      "step": 37338
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.771014880296705,
+      "learning_rate": 7.497125308776043e-07,
+      "loss": 0.7406,
+      "step": 37339
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4180049632590321,
+      "learning_rate": 7.494226807100935e-07,
+      "loss": 0.6734,
+      "step": 37340
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1749220366692747,
+      "learning_rate": 7.491328844023704e-07,
+      "loss": 0.6079,
+      "step": 37341
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6660453614167812,
+      "learning_rate": 7.48843141956127e-07,
+      "loss": 0.7406,
+      "step": 37342
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3017055333587508,
+      "learning_rate": 7.485534533730454e-07,
+      "loss": 0.7403,
+      "step": 37343
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4130385417237936,
+      "learning_rate": 7.482638186548153e-07,
+      "loss": 0.7282,
+      "step": 37344
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4431017029984108,
+      "learning_rate": 7.47974237803123e-07,
+      "loss": 0.6075,
+      "step": 37345
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6209201491964542,
+      "learning_rate": 7.47684710819655e-07,
+      "loss": 0.634,
+      "step": 37346
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4949472341362704,
+      "learning_rate": 7.473952377060934e-07,
+      "loss": 0.7419,
+      "step": 37347
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.46639992797741,
+      "learning_rate": 7.471058184641255e-07,
+      "loss": 0.7801,
+      "step": 37348
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4390758520066504,
+      "learning_rate": 7.468164530954414e-07,
+      "loss": 0.7106,
+      "step": 37349
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4440384542471616,
+      "learning_rate": 7.465271416017172e-07,
+      "loss": 0.5317,
+      "step": 37350
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2493243200854636,
+      "learning_rate": 7.462378839846419e-07,
+      "loss": 0.5721,
+      "step": 37351
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3851661168040414,
+      "learning_rate": 7.459486802458993e-07,
+      "loss": 0.6842,
+      "step": 37352
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.8767966959340483,
+      "learning_rate": 7.456595303871772e-07,
+      "loss": 0.6009,
+      "step": 37353
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3453732873093271,
+      "learning_rate": 7.453704344101509e-07,
+      "loss": 0.6953,
+      "step": 37354
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6362363677755909,
+      "learning_rate": 7.45081392316509e-07,
+      "loss": 0.5086,
+      "step": 37355
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.9770204662011523,
+      "learning_rate": 7.447924041079347e-07,
+      "loss": 0.6825,
+      "step": 37356
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5512717545560646,
+      "learning_rate": 7.445034697861064e-07,
+      "loss": 0.7666,
+      "step": 37357
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.381918868535072,
+      "learning_rate": 7.442145893527119e-07,
+      "loss": 0.6597,
+      "step": 37358
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.474066172831503,
+      "learning_rate": 7.439257628094288e-07,
+      "loss": 0.6891,
+      "step": 37359
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6313718212645574,
+      "learning_rate": 7.43636990157941e-07,
+      "loss": 0.5475,
+      "step": 37360
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5880845316467573,
+      "learning_rate": 7.433482713999285e-07,
+      "loss": 0.5078,
+      "step": 37361
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4929526663582873,
+      "learning_rate": 7.430596065370743e-07,
+      "loss": 0.5426,
+      "step": 37362
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4612511694719725,
+      "learning_rate": 7.427709955710572e-07,
+      "loss": 0.6545,
+      "step": 37363
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.509939978624415,
+      "learning_rate": 7.424824385035568e-07,
+      "loss": 0.6788,
+      "step": 37364
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3255859848559202,
+      "learning_rate": 7.42193935336255e-07,
+      "loss": 0.7378,
+      "step": 37365
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4371641527542403,
+      "learning_rate": 7.419054860708319e-07,
+      "loss": 0.6714,
+      "step": 37366
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3031230438081718,
+      "learning_rate": 7.416170907089671e-07,
+      "loss": 0.6916,
+      "step": 37367
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5878111015797329,
+      "learning_rate": 7.413287492523369e-07,
+      "loss": 0.4933,
+      "step": 37368
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4577341282654195,
+      "learning_rate": 7.410404617026212e-07,
+      "loss": 0.7768,
+      "step": 37369
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3150793712596947,
+      "learning_rate": 7.40752228061502e-07,
+      "loss": 0.6861,
+      "step": 37370
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.339251465962299,
+      "learning_rate": 7.404640483306536e-07,
+      "loss": 0.655,
+      "step": 37371
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4272240526875581,
+      "learning_rate": 7.401759225117544e-07,
+      "loss": 0.636,
+      "step": 37372
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4946932938272672,
+      "learning_rate": 7.398878506064821e-07,
+      "loss": 0.603,
+      "step": 37373
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2582447813244646,
+      "learning_rate": 7.395998326165165e-07,
+      "loss": 0.5864,
+      "step": 37374
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5482431610953609,
+      "learning_rate": 7.393118685435319e-07,
+      "loss": 0.7694,
+      "step": 37375
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4882059016714946,
+      "learning_rate": 7.390239583892045e-07,
+      "loss": 0.657,
+      "step": 37376
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6582925704052076,
+      "learning_rate": 7.38736102155212e-07,
+      "loss": 0.4935,
+      "step": 37377
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5619513141530488,
+      "learning_rate": 7.384482998432296e-07,
+      "loss": 0.5855,
+      "step": 37378
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3689860553833832,
+      "learning_rate": 7.381605514549339e-07,
+      "loss": 0.5262,
+      "step": 37379
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3634465131168971,
+      "learning_rate": 7.378728569919991e-07,
+      "loss": 0.6632,
+      "step": 37380
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4155201291845887,
+      "learning_rate": 7.375852164561026e-07,
+      "loss": 0.6476,
+      "step": 37381
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2764107727093266,
+      "learning_rate": 7.372976298489154e-07,
+      "loss": 0.6462,
+      "step": 37382
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.681956711209866,
+      "learning_rate": 7.370100971721151e-07,
+      "loss": 0.7549,
+      "step": 37383
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5631121535720924,
+      "learning_rate": 7.367226184273746e-07,
+      "loss": 0.6455,
+      "step": 37384
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3758984211723582,
+      "learning_rate": 7.364351936163672e-07,
+      "loss": 0.8248,
+      "step": 37385
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2017006052344872,
+      "learning_rate": 7.361478227407659e-07,
+      "loss": 0.676,
+      "step": 37386
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6625248779466819,
+      "learning_rate": 7.358605058022461e-07,
+      "loss": 0.5151,
+      "step": 37387
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6042521613515587,
+      "learning_rate": 7.355732428024808e-07,
+      "loss": 0.4813,
+      "step": 37388
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6169489312401504,
+      "learning_rate": 7.352860337431389e-07,
+      "loss": 0.5393,
+      "step": 37389
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4916206322968892,
+      "learning_rate": 7.349988786258944e-07,
+      "loss": 0.7077,
+      "step": 37390
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5583741002203002,
+      "learning_rate": 7.347117774524215e-07,
+      "loss": 0.6225,
+      "step": 37391
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2235508628236968,
+      "learning_rate": 7.344247302243901e-07,
+      "loss": 0.7258,
+      "step": 37392
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2607525127273962,
+      "learning_rate": 7.3413773694347e-07,
+      "loss": 0.6156,
+      "step": 37393
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5833426881385797,
+      "learning_rate": 7.338507976113329e-07,
+      "loss": 0.4875,
+      "step": 37394
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4927978821617278,
+      "learning_rate": 7.335639122296522e-07,
+      "loss": 0.6745,
+      "step": 37395
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5203025686679144,
+      "learning_rate": 7.332770808000955e-07,
+      "loss": 0.7396,
+      "step": 37396
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3160542375299153,
+      "learning_rate": 7.329903033243324e-07,
+      "loss": 0.6825,
+      "step": 37397
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2966582920468086,
+      "learning_rate": 7.32703579804036e-07,
+      "loss": 0.7128,
+      "step": 37398
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4734234780579256,
+      "learning_rate": 7.324169102408707e-07,
+      "loss": 0.6712,
+      "step": 37399
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4381853053971139,
+      "learning_rate": 7.321302946365094e-07,
+      "loss": 0.7067,
+      "step": 37400
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5341169070447713,
+      "learning_rate": 7.318437329926187e-07,
+      "loss": 0.6396,
+      "step": 37401
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.60052881457416,
+      "learning_rate": 7.315572253108694e-07,
+      "loss": 0.5062,
+      "step": 37402
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2901570336712294,
+      "learning_rate": 7.312707715929268e-07,
+      "loss": 0.7836,
+      "step": 37403
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5212371816403236,
+      "learning_rate": 7.309843718404619e-07,
+      "loss": 0.6555,
+      "step": 37404
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.372584825156241,
+      "learning_rate": 7.306980260551411e-07,
+      "loss": 0.7034,
+      "step": 37405
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.371662597695851,
+      "learning_rate": 7.304117342386286e-07,
+      "loss": 0.6659,
+      "step": 37406
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2881821837692968,
+      "learning_rate": 7.301254963925941e-07,
+      "loss": 0.7121,
+      "step": 37407
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2436262870152153,
+      "learning_rate": 7.298393125187054e-07,
+      "loss": 0.6197,
+      "step": 37408
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.289229298338454,
+      "learning_rate": 7.295531826186264e-07,
+      "loss": 0.7153,
+      "step": 37409
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4561571971045768,
+      "learning_rate": 7.292671066940226e-07,
+      "loss": 0.577,
+      "step": 37410
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3232328477271231,
+      "learning_rate": 7.289810847465606e-07,
+      "loss": 0.6009,
+      "step": 37411
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.9400336600335844,
+      "learning_rate": 7.286951167779077e-07,
+      "loss": 0.6988,
+      "step": 37412
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2804517208674806,
+      "learning_rate": 7.284092027897272e-07,
+      "loss": 0.5723,
+      "step": 37413
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3334868787003964,
+      "learning_rate": 7.28123342783682e-07,
+      "loss": 0.8149,
+      "step": 37414
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4109334279102397,
+      "learning_rate": 7.278375367614377e-07,
+      "loss": 0.6725,
+      "step": 37415
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3792589875949173,
+      "learning_rate": 7.275517847246604e-07,
+      "loss": 0.7247,
+      "step": 37416
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3419878228848845,
+      "learning_rate": 7.272660866750125e-07,
+      "loss": 0.6496,
+      "step": 37417
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6475709883421106,
+      "learning_rate": 7.269804426141558e-07,
+      "loss": 0.799,
+      "step": 37418
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.255085742845158,
+      "learning_rate": 7.266948525437556e-07,
+      "loss": 0.6319,
+      "step": 37419
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4454448846729773,
+      "learning_rate": 7.264093164654728e-07,
+      "loss": 0.6886,
+      "step": 37420
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5952701268116303,
+      "learning_rate": 7.261238343809728e-07,
+      "loss": 0.6259,
+      "step": 37421
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.426674693910767,
+      "learning_rate": 7.258384062919144e-07,
+      "loss": 0.6434,
+      "step": 37422
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6157184216757838,
+      "learning_rate": 7.255530321999616e-07,
+      "loss": 0.6529,
+      "step": 37423
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3084152192916523,
+      "learning_rate": 7.252677121067742e-07,
+      "loss": 0.596,
+      "step": 37424
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2725542516112303,
+      "learning_rate": 7.249824460140165e-07,
+      "loss": 0.6147,
+      "step": 37425
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.36374895228702,
+      "learning_rate": 7.246972339233449e-07,
+      "loss": 0.6819,
+      "step": 37426
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4561602301086953,
+      "learning_rate": 7.244120758364248e-07,
+      "loss": 0.7048,
+      "step": 37427
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3013623995658337,
+      "learning_rate": 7.241269717549126e-07,
+      "loss": 0.7124,
+      "step": 37428
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5084034132034922,
+      "learning_rate": 7.238419216804693e-07,
+      "loss": 0.7019,
+      "step": 37429
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4368716076981514,
+      "learning_rate": 7.23556925614759e-07,
+      "loss": 0.6996,
+      "step": 37430
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4953115625858733,
+      "learning_rate": 7.232719835594337e-07,
+      "loss": 0.681,
+      "step": 37431
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6883027143456626,
+      "learning_rate": 7.229870955161555e-07,
+      "loss": 0.7457,
+      "step": 37432
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2750210768200363,
+      "learning_rate": 7.227022614865841e-07,
+      "loss": 0.6469,
+      "step": 37433
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.079239375766276,
+      "learning_rate": 7.224174814723794e-07,
+      "loss": 0.6854,
+      "step": 37434
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5837046748210958,
+      "learning_rate": 7.221327554751956e-07,
+      "loss": 0.4939,
+      "step": 37435
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7957717918009424,
+      "learning_rate": 7.218480834966912e-07,
+      "loss": 0.7293,
+      "step": 37436
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2246050040086816,
+      "learning_rate": 7.215634655385273e-07,
+      "loss": 0.6345,
+      "step": 37437
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3977392686072794,
+      "learning_rate": 7.212789016023569e-07,
+      "loss": 0.6452,
+      "step": 37438
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5999993624443604,
+      "learning_rate": 7.209943916898377e-07,
+      "loss": 0.4537,
+      "step": 37439
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3309126908387627,
+      "learning_rate": 7.20709935802627e-07,
+      "loss": 0.622,
+      "step": 37440
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3754120792029205,
+      "learning_rate": 7.204255339423815e-07,
+      "loss": 0.7326,
+      "step": 37441
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3455236883496287,
+      "learning_rate": 7.201411861107555e-07,
+      "loss": 0.6688,
+      "step": 37442
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.468219439537977,
+      "learning_rate": 7.198568923094074e-07,
+      "loss": 0.7174,
+      "step": 37443
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.381700636535973,
+      "learning_rate": 7.195726525399905e-07,
+      "loss": 0.7219,
+      "step": 37444
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6055582872303304,
+      "learning_rate": 7.192884668041578e-07,
+      "loss": 0.7184,
+      "step": 37445
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4450839722032105,
+      "learning_rate": 7.190043351035658e-07,
+      "loss": 0.6469,
+      "step": 37446
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2109214309377994,
+      "learning_rate": 7.187202574398699e-07,
+      "loss": 0.6015,
+      "step": 37447
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5908561001199246,
+      "learning_rate": 7.184362338147244e-07,
+      "loss": 0.5297,
+      "step": 37448
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6212805428520636,
+      "learning_rate": 7.181522642297789e-07,
+      "loss": 0.663,
+      "step": 37449
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.356554603376317,
+      "learning_rate": 7.178683486866899e-07,
+      "loss": 0.6901,
+      "step": 37450
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6410946330834348,
+      "learning_rate": 7.175844871871129e-07,
+      "loss": 0.5772,
+      "step": 37451
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9423382754827359,
+      "learning_rate": 7.173006797326954e-07,
+      "loss": 0.682,
+      "step": 37452
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.938569242125982,
+      "learning_rate": 7.170169263250914e-07,
+      "loss": 0.8785,
+      "step": 37453
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5631391703750009,
+      "learning_rate": 7.167332269659544e-07,
+      "loss": 0.4402,
+      "step": 37454
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6685953839423207,
+      "learning_rate": 7.164495816569373e-07,
+      "loss": 0.6328,
+      "step": 37455
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2134791552683386,
+      "learning_rate": 7.161659903996887e-07,
+      "loss": 0.7247,
+      "step": 37456
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4947754264750137,
+      "learning_rate": 7.158824531958608e-07,
+      "loss": 0.5218,
+      "step": 37457
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3550497188785604,
+      "learning_rate": 7.155989700471056e-07,
+      "loss": 0.5921,
+      "step": 37458
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6037580486673284,
+      "learning_rate": 7.153155409550705e-07,
+      "loss": 0.6913,
+      "step": 37459
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3611566433628968,
+      "learning_rate": 7.150321659214099e-07,
+      "loss": 0.6285,
+      "step": 37460
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5794625720564888,
+      "learning_rate": 7.14748844947769e-07,
+      "loss": 0.6867,
+      "step": 37461
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3481227053516982,
+      "learning_rate": 7.144655780358034e-07,
+      "loss": 0.625,
+      "step": 37462
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.49965881405082,
+      "learning_rate": 7.14182365187156e-07,
+      "loss": 0.6788,
+      "step": 37463
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0036533229542153,
+      "learning_rate": 7.138992064034811e-07,
+      "loss": 0.7082,
+      "step": 37464
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.513150113706453,
+      "learning_rate": 7.136161016864241e-07,
+      "loss": 0.632,
+      "step": 37465
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3428727264000273,
+      "learning_rate": 7.133330510376335e-07,
+      "loss": 0.6726,
+      "step": 37466
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3217891555785226,
+      "learning_rate": 7.13050054458757e-07,
+      "loss": 0.683,
+      "step": 37467
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6512173611152536,
+      "learning_rate": 7.127671119514456e-07,
+      "loss": 0.632,
+      "step": 37468
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5751968630976265,
+      "learning_rate": 7.124842235173445e-07,
+      "loss": 0.6206,
+      "step": 37469
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.474065719550738,
+      "learning_rate": 7.122013891580981e-07,
+      "loss": 0.616,
+      "step": 37470
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2366761571593754,
+      "learning_rate": 7.11918608875356e-07,
+      "loss": 0.7499,
+      "step": 37471
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2757740179513983,
+      "learning_rate": 7.11635882670767e-07,
+      "loss": 0.6013,
+      "step": 37472
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.477578333018352,
+      "learning_rate": 7.113532105459731e-07,
+      "loss": 0.647,
+      "step": 37473
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.46769392677548,
+      "learning_rate": 7.110705925026207e-07,
+      "loss": 0.5529,
+      "step": 37474
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5846320093173385,
+      "learning_rate": 7.107880285423563e-07,
+      "loss": 0.5339,
+      "step": 37475
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.746755505094776,
+      "learning_rate": 7.105055186668264e-07,
+      "loss": 0.7139,
+      "step": 37476
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2653968843784522,
+      "learning_rate": 7.10223062877674e-07,
+      "loss": 0.6218,
+      "step": 37477
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4385723969705695,
+      "learning_rate": 7.099406611765425e-07,
+      "loss": 0.628,
+      "step": 37478
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2726069554012758,
+      "learning_rate": 7.096583135650792e-07,
+      "loss": 0.7248,
+      "step": 37479
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2859518657422504,
+      "learning_rate": 7.093760200449251e-07,
+      "loss": 0.6707,
+      "step": 37480
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4615866015850791,
+      "learning_rate": 7.090937806177267e-07,
+      "loss": 0.6808,
+      "step": 37481
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.194079018993413,
+      "learning_rate": 7.088115952851238e-07,
+      "loss": 0.6117,
+      "step": 37482
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.355442605510246,
+      "learning_rate": 7.085294640487639e-07,
+      "loss": 0.5664,
+      "step": 37483
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4240075961928225,
+      "learning_rate": 7.082473869102846e-07,
+      "loss": 0.6742,
+      "step": 37484
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4105723836458697,
+      "learning_rate": 7.079653638713324e-07,
+      "loss": 0.6198,
+      "step": 37485
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2963198151988373,
+      "learning_rate": 7.076833949335482e-07,
+      "loss": 0.6551,
+      "step": 37486
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5316781041913803,
+      "learning_rate": 7.074014800985718e-07,
+      "loss": 0.749,
+      "step": 37487
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.500462977547732,
+      "learning_rate": 7.071196193680452e-07,
+      "loss": 0.6005,
+      "step": 37488
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3717435610838569,
+      "learning_rate": 7.068378127436115e-07,
+      "loss": 0.6266,
+      "step": 37489
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.34704495260571,
+      "learning_rate": 7.065560602269106e-07,
+      "loss": 0.7384,
+      "step": 37490
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4588539767612145,
+      "learning_rate": 7.062743618195822e-07,
+      "loss": 0.6579,
+      "step": 37491
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.442739231630489,
+      "learning_rate": 7.05992717523265e-07,
+      "loss": 0.7372,
+      "step": 37492
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3431175800522397,
+      "learning_rate": 7.057111273396034e-07,
+      "loss": 0.6595,
+      "step": 37493
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.652842564070314,
+      "learning_rate": 7.054295912702336e-07,
+      "loss": 0.6834,
+      "step": 37494
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3189031358632142,
+      "learning_rate": 7.051481093167933e-07,
+      "loss": 0.671,
+      "step": 37495
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3670597526140498,
+      "learning_rate": 7.048666814809246e-07,
+      "loss": 0.704,
+      "step": 37496
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6039483121691422,
+      "learning_rate": 7.045853077642661e-07,
+      "loss": 0.5222,
+      "step": 37497
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.570625429809171,
+      "learning_rate": 7.043039881684544e-07,
+      "loss": 0.7043,
+      "step": 37498
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.355841283007711,
+      "learning_rate": 7.040227226951268e-07,
+      "loss": 0.6354,
+      "step": 37499
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3681197923802662,
+      "learning_rate": 7.037415113459223e-07,
+      "loss": 0.6297,
+      "step": 37500
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.250384748001731,
+      "learning_rate": 7.034603541224771e-07,
+      "loss": 0.5428,
+      "step": 37501
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5938485387801347,
+      "learning_rate": 7.031792510264312e-07,
+      "loss": 0.7261,
+      "step": 37502
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3403156697294307,
+      "learning_rate": 7.028982020594166e-07,
+      "loss": 0.7328,
+      "step": 37503
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5833461646404052,
+      "learning_rate": 7.026172072230742e-07,
+      "loss": 0.7027,
+      "step": 37504
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6327011276475406,
+      "learning_rate": 7.023362665190358e-07,
+      "loss": 0.4849,
+      "step": 37505
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5869696962618353,
+      "learning_rate": 7.020553799489415e-07,
+      "loss": 0.499,
+      "step": 37506
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6340822341782804,
+      "learning_rate": 7.017745475144233e-07,
+      "loss": 0.5622,
+      "step": 37507
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.359294454236216,
+      "learning_rate": 7.014937692171176e-07,
+      "loss": 0.703,
+      "step": 37508
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.25137908671046,
+      "learning_rate": 7.012130450586574e-07,
+      "loss": 0.6083,
+      "step": 37509
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5553894890136593,
+      "learning_rate": 7.009323750406816e-07,
+      "loss": 0.6657,
+      "step": 37510
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5587029051008505,
+      "learning_rate": 7.00651759164821e-07,
+      "loss": 0.713,
+      "step": 37511
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.549030342973927,
+      "learning_rate": 7.003711974327076e-07,
+      "loss": 0.755,
+      "step": 37512
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3962723704971458,
+      "learning_rate": 7.00090689845978e-07,
+      "loss": 0.6208,
+      "step": 37513
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.368471686735039,
+      "learning_rate": 6.998102364062665e-07,
+      "loss": 0.7056,
+      "step": 37514
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2877759841288103,
+      "learning_rate": 6.995298371152037e-07,
+      "loss": 0.6274,
+      "step": 37515
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5182696932021629,
+      "learning_rate": 6.992494919744219e-07,
+      "loss": 0.6856,
+      "step": 37516
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4646294750413824,
+      "learning_rate": 6.989692009855543e-07,
+      "loss": 0.6706,
+      "step": 37517
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9185489280485937,
+      "learning_rate": 6.986889641502326e-07,
+      "loss": 0.7543,
+      "step": 37518
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5132457261813723,
+      "learning_rate": 6.984087814700902e-07,
+      "loss": 0.7219,
+      "step": 37519
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4177796155839233,
+      "learning_rate": 6.981286529467546e-07,
+      "loss": 0.6548,
+      "step": 37520
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.636509548158092,
+      "learning_rate": 6.9784857858186e-07,
+      "loss": 0.4829,
+      "step": 37521
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.380267732274065,
+      "learning_rate": 6.975685583770341e-07,
+      "loss": 0.656,
+      "step": 37522
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5020137772022304,
+      "learning_rate": 6.972885923339123e-07,
+      "loss": 0.5777,
+      "step": 37523
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3599402272783299,
+      "learning_rate": 6.970086804541187e-07,
+      "loss": 0.687,
+      "step": 37524
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4116406609162266,
+      "learning_rate": 6.967288227392877e-07,
+      "loss": 0.5895,
+      "step": 37525
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3151235407936663,
+      "learning_rate": 6.964490191910456e-07,
+      "loss": 0.6592,
+      "step": 37526
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5056437284117077,
+      "learning_rate": 6.961692698110223e-07,
+      "loss": 0.624,
+      "step": 37527
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5734464999663104,
+      "learning_rate": 6.958895746008499e-07,
+      "loss": 0.4875,
+      "step": 37528
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6377235874714056,
+      "learning_rate": 6.956099335621513e-07,
+      "loss": 0.5818,
+      "step": 37529
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8289699069494205,
+      "learning_rate": 6.953303466965578e-07,
+      "loss": 0.7461,
+      "step": 37530
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3788156843744417,
+      "learning_rate": 6.950508140056966e-07,
+      "loss": 0.7064,
+      "step": 37531
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6193352901295175,
+      "learning_rate": 6.947713354911978e-07,
+      "loss": 0.731,
+      "step": 37532
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9103521501830596,
+      "learning_rate": 6.944919111546844e-07,
+      "loss": 0.55,
+      "step": 37533
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3187759827734566,
+      "learning_rate": 6.94212540997784e-07,
+      "loss": 0.6185,
+      "step": 37534
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.758556876301796,
+      "learning_rate": 6.939332250221265e-07,
+      "loss": 0.6682,
+      "step": 37535
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.324251070971447,
+      "learning_rate": 6.93653963229336e-07,
+      "loss": 0.6261,
+      "step": 37536
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.441305643015385,
+      "learning_rate": 6.933747556210368e-07,
+      "loss": 0.5574,
+      "step": 37537
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4251029184310184,
+      "learning_rate": 6.930956021988567e-07,
+      "loss": 0.7079,
+      "step": 37538
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2552909018793257,
+      "learning_rate": 6.928165029644219e-07,
+      "loss": 0.5317,
+      "step": 37539
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3962789275921543,
+      "learning_rate": 6.925374579193533e-07,
+      "loss": 0.776,
+      "step": 37540
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4167832024251346,
+      "learning_rate": 6.92258467065281e-07,
+      "loss": 0.7488,
+      "step": 37541
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5837510222049825,
+      "learning_rate": 6.919795304038268e-07,
+      "loss": 0.4706,
+      "step": 37542
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.700274839034072,
+      "learning_rate": 6.917006479366128e-07,
+      "loss": 0.6397,
+      "step": 37543
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6472228976513128,
+      "learning_rate": 6.914218196652645e-07,
+      "loss": 0.7079,
+      "step": 37544
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3646085391410359,
+      "learning_rate": 6.91143045591407e-07,
+      "loss": 0.6505,
+      "step": 37545
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2823082585892065,
+      "learning_rate": 6.908643257166625e-07,
+      "loss": 0.6343,
+      "step": 37546
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3787802519284282,
+      "learning_rate": 6.905856600426508e-07,
+      "loss": 0.7484,
+      "step": 37547
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1423273260237838,
+      "learning_rate": 6.903070485709962e-07,
+      "loss": 0.4569,
+      "step": 37548
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3899980259652087,
+      "learning_rate": 6.90028491303324e-07,
+      "loss": 0.5925,
+      "step": 37549
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5572600413106688,
+      "learning_rate": 6.897499882412529e-07,
+      "loss": 0.6638,
+      "step": 37550
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6055320918113563,
+      "learning_rate": 6.894715393864027e-07,
+      "loss": 0.4988,
+      "step": 37551
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2896984012846846,
+      "learning_rate": 6.891931447403977e-07,
+      "loss": 0.646,
+      "step": 37552
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3188764104000084,
+      "learning_rate": 6.889148043048599e-07,
+      "loss": 0.6203,
+      "step": 37553
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3182426928944277,
+      "learning_rate": 6.886365180814069e-07,
+      "loss": 0.5815,
+      "step": 37554
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5373558386615043,
+      "learning_rate": 6.883582860716598e-07,
+      "loss": 0.7251,
+      "step": 37555
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2489865334892998,
+      "learning_rate": 6.880801082772382e-07,
+      "loss": 0.7088,
+      "step": 37556
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3256766966513893,
+      "learning_rate": 6.87801984699763e-07,
+      "loss": 0.7024,
+      "step": 37557
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3060283735972584,
+      "learning_rate": 6.875239153408541e-07,
+      "loss": 0.6528,
+      "step": 37558
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4549665259442055,
+      "learning_rate": 6.87245900202127e-07,
+      "loss": 0.5678,
+      "step": 37559
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.615990361042453,
+      "learning_rate": 6.869679392852035e-07,
+      "loss": 0.7015,
+      "step": 37560
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6069608384911608,
+      "learning_rate": 6.866900325917003e-07,
+      "loss": 0.4697,
+      "step": 37561
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5715284036983381,
+      "learning_rate": 6.864121801232381e-07,
+      "loss": 0.6787,
+      "step": 37562
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4530590264251717,
+      "learning_rate": 6.861343818814315e-07,
+      "loss": 0.7956,
+      "step": 37563
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5773931564403207,
+      "learning_rate": 6.858566378678999e-07,
+      "loss": 0.5976,
+      "step": 37564
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5885525011001513,
+      "learning_rate": 6.855789480842589e-07,
+      "loss": 0.7531,
+      "step": 37565
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3878887687502464,
+      "learning_rate": 6.853013125321273e-07,
+      "loss": 0.7256,
+      "step": 37566
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.230189498650914,
+      "learning_rate": 6.850237312131213e-07,
+      "loss": 0.6382,
+      "step": 37567
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2370701747815682,
+      "learning_rate": 6.847462041288533e-07,
+      "loss": 0.6208,
+      "step": 37568
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4034291650146011,
+      "learning_rate": 6.844687312809429e-07,
+      "loss": 0.6677,
+      "step": 37569
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5827298589045383,
+      "learning_rate": 6.841913126710065e-07,
+      "loss": 0.4966,
+      "step": 37570
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3141027817253155,
+      "learning_rate": 6.839139483006574e-07,
+      "loss": 0.5979,
+      "step": 37571
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3810892049571628,
+      "learning_rate": 6.836366381715087e-07,
+      "loss": 0.7773,
+      "step": 37572
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.612021271646061,
+      "learning_rate": 6.83359382285177e-07,
+      "loss": 0.5166,
+      "step": 37573
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3697504305751336,
+      "learning_rate": 6.830821806432786e-07,
+      "loss": 0.6241,
+      "step": 37574
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6160251802714126,
+      "learning_rate": 6.828050332474256e-07,
+      "loss": 0.5235,
+      "step": 37575
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5948383661693831,
+      "learning_rate": 6.825279400992291e-07,
+      "loss": 0.4676,
+      "step": 37576
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5489946654229283,
+      "learning_rate": 6.822509012003054e-07,
+      "loss": 0.6382,
+      "step": 37577
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3774797503815717,
+      "learning_rate": 6.819739165522688e-07,
+      "loss": 0.6656,
+      "step": 37578
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2917493034272687,
+      "learning_rate": 6.816969861567291e-07,
+      "loss": 0.7786,
+      "step": 37579
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.582040569257864,
+      "learning_rate": 6.814201100152995e-07,
+      "loss": 0.508,
+      "step": 37580
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3768474424587387,
+      "learning_rate": 6.811432881295921e-07,
+      "loss": 0.6462,
+      "step": 37581
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.348252415744311,
+      "learning_rate": 6.808665205012188e-07,
+      "loss": 0.7096,
+      "step": 37582
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4735304870282786,
+      "learning_rate": 6.805898071317918e-07,
+      "loss": 0.7213,
+      "step": 37583
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.607171777476274,
+      "learning_rate": 6.803131480229197e-07,
+      "loss": 0.5195,
+      "step": 37584
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2725622222881556,
+      "learning_rate": 6.800365431762168e-07,
+      "loss": 0.7512,
+      "step": 37585
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.42695195507766,
+      "learning_rate": 6.797599925932907e-07,
+      "loss": 0.759,
+      "step": 37586
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3610774582120506,
+      "learning_rate": 6.794834962757535e-07,
+      "loss": 0.5583,
+      "step": 37587
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3832180204759066,
+      "learning_rate": 6.792070542252138e-07,
+      "loss": 0.6231,
+      "step": 37588
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3276463467669128,
+      "learning_rate": 6.789306664432815e-07,
+      "loss": 0.7253,
+      "step": 37589
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.610301278880571,
+      "learning_rate": 6.786543329315643e-07,
+      "loss": 0.6973,
+      "step": 37590
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.386368355180497,
+      "learning_rate": 6.783780536916751e-07,
+      "loss": 0.619,
+      "step": 37591
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.202824530959109,
+      "learning_rate": 6.781018287252206e-07,
+      "loss": 0.6298,
+      "step": 37592
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5128287902125837,
+      "learning_rate": 6.77825658033806e-07,
+      "loss": 0.7253,
+      "step": 37593
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3553854074118472,
+      "learning_rate": 6.775495416190414e-07,
+      "loss": 0.6706,
+      "step": 37594
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5449439428216136,
+      "learning_rate": 6.772734794825375e-07,
+      "loss": 0.6781,
+      "step": 37595
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3719436766412465,
+      "learning_rate": 6.769974716258987e-07,
+      "loss": 0.6252,
+      "step": 37596
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.346203342032608,
+      "learning_rate": 6.767215180507303e-07,
+      "loss": 0.7119,
+      "step": 37597
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3644503247304305,
+      "learning_rate": 6.76445618758641e-07,
+      "loss": 0.6681,
+      "step": 37598
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3965681643833223,
+      "learning_rate": 6.761697737512385e-07,
+      "loss": 0.5605,
+      "step": 37599
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.295381013014086,
+      "learning_rate": 6.75893983030127e-07,
+      "loss": 0.7158,
+      "step": 37600
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.45840165667131,
+      "learning_rate": 6.756182465969119e-07,
+      "loss": 0.6852,
+      "step": 37601
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4555462828550072,
+      "learning_rate": 6.753425644532008e-07,
+      "loss": 0.7826,
+      "step": 37602
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6016130727176722,
+      "learning_rate": 6.750669366005957e-07,
+      "loss": 0.4886,
+      "step": 37603
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.80388481300405,
+      "learning_rate": 6.747913630407054e-07,
+      "loss": 0.7165,
+      "step": 37604
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5892922685120141,
+      "learning_rate": 6.745158437751298e-07,
+      "loss": 0.5259,
+      "step": 37605
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4063650057397277,
+      "learning_rate": 6.742403788054775e-07,
+      "loss": 0.7859,
+      "step": 37606
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2711150639646485,
+      "learning_rate": 6.739649681333482e-07,
+      "loss": 0.7728,
+      "step": 37607
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.257056651035571,
+      "learning_rate": 6.736896117603487e-07,
+      "loss": 0.6783,
+      "step": 37608
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6344521771410523,
+      "learning_rate": 6.734143096880808e-07,
+      "loss": 0.5216,
+      "step": 37609
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3819063174198538,
+      "learning_rate": 6.731390619181466e-07,
+      "loss": 0.6795,
+      "step": 37610
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.344969299661552,
+      "learning_rate": 6.728638684521493e-07,
+      "loss": 0.7103,
+      "step": 37611
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7545131960036708,
+      "learning_rate": 6.725887292916933e-07,
+      "loss": 0.639,
+      "step": 37612
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4938787285679154,
+      "learning_rate": 6.723136444383793e-07,
+      "loss": 0.7915,
+      "step": 37613
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.266910741793456,
+      "learning_rate": 6.720386138938062e-07,
+      "loss": 0.6348,
+      "step": 37614
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2505611253603757,
+      "learning_rate": 6.717636376595771e-07,
+      "loss": 0.5891,
+      "step": 37615
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2245701428344133,
+      "learning_rate": 6.714887157372952e-07,
+      "loss": 0.6775,
+      "step": 37616
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6260044117788256,
+      "learning_rate": 6.71213848128559e-07,
+      "loss": 0.4987,
+      "step": 37617
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4786323882408234,
+      "learning_rate": 6.709390348349687e-07,
+      "loss": 0.6972,
+      "step": 37618
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.529925275520524,
+      "learning_rate": 6.706642758581239e-07,
+      "loss": 0.6175,
+      "step": 37619
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4933577583369502,
+      "learning_rate": 6.703895711996278e-07,
+      "loss": 0.5429,
+      "step": 37620
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4291290547447746,
+      "learning_rate": 6.701149208610758e-07,
+      "loss": 0.5155,
+      "step": 37621
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.233911662248179,
+      "learning_rate": 6.698403248440688e-07,
+      "loss": 0.6174,
+      "step": 37622
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2900279819613574,
+      "learning_rate": 6.695657831502056e-07,
+      "loss": 0.7673,
+      "step": 37623
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3470028284859104,
+      "learning_rate": 6.692912957810826e-07,
+      "loss": 0.7341,
+      "step": 37624
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4443269285361306,
+      "learning_rate": 6.69016862738301e-07,
+      "loss": 0.6825,
+      "step": 37625
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3689256783272852,
+      "learning_rate": 6.687424840234579e-07,
+      "loss": 0.5993,
+      "step": 37626
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4779237437120045,
+      "learning_rate": 6.684681596381503e-07,
+      "loss": 0.6618,
+      "step": 37627
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.645128510814325,
+      "learning_rate": 6.681938895839746e-07,
+      "loss": 0.5545,
+      "step": 37628
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4477876518857442,
+      "learning_rate": 6.679196738625282e-07,
+      "loss": 0.5968,
+      "step": 37629
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4878770431469106,
+      "learning_rate": 6.676455124754099e-07,
+      "loss": 0.6033,
+      "step": 37630
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.731077752385332,
+      "learning_rate": 6.67371405424212e-07,
+      "loss": 0.7111,
+      "step": 37631
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.103325331718734,
+      "learning_rate": 6.670973527105329e-07,
+      "loss": 0.6725,
+      "step": 37632
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5145482076476928,
+      "learning_rate": 6.66823354335967e-07,
+      "loss": 0.606,
+      "step": 37633
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5870788413200225,
+      "learning_rate": 6.66549410302113e-07,
+      "loss": 0.7078,
+      "step": 37634
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3886640654701772,
+      "learning_rate": 6.662755206105608e-07,
+      "loss": 0.5713,
+      "step": 37635
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2869502332135192,
+      "learning_rate": 6.66001685262908e-07,
+      "loss": 0.649,
+      "step": 37636
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.311821538820636,
+      "learning_rate": 6.65727904260749e-07,
+      "loss": 0.7594,
+      "step": 37637
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5702622185996946,
+      "learning_rate": 6.654541776056767e-07,
+      "loss": 0.6428,
+      "step": 37638
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3563363902936407,
+      "learning_rate": 6.651805052992866e-07,
+      "loss": 0.6629,
+      "step": 37639
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5138716696196193,
+      "learning_rate": 6.649068873431697e-07,
+      "loss": 0.6322,
+      "step": 37640
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5860951163720027,
+      "learning_rate": 6.646333237389213e-07,
+      "loss": 0.4992,
+      "step": 37641
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6184563309792627,
+      "learning_rate": 6.643598144881324e-07,
+      "loss": 0.5273,
+      "step": 37642
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2468434159035235,
+      "learning_rate": 6.640863595923975e-07,
+      "loss": 0.7298,
+      "step": 37643
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5131991877564757,
+      "learning_rate": 6.638129590533071e-07,
+      "loss": 0.7259,
+      "step": 37644
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5765199917557224,
+      "learning_rate": 6.635396128724513e-07,
+      "loss": 0.7161,
+      "step": 37645
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3013171798794838,
+      "learning_rate": 6.632663210514257e-07,
+      "loss": 0.6308,
+      "step": 37646
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2747932365342474,
+      "learning_rate": 6.629930835918196e-07,
+      "loss": 0.8697,
+      "step": 37647
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2677751101855763,
+      "learning_rate": 6.627199004952245e-07,
+      "loss": 0.6199,
+      "step": 37648
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.580956408231674,
+      "learning_rate": 6.624467717632277e-07,
+      "loss": 0.6694,
+      "step": 37649
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4151957458804705,
+      "learning_rate": 6.621736973974236e-07,
+      "loss": 0.607,
+      "step": 37650
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7129051954871424,
+      "learning_rate": 6.619006773994008e-07,
+      "loss": 0.5919,
+      "step": 37651
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5885576456582071,
+      "learning_rate": 6.616277117707493e-07,
+      "loss": 0.6488,
+      "step": 37652
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5588049179483057,
+      "learning_rate": 6.613548005130555e-07,
+      "loss": 0.7556,
+      "step": 37653
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2430221574671776,
+      "learning_rate": 6.610819436279115e-07,
+      "loss": 0.6512,
+      "step": 37654
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9063999692575269,
+      "learning_rate": 6.60809141116906e-07,
+      "loss": 0.6477,
+      "step": 37655
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4903630856525316,
+      "learning_rate": 6.605363929816266e-07,
+      "loss": 0.7708,
+      "step": 37656
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.555306563022893,
+      "learning_rate": 6.6026369922366e-07,
+      "loss": 0.7066,
+      "step": 37657
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3965076027681562,
+      "learning_rate": 6.599910598445958e-07,
+      "loss": 0.6396,
+      "step": 37658
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5783884711462526,
+      "learning_rate": 6.597184748460195e-07,
+      "loss": 0.5126,
+      "step": 37659
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6212327649592556,
+      "learning_rate": 6.594459442295209e-07,
+      "loss": 0.5176,
+      "step": 37660
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3617522893877354,
+      "learning_rate": 6.591734679966844e-07,
+      "loss": 0.669,
+      "step": 37661
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.467057134015141,
+      "learning_rate": 6.589010461490974e-07,
+      "loss": 0.6962,
+      "step": 37662
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6369912794355368,
+      "learning_rate": 6.586286786883455e-07,
+      "loss": 0.5814,
+      "step": 37663
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3155094989726355,
+      "learning_rate": 6.583563656160152e-07,
+      "loss": 0.6306,
+      "step": 37664
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5347611621727237,
+      "learning_rate": 6.580841069336929e-07,
+      "loss": 0.6341,
+      "step": 37665
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5582303779313211,
+      "learning_rate": 6.578119026429597e-07,
+      "loss": 0.671,
+      "step": 37666
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6292202957617691,
+      "learning_rate": 6.575397527454042e-07,
+      "loss": 0.5625,
+      "step": 37667
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3801247383989304,
+      "learning_rate": 6.572676572426118e-07,
+      "loss": 0.6145,
+      "step": 37668
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2642268555988108,
+      "learning_rate": 6.569956161361646e-07,
+      "loss": 0.5414,
+      "step": 37669
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.653670463567309,
+      "learning_rate": 6.567236294276446e-07,
+      "loss": 0.5478,
+      "step": 37670
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4013698804020345,
+      "learning_rate": 6.564516971186385e-07,
+      "loss": 0.7168,
+      "step": 37671
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4773250718740705,
+      "learning_rate": 6.561798192107305e-07,
+      "loss": 0.7119,
+      "step": 37672
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6706854564258924,
+      "learning_rate": 6.559079957055015e-07,
+      "loss": 0.6775,
+      "step": 37673
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5270340798232298,
+      "learning_rate": 6.556362266045324e-07,
+      "loss": 0.728,
+      "step": 37674
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7128529304498665,
+      "learning_rate": 6.553645119094076e-07,
+      "loss": 0.6202,
+      "step": 37675
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2747946063305495,
+      "learning_rate": 6.550928516217115e-07,
+      "loss": 0.6723,
+      "step": 37676
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3685333727266533,
+      "learning_rate": 6.548212457430225e-07,
+      "loss": 0.6448,
+      "step": 37677
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2716086425860225,
+      "learning_rate": 6.545496942749219e-07,
+      "loss": 0.6236,
+      "step": 37678
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3369290083192318,
+      "learning_rate": 6.542781972189915e-07,
+      "loss": 0.6511,
+      "step": 37679
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4831323782945636,
+      "learning_rate": 6.540067545768136e-07,
+      "loss": 0.7529,
+      "step": 37680
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.22669831671312,
+      "learning_rate": 6.537353663499669e-07,
+      "loss": 0.5637,
+      "step": 37681
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5380903990241843,
+      "learning_rate": 6.53464032540031e-07,
+      "loss": 0.7016,
+      "step": 37682
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3305635218632783,
+      "learning_rate": 6.531927531485882e-07,
+      "loss": 0.711,
+      "step": 37683
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2723375405931698,
+      "learning_rate": 6.529215281772139e-07,
+      "loss": 0.6366,
+      "step": 37684
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2783913024952769,
+      "learning_rate": 6.526503576274923e-07,
+      "loss": 0.6593,
+      "step": 37685
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6715152896822463,
+      "learning_rate": 6.523792415009977e-07,
+      "loss": 0.7176,
+      "step": 37686
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3238366335787881,
+      "learning_rate": 6.521081797993123e-07,
+      "loss": 0.7462,
+      "step": 37687
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5057486490685115,
+      "learning_rate": 6.518371725240103e-07,
+      "loss": 0.7653,
+      "step": 37688
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2833401699552294,
+      "learning_rate": 6.515662196766737e-07,
+      "loss": 0.6752,
+      "step": 37689
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5101115695573892,
+      "learning_rate": 6.512953212588791e-07,
+      "loss": 0.7016,
+      "step": 37690
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4025760361084991,
+      "learning_rate": 6.510244772721996e-07,
+      "loss": 0.6508,
+      "step": 37691
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6420622481272755,
+      "learning_rate": 6.507536877182175e-07,
+      "loss": 0.8125,
+      "step": 37692
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8722330850788884,
+      "learning_rate": 6.504829525985079e-07,
+      "loss": 0.6472,
+      "step": 37693
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4093231307350687,
+      "learning_rate": 6.502122719146464e-07,
+      "loss": 0.6278,
+      "step": 37694
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3830263052527605,
+      "learning_rate": 6.499416456682072e-07,
+      "loss": 0.6572,
+      "step": 37695
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5921860137454069,
+      "learning_rate": 6.496710738607681e-07,
+      "loss": 0.4653,
+      "step": 37696
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2302337604814664,
+      "learning_rate": 6.494005564939066e-07,
+      "loss": 0.6166,
+      "step": 37697
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2705650502417203,
+      "learning_rate": 6.491300935691947e-07,
+      "loss": 0.5617,
+      "step": 37698
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.455893560613953,
+      "learning_rate": 6.488596850882068e-07,
+      "loss": 0.6033,
+      "step": 37699
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6764178464284154,
+      "learning_rate": 6.485893310525182e-07,
+      "loss": 0.6342,
+      "step": 37700
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6070913394744253,
+      "learning_rate": 6.483190314637045e-07,
+      "loss": 0.4811,
+      "step": 37701
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4093121748692492,
+      "learning_rate": 6.480487863233387e-07,
+      "loss": 0.7725,
+      "step": 37702
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1553711990508626,
+      "learning_rate": 6.477785956329907e-07,
+      "loss": 0.565,
+      "step": 37703
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4787621699624294,
+      "learning_rate": 6.475084593942393e-07,
+      "loss": 0.7612,
+      "step": 37704
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3909294009023436,
+      "learning_rate": 6.472383776086533e-07,
+      "loss": 0.6321,
+      "step": 37705
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5890116243106298,
+      "learning_rate": 6.469683502778068e-07,
+      "loss": 0.8754,
+      "step": 37706
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4199088761984633,
+      "learning_rate": 6.466983774032709e-07,
+      "loss": 0.6269,
+      "step": 37707
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.550875653081487,
+      "learning_rate": 6.464284589866199e-07,
+      "loss": 0.6461,
+      "step": 37708
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3925697802986055,
+      "learning_rate": 6.461585950294214e-07,
+      "loss": 0.6346,
+      "step": 37709
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6725932674389523,
+      "learning_rate": 6.458887855332518e-07,
+      "loss": 0.7795,
+      "step": 37710
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3464988486107192,
+      "learning_rate": 6.456190304996779e-07,
+      "loss": 0.6995,
+      "step": 37711
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3829746267263885,
+      "learning_rate": 6.453493299302704e-07,
+      "loss": 0.6857,
+      "step": 37712
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5452292904847196,
+      "learning_rate": 6.450796838266015e-07,
+      "loss": 0.658,
+      "step": 37713
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3378487999083917,
+      "learning_rate": 6.44810092190239e-07,
+      "loss": 0.5445,
+      "step": 37714
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.25636129079063,
+      "learning_rate": 6.44540555022758e-07,
+      "loss": 0.5023,
+      "step": 37715
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.698260376373475,
+      "learning_rate": 6.442710723257206e-07,
+      "loss": 0.6355,
+      "step": 37716
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.289259269942451,
+      "learning_rate": 6.440016441006991e-07,
+      "loss": 0.7672,
+      "step": 37717
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.304797049061366,
+      "learning_rate": 6.437322703492633e-07,
+      "loss": 0.5576,
+      "step": 37718
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.20381668410744,
+      "learning_rate": 6.434629510729806e-07,
+      "loss": 0.5924,
+      "step": 37719
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3470784085304095,
+      "learning_rate": 6.431936862734167e-07,
+      "loss": 0.6502,
+      "step": 37720
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.589282389183862,
+      "learning_rate": 6.429244759521425e-07,
+      "loss": 0.4808,
+      "step": 37721
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1750489627958234,
+      "learning_rate": 6.426553201107256e-07,
+      "loss": 0.591,
+      "step": 37722
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5397788182828516,
+      "learning_rate": 6.423862187507291e-07,
+      "loss": 0.6982,
+      "step": 37723
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.574087182800771,
+      "learning_rate": 6.421171718737252e-07,
+      "loss": 0.4844,
+      "step": 37724
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5983812077754015,
+      "learning_rate": 6.418481794812781e-07,
+      "loss": 0.511,
+      "step": 37725
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2593376190934065,
+      "learning_rate": 6.415792415749522e-07,
+      "loss": 0.6376,
+      "step": 37726
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.341442710557855,
+      "learning_rate": 6.41310358156314e-07,
+      "loss": 0.65,
+      "step": 37727
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.396568337580683,
+      "learning_rate": 6.410415292269312e-07,
+      "loss": 0.6361,
+      "step": 37728
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.242137925049828,
+      "learning_rate": 6.407727547883691e-07,
+      "loss": 0.7068,
+      "step": 37729
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3728112986283671,
+      "learning_rate": 6.405040348421876e-07,
+      "loss": 0.6522,
+      "step": 37730
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6979546799956262,
+      "learning_rate": 6.402353693899566e-07,
+      "loss": 0.5944,
+      "step": 37731
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4148869978723133,
+      "learning_rate": 6.399667584332403e-07,
+      "loss": 0.6297,
+      "step": 37732
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4947003782017703,
+      "learning_rate": 6.396982019735987e-07,
+      "loss": 0.7321,
+      "step": 37733
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.308110658515458,
+      "learning_rate": 6.394297000125971e-07,
+      "loss": 0.7722,
+      "step": 37734
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4217305930360375,
+      "learning_rate": 6.391612525517988e-07,
+      "loss": 0.6275,
+      "step": 37735
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4424144490232722,
+      "learning_rate": 6.388928595927691e-07,
+      "loss": 0.4974,
+      "step": 37736
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3123566034923604,
+      "learning_rate": 6.38624521137069e-07,
+      "loss": 0.6449,
+      "step": 37737
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8774411466326117,
+      "learning_rate": 6.383562371862595e-07,
+      "loss": 0.7445,
+      "step": 37738
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.370703610898371,
+      "learning_rate": 6.380880077419049e-07,
+      "loss": 0.6284,
+      "step": 37739
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.6655952453803042,
+      "learning_rate": 6.378198328055652e-07,
+      "loss": 0.6608,
+      "step": 37740
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.452970376383548,
+      "learning_rate": 6.375517123788033e-07,
+      "loss": 0.6343,
+      "step": 37741
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.151687887967473,
+      "learning_rate": 6.372836464631771e-07,
+      "loss": 0.62,
+      "step": 37742
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3628833865412093,
+      "learning_rate": 6.37015635060253e-07,
+      "loss": 0.7042,
+      "step": 37743
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5224764022174002,
+      "learning_rate": 6.367476781715853e-07,
+      "loss": 0.5678,
+      "step": 37744
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.438642310331988,
+      "learning_rate": 6.364797757987384e-07,
+      "loss": 0.6778,
+      "step": 37745
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.261605815630602,
+      "learning_rate": 6.36211927943271e-07,
+      "loss": 0.6929,
+      "step": 37746
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3846795238101524,
+      "learning_rate": 6.359441346067419e-07,
+      "loss": 0.7267,
+      "step": 37747
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4167345720218463,
+      "learning_rate": 6.356763957907097e-07,
+      "loss": 0.6457,
+      "step": 37748
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2431637817541317,
+      "learning_rate": 6.354087114967355e-07,
+      "loss": 0.7442,
+      "step": 37749
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3536136991911167,
+      "learning_rate": 6.35141081726377e-07,
+      "loss": 0.5247,
+      "step": 37750
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.378640334139086,
+      "learning_rate": 6.348735064811895e-07,
+      "loss": 0.6369,
+      "step": 37751
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.733173531692585,
+      "learning_rate": 6.34605985762734e-07,
+      "loss": 0.6434,
+      "step": 37752
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.41595514902097,
+      "learning_rate": 6.343385195725693e-07,
+      "loss": 0.638,
+      "step": 37753
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4964385282198502,
+      "learning_rate": 6.340711079122508e-07,
+      "loss": 0.6572,
+      "step": 37754
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5969494471822676,
+      "learning_rate": 6.338037507833339e-07,
+      "loss": 0.6231,
+      "step": 37755
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.445592851085406,
+      "learning_rate": 6.335364481873762e-07,
+      "loss": 0.6235,
+      "step": 37756
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.363950015136078,
+      "learning_rate": 6.332692001259377e-07,
+      "loss": 0.7104,
+      "step": 37757
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7694965863441197,
+      "learning_rate": 6.330020066005704e-07,
+      "loss": 0.7129,
+      "step": 37758
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3762020148959406,
+      "learning_rate": 6.327348676128298e-07,
+      "loss": 0.6102,
+      "step": 37759
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.225903556139519,
+      "learning_rate": 6.324677831642734e-07,
+      "loss": 0.6105,
+      "step": 37760
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.438611785021751,
+      "learning_rate": 6.322007532564533e-07,
+      "loss": 0.757,
+      "step": 37761
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.224688830969084,
+      "learning_rate": 6.319337778909295e-07,
+      "loss": 0.726,
+      "step": 37762
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2737036585298562,
+      "learning_rate": 6.316668570692497e-07,
+      "loss": 0.5639,
+      "step": 37763
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4434350801183184,
+      "learning_rate": 6.313999907929746e-07,
+      "loss": 0.7192,
+      "step": 37764
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.571240627534071,
+      "learning_rate": 6.311331790636521e-07,
+      "loss": 0.7319,
+      "step": 37765
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.14830196040344,
+      "learning_rate": 6.308664218828397e-07,
+      "loss": 0.7366,
+      "step": 37766
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3987240664615843,
+      "learning_rate": 6.305997192520908e-07,
+      "loss": 0.5679,
+      "step": 37767
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2248301125292835,
+      "learning_rate": 6.303330711729539e-07,
+      "loss": 0.6082,
+      "step": 37768
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3158935979433593,
+      "learning_rate": 6.300664776469845e-07,
+      "loss": 0.6242,
+      "step": 37769
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2983506755880934,
+      "learning_rate": 6.297999386757359e-07,
+      "loss": 0.6545,
+      "step": 37770
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4262498382781195,
+      "learning_rate": 6.29533454260759e-07,
+      "loss": 0.5395,
+      "step": 37771
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3395851151676592,
+      "learning_rate": 6.292670244036026e-07,
+      "loss": 0.6721,
+      "step": 37772
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.374528334233034,
+      "learning_rate": 6.290006491058209e-07,
+      "loss": 0.6249,
+      "step": 37773
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4312874158276843,
+      "learning_rate": 6.287343283689662e-07,
+      "loss": 0.6852,
+      "step": 37774
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4876041768899386,
+      "learning_rate": 6.28468062194586e-07,
+      "loss": 0.6042,
+      "step": 37775
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5069522692430883,
+      "learning_rate": 6.282018505842314e-07,
+      "loss": 0.687,
+      "step": 37776
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.246575228363986,
+      "learning_rate": 6.27935693539452e-07,
+      "loss": 0.6045,
+      "step": 37777
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2753593773899214,
+      "learning_rate": 6.276695910618002e-07,
+      "loss": 0.7256,
+      "step": 37778
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3234272771314741,
+      "learning_rate": 6.274035431528225e-07,
+      "loss": 0.7357,
+      "step": 37779
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4082935982449822,
+      "learning_rate": 6.271375498140675e-07,
+      "loss": 0.8084,
+      "step": 37780
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5252056923826742,
+      "learning_rate": 6.268716110470863e-07,
+      "loss": 0.7318,
+      "step": 37781
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2789617702884464,
+      "learning_rate": 6.266057268534254e-07,
+      "loss": 0.6592,
+      "step": 37782
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3646163110257457,
+      "learning_rate": 6.263398972346346e-07,
+      "loss": 0.5736,
+      "step": 37783
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6600000164817026,
+      "learning_rate": 6.260741221922583e-07,
+      "loss": 0.6888,
+      "step": 37784
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3757045050397396,
+      "learning_rate": 6.258084017278487e-07,
+      "loss": 0.6457,
+      "step": 37785
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3317620338855511,
+      "learning_rate": 6.255427358429489e-07,
+      "loss": 0.6883,
+      "step": 37786
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6987044484163063,
+      "learning_rate": 6.252771245391087e-07,
+      "loss": 0.8085,
+      "step": 37787
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.231727116151189,
+      "learning_rate": 6.250115678178725e-07,
+      "loss": 0.5405,
+      "step": 37788
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4769333434412588,
+      "learning_rate": 6.247460656807857e-07,
+      "loss": 0.6039,
+      "step": 37789
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.21211469999341,
+      "learning_rate": 6.24480618129395e-07,
+      "loss": 0.6378,
+      "step": 37790
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.343004984680454,
+      "learning_rate": 6.242152251652489e-07,
+      "loss": 0.5835,
+      "step": 37791
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5010124291082982,
+      "learning_rate": 6.239498867898897e-07,
+      "loss": 0.7569,
+      "step": 37792
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2033791300813583,
+      "learning_rate": 6.236846030048605e-07,
+      "loss": 0.757,
+      "step": 37793
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4970098396351073,
+      "learning_rate": 6.234193738117089e-07,
+      "loss": 0.7327,
+      "step": 37794
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.114587106634608,
+      "learning_rate": 6.231541992119794e-07,
+      "loss": 0.7784,
+      "step": 37795
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.128932612089398,
+      "learning_rate": 6.22889079207215e-07,
+      "loss": 0.6641,
+      "step": 37796
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4606656212638092,
+      "learning_rate": 6.226240137989581e-07,
+      "loss": 0.6214,
+      "step": 37797
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2720717103132622,
+      "learning_rate": 6.223590029887516e-07,
+      "loss": 0.676,
+      "step": 37798
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.509180301086766,
+      "learning_rate": 6.220940467781422e-07,
+      "loss": 0.6417,
+      "step": 37799
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7142979616367284,
+      "learning_rate": 6.218291451686709e-07,
+      "loss": 0.6146,
+      "step": 37800
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.432145572001893,
+      "learning_rate": 6.215642981618775e-07,
+      "loss": 0.6638,
+      "step": 37801
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2090270822495337,
+      "learning_rate": 6.212995057593052e-07,
+      "loss": 0.7355,
+      "step": 37802
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6128337491047448,
+      "learning_rate": 6.210347679624984e-07,
+      "loss": 0.5158,
+      "step": 37803
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5760960179872399,
+      "learning_rate": 6.207700847729969e-07,
+      "loss": 0.8402,
+      "step": 37804
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4746197361949198,
+      "learning_rate": 6.205054561923396e-07,
+      "loss": 0.7366,
+      "step": 37805
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3208648488720436,
+      "learning_rate": 6.202408822220696e-07,
+      "loss": 0.6355,
+      "step": 37806
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4644469053828146,
+      "learning_rate": 6.199763628637267e-07,
+      "loss": 0.7646,
+      "step": 37807
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2992425106510794,
+      "learning_rate": 6.197118981188499e-07,
+      "loss": 0.5556,
+      "step": 37808
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.473402672904967,
+      "learning_rate": 6.194474879889823e-07,
+      "loss": 0.646,
+      "step": 37809
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5825753587323508,
+      "learning_rate": 6.191831324756615e-07,
+      "loss": 0.6555,
+      "step": 37810
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.103311251028042,
+      "learning_rate": 6.189188315804251e-07,
+      "loss": 0.5368,
+      "step": 37811
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6074447859333414,
+      "learning_rate": 6.18654585304812e-07,
+      "loss": 0.756,
+      "step": 37812
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9697511809261683,
+      "learning_rate": 6.183903936503666e-07,
+      "loss": 0.6004,
+      "step": 37813
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6162598624932842,
+      "learning_rate": 6.181262566186186e-07,
+      "loss": 0.5299,
+      "step": 37814
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4514042985377416,
+      "learning_rate": 6.178621742111102e-07,
+      "loss": 0.7786,
+      "step": 37815
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3639600198330561,
+      "learning_rate": 6.175981464293779e-07,
+      "loss": 0.6916,
+      "step": 37816
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.406220507974784,
+      "learning_rate": 6.173341732749627e-07,
+      "loss": 0.8002,
+      "step": 37817
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3765919692691797,
+      "learning_rate": 6.170702547493956e-07,
+      "loss": 0.6664,
+      "step": 37818
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.424216859796257,
+      "learning_rate": 6.168063908542166e-07,
+      "loss": 0.7173,
+      "step": 37819
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.500933995115237,
+      "learning_rate": 6.16542581590962e-07,
+      "loss": 0.7025,
+      "step": 37820
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8020642678501189,
+      "learning_rate": 6.162788269611664e-07,
+      "loss": 0.5442,
+      "step": 37821
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.497072606217558,
+      "learning_rate": 6.160151269663684e-07,
+      "loss": 0.6194,
+      "step": 37822
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4912361118186184,
+      "learning_rate": 6.157514816080989e-07,
+      "loss": 0.6593,
+      "step": 37823
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6000029047317192,
+      "learning_rate": 6.15487890887897e-07,
+      "loss": 0.7005,
+      "step": 37824
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5043277843971454,
+      "learning_rate": 6.152243548072933e-07,
+      "loss": 0.7107,
+      "step": 37825
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5836748898616202,
+      "learning_rate": 6.149608733678269e-07,
+      "loss": 0.6518,
+      "step": 37826
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4426381271319,
+      "learning_rate": 6.146974465710298e-07,
+      "loss": 0.7108,
+      "step": 37827
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3350145418271384,
+      "learning_rate": 6.144340744184341e-07,
+      "loss": 0.5525,
+      "step": 37828
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2205823055032825,
+      "learning_rate": 6.14170756911574e-07,
+      "loss": 0.6369,
+      "step": 37829
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2373921124290703,
+      "learning_rate": 6.139074940519851e-07,
+      "loss": 0.7034,
+      "step": 37830
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5085000079152282,
+      "learning_rate": 6.136442858411984e-07,
+      "loss": 0.8815,
+      "step": 37831
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5137052927635262,
+      "learning_rate": 6.133811322807437e-07,
+      "loss": 0.7264,
+      "step": 37832
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2927920050833992,
+      "learning_rate": 6.131180333721576e-07,
+      "loss": 0.5887,
+      "step": 37833
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3919486619339372,
+      "learning_rate": 6.128549891169711e-07,
+      "loss": 0.6367,
+      "step": 37834
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6490494723015188,
+      "learning_rate": 6.12591999516714e-07,
+      "loss": 0.8121,
+      "step": 37835
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5753140743657964,
+      "learning_rate": 6.123290645729174e-07,
+      "loss": 0.5284,
+      "step": 37836
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4059793648245433,
+      "learning_rate": 6.120661842871134e-07,
+      "loss": 0.6739,
+      "step": 37837
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.450116694772596,
+      "learning_rate": 6.11803358660833e-07,
+      "loss": 0.5957,
+      "step": 37838
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2923319433317195,
+      "learning_rate": 6.11540587695606e-07,
+      "loss": 0.6181,
+      "step": 37839
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4663055126848221,
+      "learning_rate": 6.112778713929601e-07,
+      "loss": 0.5749,
+      "step": 37840
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3790648275554804,
+      "learning_rate": 6.110152097544286e-07,
+      "loss": 0.5518,
+      "step": 37841
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3566900555854902,
+      "learning_rate": 6.10752602781538e-07,
+      "loss": 0.5889,
+      "step": 37842
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4186050667258367,
+      "learning_rate": 6.104900504758193e-07,
+      "loss": 0.7493,
+      "step": 37843
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3329834505927367,
+      "learning_rate": 6.10227552838799e-07,
+      "loss": 0.5894,
+      "step": 37844
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3045501093242828,
+      "learning_rate": 6.099651098720083e-07,
+      "loss": 0.7159,
+      "step": 37845
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3975674350192275,
+      "learning_rate": 6.097027215769725e-07,
+      "loss": 0.6317,
+      "step": 37846
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.762282069199853,
+      "learning_rate": 6.094403879552213e-07,
+      "loss": 0.6779,
+      "step": 37847
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2707728076628895,
+      "learning_rate": 6.091781090082804e-07,
+      "loss": 0.6381,
+      "step": 37848
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5387349659560029,
+      "learning_rate": 6.089158847376775e-07,
+      "loss": 0.4757,
+      "step": 37849
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6259563408362414,
+      "learning_rate": 6.08653715144939e-07,
+      "loss": 0.7856,
+      "step": 37850
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4535175697275815,
+      "learning_rate": 6.083916002315926e-07,
+      "loss": 0.6905,
+      "step": 37851
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3308138384721027,
+      "learning_rate": 6.081295399991649e-07,
+      "loss": 0.6443,
+      "step": 37852
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.649297109205891,
+      "learning_rate": 6.07867534449178e-07,
+      "loss": 0.7072,
+      "step": 37853
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.398577410253361,
+      "learning_rate": 6.076055835831596e-07,
+      "loss": 0.6858,
+      "step": 37854
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3870809964125266,
+      "learning_rate": 6.073436874026373e-07,
+      "loss": 0.657,
+      "step": 37855
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.46743550632136,
+      "learning_rate": 6.070818459091333e-07,
+      "loss": 0.6042,
+      "step": 37856
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.681666189106301,
+      "learning_rate": 6.068200591041717e-07,
+      "loss": 0.6302,
+      "step": 37857
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5273087920569628,
+      "learning_rate": 6.065583269892761e-07,
+      "loss": 0.6989,
+      "step": 37858
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.616900631483958,
+      "learning_rate": 6.06296649565975e-07,
+      "loss": 0.5155,
+      "step": 37859
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2645935487963753,
+      "learning_rate": 6.060350268357873e-07,
+      "loss": 0.6129,
+      "step": 37860
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4584576401156655,
+      "learning_rate": 6.057734588002373e-07,
+      "loss": 0.6703,
+      "step": 37861
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8149871253271346,
+      "learning_rate": 6.055119454608493e-07,
+      "loss": 0.486,
+      "step": 37862
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3784200120286463,
+      "learning_rate": 6.052504868191444e-07,
+      "loss": 0.7617,
+      "step": 37863
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4224924761845719,
+      "learning_rate": 6.049890828766458e-07,
+      "loss": 0.594,
+      "step": 37864
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5701817649192829,
+      "learning_rate": 6.047277336348745e-07,
+      "loss": 0.8258,
+      "step": 37865
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.506487172689327,
+      "learning_rate": 6.044664390953547e-07,
+      "loss": 0.6644,
+      "step": 37866
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2216716287205704,
+      "learning_rate": 6.042051992596032e-07,
+      "loss": 0.6463,
+      "step": 37867
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8921637411453454,
+      "learning_rate": 6.039440141291464e-07,
+      "loss": 0.508,
+      "step": 37868
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6329935848732396,
+      "learning_rate": 6.036828837055008e-07,
+      "loss": 0.4816,
+      "step": 37869
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5680925732847292,
+      "learning_rate": 6.034218079901877e-07,
+      "loss": 0.4964,
+      "step": 37870
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6121662912397745,
+      "learning_rate": 6.031607869847278e-07,
+      "loss": 0.4783,
+      "step": 37871
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2672990128564756,
+      "learning_rate": 6.028998206906423e-07,
+      "loss": 0.6349,
+      "step": 37872
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.435922197783076,
+      "learning_rate": 6.026389091094487e-07,
+      "loss": 0.9075,
+      "step": 37873
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6107734529318183,
+      "learning_rate": 6.023780522426659e-07,
+      "loss": 0.4917,
+      "step": 37874
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.265551372140712,
+      "learning_rate": 6.021172500918127e-07,
+      "loss": 0.7339,
+      "step": 37875
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3970341426452901,
+      "learning_rate": 6.018565026584089e-07,
+      "loss": 0.5593,
+      "step": 37876
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7018488186761696,
+      "learning_rate": 6.015958099439734e-07,
+      "loss": 0.7127,
+      "step": 37877
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.28533496454707,
+      "learning_rate": 6.013351719500205e-07,
+      "loss": 0.6668,
+      "step": 37878
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4744417638051812,
+      "learning_rate": 6.010745886780689e-07,
+      "loss": 0.6967,
+      "step": 37879
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6214231922136108,
+      "learning_rate": 6.008140601296398e-07,
+      "loss": 0.5331,
+      "step": 37880
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4567151821041096,
+      "learning_rate": 6.005535863062461e-07,
+      "loss": 0.7078,
+      "step": 37881
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4146981963578156,
+      "learning_rate": 6.002931672094037e-07,
+      "loss": 0.8098,
+      "step": 37882
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4079773687479886,
+      "learning_rate": 6.000328028406322e-07,
+      "loss": 0.6539,
+      "step": 37883
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3847810986029954,
+      "learning_rate": 5.997724932014437e-07,
+      "loss": 0.736,
+      "step": 37884
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2088955182890535,
+      "learning_rate": 5.995122382933571e-07,
+      "loss": 0.5559,
+      "step": 37885
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6263878451178893,
+      "learning_rate": 5.992520381178857e-07,
+      "loss": 0.7156,
+      "step": 37886
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.389153176277153,
+      "learning_rate": 5.98991892676547e-07,
+      "loss": 0.6933,
+      "step": 37887
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.585823422755289,
+      "learning_rate": 5.98731801970851e-07,
+      "loss": 0.5901,
+      "step": 37888
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4196012170053092,
+      "learning_rate": 5.984717660023176e-07,
+      "loss": 0.6625,
+      "step": 37889
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2458666321431824,
+      "learning_rate": 5.982117847724567e-07,
+      "loss": 0.5398,
+      "step": 37890
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.427685822613778,
+      "learning_rate": 5.979518582827826e-07,
+      "loss": 0.8679,
+      "step": 37891
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7888451640182659,
+      "learning_rate": 5.976919865348097e-07,
+      "loss": 0.5558,
+      "step": 37892
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5535692471593947,
+      "learning_rate": 5.974321695300523e-07,
+      "loss": 0.4943,
+      "step": 37893
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.342281680746148,
+      "learning_rate": 5.971724072700213e-07,
+      "loss": 0.7463,
+      "step": 37894
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3482914318139956,
+      "learning_rate": 5.96912699756228e-07,
+      "loss": 0.7458,
+      "step": 37895
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5024852077128643,
+      "learning_rate": 5.966530469901854e-07,
+      "loss": 0.7023,
+      "step": 37896
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.303688330266797,
+      "learning_rate": 5.96393448973408e-07,
+      "loss": 0.5262,
+      "step": 37897
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5670808116524095,
+      "learning_rate": 5.961339057074045e-07,
+      "loss": 0.5905,
+      "step": 37898
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4188092592012371,
+      "learning_rate": 5.958744171936859e-07,
+      "loss": 0.691,
+      "step": 37899
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4883400352203684,
+      "learning_rate": 5.956149834337632e-07,
+      "loss": 0.6271,
+      "step": 37900
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.427857455650873,
+      "learning_rate": 5.953556044291487e-07,
+      "loss": 0.5592,
+      "step": 37901
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6119168016131262,
+      "learning_rate": 5.950962801813521e-07,
+      "loss": 0.6157,
+      "step": 37902
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5356430250048263,
+      "learning_rate": 5.948370106918799e-07,
+      "loss": 0.6145,
+      "step": 37903
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.421051694406364,
+      "learning_rate": 5.945777959622467e-07,
+      "loss": 0.6768,
+      "step": 37904
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3066943271690816,
+      "learning_rate": 5.943186359939579e-07,
+      "loss": 0.6471,
+      "step": 37905
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5698043020167997,
+      "learning_rate": 5.940595307885233e-07,
+      "loss": 0.5117,
+      "step": 37906
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6056744319274869,
+      "learning_rate": 5.93800480347454e-07,
+      "loss": 0.5113,
+      "step": 37907
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2991626412584654,
+      "learning_rate": 5.935414846722554e-07,
+      "loss": 0.7392,
+      "step": 37908
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4560490302546676,
+      "learning_rate": 5.932825437644352e-07,
+      "loss": 0.6836,
+      "step": 37909
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.519926586606212,
+      "learning_rate": 5.930236576255022e-07,
+      "loss": 0.7558,
+      "step": 37910
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.643956463653481,
+      "learning_rate": 5.927648262569663e-07,
+      "loss": 0.5409,
+      "step": 37911
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3256928932274745,
+      "learning_rate": 5.925060496603297e-07,
+      "loss": 0.6855,
+      "step": 37912
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.27265544598285,
+      "learning_rate": 5.92247327837101e-07,
+      "loss": 0.6485,
+      "step": 37913
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4366742676646929,
+      "learning_rate": 5.919886607887859e-07,
+      "loss": 0.6472,
+      "step": 37914
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3677791003765871,
+      "learning_rate": 5.917300485168953e-07,
+      "loss": 0.588,
+      "step": 37915
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4505367867823553,
+      "learning_rate": 5.91471491022928e-07,
+      "loss": 0.661,
+      "step": 37916
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5858506422932237,
+      "learning_rate": 5.912129883083917e-07,
+      "loss": 0.5186,
+      "step": 37917
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.468081479416106,
+      "learning_rate": 5.909545403747941e-07,
+      "loss": 0.7016,
+      "step": 37918
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3516211945932726,
+      "learning_rate": 5.906961472236361e-07,
+      "loss": 0.7411,
+      "step": 37919
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2870782117610233,
+      "learning_rate": 5.904378088564255e-07,
+      "loss": 0.6054,
+      "step": 37920
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6243679602914423,
+      "learning_rate": 5.901795252746644e-07,
+      "loss": 0.6787,
+      "step": 37921
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2756849602127847,
+      "learning_rate": 5.899212964798584e-07,
+      "loss": 0.7658,
+      "step": 37922
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4621004670690532,
+      "learning_rate": 5.896631224735072e-07,
+      "loss": 0.7395,
+      "step": 37923
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.325717001501937,
+      "learning_rate": 5.894050032571197e-07,
+      "loss": 0.7518,
+      "step": 37924
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3962500751203268,
+      "learning_rate": 5.891469388321936e-07,
+      "loss": 0.6348,
+      "step": 37925
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5246637320399157,
+      "learning_rate": 5.888889292002342e-07,
+      "loss": 0.7207,
+      "step": 37926
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4218302103851543,
+      "learning_rate": 5.886309743627416e-07,
+      "loss": 0.8143,
+      "step": 37927
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5212750554767633,
+      "learning_rate": 5.883730743212213e-07,
+      "loss": 0.7513,
+      "step": 37928
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.420443858701328,
+      "learning_rate": 5.881152290771719e-07,
+      "loss": 0.6527,
+      "step": 37929
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3787604506341176,
+      "learning_rate": 5.878574386320945e-07,
+      "loss": 0.7208,
+      "step": 37930
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5802829215945071,
+      "learning_rate": 5.875997029874903e-07,
+      "loss": 0.6457,
+      "step": 37931
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.344703048424373,
+      "learning_rate": 5.873420221448623e-07,
+      "loss": 0.6529,
+      "step": 37932
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.276074820731682,
+      "learning_rate": 5.870843961057093e-07,
+      "loss": 0.5282,
+      "step": 37933
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.234668270906335,
+      "learning_rate": 5.868268248715292e-07,
+      "loss": 0.7211,
+      "step": 37934
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.293193624717578,
+      "learning_rate": 5.86569308443824e-07,
+      "loss": 0.7468,
+      "step": 37935
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.212236868955054,
+      "learning_rate": 5.863118468240936e-07,
+      "loss": 0.7096,
+      "step": 37936
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4504738800873629,
+      "learning_rate": 5.860544400138357e-07,
+      "loss": 0.7208,
+      "step": 37937
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4687331322147006,
+      "learning_rate": 5.857970880145492e-07,
+      "loss": 0.7134,
+      "step": 37938
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6910796349630985,
+      "learning_rate": 5.855397908277305e-07,
+      "loss": 0.6896,
+      "step": 37939
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3086126874151263,
+      "learning_rate": 5.85282548454883e-07,
+      "loss": 0.7628,
+      "step": 37940
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2959952442355436,
+      "learning_rate": 5.850253608974998e-07,
+      "loss": 0.4669,
+      "step": 37941
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3994006258723102,
+      "learning_rate": 5.847682281570788e-07,
+      "loss": 0.6406,
+      "step": 37942
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8985255879797338,
+      "learning_rate": 5.845111502351197e-07,
+      "loss": 0.6218,
+      "step": 37943
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3975491163536051,
+      "learning_rate": 5.842541271331148e-07,
+      "loss": 0.6404,
+      "step": 37944
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4774135443234235,
+      "learning_rate": 5.839971588525661e-07,
+      "loss": 0.7627,
+      "step": 37945
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3460276613485092,
+      "learning_rate": 5.837402453949648e-07,
+      "loss": 0.7135,
+      "step": 37946
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2794470619151836,
+      "learning_rate": 5.834833867618117e-07,
+      "loss": 0.8137,
+      "step": 37947
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5961453136145548,
+      "learning_rate": 5.832265829545969e-07,
+      "loss": 0.5042,
+      "step": 37948
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5910745544256838,
+      "learning_rate": 5.829698339748202e-07,
+      "loss": 0.4829,
+      "step": 37949
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.421195547962792,
+      "learning_rate": 5.827131398239738e-07,
+      "loss": 0.6711,
+      "step": 37950
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5006501670824597,
+      "learning_rate": 5.824565005035521e-07,
+      "loss": 0.7224,
+      "step": 37951
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5838557867966035,
+      "learning_rate": 5.821999160150504e-07,
+      "loss": 0.4581,
+      "step": 37952
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4683307932195846,
+      "learning_rate": 5.819433863599633e-07,
+      "loss": 0.6383,
+      "step": 37953
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5670111642069698,
+      "learning_rate": 5.816869115397838e-07,
+      "loss": 0.4907,
+      "step": 37954
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.25085695045351,
+      "learning_rate": 5.814304915560031e-07,
+      "loss": 0.6099,
+      "step": 37955
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3702671016548853,
+      "learning_rate": 5.811741264101167e-07,
+      "loss": 0.6368,
+      "step": 37956
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2913938577366728,
+      "learning_rate": 5.809178161036166e-07,
+      "loss": 0.6225,
+      "step": 37957
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2778572474556398,
+      "learning_rate": 5.806615606379961e-07,
+      "loss": 0.7719,
+      "step": 37958
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5656248987245782,
+      "learning_rate": 5.804053600147452e-07,
+      "loss": 0.4873,
+      "step": 37959
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3990500039224798,
+      "learning_rate": 5.801492142353549e-07,
+      "loss": 0.7274,
+      "step": 37960
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3911053014539265,
+      "learning_rate": 5.798931233013205e-07,
+      "loss": 0.7716,
+      "step": 37961
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.285558507120713,
+      "learning_rate": 5.796370872141311e-07,
+      "loss": 0.5361,
+      "step": 37962
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4231731940272423,
+      "learning_rate": 5.793811059752752e-07,
+      "loss": 0.6418,
+      "step": 37963
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.518228960482804,
+      "learning_rate": 5.791251795862463e-07,
+      "loss": 0.708,
+      "step": 37964
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5691201292934152,
+      "learning_rate": 5.788693080485319e-07,
+      "loss": 0.6436,
+      "step": 37965
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4632250073375515,
+      "learning_rate": 5.786134913636243e-07,
+      "loss": 0.5828,
+      "step": 37966
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5180846429474366,
+      "learning_rate": 5.7835772953301e-07,
+      "loss": 0.6552,
+      "step": 37967
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5840899394386445,
+      "learning_rate": 5.781020225581813e-07,
+      "loss": 0.6811,
+      "step": 37968
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.415581002333588,
+      "learning_rate": 5.778463704406245e-07,
+      "loss": 0.675,
+      "step": 37969
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5805154402162227,
+      "learning_rate": 5.775907731818308e-07,
+      "loss": 0.7648,
+      "step": 37970
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6037093370977398,
+      "learning_rate": 5.773352307832869e-07,
+      "loss": 0.4615,
+      "step": 37971
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.701624140074381,
+      "learning_rate": 5.770797432464781e-07,
+      "loss": 0.6446,
+      "step": 37972
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.229648227505504,
+      "learning_rate": 5.768243105728954e-07,
+      "loss": 0.7009,
+      "step": 37973
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3702281256692923,
+      "learning_rate": 5.765689327640256e-07,
+      "loss": 0.6784,
+      "step": 37974
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3008629548301904,
+      "learning_rate": 5.763136098213551e-07,
+      "loss": 0.6734,
+      "step": 37975
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4525646748629129,
+      "learning_rate": 5.760583417463683e-07,
+      "loss": 0.7392,
+      "step": 37976
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2667765623142104,
+      "learning_rate": 5.758031285405541e-07,
+      "loss": 0.7053,
+      "step": 37977
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8338172557506653,
+      "learning_rate": 5.755479702053979e-07,
+      "loss": 0.6565,
+      "step": 37978
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2931427382131435,
+      "learning_rate": 5.752928667423862e-07,
+      "loss": 0.6182,
+      "step": 37979
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5212904747031863,
+      "learning_rate": 5.750378181530025e-07,
+      "loss": 0.7104,
+      "step": 37980
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3808615453619069,
+      "learning_rate": 5.74782824438731e-07,
+      "loss": 0.7451,
+      "step": 37981
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8258321948467715,
+      "learning_rate": 5.745278856010605e-07,
+      "loss": 0.6638,
+      "step": 37982
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2672590912511013,
+      "learning_rate": 5.742730016414722e-07,
+      "loss": 0.6197,
+      "step": 37983
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2027856802506416,
+      "learning_rate": 5.740181725614491e-07,
+      "loss": 0.5455,
+      "step": 37984
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.156318975879282,
+      "learning_rate": 5.737633983624779e-07,
+      "loss": 0.5867,
+      "step": 37985
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3548961053609365,
+      "learning_rate": 5.735086790460387e-07,
+      "loss": 0.6748,
+      "step": 37986
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4165603194025136,
+      "learning_rate": 5.73254014613619e-07,
+      "loss": 0.5127,
+      "step": 37987
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5960877049197862,
+      "learning_rate": 5.729994050666965e-07,
+      "loss": 0.486,
+      "step": 37988
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3523551458031022,
+      "learning_rate": 5.727448504067578e-07,
+      "loss": 0.681,
+      "step": 37989
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.286157106005903,
+      "learning_rate": 5.724903506352808e-07,
+      "loss": 0.8325,
+      "step": 37990
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4480673275020421,
+      "learning_rate": 5.722359057537519e-07,
+      "loss": 0.714,
+      "step": 37991
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5895586854047976,
+      "learning_rate": 5.71981515763651e-07,
+      "loss": 0.5211,
+      "step": 37992
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.002779550748645,
+      "learning_rate": 5.71727180666456e-07,
+      "loss": 0.7897,
+      "step": 37993
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2851801577226356,
+      "learning_rate": 5.71472900463651e-07,
+      "loss": 0.5968,
+      "step": 37994
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3393230529920515,
+      "learning_rate": 5.71218675156715e-07,
+      "loss": 0.5849,
+      "step": 37995
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3912413096712448,
+      "learning_rate": 5.709645047471324e-07,
+      "loss": 0.68,
+      "step": 37996
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.370275053234296,
+      "learning_rate": 5.707103892363774e-07,
+      "loss": 0.5645,
+      "step": 37997
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4790648982837045,
+      "learning_rate": 5.704563286259313e-07,
+      "loss": 0.7382,
+      "step": 37998
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2367565798822975,
+      "learning_rate": 5.702023229172749e-07,
+      "loss": 0.5111,
+      "step": 37999
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3746559520407855,
+      "learning_rate": 5.699483721118859e-07,
+      "loss": 0.7322,
+      "step": 38000
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2766776761393404,
+      "learning_rate": 5.696944762112422e-07,
+      "loss": 0.6286,
+      "step": 38001
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4961582058213967,
+      "learning_rate": 5.694406352168214e-07,
+      "loss": 0.6862,
+      "step": 38002
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4993326281127504,
+      "learning_rate": 5.691868491301056e-07,
+      "loss": 0.6168,
+      "step": 38003
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4499759124389637,
+      "learning_rate": 5.689331179525681e-07,
+      "loss": 0.7671,
+      "step": 38004
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2843751093372655,
+      "learning_rate": 5.686794416856889e-07,
+      "loss": 0.6045,
+      "step": 38005
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5839042757974449,
+      "learning_rate": 5.684258203309446e-07,
+      "loss": 0.7965,
+      "step": 38006
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5609210476844164,
+      "learning_rate": 5.681722538898093e-07,
+      "loss": 0.5963,
+      "step": 38007
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5118653211313935,
+      "learning_rate": 5.679187423637611e-07,
+      "loss": 0.761,
+      "step": 38008
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.731504582984719,
+      "learning_rate": 5.676652857542775e-07,
+      "loss": 0.5998,
+      "step": 38009
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5438320618774823,
+      "learning_rate": 5.674118840628329e-07,
+      "loss": 0.4324,
+      "step": 38010
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5598846594353595,
+      "learning_rate": 5.671585372909016e-07,
+      "loss": 0.7013,
+      "step": 38011
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4314476008651646,
+      "learning_rate": 5.669052454399593e-07,
+      "loss": 0.6847,
+      "step": 38012
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7422553317346205,
+      "learning_rate": 5.666520085114846e-07,
+      "loss": 0.6748,
+      "step": 38013
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5553233645738587,
+      "learning_rate": 5.663988265069454e-07,
+      "loss": 0.6683,
+      "step": 38014
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7294638913853804,
+      "learning_rate": 5.661456994278181e-07,
+      "loss": 0.5924,
+      "step": 38015
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3345220770942208,
+      "learning_rate": 5.658926272755772e-07,
+      "loss": 0.7384,
+      "step": 38016
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4816553758767537,
+      "learning_rate": 5.656396100516981e-07,
+      "loss": 0.723,
+      "step": 38017
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4359324044213333,
+      "learning_rate": 5.65386647757652e-07,
+      "loss": 0.762,
+      "step": 38018
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6035711347701419,
+      "learning_rate": 5.651337403949109e-07,
+      "loss": 0.4905,
+      "step": 38019
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4415564602542486,
+      "learning_rate": 5.648808879649493e-07,
+      "loss": 0.6011,
+      "step": 38020
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.320467387174059,
+      "learning_rate": 5.64628090469237e-07,
+      "loss": 0.6367,
+      "step": 38021
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3880760820013793,
+      "learning_rate": 5.643753479092473e-07,
+      "loss": 0.6361,
+      "step": 38022
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4872672530812105,
+      "learning_rate": 5.641226602864514e-07,
+      "loss": 0.55,
+      "step": 38023
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.518549249912804,
+      "learning_rate": 5.638700276023223e-07,
+      "loss": 0.8693,
+      "step": 38024
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.9397474846660234,
+      "learning_rate": 5.636174498583269e-07,
+      "loss": 0.6243,
+      "step": 38025
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4128168415792735,
+      "learning_rate": 5.633649270559405e-07,
+      "loss": 0.6875,
+      "step": 38026
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3299055477231387,
+      "learning_rate": 5.631124591966308e-07,
+      "loss": 0.6276,
+      "step": 38027
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6483021969060665,
+      "learning_rate": 5.628600462818668e-07,
+      "loss": 0.5258,
+      "step": 38028
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2647649463894959,
+      "learning_rate": 5.626076883131193e-07,
+      "loss": 0.5527,
+      "step": 38029
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.311530316869868,
+      "learning_rate": 5.623553852918584e-07,
+      "loss": 0.513,
+      "step": 38030
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6821526573221395,
+      "learning_rate": 5.621031372195529e-07,
+      "loss": 0.8242,
+      "step": 38031
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.331313948074972,
+      "learning_rate": 5.618509440976694e-07,
+      "loss": 0.7255,
+      "step": 38032
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.644188784383159,
+      "learning_rate": 5.615988059276777e-07,
+      "loss": 0.7467,
+      "step": 38033
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.533616171084257,
+      "learning_rate": 5.613467227110481e-07,
+      "loss": 0.6553,
+      "step": 38034
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.273527302418233,
+      "learning_rate": 5.610946944492457e-07,
+      "loss": 0.5474,
+      "step": 38035
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4452356485706979,
+      "learning_rate": 5.608427211437362e-07,
+      "loss": 0.6147,
+      "step": 38036
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4782256374841283,
+      "learning_rate": 5.605908027959894e-07,
+      "loss": 0.6025,
+      "step": 38037
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2132398666481277,
+      "learning_rate": 5.603389394074732e-07,
+      "loss": 0.6036,
+      "step": 38038
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4371960152682541,
+      "learning_rate": 5.600871309796519e-07,
+      "loss": 0.6125,
+      "step": 38039
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.370201550306832,
+      "learning_rate": 5.598353775139898e-07,
+      "loss": 0.6965,
+      "step": 38040
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1977783259851527,
+      "learning_rate": 5.595836790119569e-07,
+      "loss": 0.6983,
+      "step": 38041
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3502058046002743,
+      "learning_rate": 5.593320354750154e-07,
+      "loss": 0.6813,
+      "step": 38042
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3407383041809686,
+      "learning_rate": 5.59080446904633e-07,
+      "loss": 0.673,
+      "step": 38043
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.516070075338963,
+      "learning_rate": 5.588289133022706e-07,
+      "loss": 0.6502,
+      "step": 38044
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8412984027026682,
+      "learning_rate": 5.585774346693984e-07,
+      "loss": 0.6275,
+      "step": 38045
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5818881654385974,
+      "learning_rate": 5.583260110074751e-07,
+      "loss": 0.4759,
+      "step": 38046
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4589432606058998,
+      "learning_rate": 5.580746423179684e-07,
+      "loss": 0.6519,
+      "step": 38047
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.651684563120437,
+      "learning_rate": 5.578233286023393e-07,
+      "loss": 0.6554,
+      "step": 38048
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3479240372243917,
+      "learning_rate": 5.575720698620535e-07,
+      "loss": 0.5829,
+      "step": 38049
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3624276985305854,
+      "learning_rate": 5.573208660985707e-07,
+      "loss": 0.5925,
+      "step": 38050
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3092860682044336,
+      "learning_rate": 5.570697173133565e-07,
+      "loss": 0.5511,
+      "step": 38051
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.457018858203578,
+      "learning_rate": 5.568186235078732e-07,
+      "loss": 0.7542,
+      "step": 38052
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5499189410170648,
+      "learning_rate": 5.565675846835794e-07,
+      "loss": 0.5375,
+      "step": 38053
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.47781580261651,
+      "learning_rate": 5.563166008419385e-07,
+      "loss": 0.6732,
+      "step": 38054
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.6095201328612174,
+      "learning_rate": 5.560656719844149e-07,
+      "loss": 0.6708,
+      "step": 38055
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7090403655451045,
+      "learning_rate": 5.55814798112465e-07,
+      "loss": 0.7186,
+      "step": 38056
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4233356829921047,
+      "learning_rate": 5.555639792275514e-07,
+      "loss": 0.5692,
+      "step": 38057
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.37723911190887,
+      "learning_rate": 5.553132153311335e-07,
+      "loss": 0.6281,
+      "step": 38058
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6532246400610229,
+      "learning_rate": 5.55062506424674e-07,
+      "loss": 0.6108,
+      "step": 38059
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.34051226596623,
+      "learning_rate": 5.548118525096313e-07,
+      "loss": 0.7071,
+      "step": 38060
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3214028873803045,
+      "learning_rate": 5.545612535874622e-07,
+      "loss": 0.7128,
+      "step": 38061
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2769463805067216,
+      "learning_rate": 5.543107096596279e-07,
+      "loss": 0.6721,
+      "step": 38062
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5197489352777025,
+      "learning_rate": 5.54060220727588e-07,
+      "loss": 0.7697,
+      "step": 38063
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.359908803626459,
+      "learning_rate": 5.538097867928016e-07,
+      "loss": 0.7364,
+      "step": 38064
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4053271254307047,
+      "learning_rate": 5.53559407856722e-07,
+      "loss": 0.7148,
+      "step": 38065
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4863500271566974,
+      "learning_rate": 5.533090839208133e-07,
+      "loss": 0.5827,
+      "step": 38066
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2717023230792215,
+      "learning_rate": 5.530588149865268e-07,
+      "loss": 0.7418,
+      "step": 38067
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3018458206233843,
+      "learning_rate": 5.528086010553257e-07,
+      "loss": 0.6193,
+      "step": 38068
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2816303076224522,
+      "learning_rate": 5.525584421286612e-07,
+      "loss": 0.4986,
+      "step": 38069
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4840027800568796,
+      "learning_rate": 5.523083382079952e-07,
+      "loss": 0.5864,
+      "step": 38070
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3607718178799864,
+      "learning_rate": 5.520582892947779e-07,
+      "loss": 0.5326,
+      "step": 38071
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.489412299159766,
+      "learning_rate": 5.518082953904714e-07,
+      "loss": 0.7363,
+      "step": 38072
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2887543124855436,
+      "learning_rate": 5.515583564965277e-07,
+      "loss": 0.6631,
+      "step": 38073
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8979652623219692,
+      "learning_rate": 5.513084726144013e-07,
+      "loss": 0.728,
+      "step": 38074
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.286543262693189,
+      "learning_rate": 5.510586437455478e-07,
+      "loss": 0.6808,
+      "step": 38075
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.261195424054901,
+      "learning_rate": 5.508088698914249e-07,
+      "loss": 0.6557,
+      "step": 38076
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6529207108361739,
+      "learning_rate": 5.505591510534836e-07,
+      "loss": 0.7778,
+      "step": 38077
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4862886189119506,
+      "learning_rate": 5.503094872331771e-07,
+      "loss": 0.6679,
+      "step": 38078
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2677720474360024,
+      "learning_rate": 5.5005987843196e-07,
+      "loss": 0.6511,
+      "step": 38079
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3586676284838297,
+      "learning_rate": 5.498103246512887e-07,
+      "loss": 0.7138,
+      "step": 38080
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4708061366026166,
+      "learning_rate": 5.495608258926133e-07,
+      "loss": 0.7603,
+      "step": 38081
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5497443828297093,
+      "learning_rate": 5.493113821573847e-07,
+      "loss": 0.6104,
+      "step": 38082
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.355659390874654,
+      "learning_rate": 5.490619934470576e-07,
+      "loss": 0.7166,
+      "step": 38083
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8346433386005627,
+      "learning_rate": 5.48812659763085e-07,
+      "loss": 0.7383,
+      "step": 38084
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9328737499655944,
+      "learning_rate": 5.485633811069169e-07,
+      "loss": 0.5389,
+      "step": 38085
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6207542975485233,
+      "learning_rate": 5.483141574800044e-07,
+      "loss": 0.4573,
+      "step": 38086
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4894383690088508,
+      "learning_rate": 5.480649888838008e-07,
+      "loss": 0.7598,
+      "step": 38087
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4163053377955819,
+      "learning_rate": 5.478158753197527e-07,
+      "loss": 0.7494,
+      "step": 38088
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2710700583744856,
+      "learning_rate": 5.475668167893156e-07,
+      "loss": 0.6245,
+      "step": 38089
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5837269068783906,
+      "learning_rate": 5.47317813293935e-07,
+      "loss": 0.788,
+      "step": 38090
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.59623969782264,
+      "learning_rate": 5.470688648350642e-07,
+      "loss": 0.5451,
+      "step": 38091
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4178877241193355,
+      "learning_rate": 5.468199714141509e-07,
+      "loss": 0.7839,
+      "step": 38092
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4290327819247501,
+      "learning_rate": 5.465711330326428e-07,
+      "loss": 0.6689,
+      "step": 38093
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3479504301435814,
+      "learning_rate": 5.463223496919945e-07,
+      "loss": 0.7412,
+      "step": 38094
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3718494481770587,
+      "learning_rate": 5.460736213936479e-07,
+      "loss": 0.6812,
+      "step": 38095
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.577252981468033,
+      "learning_rate": 5.458249481390532e-07,
+      "loss": 0.4919,
+      "step": 38096
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.34501741314537,
+      "learning_rate": 5.45576329929658e-07,
+      "loss": 0.6875,
+      "step": 38097
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5931638419097827,
+      "learning_rate": 5.453277667669143e-07,
+      "loss": 0.5009,
+      "step": 38098
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5356434938161192,
+      "learning_rate": 5.450792586522624e-07,
+      "loss": 0.5773,
+      "step": 38099
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5770333126842933,
+      "learning_rate": 5.448308055871532e-07,
+      "loss": 0.6109,
+      "step": 38100
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5035827505239003,
+      "learning_rate": 5.445824075730333e-07,
+      "loss": 0.8492,
+      "step": 38101
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5934787731902187,
+      "learning_rate": 5.443340646113471e-07,
+      "loss": 0.5081,
+      "step": 38102
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.532351839427925,
+      "learning_rate": 5.440857767035434e-07,
+      "loss": 0.6591,
+      "step": 38103
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2685031497030599,
+      "learning_rate": 5.438375438510635e-07,
+      "loss": 0.6498,
+      "step": 38104
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.546279954826991,
+      "learning_rate": 5.43589366055357e-07,
+      "loss": 0.5963,
+      "step": 38105
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4429662262832406,
+      "learning_rate": 5.433412433178664e-07,
+      "loss": 0.7168,
+      "step": 38106
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2285517167405708,
+      "learning_rate": 5.43093175640037e-07,
+      "loss": 0.653,
+      "step": 38107
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6032115060316496,
+      "learning_rate": 5.428451630233144e-07,
+      "loss": 0.6922,
+      "step": 38108
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5462994459123445,
+      "learning_rate": 5.425972054691397e-07,
+      "loss": 0.718,
+      "step": 38109
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3064987437421305,
+      "learning_rate": 5.423493029789573e-07,
+      "loss": 0.7269,
+      "step": 38110
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7916116064857552,
+      "learning_rate": 5.421014555542137e-07,
+      "loss": 0.7055,
+      "step": 38111
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5980930997063065,
+      "learning_rate": 5.418536631963501e-07,
+      "loss": 0.445,
+      "step": 38112
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5579649702704021,
+      "learning_rate": 5.416059259068063e-07,
+      "loss": 0.7391,
+      "step": 38113
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.635708198656842,
+      "learning_rate": 5.41358243687028e-07,
+      "loss": 0.7062,
+      "step": 38114
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2588904148581423,
+      "learning_rate": 5.411106165384583e-07,
+      "loss": 0.6774,
+      "step": 38115
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3405737604953383,
+      "learning_rate": 5.408630444625373e-07,
+      "loss": 0.6916,
+      "step": 38116
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2805020632090307,
+      "learning_rate": 5.406155274607039e-07,
+      "loss": 0.7155,
+      "step": 38117
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6581913486091042,
+      "learning_rate": 5.403680655344023e-07,
+      "loss": 0.6588,
+      "step": 38118
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.428890877857021,
+      "learning_rate": 5.40120658685075e-07,
+      "loss": 0.8031,
+      "step": 38119
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.488959940294108,
+      "learning_rate": 5.398733069141593e-07,
+      "loss": 0.6503,
+      "step": 38120
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5927115958422113,
+      "learning_rate": 5.396260102230944e-07,
+      "loss": 0.5039,
+      "step": 38121
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5869936454212956,
+      "learning_rate": 5.393787686133234e-07,
+      "loss": 0.4866,
+      "step": 38122
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4756624413031507,
+      "learning_rate": 5.391315820862841e-07,
+      "loss": 0.8037,
+      "step": 38123
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.491154645695843,
+      "learning_rate": 5.388844506434166e-07,
+      "loss": 0.5941,
+      "step": 38124
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4216326616185078,
+      "learning_rate": 5.386373742861572e-07,
+      "loss": 0.686,
+      "step": 38125
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2710417613420513,
+      "learning_rate": 5.383903530159484e-07,
+      "loss": 0.4733,
+      "step": 38126
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3230478198098339,
+      "learning_rate": 5.381433868342256e-07,
+      "loss": 0.6415,
+      "step": 38127
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5879583879247475,
+      "learning_rate": 5.378964757424287e-07,
+      "loss": 0.4962,
+      "step": 38128
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6023479101183637,
+      "learning_rate": 5.376496197419934e-07,
+      "loss": 0.8389,
+      "step": 38129
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.429857930921334,
+      "learning_rate": 5.374028188343572e-07,
+      "loss": 0.7443,
+      "step": 38130
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3010989715464139,
+      "learning_rate": 5.371560730209579e-07,
+      "loss": 0.6355,
+      "step": 38131
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4583918630729065,
+      "learning_rate": 5.369093823032323e-07,
+      "loss": 0.5868,
+      "step": 38132
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2870933292323785,
+      "learning_rate": 5.366627466826169e-07,
+      "loss": 0.5733,
+      "step": 38133
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1006612691462885,
+      "learning_rate": 5.364161661605449e-07,
+      "loss": 0.7346,
+      "step": 38134
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3911632460749233,
+      "learning_rate": 5.361696407384543e-07,
+      "loss": 0.7625,
+      "step": 38135
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3230335606693682,
+      "learning_rate": 5.359231704177825e-07,
+      "loss": 0.6049,
+      "step": 38136
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.6630526029074573,
+      "learning_rate": 5.356767551999608e-07,
+      "loss": 0.5845,
+      "step": 38137
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4621148484183581,
+      "learning_rate": 5.354303950864248e-07,
+      "loss": 0.7306,
+      "step": 38138
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3917969413639069,
+      "learning_rate": 5.351840900786098e-07,
+      "loss": 0.6263,
+      "step": 38139
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3755602024981575,
+      "learning_rate": 5.349378401779504e-07,
+      "loss": 0.5873,
+      "step": 38140
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3226110531446422,
+      "learning_rate": 5.346916453858797e-07,
+      "loss": 0.7158,
+      "step": 38141
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5379418952417585,
+      "learning_rate": 5.3444550570383e-07,
+      "loss": 0.5609,
+      "step": 38142
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6199206689117092,
+      "learning_rate": 5.341994211332358e-07,
+      "loss": 0.8184,
+      "step": 38143
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5865235434163326,
+      "learning_rate": 5.339533916755291e-07,
+      "loss": 0.4868,
+      "step": 38144
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6621527674204453,
+      "learning_rate": 5.337074173321433e-07,
+      "loss": 0.6601,
+      "step": 38145
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6212107872075053,
+      "learning_rate": 5.334614981045083e-07,
+      "loss": 0.4774,
+      "step": 38146
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2567741452073513,
+      "learning_rate": 5.332156339940598e-07,
+      "loss": 0.6684,
+      "step": 38147
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.353204172185309,
+      "learning_rate": 5.329698250022241e-07,
+      "loss": 0.7229,
+      "step": 38148
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4051548723670555,
+      "learning_rate": 5.327240711304383e-07,
+      "loss": 0.6481,
+      "step": 38149
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.385885527524505,
+      "learning_rate": 5.324783723801296e-07,
+      "loss": 0.752,
+      "step": 38150
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1631079627028211,
+      "learning_rate": 5.322327287527274e-07,
+      "loss": 0.7483,
+      "step": 38151
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6094957629608238,
+      "learning_rate": 5.319871402496645e-07,
+      "loss": 0.7031,
+      "step": 38152
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4324552491556044,
+      "learning_rate": 5.317416068723702e-07,
+      "loss": 0.7085,
+      "step": 38153
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5306400353186083,
+      "learning_rate": 5.314961286222752e-07,
+      "loss": 0.657,
+      "step": 38154
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.539933826639522,
+      "learning_rate": 5.312507055008054e-07,
+      "loss": 0.6927,
+      "step": 38155
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3373216262123027,
+      "learning_rate": 5.310053375093915e-07,
+      "loss": 0.6194,
+      "step": 38156
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3851737138104605,
+      "learning_rate": 5.307600246494638e-07,
+      "loss": 0.7062,
+      "step": 38157
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4813784210622445,
+      "learning_rate": 5.305147669224497e-07,
+      "loss": 0.8562,
+      "step": 38158
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5684447084740194,
+      "learning_rate": 5.302695643297751e-07,
+      "loss": 0.7134,
+      "step": 38159
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3760032274624225,
+      "learning_rate": 5.300244168728686e-07,
+      "loss": 0.6345,
+      "step": 38160
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3883713430120812,
+      "learning_rate": 5.297793245531602e-07,
+      "loss": 0.6842,
+      "step": 38161
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7003356890809296,
+      "learning_rate": 5.295342873720743e-07,
+      "loss": 0.6604,
+      "step": 38162
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3800300671136612,
+      "learning_rate": 5.292893053310366e-07,
+      "loss": 0.6081,
+      "step": 38163
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7740668553228465,
+      "learning_rate": 5.29044378431477e-07,
+      "loss": 0.6899,
+      "step": 38164
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3895640407080834,
+      "learning_rate": 5.287995066748175e-07,
+      "loss": 0.6576,
+      "step": 38165
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.142573216049863,
+      "learning_rate": 5.285546900624872e-07,
+      "loss": 0.5614,
+      "step": 38166
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.253826973327969,
+      "learning_rate": 5.283099285959081e-07,
+      "loss": 0.7856,
+      "step": 38167
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.927445202747455,
+      "learning_rate": 5.280652222765103e-07,
+      "loss": 0.6116,
+      "step": 38168
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5869202254615833,
+      "learning_rate": 5.278205711057127e-07,
+      "loss": 0.6658,
+      "step": 38169
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.235075039468771,
+      "learning_rate": 5.275759750849441e-07,
+      "loss": 0.6153,
+      "step": 38170
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.734120036889878,
+      "learning_rate": 5.273314342156266e-07,
+      "loss": 0.7059,
+      "step": 38171
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.573042143554868,
+      "learning_rate": 5.270869484991858e-07,
+      "loss": 0.7311,
+      "step": 38172
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6133973147398908,
+      "learning_rate": 5.268425179370418e-07,
+      "loss": 0.6678,
+      "step": 38173
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2735841812341608,
+      "learning_rate": 5.26598142530621e-07,
+      "loss": 0.6372,
+      "step": 38174
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5749680015759354,
+      "learning_rate": 5.263538222813457e-07,
+      "loss": 0.5427,
+      "step": 38175
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4750248770057455,
+      "learning_rate": 5.261095571906361e-07,
+      "loss": 0.6066,
+      "step": 38176
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6180327492662645,
+      "learning_rate": 5.258653472599162e-07,
+      "loss": 0.4726,
+      "step": 38177
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.513420137287714,
+      "learning_rate": 5.256211924906074e-07,
+      "loss": 0.7519,
+      "step": 38178
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9022458067052208,
+      "learning_rate": 5.253770928841351e-07,
+      "loss": 0.5554,
+      "step": 38179
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.598424662005235,
+      "learning_rate": 5.251330484419137e-07,
+      "loss": 0.8219,
+      "step": 38180
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4730281477539644,
+      "learning_rate": 5.248890591653666e-07,
+      "loss": 0.6713,
+      "step": 38181
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.438919838480287,
+      "learning_rate": 5.246451250559182e-07,
+      "loss": 0.6396,
+      "step": 38182
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6332102702918163,
+      "learning_rate": 5.24401246114985e-07,
+      "loss": 0.8706,
+      "step": 38183
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4298121914285498,
+      "learning_rate": 5.24157422343986e-07,
+      "loss": 0.7037,
+      "step": 38184
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4264102615250767,
+      "learning_rate": 5.239136537443423e-07,
+      "loss": 0.617,
+      "step": 38185
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3304851455766393,
+      "learning_rate": 5.23669940317475e-07,
+      "loss": 0.6906,
+      "step": 38186
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4710491467942215,
+      "learning_rate": 5.234262820648006e-07,
+      "loss": 0.789,
+      "step": 38187
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.426936005715608,
+      "learning_rate": 5.231826789877392e-07,
+      "loss": 0.6523,
+      "step": 38188
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4153185703796929,
+      "learning_rate": 5.229391310877085e-07,
+      "loss": 0.6123,
+      "step": 38189
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.355772219916871,
+      "learning_rate": 5.226956383661263e-07,
+      "loss": 0.6984,
+      "step": 38190
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3087609488178793,
+      "learning_rate": 5.224522008244092e-07,
+      "loss": 0.4934,
+      "step": 38191
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2879021128093897,
+      "learning_rate": 5.222088184639785e-07,
+      "loss": 0.5726,
+      "step": 38192
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4347303252593686,
+      "learning_rate": 5.219654912862482e-07,
+      "loss": 0.7184,
+      "step": 38193
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4947018434783643,
+      "learning_rate": 5.217222192926342e-07,
+      "loss": 0.6364,
+      "step": 38194
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2772050076075319,
+      "learning_rate": 5.214790024845539e-07,
+      "loss": 0.6265,
+      "step": 38195
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5355004587385608,
+      "learning_rate": 5.212358408634266e-07,
+      "loss": 0.6569,
+      "step": 38196
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6540702737691209,
+      "learning_rate": 5.20992734430662e-07,
+      "loss": 0.6671,
+      "step": 38197
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6705058325723225,
+      "learning_rate": 5.207496831876791e-07,
+      "loss": 0.6814,
+      "step": 38198
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.339747032841291,
+      "learning_rate": 5.205066871358932e-07,
+      "loss": 0.5201,
+      "step": 38199
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.54996836580268,
+      "learning_rate": 5.20263746276719e-07,
+      "loss": 0.7058,
+      "step": 38200
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.281690268729912,
+      "learning_rate": 5.200208606115709e-07,
+      "loss": 0.7352,
+      "step": 38201
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4806303214137109,
+      "learning_rate": 5.197780301418609e-07,
+      "loss": 0.7133,
+      "step": 38202
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3818574725498074,
+      "learning_rate": 5.195352548690069e-07,
+      "loss": 0.6481,
+      "step": 38203
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3664119681966076,
+      "learning_rate": 5.192925347944178e-07,
+      "loss": 0.6528,
+      "step": 38204
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4503325792634143,
+      "learning_rate": 5.190498699195112e-07,
+      "loss": 0.697,
+      "step": 38205
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5195720366505472,
+      "learning_rate": 5.188072602456962e-07,
+      "loss": 0.5389,
+      "step": 38206
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4549160838507653,
+      "learning_rate": 5.185647057743881e-07,
+      "loss": 0.8037,
+      "step": 38207
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.152506976917796,
+      "learning_rate": 5.18322206506997e-07,
+      "loss": 0.594,
+      "step": 38208
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3829901387816095,
+      "learning_rate": 5.180797624449363e-07,
+      "loss": 0.7735,
+      "step": 38209
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3057476650786555,
+      "learning_rate": 5.17837373589618e-07,
+      "loss": 0.5968,
+      "step": 38210
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8696183689316876,
+      "learning_rate": 5.175950399424512e-07,
+      "loss": 0.7531,
+      "step": 38211
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4084088680663267,
+      "learning_rate": 5.173527615048491e-07,
+      "loss": 0.6301,
+      "step": 38212
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6175858092840357,
+      "learning_rate": 5.171105382782215e-07,
+      "loss": 0.5395,
+      "step": 38213
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5630676291016508,
+      "learning_rate": 5.168683702639788e-07,
+      "loss": 0.6129,
+      "step": 38214
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4373312988431788,
+      "learning_rate": 5.166262574635295e-07,
+      "loss": 0.681,
+      "step": 38215
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.374419877120271,
+      "learning_rate": 5.163841998782837e-07,
+      "loss": 0.715,
+      "step": 38216
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.451125207994329,
+      "learning_rate": 5.161421975096537e-07,
+      "loss": 0.6545,
+      "step": 38217
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.489764129769943,
+      "learning_rate": 5.159002503590472e-07,
+      "loss": 0.6575,
+      "step": 38218
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5798403890013033,
+      "learning_rate": 5.156583584278696e-07,
+      "loss": 0.7575,
+      "step": 38219
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5929074544679334,
+      "learning_rate": 5.154165217175311e-07,
+      "loss": 0.6739,
+      "step": 38220
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3338140764303703,
+      "learning_rate": 5.151747402294427e-07,
+      "loss": 0.5592,
+      "step": 38221
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6057147379872967,
+      "learning_rate": 5.1493301396501e-07,
+      "loss": 0.5946,
+      "step": 38222
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.447878835029794,
+      "learning_rate": 5.146913429256384e-07,
+      "loss": 0.702,
+      "step": 38223
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.464839972719647,
+      "learning_rate": 5.14449727112738e-07,
+      "loss": 0.7255,
+      "step": 38224
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6669558986746678,
+      "learning_rate": 5.142081665277132e-07,
+      "loss": 0.5346,
+      "step": 38225
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3878107727365463,
+      "learning_rate": 5.13966661171974e-07,
+      "loss": 0.6641,
+      "step": 38226
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.438971219636795,
+      "learning_rate": 5.137252110469215e-07,
+      "loss": 0.7272,
+      "step": 38227
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4933341306144665,
+      "learning_rate": 5.134838161539657e-07,
+      "loss": 0.7281,
+      "step": 38228
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4804900191906771,
+      "learning_rate": 5.132424764945088e-07,
+      "loss": 0.589,
+      "step": 38229
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3623997269634458,
+      "learning_rate": 5.130011920699596e-07,
+      "loss": 0.5739,
+      "step": 38230
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.021890770170751,
+      "learning_rate": 5.127599628817215e-07,
+      "loss": 0.7148,
+      "step": 38231
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4001312187367851,
+      "learning_rate": 5.125187889311966e-07,
+      "loss": 0.6184,
+      "step": 38232
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2203815099830457,
+      "learning_rate": 5.122776702197907e-07,
+      "loss": 0.7358,
+      "step": 38233
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4390723540742276,
+      "learning_rate": 5.120366067489091e-07,
+      "loss": 0.5533,
+      "step": 38234
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3262647551621622,
+      "learning_rate": 5.117955985199552e-07,
+      "loss": 0.7088,
+      "step": 38235
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3593114527161452,
+      "learning_rate": 5.115546455343289e-07,
+      "loss": 0.6619,
+      "step": 38236
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3170463562031605,
+      "learning_rate": 5.11313747793436e-07,
+      "loss": 0.6503,
+      "step": 38237
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4118338762529883,
+      "learning_rate": 5.110729052986796e-07,
+      "loss": 0.7282,
+      "step": 38238
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6059871797493264,
+      "learning_rate": 5.108321180514608e-07,
+      "loss": 0.628,
+      "step": 38239
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3248522905704971,
+      "learning_rate": 5.105913860531808e-07,
+      "loss": 0.6963,
+      "step": 38240
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.1685520954340967,
+      "learning_rate": 5.103507093052418e-07,
+      "loss": 0.6276,
+      "step": 38241
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.905511084621272,
+      "learning_rate": 5.101100878090471e-07,
+      "loss": 0.6271,
+      "step": 38242
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3822065154828413,
+      "learning_rate": 5.098695215659954e-07,
+      "loss": 0.6329,
+      "step": 38243
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4059697948835823,
+      "learning_rate": 5.096290105774859e-07,
+      "loss": 0.7379,
+      "step": 38244
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4869851357517028,
+      "learning_rate": 5.093885548449229e-07,
+      "loss": 0.7729,
+      "step": 38245
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2887010149676228,
+      "learning_rate": 5.091481543697042e-07,
+      "loss": 0.7461,
+      "step": 38246
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4462767952305224,
+      "learning_rate": 5.089078091532296e-07,
+      "loss": 0.7771,
+      "step": 38247
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.267115779291548,
+      "learning_rate": 5.086675191968971e-07,
+      "loss": 0.7175,
+      "step": 38248
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5165996943653137,
+      "learning_rate": 5.084272845021088e-07,
+      "loss": 0.6581,
+      "step": 38249
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5272651066779028,
+      "learning_rate": 5.081871050702614e-07,
+      "loss": 0.6798,
+      "step": 38250
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5627531431166322,
+      "learning_rate": 5.079469809027537e-07,
+      "loss": 0.6859,
+      "step": 38251
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4623069593961042,
+      "learning_rate": 5.077069120009848e-07,
+      "loss": 0.6909,
+      "step": 38252
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.625239603290248,
+      "learning_rate": 5.0746689836635e-07,
+      "loss": 0.6866,
+      "step": 38253
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.628170363311982,
+      "learning_rate": 5.072269400002483e-07,
+      "loss": 0.6345,
+      "step": 38254
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4518502985327946,
+      "learning_rate": 5.069870369040775e-07,
+      "loss": 0.7453,
+      "step": 38255
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4116778362829376,
+      "learning_rate": 5.067471890792341e-07,
+      "loss": 0.6249,
+      "step": 38256
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.754524787397245,
+      "learning_rate": 5.065073965271128e-07,
+      "loss": 0.7051,
+      "step": 38257
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3971214120127842,
+      "learning_rate": 5.062676592491101e-07,
+      "loss": 0.7103,
+      "step": 38258
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4774708381300166,
+      "learning_rate": 5.060279772466248e-07,
+      "loss": 0.7283,
+      "step": 38259
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.484752152541868,
+      "learning_rate": 5.057883505210504e-07,
+      "loss": 0.6869,
+      "step": 38260
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3411537058372167,
+      "learning_rate": 5.055487790737812e-07,
+      "loss": 0.5935,
+      "step": 38261
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5026866515602557,
+      "learning_rate": 5.053092629062117e-07,
+      "loss": 0.7631,
+      "step": 38262
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3089515937593605,
+      "learning_rate": 5.050698020197398e-07,
+      "loss": 0.7802,
+      "step": 38263
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3029405694846503,
+      "learning_rate": 5.048303964157575e-07,
+      "loss": 0.7065,
+      "step": 38264
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.489127751782076,
+      "learning_rate": 5.045910460956572e-07,
+      "loss": 0.6184,
+      "step": 38265
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3133950637483798,
+      "learning_rate": 5.043517510608354e-07,
+      "loss": 0.5651,
+      "step": 38266
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2788752441105922,
+      "learning_rate": 5.041125113126821e-07,
+      "loss": 0.5413,
+      "step": 38267
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6358055201859996,
+      "learning_rate": 5.038733268525941e-07,
+      "loss": 0.4752,
+      "step": 38268
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4700687241351078,
+      "learning_rate": 5.036341976819615e-07,
+      "loss": 0.8119,
+      "step": 38269
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.91105881301898,
+      "learning_rate": 5.033951238021773e-07,
+      "loss": 0.6919,
+      "step": 38270
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4146981838600423,
+      "learning_rate": 5.031561052146316e-07,
+      "loss": 0.6069,
+      "step": 38271
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.264025861141304,
+      "learning_rate": 5.029171419207202e-07,
+      "loss": 0.6875,
+      "step": 38272
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2936344699033575,
+      "learning_rate": 5.026782339218306e-07,
+      "loss": 0.8524,
+      "step": 38273
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.593023992037741,
+      "learning_rate": 5.024393812193551e-07,
+      "loss": 0.7471,
+      "step": 38274
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3824845019905454,
+      "learning_rate": 5.022005838146837e-07,
+      "loss": 0.5579,
+      "step": 38275
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5068731629993959,
+      "learning_rate": 5.019618417092075e-07,
+      "loss": 0.7674,
+      "step": 38276
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5290004796041787,
+      "learning_rate": 5.017231549043189e-07,
+      "loss": 0.7104,
+      "step": 38277
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6185323450453949,
+      "learning_rate": 5.014845234014032e-07,
+      "loss": 0.5115,
+      "step": 38278
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4622422718603751,
+      "learning_rate": 5.012459472018516e-07,
+      "loss": 0.6813,
+      "step": 38279
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5329370035546401,
+      "learning_rate": 5.010074263070541e-07,
+      "loss": 0.6108,
+      "step": 38280
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6662554255977464,
+      "learning_rate": 5.007689607183997e-07,
+      "loss": 0.7817,
+      "step": 38281
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3232606216805878,
+      "learning_rate": 5.005305504372738e-07,
+      "loss": 0.6172,
+      "step": 38282
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5540219112718826,
+      "learning_rate": 5.002921954650664e-07,
+      "loss": 0.5067,
+      "step": 38283
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5728711175377214,
+      "learning_rate": 5.000538958031664e-07,
+      "loss": 0.5311,
+      "step": 38284
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5275186298103314,
+      "learning_rate": 4.998156514529595e-07,
+      "loss": 0.5713,
+      "step": 38285
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2796120213584825,
+      "learning_rate": 4.995774624158356e-07,
+      "loss": 0.5973,
+      "step": 38286
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9139808996919763,
+      "learning_rate": 4.993393286931781e-07,
+      "loss": 0.6619,
+      "step": 38287
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5464645664509262,
+      "learning_rate": 4.991012502863745e-07,
+      "loss": 0.655,
+      "step": 38288
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.058846738280444,
+      "learning_rate": 4.988632271968108e-07,
+      "loss": 0.6233,
+      "step": 38289
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3956618475108347,
+      "learning_rate": 4.986252594258756e-07,
+      "loss": 0.5805,
+      "step": 38290
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2784509087239746,
+      "learning_rate": 4.983873469749511e-07,
+      "loss": 0.6322,
+      "step": 38291
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4623957586757135,
+      "learning_rate": 4.981494898454232e-07,
+      "loss": 0.6288,
+      "step": 38292
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7190051567656928,
+      "learning_rate": 4.97911688038677e-07,
+      "loss": 0.767,
+      "step": 38293
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3271646326936894,
+      "learning_rate": 4.976739415560982e-07,
+      "loss": 0.6209,
+      "step": 38294
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5876586593580891,
+      "learning_rate": 4.974362503990704e-07,
+      "loss": 0.48,
+      "step": 38295
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1036452928036335,
+      "learning_rate": 4.971986145689756e-07,
+      "loss": 0.6947,
+      "step": 38296
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4878201273281912,
+      "learning_rate": 4.969610340671993e-07,
+      "loss": 0.6685,
+      "step": 38297
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.238581826319628,
+      "learning_rate": 4.967235088951261e-07,
+      "loss": 0.7549,
+      "step": 38298
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6683010614991576,
+      "learning_rate": 4.96486039054136e-07,
+      "loss": 0.7134,
+      "step": 38299
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2368744325518866,
+      "learning_rate": 4.962486245456133e-07,
+      "loss": 0.6441,
+      "step": 38300
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3048159252930602,
+      "learning_rate": 4.960112653709392e-07,
+      "loss": 0.53,
+      "step": 38301
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5761339349272117,
+      "learning_rate": 4.957739615314972e-07,
+      "loss": 0.5036,
+      "step": 38302
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4174544412179317,
+      "learning_rate": 4.955367130286692e-07,
+      "loss": 0.539,
+      "step": 38303
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6292735121868059,
+      "learning_rate": 4.952995198638333e-07,
+      "loss": 0.5237,
+      "step": 38304
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2999912598785468,
+      "learning_rate": 4.950623820383749e-07,
+      "loss": 0.6763,
+      "step": 38305
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1675947491986178,
+      "learning_rate": 4.948252995536707e-07,
+      "loss": 0.6254,
+      "step": 38306
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2716579470427016,
+      "learning_rate": 4.94588272411104e-07,
+      "loss": 0.6282,
+      "step": 38307
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3040538561871828,
+      "learning_rate": 4.943513006120526e-07,
+      "loss": 0.5311,
+      "step": 38308
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6452240603217332,
+      "learning_rate": 4.941143841578988e-07,
+      "loss": 0.6878,
+      "step": 38309
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2238118899057324,
+      "learning_rate": 4.938775230500192e-07,
+      "loss": 0.5309,
+      "step": 38310
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2995733870831003,
+      "learning_rate": 4.936407172897961e-07,
+      "loss": 0.7289,
+      "step": 38311
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6234165931130186,
+      "learning_rate": 4.934039668786061e-07,
+      "loss": 0.5183,
+      "step": 38312
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.2593620474931377,
+      "learning_rate": 4.931672718178259e-07,
+      "loss": 0.637,
+      "step": 38313
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6011997262014127,
+      "learning_rate": 4.929306321088368e-07,
+      "loss": 0.5056,
+      "step": 38314
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6953483766895416,
+      "learning_rate": 4.926940477530151e-07,
+      "loss": 0.5542,
+      "step": 38315
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3974650792877565,
+      "learning_rate": 4.924575187517399e-07,
+      "loss": 0.6571,
+      "step": 38316
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2592281189310535,
+      "learning_rate": 4.922210451063858e-07,
+      "loss": 0.5938,
+      "step": 38317
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.060556211636846,
+      "learning_rate": 4.919846268183303e-07,
+      "loss": 0.7726,
+      "step": 38318
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6019672467117155,
+      "learning_rate": 4.917482638889525e-07,
+      "loss": 0.4831,
+      "step": 38319
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5552785975788498,
+      "learning_rate": 4.915119563196258e-07,
+      "loss": 0.4745,
+      "step": 38320
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3968231930243244,
+      "learning_rate": 4.912757041117266e-07,
+      "loss": 0.683,
+      "step": 38321
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3924307140231782,
+      "learning_rate": 4.910395072666297e-07,
+      "loss": 0.6093,
+      "step": 38322
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6345272208957844,
+      "learning_rate": 4.908033657857125e-07,
+      "loss": 0.6496,
+      "step": 38323
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2727628953740724,
+      "learning_rate": 4.905672796703498e-07,
+      "loss": 0.6889,
+      "step": 38324
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3508290459397794,
+      "learning_rate": 4.903312489219125e-07,
+      "loss": 0.5598,
+      "step": 38325
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3015026059366315,
+      "learning_rate": 4.900952735417786e-07,
+      "loss": 0.5763,
+      "step": 38326
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3733875455520952,
+      "learning_rate": 4.898593535313201e-07,
+      "loss": 0.6273,
+      "step": 38327
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4339775195454898,
+      "learning_rate": 4.896234888919127e-07,
+      "loss": 0.4544,
+      "step": 38328
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.940957105837309,
+      "learning_rate": 4.893876796249264e-07,
+      "loss": 0.7733,
+      "step": 38329
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.304273184599698,
+      "learning_rate": 4.891519257317379e-07,
+      "loss": 0.7968,
+      "step": 38330
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2800698982148078,
+      "learning_rate": 4.889162272137171e-07,
+      "loss": 0.6138,
+      "step": 38331
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2855357492350725,
+      "learning_rate": 4.886805840722375e-07,
+      "loss": 0.5946,
+      "step": 38332
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0856719270867785,
+      "learning_rate": 4.884449963086712e-07,
+      "loss": 0.6897,
+      "step": 38333
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4119722467744822,
+      "learning_rate": 4.882094639243884e-07,
+      "loss": 0.7144,
+      "step": 38334
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4681113713490193,
+      "learning_rate": 4.879739869207611e-07,
+      "loss": 0.7542,
+      "step": 38335
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3207944181799318,
+      "learning_rate": 4.877385652991618e-07,
+      "loss": 0.7226,
+      "step": 38336
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.346809051060955,
+      "learning_rate": 4.875031990609602e-07,
+      "loss": 0.6199,
+      "step": 38337
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3354076879087264,
+      "learning_rate": 4.872678882075255e-07,
+      "loss": 0.7546,
+      "step": 38338
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6159308492168559,
+      "learning_rate": 4.870326327402297e-07,
+      "loss": 0.5176,
+      "step": 38339
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6130540362551238,
+      "learning_rate": 4.867974326604419e-07,
+      "loss": 0.5018,
+      "step": 38340
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3824962472204536,
+      "learning_rate": 4.865622879695309e-07,
+      "loss": 0.6747,
+      "step": 38341
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.309912322572845,
+      "learning_rate": 4.863271986688656e-07,
+      "loss": 0.701,
+      "step": 38342
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6253975618904302,
+      "learning_rate": 4.860921647598149e-07,
+      "loss": 0.6825,
+      "step": 38343
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.653803889983023,
+      "learning_rate": 4.8585718624375e-07,
+      "loss": 0.7444,
+      "step": 38344
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2247535155379063,
+      "learning_rate": 4.856222631220354e-07,
+      "loss": 0.599,
+      "step": 38345
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4839890476002404,
+      "learning_rate": 4.853873953960397e-07,
+      "loss": 0.6679,
+      "step": 38346
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2641544996149683,
+      "learning_rate": 4.851525830671322e-07,
+      "loss": 0.6857,
+      "step": 38347
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.02351489956797,
+      "learning_rate": 4.849178261366771e-07,
+      "loss": 0.7768,
+      "step": 38348
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4294170002986877,
+      "learning_rate": 4.846831246060435e-07,
+      "loss": 0.5649,
+      "step": 38349
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.492610289546153,
+      "learning_rate": 4.844484784765968e-07,
+      "loss": 0.7554,
+      "step": 38350
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5908434956165661,
+      "learning_rate": 4.842138877497049e-07,
+      "loss": 0.4714,
+      "step": 38351
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4020774296916438,
+      "learning_rate": 4.839793524267311e-07,
+      "loss": 0.6528,
+      "step": 38352
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3828877918234852,
+      "learning_rate": 4.837448725090433e-07,
+      "loss": 0.7002,
+      "step": 38353
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5227585919969757,
+      "learning_rate": 4.835104479980058e-07,
+      "loss": 0.7128,
+      "step": 38354
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6414018535972879,
+      "learning_rate": 4.832760788949809e-07,
+      "loss": 0.7204,
+      "step": 38355
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3274436094196598,
+      "learning_rate": 4.830417652013364e-07,
+      "loss": 0.7833,
+      "step": 38356
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4612521806972087,
+      "learning_rate": 4.828075069184379e-07,
+      "loss": 0.4777,
+      "step": 38357
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2436114955152968,
+      "learning_rate": 4.825733040476465e-07,
+      "loss": 0.6334,
+      "step": 38358
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8013210688305303,
+      "learning_rate": 4.823391565903257e-07,
+      "loss": 0.7815,
+      "step": 38359
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3354450064057442,
+      "learning_rate": 4.821050645478386e-07,
+      "loss": 0.6885,
+      "step": 38360
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3107118720112678,
+      "learning_rate": 4.81871027921551e-07,
+      "loss": 0.6045,
+      "step": 38361
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.467988069901245,
+      "learning_rate": 4.81637046712824e-07,
+      "loss": 0.752,
+      "step": 38362
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4543954896013624,
+      "learning_rate": 4.814031209230175e-07,
+      "loss": 0.6386,
+      "step": 38363
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3842402840502637,
+      "learning_rate": 4.811692505534959e-07,
+      "loss": 0.7183,
+      "step": 38364
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2650053378045694,
+      "learning_rate": 4.809354356056217e-07,
+      "loss": 0.7209,
+      "step": 38365
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5311022574338629,
+      "learning_rate": 4.807016760807559e-07,
+      "loss": 0.689,
+      "step": 38366
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3122360657152823,
+      "learning_rate": 4.804679719802563e-07,
+      "loss": 0.6515,
+      "step": 38367
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5306734967609537,
+      "learning_rate": 4.802343233054884e-07,
+      "loss": 0.6106,
+      "step": 38368
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3986390780265923,
+      "learning_rate": 4.80000730057808e-07,
+      "loss": 0.8415,
+      "step": 38369
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5660110760749858,
+      "learning_rate": 4.797671922385783e-07,
+      "loss": 0.4842,
+      "step": 38370
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2827977263338106,
+      "learning_rate": 4.795337098491582e-07,
+      "loss": 0.6643,
+      "step": 38371
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.722878793641476,
+      "learning_rate": 4.793002828909066e-07,
+      "loss": 0.704,
+      "step": 38372
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4377040823954947,
+      "learning_rate": 4.790669113651825e-07,
+      "loss": 0.6649,
+      "step": 38373
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.286620076099252,
+      "learning_rate": 4.788335952733459e-07,
+      "loss": 0.6269,
+      "step": 38374
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2725931844944753,
+      "learning_rate": 4.786003346167556e-07,
+      "loss": 0.5088,
+      "step": 38375
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5625356635199703,
+      "learning_rate": 4.783671293967673e-07,
+      "loss": 0.7263,
+      "step": 38376
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3309165727427974,
+      "learning_rate": 4.781339796147389e-07,
+      "loss": 0.5721,
+      "step": 38377
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3623510241212256,
+      "learning_rate": 4.77900885272029e-07,
+      "loss": 0.5293,
+      "step": 38378
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4833884083273534,
+      "learning_rate": 4.77667846369998e-07,
+      "loss": 0.6454,
+      "step": 38379
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3438104831235107,
+      "learning_rate": 4.774348629099967e-07,
+      "loss": 0.6146,
+      "step": 38380
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5038647143651358,
+      "learning_rate": 4.772019348933854e-07,
+      "loss": 0.7133,
+      "step": 38381
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5492037006907753,
+      "learning_rate": 4.769690623215195e-07,
+      "loss": 0.6136,
+      "step": 38382
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5533588472204163,
+      "learning_rate": 4.7673624519575355e-07,
+      "loss": 0.5672,
+      "step": 38383
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2773744487231453,
+      "learning_rate": 4.7650348351744533e-07,
+      "loss": 0.6941,
+      "step": 38384
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3069098670789345,
+      "learning_rate": 4.7627077728794824e-07,
+      "loss": 0.5761,
+      "step": 38385
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.329608636547978,
+      "learning_rate": 4.7603812650861893e-07,
+      "loss": 0.6602,
+      "step": 38386
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5280064408489327,
+      "learning_rate": 4.758055311808096e-07,
+      "loss": 0.7772,
+      "step": 38387
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3634368464536637,
+      "learning_rate": 4.755729913058771e-07,
+      "loss": 0.5325,
+      "step": 38388
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5704449589252638,
+      "learning_rate": 4.753405068851735e-07,
+      "loss": 0.9042,
+      "step": 38389
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4557278635886375,
+      "learning_rate": 4.751080779200512e-07,
+      "loss": 0.7438,
+      "step": 38390
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3222747377220019,
+      "learning_rate": 4.748757044118668e-07,
+      "loss": 0.7369,
+      "step": 38391
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3008762466245831,
+      "learning_rate": 4.746433863619715e-07,
+      "loss": 0.7912,
+      "step": 38392
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2664508433031627,
+      "learning_rate": 4.744111237717186e-07,
+      "loss": 0.6051,
+      "step": 38393
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6834740775809542,
+      "learning_rate": 4.741789166424593e-07,
+      "loss": 0.6862,
+      "step": 38394
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.060141406932225,
+      "learning_rate": 4.739467649755447e-07,
+      "loss": 0.6245,
+      "step": 38395
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6017657061647357,
+      "learning_rate": 4.7371466877233043e-07,
+      "loss": 0.6847,
+      "step": 38396
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3865963459875057,
+      "learning_rate": 4.734826280341653e-07,
+      "loss": 0.7357,
+      "step": 38397
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.518854968722449,
+      "learning_rate": 4.732506427623984e-07,
+      "loss": 0.6356,
+      "step": 38398
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3942326729484764,
+      "learning_rate": 4.7301871295838297e-07,
+      "loss": 0.6399,
+      "step": 38399
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.649601683748249,
+      "learning_rate": 4.727868386234702e-07,
+      "loss": 0.6631,
+      "step": 38400
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6271635265780273,
+      "learning_rate": 4.725550197590079e-07,
+      "loss": 0.801,
+      "step": 38401
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4949230164791298,
+      "learning_rate": 4.7232325636634603e-07,
+      "loss": 0.6223,
+      "step": 38402
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8630077547739634,
+      "learning_rate": 4.720915484468358e-07,
+      "loss": 0.6871,
+      "step": 38403
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6050223771811258,
+      "learning_rate": 4.718598960018228e-07,
+      "loss": 0.4839,
+      "step": 38404
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6098660003061857,
+      "learning_rate": 4.716282990326604e-07,
+      "loss": 0.5335,
+      "step": 38405
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.468840299014019,
+      "learning_rate": 4.7139675754069304e-07,
+      "loss": 0.698,
+      "step": 38406
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.513129177808392,
+      "learning_rate": 4.7116527152727187e-07,
+      "loss": 0.7602,
+      "step": 38407
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7556204687673629,
+      "learning_rate": 4.709338409937414e-07,
+      "loss": 0.7165,
+      "step": 38408
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4728971693821602,
+      "learning_rate": 4.7070246594145276e-07,
+      "loss": 0.6853,
+      "step": 38409
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4797522176978426,
+      "learning_rate": 4.704711463717515e-07,
+      "loss": 0.6429,
+      "step": 38410
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7594206633767093,
+      "learning_rate": 4.7023988228598326e-07,
+      "loss": 0.8217,
+      "step": 38411
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3122398324919227,
+      "learning_rate": 4.7000867368549473e-07,
+      "loss": 0.7126,
+      "step": 38412
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.350744817297271,
+      "learning_rate": 4.697775205716348e-07,
+      "loss": 0.634,
+      "step": 38413
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5656928645172398,
+      "learning_rate": 4.695464229457469e-07,
+      "loss": 0.4537,
+      "step": 38414
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7038350758711294,
+      "learning_rate": 4.693153808091755e-07,
+      "loss": 0.6208,
+      "step": 38415
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2222039703062846,
+      "learning_rate": 4.690843941632672e-07,
+      "loss": 0.5503,
+      "step": 38416
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4218781903186575,
+      "learning_rate": 4.6885346300936885e-07,
+      "loss": 0.5343,
+      "step": 38417
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2424330357933813,
+      "learning_rate": 4.686225873488226e-07,
+      "loss": 0.6942,
+      "step": 38418
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.389753885833608,
+      "learning_rate": 4.6839176718297297e-07,
+      "loss": 0.6303,
+      "step": 38419
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5818550136368372,
+      "learning_rate": 4.6816100251316333e-07,
+      "loss": 0.5102,
+      "step": 38420
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6428792295215654,
+      "learning_rate": 4.6793029334073926e-07,
+      "loss": 0.5588,
+      "step": 38421
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3792113505586405,
+      "learning_rate": 4.67699639667043e-07,
+      "loss": 0.6736,
+      "step": 38422
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8905478698624183,
+      "learning_rate": 4.6746904149341576e-07,
+      "loss": 0.7245,
+      "step": 38423
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5904411211128803,
+      "learning_rate": 4.67238498821202e-07,
+      "loss": 0.5144,
+      "step": 38424
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3895889955909353,
+      "learning_rate": 4.6700801165174504e-07,
+      "loss": 0.6606,
+      "step": 38425
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6372688229284467,
+      "learning_rate": 4.667775799863861e-07,
+      "loss": 0.7783,
+      "step": 38426
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4703703465388007,
+      "learning_rate": 4.66547203826464e-07,
+      "loss": 0.6859,
+      "step": 38427
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.42459205506382,
+      "learning_rate": 4.6631688317332447e-07,
+      "loss": 0.7096,
+      "step": 38428
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3196633044489339,
+      "learning_rate": 4.660866180283041e-07,
+      "loss": 0.6799,
+      "step": 38429
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4632946857747704,
+      "learning_rate": 4.658564083927475e-07,
+      "loss": 0.7126,
+      "step": 38430
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4957324474261653,
+      "learning_rate": 4.6562625426799237e-07,
+      "loss": 0.7971,
+      "step": 38431
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.260452807476963,
+      "learning_rate": 4.65396155655381e-07,
+      "loss": 0.7284,
+      "step": 38432
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5948218638290389,
+      "learning_rate": 4.651661125562501e-07,
+      "loss": 0.4892,
+      "step": 38433
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5620702882148148,
+      "learning_rate": 4.64936124971942e-07,
+      "loss": 0.5324,
+      "step": 38434
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4792879413027147,
+      "learning_rate": 4.6470619290379547e-07,
+      "loss": 0.6922,
+      "step": 38435
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.491246898876946,
+      "learning_rate": 4.6447631635314626e-07,
+      "loss": 0.7685,
+      "step": 38436
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5808828889292066,
+      "learning_rate": 4.6424649532133437e-07,
+      "loss": 0.4605,
+      "step": 38437
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1797218903610895,
+      "learning_rate": 4.640167298097009e-07,
+      "loss": 0.6427,
+      "step": 38438
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.399754722371496,
+      "learning_rate": 4.637870198195804e-07,
+      "loss": 0.7234,
+      "step": 38439
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4367801174290773,
+      "learning_rate": 4.6355736535230956e-07,
+      "loss": 0.8174,
+      "step": 38440
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5933823695461784,
+      "learning_rate": 4.633277664092262e-07,
+      "loss": 0.6269,
+      "step": 38441
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7505996216794348,
+      "learning_rate": 4.630982229916703e-07,
+      "loss": 0.6085,
+      "step": 38442
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5654282002276878,
+      "learning_rate": 4.628687351009753e-07,
+      "loss": 0.6109,
+      "step": 38443
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2455796180804968,
+      "learning_rate": 4.626393027384768e-07,
+      "loss": 0.6364,
+      "step": 38444
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.607668404180929,
+      "learning_rate": 4.624099259055126e-07,
+      "loss": 0.5989,
+      "step": 38445
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4832061272461858,
+      "learning_rate": 4.6218060460341716e-07,
+      "loss": 0.6557,
+      "step": 38446
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4331868549568172,
+      "learning_rate": 4.619513388335273e-07,
+      "loss": 0.6807,
+      "step": 38447
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3012015425037617,
+      "learning_rate": 4.6172212859717403e-07,
+      "loss": 0.5767,
+      "step": 38448
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6254421643412542,
+      "learning_rate": 4.6149297389569634e-07,
+      "loss": 0.5131,
+      "step": 38449
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4440851106089843,
+      "learning_rate": 4.612638747304243e-07,
+      "loss": 0.7294,
+      "step": 38450
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3882745103100231,
+      "learning_rate": 4.6103483110269575e-07,
+      "loss": 0.5936,
+      "step": 38451
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4446792891224516,
+      "learning_rate": 4.608058430138407e-07,
+      "loss": 0.7155,
+      "step": 38452
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.269766992049505,
+      "learning_rate": 4.605769104651958e-07,
+      "loss": 0.6992,
+      "step": 38453
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2206412966686675,
+      "learning_rate": 4.603480334580912e-07,
+      "loss": 0.5545,
+      "step": 38454
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8601208675046021,
+      "learning_rate": 4.6011921199386136e-07,
+      "loss": 0.6331,
+      "step": 38455
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2845239684423335,
+      "learning_rate": 4.5989044607383846e-07,
+      "loss": 0.6625,
+      "step": 38456
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.557860407386625,
+      "learning_rate": 4.5966173569935156e-07,
+      "loss": 0.7659,
+      "step": 38457
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4598551132528792,
+      "learning_rate": 4.594330808717351e-07,
+      "loss": 0.74,
+      "step": 38458
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3230069541261733,
+      "learning_rate": 4.592044815923191e-07,
+      "loss": 0.7131,
+      "step": 38459
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.340292863369074,
+      "learning_rate": 4.589759378624381e-07,
+      "loss": 0.6315,
+      "step": 38460
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8237412007780804,
+      "learning_rate": 4.587474496834166e-07,
+      "loss": 0.7834,
+      "step": 38461
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5218793919437836,
+      "learning_rate": 4.58519017056589e-07,
+      "loss": 0.7068,
+      "step": 38462
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7943503240724412,
+      "learning_rate": 4.582906399832854e-07,
+      "loss": 0.6529,
+      "step": 38463
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5214090110750484,
+      "learning_rate": 4.5806231846483475e-07,
+      "loss": 0.7818,
+      "step": 38464
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3579039621966316,
+      "learning_rate": 4.5783405250256487e-07,
+      "loss": 0.6562,
+      "step": 38465
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4473508460324838,
+      "learning_rate": 4.5760584209780577e-07,
+      "loss": 0.6849,
+      "step": 38466
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9202172983294206,
+      "learning_rate": 4.573776872518887e-07,
+      "loss": 0.722,
+      "step": 38467
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5817133557255239,
+      "learning_rate": 4.571495879661392e-07,
+      "loss": 0.5381,
+      "step": 38468
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2586787164607476,
+      "learning_rate": 4.5692154424188506e-07,
+      "loss": 0.6679,
+      "step": 38469
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6159772067715183,
+      "learning_rate": 4.566935560804564e-07,
+      "loss": 0.7731,
+      "step": 38470
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6640274443604663,
+      "learning_rate": 4.5646562348317657e-07,
+      "loss": 0.4805,
+      "step": 38471
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.53087761209632,
+      "learning_rate": 4.5623774645137675e-07,
+      "loss": 0.738,
+      "step": 38472
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6202565250412542,
+      "learning_rate": 4.5600992498638364e-07,
+      "loss": 0.4846,
+      "step": 38473
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2674649480504752,
+      "learning_rate": 4.557821590895217e-07,
+      "loss": 0.5826,
+      "step": 38474
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5386358613159767,
+      "learning_rate": 4.5555444876211664e-07,
+      "loss": 0.655,
+      "step": 38475
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5933226100481752,
+      "learning_rate": 4.553267940054951e-07,
+      "loss": 0.5187,
+      "step": 38476
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1698637507932792,
+      "learning_rate": 4.5509919482098486e-07,
+      "loss": 0.6246,
+      "step": 38477
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.520708059885678,
+      "learning_rate": 4.5487165120990716e-07,
+      "loss": 0.6601,
+      "step": 38478
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3969486592680722,
+      "learning_rate": 4.546441631735887e-07,
+      "loss": 0.7154,
+      "step": 38479
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.599098248218462,
+      "learning_rate": 4.544167307133529e-07,
+      "loss": 0.791,
+      "step": 38480
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3722326488099823,
+      "learning_rate": 4.5418935383052756e-07,
+      "loss": 0.7712,
+      "step": 38481
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.730246495242968,
+      "learning_rate": 4.5396203252643377e-07,
+      "loss": 0.7403,
+      "step": 38482
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3851103194338874,
+      "learning_rate": 4.5373476680239283e-07,
+      "loss": 0.7517,
+      "step": 38483
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.341550620887001,
+      "learning_rate": 4.5350755665973243e-07,
+      "loss": 0.6851,
+      "step": 38484
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3691780031704683,
+      "learning_rate": 4.532804020997728e-07,
+      "loss": 0.6593,
+      "step": 38485
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5981762756711466,
+      "learning_rate": 4.5305330312383713e-07,
+      "loss": 0.5134,
+      "step": 38486
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3176914436398925,
+      "learning_rate": 4.5282625973324666e-07,
+      "loss": 0.6306,
+      "step": 38487
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.7282767446150835,
+      "learning_rate": 4.52599271929326e-07,
+      "loss": 0.675,
+      "step": 38488
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.454112311521479,
+      "learning_rate": 4.523723397133939e-07,
+      "loss": 0.7049,
+      "step": 38489
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3599326914309446,
+      "learning_rate": 4.5214546308677386e-07,
+      "loss": 0.5506,
+      "step": 38490
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3370064683297478,
+      "learning_rate": 4.5191864205078596e-07,
+      "loss": 0.6344,
+      "step": 38491
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3130980270904404,
+      "learning_rate": 4.516918766067491e-07,
+      "loss": 0.6931,
+      "step": 38492
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3831763162816064,
+      "learning_rate": 4.514651667559844e-07,
+      "loss": 0.7369,
+      "step": 38493
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.351649027562684,
+      "learning_rate": 4.512385124998142e-07,
+      "loss": 0.7264,
+      "step": 38494
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6357885531184871,
+      "learning_rate": 4.510119138395563e-07,
+      "loss": 0.4698,
+      "step": 38495
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.646153919468492,
+      "learning_rate": 4.5078537077652975e-07,
+      "loss": 0.7165,
+      "step": 38496
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4625586100827057,
+      "learning_rate": 4.5055888331205335e-07,
+      "loss": 0.7575,
+      "step": 38497
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6000376974443854,
+      "learning_rate": 4.503324514474483e-07,
+      "loss": 0.5125,
+      "step": 38498
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3049537862445508,
+      "learning_rate": 4.5010607518403024e-07,
+      "loss": 0.6895,
+      "step": 38499
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.199813717662121,
+      "learning_rate": 4.498797545231171e-07,
+      "loss": 0.5316,
+      "step": 38500
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4327055519517546,
+      "learning_rate": 4.496534894660276e-07,
+      "loss": 0.6741,
+      "step": 38501
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4237313894053596,
+      "learning_rate": 4.4942728001408087e-07,
+      "loss": 0.7117,
+      "step": 38502
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6106313372023063,
+      "learning_rate": 4.4920112616859135e-07,
+      "loss": 0.4985,
+      "step": 38503
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.608732584862381,
+      "learning_rate": 4.4897502793087576e-07,
+      "loss": 0.6803,
+      "step": 38504
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3751981715970691,
+      "learning_rate": 4.487489853022531e-07,
+      "loss": 0.6835,
+      "step": 38505
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.906595863713573,
+      "learning_rate": 4.485229982840356e-07,
+      "loss": 0.6696,
+      "step": 38506
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.287187065579942,
+      "learning_rate": 4.4829706687754213e-07,
+      "loss": 0.6424,
+      "step": 38507
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6077356642210712,
+      "learning_rate": 4.480711910840874e-07,
+      "loss": 0.4924,
+      "step": 38508
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5061530674390236,
+      "learning_rate": 4.478453709049857e-07,
+      "loss": 0.6711,
+      "step": 38509
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4601139793052451,
+      "learning_rate": 4.476196063415528e-07,
+      "loss": 0.8275,
+      "step": 38510
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.331056637703075,
+      "learning_rate": 4.4739389739510306e-07,
+      "loss": 0.6762,
+      "step": 38511
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2450971101446429,
+      "learning_rate": 4.4716824406695005e-07,
+      "loss": 0.7047,
+      "step": 38512
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3275034850791194,
+      "learning_rate": 4.4694264635840814e-07,
+      "loss": 0.6256,
+      "step": 38513
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4478914352343044,
+      "learning_rate": 4.467171042707896e-07,
+      "loss": 0.7001,
+      "step": 38514
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6162555192405972,
+      "learning_rate": 4.4649161780540907e-07,
+      "loss": 0.5205,
+      "step": 38515
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5835545613809687,
+      "learning_rate": 4.4626618696357983e-07,
+      "loss": 0.6657,
+      "step": 38516
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.476079183090861,
+      "learning_rate": 4.460408117466131e-07,
+      "loss": 0.6734,
+      "step": 38517
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3990532635248019,
+      "learning_rate": 4.4581549215582e-07,
+      "loss": 0.617,
+      "step": 38518
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3171532635857177,
+      "learning_rate": 4.455902281925162e-07,
+      "loss": 0.5576,
+      "step": 38519
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4183060918863042,
+      "learning_rate": 4.453650198580106e-07,
+      "loss": 0.5159,
+      "step": 38520
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4086958653211679,
+      "learning_rate": 4.4513986715361336e-07,
+      "loss": 0.6856,
+      "step": 38521
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6178256244170934,
+      "learning_rate": 4.4491477008063666e-07,
+      "loss": 0.6805,
+      "step": 38522
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4628221039766118,
+      "learning_rate": 4.446897286403928e-07,
+      "loss": 0.6695,
+      "step": 38523
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5288826425895417,
+      "learning_rate": 4.444647428341908e-07,
+      "loss": 0.781,
+      "step": 38524
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2410021163000464,
+      "learning_rate": 4.4423981266333847e-07,
+      "loss": 0.6489,
+      "step": 38525
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5942913465372597,
+      "learning_rate": 4.440149381291492e-07,
+      "loss": 0.6776,
+      "step": 38526
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4883099762771383,
+      "learning_rate": 4.437901192329286e-07,
+      "loss": 0.7159,
+      "step": 38527
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4976440477077,
+      "learning_rate": 4.435653559759878e-07,
+      "loss": 0.6254,
+      "step": 38528
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.596633726991094,
+      "learning_rate": 4.4334064835963475e-07,
+      "loss": 0.6381,
+      "step": 38529
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5366313629027712,
+      "learning_rate": 4.4311599638517944e-07,
+      "loss": 0.6455,
+      "step": 38530
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5797626223318606,
+      "learning_rate": 4.4289140005392637e-07,
+      "loss": 0.5067,
+      "step": 38531
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2692806302395663,
+      "learning_rate": 4.426668593671879e-07,
+      "loss": 0.7341,
+      "step": 38532
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.359191121718228,
+      "learning_rate": 4.4244237432626737e-07,
+      "loss": 0.6982,
+      "step": 38533
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.790764292133235,
+      "learning_rate": 4.422179449324726e-07,
+      "loss": 0.5696,
+      "step": 38534
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6190682753649089,
+      "learning_rate": 4.419935711871115e-07,
+      "loss": 0.5422,
+      "step": 38535
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4997695350710418,
+      "learning_rate": 4.417692530914908e-07,
+      "loss": 0.6839,
+      "step": 38536
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4775479530299742,
+      "learning_rate": 4.415449906469149e-07,
+      "loss": 0.6988,
+      "step": 38537
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1436179701916434,
+      "learning_rate": 4.4132078385468956e-07,
+      "loss": 0.643,
+      "step": 38538
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4156831630775246,
+      "learning_rate": 4.4109663271612036e-07,
+      "loss": 0.5353,
+      "step": 38539
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3808985988201108,
+      "learning_rate": 4.40872537232514e-07,
+      "loss": 0.707,
+      "step": 38540
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4781236113414702,
+      "learning_rate": 4.40648497405175e-07,
+      "loss": 0.6733,
+      "step": 38541
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.275783589753,
+      "learning_rate": 4.4042451323540457e-07,
+      "loss": 0.5854,
+      "step": 38542
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4432226106233952,
+      "learning_rate": 4.4020058472450946e-07,
+      "loss": 0.6735,
+      "step": 38543
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2622230828216665,
+      "learning_rate": 4.399767118737952e-07,
+      "loss": 0.6268,
+      "step": 38544
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2956530876088432,
+      "learning_rate": 4.3975289468456194e-07,
+      "loss": 0.7851,
+      "step": 38545
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3603264452858628,
+      "learning_rate": 4.395291331581142e-07,
+      "loss": 0.7208,
+      "step": 38546
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1943882763833704,
+      "learning_rate": 4.393054272957531e-07,
+      "loss": 0.7031,
+      "step": 38547
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5123721418927047,
+      "learning_rate": 4.390817770987854e-07,
+      "loss": 0.5831,
+      "step": 38548
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3709879199809325,
+      "learning_rate": 4.38858182568509e-07,
+      "loss": 0.6636,
+      "step": 38549
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4891998980646703,
+      "learning_rate": 4.3863464370622723e-07,
+      "loss": 0.6121,
+      "step": 38550
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5731061322392859,
+      "learning_rate": 4.384111605132424e-07,
+      "loss": 0.6759,
+      "step": 38551
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2877913947434605,
+      "learning_rate": 4.381877329908535e-07,
+      "loss": 0.7187,
+      "step": 38552
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6611193329740326,
+      "learning_rate": 4.37964361140365e-07,
+      "loss": 0.5722,
+      "step": 38553
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4364193462611614,
+      "learning_rate": 4.377410449630737e-07,
+      "loss": 0.6218,
+      "step": 38554
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7299770620459494,
+      "learning_rate": 4.375177844602818e-07,
+      "loss": 0.7218,
+      "step": 38555
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6858207273074786,
+      "learning_rate": 4.3729457963328837e-07,
+      "loss": 0.6964,
+      "step": 38556
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3845834685177965,
+      "learning_rate": 4.370714304833923e-07,
+      "loss": 0.7226,
+      "step": 38557
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.573844999107816,
+      "learning_rate": 4.368483370118981e-07,
+      "loss": 0.4608,
+      "step": 38558
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3461846550936318,
+      "learning_rate": 4.36625299220097e-07,
+      "loss": 0.7968,
+      "step": 38559
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3974333046738439,
+      "learning_rate": 4.3640231710929124e-07,
+      "loss": 0.7147,
+      "step": 38560
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5661816464756855,
+      "learning_rate": 4.3617939068077874e-07,
+      "loss": 0.7331,
+      "step": 38561
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3273810995933266,
+      "learning_rate": 4.359565199358606e-07,
+      "loss": 0.685,
+      "step": 38562
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4220415122820351,
+      "learning_rate": 4.357337048758292e-07,
+      "loss": 0.6818,
+      "step": 38563
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.374573708003695,
+      "learning_rate": 4.355109455019835e-07,
+      "loss": 0.6964,
+      "step": 38564
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.757784477646309,
+      "learning_rate": 4.3528824181562237e-07,
+      "loss": 0.7331,
+      "step": 38565
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4599482010481974,
+      "learning_rate": 4.350655938180426e-07,
+      "loss": 0.713,
+      "step": 38566
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.375447138682033,
+      "learning_rate": 4.348430015105365e-07,
+      "loss": 0.7132,
+      "step": 38567
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4741247850379298,
+      "learning_rate": 4.34620464894403e-07,
+      "loss": 0.8557,
+      "step": 38568
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1155310552184394,
+      "learning_rate": 4.343979839709389e-07,
+      "loss": 0.6326,
+      "step": 38569
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3311302194414198,
+      "learning_rate": 4.3417555874143644e-07,
+      "loss": 0.7,
+      "step": 38570
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3205757964370548,
+      "learning_rate": 4.3395318920719465e-07,
+      "loss": 0.7887,
+      "step": 38571
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2096705247659627,
+      "learning_rate": 4.3373087536950574e-07,
+      "loss": 0.6118,
+      "step": 38572
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4546683444303632,
+      "learning_rate": 4.335086172296621e-07,
+      "loss": 0.5512,
+      "step": 38573
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5926474630624925,
+      "learning_rate": 4.3328641478896037e-07,
+      "loss": 0.7085,
+      "step": 38574
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3440405332430223,
+      "learning_rate": 4.330642680486952e-07,
+      "loss": 0.6562,
+      "step": 38575
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2030136753862304,
+      "learning_rate": 4.3284217701015985e-07,
+      "loss": 0.5547,
+      "step": 38576
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6314344191391432,
+      "learning_rate": 4.3262014167464337e-07,
+      "loss": 0.5224,
+      "step": 38577
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3658950640773724,
+      "learning_rate": 4.3239816204344256e-07,
+      "loss": 0.6411,
+      "step": 38578
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2776191079513546,
+      "learning_rate": 4.321762381178496e-07,
+      "loss": 0.6456,
+      "step": 38579
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6479747321661673,
+      "learning_rate": 4.319543698991546e-07,
+      "loss": 0.6888,
+      "step": 38580
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.628196291527902,
+      "learning_rate": 4.3173255738865104e-07,
+      "loss": 0.5112,
+      "step": 38581
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.429293553030061,
+      "learning_rate": 4.3151080058762897e-07,
+      "loss": 0.5795,
+      "step": 38582
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4952724709677154,
+      "learning_rate": 4.312890994973806e-07,
+      "loss": 0.6586,
+      "step": 38583
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6340087223352906,
+      "learning_rate": 4.310674541191984e-07,
+      "loss": 0.7845,
+      "step": 38584
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5207330738465967,
+      "learning_rate": 4.308458644543678e-07,
+      "loss": 0.6575,
+      "step": 38585
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5507660169980579,
+      "learning_rate": 4.306243305041846e-07,
+      "loss": 0.4702,
+      "step": 38586
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.31703330371226,
+      "learning_rate": 4.304028522699344e-07,
+      "loss": 0.6424,
+      "step": 38587
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.310304715186776,
+      "learning_rate": 4.3018142975291054e-07,
+      "loss": 0.7851,
+      "step": 38588
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.488914654664231,
+      "learning_rate": 4.2996006295439764e-07,
+      "loss": 0.7117,
+      "step": 38589
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.35715519793868,
+      "learning_rate": 4.2973875187568903e-07,
+      "loss": 0.7019,
+      "step": 38590
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5651841330793352,
+      "learning_rate": 4.295174965180704e-07,
+      "loss": 0.6638,
+      "step": 38591
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.407585693084685,
+      "learning_rate": 4.292962968828318e-07,
+      "loss": 0.755,
+      "step": 38592
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.458262085130987,
+      "learning_rate": 4.2907515297126003e-07,
+      "loss": 0.6042,
+      "step": 38593
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6951208342756787,
+      "learning_rate": 4.288540647846418e-07,
+      "loss": 0.6233,
+      "step": 38594
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2655461578804348,
+      "learning_rate": 4.286330323242649e-07,
+      "loss": 0.5758,
+      "step": 38595
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5629207124222317,
+      "learning_rate": 4.284120555914184e-07,
+      "loss": 0.6293,
+      "step": 38596
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5571040512325016,
+      "learning_rate": 4.281911345873868e-07,
+      "loss": 0.6548,
+      "step": 38597
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5941452614079554,
+      "learning_rate": 4.279702693134569e-07,
+      "loss": 0.6157,
+      "step": 38598
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2432448321530098,
+      "learning_rate": 4.277494597709131e-07,
+      "loss": 0.6731,
+      "step": 38599
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.447598363514053,
+      "learning_rate": 4.2752870596104446e-07,
+      "loss": 0.7545,
+      "step": 38600
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3286478602561538,
+      "learning_rate": 4.273080078851333e-07,
+      "loss": 0.6075,
+      "step": 38601
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5714206834379283,
+      "learning_rate": 4.270873655444652e-07,
+      "loss": 0.4679,
+      "step": 38602
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4894022863597132,
+      "learning_rate": 4.268667789403247e-07,
+      "loss": 0.6831,
+      "step": 38603
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6138598345820385,
+      "learning_rate": 4.2664624807399857e-07,
+      "loss": 0.5063,
+      "step": 38604
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.681623147820076,
+      "learning_rate": 4.264257729467691e-07,
+      "loss": 0.8242,
+      "step": 38605
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5310695624599133,
+      "learning_rate": 4.2620535355991756e-07,
+      "loss": 0.6459,
+      "step": 38606
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3723736012157202,
+      "learning_rate": 4.2598498991473057e-07,
+      "loss": 0.6269,
+      "step": 38607
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.208994683785187,
+      "learning_rate": 4.2576468201248946e-07,
+      "loss": 0.8386,
+      "step": 38608
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2667695546793707,
+      "learning_rate": 4.2554442985447864e-07,
+      "loss": 0.5073,
+      "step": 38609
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9987838908902544,
+      "learning_rate": 4.2532423344197825e-07,
+      "loss": 0.6753,
+      "step": 38610
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4853433729128755,
+      "learning_rate": 4.2510409277627174e-07,
+      "loss": 0.6951,
+      "step": 38611
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8242642469215666,
+      "learning_rate": 4.2488400785864027e-07,
+      "loss": 0.6274,
+      "step": 38612
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2890422665929868,
+      "learning_rate": 4.2466397869036724e-07,
+      "loss": 0.5364,
+      "step": 38613
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2905606908493652,
+      "learning_rate": 4.2444400527273056e-07,
+      "loss": 0.7708,
+      "step": 38614
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3906086508395612,
+      "learning_rate": 4.2422408760701253e-07,
+      "loss": 0.812,
+      "step": 38615
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5815154825645557,
+      "learning_rate": 4.240042256944932e-07,
+      "loss": 0.5032,
+      "step": 38616
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3252447021537368,
+      "learning_rate": 4.2378441953645375e-07,
+      "loss": 0.5998,
+      "step": 38617
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6247555163512233,
+      "learning_rate": 4.2356466913417326e-07,
+      "loss": 0.5196,
+      "step": 38618
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6287409681992828,
+      "learning_rate": 4.233449744889295e-07,
+      "loss": 0.6091,
+      "step": 38619
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3516398732599029,
+      "learning_rate": 4.2312533560200376e-07,
+      "loss": 0.6163,
+      "step": 38620
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4492329284859302,
+      "learning_rate": 4.22905752474676e-07,
+      "loss": 0.6765,
+      "step": 38621
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5790835234861864,
+      "learning_rate": 4.226862251082231e-07,
+      "loss": 0.5865,
+      "step": 38622
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.378964796717466,
+      "learning_rate": 4.224667535039206e-07,
+      "loss": 0.7322,
+      "step": 38623
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.588678240623347,
+      "learning_rate": 4.222473376630498e-07,
+      "loss": 0.5741,
+      "step": 38624
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3598777576537209,
+      "learning_rate": 4.2202797758688853e-07,
+      "loss": 0.668,
+      "step": 38625
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1976846930522023,
+      "learning_rate": 4.2180867327671237e-07,
+      "loss": 0.649,
+      "step": 38626
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.330386542294295,
+      "learning_rate": 4.2158942473379706e-07,
+      "loss": 0.7148,
+      "step": 38627
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3632654615625588,
+      "learning_rate": 4.2137023195942264e-07,
+      "loss": 0.7084,
+      "step": 38628
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.408038145840999,
+      "learning_rate": 4.211510949548614e-07,
+      "loss": 0.613,
+      "step": 38629
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4318473899049564,
+      "learning_rate": 4.209320137213924e-07,
+      "loss": 0.6096,
+      "step": 38630
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4732618415196803,
+      "learning_rate": 4.2071298826028894e-07,
+      "loss": 0.6345,
+      "step": 38631
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6522620726037573,
+      "learning_rate": 4.204940185728279e-07,
+      "loss": 0.4756,
+      "step": 38632
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2595272476538033,
+      "learning_rate": 4.2027510466028265e-07,
+      "loss": 0.771,
+      "step": 38633
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.432960455213044,
+      "learning_rate": 4.2005624652392996e-07,
+      "loss": 0.6867,
+      "step": 38634
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4463094870987245,
+      "learning_rate": 4.1983744416504323e-07,
+      "loss": 0.6066,
+      "step": 38635
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.604970009678373,
+      "learning_rate": 4.196186975848937e-07,
+      "loss": 0.7709,
+      "step": 38636
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3767319744312654,
+      "learning_rate": 4.19400006784757e-07,
+      "loss": 0.6863,
+      "step": 38637
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2721389565238408,
+      "learning_rate": 4.1918137176590876e-07,
+      "loss": 0.7267,
+      "step": 38638
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3355099096206682,
+      "learning_rate": 4.189627925296202e-07,
+      "loss": 0.6891,
+      "step": 38639
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4105332087875277,
+      "learning_rate": 4.187442690771615e-07,
+      "loss": 0.6916,
+      "step": 38640
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2929467081135235,
+      "learning_rate": 4.18525801409807e-07,
+      "loss": 0.7224,
+      "step": 38641
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.775914376854223,
+      "learning_rate": 4.183073895288303e-07,
+      "loss": 0.6623,
+      "step": 38642
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2917121812225438,
+      "learning_rate": 4.1808903343550145e-07,
+      "loss": 0.7496,
+      "step": 38643
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.899135927643493,
+      "learning_rate": 4.178707331310905e-07,
+      "loss": 0.6584,
+      "step": 38644
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4218627001026032,
+      "learning_rate": 4.176524886168698e-07,
+      "loss": 0.5892,
+      "step": 38645
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.540845461691125,
+      "learning_rate": 4.1743429989411054e-07,
+      "loss": 0.6379,
+      "step": 38646
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.264160452627097,
+      "learning_rate": 4.1721616696408394e-07,
+      "loss": 0.7269,
+      "step": 38647
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1575865764800513,
+      "learning_rate": 4.169980898280568e-07,
+      "loss": 0.6414,
+      "step": 38648
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1341649325372314,
+      "learning_rate": 4.167800684873013e-07,
+      "loss": 0.6144,
+      "step": 38649
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8489125352286262,
+      "learning_rate": 4.165621029430855e-07,
+      "loss": 0.6527,
+      "step": 38650
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6763158165041285,
+      "learning_rate": 4.1634419319668163e-07,
+      "loss": 0.6239,
+      "step": 38651
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2705619068242955,
+      "learning_rate": 4.1612633924935307e-07,
+      "loss": 0.6391,
+      "step": 38652
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2902957275057056,
+      "learning_rate": 4.1590854110237333e-07,
+      "loss": 0.704,
+      "step": 38653
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.331851913169643,
+      "learning_rate": 4.156907987570069e-07,
+      "loss": 0.722,
+      "step": 38654
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5645491826083238,
+      "learning_rate": 4.154731122145228e-07,
+      "loss": 0.4969,
+      "step": 38655
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5843235346397904,
+      "learning_rate": 4.152554814761911e-07,
+      "loss": 0.6761,
+      "step": 38656
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.379798534568947,
+      "learning_rate": 4.1503790654327414e-07,
+      "loss": 0.6008,
+      "step": 38657
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3943056050818707,
+      "learning_rate": 4.148203874170409e-07,
+      "loss": 0.626,
+      "step": 38658
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2218451667367305,
+      "learning_rate": 4.1460292409875815e-07,
+      "loss": 0.6335,
+      "step": 38659
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3957726524959029,
+      "learning_rate": 4.1438551658969484e-07,
+      "loss": 0.5622,
+      "step": 38660
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4817055512316843,
+      "learning_rate": 4.1416816489111e-07,
+      "loss": 0.6679,
+      "step": 38661
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2632118751922272,
+      "learning_rate": 4.139508690042737e-07,
+      "loss": 0.5835,
+      "step": 38662
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.310223623329571,
+      "learning_rate": 4.137336289304517e-07,
+      "loss": 0.6951,
+      "step": 38663
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.483675768330962,
+      "learning_rate": 4.1351644467090614e-07,
+      "loss": 0.6707,
+      "step": 38664
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4700110360676162,
+      "learning_rate": 4.132993162269039e-07,
+      "loss": 0.5234,
+      "step": 38665
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3039634618523568,
+      "learning_rate": 4.130822435997073e-07,
+      "loss": 0.6246,
+      "step": 38666
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3470843783008664,
+      "learning_rate": 4.128652267905819e-07,
+      "loss": 0.7346,
+      "step": 38667
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6048147802837205,
+      "learning_rate": 4.126482658007891e-07,
+      "loss": 0.7104,
+      "step": 38668
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4116173319105063,
+      "learning_rate": 4.124313606315955e-07,
+      "loss": 0.68,
+      "step": 38669
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5662112581917622,
+      "learning_rate": 4.1221451128426014e-07,
+      "loss": 0.4748,
+      "step": 38670
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3765021115835152,
+      "learning_rate": 4.119977177600476e-07,
+      "loss": 0.6304,
+      "step": 38671
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.326674649145111,
+      "learning_rate": 4.117809800602202e-07,
+      "loss": 0.5723,
+      "step": 38672
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.376213099033769,
+      "learning_rate": 4.11564298186039e-07,
+      "loss": 0.7196,
+      "step": 38673
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.408636570748285,
+      "learning_rate": 4.113476721387677e-07,
+      "loss": 0.6431,
+      "step": 38674
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2953089443282084,
+      "learning_rate": 4.11131101919664e-07,
+      "loss": 0.7191,
+      "step": 38675
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3678974775262431,
+      "learning_rate": 4.1091458752999136e-07,
+      "loss": 0.6993,
+      "step": 38676
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4971905041156837,
+      "learning_rate": 4.106981289710099e-07,
+      "loss": 0.603,
+      "step": 38677
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.093447781068601,
+      "learning_rate": 4.104817262439809e-07,
+      "loss": 0.6766,
+      "step": 38678
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8536491633441377,
+      "learning_rate": 4.102653793501621e-07,
+      "loss": 0.7039,
+      "step": 38679
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5603064580800403,
+      "learning_rate": 4.100490882908137e-07,
+      "loss": 0.633,
+      "step": 38680
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4470926143090281,
+      "learning_rate": 4.098328530671969e-07,
+      "loss": 0.7234,
+      "step": 38681
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7166568746622535,
+      "learning_rate": 4.0961667368056957e-07,
+      "loss": 0.6095,
+      "step": 38682
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3727874549790162,
+      "learning_rate": 4.094005501321896e-07,
+      "loss": 0.698,
+      "step": 38683
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5850216226770729,
+      "learning_rate": 4.09184482423316e-07,
+      "loss": 0.506,
+      "step": 38684
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.613479570636195,
+      "learning_rate": 4.089684705552066e-07,
+      "loss": 0.5026,
+      "step": 38685
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3471969307654912,
+      "learning_rate": 4.087525145291205e-07,
+      "loss": 0.6397,
+      "step": 38686
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.237366616199321,
+      "learning_rate": 4.085366143463121e-07,
+      "loss": 0.4891,
+      "step": 38687
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3101793244595024,
+      "learning_rate": 4.083207700080416e-07,
+      "loss": 0.6427,
+      "step": 38688
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4313166194053124,
+      "learning_rate": 4.0810498151556246e-07,
+      "loss": 0.6332,
+      "step": 38689
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8716065673341926,
+      "learning_rate": 4.078892488701347e-07,
+      "loss": 0.5903,
+      "step": 38690
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4862718655018914,
+      "learning_rate": 4.0767357207301073e-07,
+      "loss": 0.6636,
+      "step": 38691
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4384402660910551,
+      "learning_rate": 4.074579511254495e-07,
+      "loss": 0.7507,
+      "step": 38692
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3777588535465848,
+      "learning_rate": 4.072423860287045e-07,
+      "loss": 0.5726,
+      "step": 38693
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5347155812845243,
+      "learning_rate": 4.070268767840324e-07,
+      "loss": 0.6082,
+      "step": 38694
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7426672541241917,
+      "learning_rate": 4.0681142339268564e-07,
+      "loss": 0.605,
+      "step": 38695
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8438378881189084,
+      "learning_rate": 4.0659602585591984e-07,
+      "loss": 0.7341,
+      "step": 38696
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6279492989637008,
+      "learning_rate": 4.0638068417498955e-07,
+      "loss": 0.6337,
+      "step": 38697
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2911287222999288,
+      "learning_rate": 4.0616539835114823e-07,
+      "loss": 0.706,
+      "step": 38698
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.458883717954085,
+      "learning_rate": 4.059501683856504e-07,
+      "loss": 0.7495,
+      "step": 38699
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7310985300289778,
+      "learning_rate": 4.0573499427974616e-07,
+      "loss": 0.7224,
+      "step": 38700
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5549176649622132,
+      "learning_rate": 4.055198760346901e-07,
+      "loss": 0.6271,
+      "step": 38701
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5363690614611296,
+      "learning_rate": 4.053048136517368e-07,
+      "loss": 0.7232,
+      "step": 38702
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.570707805841316,
+      "learning_rate": 4.050898071321363e-07,
+      "loss": 0.5621,
+      "step": 38703
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5312141891258753,
+      "learning_rate": 4.048748564771399e-07,
+      "loss": 0.743,
+      "step": 38704
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3984799022138346,
+      "learning_rate": 4.046599616879987e-07,
+      "loss": 0.6609,
+      "step": 38705
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4173017041540024,
+      "learning_rate": 4.044451227659674e-07,
+      "loss": 0.6718,
+      "step": 38706
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3801565365237254,
+      "learning_rate": 4.0423033971229375e-07,
+      "loss": 0.7039,
+      "step": 38707
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6012593981818672,
+      "learning_rate": 4.0401561252822796e-07,
+      "loss": 0.6399,
+      "step": 38708
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.44599789859056,
+      "learning_rate": 4.0380094121502236e-07,
+      "loss": 0.5967,
+      "step": 38709
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4211401100953152,
+      "learning_rate": 4.035863257739248e-07,
+      "loss": 0.6615,
+      "step": 38710
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4605358299011244,
+      "learning_rate": 4.0337176620618756e-07,
+      "loss": 0.689,
+      "step": 38711
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5209187260249135,
+      "learning_rate": 4.031572625130564e-07,
+      "loss": 0.6596,
+      "step": 38712
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.81007018977479,
+      "learning_rate": 4.0294281469578364e-07,
+      "loss": 0.6563,
+      "step": 38713
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5957056527751012,
+      "learning_rate": 4.0272842275561385e-07,
+      "loss": 0.4955,
+      "step": 38714
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3928044281443368,
+      "learning_rate": 4.025140866937993e-07,
+      "loss": 0.8024,
+      "step": 38715
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6404278272868777,
+      "learning_rate": 4.0229980651158686e-07,
+      "loss": 0.7196,
+      "step": 38716
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4300556468991221,
+      "learning_rate": 4.020855822102221e-07,
+      "loss": 0.7003,
+      "step": 38717
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3494702615873744,
+      "learning_rate": 4.0187141379095294e-07,
+      "loss": 0.6441,
+      "step": 38718
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4228524258875404,
+      "learning_rate": 4.0165730125502956e-07,
+      "loss": 0.6195,
+      "step": 38719
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4656005078830256,
+      "learning_rate": 4.0144324460369535e-07,
+      "loss": 0.5868,
+      "step": 38720
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4925337816182973,
+      "learning_rate": 4.012292438381971e-07,
+      "loss": 0.7112,
+      "step": 38721
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3333590449020358,
+      "learning_rate": 4.010152989597804e-07,
+      "loss": 0.519,
+      "step": 38722
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5062577879457848,
+      "learning_rate": 4.008014099696922e-07,
+      "loss": 0.6908,
+      "step": 38723
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2872300091252062,
+      "learning_rate": 4.0058757686917805e-07,
+      "loss": 0.6319,
+      "step": 38724
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3006407378181295,
+      "learning_rate": 4.003737996594803e-07,
+      "loss": 0.6011,
+      "step": 38725
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.77911654936424,
+      "learning_rate": 4.001600783418469e-07,
+      "loss": 0.7483,
+      "step": 38726
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6541840529924017,
+      "learning_rate": 3.999464129175201e-07,
+      "loss": 0.6429,
+      "step": 38727
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4710376016292743,
+      "learning_rate": 3.997328033877457e-07,
+      "loss": 0.6033,
+      "step": 38728
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5861680438807453,
+      "learning_rate": 3.995192497537659e-07,
+      "loss": 0.4718,
+      "step": 38729
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3263602028467847,
+      "learning_rate": 3.993057520168242e-07,
+      "loss": 0.7582,
+      "step": 38730
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4336274372521098,
+      "learning_rate": 3.99092310178163e-07,
+      "loss": 0.6718,
+      "step": 38731
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3599240247631879,
+      "learning_rate": 3.988789242390279e-07,
+      "loss": 0.6527,
+      "step": 38732
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.51318409670618,
+      "learning_rate": 3.986655942006579e-07,
+      "loss": 0.6423,
+      "step": 38733
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5536777464058065,
+      "learning_rate": 3.9845232006429755e-07,
+      "loss": 0.682,
+      "step": 38734
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3482315503810114,
+      "learning_rate": 3.98239101831186e-07,
+      "loss": 0.7384,
+      "step": 38735
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3976738491900704,
+      "learning_rate": 3.9802593950256875e-07,
+      "loss": 0.8231,
+      "step": 38736
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.771166613786062,
+      "learning_rate": 3.9781283307968266e-07,
+      "loss": 0.6722,
+      "step": 38737
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4402669687315928,
+      "learning_rate": 3.9759978256377006e-07,
+      "loss": 0.6582,
+      "step": 38738
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3314487649430908,
+      "learning_rate": 3.973867879560722e-07,
+      "loss": 0.6267,
+      "step": 38739
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5822556238023169,
+      "learning_rate": 3.9717384925782697e-07,
+      "loss": 0.478,
+      "step": 38740
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3958377050704887,
+      "learning_rate": 3.9696096647028004e-07,
+      "loss": 0.6911,
+      "step": 38741
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.680704602364764,
+      "learning_rate": 3.9674813959466263e-07,
+      "loss": 0.7879,
+      "step": 38742
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3717988455198245,
+      "learning_rate": 3.965353686322193e-07,
+      "loss": 0.7384,
+      "step": 38743
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6415112822321576,
+      "learning_rate": 3.9632265358418796e-07,
+      "loss": 0.6081,
+      "step": 38744
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3298763080110254,
+      "learning_rate": 3.961099944518065e-07,
+      "loss": 0.5676,
+      "step": 38745
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5896991557466036,
+      "learning_rate": 3.9589739123631177e-07,
+      "loss": 0.4925,
+      "step": 38746
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6588729321631124,
+      "learning_rate": 3.956848439389427e-07,
+      "loss": 0.5937,
+      "step": 38747
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2355737008655392,
+      "learning_rate": 3.954723525609394e-07,
+      "loss": 0.6843,
+      "step": 38748
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3174935232216354,
+      "learning_rate": 3.952599171035365e-07,
+      "loss": 0.6377,
+      "step": 38749
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3365443923764153,
+      "learning_rate": 3.950475375679697e-07,
+      "loss": 0.6813,
+      "step": 38750
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2163934045955214,
+      "learning_rate": 3.9483521395547895e-07,
+      "loss": 0.5772,
+      "step": 38751
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2633361749367762,
+      "learning_rate": 3.946229462672957e-07,
+      "loss": 0.6631,
+      "step": 38752
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5661341749514999,
+      "learning_rate": 3.9441073450465997e-07,
+      "loss": 0.7515,
+      "step": 38753
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6493023561859681,
+      "learning_rate": 3.941985786688074e-07,
+      "loss": 0.5057,
+      "step": 38754
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.380472536706224,
+      "learning_rate": 3.939864787609715e-07,
+      "loss": 0.6368,
+      "step": 38755
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2736003645165872,
+      "learning_rate": 3.9377443478238575e-07,
+      "loss": 0.5775,
+      "step": 38756
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6351770036289255,
+      "learning_rate": 3.935624467342869e-07,
+      "loss": 0.5927,
+      "step": 38757
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3510659922607666,
+      "learning_rate": 3.933505146179117e-07,
+      "loss": 0.5864,
+      "step": 38758
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2793046421668812,
+      "learning_rate": 3.9313863843448817e-07,
+      "loss": 0.5898,
+      "step": 38759
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5619801368938562,
+      "learning_rate": 3.92926818185253e-07,
+      "loss": 0.7804,
+      "step": 38760
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.26482942949437,
+      "learning_rate": 3.927150538714397e-07,
+      "loss": 0.5867,
+      "step": 38761
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6306082387612213,
+      "learning_rate": 3.925033454942828e-07,
+      "loss": 0.5637,
+      "step": 38762
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.589651679715302,
+      "learning_rate": 3.922916930550125e-07,
+      "loss": 0.6596,
+      "step": 38763
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4530407180701967,
+      "learning_rate": 3.920800965548599e-07,
+      "loss": 0.6676,
+      "step": 38764
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3189627762288771,
+      "learning_rate": 3.9186855599506077e-07,
+      "loss": 0.6315,
+      "step": 38765
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4426457716784788,
+      "learning_rate": 3.916570713768442e-07,
+      "loss": 0.5655,
+      "step": 38766
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4974842156161134,
+      "learning_rate": 3.914456427014424e-07,
+      "loss": 0.709,
+      "step": 38767
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3809174583782873,
+      "learning_rate": 3.912342699700844e-07,
+      "loss": 0.5558,
+      "step": 38768
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5605794296133488,
+      "learning_rate": 3.9102295318400487e-07,
+      "loss": 0.7003,
+      "step": 38769
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.372198949859552,
+      "learning_rate": 3.908116923444294e-07,
+      "loss": 0.653,
+      "step": 38770
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.039123481483702,
+      "learning_rate": 3.906004874525926e-07,
+      "loss": 0.5556,
+      "step": 38771
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6532207512685027,
+      "learning_rate": 3.903893385097224e-07,
+      "loss": 0.7149,
+      "step": 38772
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5856521810150739,
+      "learning_rate": 3.9017824551704554e-07,
+      "loss": 0.6266,
+      "step": 38773
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4638879663110502,
+      "learning_rate": 3.8996720847579437e-07,
+      "loss": 0.7278,
+      "step": 38774
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4489420452341288,
+      "learning_rate": 3.8975622738719685e-07,
+      "loss": 0.7096,
+      "step": 38775
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.464091352928734,
+      "learning_rate": 3.89545302252482e-07,
+      "loss": 0.6894,
+      "step": 38776
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5732246090541373,
+      "learning_rate": 3.893344330728754e-07,
+      "loss": 0.6598,
+      "step": 38777
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3724726249918482,
+      "learning_rate": 3.891236198496062e-07,
+      "loss": 0.578,
+      "step": 38778
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2688195326163445,
+      "learning_rate": 3.889128625839034e-07,
+      "loss": 0.6368,
+      "step": 38779
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5574718467605084,
+      "learning_rate": 3.887021612769937e-07,
+      "loss": 0.5897,
+      "step": 38780
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6682845364515346,
+      "learning_rate": 3.884915159301006e-07,
+      "loss": 0.7463,
+      "step": 38781
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6045671909592226,
+      "learning_rate": 3.8828092654445314e-07,
+      "loss": 0.516,
+      "step": 38782
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4286665306620105,
+      "learning_rate": 3.8807039312127813e-07,
+      "loss": 0.5384,
+      "step": 38783
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.335968020638893,
+      "learning_rate": 3.8785991566180017e-07,
+      "loss": 0.8076,
+      "step": 38784
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.4332666958334936,
+      "learning_rate": 3.876494941672437e-07,
+      "loss": 0.6876,
+      "step": 38785
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3237266941282333,
+      "learning_rate": 3.874391286388368e-07,
+      "loss": 0.7141,
+      "step": 38786
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.808897717558476,
+      "learning_rate": 3.8722881907780173e-07,
+      "loss": 0.6288,
+      "step": 38787
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5868691569014602,
+      "learning_rate": 3.870185654853642e-07,
+      "loss": 0.4428,
+      "step": 38788
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2408173168366465,
+      "learning_rate": 3.8680836786274764e-07,
+      "loss": 0.7312,
+      "step": 38789
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.496945450861908,
+      "learning_rate": 3.865982262111767e-07,
+      "loss": 0.629,
+      "step": 38790
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3625845207685607,
+      "learning_rate": 3.863881405318737e-07,
+      "loss": 0.6762,
+      "step": 38791
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6545605561760361,
+      "learning_rate": 3.8617811082606317e-07,
+      "loss": 0.6019,
+      "step": 38792
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5788395927780563,
+      "learning_rate": 3.859681370949664e-07,
+      "loss": 0.6392,
+      "step": 38793
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.397116695250494,
+      "learning_rate": 3.8575821933980906e-07,
+      "loss": 0.6282,
+      "step": 38794
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.564479018594912,
+      "learning_rate": 3.855483575618091e-07,
+      "loss": 0.6812,
+      "step": 38795
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3723263593688935,
+      "learning_rate": 3.85338551762191e-07,
+      "loss": 0.6857,
+      "step": 38796
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6057979125935853,
+      "learning_rate": 3.851288019421773e-07,
+      "loss": 0.4625,
+      "step": 38797
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3389522291063198,
+      "learning_rate": 3.8491910810298574e-07,
+      "loss": 0.5612,
+      "step": 38798
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4843886415974668,
+      "learning_rate": 3.847094702458387e-07,
+      "loss": 0.7155,
+      "step": 38799
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5776655631446962,
+      "learning_rate": 3.8449988837195975e-07,
+      "loss": 0.484,
+      "step": 38800
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.643375384867884,
+      "learning_rate": 3.842903624825656e-07,
+      "loss": 0.6741,
+      "step": 38801
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3513083230785674,
+      "learning_rate": 3.840808925788764e-07,
+      "loss": 0.6025,
+      "step": 38802
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5205333937599763,
+      "learning_rate": 3.838714786621123e-07,
+      "loss": 0.6569,
+      "step": 38803
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5791528394098929,
+      "learning_rate": 3.8366212073349453e-07,
+      "loss": 0.4925,
+      "step": 38804
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4057840356677647,
+      "learning_rate": 3.834528187942399e-07,
+      "loss": 0.7261,
+      "step": 38805
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5552721630473183,
+      "learning_rate": 3.8324357284556744e-07,
+      "loss": 0.7343,
+      "step": 38806
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.581961951462181,
+      "learning_rate": 3.8303438288869397e-07,
+      "loss": 0.4814,
+      "step": 38807
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.838410037120349,
+      "learning_rate": 3.828252489248407e-07,
+      "loss": 0.6392,
+      "step": 38808
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5364668624670326,
+      "learning_rate": 3.826161709552245e-07,
+      "loss": 0.5991,
+      "step": 38809
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.635269450512188,
+      "learning_rate": 3.824071489810599e-07,
+      "loss": 0.6024,
+      "step": 38810
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6162754517756885,
+      "learning_rate": 3.8219818300356706e-07,
+      "loss": 0.6475,
+      "step": 38811
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.011944771555928,
+      "learning_rate": 3.819892730239594e-07,
+      "loss": 0.6229,
+      "step": 38812
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2595146558119987,
+      "learning_rate": 3.817804190434571e-07,
+      "loss": 0.5391,
+      "step": 38813
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4975819930933398,
+      "learning_rate": 3.815716210632736e-07,
+      "loss": 0.6359,
+      "step": 38814
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3260651117830382,
+      "learning_rate": 3.813628790846258e-07,
+      "loss": 0.615,
+      "step": 38815
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5830765195553955,
+      "learning_rate": 3.81154193108727e-07,
+      "loss": 0.7378,
+      "step": 38816
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2695136350202858,
+      "learning_rate": 3.809455631367953e-07,
+      "loss": 0.5806,
+      "step": 38817
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6820310518068463,
+      "learning_rate": 3.807369891700441e-07,
+      "loss": 0.5455,
+      "step": 38818
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1990505752000393,
+      "learning_rate": 3.805284712096857e-07,
+      "loss": 0.6206,
+      "step": 38819
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6296551797300463,
+      "learning_rate": 3.80320009256937e-07,
+      "loss": 0.4962,
+      "step": 38820
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7807567262697226,
+      "learning_rate": 3.8011160331301147e-07,
+      "loss": 0.6977,
+      "step": 38821
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.369523858459227,
+      "learning_rate": 3.7990325337912136e-07,
+      "loss": 0.7124,
+      "step": 38822
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2950508477187244,
+      "learning_rate": 3.7969495945647915e-07,
+      "loss": 0.5942,
+      "step": 38823
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9453532139972234,
+      "learning_rate": 3.794867215462994e-07,
+      "loss": 0.6777,
+      "step": 38824
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4606113103157,
+      "learning_rate": 3.7927853964979443e-07,
+      "loss": 0.7426,
+      "step": 38825
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3521186947534611,
+      "learning_rate": 3.7907041376817554e-07,
+      "loss": 0.7046,
+      "step": 38826
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4032697105418073,
+      "learning_rate": 3.7886234390265286e-07,
+      "loss": 0.6071,
+      "step": 38827
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.390876516421288,
+      "learning_rate": 3.78654330054441e-07,
+      "loss": 0.602,
+      "step": 38828
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3424800657766678,
+      "learning_rate": 3.7844637222475e-07,
+      "loss": 0.714,
+      "step": 38829
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6883794365578435,
+      "learning_rate": 3.782384704147901e-07,
+      "loss": 0.6537,
+      "step": 38830
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.264037138640612,
+      "learning_rate": 3.7803062462577147e-07,
+      "loss": 0.5903,
+      "step": 38831
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2767562374975794,
+      "learning_rate": 3.778228348589064e-07,
+      "loss": 0.5938,
+      "step": 38832
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4687053854142125,
+      "learning_rate": 3.7761510111540054e-07,
+      "loss": 0.7825,
+      "step": 38833
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3963199116928555,
+      "learning_rate": 3.774074233964686e-07,
+      "loss": 0.5998,
+      "step": 38834
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7220175445783448,
+      "learning_rate": 3.7719980170331516e-07,
+      "loss": 0.5445,
+      "step": 38835
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5867089281596592,
+      "learning_rate": 3.769922360371525e-07,
+      "loss": 0.5213,
+      "step": 38836
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3665473766286962,
+      "learning_rate": 3.767847263991853e-07,
+      "loss": 0.6583,
+      "step": 38837
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3110283642775848,
+      "learning_rate": 3.7657727279062585e-07,
+      "loss": 0.746,
+      "step": 38838
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7331854722503341,
+      "learning_rate": 3.7636987521268097e-07,
+      "loss": 0.5602,
+      "step": 38839
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6089297899471988,
+      "learning_rate": 3.761625336665564e-07,
+      "loss": 0.5173,
+      "step": 38840
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2768928131464765,
+      "learning_rate": 3.759552481534601e-07,
+      "loss": 0.6011,
+      "step": 38841
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.435323400562782,
+      "learning_rate": 3.7574801867459877e-07,
+      "loss": 0.7294,
+      "step": 38842
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3899783717513183,
+      "learning_rate": 3.755408452311826e-07,
+      "loss": 0.7237,
+      "step": 38843
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7277631205383726,
+      "learning_rate": 3.753337278244118e-07,
+      "loss": 0.7008,
+      "step": 38844
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7459490124144452,
+      "learning_rate": 3.7512666645549533e-07,
+      "loss": 0.6479,
+      "step": 38845
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2448978499687067,
+      "learning_rate": 3.7491966112563893e-07,
+      "loss": 0.7083,
+      "step": 38846
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5997610293476551,
+      "learning_rate": 3.747127118360483e-07,
+      "loss": 0.5063,
+      "step": 38847
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5453398287678433,
+      "learning_rate": 3.745058185879269e-07,
+      "loss": 0.7725,
+      "step": 38848
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3931448900220589,
+      "learning_rate": 3.7429898138247933e-07,
+      "loss": 0.6503,
+      "step": 38849
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7999139326824853,
+      "learning_rate": 3.740922002209113e-07,
+      "loss": 0.7263,
+      "step": 38850
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3308258926808134,
+      "learning_rate": 3.738854751044252e-07,
+      "loss": 0.7373,
+      "step": 38851
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.557634776264245,
+      "learning_rate": 3.736788060342267e-07,
+      "loss": 0.6883,
+      "step": 38852
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6290028979493145,
+      "learning_rate": 3.7347219301151815e-07,
+      "loss": 0.4837,
+      "step": 38853
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2823975360924336,
+      "learning_rate": 3.7326563603750085e-07,
+      "loss": 0.5981,
+      "step": 38854
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4027667669922075,
+      "learning_rate": 3.7305913511337834e-07,
+      "loss": 0.6467,
+      "step": 38855
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4539122763810186,
+      "learning_rate": 3.728526902403551e-07,
+      "loss": 0.6171,
+      "step": 38856
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.392955635450952,
+      "learning_rate": 3.726463014196313e-07,
+      "loss": 0.6153,
+      "step": 38857
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4968126214447999,
+      "learning_rate": 3.724399686524072e-07,
+      "loss": 0.8448,
+      "step": 38858
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3494792639834863,
+      "learning_rate": 3.722336919398861e-07,
+      "loss": 0.7542,
+      "step": 38859
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4659786395682783,
+      "learning_rate": 3.7202747128326943e-07,
+      "loss": 0.6851,
+      "step": 38860
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2509750700565014,
+      "learning_rate": 3.7182130668375726e-07,
+      "loss": 0.7578,
+      "step": 38861
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3484248895417825,
+      "learning_rate": 3.7161519814254864e-07,
+      "loss": 0.6855,
+      "step": 38862
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4387859256854014,
+      "learning_rate": 3.7140914566084376e-07,
+      "loss": 0.6195,
+      "step": 38863
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8831158217519064,
+      "learning_rate": 3.712031492398449e-07,
+      "loss": 0.7806,
+      "step": 38864
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6149395907496625,
+      "learning_rate": 3.7099720888075006e-07,
+      "loss": 0.5467,
+      "step": 38865
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2698144494086907,
+      "learning_rate": 3.7079132458475606e-07,
+      "loss": 0.5288,
+      "step": 38866
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4644233627096708,
+      "learning_rate": 3.705854963530653e-07,
+      "loss": 0.6872,
+      "step": 38867
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5434080592988288,
+      "learning_rate": 3.703797241868734e-07,
+      "loss": 0.7011,
+      "step": 38868
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3844778634669104,
+      "learning_rate": 3.7017400808737946e-07,
+      "loss": 0.6585,
+      "step": 38869
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4544399300130075,
+      "learning_rate": 3.699683480557814e-07,
+      "loss": 0.6713,
+      "step": 38870
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5889424606618013,
+      "learning_rate": 3.6976274409327717e-07,
+      "loss": 0.4763,
+      "step": 38871
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5382893345330015,
+      "learning_rate": 3.6955719620106135e-07,
+      "loss": 0.7583,
+      "step": 38872
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3917917233759065,
+      "learning_rate": 3.693517043803341e-07,
+      "loss": 0.7446,
+      "step": 38873
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8766343254011453,
+      "learning_rate": 3.6914626863229e-07,
+      "loss": 0.6675,
+      "step": 38874
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6104155613962976,
+      "learning_rate": 3.689408889581247e-07,
+      "loss": 0.501,
+      "step": 38875
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2795567433983364,
+      "learning_rate": 3.687355653590341e-07,
+      "loss": 0.6757,
+      "step": 38876
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.270027138610882,
+      "learning_rate": 3.6853029783621596e-07,
+      "loss": 0.6407,
+      "step": 38877
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3314869828557956,
+      "learning_rate": 3.6832508639086386e-07,
+      "loss": 0.6408,
+      "step": 38878
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.147387675365964,
+      "learning_rate": 3.681199310241701e-07,
+      "loss": 0.7472,
+      "step": 38879
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.540315583222856,
+      "learning_rate": 3.6791483173733267e-07,
+      "loss": 0.7935,
+      "step": 38880
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3774785596313008,
+      "learning_rate": 3.67709788531545e-07,
+      "loss": 0.7014,
+      "step": 38881
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2714609937021066,
+      "learning_rate": 3.675048014080007e-07,
+      "loss": 0.7889,
+      "step": 38882
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4226060633753532,
+      "learning_rate": 3.6729987036789205e-07,
+      "loss": 0.8096,
+      "step": 38883
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6535682597039787,
+      "learning_rate": 3.6709499541241367e-07,
+      "loss": 0.694,
+      "step": 38884
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.640476509755841,
+      "learning_rate": 3.6689017654275905e-07,
+      "loss": 0.4691,
+      "step": 38885
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5557960962453148,
+      "learning_rate": 3.6668541376011836e-07,
+      "loss": 0.6566,
+      "step": 38886
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5678672269426288,
+      "learning_rate": 3.664807070656851e-07,
+      "loss": 0.6159,
+      "step": 38887
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3794006713798046,
+      "learning_rate": 3.6627605646065157e-07,
+      "loss": 0.6616,
+      "step": 38888
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3281402688472308,
+      "learning_rate": 3.66071461946208e-07,
+      "loss": 0.6389,
+      "step": 38889
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.406516600026251,
+      "learning_rate": 3.6586692352354793e-07,
+      "loss": 0.6622,
+      "step": 38890
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4176322086230602,
+      "learning_rate": 3.656624411938592e-07,
+      "loss": 0.6266,
+      "step": 38891
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2089822084197743,
+      "learning_rate": 3.6545801495833534e-07,
+      "loss": 0.6264,
+      "step": 38892
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6188411496045945,
+      "learning_rate": 3.6525364481816315e-07,
+      "loss": 0.5232,
+      "step": 38893
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.407503261576262,
+      "learning_rate": 3.650493307745362e-07,
+      "loss": 0.7927,
+      "step": 38894
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.457045839781717,
+      "learning_rate": 3.6484507282864233e-07,
+      "loss": 0.6585,
+      "step": 38895
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5222490657419985,
+      "learning_rate": 3.646408709816707e-07,
+      "loss": 0.6631,
+      "step": 38896
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.621365942495822,
+      "learning_rate": 3.644367252348091e-07,
+      "loss": 0.7193,
+      "step": 38897
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.383461932155664,
+      "learning_rate": 3.6423263558924895e-07,
+      "loss": 0.5963,
+      "step": 38898
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3894270625574034,
+      "learning_rate": 3.64028602046177e-07,
+      "loss": 0.7127,
+      "step": 38899
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4229854655099037,
+      "learning_rate": 3.6382462460678113e-07,
+      "loss": 0.6014,
+      "step": 38900
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6677448560363586,
+      "learning_rate": 3.636207032722483e-07,
+      "loss": 0.7121,
+      "step": 38901
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8016650106290448,
+      "learning_rate": 3.634168380437675e-07,
+      "loss": 0.688,
+      "step": 38902
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4994566053436689,
+      "learning_rate": 3.632130289225244e-07,
+      "loss": 0.6707,
+      "step": 38903
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3130180512270442,
+      "learning_rate": 3.630092759097059e-07,
+      "loss": 0.6756,
+      "step": 38904
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.047324285844041,
+      "learning_rate": 3.628055790064977e-07,
+      "loss": 0.821,
+      "step": 38905
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.483323838959465,
+      "learning_rate": 3.6260193821408776e-07,
+      "loss": 0.6726,
+      "step": 38906
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5327233225299914,
+      "learning_rate": 3.623983535336606e-07,
+      "loss": 0.7463,
+      "step": 38907
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4249444253611836,
+      "learning_rate": 3.621948249663998e-07,
+      "loss": 0.5482,
+      "step": 38908
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6688955040456324,
+      "learning_rate": 3.6199135251349324e-07,
+      "loss": 0.5779,
+      "step": 38909
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 11.194065255075001,
+      "learning_rate": 3.617879361761245e-07,
+      "loss": 0.6818,
+      "step": 38910
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3978505102752492,
+      "learning_rate": 3.615845759554781e-07,
+      "loss": 0.7987,
+      "step": 38911
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1601111897820573,
+      "learning_rate": 3.613812718527365e-07,
+      "loss": 0.633,
+      "step": 38912
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2849642037864422,
+      "learning_rate": 3.611780238690854e-07,
+      "loss": 0.688,
+      "step": 38913
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5915628484437354,
+      "learning_rate": 3.6097483200570714e-07,
+      "loss": 0.7487,
+      "step": 38914
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4112711426761968,
+      "learning_rate": 3.6077169626378526e-07,
+      "loss": 0.6911,
+      "step": 38915
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3775375267004968,
+      "learning_rate": 3.6056861664450214e-07,
+      "loss": 0.6786,
+      "step": 38916
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.326420793466794,
+      "learning_rate": 3.603655931490413e-07,
+      "loss": 0.7899,
+      "step": 38917
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6514711897743024,
+      "learning_rate": 3.6016262577858174e-07,
+      "loss": 0.7628,
+      "step": 38918
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9397371225976499,
+      "learning_rate": 3.599597145343092e-07,
+      "loss": 0.7719,
+      "step": 38919
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2073654930626803,
+      "learning_rate": 3.597568594174028e-07,
+      "loss": 0.7352,
+      "step": 38920
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5667775127135222,
+      "learning_rate": 3.595540604290437e-07,
+      "loss": 0.4376,
+      "step": 38921
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4991166329788852,
+      "learning_rate": 3.593513175704122e-07,
+      "loss": 0.6988,
+      "step": 38922
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5943497033709289,
+      "learning_rate": 3.5914863084268946e-07,
+      "loss": 0.5088,
+      "step": 38923
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3337904518975985,
+      "learning_rate": 3.5894600024705904e-07,
+      "loss": 0.6845,
+      "step": 38924
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4458850554739144,
+      "learning_rate": 3.5874342578469337e-07,
+      "loss": 0.7034,
+      "step": 38925
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3028189533342105,
+      "learning_rate": 3.5854090745677696e-07,
+      "loss": 0.6298,
+      "step": 38926
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.358130914713382,
+      "learning_rate": 3.583384452644889e-07,
+      "loss": 0.7009,
+      "step": 38927
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5864160657861244,
+      "learning_rate": 3.581360392090072e-07,
+      "loss": 0.5097,
+      "step": 38928
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2162927988951393,
+      "learning_rate": 3.579336892915075e-07,
+      "loss": 0.5556,
+      "step": 38929
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.364248974184138,
+      "learning_rate": 3.5773139551317226e-07,
+      "loss": 0.7381,
+      "step": 38930
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3269453475545323,
+      "learning_rate": 3.5752915787517716e-07,
+      "loss": 0.6873,
+      "step": 38931
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8110405954750204,
+      "learning_rate": 3.573269763787013e-07,
+      "loss": 0.7309,
+      "step": 38932
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8228161406916235,
+      "learning_rate": 3.571248510249181e-07,
+      "loss": 0.6563,
+      "step": 38933
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5125885278178635,
+      "learning_rate": 3.5692278181500893e-07,
+      "loss": 0.7234,
+      "step": 38934
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6423493157929792,
+      "learning_rate": 3.5672076875014727e-07,
+      "loss": 0.6636,
+      "step": 38935
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3496011285460483,
+      "learning_rate": 3.5651881183150995e-07,
+      "loss": 0.8757,
+      "step": 38936
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.467889014544966,
+      "learning_rate": 3.563169110602749e-07,
+      "loss": 0.7464,
+      "step": 38937
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6058488671053364,
+      "learning_rate": 3.5611506643761453e-07,
+      "loss": 0.6423,
+      "step": 38938
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.625157381569619,
+      "learning_rate": 3.5591327796470567e-07,
+      "loss": 0.4958,
+      "step": 38939
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6025365266490739,
+      "learning_rate": 3.5571154564272293e-07,
+      "loss": 0.6933,
+      "step": 38940
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2653707556573712,
+      "learning_rate": 3.5550986947284205e-07,
+      "loss": 0.715,
+      "step": 38941
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5989208017195808,
+      "learning_rate": 3.553082494562354e-07,
+      "loss": 0.7402,
+      "step": 38942
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2285896989502976,
+      "learning_rate": 3.5510668559407656e-07,
+      "loss": 0.6208,
+      "step": 38943
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.093276711037179,
+      "learning_rate": 3.549051778875401e-07,
+      "loss": 0.6819,
+      "step": 38944
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.591160783789975,
+      "learning_rate": 3.547037263378017e-07,
+      "loss": 0.6597,
+      "step": 38945
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6341590774534174,
+      "learning_rate": 3.545023309460294e-07,
+      "loss": 0.6974,
+      "step": 38946
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5724883170954751,
+      "learning_rate": 3.543009917133988e-07,
+      "loss": 0.4957,
+      "step": 38947
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8746405756893296,
+      "learning_rate": 3.5409970864108247e-07,
+      "loss": 0.6274,
+      "step": 38948
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.440807010451761,
+      "learning_rate": 3.538984817302504e-07,
+      "loss": 0.6175,
+      "step": 38949
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1757955577257435,
+      "learning_rate": 3.536973109820763e-07,
+      "loss": 0.5881,
+      "step": 38950
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4919643869964125,
+      "learning_rate": 3.5349619639772903e-07,
+      "loss": 0.5843,
+      "step": 38951
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3908066497199483,
+      "learning_rate": 3.532951379783822e-07,
+      "loss": 0.6849,
+      "step": 38952
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.666715564089005,
+      "learning_rate": 3.530941357252038e-07,
+      "loss": 0.6752,
+      "step": 38953
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6010662986398585,
+      "learning_rate": 3.5289318963936726e-07,
+      "loss": 0.4537,
+      "step": 38954
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.192081620021138,
+      "learning_rate": 3.526922997220406e-07,
+      "loss": 0.7022,
+      "step": 38955
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.606123073703965,
+      "learning_rate": 3.524914659743928e-07,
+      "loss": 0.7042,
+      "step": 38956
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1911201611437785,
+      "learning_rate": 3.5229068839759297e-07,
+      "loss": 0.6366,
+      "step": 38957
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.534621712572844,
+      "learning_rate": 3.520899669928135e-07,
+      "loss": 0.6545,
+      "step": 38958
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.281914637024805,
+      "learning_rate": 3.5188930176122016e-07,
+      "loss": 0.6608,
+      "step": 38959
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4336447361976343,
+      "learning_rate": 3.5168869270398085e-07,
+      "loss": 0.6765,
+      "step": 38960
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.44383142428406,
+      "learning_rate": 3.514881398222647e-07,
+      "loss": 0.5568,
+      "step": 38961
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5225216801311054,
+      "learning_rate": 3.512876431172396e-07,
+      "loss": 0.7632,
+      "step": 38962
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.307505825467832,
+      "learning_rate": 3.5108720259007356e-07,
+      "loss": 0.7444,
+      "step": 38963
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.460209814598771,
+      "learning_rate": 3.508868182419312e-07,
+      "loss": 0.6326,
+      "step": 38964
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6052050618134496,
+      "learning_rate": 3.506864900739804e-07,
+      "loss": 0.4617,
+      "step": 38965
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4736936823480844,
+      "learning_rate": 3.5048621808738913e-07,
+      "loss": 0.6637,
+      "step": 38966
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7318393822319706,
+      "learning_rate": 3.502860022833221e-07,
+      "loss": 0.6847,
+      "step": 38967
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4946749698439445,
+      "learning_rate": 3.500858426629439e-07,
+      "loss": 0.7225,
+      "step": 38968
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5709965569646535,
+      "learning_rate": 3.498857392274213e-07,
+      "loss": 0.512,
+      "step": 38969
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4321116335001494,
+      "learning_rate": 3.49685691977919e-07,
+      "loss": 0.8285,
+      "step": 38970
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8378703252070214,
+      "learning_rate": 3.4948570091560276e-07,
+      "loss": 0.6901,
+      "step": 38971
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6016952002279945,
+      "learning_rate": 3.4928576604163487e-07,
+      "loss": 0.5939,
+      "step": 38972
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3706845361607454,
+      "learning_rate": 3.4908588735718116e-07,
+      "loss": 0.6346,
+      "step": 38973
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6613850161204717,
+      "learning_rate": 3.4888606486340515e-07,
+      "loss": 0.6557,
+      "step": 38974
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3606357641185935,
+      "learning_rate": 3.486862985614703e-07,
+      "loss": 0.6967,
+      "step": 38975
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5250350448547336,
+      "learning_rate": 3.4848658845253903e-07,
+      "loss": 0.6296,
+      "step": 38976
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6164178403735175,
+      "learning_rate": 3.4828693453777375e-07,
+      "loss": 0.4569,
+      "step": 38977
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6803952952600838,
+      "learning_rate": 3.4808733681833686e-07,
+      "loss": 0.6517,
+      "step": 38978
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.505714008190123,
+      "learning_rate": 3.4788779529539297e-07,
+      "loss": 0.5801,
+      "step": 38979
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6541964736204258,
+      "learning_rate": 3.476883099701023e-07,
+      "loss": 0.5249,
+      "step": 38980
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2993148753101809,
+      "learning_rate": 3.4748888084362497e-07,
+      "loss": 0.6624,
+      "step": 38981
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3169554625836122,
+      "learning_rate": 3.4728950791712343e-07,
+      "loss": 0.6552,
+      "step": 38982
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5578953993777565,
+      "learning_rate": 3.47090191191759e-07,
+      "loss": 0.6484,
+      "step": 38983
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4498416939111722,
+      "learning_rate": 3.4689093066869184e-07,
+      "loss": 0.7396,
+      "step": 38984
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3101089753900268,
+      "learning_rate": 3.4669172634908097e-07,
+      "loss": 0.7044,
+      "step": 38985
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.172643104161451,
+      "learning_rate": 3.4649257823408667e-07,
+      "loss": 0.6372,
+      "step": 38986
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4378102424176216,
+      "learning_rate": 3.462934863248701e-07,
+      "loss": 0.6726,
+      "step": 38987
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4865700056391224,
+      "learning_rate": 3.460944506225894e-07,
+      "loss": 0.6611,
+      "step": 38988
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4004667192141151,
+      "learning_rate": 3.458954711284013e-07,
+      "loss": 0.6314,
+      "step": 38989
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.459459211681443,
+      "learning_rate": 3.4569654784346816e-07,
+      "loss": 0.7064,
+      "step": 38990
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5639159138710522,
+      "learning_rate": 3.4549768076894473e-07,
+      "loss": 0.4897,
+      "step": 38991
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9617020014778928,
+      "learning_rate": 3.4529886990599113e-07,
+      "loss": 0.6766,
+      "step": 38992
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3962302918492029,
+      "learning_rate": 3.4510011525576317e-07,
+      "loss": 0.6418,
+      "step": 38993
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5809072167747572,
+      "learning_rate": 3.449014168194209e-07,
+      "loss": 0.4753,
+      "step": 38994
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3452839812410362,
+      "learning_rate": 3.44702774598118e-07,
+      "loss": 0.6385,
+      "step": 38995
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3976438036859755,
+      "learning_rate": 3.4450418859301226e-07,
+      "loss": 0.747,
+      "step": 38996
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.817463134725803,
+      "learning_rate": 3.443056588052618e-07,
+      "loss": 0.8256,
+      "step": 38997
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.938633310021009,
+      "learning_rate": 3.441071852360178e-07,
+      "loss": 0.6435,
+      "step": 38998
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.405275239493472,
+      "learning_rate": 3.4390876788644056e-07,
+      "loss": 0.7196,
+      "step": 38999
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8642802900285118,
+      "learning_rate": 3.437104067576835e-07,
+      "loss": 0.742,
+      "step": 39000
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5627386076913714,
+      "learning_rate": 3.4351210185090133e-07,
+      "loss": 0.6499,
+      "step": 39001
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3349910632420787,
+      "learning_rate": 3.4331385316724864e-07,
+      "loss": 0.6623,
+      "step": 39002
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.94373385659498,
+      "learning_rate": 3.4311566070788004e-07,
+      "loss": 0.6269,
+      "step": 39003
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.347091782440478,
+      "learning_rate": 3.429175244739502e-07,
+      "loss": 0.8178,
+      "step": 39004
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.578918392380473,
+      "learning_rate": 3.4271944446661153e-07,
+      "loss": 0.508,
+      "step": 39005
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2646688993903885,
+      "learning_rate": 3.4252142068701645e-07,
+      "loss": 0.7245,
+      "step": 39006
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3904877995382934,
+      "learning_rate": 3.423234531363195e-07,
+      "loss": 0.6106,
+      "step": 39007
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2882052244482387,
+      "learning_rate": 3.421255418156744e-07,
+      "loss": 0.6825,
+      "step": 39008
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5183873141832696,
+      "learning_rate": 3.4192768672623225e-07,
+      "loss": 0.6662,
+      "step": 39009
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5770380069411036,
+      "learning_rate": 3.417298878691433e-07,
+      "loss": 0.7958,
+      "step": 39010
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6079020488391806,
+      "learning_rate": 3.415321452455611e-07,
+      "loss": 0.4763,
+      "step": 39011
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5776321109474482,
+      "learning_rate": 3.4133445885663584e-07,
+      "loss": 0.5308,
+      "step": 39012
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3800583069320551,
+      "learning_rate": 3.4113682870352103e-07,
+      "loss": 0.6493,
+      "step": 39013
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5683117578949044,
+      "learning_rate": 3.409392547873636e-07,
+      "loss": 0.8255,
+      "step": 39014
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.274760019917392,
+      "learning_rate": 3.4074173710931804e-07,
+      "loss": 0.6454,
+      "step": 39015
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.793975610644603,
+      "learning_rate": 3.405442756705302e-07,
+      "loss": 0.5529,
+      "step": 39016
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.4450537399158696,
+      "learning_rate": 3.4034687047215245e-07,
+      "loss": 0.6437,
+      "step": 39017
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3865374605393048,
+      "learning_rate": 3.40149521515335e-07,
+      "loss": 0.7549,
+      "step": 39018
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2788186751793413,
+      "learning_rate": 3.3995222880122245e-07,
+      "loss": 0.6251,
+      "step": 39019
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.168476622474085,
+      "learning_rate": 3.3975499233096734e-07,
+      "loss": 0.6986,
+      "step": 39020
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3297678587146313,
+      "learning_rate": 3.395578121057164e-07,
+      "loss": 0.675,
+      "step": 39021
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5268506080378061,
+      "learning_rate": 3.3936068812662094e-07,
+      "loss": 0.5977,
+      "step": 39022
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2720273667259683,
+      "learning_rate": 3.391636203948234e-07,
+      "loss": 0.5934,
+      "step": 39023
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5701960520872387,
+      "learning_rate": 3.3896660891147516e-07,
+      "loss": 0.6488,
+      "step": 39024
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6231962610846854,
+      "learning_rate": 3.3876965367772183e-07,
+      "loss": 0.4837,
+      "step": 39025
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.52567343375557,
+      "learning_rate": 3.385727546947104e-07,
+      "loss": 0.7532,
+      "step": 39026
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3686583413931037,
+      "learning_rate": 3.3837591196358657e-07,
+      "loss": 0.5148,
+      "step": 39027
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.329514027873105,
+      "learning_rate": 3.38179125485496e-07,
+      "loss": 0.7504,
+      "step": 39028
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5800088645609558,
+      "learning_rate": 3.3798239526158793e-07,
+      "loss": 0.6833,
+      "step": 39029
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7260632914726035,
+      "learning_rate": 3.377857212930047e-07,
+      "loss": 0.6711,
+      "step": 39030
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.218858764149113,
+      "learning_rate": 3.375891035808898e-07,
+      "loss": 0.7523,
+      "step": 39031
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3390785239564909,
+      "learning_rate": 3.373925421263924e-07,
+      "loss": 0.6151,
+      "step": 39032
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.437312067045541,
+      "learning_rate": 3.3719603693065373e-07,
+      "loss": 0.4896,
+      "step": 39033
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.925707704171482,
+      "learning_rate": 3.369995879948185e-07,
+      "loss": 0.5658,
+      "step": 39034
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4325005297564668,
+      "learning_rate": 3.3680319532003234e-07,
+      "loss": 0.5935,
+      "step": 39035
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4284370367431325,
+      "learning_rate": 3.366068589074378e-07,
+      "loss": 0.7255,
+      "step": 39036
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2138210956789972,
+      "learning_rate": 3.3641057875817507e-07,
+      "loss": 0.5788,
+      "step": 39037
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.307139648868768,
+      "learning_rate": 3.3621435487339095e-07,
+      "loss": 0.588,
+      "step": 39038
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.380355578365036,
+      "learning_rate": 3.360181872542256e-07,
+      "loss": 0.6581,
+      "step": 39039
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.877051508214416,
+      "learning_rate": 3.358220759018238e-07,
+      "loss": 0.7047,
+      "step": 39040
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4100987586002611,
+      "learning_rate": 3.3562602081732344e-07,
+      "loss": 0.6869,
+      "step": 39041
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4116459047127239,
+      "learning_rate": 3.3543002200186803e-07,
+      "loss": 0.6601,
+      "step": 39042
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5166520502287948,
+      "learning_rate": 3.3523407945660004e-07,
+      "loss": 0.7046,
+      "step": 39043
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4864859647959006,
+      "learning_rate": 3.350381931826585e-07,
+      "loss": 0.7816,
+      "step": 39044
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2606985066266805,
+      "learning_rate": 3.348423631811848e-07,
+      "loss": 0.6978,
+      "step": 39045
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4196661507364765,
+      "learning_rate": 3.3464658945331796e-07,
+      "loss": 0.7688,
+      "step": 39046
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7691767378449685,
+      "learning_rate": 3.3445087200019934e-07,
+      "loss": 0.7569,
+      "step": 39047
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.875765403046802,
+      "learning_rate": 3.3425521082296797e-07,
+      "loss": 0.6083,
+      "step": 39048
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4935029034743235,
+      "learning_rate": 3.340596059227619e-07,
+      "loss": 0.7102,
+      "step": 39049
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5946012788142252,
+      "learning_rate": 3.3386405730072237e-07,
+      "loss": 0.5247,
+      "step": 39050
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4475827959294474,
+      "learning_rate": 3.336685649579852e-07,
+      "loss": 0.6856,
+      "step": 39051
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.245060909240554,
+      "learning_rate": 3.334731288956916e-07,
+      "loss": 0.6944,
+      "step": 39052
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3159873840366336,
+      "learning_rate": 3.332777491149763e-07,
+      "loss": 0.7789,
+      "step": 39053
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7705835061507178,
+      "learning_rate": 3.3308242561697956e-07,
+      "loss": 0.6285,
+      "step": 39054
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5958527860679496,
+      "learning_rate": 3.328871584028359e-07,
+      "loss": 0.4818,
+      "step": 39055
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6061267612432722,
+      "learning_rate": 3.326919474736856e-07,
+      "loss": 0.5,
+      "step": 39056
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.597166557161602,
+      "learning_rate": 3.324967928306622e-07,
+      "loss": 0.6346,
+      "step": 39057
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9814732561075785,
+      "learning_rate": 3.3230169447490356e-07,
+      "loss": 0.5883,
+      "step": 39058
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4515663093020248,
+      "learning_rate": 3.3210665240754446e-07,
+      "loss": 0.6713,
+      "step": 39059
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5847983609638271,
+      "learning_rate": 3.3191166662972174e-07,
+      "loss": 0.5013,
+      "step": 39060
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3548182430433646,
+      "learning_rate": 3.3171673714257113e-07,
+      "loss": 0.6791,
+      "step": 39061
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5929277777504799,
+      "learning_rate": 3.3152186394722506e-07,
+      "loss": 0.4886,
+      "step": 39062
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3631716772942846,
+      "learning_rate": 3.3132704704481934e-07,
+      "loss": 0.6396,
+      "step": 39063
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2715178489593466,
+      "learning_rate": 3.3113228643649075e-07,
+      "loss": 0.7217,
+      "step": 39064
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3935356674528692,
+      "learning_rate": 3.309375821233696e-07,
+      "loss": 0.5354,
+      "step": 39065
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6169378967976213,
+      "learning_rate": 3.3074293410659154e-07,
+      "loss": 0.5914,
+      "step": 39066
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5775312482499856,
+      "learning_rate": 3.3054834238728795e-07,
+      "loss": 0.4624,
+      "step": 39067
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3602875717182497,
+      "learning_rate": 3.303538069665957e-07,
+      "loss": 0.6257,
+      "step": 39068
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5711020826079064,
+      "learning_rate": 3.3015932784564385e-07,
+      "loss": 0.4937,
+      "step": 39069
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.530416063265909,
+      "learning_rate": 3.2996490502556486e-07,
+      "loss": 0.6889,
+      "step": 39070
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5296878363531516,
+      "learning_rate": 3.297705385074934e-07,
+      "loss": 0.826,
+      "step": 39071
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6126570162646388,
+      "learning_rate": 3.2957622829255744e-07,
+      "loss": 0.4892,
+      "step": 39072
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5690485422359668,
+      "learning_rate": 3.2938197438189154e-07,
+      "loss": 0.4919,
+      "step": 39073
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3839594033524283,
+      "learning_rate": 3.2918777677662495e-07,
+      "loss": 0.6909,
+      "step": 39074
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.052339500937712,
+      "learning_rate": 3.2899363547789e-07,
+      "loss": 0.5291,
+      "step": 39075
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7793181277564571,
+      "learning_rate": 3.2879955048681465e-07,
+      "loss": 0.8087,
+      "step": 39076
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5829997623604078,
+      "learning_rate": 3.286055218045303e-07,
+      "loss": 0.4957,
+      "step": 39077
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6036068310575813,
+      "learning_rate": 3.2841154943216714e-07,
+      "loss": 0.5039,
+      "step": 39078
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3516737094445568,
+      "learning_rate": 3.282176333708531e-07,
+      "loss": 0.7124,
+      "step": 39079
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3584742138826593,
+      "learning_rate": 3.2802377362171845e-07,
+      "loss": 0.6244,
+      "step": 39080
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5106955613360797,
+      "learning_rate": 3.278299701858922e-07,
+      "loss": 0.7473,
+      "step": 39081
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6220972572511313,
+      "learning_rate": 3.276362230645014e-07,
+      "loss": 0.5169,
+      "step": 39082
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5323197493820135,
+      "learning_rate": 3.2744253225867386e-07,
+      "loss": 0.721,
+      "step": 39083
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.317835877446382,
+      "learning_rate": 3.2724889776953874e-07,
+      "loss": 0.7165,
+      "step": 39084
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3230355711569646,
+      "learning_rate": 3.270553195982229e-07,
+      "loss": 0.6552,
+      "step": 39085
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5534743071544959,
+      "learning_rate": 3.2686179774585434e-07,
+      "loss": 0.66,
+      "step": 39086
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.45374556478676,
+      "learning_rate": 3.2666833221355774e-07,
+      "loss": 0.6454,
+      "step": 39087
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6262675501261248,
+      "learning_rate": 3.2647492300246e-07,
+      "loss": 0.6889,
+      "step": 39088
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.438151009450522,
+      "learning_rate": 3.262815701136901e-07,
+      "loss": 0.6906,
+      "step": 39089
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5380794814235432,
+      "learning_rate": 3.2608827354837057e-07,
+      "loss": 0.6811,
+      "step": 39090
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3427532999305982,
+      "learning_rate": 3.258950333076272e-07,
+      "loss": 0.6377,
+      "step": 39091
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.447649390838926,
+      "learning_rate": 3.2570184939258674e-07,
+      "loss": 0.6188,
+      "step": 39092
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4003246673430458,
+      "learning_rate": 3.255087218043718e-07,
+      "loss": 0.7078,
+      "step": 39093
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0912000329704843,
+      "learning_rate": 3.2531565054410797e-07,
+      "loss": 0.7446,
+      "step": 39094
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3157662256400267,
+      "learning_rate": 3.2512263561292e-07,
+      "loss": 0.6912,
+      "step": 39095
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4871886128850578,
+      "learning_rate": 3.2492967701193144e-07,
+      "loss": 0.681,
+      "step": 39096
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.736022379077796,
+      "learning_rate": 3.2473677474226363e-07,
+      "loss": 0.6404,
+      "step": 39097
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.366369374543659,
+      "learning_rate": 3.245439288050434e-07,
+      "loss": 0.6162,
+      "step": 39098
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.562758833096214,
+      "learning_rate": 3.2435113920139094e-07,
+      "loss": 0.6474,
+      "step": 39099
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4260513397878658,
+      "learning_rate": 3.2415840593242874e-07,
+      "loss": 0.7403,
+      "step": 39100
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3232062811858236,
+      "learning_rate": 3.2396572899928037e-07,
+      "loss": 0.6181,
+      "step": 39101
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5813059222489387,
+      "learning_rate": 3.237731084030671e-07,
+      "loss": 0.7144,
+      "step": 39102
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4753178431194052,
+      "learning_rate": 3.2358054414491024e-07,
+      "loss": 0.7408,
+      "step": 39103
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.57461409093062,
+      "learning_rate": 3.2338803622593006e-07,
+      "loss": 0.6602,
+      "step": 39104
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6575989727040168,
+      "learning_rate": 3.231955846472479e-07,
+      "loss": 0.6629,
+      "step": 39105
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2564100730344965,
+      "learning_rate": 3.2300318940998607e-07,
+      "loss": 0.7446,
+      "step": 39106
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5048494757238515,
+      "learning_rate": 3.2281085051526383e-07,
+      "loss": 0.8786,
+      "step": 39107
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6240212073690004,
+      "learning_rate": 3.226185679641991e-07,
+      "loss": 0.714,
+      "step": 39108
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3291906756279157,
+      "learning_rate": 3.224263417579132e-07,
+      "loss": 0.6036,
+      "step": 39109
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5603801356568416,
+      "learning_rate": 3.2223417189752525e-07,
+      "loss": 0.5495,
+      "step": 39110
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4875335982532418,
+      "learning_rate": 3.2204205838415437e-07,
+      "loss": 0.6662,
+      "step": 39111
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.35297093912486,
+      "learning_rate": 3.218500012189174e-07,
+      "loss": 0.5979,
+      "step": 39112
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6002676204080861,
+      "learning_rate": 3.216580004029357e-07,
+      "loss": 0.7255,
+      "step": 39113
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2466166265268248,
+      "learning_rate": 3.214660559373228e-07,
+      "loss": 0.6708,
+      "step": 39114
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3402635211146359,
+      "learning_rate": 3.212741678232012e-07,
+      "loss": 0.7191,
+      "step": 39115
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4148889520664303,
+      "learning_rate": 3.210823360616833e-07,
+      "loss": 0.5724,
+      "step": 39116
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.469067943308627,
+      "learning_rate": 3.2089056065388924e-07,
+      "loss": 0.6408,
+      "step": 39117
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3855534751135552,
+      "learning_rate": 3.206988416009349e-07,
+      "loss": 0.7897,
+      "step": 39118
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1780493435357047,
+      "learning_rate": 3.20507178903936e-07,
+      "loss": 0.551,
+      "step": 39119
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6694839138512347,
+      "learning_rate": 3.203155725640106e-07,
+      "loss": 0.6433,
+      "step": 39120
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.949530992485871,
+      "learning_rate": 3.2012402258226995e-07,
+      "loss": 0.6862,
+      "step": 39121
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.424376765033718,
+      "learning_rate": 3.1993252895983205e-07,
+      "loss": 0.6847,
+      "step": 39122
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.095039972072405,
+      "learning_rate": 3.1974109169781165e-07,
+      "loss": 0.6552,
+      "step": 39123
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.191094678224607,
+      "learning_rate": 3.1954971079732554e-07,
+      "loss": 0.7639,
+      "step": 39124
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4638577342384242,
+      "learning_rate": 3.1935838625948403e-07,
+      "loss": 0.6053,
+      "step": 39125
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6401238316836408,
+      "learning_rate": 3.191671180854028e-07,
+      "loss": 0.6156,
+      "step": 39126
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3688712180997817,
+      "learning_rate": 3.1897590627619546e-07,
+      "loss": 0.7619,
+      "step": 39127
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.627343358716778,
+      "learning_rate": 3.1878475083297667e-07,
+      "loss": 0.6414,
+      "step": 39128
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.548057641166959,
+      "learning_rate": 3.1859365175685666e-07,
+      "loss": 0.6732,
+      "step": 39129
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7516959465850885,
+      "learning_rate": 3.1840260904894896e-07,
+      "loss": 0.6847,
+      "step": 39130
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5578947505969802,
+      "learning_rate": 3.182116227103682e-07,
+      "loss": 0.708,
+      "step": 39131
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.342435259574893,
+      "learning_rate": 3.180206927422236e-07,
+      "loss": 0.6891,
+      "step": 39132
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3727115181753968,
+      "learning_rate": 3.178298191456275e-07,
+      "loss": 0.5631,
+      "step": 39133
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5625237935739362,
+      "learning_rate": 3.176390019216924e-07,
+      "loss": 0.6421,
+      "step": 39134
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3654359668263625,
+      "learning_rate": 3.174482410715274e-07,
+      "loss": 0.5534,
+      "step": 39135
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6503349480146428,
+      "learning_rate": 3.172575365962449e-07,
+      "loss": 0.6846,
+      "step": 39136
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3028062104451912,
+      "learning_rate": 3.170668884969552e-07,
+      "loss": 0.6037,
+      "step": 39137
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4482489720541405,
+      "learning_rate": 3.168762967747674e-07,
+      "loss": 0.6794,
+      "step": 39138
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4522046582423518,
+      "learning_rate": 3.1668576143078946e-07,
+      "loss": 0.6996,
+      "step": 39139
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4960905791715462,
+      "learning_rate": 3.164952824661338e-07,
+      "loss": 0.6328,
+      "step": 39140
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3848705240246788,
+      "learning_rate": 3.163048598819096e-07,
+      "loss": 0.6433,
+      "step": 39141
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3109780767116403,
+      "learning_rate": 3.1611449367922374e-07,
+      "loss": 0.6619,
+      "step": 39142
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6111879950347178,
+      "learning_rate": 3.1592418385918424e-07,
+      "loss": 0.505,
+      "step": 39143
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5534781479859916,
+      "learning_rate": 3.1573393042290015e-07,
+      "loss": 0.4489,
+      "step": 39144
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3379508392047292,
+      "learning_rate": 3.155437333714795e-07,
+      "loss": 0.6853,
+      "step": 39145
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2442038505935848,
+      "learning_rate": 3.1535359270603027e-07,
+      "loss": 0.702,
+      "step": 39146
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2856440650071264,
+      "learning_rate": 3.151635084276572e-07,
+      "loss": 0.5892,
+      "step": 39147
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5833509777113431,
+      "learning_rate": 3.1497348053746934e-07,
+      "loss": 0.4922,
+      "step": 39148
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3849059767901353,
+      "learning_rate": 3.1478350903657026e-07,
+      "loss": 0.6583,
+      "step": 39149
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5997827247825043,
+      "learning_rate": 3.1459359392607025e-07,
+      "loss": 0.4977,
+      "step": 39150
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.365745703170645,
+      "learning_rate": 3.1440373520707055e-07,
+      "loss": 0.8135,
+      "step": 39151
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3999925422246136,
+      "learning_rate": 3.1421393288068035e-07,
+      "loss": 0.7368,
+      "step": 39152
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2757834075722536,
+      "learning_rate": 3.1402418694800207e-07,
+      "loss": 0.6131,
+      "step": 39153
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4269673318391305,
+      "learning_rate": 3.138344974101426e-07,
+      "loss": 0.6009,
+      "step": 39154
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3200152004638601,
+      "learning_rate": 3.136448642682044e-07,
+      "loss": 0.673,
+      "step": 39155
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5939939679968904,
+      "learning_rate": 3.1345528752329214e-07,
+      "loss": 0.4611,
+      "step": 39156
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.332570232805043,
+      "learning_rate": 3.132657671765105e-07,
+      "loss": 0.6656,
+      "step": 39157
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.256116092651821,
+      "learning_rate": 3.130763032289619e-07,
+      "loss": 0.5824,
+      "step": 39158
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.443877114975719,
+      "learning_rate": 3.12886895681751e-07,
+      "loss": 0.6233,
+      "step": 39159
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.239842264438314,
+      "learning_rate": 3.126975445359781e-07,
+      "loss": 0.6901,
+      "step": 39160
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7716123593171282,
+      "learning_rate": 3.1250824979274675e-07,
+      "loss": 0.6732,
+      "step": 39161
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3678302770305009,
+      "learning_rate": 3.1231901145316156e-07,
+      "loss": 0.8029,
+      "step": 39162
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4872328389989093,
+      "learning_rate": 3.1212982951832063e-07,
+      "loss": 0.7271,
+      "step": 39163
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6813088038121597,
+      "learning_rate": 3.119407039893274e-07,
+      "loss": 0.6313,
+      "step": 39164
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4659252792718291,
+      "learning_rate": 3.1175163486728113e-07,
+      "loss": 0.7803,
+      "step": 39165
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.43815388903808,
+      "learning_rate": 3.115626221532864e-07,
+      "loss": 0.7445,
+      "step": 39166
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.44314458730192,
+      "learning_rate": 3.113736658484412e-07,
+      "loss": 0.6594,
+      "step": 39167
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4329041452115368,
+      "learning_rate": 3.1118476595384363e-07,
+      "loss": 0.7024,
+      "step": 39168
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5797292519963524,
+      "learning_rate": 3.1099592247059715e-07,
+      "loss": 0.6163,
+      "step": 39169
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2541969850842525,
+      "learning_rate": 3.108071353997999e-07,
+      "loss": 0.7288,
+      "step": 39170
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3585179341148137,
+      "learning_rate": 3.10618404742552e-07,
+      "loss": 0.6216,
+      "step": 39171
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6108096343351068,
+      "learning_rate": 3.1042973049994927e-07,
+      "loss": 0.4636,
+      "step": 39172
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.26319247049087,
+      "learning_rate": 3.102411126730942e-07,
+      "loss": 0.5573,
+      "step": 39173
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6083851424517042,
+      "learning_rate": 3.1005255126308255e-07,
+      "loss": 0.6358,
+      "step": 39174
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6260447942730514,
+      "learning_rate": 3.098640462710123e-07,
+      "loss": 0.6221,
+      "step": 39175
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.35409317924282,
+      "learning_rate": 3.096755976979815e-07,
+      "loss": 0.8141,
+      "step": 39176
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4043913469354905,
+      "learning_rate": 3.094872055450882e-07,
+      "loss": 0.654,
+      "step": 39177
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3587159577618557,
+      "learning_rate": 3.0929886981342694e-07,
+      "loss": 0.6791,
+      "step": 39178
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2803844548849483,
+      "learning_rate": 3.0911059050409696e-07,
+      "loss": 0.7185,
+      "step": 39179
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6012872483094162,
+      "learning_rate": 3.0892236761819297e-07,
+      "loss": 0.4868,
+      "step": 39180
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6104351608007175,
+      "learning_rate": 3.0873420115680955e-07,
+      "loss": 0.4891,
+      "step": 39181
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4414820020703492,
+      "learning_rate": 3.085460911210447e-07,
+      "loss": 0.6943,
+      "step": 39182
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2642166553963012,
+      "learning_rate": 3.083580375119932e-07,
+      "loss": 0.6228,
+      "step": 39183
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.637788924490426,
+      "learning_rate": 3.0817004033075083e-07,
+      "loss": 0.6407,
+      "step": 39184
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.673294622156362,
+      "learning_rate": 3.0798209957840886e-07,
+      "loss": 0.5693,
+      "step": 39185
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.508729886407593,
+      "learning_rate": 3.077942152560631e-07,
+      "loss": 0.653,
+      "step": 39186
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5668288454362223,
+      "learning_rate": 3.076063873648094e-07,
+      "loss": 0.4624,
+      "step": 39187
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5841093708749738,
+      "learning_rate": 3.074186159057402e-07,
+      "loss": 0.734,
+      "step": 39188
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4610306249963652,
+      "learning_rate": 3.0723090087994676e-07,
+      "loss": 0.6285,
+      "step": 39189
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4473963549662872,
+      "learning_rate": 3.070432422885239e-07,
+      "loss": 0.7448,
+      "step": 39190
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4349013640446044,
+      "learning_rate": 3.068556401325651e-07,
+      "loss": 0.6915,
+      "step": 39191
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.404345567668822,
+      "learning_rate": 3.066680944131617e-07,
+      "loss": 0.7548,
+      "step": 39192
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.103791967306838,
+      "learning_rate": 3.06480605131404e-07,
+      "loss": 0.667,
+      "step": 39193
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2422091187550923,
+      "learning_rate": 3.062931722883866e-07,
+      "loss": 0.6773,
+      "step": 39194
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.309189023506818,
+      "learning_rate": 3.0610579588519874e-07,
+      "loss": 0.6974,
+      "step": 39195
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.4888894261643384,
+      "learning_rate": 3.0591847592293277e-07,
+      "loss": 0.706,
+      "step": 39196
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.436185103706009,
+      "learning_rate": 3.057312124026768e-07,
+      "loss": 0.654,
+      "step": 39197
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.064196724646587,
+      "learning_rate": 3.055440053255243e-07,
+      "loss": 0.5671,
+      "step": 39198
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4311593935528326,
+      "learning_rate": 3.0535685469256336e-07,
+      "loss": 0.6659,
+      "step": 39199
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3355757525663803,
+      "learning_rate": 3.051697605048842e-07,
+      "loss": 0.6577,
+      "step": 39200
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8838776977203662,
+      "learning_rate": 3.049827227635771e-07,
+      "loss": 0.7285,
+      "step": 39201
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5792689705076743,
+      "learning_rate": 3.0479574146972777e-07,
+      "loss": 0.4872,
+      "step": 39202
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3879044331143502,
+      "learning_rate": 3.046088166244276e-07,
+      "loss": 0.6731,
+      "step": 39203
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6418795205809832,
+      "learning_rate": 3.044219482287647e-07,
+      "loss": 0.4742,
+      "step": 39204
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3064805463861424,
+      "learning_rate": 3.0423513628382805e-07,
+      "loss": 0.663,
+      "step": 39205
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2796908343137812,
+      "learning_rate": 3.040483807907024e-07,
+      "loss": 0.4936,
+      "step": 39206
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4490548820643838,
+      "learning_rate": 3.0386168175047694e-07,
+      "loss": 0.7707,
+      "step": 39207
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2610269998438073,
+      "learning_rate": 3.0367503916423956e-07,
+      "loss": 0.5967,
+      "step": 39208
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3989611304354008,
+      "learning_rate": 3.034884530330762e-07,
+      "loss": 0.6163,
+      "step": 39209
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3966787194740984,
+      "learning_rate": 3.033019233580714e-07,
+      "loss": 0.7035,
+      "step": 39210
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6320206371063335,
+      "learning_rate": 3.0311545014031327e-07,
+      "loss": 0.6643,
+      "step": 39211
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4385735328626683,
+      "learning_rate": 3.0292903338088654e-07,
+      "loss": 0.5981,
+      "step": 39212
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7083347559913868,
+      "learning_rate": 3.0274267308087913e-07,
+      "loss": 0.7112,
+      "step": 39213
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.246402469831438,
+      "learning_rate": 3.0255636924137134e-07,
+      "loss": 0.6271,
+      "step": 39214
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5483350314616207,
+      "learning_rate": 3.0237012186345225e-07,
+      "loss": 0.5912,
+      "step": 39215
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6302669899816176,
+      "learning_rate": 3.021839309482033e-07,
+      "loss": 0.5072,
+      "step": 39216
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4414810278265644,
+      "learning_rate": 3.0199779649670915e-07,
+      "loss": 0.8202,
+      "step": 39217
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2891070520990695,
+      "learning_rate": 3.018117185100566e-07,
+      "loss": 0.7394,
+      "step": 39218
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2932804543876204,
+      "learning_rate": 3.016256969893261e-07,
+      "loss": 0.7786,
+      "step": 39219
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.339509159678488,
+      "learning_rate": 3.0143973193559994e-07,
+      "loss": 0.726,
+      "step": 39220
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5912611422544193,
+      "learning_rate": 3.012538233499618e-07,
+      "loss": 0.4735,
+      "step": 39221
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9097757674208506,
+      "learning_rate": 3.010679712334974e-07,
+      "loss": 0.6406,
+      "step": 39222
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5374946869708175,
+      "learning_rate": 3.008821755872826e-07,
+      "loss": 0.7071,
+      "step": 39223
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4014673579365367,
+      "learning_rate": 3.006964364124032e-07,
+      "loss": 0.705,
+      "step": 39224
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2836843730440166,
+      "learning_rate": 3.005107537099405e-07,
+      "loss": 0.5931,
+      "step": 39225
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5844375676802034,
+      "learning_rate": 3.0032512748097595e-07,
+      "loss": 0.5167,
+      "step": 39226
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3813160335802754,
+      "learning_rate": 3.0013955772658864e-07,
+      "loss": 0.6788,
+      "step": 39227
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.460069768166057,
+      "learning_rate": 2.999540444478588e-07,
+      "loss": 0.6515,
+      "step": 39228
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4804048695651935,
+      "learning_rate": 2.997685876458689e-07,
+      "loss": 0.7396,
+      "step": 39229
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2727036432221193,
+      "learning_rate": 2.9958318732169587e-07,
+      "loss": 0.6323,
+      "step": 39230
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.681581437480715,
+      "learning_rate": 2.993978434764222e-07,
+      "loss": 0.7679,
+      "step": 39231
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6014410212212695,
+      "learning_rate": 2.9921255611112363e-07,
+      "loss": 0.5331,
+      "step": 39232
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4093038448357111,
+      "learning_rate": 2.9902732522688273e-07,
+      "loss": 0.6711,
+      "step": 39233
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4326848167689732,
+      "learning_rate": 2.988421508247741e-07,
+      "loss": 0.7357,
+      "step": 39234
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5301629667257086,
+      "learning_rate": 2.986570329058791e-07,
+      "loss": 0.8275,
+      "step": 39235
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.371457292705261,
+      "learning_rate": 2.9847197147127473e-07,
+      "loss": 0.6509,
+      "step": 39236
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6038022058098869,
+      "learning_rate": 2.982869665220367e-07,
+      "loss": 0.4833,
+      "step": 39237
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.562806912041124,
+      "learning_rate": 2.981020180592431e-07,
+      "loss": 0.6223,
+      "step": 39238
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5571452657935637,
+      "learning_rate": 2.9791712608397306e-07,
+      "loss": 0.4668,
+      "step": 39239
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5601611756283569,
+      "learning_rate": 2.977322905973001e-07,
+      "loss": 0.4756,
+      "step": 39240
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.701605927678716,
+      "learning_rate": 2.975475116003013e-07,
+      "loss": 0.8159,
+      "step": 39241
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2306065287352292,
+      "learning_rate": 2.973627890940522e-07,
+      "loss": 0.5434,
+      "step": 39242
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4617970151157929,
+      "learning_rate": 2.971781230796289e-07,
+      "loss": 0.6026,
+      "step": 39243
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6377709912700413,
+      "learning_rate": 2.969935135581081e-07,
+      "loss": 0.6744,
+      "step": 39244
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4560610996024559,
+      "learning_rate": 2.968089605305613e-07,
+      "loss": 0.6667,
+      "step": 39245
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2362849833317175,
+      "learning_rate": 2.966244639980642e-07,
+      "loss": 0.734,
+      "step": 39246
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5512026972158541,
+      "learning_rate": 2.964400239616927e-07,
+      "loss": 0.713,
+      "step": 39247
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.365429751035941,
+      "learning_rate": 2.962556404225192e-07,
+      "loss": 0.6599,
+      "step": 39248
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4503293766462213,
+      "learning_rate": 2.9607131338161733e-07,
+      "loss": 0.6325,
+      "step": 39249
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2415358961926954,
+      "learning_rate": 2.9588704284006176e-07,
+      "loss": 0.6541,
+      "step": 39250
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.367156959377779,
+      "learning_rate": 2.9570282879892166e-07,
+      "loss": 0.5639,
+      "step": 39251
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4134753051696447,
+      "learning_rate": 2.955186712592739e-07,
+      "loss": 0.6797,
+      "step": 39252
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6433773928551052,
+      "learning_rate": 2.953345702221877e-07,
+      "loss": 0.6797,
+      "step": 39253
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2766263605011703,
+      "learning_rate": 2.951505256887377e-07,
+      "loss": 0.7417,
+      "step": 39254
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5943357308070256,
+      "learning_rate": 2.9496653765999196e-07,
+      "loss": 0.7159,
+      "step": 39255
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5697255880219734,
+      "learning_rate": 2.9478260613702513e-07,
+      "loss": 0.7036,
+      "step": 39256
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5467887004150829,
+      "learning_rate": 2.9459873112090644e-07,
+      "loss": 0.738,
+      "step": 39257
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5299627109889828,
+      "learning_rate": 2.944149126127049e-07,
+      "loss": 0.6541,
+      "step": 39258
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5467245276982526,
+      "learning_rate": 2.9423115061349315e-07,
+      "loss": 0.6534,
+      "step": 39259
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2564539361487976,
+      "learning_rate": 2.940474451243414e-07,
+      "loss": 0.6492,
+      "step": 39260
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2390259945543238,
+      "learning_rate": 2.9386379614631866e-07,
+      "loss": 0.5422,
+      "step": 39261
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.506607812638092,
+      "learning_rate": 2.93680203680492e-07,
+      "loss": 0.7451,
+      "step": 39262
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1877970544897023,
+      "learning_rate": 2.934966677279327e-07,
+      "loss": 0.613,
+      "step": 39263
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5191627188044508,
+      "learning_rate": 2.9331318828971e-07,
+      "loss": 0.5506,
+      "step": 39264
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6207963400851872,
+      "learning_rate": 2.9312976536689076e-07,
+      "loss": 0.5056,
+      "step": 39265
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6789231164962963,
+      "learning_rate": 2.9294639896054187e-07,
+      "loss": 0.5944,
+      "step": 39266
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3257051146006977,
+      "learning_rate": 2.927630890717326e-07,
+      "loss": 0.6952,
+      "step": 39267
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2261433724301665,
+      "learning_rate": 2.925798357015308e-07,
+      "loss": 0.5337,
+      "step": 39268
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4418501226230902,
+      "learning_rate": 2.9239663885100244e-07,
+      "loss": 0.6496,
+      "step": 39269
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4034026578464334,
+      "learning_rate": 2.922134985212133e-07,
+      "loss": 0.7081,
+      "step": 39270
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8279340346700217,
+      "learning_rate": 2.9203041471323136e-07,
+      "loss": 0.6678,
+      "step": 39271
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.505628345812676,
+      "learning_rate": 2.918473874281214e-07,
+      "loss": 0.6516,
+      "step": 39272
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6076239299357931,
+      "learning_rate": 2.9166441666695023e-07,
+      "loss": 0.522,
+      "step": 39273
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3360054034104865,
+      "learning_rate": 2.9148150243078154e-07,
+      "loss": 0.6954,
+      "step": 39274
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0790957639874508,
+      "learning_rate": 2.912986447206811e-07,
+      "loss": 0.5927,
+      "step": 39275
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4341370606340302,
+      "learning_rate": 2.911158435377137e-07,
+      "loss": 0.6384,
+      "step": 39276
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.293229234290861,
+      "learning_rate": 2.9093309888294507e-07,
+      "loss": 0.6374,
+      "step": 39277
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5606389985513993,
+      "learning_rate": 2.907504107574377e-07,
+      "loss": 0.6961,
+      "step": 39278
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3807693083589978,
+      "learning_rate": 2.90567779162253e-07,
+      "loss": 0.6687,
+      "step": 39279
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6914838301873882,
+      "learning_rate": 2.9038520409845785e-07,
+      "loss": 0.6981,
+      "step": 39280
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4035671851695064,
+      "learning_rate": 2.9020268556711475e-07,
+      "loss": 0.6321,
+      "step": 39281
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2458245459596209,
+      "learning_rate": 2.9002022356928615e-07,
+      "loss": 0.6087,
+      "step": 39282
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.664232661222831,
+      "learning_rate": 2.898378181060324e-07,
+      "loss": 0.6715,
+      "step": 39283
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3634656739904132,
+      "learning_rate": 2.8965546917841705e-07,
+      "loss": 0.6376,
+      "step": 39284
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4164434275098188,
+      "learning_rate": 2.894731767875025e-07,
+      "loss": 0.6864,
+      "step": 39285
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.545875406816393,
+      "learning_rate": 2.892909409343503e-07,
+      "loss": 0.4809,
+      "step": 39286
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2873667943364768,
+      "learning_rate": 2.8910876162001835e-07,
+      "loss": 0.7471,
+      "step": 39287
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2814496974320602,
+      "learning_rate": 2.8892663884557024e-07,
+      "loss": 0.708,
+      "step": 39288
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5647422603746695,
+      "learning_rate": 2.8874457261206743e-07,
+      "loss": 0.6439,
+      "step": 39289
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3872978097937865,
+      "learning_rate": 2.8856256292056797e-07,
+      "loss": 0.6986,
+      "step": 39290
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8212310905792577,
+      "learning_rate": 2.883806097721298e-07,
+      "loss": 0.6594,
+      "step": 39291
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4173049640710444,
+      "learning_rate": 2.8819871316781543e-07,
+      "loss": 0.5971,
+      "step": 39292
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4581559532468795,
+      "learning_rate": 2.88016873108683e-07,
+      "loss": 0.607,
+      "step": 39293
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.269976140756482,
+      "learning_rate": 2.8783508959579153e-07,
+      "loss": 0.7582,
+      "step": 39294
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3215458051693596,
+      "learning_rate": 2.8765336263019805e-07,
+      "loss": 0.5896,
+      "step": 39295
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5798415673737425,
+      "learning_rate": 2.874716922129628e-07,
+      "loss": 0.6168,
+      "step": 39296
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6031386100357852,
+      "learning_rate": 2.8729007834514157e-07,
+      "loss": 0.5163,
+      "step": 39297
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.498531608837204,
+      "learning_rate": 2.871085210277924e-07,
+      "loss": 0.8161,
+      "step": 39298
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.37963640658021,
+      "learning_rate": 2.869270202619723e-07,
+      "loss": 0.839,
+      "step": 39299
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5731103140221465,
+      "learning_rate": 2.867455760487392e-07,
+      "loss": 0.842,
+      "step": 39300
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.747020916459928,
+      "learning_rate": 2.865641883891468e-07,
+      "loss": 0.6537,
+      "step": 39301
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3456602987921868,
+      "learning_rate": 2.863828572842542e-07,
+      "loss": 0.5591,
+      "step": 39302
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4500665227178935,
+      "learning_rate": 2.862015827351172e-07,
+      "loss": 0.8388,
+      "step": 39303
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.477349521281742,
+      "learning_rate": 2.860203647427873e-07,
+      "loss": 0.7683,
+      "step": 39304
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3122483445994757,
+      "learning_rate": 2.8583920330832237e-07,
+      "loss": 0.6276,
+      "step": 39305
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.266162422475533,
+      "learning_rate": 2.8565809843277726e-07,
+      "loss": 0.6502,
+      "step": 39306
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.581891206577703,
+      "learning_rate": 2.8547705011720885e-07,
+      "loss": 0.5013,
+      "step": 39307
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.557310277471412,
+      "learning_rate": 2.8529605836266625e-07,
+      "loss": 0.7114,
+      "step": 39308
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3950638716130566,
+      "learning_rate": 2.851151231702054e-07,
+      "loss": 0.5973,
+      "step": 39309
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3229746009453924,
+      "learning_rate": 2.849342445408809e-07,
+      "loss": 0.7307,
+      "step": 39310
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2659148526536836,
+      "learning_rate": 2.8475342247574535e-07,
+      "loss": 0.6448,
+      "step": 39311
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1734386859396833,
+      "learning_rate": 2.8457265697585e-07,
+      "loss": 0.551,
+      "step": 39312
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9804782244706798,
+      "learning_rate": 2.8439194804224857e-07,
+      "loss": 0.646,
+      "step": 39313
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6298883588116837,
+      "learning_rate": 2.842112956759946e-07,
+      "loss": 0.6606,
+      "step": 39314
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2595721022414081,
+      "learning_rate": 2.8403069987813613e-07,
+      "loss": 0.7542,
+      "step": 39315
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.350560656048114,
+      "learning_rate": 2.8385016064972905e-07,
+      "loss": 0.7686,
+      "step": 39316
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4285780733429758,
+      "learning_rate": 2.836696779918224e-07,
+      "loss": 0.6224,
+      "step": 39317
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8104335563691973,
+      "learning_rate": 2.834892519054655e-07,
+      "loss": 0.7125,
+      "step": 39318
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5928533796255565,
+      "learning_rate": 2.8330888239171074e-07,
+      "loss": 0.5156,
+      "step": 39319
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4467837440234572,
+      "learning_rate": 2.831285694516084e-07,
+      "loss": 0.6426,
+      "step": 39320
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.371768420430837,
+      "learning_rate": 2.829483130862076e-07,
+      "loss": 0.5871,
+      "step": 39321
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6288421992800102,
+      "learning_rate": 2.8276811329655764e-07,
+      "loss": 0.4847,
+      "step": 39322
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.747469690286592,
+      "learning_rate": 2.8258797008370865e-07,
+      "loss": 0.8439,
+      "step": 39323
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6142186011525823,
+      "learning_rate": 2.8240788344871094e-07,
+      "loss": 0.5066,
+      "step": 39324
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4151941980784821,
+      "learning_rate": 2.822278533926082e-07,
+      "loss": 0.6732,
+      "step": 39325
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3810911547021865,
+      "learning_rate": 2.8204787991645163e-07,
+      "loss": 0.7165,
+      "step": 39326
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7810930189568037,
+      "learning_rate": 2.818679630212895e-07,
+      "loss": 0.7179,
+      "step": 39327
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.215180140314139,
+      "learning_rate": 2.816881027081697e-07,
+      "loss": 0.5993,
+      "step": 39328
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9108063451439963,
+      "learning_rate": 2.8150829897813814e-07,
+      "loss": 0.6401,
+      "step": 39329
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6029114754987871,
+      "learning_rate": 2.813285518322417e-07,
+      "loss": 0.5002,
+      "step": 39330
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.507846154138254,
+      "learning_rate": 2.811488612715274e-07,
+      "loss": 0.6661,
+      "step": 39331
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5769493860267545,
+      "learning_rate": 2.80969227297041e-07,
+      "loss": 0.4578,
+      "step": 39332
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.597290346766859,
+      "learning_rate": 2.807896499098295e-07,
+      "loss": 0.7053,
+      "step": 39333
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.479525144340969,
+      "learning_rate": 2.8061012911093756e-07,
+      "loss": 0.6187,
+      "step": 39334
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2283519853980545,
+      "learning_rate": 2.804306649014099e-07,
+      "loss": 0.5988,
+      "step": 39335
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3525947215276939,
+      "learning_rate": 2.802512572822924e-07,
+      "loss": 0.7168,
+      "step": 39336
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5853947928296874,
+      "learning_rate": 2.800719062546298e-07,
+      "loss": 0.4826,
+      "step": 39337
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.596207347495547,
+      "learning_rate": 2.798926118194667e-07,
+      "loss": 0.8106,
+      "step": 39338
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2204039561478452,
+      "learning_rate": 2.797133739778435e-07,
+      "loss": 0.575,
+      "step": 39339
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5883573289482127,
+      "learning_rate": 2.795341927308082e-07,
+      "loss": 0.4855,
+      "step": 39340
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6565606307008114,
+      "learning_rate": 2.793550680794022e-07,
+      "loss": 0.7251,
+      "step": 39341
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5225763888678925,
+      "learning_rate": 2.79176000024669e-07,
+      "loss": 0.7393,
+      "step": 39342
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8434467785664534,
+      "learning_rate": 2.789969885676491e-07,
+      "loss": 0.7405,
+      "step": 39343
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4482713498439308,
+      "learning_rate": 2.78818033709386e-07,
+      "loss": 0.5618,
+      "step": 39344
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2675110970930095,
+      "learning_rate": 2.7863913545092434e-07,
+      "loss": 0.7449,
+      "step": 39345
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.285025736806498,
+      "learning_rate": 2.784602937933023e-07,
+      "loss": 0.6451,
+      "step": 39346
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4680563305650698,
+      "learning_rate": 2.7828150873756124e-07,
+      "loss": 0.6051,
+      "step": 39347
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4206223252352455,
+      "learning_rate": 2.781027802847436e-07,
+      "loss": 0.6329,
+      "step": 39348
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2705409799501488,
+      "learning_rate": 2.779241084358908e-07,
+      "loss": 0.624,
+      "step": 39349
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.829138343438871,
+      "learning_rate": 2.777454931920409e-07,
+      "loss": 0.7234,
+      "step": 39350
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.301256384244121,
+      "learning_rate": 2.775669345542342e-07,
+      "loss": 0.6021,
+      "step": 39351
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3291790347951369,
+      "learning_rate": 2.773884325235121e-07,
+      "loss": 0.7351,
+      "step": 39352
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3519556366949461,
+      "learning_rate": 2.772099871009115e-07,
+      "loss": 0.7525,
+      "step": 39353
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6217132825176355,
+      "learning_rate": 2.770315982874727e-07,
+      "loss": 0.5024,
+      "step": 39354
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2595742970036627,
+      "learning_rate": 2.768532660842349e-07,
+      "loss": 0.6735,
+      "step": 39355
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4137844181004928,
+      "learning_rate": 2.76674990492235e-07,
+      "loss": 0.6627,
+      "step": 39356
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.361192059716501,
+      "learning_rate": 2.764967715125111e-07,
+      "loss": 0.6407,
+      "step": 39357
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5453926026295028,
+      "learning_rate": 2.763186091461034e-07,
+      "loss": 0.6121,
+      "step": 39358
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3667012895253765,
+      "learning_rate": 2.761405033940467e-07,
+      "loss": 0.7536,
+      "step": 39359
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3783085887258164,
+      "learning_rate": 2.759624542573769e-07,
+      "loss": 0.7746,
+      "step": 39360
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.352767910558515,
+      "learning_rate": 2.7578446173713305e-07,
+      "loss": 0.5927,
+      "step": 39361
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.335274058916124,
+      "learning_rate": 2.7560652583435097e-07,
+      "loss": 0.6757,
+      "step": 39362
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7023618762675066,
+      "learning_rate": 2.754286465500666e-07,
+      "loss": 0.8047,
+      "step": 39363
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0363847272001303,
+      "learning_rate": 2.7525082388531465e-07,
+      "loss": 0.6448,
+      "step": 39364
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4115823134478505,
+      "learning_rate": 2.7507305784113094e-07,
+      "loss": 0.6275,
+      "step": 39365
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.502658913879527,
+      "learning_rate": 2.7489534841855235e-07,
+      "loss": 0.5201,
+      "step": 39366
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2669017274954761,
+      "learning_rate": 2.747176956186115e-07,
+      "loss": 0.5796,
+      "step": 39367
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0282074444923945,
+      "learning_rate": 2.7454009944234195e-07,
+      "loss": 0.7555,
+      "step": 39368
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2935160062489384,
+      "learning_rate": 2.7436255989077844e-07,
+      "loss": 0.7037,
+      "step": 39369
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3378146259948214,
+      "learning_rate": 2.741850769649579e-07,
+      "loss": 0.6814,
+      "step": 39370
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6139072497571922,
+      "learning_rate": 2.740076506659095e-07,
+      "loss": 0.5414,
+      "step": 39371
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.27665320988583,
+      "learning_rate": 2.738302809946669e-07,
+      "loss": 0.6958,
+      "step": 39372
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.9666968414015353,
+      "learning_rate": 2.736529679522659e-07,
+      "loss": 0.6106,
+      "step": 39373
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3960789437178793,
+      "learning_rate": 2.734757115397346e-07,
+      "loss": 0.6661,
+      "step": 39374
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2898543135180767,
+      "learning_rate": 2.732985117581077e-07,
+      "loss": 0.6802,
+      "step": 39375
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4747248442942356,
+      "learning_rate": 2.7312136860841666e-07,
+      "loss": 0.7957,
+      "step": 39376
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5016888676968738,
+      "learning_rate": 2.729442820916928e-07,
+      "loss": 0.684,
+      "step": 39377
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3269231326425377,
+      "learning_rate": 2.7276725220896526e-07,
+      "loss": 0.5503,
+      "step": 39378
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3069673800286945,
+      "learning_rate": 2.7259027896126776e-07,
+      "loss": 0.5419,
+      "step": 39379
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.561159123647299,
+      "learning_rate": 2.7241336234962943e-07,
+      "loss": 0.6103,
+      "step": 39380
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4063096204137748,
+      "learning_rate": 2.722365023750795e-07,
+      "loss": 0.6275,
+      "step": 39381
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3374727217962838,
+      "learning_rate": 2.7205969903864814e-07,
+      "loss": 0.7165,
+      "step": 39382
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3059906799659784,
+      "learning_rate": 2.7188295234136575e-07,
+      "loss": 0.7055,
+      "step": 39383
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4161810832738124,
+      "learning_rate": 2.717062622842603e-07,
+      "loss": 0.7322,
+      "step": 39384
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3111504579967261,
+      "learning_rate": 2.715296288683611e-07,
+      "loss": 0.5183,
+      "step": 39385
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.407969123236368,
+      "learning_rate": 2.7135305209469607e-07,
+      "loss": 0.6627,
+      "step": 39386
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2329840703681265,
+      "learning_rate": 2.7117653196429337e-07,
+      "loss": 0.6734,
+      "step": 39387
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3861064562223324,
+      "learning_rate": 2.7100006847818214e-07,
+      "loss": 0.7447,
+      "step": 39388
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.227278998954652,
+      "learning_rate": 2.7082366163738825e-07,
+      "loss": 0.5531,
+      "step": 39389
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4558188939802843,
+      "learning_rate": 2.7064731144293755e-07,
+      "loss": 0.6961,
+      "step": 39390
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.797442092658316,
+      "learning_rate": 2.7047101789586026e-07,
+      "loss": 0.6091,
+      "step": 39391
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3550149982365765,
+      "learning_rate": 2.702947809971812e-07,
+      "loss": 0.7519,
+      "step": 39392
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5877082745518194,
+      "learning_rate": 2.7011860074792616e-07,
+      "loss": 0.515,
+      "step": 39393
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7345922095273887,
+      "learning_rate": 2.6994247714912103e-07,
+      "loss": 0.6962,
+      "step": 39394
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.324603474117356,
+      "learning_rate": 2.697664102017905e-07,
+      "loss": 0.5921,
+      "step": 39395
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6236085396255634,
+      "learning_rate": 2.6959039990696046e-07,
+      "loss": 0.4959,
+      "step": 39396
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5910308894268348,
+      "learning_rate": 2.694144462656556e-07,
+      "loss": 0.4819,
+      "step": 39397
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3825557061025078,
+      "learning_rate": 2.692385492789018e-07,
+      "loss": 0.6426,
+      "step": 39398
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7107114769219294,
+      "learning_rate": 2.6906270894772047e-07,
+      "loss": 0.6907,
+      "step": 39399
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.448466488061206,
+      "learning_rate": 2.688869252731374e-07,
+      "loss": 0.6552,
+      "step": 39400
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6708456696083696,
+      "learning_rate": 2.687111982561763e-07,
+      "loss": 0.6514,
+      "step": 39401
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.438081419042308,
+      "learning_rate": 2.6853552789785855e-07,
+      "loss": 0.6407,
+      "step": 39402
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3850571050354685,
+      "learning_rate": 2.683599141992077e-07,
+      "loss": 0.6288,
+      "step": 39403
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6133480827170662,
+      "learning_rate": 2.6818435716124525e-07,
+      "loss": 0.4959,
+      "step": 39404
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.242501432436032,
+      "learning_rate": 2.680088567849981e-07,
+      "loss": 0.6528,
+      "step": 39405
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3313071795697775,
+      "learning_rate": 2.6783341307148103e-07,
+      "loss": 0.5987,
+      "step": 39406
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3230465296355174,
+      "learning_rate": 2.6765802602171987e-07,
+      "loss": 0.6798,
+      "step": 39407
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1492379548972371,
+      "learning_rate": 2.674826956367349e-07,
+      "loss": 0.5599,
+      "step": 39408
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6230796462855033,
+      "learning_rate": 2.6730742191754864e-07,
+      "loss": 0.7502,
+      "step": 39409
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3781772631221432,
+      "learning_rate": 2.671322048651781e-07,
+      "loss": 0.6075,
+      "step": 39410
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6610520062241498,
+      "learning_rate": 2.6695704448064575e-07,
+      "loss": 0.7165,
+      "step": 39411
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4584238140437418,
+      "learning_rate": 2.6678194076497187e-07,
+      "loss": 0.6575,
+      "step": 39412
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4317670151867885,
+      "learning_rate": 2.6660689371917346e-07,
+      "loss": 0.6028,
+      "step": 39413
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4177903411405957,
+      "learning_rate": 2.6643190334427194e-07,
+      "loss": 0.6081,
+      "step": 39414
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4676876995836141,
+      "learning_rate": 2.662569696412842e-07,
+      "loss": 0.6684,
+      "step": 39415
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3945258319356126,
+      "learning_rate": 2.6608209261123176e-07,
+      "loss": 0.6178,
+      "step": 39416
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2947649689762708,
+      "learning_rate": 2.659072722551304e-07,
+      "loss": 0.5809,
+      "step": 39417
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6168691998640927,
+      "learning_rate": 2.657325085739992e-07,
+      "loss": 0.5087,
+      "step": 39418
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6767204127020885,
+      "learning_rate": 2.6555780156885533e-07,
+      "loss": 0.5715,
+      "step": 39419
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6165240432384855,
+      "learning_rate": 2.6538315124071454e-07,
+      "loss": 0.4889,
+      "step": 39420
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4503281751010728,
+      "learning_rate": 2.652085575905949e-07,
+      "loss": 0.7219,
+      "step": 39421
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4085694242261348,
+      "learning_rate": 2.6503402061951456e-07,
+      "loss": 0.643,
+      "step": 39422
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4726223088348092,
+      "learning_rate": 2.6485954032848814e-07,
+      "loss": 0.6758,
+      "step": 39423
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4111800002854962,
+      "learning_rate": 2.6468511671853047e-07,
+      "loss": 0.5951,
+      "step": 39424
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5317625222691131,
+      "learning_rate": 2.645107497906574e-07,
+      "loss": 0.6839,
+      "step": 39425
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4392281597439782,
+      "learning_rate": 2.6433643954588695e-07,
+      "loss": 0.7164,
+      "step": 39426
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4413042832898353,
+      "learning_rate": 2.641621859852317e-07,
+      "loss": 0.7264,
+      "step": 39427
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2765687831617485,
+      "learning_rate": 2.6398798910970525e-07,
+      "loss": 0.6582,
+      "step": 39428
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3015230837397394,
+      "learning_rate": 2.638138489203235e-07,
+      "loss": 0.7232,
+      "step": 39429
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7313754400159613,
+      "learning_rate": 2.636397654181011e-07,
+      "loss": 0.8542,
+      "step": 39430
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4333157192413661,
+      "learning_rate": 2.6346573860404954e-07,
+      "loss": 0.596,
+      "step": 39431
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3331299979720963,
+      "learning_rate": 2.6329176847918245e-07,
+      "loss": 0.6525,
+      "step": 39432
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4621152784821916,
+      "learning_rate": 2.631178550445146e-07,
+      "loss": 0.6797,
+      "step": 39433
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.562960112730388,
+      "learning_rate": 2.629439983010562e-07,
+      "loss": 0.4431,
+      "step": 39434
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4406154400660935,
+      "learning_rate": 2.627701982498221e-07,
+      "loss": 0.638,
+      "step": 39435
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2858507954115408,
+      "learning_rate": 2.625964548918203e-07,
+      "loss": 0.5896,
+      "step": 39436
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6031641288837286,
+      "learning_rate": 2.6242276822806666e-07,
+      "loss": 0.5279,
+      "step": 39437
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.314014860320336,
+      "learning_rate": 2.6224913825956933e-07,
+      "loss": 0.708,
+      "step": 39438
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.482856643188202,
+      "learning_rate": 2.6207556498734076e-07,
+      "loss": 0.7781,
+      "step": 39439
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6153257136388761,
+      "learning_rate": 2.6190204841239244e-07,
+      "loss": 0.5468,
+      "step": 39440
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6087830809934446,
+      "learning_rate": 2.6172858853573125e-07,
+      "loss": 0.4524,
+      "step": 39441
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4113860030936596,
+      "learning_rate": 2.6155518535836976e-07,
+      "loss": 0.6818,
+      "step": 39442
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.477091008724234,
+      "learning_rate": 2.6138183888131827e-07,
+      "loss": 0.7304,
+      "step": 39443
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4995805698211895,
+      "learning_rate": 2.612085491055849e-07,
+      "loss": 0.7387,
+      "step": 39444
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.21094846216987,
+      "learning_rate": 2.610353160321777e-07,
+      "loss": 0.5661,
+      "step": 39445
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3401208602425558,
+      "learning_rate": 2.6086213966210697e-07,
+      "loss": 0.7591,
+      "step": 39446
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4503968223054555,
+      "learning_rate": 2.6068901999637965e-07,
+      "loss": 0.5198,
+      "step": 39447
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7010347729413304,
+      "learning_rate": 2.6051595703600494e-07,
+      "loss": 0.6435,
+      "step": 39448
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6539324229374865,
+      "learning_rate": 2.603429507819888e-07,
+      "loss": 0.6205,
+      "step": 39449
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3618031828344477,
+      "learning_rate": 2.601700012353392e-07,
+      "loss": 0.6207,
+      "step": 39450
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1830542454929889,
+      "learning_rate": 2.5999710839706537e-07,
+      "loss": 0.6222,
+      "step": 39451
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.339218016982403,
+      "learning_rate": 2.59824272268171e-07,
+      "loss": 0.6942,
+      "step": 39452
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3561548607861007,
+      "learning_rate": 2.5965149284966297e-07,
+      "loss": 0.6218,
+      "step": 39453
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2636046284043538,
+      "learning_rate": 2.594787701425483e-07,
+      "loss": 0.6296,
+      "step": 39454
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5567535100848529,
+      "learning_rate": 2.5930610414783177e-07,
+      "loss": 0.5599,
+      "step": 39455
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3699250983429947,
+      "learning_rate": 2.591334948665192e-07,
+      "loss": 0.7349,
+      "step": 39456
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3259306902067567,
+      "learning_rate": 2.589609422996142e-07,
+      "loss": 0.7043,
+      "step": 39457
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.349445677498687,
+      "learning_rate": 2.587884464481238e-07,
+      "loss": 0.7369,
+      "step": 39458
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2530274033270035,
+      "learning_rate": 2.586160073130506e-07,
+      "loss": 0.7398,
+      "step": 39459
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3011781775062412,
+      "learning_rate": 2.584436248954003e-07,
+      "loss": 0.646,
+      "step": 39460
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5738982323498937,
+      "learning_rate": 2.582712991961744e-07,
+      "loss": 0.4612,
+      "step": 39461
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3603347987294292,
+      "learning_rate": 2.580990302163766e-07,
+      "loss": 0.5566,
+      "step": 39462
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7620585653255882,
+      "learning_rate": 2.5792681795701046e-07,
+      "loss": 0.6485,
+      "step": 39463
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.541475747952445,
+      "learning_rate": 2.577546624190797e-07,
+      "loss": 0.7476,
+      "step": 39464
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.45153612598548,
+      "learning_rate": 2.5758256360358444e-07,
+      "loss": 0.6003,
+      "step": 39465
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4385545530938098,
+      "learning_rate": 2.5741052151152855e-07,
+      "loss": 0.7525,
+      "step": 39466
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.47627878074921,
+      "learning_rate": 2.5723853614391225e-07,
+      "loss": 0.634,
+      "step": 39467
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8670416339509528,
+      "learning_rate": 2.5706660750173916e-07,
+      "loss": 0.6955,
+      "step": 39468
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3465463813489857,
+      "learning_rate": 2.568947355860085e-07,
+      "loss": 0.6693,
+      "step": 39469
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.531708638758207,
+      "learning_rate": 2.567229203977206e-07,
+      "loss": 0.5996,
+      "step": 39470
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4424677214985104,
+      "learning_rate": 2.565511619378758e-07,
+      "loss": 0.7335,
+      "step": 39471
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4302941483272813,
+      "learning_rate": 2.563794602074765e-07,
+      "loss": 0.7473,
+      "step": 39472
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2627571652326934,
+      "learning_rate": 2.562078152075209e-07,
+      "loss": 0.611,
+      "step": 39473
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2363733959174603,
+      "learning_rate": 2.560362269390071e-07,
+      "loss": 0.5677,
+      "step": 39474
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3645759871240328,
+      "learning_rate": 2.5586469540293643e-07,
+      "loss": 0.5815,
+      "step": 39475
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1944634264928822,
+      "learning_rate": 2.5569322060030487e-07,
+      "loss": 0.6825,
+      "step": 39476
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6278025101498252,
+      "learning_rate": 2.555218025321138e-07,
+      "loss": 0.6276,
+      "step": 39477
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5418057326623287,
+      "learning_rate": 2.5535044119935903e-07,
+      "loss": 0.6544,
+      "step": 39478
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2545748989211758,
+      "learning_rate": 2.5517913660303984e-07,
+      "loss": 0.6642,
+      "step": 39479
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4314904769903343,
+      "learning_rate": 2.5500788874415097e-07,
+      "loss": 0.5914,
+      "step": 39480
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.459327923223025,
+      "learning_rate": 2.5483669762369377e-07,
+      "loss": 0.758,
+      "step": 39481
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4428906538180897,
+      "learning_rate": 2.54665563242662e-07,
+      "loss": 0.5767,
+      "step": 39482
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3542170528088657,
+      "learning_rate": 2.544944856020515e-07,
+      "loss": 0.7259,
+      "step": 39483
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.407546518532391,
+      "learning_rate": 2.543234647028603e-07,
+      "loss": 0.6545,
+      "step": 39484
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2327604670761427,
+      "learning_rate": 2.541525005460821e-07,
+      "loss": 0.5705,
+      "step": 39485
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2257471648801816,
+      "learning_rate": 2.539815931327161e-07,
+      "loss": 0.6165,
+      "step": 39486
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4548603257186195,
+      "learning_rate": 2.538107424637526e-07,
+      "loss": 0.7358,
+      "step": 39487
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4067075175723633,
+      "learning_rate": 2.536399485401897e-07,
+      "loss": 0.8223,
+      "step": 39488
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5898584376252769,
+      "learning_rate": 2.5346921136301995e-07,
+      "loss": 0.5257,
+      "step": 39489
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3460466135924374,
+      "learning_rate": 2.532985309332392e-07,
+      "loss": 0.5194,
+      "step": 39490
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3128135349589802,
+      "learning_rate": 2.531279072518389e-07,
+      "loss": 0.5965,
+      "step": 39491
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.33329801219862,
+      "learning_rate": 2.529573403198138e-07,
+      "loss": 0.7883,
+      "step": 39492
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.363803616279274,
+      "learning_rate": 2.5278683013815755e-07,
+      "loss": 0.6827,
+      "step": 39493
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6084972424093684,
+      "learning_rate": 2.5261637670786266e-07,
+      "loss": 0.6706,
+      "step": 39494
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4072947673500813,
+      "learning_rate": 2.524459800299206e-07,
+      "loss": 0.5325,
+      "step": 39495
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6589545621126012,
+      "learning_rate": 2.522756401053239e-07,
+      "loss": 0.6349,
+      "step": 39496
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6836061500714437,
+      "learning_rate": 2.521053569350651e-07,
+      "loss": 0.5832,
+      "step": 39497
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6268167846855942,
+      "learning_rate": 2.5193513052013343e-07,
+      "loss": 0.5268,
+      "step": 39498
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5983409162971118,
+      "learning_rate": 2.517649608615236e-07,
+      "loss": 0.862,
+      "step": 39499
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6807031404404926,
+      "learning_rate": 2.515948479602237e-07,
+      "loss": 0.5619,
+      "step": 39500
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6619899951233691,
+      "learning_rate": 2.5142479181722525e-07,
+      "loss": 0.5331,
+      "step": 39501
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3600949831493772,
+      "learning_rate": 2.512547924335162e-07,
+      "loss": 0.7096,
+      "step": 39502
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2652967832068898,
+      "learning_rate": 2.510848498100915e-07,
+      "loss": 0.7134,
+      "step": 39503
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4352181542201279,
+      "learning_rate": 2.509149639479347e-07,
+      "loss": 0.5577,
+      "step": 39504
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2080699283208307,
+      "learning_rate": 2.507451348480372e-07,
+      "loss": 0.6134,
+      "step": 39505
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.491051765447579,
+      "learning_rate": 2.505753625113894e-07,
+      "loss": 0.7184,
+      "step": 39506
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5935022145189393,
+      "learning_rate": 2.504056469389793e-07,
+      "loss": 0.4752,
+      "step": 39507
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6951951405124006,
+      "learning_rate": 2.502359881317917e-07,
+      "loss": 0.6375,
+      "step": 39508
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3523145879611789,
+      "learning_rate": 2.500663860908181e-07,
+      "loss": 0.66,
+      "step": 39509
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7559088075969784,
+      "learning_rate": 2.4989684081704545e-07,
+      "loss": 0.6854,
+      "step": 39510
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2477450610496237,
+      "learning_rate": 2.497273523114585e-07,
+      "loss": 0.8163,
+      "step": 39511
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5619806644493257,
+      "learning_rate": 2.495579205750476e-07,
+      "loss": 0.7131,
+      "step": 39512
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.397927577245617,
+      "learning_rate": 2.493885456087963e-07,
+      "loss": 0.641,
+      "step": 39513
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2595291170391596,
+      "learning_rate": 2.4921922741369286e-07,
+      "loss": 0.6625,
+      "step": 39514
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6864510743039356,
+      "learning_rate": 2.4904996599072087e-07,
+      "loss": 0.5599,
+      "step": 39515
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3384122503929978,
+      "learning_rate": 2.4888076134086834e-07,
+      "loss": 0.7394,
+      "step": 39516
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.086117549978156,
+      "learning_rate": 2.4871161346511906e-07,
+      "loss": 0.5448,
+      "step": 39517
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.523204352757161,
+      "learning_rate": 2.4854252236445664e-07,
+      "loss": 0.7305,
+      "step": 39518
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.216098847402547,
+      "learning_rate": 2.4837348803986803e-07,
+      "loss": 0.4987,
+      "step": 39519
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2448130918906337,
+      "learning_rate": 2.4820451049233585e-07,
+      "loss": 0.6724,
+      "step": 39520
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2772531176146704,
+      "learning_rate": 2.480355897228448e-07,
+      "loss": 0.6675,
+      "step": 39521
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2972149050607398,
+      "learning_rate": 2.4786672573237744e-07,
+      "loss": 0.75,
+      "step": 39522
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4453210264998313,
+      "learning_rate": 2.476979185219175e-07,
+      "loss": 0.6556,
+      "step": 39523
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6273110683127556,
+      "learning_rate": 2.4752916809244856e-07,
+      "loss": 0.7477,
+      "step": 39524
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5916670094873362,
+      "learning_rate": 2.4736047444495206e-07,
+      "loss": 0.4464,
+      "step": 39525
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3836869284750508,
+      "learning_rate": 2.4719183758041056e-07,
+      "loss": 0.6432,
+      "step": 39526
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1838700336927177,
+      "learning_rate": 2.470232574998044e-07,
+      "loss": 0.6405,
+      "step": 39527
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6072927056486281,
+      "learning_rate": 2.468547342041183e-07,
+      "loss": 0.7212,
+      "step": 39528
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3472426659679346,
+      "learning_rate": 2.4668626769433155e-07,
+      "loss": 0.7229,
+      "step": 39529
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2639904281928434,
+      "learning_rate": 2.4651785797142447e-07,
+      "loss": 0.554,
+      "step": 39530
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6010702845932125,
+      "learning_rate": 2.4634950503637957e-07,
+      "loss": 0.7292,
+      "step": 39531
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3970350442115078,
+      "learning_rate": 2.461812088901749e-07,
+      "loss": 0.7941,
+      "step": 39532
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3674437839985027,
+      "learning_rate": 2.4601296953379094e-07,
+      "loss": 0.6442,
+      "step": 39533
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3652150046853966,
+      "learning_rate": 2.458447869682079e-07,
+      "loss": 0.635,
+      "step": 39534
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7271159732949357,
+      "learning_rate": 2.4567666119440504e-07,
+      "loss": 0.7303,
+      "step": 39535
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.358372422046292,
+      "learning_rate": 2.455085922133604e-07,
+      "loss": 0.5642,
+      "step": 39536
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4229477372348969,
+      "learning_rate": 2.4534058002605334e-07,
+      "loss": 0.7617,
+      "step": 39537
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5096190077847351,
+      "learning_rate": 2.451726246334607e-07,
+      "loss": 0.6549,
+      "step": 39538
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3635687610495757,
+      "learning_rate": 2.450047260365618e-07,
+      "loss": 0.6145,
+      "step": 39539
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.424889310887469,
+      "learning_rate": 2.4483688423633356e-07,
+      "loss": 0.7459,
+      "step": 39540
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5920248469376677,
+      "learning_rate": 2.4466909923375414e-07,
+      "loss": 0.5148,
+      "step": 39541
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5740474767102284,
+      "learning_rate": 2.445013710298005e-07,
+      "loss": 0.7515,
+      "step": 39542
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.18619462877378,
+      "learning_rate": 2.443336996254464e-07,
+      "loss": 0.6964,
+      "step": 39543
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5910322392726376,
+      "learning_rate": 2.4416608502167095e-07,
+      "loss": 0.4843,
+      "step": 39544
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.395344029116707,
+      "learning_rate": 2.4399852721944894e-07,
+      "loss": 0.6464,
+      "step": 39545
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4593504372125612,
+      "learning_rate": 2.438310262197574e-07,
+      "loss": 0.747,
+      "step": 39546
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5133159349988727,
+      "learning_rate": 2.4366358202356887e-07,
+      "loss": 0.6458,
+      "step": 39547
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2345487763091068,
+      "learning_rate": 2.4349619463186036e-07,
+      "loss": 0.6675,
+      "step": 39548
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3453002304702713,
+      "learning_rate": 2.4332886404560553e-07,
+      "loss": 0.5993,
+      "step": 39549
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5868634677157633,
+      "learning_rate": 2.431615902657802e-07,
+      "loss": 0.4872,
+      "step": 39550
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.676178689152049,
+      "learning_rate": 2.4299437329335595e-07,
+      "loss": 0.7309,
+      "step": 39551
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3996166246756196,
+      "learning_rate": 2.428272131293075e-07,
+      "loss": 0.7303,
+      "step": 39552
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3787531310225076,
+      "learning_rate": 2.426601097746084e-07,
+      "loss": 0.7344,
+      "step": 39553
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6066888367742826,
+      "learning_rate": 2.424930632302314e-07,
+      "loss": 0.5315,
+      "step": 39554
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5775311658898232,
+      "learning_rate": 2.4232607349714775e-07,
+      "loss": 0.7612,
+      "step": 39555
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2779799896545532,
+      "learning_rate": 2.421591405763324e-07,
+      "loss": 0.6706,
+      "step": 39556
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2728186982463268,
+      "learning_rate": 2.4199226446875557e-07,
+      "loss": 0.6204,
+      "step": 39557
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4794368648375487,
+      "learning_rate": 2.4182544517538874e-07,
+      "loss": 0.7729,
+      "step": 39558
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2641243697098699,
+      "learning_rate": 2.416586826972034e-07,
+      "loss": 0.7266,
+      "step": 39559
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4084215926300925,
+      "learning_rate": 2.414919770351709e-07,
+      "loss": 0.6426,
+      "step": 39560
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6231441067738971,
+      "learning_rate": 2.4132532819026165e-07,
+      "loss": 0.4975,
+      "step": 39561
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.628598203132208,
+      "learning_rate": 2.41158736163446e-07,
+      "loss": 0.7665,
+      "step": 39562
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4493860598380657,
+      "learning_rate": 2.409922009556942e-07,
+      "loss": 0.747,
+      "step": 39563
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3936564059119285,
+      "learning_rate": 2.408257225679744e-07,
+      "loss": 0.8061,
+      "step": 39564
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8498623466413064,
+      "learning_rate": 2.40659301001257e-07,
+      "loss": 0.6332,
+      "step": 39565
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.313214894701418,
+      "learning_rate": 2.4049293625651224e-07,
+      "loss": 0.616,
+      "step": 39566
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3089610786338943,
+      "learning_rate": 2.403266283347061e-07,
+      "loss": 0.756,
+      "step": 39567
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.410103632904662,
+      "learning_rate": 2.4016037723680886e-07,
+      "loss": 0.8335,
+      "step": 39568
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6082868038078695,
+      "learning_rate": 2.3999418296378643e-07,
+      "loss": 0.5487,
+      "step": 39569
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4057643628315009,
+      "learning_rate": 2.398280455166091e-07,
+      "loss": 0.6383,
+      "step": 39570
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.305587054382691,
+      "learning_rate": 2.396619648962439e-07,
+      "loss": 0.6177,
+      "step": 39571
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7938173811369964,
+      "learning_rate": 2.394959411036546e-07,
+      "loss": 0.7936,
+      "step": 39572
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7379292637446577,
+      "learning_rate": 2.393299741398114e-07,
+      "loss": 0.6311,
+      "step": 39573
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6421886850293033,
+      "learning_rate": 2.3916406400567915e-07,
+      "loss": 0.6457,
+      "step": 39574
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5880272588595331,
+      "learning_rate": 2.3899821070222375e-07,
+      "loss": 0.4887,
+      "step": 39575
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3444237315791794,
+      "learning_rate": 2.38832414230411e-07,
+      "loss": 0.6553,
+      "step": 39576
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6547842839008822,
+      "learning_rate": 2.38666674591207e-07,
+      "loss": 0.7426,
+      "step": 39577
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.581894813362049,
+      "learning_rate": 2.3850099178557517e-07,
+      "loss": 0.695,
+      "step": 39578
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3841296155211555,
+      "learning_rate": 2.3833536581448157e-07,
+      "loss": 0.6479,
+      "step": 39579
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.372310187213625,
+      "learning_rate": 2.3816979667888872e-07,
+      "loss": 0.7292,
+      "step": 39580
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1142309207161367,
+      "learning_rate": 2.3800428437976364e-07,
+      "loss": 0.6743,
+      "step": 39581
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5692155272884185,
+      "learning_rate": 2.3783882891806664e-07,
+      "loss": 0.7246,
+      "step": 39582
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2133243739422306,
+      "learning_rate": 2.3767343029476253e-07,
+      "loss": 0.704,
+      "step": 39583
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3725183408295438,
+      "learning_rate": 2.3750808851081607e-07,
+      "loss": 0.702,
+      "step": 39584
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5952899157029925,
+      "learning_rate": 2.3734280356718764e-07,
+      "loss": 0.6616,
+      "step": 39585
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4613835304951965,
+      "learning_rate": 2.3717757546483866e-07,
+      "loss": 0.5662,
+      "step": 39586
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5355014274377812,
+      "learning_rate": 2.370124042047339e-07,
+      "loss": 0.811,
+      "step": 39587
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2628132251497746,
+      "learning_rate": 2.368472897878349e-07,
+      "loss": 0.6133,
+      "step": 39588
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2723891053913017,
+      "learning_rate": 2.3668223221510078e-07,
+      "loss": 0.5694,
+      "step": 39589
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.861825752681915,
+      "learning_rate": 2.3651723148749305e-07,
+      "loss": 0.6467,
+      "step": 39590
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3526936226177952,
+      "learning_rate": 2.3635228760597429e-07,
+      "loss": 0.6905,
+      "step": 39591
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3582712668480572,
+      "learning_rate": 2.3618740057150368e-07,
+      "loss": 0.6499,
+      "step": 39592
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.539997949326216,
+      "learning_rate": 2.360225703850405e-07,
+      "loss": 0.5858,
+      "step": 39593
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3422543349826317,
+      "learning_rate": 2.3585779704754507e-07,
+      "loss": 0.6516,
+      "step": 39594
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2289925157543358,
+      "learning_rate": 2.3569308055997776e-07,
+      "loss": 0.5905,
+      "step": 39595
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.66182176572485,
+      "learning_rate": 2.3552842092329554e-07,
+      "loss": 0.6634,
+      "step": 39596
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.304629660931515,
+      "learning_rate": 2.3536381813845878e-07,
+      "loss": 0.7349,
+      "step": 39597
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2773750594850448,
+      "learning_rate": 2.3519927220642558e-07,
+      "loss": 0.6921,
+      "step": 39598
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4397282907060678,
+      "learning_rate": 2.3503478312815298e-07,
+      "loss": 0.6445,
+      "step": 39599
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5893317127523873,
+      "learning_rate": 2.348703509045991e-07,
+      "loss": 0.4664,
+      "step": 39600
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3016928251979367,
+      "learning_rate": 2.3470597553672314e-07,
+      "loss": 0.6999,
+      "step": 39601
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9060382312095197,
+      "learning_rate": 2.3454165702547993e-07,
+      "loss": 0.626,
+      "step": 39602
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6211077349337762,
+      "learning_rate": 2.3437739537182646e-07,
+      "loss": 0.5361,
+      "step": 39603
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.215405879616215,
+      "learning_rate": 2.3421319057671975e-07,
+      "loss": 0.6712,
+      "step": 39604
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6500200610619516,
+      "learning_rate": 2.3404904264111683e-07,
+      "loss": 0.5124,
+      "step": 39605
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3062018917781906,
+      "learning_rate": 2.3388495156597136e-07,
+      "loss": 0.5552,
+      "step": 39606
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4004795282417009,
+      "learning_rate": 2.3372091735223812e-07,
+      "loss": 0.5392,
+      "step": 39607
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6295063217635637,
+      "learning_rate": 2.3355694000087525e-07,
+      "loss": 0.5127,
+      "step": 39608
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5038823011661997,
+      "learning_rate": 2.3339301951283532e-07,
+      "loss": 0.7261,
+      "step": 39609
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3085049232984587,
+      "learning_rate": 2.3322915588907425e-07,
+      "loss": 0.682,
+      "step": 39610
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3598634089196184,
+      "learning_rate": 2.3306534913054347e-07,
+      "loss": 0.739,
+      "step": 39611
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4679539233756074,
+      "learning_rate": 2.329015992382e-07,
+      "loss": 0.5857,
+      "step": 39612
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5619308270823002,
+      "learning_rate": 2.327379062129942e-07,
+      "loss": 0.4906,
+      "step": 39613
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1936654656392034,
+      "learning_rate": 2.3257427005588197e-07,
+      "loss": 0.5964,
+      "step": 39614
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4077923361731945,
+      "learning_rate": 2.3241069076781365e-07,
+      "loss": 0.7317,
+      "step": 39615
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5799517126152676,
+      "learning_rate": 2.3224716834974404e-07,
+      "loss": 0.5406,
+      "step": 39616
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3842135761179686,
+      "learning_rate": 2.3208370280262127e-07,
+      "loss": 0.7541,
+      "step": 39617
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2936373402722103,
+      "learning_rate": 2.3192029412740237e-07,
+      "loss": 0.7647,
+      "step": 39618
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.492337198391749,
+      "learning_rate": 2.3175694232503542e-07,
+      "loss": 0.6611,
+      "step": 39619
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6200936520303213,
+      "learning_rate": 2.3159364739647084e-07,
+      "loss": 0.5016,
+      "step": 39620
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.357530991999956,
+      "learning_rate": 2.3143040934266115e-07,
+      "loss": 0.6417,
+      "step": 39621
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3863760808236703,
+      "learning_rate": 2.312672281645567e-07,
+      "loss": 0.5694,
+      "step": 39622
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6213550892311056,
+      "learning_rate": 2.311041038631079e-07,
+      "loss": 0.5194,
+      "step": 39623
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.330660033748771,
+      "learning_rate": 2.3094103643926168e-07,
+      "loss": 0.5252,
+      "step": 39624
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7730134677486857,
+      "learning_rate": 2.3077802589397068e-07,
+      "loss": 0.599,
+      "step": 39625
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6503143282154,
+      "learning_rate": 2.3061507222818303e-07,
+      "loss": 0.6898,
+      "step": 39626
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3012308294792678,
+      "learning_rate": 2.3045217544284791e-07,
+      "loss": 0.5959,
+      "step": 39627
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3267253188425594,
+      "learning_rate": 2.302893355389124e-07,
+      "loss": 0.6069,
+      "step": 39628
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4130944281190674,
+      "learning_rate": 2.3012655251732574e-07,
+      "loss": 0.6914,
+      "step": 39629
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5046911165685966,
+      "learning_rate": 2.2996382637903492e-07,
+      "loss": 0.7846,
+      "step": 39630
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.561254731086427,
+      "learning_rate": 2.2980115712498918e-07,
+      "loss": 0.6563,
+      "step": 39631
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1912205111861545,
+      "learning_rate": 2.2963854475613223e-07,
+      "loss": 0.6609,
+      "step": 39632
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5181573047817116,
+      "learning_rate": 2.2947598927341442e-07,
+      "loss": 0.5714,
+      "step": 39633
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5367655382967638,
+      "learning_rate": 2.2931349067778053e-07,
+      "loss": 0.6295,
+      "step": 39634
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5133594463533828,
+      "learning_rate": 2.2915104897017648e-07,
+      "loss": 0.719,
+      "step": 39635
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4993432298169491,
+      "learning_rate": 2.2898866415154818e-07,
+      "loss": 0.6744,
+      "step": 39636
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2994932642545503,
+      "learning_rate": 2.2882633622284268e-07,
+      "loss": 0.6291,
+      "step": 39637
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4035088461256473,
+      "learning_rate": 2.2866406518500252e-07,
+      "loss": 0.6629,
+      "step": 39638
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5383139242544144,
+      "learning_rate": 2.285018510389736e-07,
+      "loss": 0.684,
+      "step": 39639
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3779248298818598,
+      "learning_rate": 2.2833969378570188e-07,
+      "loss": 0.7461,
+      "step": 39640
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5384972300193978,
+      "learning_rate": 2.2817759342612878e-07,
+      "loss": 0.5396,
+      "step": 39641
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6826055370477577,
+      "learning_rate": 2.2801554996119912e-07,
+      "loss": 0.7079,
+      "step": 39642
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5972849939187923,
+      "learning_rate": 2.2785356339185772e-07,
+      "loss": 0.7208,
+      "step": 39643
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4839277401824507,
+      "learning_rate": 2.2769163371904713e-07,
+      "loss": 0.719,
+      "step": 39644
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6603760860550605,
+      "learning_rate": 2.275297609437077e-07,
+      "loss": 0.7237,
+      "step": 39645
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2762304327419636,
+      "learning_rate": 2.2736794506678539e-07,
+      "loss": 0.5864,
+      "step": 39646
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2527506493004068,
+      "learning_rate": 2.272061860892205e-07,
+      "loss": 0.5331,
+      "step": 39647
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.48794120811254,
+      "learning_rate": 2.2704448401195567e-07,
+      "loss": 0.7203,
+      "step": 39648
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6085813134789777,
+      "learning_rate": 2.268828388359312e-07,
+      "loss": 0.6561,
+      "step": 39649
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6424906989958459,
+      "learning_rate": 2.267212505620886e-07,
+      "loss": 0.7331,
+      "step": 39650
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6055138365312375,
+      "learning_rate": 2.2655971919137044e-07,
+      "loss": 0.4993,
+      "step": 39651
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4636669244425133,
+      "learning_rate": 2.2639824472471594e-07,
+      "loss": 0.7385,
+      "step": 39652
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5497520934802718,
+      "learning_rate": 2.262368271630633e-07,
+      "loss": 0.5748,
+      "step": 39653
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2830149262658237,
+      "learning_rate": 2.2607546650735613e-07,
+      "loss": 0.8134,
+      "step": 39654
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.500386195503226,
+      "learning_rate": 2.259141627585304e-07,
+      "loss": 0.7248,
+      "step": 39655
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3679719861850015,
+      "learning_rate": 2.2575291591752756e-07,
+      "loss": 0.5652,
+      "step": 39656
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5723455396367501,
+      "learning_rate": 2.2559172598528578e-07,
+      "loss": 0.516,
+      "step": 39657
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3972613662220614,
+      "learning_rate": 2.2543059296274316e-07,
+      "loss": 0.6302,
+      "step": 39658
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2548179283430498,
+      "learning_rate": 2.2526951685083787e-07,
+      "loss": 0.6586,
+      "step": 39659
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.258261320297878,
+      "learning_rate": 2.2510849765050914e-07,
+      "loss": 0.6505,
+      "step": 39660
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.396389616134654,
+      "learning_rate": 2.2494753536269288e-07,
+      "loss": 0.6077,
+      "step": 39661
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.541373200381801,
+      "learning_rate": 2.2478662998832725e-07,
+      "loss": 0.8089,
+      "step": 39662
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3976023667922488,
+      "learning_rate": 2.2462578152834812e-07,
+      "loss": 0.7765,
+      "step": 39663
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4381275434003125,
+      "learning_rate": 2.2446498998369259e-07,
+      "loss": 0.7361,
+      "step": 39664
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2793235483009653,
+      "learning_rate": 2.2430425535529765e-07,
+      "loss": 0.6758,
+      "step": 39665
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2543432108678438,
+      "learning_rate": 2.24143577644097e-07,
+      "loss": 0.5306,
+      "step": 39666
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5413823099706605,
+      "learning_rate": 2.2398295685102655e-07,
+      "loss": 0.7723,
+      "step": 39667
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.32892036843008,
+      "learning_rate": 2.2382239297702447e-07,
+      "loss": 0.8282,
+      "step": 39668
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.303224660708668,
+      "learning_rate": 2.236618860230233e-07,
+      "loss": 0.6943,
+      "step": 39669
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9565897472830354,
+      "learning_rate": 2.2350143598995676e-07,
+      "loss": 0.6568,
+      "step": 39670
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4978594453186607,
+      "learning_rate": 2.2334104287875969e-07,
+      "loss": 0.7084,
+      "step": 39671
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.277880263844041,
+      "learning_rate": 2.2318070669036686e-07,
+      "loss": 0.5754,
+      "step": 39672
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4315362382247752,
+      "learning_rate": 2.2302042742571194e-07,
+      "loss": 0.6112,
+      "step": 39673
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5859989731362116,
+      "learning_rate": 2.2286020508572648e-07,
+      "loss": 0.7031,
+      "step": 39674
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.316989234763204,
+      "learning_rate": 2.2270003967134414e-07,
+      "loss": 0.7001,
+      "step": 39675
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6577484164591936,
+      "learning_rate": 2.225399311834986e-07,
+      "loss": 0.8738,
+      "step": 39676
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3928151901643275,
+      "learning_rate": 2.2237987962312134e-07,
+      "loss": 0.7967,
+      "step": 39677
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4139136862924182,
+      "learning_rate": 2.2221988499114278e-07,
+      "loss": 0.7001,
+      "step": 39678
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4143456265772898,
+      "learning_rate": 2.2205994728849655e-07,
+      "loss": 0.5942,
+      "step": 39679
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2658723561794196,
+      "learning_rate": 2.2190006651611197e-07,
+      "loss": 0.6303,
+      "step": 39680
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4646316045385206,
+      "learning_rate": 2.2174024267492155e-07,
+      "loss": 0.709,
+      "step": 39681
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4300621550718133,
+      "learning_rate": 2.2158047576585572e-07,
+      "loss": 0.5747,
+      "step": 39682
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2843767566327002,
+      "learning_rate": 2.2142076578984372e-07,
+      "loss": 0.699,
+      "step": 39683
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5737601791749924,
+      "learning_rate": 2.212611127478148e-07,
+      "loss": 0.5423,
+      "step": 39684
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.506776089143782,
+      "learning_rate": 2.2110151664070044e-07,
+      "loss": 0.6832,
+      "step": 39685
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3395129121079643,
+      "learning_rate": 2.209419774694299e-07,
+      "loss": 0.5977,
+      "step": 39686
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5484369884540283,
+      "learning_rate": 2.2078249523493022e-07,
+      "loss": 0.5937,
+      "step": 39687
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3703005808742073,
+      "learning_rate": 2.2062306993813064e-07,
+      "loss": 0.6569,
+      "step": 39688
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5855533209623104,
+      "learning_rate": 2.2046370157995934e-07,
+      "loss": 0.483,
+      "step": 39689
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2928010357597512,
+      "learning_rate": 2.2030439016134664e-07,
+      "loss": 0.5965,
+      "step": 39690
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4822389526477031,
+      "learning_rate": 2.2014513568321628e-07,
+      "loss": 0.6115,
+      "step": 39691
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3778213455046537,
+      "learning_rate": 2.1998593814649637e-07,
+      "loss": 0.6254,
+      "step": 39692
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.186505791364639,
+      "learning_rate": 2.1982679755211622e-07,
+      "loss": 0.5813,
+      "step": 39693
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.341932258179462,
+      "learning_rate": 2.196677139009995e-07,
+      "loss": 0.696,
+      "step": 39694
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5899586404311821,
+      "learning_rate": 2.1950868719407548e-07,
+      "loss": 0.4805,
+      "step": 39695
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4124820999050194,
+      "learning_rate": 2.1934971743226674e-07,
+      "loss": 0.7407,
+      "step": 39696
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3061361821346786,
+      "learning_rate": 2.1919080461650144e-07,
+      "loss": 0.6249,
+      "step": 39697
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.290963924262031,
+      "learning_rate": 2.190319487477033e-07,
+      "loss": 0.6211,
+      "step": 39698
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.651698619133339,
+      "learning_rate": 2.188731498267982e-07,
+      "loss": 0.6503,
+      "step": 39699
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.453753825341454,
+      "learning_rate": 2.1871440785470987e-07,
+      "loss": 0.6624,
+      "step": 39700
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3541111105784875,
+      "learning_rate": 2.1855572283236314e-07,
+      "loss": 0.602,
+      "step": 39701
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.187442294303441,
+      "learning_rate": 2.183970947606806e-07,
+      "loss": 0.7047,
+      "step": 39702
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2255826316937093,
+      "learning_rate": 2.1823852364058928e-07,
+      "loss": 0.5296,
+      "step": 39703
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4130849141080628,
+      "learning_rate": 2.1808000947300843e-07,
+      "loss": 0.691,
+      "step": 39704
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6232553929055069,
+      "learning_rate": 2.179215522588629e-07,
+      "loss": 0.477,
+      "step": 39705
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3889922363912677,
+      "learning_rate": 2.1776315199907526e-07,
+      "loss": 0.5989,
+      "step": 39706
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3941668736798383,
+      "learning_rate": 2.1760480869456812e-07,
+      "loss": 0.5997,
+      "step": 39707
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.311120768678485,
+      "learning_rate": 2.1744652234626295e-07,
+      "loss": 0.6735,
+      "step": 39708
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6187013606421428,
+      "learning_rate": 2.1728829295508013e-07,
+      "loss": 0.5086,
+      "step": 39709
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4783441674156443,
+      "learning_rate": 2.1713012052194114e-07,
+      "loss": 0.6373,
+      "step": 39710
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4590682461189644,
+      "learning_rate": 2.1697200504776973e-07,
+      "loss": 0.6506,
+      "step": 39711
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5862196069832284,
+      "learning_rate": 2.1681394653348508e-07,
+      "loss": 0.5305,
+      "step": 39712
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.185498793371013,
+      "learning_rate": 2.166559449800043e-07,
+      "loss": 0.8028,
+      "step": 39713
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2611943198609605,
+      "learning_rate": 2.164980003882522e-07,
+      "loss": 0.7691,
+      "step": 39714
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5627198208290649,
+      "learning_rate": 2.1634011275914468e-07,
+      "loss": 0.4798,
+      "step": 39715
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4389484068612486,
+      "learning_rate": 2.1618228209360438e-07,
+      "loss": 0.7766,
+      "step": 39716
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6237096594392728,
+      "learning_rate": 2.1602450839254608e-07,
+      "loss": 0.6478,
+      "step": 39717
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7039986858425138,
+      "learning_rate": 2.158667916568924e-07,
+      "loss": 0.7279,
+      "step": 39718
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4139078388321178,
+      "learning_rate": 2.1570913188755925e-07,
+      "loss": 0.6453,
+      "step": 39719
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.304567784289147,
+      "learning_rate": 2.1555152908546484e-07,
+      "loss": 0.7144,
+      "step": 39720
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.358975544758204,
+      "learning_rate": 2.1539398325152839e-07,
+      "loss": 0.6234,
+      "step": 39721
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3720660665374693,
+      "learning_rate": 2.1523649438666472e-07,
+      "loss": 0.7079,
+      "step": 39722
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4592484902642588,
+      "learning_rate": 2.15079062491792e-07,
+      "loss": 0.6403,
+      "step": 39723
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3789001418400888,
+      "learning_rate": 2.1492168756782728e-07,
+      "loss": 0.7236,
+      "step": 39724
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6106425509241694,
+      "learning_rate": 2.1476436961568648e-07,
+      "loss": 0.5281,
+      "step": 39725
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3390778740270135,
+      "learning_rate": 2.1460710863628443e-07,
+      "loss": 0.7307,
+      "step": 39726
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3250843831904766,
+      "learning_rate": 2.1444990463053817e-07,
+      "loss": 0.7366,
+      "step": 39727
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3590164608166193,
+      "learning_rate": 2.1429275759936363e-07,
+      "loss": 0.6172,
+      "step": 39728
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.716918220827052,
+      "learning_rate": 2.1413566754367454e-07,
+      "loss": 0.8461,
+      "step": 39729
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4986147999613169,
+      "learning_rate": 2.139786344643846e-07,
+      "loss": 0.6311,
+      "step": 39730
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4538333798149405,
+      "learning_rate": 2.1382165836240977e-07,
+      "loss": 0.692,
+      "step": 39731
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2773137594319228,
+      "learning_rate": 2.1366473923866372e-07,
+      "loss": 0.5644,
+      "step": 39732
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7351651316805496,
+      "learning_rate": 2.135078770940602e-07,
+      "loss": 0.6532,
+      "step": 39733
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.324330377074963,
+      "learning_rate": 2.1335107192951176e-07,
+      "loss": 0.7572,
+      "step": 39734
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4992435421676307,
+      "learning_rate": 2.1319432374593218e-07,
+      "loss": 0.7093,
+      "step": 39735
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.038737850139313,
+      "learning_rate": 2.1303763254423293e-07,
+      "loss": 0.7405,
+      "step": 39736
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.579774497531884,
+      "learning_rate": 2.128809983253277e-07,
+      "loss": 0.4601,
+      "step": 39737
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.386482749777533,
+      "learning_rate": 2.127244210901269e-07,
+      "loss": 0.8092,
+      "step": 39738
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.371113810403532,
+      "learning_rate": 2.1256790083954537e-07,
+      "loss": 0.7017,
+      "step": 39739
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5078808601041664,
+      "learning_rate": 2.1241143757449012e-07,
+      "loss": 0.6447,
+      "step": 39740
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3104834032326993,
+      "learning_rate": 2.1225503129587488e-07,
+      "loss": 0.6529,
+      "step": 39741
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4145005010095169,
+      "learning_rate": 2.1209868200461004e-07,
+      "loss": 0.7283,
+      "step": 39742
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4482692148594316,
+      "learning_rate": 2.1194238970160376e-07,
+      "loss": 0.5984,
+      "step": 39743
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3473368596010806,
+      "learning_rate": 2.117861543877686e-07,
+      "loss": 0.5648,
+      "step": 39744
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3607531289434258,
+      "learning_rate": 2.1162997606401392e-07,
+      "loss": 0.7306,
+      "step": 39745
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.31937468947967,
+      "learning_rate": 2.1147385473124892e-07,
+      "loss": 0.6941,
+      "step": 39746
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.691784906396602,
+      "learning_rate": 2.113177903903807e-07,
+      "loss": 0.7235,
+      "step": 39747
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.402685390983335,
+      "learning_rate": 2.111617830423196e-07,
+      "loss": 0.6333,
+      "step": 39748
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2653290177051264,
+      "learning_rate": 2.1100583268797493e-07,
+      "loss": 0.5079,
+      "step": 39749
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3790904222899174,
+      "learning_rate": 2.1084993932825259e-07,
+      "loss": 0.6293,
+      "step": 39750
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5843812390540004,
+      "learning_rate": 2.1069410296406079e-07,
+      "loss": 0.4972,
+      "step": 39751
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8520797756891352,
+      "learning_rate": 2.1053832359630655e-07,
+      "loss": 0.5161,
+      "step": 39752
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2023650334744405,
+      "learning_rate": 2.1038260122589915e-07,
+      "loss": 0.7873,
+      "step": 39753
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.587929401475507,
+      "learning_rate": 2.1022693585374343e-07,
+      "loss": 0.6873,
+      "step": 39754
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5430802740689198,
+      "learning_rate": 2.1007132748074533e-07,
+      "loss": 0.6636,
+      "step": 39755
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3828862239395925,
+      "learning_rate": 2.0991577610781188e-07,
+      "loss": 0.6973,
+      "step": 39756
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1952759960047148,
+      "learning_rate": 2.0976028173584683e-07,
+      "loss": 0.5667,
+      "step": 39757
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1885464066409803,
+      "learning_rate": 2.0960484436575835e-07,
+      "loss": 0.7195,
+      "step": 39758
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3571270120296228,
+      "learning_rate": 2.09449463998449e-07,
+      "loss": 0.764,
+      "step": 39759
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6832557587658312,
+      "learning_rate": 2.092941406348259e-07,
+      "loss": 0.6507,
+      "step": 39760
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.581138571329473,
+      "learning_rate": 2.0913887427579048e-07,
+      "loss": 0.6465,
+      "step": 39761
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4817405201866194,
+      "learning_rate": 2.0898366492224986e-07,
+      "loss": 0.6126,
+      "step": 39762
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3032635296226027,
+      "learning_rate": 2.0882851257510551e-07,
+      "loss": 0.6432,
+      "step": 39763
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4236244512648342,
+      "learning_rate": 2.0867341723526002e-07,
+      "loss": 0.6401,
+      "step": 39764
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3045147484692536,
+      "learning_rate": 2.0851837890361936e-07,
+      "loss": 0.6733,
+      "step": 39765
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4398696276038476,
+      "learning_rate": 2.0836339758108393e-07,
+      "loss": 0.6667,
+      "step": 39766
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3452772938620636,
+      "learning_rate": 2.082084732685574e-07,
+      "loss": 0.6318,
+      "step": 39767
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.623163373018061,
+      "learning_rate": 2.0805360596694024e-07,
+      "loss": 0.5242,
+      "step": 39768
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.293666698840493,
+      "learning_rate": 2.078987956771361e-07,
+      "loss": 0.5188,
+      "step": 39769
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3538688969223847,
+      "learning_rate": 2.0774404240004432e-07,
+      "loss": 0.7301,
+      "step": 39770
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3974915480964072,
+      "learning_rate": 2.0758934613656856e-07,
+      "loss": 0.5861,
+      "step": 39771
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.279899686203112,
+      "learning_rate": 2.0743470688760593e-07,
+      "loss": 0.593,
+      "step": 39772
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3691086626883842,
+      "learning_rate": 2.07280124654059e-07,
+      "loss": 0.7008,
+      "step": 39773
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6954440126077859,
+      "learning_rate": 2.071255994368282e-07,
+      "loss": 0.8434,
+      "step": 39774
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5322909039642334,
+      "learning_rate": 2.069711312368128e-07,
+      "loss": 0.7989,
+      "step": 39775
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4297207624044332,
+      "learning_rate": 2.0681672005491093e-07,
+      "loss": 0.7,
+      "step": 39776
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.539048814149189,
+      "learning_rate": 2.0666236589202305e-07,
+      "loss": 0.6939,
+      "step": 39777
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3556396216239435,
+      "learning_rate": 2.0650806874904727e-07,
+      "loss": 0.5636,
+      "step": 39778
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7587876473028086,
+      "learning_rate": 2.063538286268818e-07,
+      "loss": 0.6614,
+      "step": 39779
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3625389288906824,
+      "learning_rate": 2.061996455264259e-07,
+      "loss": 0.6688,
+      "step": 39780
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4024107996273771,
+      "learning_rate": 2.0604551944857665e-07,
+      "loss": 0.7599,
+      "step": 39781
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5719691817445705,
+      "learning_rate": 2.0589145039422887e-07,
+      "loss": 0.4855,
+      "step": 39782
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6141026213592227,
+      "learning_rate": 2.0573743836428294e-07,
+      "loss": 0.5481,
+      "step": 39783
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5043928319653068,
+      "learning_rate": 2.0558348335963595e-07,
+      "loss": 0.5779,
+      "step": 39784
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3699685453034463,
+      "learning_rate": 2.0542958538118162e-07,
+      "loss": 0.6805,
+      "step": 39785
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6304884590285414,
+      "learning_rate": 2.05275744429817e-07,
+      "loss": 0.5215,
+      "step": 39786
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3598825099683933,
+      "learning_rate": 2.0512196050643697e-07,
+      "loss": 0.7982,
+      "step": 39787
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7238318078948738,
+      "learning_rate": 2.0496823361194074e-07,
+      "loss": 0.7343,
+      "step": 39788
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5377363350986448,
+      "learning_rate": 2.0481456374721763e-07,
+      "loss": 0.6356,
+      "step": 39789
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4443795159987851,
+      "learning_rate": 2.0466095091316585e-07,
+      "loss": 0.7123,
+      "step": 39790
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1173108553806914,
+      "learning_rate": 2.0450739511067907e-07,
+      "loss": 0.6032,
+      "step": 39791
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.383038002332163,
+      "learning_rate": 2.0435389634065328e-07,
+      "loss": 0.5942,
+      "step": 39792
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3288237354280292,
+      "learning_rate": 2.0420045460397885e-07,
+      "loss": 0.6958,
+      "step": 39793
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4401310780616605,
+      "learning_rate": 2.0404706990155065e-07,
+      "loss": 0.5625,
+      "step": 39794
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0850947206767962,
+      "learning_rate": 2.038937422342624e-07,
+      "loss": 0.6013,
+      "step": 39795
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9266154376032378,
+      "learning_rate": 2.0374047160300557e-07,
+      "loss": 0.6773,
+      "step": 39796
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3511495589560394,
+      "learning_rate": 2.0358725800867395e-07,
+      "loss": 0.5646,
+      "step": 39797
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8662230481591833,
+      "learning_rate": 2.0343410145215793e-07,
+      "loss": 0.6531,
+      "step": 39798
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3184811835840147,
+      "learning_rate": 2.0328100193435006e-07,
+      "loss": 0.6024,
+      "step": 39799
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2933750905836172,
+      "learning_rate": 2.0312795945614193e-07,
+      "loss": 0.6225,
+      "step": 39800
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2897957354110148,
+      "learning_rate": 2.0297497401842502e-07,
+      "loss": 0.6642,
+      "step": 39801
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.20623113974743,
+      "learning_rate": 2.0282204562208974e-07,
+      "loss": 0.5939,
+      "step": 39802
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5394579435455256,
+      "learning_rate": 2.0266917426802536e-07,
+      "loss": 0.5906,
+      "step": 39803
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0469867906483503,
+      "learning_rate": 2.0251635995712226e-07,
+      "loss": 0.6656,
+      "step": 39804
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.542018885648242,
+      "learning_rate": 2.02363602690272e-07,
+      "loss": 0.6853,
+      "step": 39805
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6500066702239236,
+      "learning_rate": 2.0221090246836274e-07,
+      "loss": 0.602,
+      "step": 39806
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6083451952131487,
+      "learning_rate": 2.0205825929228262e-07,
+      "loss": 0.6083,
+      "step": 39807
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.77793664306655,
+      "learning_rate": 2.019056731629221e-07,
+      "loss": 0.7497,
+      "step": 39808
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9212185306467118,
+      "learning_rate": 2.017531440811682e-07,
+      "loss": 0.6739,
+      "step": 39809
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4010338095030732,
+      "learning_rate": 2.0160067204791023e-07,
+      "loss": 0.6168,
+      "step": 39810
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4912759812016658,
+      "learning_rate": 2.0144825706403526e-07,
+      "loss": 0.5817,
+      "step": 39811
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3155052970712857,
+      "learning_rate": 2.0129589913042924e-07,
+      "loss": 0.683,
+      "step": 39812
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3629179979160682,
+      "learning_rate": 2.0114359824798257e-07,
+      "loss": 0.6322,
+      "step": 39813
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8072939039579738,
+      "learning_rate": 2.00991354417579e-07,
+      "loss": 0.7059,
+      "step": 39814
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1239555319382661,
+      "learning_rate": 2.008391676401067e-07,
+      "loss": 0.6452,
+      "step": 39815
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.607979236576503,
+      "learning_rate": 2.0068703791645162e-07,
+      "loss": 0.6188,
+      "step": 39816
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7977348952896357,
+      "learning_rate": 2.005349652474975e-07,
+      "loss": 0.7214,
+      "step": 39817
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2782222593378318,
+      "learning_rate": 2.0038294963413251e-07,
+      "loss": 0.6108,
+      "step": 39818
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.128578883735634,
+      "learning_rate": 2.002309910772393e-07,
+      "loss": 0.6612,
+      "step": 39819
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.463840926328763,
+      "learning_rate": 2.0007908957770493e-07,
+      "loss": 0.6248,
+      "step": 39820
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.578833464196002,
+      "learning_rate": 1.9992724513641204e-07,
+      "loss": 0.4732,
+      "step": 39821
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5854352795909203,
+      "learning_rate": 1.9977545775424545e-07,
+      "loss": 0.708,
+      "step": 39822
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6082291589195026,
+      "learning_rate": 1.9962372743209e-07,
+      "loss": 0.5305,
+      "step": 39823
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5739814517496835,
+      "learning_rate": 1.9947205417082615e-07,
+      "loss": 0.483,
+      "step": 39824
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3639666378038187,
+      "learning_rate": 1.993204379713398e-07,
+      "loss": 0.6781,
+      "step": 39825
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.701625409739575,
+      "learning_rate": 1.991688788345125e-07,
+      "loss": 0.701,
+      "step": 39826
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4161201543753732,
+      "learning_rate": 1.99017376761228e-07,
+      "loss": 0.6788,
+      "step": 39827
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5875765247624175,
+      "learning_rate": 1.9886593175236557e-07,
+      "loss": 0.4874,
+      "step": 39828
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.344460088466611,
+      "learning_rate": 1.9871454380880894e-07,
+      "loss": 0.7645,
+      "step": 39829
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1649007144355816,
+      "learning_rate": 1.9856321293143965e-07,
+      "loss": 0.6043,
+      "step": 39830
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3000059235683692,
+      "learning_rate": 1.9841193912113922e-07,
+      "loss": 0.6331,
+      "step": 39831
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.267195840635932,
+      "learning_rate": 1.982607223787858e-07,
+      "loss": 0.7213,
+      "step": 39832
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3485512137544735,
+      "learning_rate": 1.9810956270526204e-07,
+      "loss": 0.7658,
+      "step": 39833
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6075045412686727,
+      "learning_rate": 1.9795846010144837e-07,
+      "loss": 0.5356,
+      "step": 39834
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2961272493004956,
+      "learning_rate": 1.9780741456822404e-07,
+      "loss": 0.6454,
+      "step": 39835
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4342241218056604,
+      "learning_rate": 1.976564261064684e-07,
+      "loss": 0.7503,
+      "step": 39836
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4011674654860573,
+      "learning_rate": 1.975054947170596e-07,
+      "loss": 0.6838,
+      "step": 39837
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4030961221762934,
+      "learning_rate": 1.9735462040087693e-07,
+      "loss": 0.8192,
+      "step": 39838
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5989138259914962,
+      "learning_rate": 1.9720380315880083e-07,
+      "loss": 0.5021,
+      "step": 39839
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.185328575413897,
+      "learning_rate": 1.9705304299170613e-07,
+      "loss": 0.5737,
+      "step": 39840
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.368844874092688,
+      "learning_rate": 1.9690233990047325e-07,
+      "loss": 0.7321,
+      "step": 39841
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.352506792838179,
+      "learning_rate": 1.9675169388597705e-07,
+      "loss": 0.5839,
+      "step": 39842
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.436667768144032,
+      "learning_rate": 1.966011049490979e-07,
+      "loss": 0.6709,
+      "step": 39843
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.79145529520033,
+      "learning_rate": 1.9645057309070958e-07,
+      "loss": 0.7497,
+      "step": 39844
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4381068604796627,
+      "learning_rate": 1.9630009831169027e-07,
+      "loss": 0.7602,
+      "step": 39845
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4028980482721238,
+      "learning_rate": 1.9614968061291484e-07,
+      "loss": 0.6798,
+      "step": 39846
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3653468716272537,
+      "learning_rate": 1.9599931999526035e-07,
+      "loss": 0.6162,
+      "step": 39847
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3005017020220766,
+      "learning_rate": 1.9584901645960275e-07,
+      "loss": 0.6577,
+      "step": 39848
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3687989936746423,
+      "learning_rate": 1.956987700068147e-07,
+      "loss": 0.6629,
+      "step": 39849
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.58257307993114,
+      "learning_rate": 1.9554858063777215e-07,
+      "loss": 0.6965,
+      "step": 39850
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4612740640402846,
+      "learning_rate": 1.953984483533511e-07,
+      "loss": 0.6362,
+      "step": 39851
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.333603229155377,
+      "learning_rate": 1.9524837315442414e-07,
+      "loss": 0.7982,
+      "step": 39852
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3026763498706446,
+      "learning_rate": 1.9509835504186504e-07,
+      "loss": 0.5845,
+      "step": 39853
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3729409723906212,
+      "learning_rate": 1.9494839401654642e-07,
+      "loss": 0.7243,
+      "step": 39854
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7144093528702855,
+      "learning_rate": 1.9479849007934425e-07,
+      "loss": 0.6864,
+      "step": 39855
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2539215815647962,
+      "learning_rate": 1.9464864323112897e-07,
+      "loss": 0.6476,
+      "step": 39856
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1967999303777572,
+      "learning_rate": 1.944988534727732e-07,
+      "loss": 0.6277,
+      "step": 39857
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.372085813607863,
+      "learning_rate": 1.9434912080514955e-07,
+      "loss": 0.71,
+      "step": 39858
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5320198696335583,
+      "learning_rate": 1.941994452291296e-07,
+      "loss": 0.6729,
+      "step": 39859
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2837759912055873,
+      "learning_rate": 1.9404982674558593e-07,
+      "loss": 0.7056,
+      "step": 39860
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0476726079649312,
+      "learning_rate": 1.9390026535538675e-07,
+      "loss": 0.6439,
+      "step": 39861
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.250783719561143,
+      "learning_rate": 1.9375076105940693e-07,
+      "loss": 0.6945,
+      "step": 39862
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.3587598927081785,
+      "learning_rate": 1.9360131385851356e-07,
+      "loss": 0.6359,
+      "step": 39863
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.162740737150749,
+      "learning_rate": 1.9345192375357812e-07,
+      "loss": 0.7002,
+      "step": 39864
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6202421829236313,
+      "learning_rate": 1.933025907454711e-07,
+      "loss": 0.7507,
+      "step": 39865
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5838271294727511,
+      "learning_rate": 1.9315331483506062e-07,
+      "loss": 0.4939,
+      "step": 39866
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6497203834380398,
+      "learning_rate": 1.930040960232149e-07,
+      "loss": 0.7309,
+      "step": 39867
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4612344066932748,
+      "learning_rate": 1.928549343108055e-07,
+      "loss": 0.615,
+      "step": 39868
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2756728963309227,
+      "learning_rate": 1.9270582969870056e-07,
+      "loss": 0.4887,
+      "step": 39869
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.411509071292231,
+      "learning_rate": 1.9255678218776608e-07,
+      "loss": 0.7997,
+      "step": 39870
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3059773556080445,
+      "learning_rate": 1.9240779177887027e-07,
+      "loss": 0.6386,
+      "step": 39871
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5490464366926882,
+      "learning_rate": 1.922588584728824e-07,
+      "loss": 0.7065,
+      "step": 39872
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.099420931171877,
+      "learning_rate": 1.9210998227066846e-07,
+      "loss": 0.7428,
+      "step": 39873
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.5845444688269374,
+      "learning_rate": 1.9196116317309554e-07,
+      "loss": 0.5509,
+      "step": 39874
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8161678118054714,
+      "learning_rate": 1.918124011810285e-07,
+      "loss": 0.5576,
+      "step": 39875
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4857923735431204,
+      "learning_rate": 1.9166369629533664e-07,
+      "loss": 0.8151,
+      "step": 39876
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5554299087597485,
+      "learning_rate": 1.9151504851688263e-07,
+      "loss": 0.6502,
+      "step": 39877
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5969478598683501,
+      "learning_rate": 1.9136645784653464e-07,
+      "loss": 0.4721,
+      "step": 39878
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4346997821676775,
+      "learning_rate": 1.912179242851564e-07,
+      "loss": 0.7347,
+      "step": 39879
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.945815908460981,
+      "learning_rate": 1.910694478336117e-07,
+      "loss": 0.6447,
+      "step": 39880
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4073609076864704,
+      "learning_rate": 1.9092102849276651e-07,
+      "loss": 0.62,
+      "step": 39881
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3007862285556864,
+      "learning_rate": 1.9077266626348457e-07,
+      "loss": 0.5539,
+      "step": 39882
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3675188661385844,
+      "learning_rate": 1.9062436114663073e-07,
+      "loss": 0.6426,
+      "step": 39883
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6830780092955744,
+      "learning_rate": 1.9047611314306546e-07,
+      "loss": 0.5881,
+      "step": 39884
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3167792782611893,
+      "learning_rate": 1.903279222536547e-07,
+      "loss": 0.6465,
+      "step": 39885
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4156837012281822,
+      "learning_rate": 1.9017978847926222e-07,
+      "loss": 0.71,
+      "step": 39886
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3006398095944542,
+      "learning_rate": 1.9003171182074732e-07,
+      "loss": 0.5886,
+      "step": 39887
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5007000449459502,
+      "learning_rate": 1.8988369227897263e-07,
+      "loss": 0.7604,
+      "step": 39888
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4136331360443972,
+      "learning_rate": 1.8973572985480193e-07,
+      "loss": 0.7081,
+      "step": 39889
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6272002070780252,
+      "learning_rate": 1.8958782454909563e-07,
+      "loss": 0.8247,
+      "step": 39890
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5898254968806249,
+      "learning_rate": 1.8943997636271528e-07,
+      "loss": 0.4842,
+      "step": 39891
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.423498028744082,
+      "learning_rate": 1.8929218529652128e-07,
+      "loss": 0.6646,
+      "step": 39892
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5540535995037699,
+      "learning_rate": 1.8914445135137405e-07,
+      "loss": 0.5514,
+      "step": 39893
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.260417382065876,
+      "learning_rate": 1.889967745281329e-07,
+      "loss": 0.5656,
+      "step": 39894
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.434675003477738,
+      "learning_rate": 1.8884915482766054e-07,
+      "loss": 0.7458,
+      "step": 39895
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3230590576141605,
+      "learning_rate": 1.8870159225081285e-07,
+      "loss": 0.6252,
+      "step": 39896
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.581548963247141,
+      "learning_rate": 1.8855408679845143e-07,
+      "loss": 0.6184,
+      "step": 39897
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4040159937715924,
+      "learning_rate": 1.884066384714345e-07,
+      "loss": 0.6813,
+      "step": 39898
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3980904007737593,
+      "learning_rate": 1.8825924727062016e-07,
+      "loss": 0.6595,
+      "step": 39899
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2953616611054997,
+      "learning_rate": 1.8811191319686783e-07,
+      "loss": 0.6408,
+      "step": 39900
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.987272018467538,
+      "learning_rate": 1.8796463625103233e-07,
+      "loss": 0.7208,
+      "step": 39901
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5869003889888613,
+      "learning_rate": 1.8781741643397412e-07,
+      "loss": 0.6642,
+      "step": 39902
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6101956731798053,
+      "learning_rate": 1.8767025374655023e-07,
+      "loss": 0.4676,
+      "step": 39903
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.324324095436949,
+      "learning_rate": 1.875231481896167e-07,
+      "loss": 0.623,
+      "step": 39904
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3031886772290417,
+      "learning_rate": 1.8737609976402836e-07,
+      "loss": 0.7831,
+      "step": 39905
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4106041938540295,
+      "learning_rate": 1.8722910847064346e-07,
+      "loss": 0.5397,
+      "step": 39906
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3315062463618472,
+      "learning_rate": 1.8708217431031793e-07,
+      "loss": 0.7438,
+      "step": 39907
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.419199552989993,
+      "learning_rate": 1.869352972839067e-07,
+      "loss": 0.7344,
+      "step": 39908
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.313549171779809,
+      "learning_rate": 1.8678847739226459e-07,
+      "loss": 0.7022,
+      "step": 39909
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.498692591580745,
+      "learning_rate": 1.8664171463624648e-07,
+      "loss": 0.5643,
+      "step": 39910
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.556007330441328,
+      "learning_rate": 1.8649500901670837e-07,
+      "loss": 0.6281,
+      "step": 39911
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4596945393101253,
+      "learning_rate": 1.863483605345029e-07,
+      "loss": 0.7854,
+      "step": 39912
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.063642226826058,
+      "learning_rate": 1.8620176919048273e-07,
+      "loss": 0.6627,
+      "step": 39913
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.593884333210562,
+      "learning_rate": 1.860552349855038e-07,
+      "loss": 0.7731,
+      "step": 39914
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7372518876535956,
+      "learning_rate": 1.8590875792041884e-07,
+      "loss": 0.662,
+      "step": 39915
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4505286940435256,
+      "learning_rate": 1.8576233799608046e-07,
+      "loss": 0.6411,
+      "step": 39916
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.459493565854295,
+      "learning_rate": 1.8561597521334018e-07,
+      "loss": 0.6247,
+      "step": 39917
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3232343691070547,
+      "learning_rate": 1.854696695730507e-07,
+      "loss": 0.604,
+      "step": 39918
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5863837296891603,
+      "learning_rate": 1.853234210760635e-07,
+      "loss": 0.6914,
+      "step": 39919
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5214576071430062,
+      "learning_rate": 1.8517722972323239e-07,
+      "loss": 0.7832,
+      "step": 39920
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4592660048893638,
+      "learning_rate": 1.8503109551540556e-07,
+      "loss": 0.5622,
+      "step": 39921
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.8564718126734636,
+      "learning_rate": 1.8488501845343565e-07,
+      "loss": 0.6409,
+      "step": 39922
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.417764839609434,
+      "learning_rate": 1.847389985381709e-07,
+      "loss": 0.5806,
+      "step": 39923
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4133290399700622,
+      "learning_rate": 1.8459303577046505e-07,
+      "loss": 0.6089,
+      "step": 39924
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7013284871300924,
+      "learning_rate": 1.8444713015116632e-07,
+      "loss": 0.4496,
+      "step": 39925
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3567309356127695,
+      "learning_rate": 1.8430128168112293e-07,
+      "loss": 0.621,
+      "step": 39926
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4543039668075435,
+      "learning_rate": 1.841554903611842e-07,
+      "loss": 0.6113,
+      "step": 39927
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5847171635844768,
+      "learning_rate": 1.8400975619220163e-07,
+      "loss": 0.7284,
+      "step": 39928
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5807499598156307,
+      "learning_rate": 1.8386407917502126e-07,
+      "loss": 0.6481,
+      "step": 39929
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3544910260737764,
+      "learning_rate": 1.8371845931049126e-07,
+      "loss": 0.7806,
+      "step": 39930
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6134504277230826,
+      "learning_rate": 1.8357289659946098e-07,
+      "loss": 0.784,
+      "step": 39931
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3121404838464417,
+      "learning_rate": 1.8342739104277753e-07,
+      "loss": 0.7154,
+      "step": 39932
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6087510188767341,
+      "learning_rate": 1.8328194264128686e-07,
+      "loss": 0.5123,
+      "step": 39933
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3449406965094244,
+      "learning_rate": 1.831365513958361e-07,
+      "loss": 0.6753,
+      "step": 39934
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.420017153107533,
+      "learning_rate": 1.8299121730727342e-07,
+      "loss": 0.5229,
+      "step": 39935
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3033204083886225,
+      "learning_rate": 1.8284594037644377e-07,
+      "loss": 0.6401,
+      "step": 39936
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6001555167992902,
+      "learning_rate": 1.8270072060419308e-07,
+      "loss": 0.6875,
+      "step": 39937
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.388502991411201,
+      "learning_rate": 1.8255555799136627e-07,
+      "loss": 0.5633,
+      "step": 39938
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3685505326888538,
+      "learning_rate": 1.824104525388104e-07,
+      "loss": 0.6914,
+      "step": 39939
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7316223340951002,
+      "learning_rate": 1.8226540424736706e-07,
+      "loss": 0.5734,
+      "step": 39940
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4049727931248486,
+      "learning_rate": 1.8212041311788442e-07,
+      "loss": 0.6436,
+      "step": 39941
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6212479145642619,
+      "learning_rate": 1.8197547915120407e-07,
+      "loss": 0.4988,
+      "step": 39942
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5799882773229018,
+      "learning_rate": 1.8183060234817196e-07,
+      "loss": 0.5259,
+      "step": 39943
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3662105197561605,
+      "learning_rate": 1.8168578270962968e-07,
+      "loss": 0.6871,
+      "step": 39944
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.366820406634533,
+      "learning_rate": 1.8154102023642205e-07,
+      "loss": 0.6936,
+      "step": 39945
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2303539836269153,
+      "learning_rate": 1.813963149293907e-07,
+      "loss": 0.5965,
+      "step": 39946
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.505211539039973,
+      "learning_rate": 1.812516667893771e-07,
+      "loss": 0.6963,
+      "step": 39947
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.423540577047211,
+      "learning_rate": 1.811070758172262e-07,
+      "loss": 0.7839,
+      "step": 39948
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.258676332630184,
+      "learning_rate": 1.8096254201377838e-07,
+      "loss": 0.6258,
+      "step": 39949
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4603742956429906,
+      "learning_rate": 1.8081806537987524e-07,
+      "loss": 0.7591,
+      "step": 39950
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5197040132437019,
+      "learning_rate": 1.8067364591635828e-07,
+      "loss": 0.6903,
+      "step": 39951
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3581042047349212,
+      "learning_rate": 1.8052928362406686e-07,
+      "loss": 0.5614,
+      "step": 39952
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.465553095968087,
+      "learning_rate": 1.803849785038436e-07,
+      "loss": 0.5901,
+      "step": 39953
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2797279373886552,
+      "learning_rate": 1.8024073055652903e-07,
+      "loss": 0.5872,
+      "step": 39954
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5900145046043939,
+      "learning_rate": 1.8009653978296017e-07,
+      "loss": 0.6939,
+      "step": 39955
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2378973998294291,
+      "learning_rate": 1.7995240618397859e-07,
+      "loss": 0.6179,
+      "step": 39956
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.241654166308052,
+      "learning_rate": 1.798083297604236e-07,
+      "loss": 0.6426,
+      "step": 39957
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4927706689747624,
+      "learning_rate": 1.7966431051313349e-07,
+      "loss": 0.618,
+      "step": 39958
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4748912236105245,
+      "learning_rate": 1.795203484429453e-07,
+      "loss": 0.7386,
+      "step": 39959
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7618720192870179,
+      "learning_rate": 1.793764435507006e-07,
+      "loss": 0.6333,
+      "step": 39960
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.563843666367026,
+      "learning_rate": 1.792325958372343e-07,
+      "loss": 0.7557,
+      "step": 39961
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8578572971591814,
+      "learning_rate": 1.7908880530338457e-07,
+      "loss": 0.6591,
+      "step": 39962
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8811948248779196,
+      "learning_rate": 1.7894507194999077e-07,
+      "loss": 0.6187,
+      "step": 39963
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4471922870066554,
+      "learning_rate": 1.7880139577788779e-07,
+      "loss": 0.6823,
+      "step": 39964
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6030107506624531,
+      "learning_rate": 1.7865777678791163e-07,
+      "loss": 0.5158,
+      "step": 39965
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3140035132392203,
+      "learning_rate": 1.7851421498089826e-07,
+      "loss": 0.6141,
+      "step": 39966
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4084084335469884,
+      "learning_rate": 1.7837071035768704e-07,
+      "loss": 0.847,
+      "step": 39967
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4397494292358848,
+      "learning_rate": 1.7822726291910952e-07,
+      "loss": 0.5726,
+      "step": 39968
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8242825461320828,
+      "learning_rate": 1.780838726660028e-07,
+      "loss": 0.7289,
+      "step": 39969
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7038904346796038,
+      "learning_rate": 1.7794053959920065e-07,
+      "loss": 0.704,
+      "step": 39970
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2499224606484314,
+      "learning_rate": 1.777972637195402e-07,
+      "loss": 0.6081,
+      "step": 39971
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5077133558792752,
+      "learning_rate": 1.776540450278519e-07,
+      "loss": 0.7212,
+      "step": 39972
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2826623496797513,
+      "learning_rate": 1.775108835249717e-07,
+      "loss": 0.6265,
+      "step": 39973
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4341522244123972,
+      "learning_rate": 1.7736777921173343e-07,
+      "loss": 0.6895,
+      "step": 39974
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1324417751002875,
+      "learning_rate": 1.7722473208896862e-07,
+      "loss": 0.7318,
+      "step": 39975
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6195711301339824,
+      "learning_rate": 1.7708174215751217e-07,
+      "loss": 0.5354,
+      "step": 39976
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.531903635237014,
+      "learning_rate": 1.7693880941819564e-07,
+      "loss": 0.6667,
+      "step": 39977
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5904869570218338,
+      "learning_rate": 1.7679593387185169e-07,
+      "loss": 0.5174,
+      "step": 39978
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.258985058664905,
+      "learning_rate": 1.7665311551931075e-07,
+      "loss": 0.6107,
+      "step": 39979
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5668824018435832,
+      "learning_rate": 1.7651035436140662e-07,
+      "loss": 0.8364,
+      "step": 39980
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1841871510992175,
+      "learning_rate": 1.7636765039896865e-07,
+      "loss": 0.6014,
+      "step": 39981
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.390250528814958,
+      "learning_rate": 1.7622500363282836e-07,
+      "loss": 0.6954,
+      "step": 39982
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.634805138921564,
+      "learning_rate": 1.7608241406381509e-07,
+      "loss": 0.5071,
+      "step": 39983
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2620134804744418,
+      "learning_rate": 1.7593988169276156e-07,
+      "loss": 0.6838,
+      "step": 39984
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2149719971058357,
+      "learning_rate": 1.7579740652049705e-07,
+      "loss": 0.6917,
+      "step": 39985
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5994624677855145,
+      "learning_rate": 1.7565498854784868e-07,
+      "loss": 0.7162,
+      "step": 39986
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5847737764439958,
+      "learning_rate": 1.7551262777564693e-07,
+      "loss": 0.4794,
+      "step": 39987
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3972468589452034,
+      "learning_rate": 1.7537032420472334e-07,
+      "loss": 0.6398,
+      "step": 39988
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4735333831817952,
+      "learning_rate": 1.7522807783590278e-07,
+      "loss": 0.7855,
+      "step": 39989
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.153116677672413,
+      "learning_rate": 1.7508588867001463e-07,
+      "loss": 0.678,
+      "step": 39990
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3358656547772354,
+      "learning_rate": 1.749437567078871e-07,
+      "loss": 0.5947,
+      "step": 39991
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7319672790873977,
+      "learning_rate": 1.748016819503473e-07,
+      "loss": 0.7721,
+      "step": 39992
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.383775057219967,
+      "learning_rate": 1.7465966439822346e-07,
+      "loss": 0.573,
+      "step": 39993
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3747465442414877,
+      "learning_rate": 1.745177040523416e-07,
+      "loss": 0.7243,
+      "step": 39994
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6337884050317777,
+      "learning_rate": 1.7437580091352878e-07,
+      "loss": 0.659,
+      "step": 39995
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.536682598366442,
+      "learning_rate": 1.7423395498260887e-07,
+      "loss": 0.7264,
+      "step": 39996
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2758548924463307,
+      "learning_rate": 1.7409216626041225e-07,
+      "loss": 0.7872,
+      "step": 39997
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4771147947981627,
+      "learning_rate": 1.7395043474775942e-07,
+      "loss": 0.6955,
+      "step": 39998
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5528518508048197,
+      "learning_rate": 1.7380876044547966e-07,
+      "loss": 0.7122,
+      "step": 39999
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6027871901531701,
+      "learning_rate": 1.736671433543957e-07,
+      "loss": 0.5297,
+      "step": 40000
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.228819349912214,
+      "learning_rate": 1.7352558347533243e-07,
+      "loss": 0.5545,
+      "step": 40001
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5731968207823073,
+      "learning_rate": 1.7338408080911473e-07,
+      "loss": 0.4176,
+      "step": 40002
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7553620692851455,
+      "learning_rate": 1.7324263535656528e-07,
+      "loss": 0.6942,
+      "step": 40003
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6084271328241505,
+      "learning_rate": 1.7310124711850894e-07,
+      "loss": 0.5278,
+      "step": 40004
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2331458815932481,
+      "learning_rate": 1.7295991609576845e-07,
+      "loss": 0.6573,
+      "step": 40005
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2376658512382788,
+      "learning_rate": 1.7281864228916644e-07,
+      "loss": 0.6913,
+      "step": 40006
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4635613248694523,
+      "learning_rate": 1.726774256995245e-07,
+      "loss": 0.8169,
+      "step": 40007
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6355081805913466,
+      "learning_rate": 1.725362663276653e-07,
+      "loss": 0.5563,
+      "step": 40008
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4225165281788754,
+      "learning_rate": 1.723951641744137e-07,
+      "loss": 0.6731,
+      "step": 40009
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.401526176259433,
+      "learning_rate": 1.72254119240588e-07,
+      "loss": 0.6108,
+      "step": 40010
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7165661551590141,
+      "learning_rate": 1.721131315270097e-07,
+      "loss": 0.6918,
+      "step": 40011
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2983373013930948,
+      "learning_rate": 1.719722010344993e-07,
+      "loss": 0.652,
+      "step": 40012
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3402465804581951,
+      "learning_rate": 1.7183132776387945e-07,
+      "loss": 0.7288,
+      "step": 40013
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8741318452895264,
+      "learning_rate": 1.716905117159695e-07,
+      "loss": 0.6829,
+      "step": 40014
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.741323025675153,
+      "learning_rate": 1.715497528915888e-07,
+      "loss": 0.6683,
+      "step": 40015
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2915638095963333,
+      "learning_rate": 1.714090512915567e-07,
+      "loss": 0.7667,
+      "step": 40016
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8426659239070486,
+      "learning_rate": 1.7126840691669255e-07,
+      "loss": 0.5526,
+      "step": 40017
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6895257426032633,
+      "learning_rate": 1.7112781976781567e-07,
+      "loss": 0.8078,
+      "step": 40018
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.297205243009576,
+      "learning_rate": 1.709872898457443e-07,
+      "loss": 0.6852,
+      "step": 40019
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6409068803006582,
+      "learning_rate": 1.708468171512978e-07,
+      "loss": 0.5095,
+      "step": 40020
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.562092666188288,
+      "learning_rate": 1.7070640168529107e-07,
+      "loss": 0.6938,
+      "step": 40021
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.308794103637847,
+      "learning_rate": 1.7056604344854565e-07,
+      "loss": 0.6164,
+      "step": 40022
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3202600068938544,
+      "learning_rate": 1.7042574244187537e-07,
+      "loss": 0.6198,
+      "step": 40023
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3813340641539968,
+      "learning_rate": 1.7028549866609845e-07,
+      "loss": 0.6576,
+      "step": 40024
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3480328427064128,
+      "learning_rate": 1.7014531212203088e-07,
+      "loss": 0.7407,
+      "step": 40025
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3454699782126736,
+      "learning_rate": 1.7000518281049093e-07,
+      "loss": 0.6116,
+      "step": 40026
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.345803351062815,
+      "learning_rate": 1.698651107322924e-07,
+      "loss": 0.6854,
+      "step": 40027
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2884647993158347,
+      "learning_rate": 1.6972509588825015e-07,
+      "loss": 0.655,
+      "step": 40028
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4498153481357456,
+      "learning_rate": 1.6958513827918134e-07,
+      "loss": 0.6918,
+      "step": 40029
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4856611855387323,
+      "learning_rate": 1.6944523790590084e-07,
+      "loss": 0.6954,
+      "step": 40030
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4375152965780866,
+      "learning_rate": 1.6930539476922248e-07,
+      "loss": 0.672,
+      "step": 40031
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3217235373526353,
+      "learning_rate": 1.6916560886995893e-07,
+      "loss": 0.6107,
+      "step": 40032
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6125253703617247,
+      "learning_rate": 1.690258802089262e-07,
+      "loss": 0.4767,
+      "step": 40033
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5700920700626931,
+      "learning_rate": 1.6888620878693806e-07,
+      "loss": 0.4617,
+      "step": 40034
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4413905629493906,
+      "learning_rate": 1.6874659460480726e-07,
+      "loss": 0.5999,
+      "step": 40035
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5010542298217868,
+      "learning_rate": 1.686070376633453e-07,
+      "loss": 0.7642,
+      "step": 40036
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4024586223885664,
+      "learning_rate": 1.6846753796336491e-07,
+      "loss": 0.652,
+      "step": 40037
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3656837653170872,
+      "learning_rate": 1.6832809550568097e-07,
+      "loss": 0.7397,
+      "step": 40038
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1823982562203066,
+      "learning_rate": 1.6818871029110284e-07,
+      "loss": 0.6624,
+      "step": 40039
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5478471653632153,
+      "learning_rate": 1.680493823204421e-07,
+      "loss": 0.5243,
+      "step": 40040
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.321252297960363,
+      "learning_rate": 1.6791011159451141e-07,
+      "loss": 0.6503,
+      "step": 40041
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4685490200148286,
+      "learning_rate": 1.6777089811412017e-07,
+      "loss": 0.7026,
+      "step": 40042
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6379057866381608,
+      "learning_rate": 1.676317418800799e-07,
+      "loss": 0.7348,
+      "step": 40043
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5171244724735862,
+      "learning_rate": 1.6749264289320111e-07,
+      "loss": 0.6119,
+      "step": 40044
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6933910668007135,
+      "learning_rate": 1.67353601154292e-07,
+      "loss": 0.6197,
+      "step": 40045
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6332953467720466,
+      "learning_rate": 1.6721461666416305e-07,
+      "loss": 0.5401,
+      "step": 40046
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.924325654717977,
+      "learning_rate": 1.6707568942362473e-07,
+      "loss": 0.6663,
+      "step": 40047
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.437271599693931,
+      "learning_rate": 1.6693681943348527e-07,
+      "loss": 0.648,
+      "step": 40048
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4527206448946062,
+      "learning_rate": 1.6679800669455072e-07,
+      "loss": 0.5957,
+      "step": 40049
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.582424975155042,
+      "learning_rate": 1.666592512076326e-07,
+      "loss": 0.6504,
+      "step": 40050
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4308493030862328,
+      "learning_rate": 1.6652055297353698e-07,
+      "loss": 0.7566,
+      "step": 40051
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3663804290584198,
+      "learning_rate": 1.663819119930732e-07,
+      "loss": 0.6247,
+      "step": 40052
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1207832162148148,
+      "learning_rate": 1.6624332826704616e-07,
+      "loss": 0.4935,
+      "step": 40053
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4943498975137457,
+      "learning_rate": 1.66104801796263e-07,
+      "loss": 0.6134,
+      "step": 40054
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3115435020152626,
+      "learning_rate": 1.6596633258153193e-07,
+      "loss": 0.677,
+      "step": 40055
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.881532041979757,
+      "learning_rate": 1.6582792062365905e-07,
+      "loss": 0.5949,
+      "step": 40056
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.542683846862133,
+      "learning_rate": 1.6568956592344808e-07,
+      "loss": 0.5501,
+      "step": 40057
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4720587514836883,
+      "learning_rate": 1.6555126848170622e-07,
+      "loss": 0.6207,
+      "step": 40058
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.926008364428942,
+      "learning_rate": 1.6541302829923945e-07,
+      "loss": 0.6927,
+      "step": 40059
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.144036485808635,
+      "learning_rate": 1.6527484537684935e-07,
+      "loss": 0.7053,
+      "step": 40060
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.297807005783164,
+      "learning_rate": 1.6513671971534528e-07,
+      "loss": 0.816,
+      "step": 40061
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5372431164379268,
+      "learning_rate": 1.6499865131552773e-07,
+      "loss": 0.6353,
+      "step": 40062
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7506244990486908,
+      "learning_rate": 1.6486064017820157e-07,
+      "loss": 0.6668,
+      "step": 40063
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6870779178199886,
+      "learning_rate": 1.6472268630417065e-07,
+      "loss": 0.7743,
+      "step": 40064
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5450764841894808,
+      "learning_rate": 1.6458478969423876e-07,
+      "loss": 0.6995,
+      "step": 40065
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3486160899916864,
+      "learning_rate": 1.6444695034920743e-07,
+      "loss": 0.7364,
+      "step": 40066
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3748964019758159,
+      "learning_rate": 1.643091682698794e-07,
+      "loss": 0.5716,
+      "step": 40067
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3587202579256246,
+      "learning_rate": 1.6417144345705738e-07,
+      "loss": 0.7289,
+      "step": 40068
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9780570497991838,
+      "learning_rate": 1.6403377591154512e-07,
+      "loss": 0.7457,
+      "step": 40069
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.831376383115268,
+      "learning_rate": 1.6389616563413978e-07,
+      "loss": 0.6165,
+      "step": 40070
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8915165583661395,
+      "learning_rate": 1.637586126256463e-07,
+      "loss": 0.6391,
+      "step": 40071
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.429204826529085,
+      "learning_rate": 1.6362111688686288e-07,
+      "loss": 0.4515,
+      "step": 40072
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3221915431334923,
+      "learning_rate": 1.6348367841859337e-07,
+      "loss": 0.6366,
+      "step": 40073
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6242763943922436,
+      "learning_rate": 1.6334629722163597e-07,
+      "loss": 0.6638,
+      "step": 40074
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4670912907308935,
+      "learning_rate": 1.6320897329678899e-07,
+      "loss": 0.6696,
+      "step": 40075
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.619390607651438,
+      "learning_rate": 1.6307170664485506e-07,
+      "loss": 0.4464,
+      "step": 40076
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3496413987573443,
+      "learning_rate": 1.6293449726663136e-07,
+      "loss": 0.6925,
+      "step": 40077
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3595489490085098,
+      "learning_rate": 1.6279734516291724e-07,
+      "loss": 0.6476,
+      "step": 40078
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.648519322858328,
+      "learning_rate": 1.6266025033451206e-07,
+      "loss": 0.6183,
+      "step": 40079
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4323189653722064,
+      "learning_rate": 1.6252321278221294e-07,
+      "loss": 0.8281,
+      "step": 40080
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5639186293604623,
+      "learning_rate": 1.6238623250681817e-07,
+      "loss": 0.6181,
+      "step": 40081
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4575008565658427,
+      "learning_rate": 1.6224930950912487e-07,
+      "loss": 0.8349,
+      "step": 40082
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6577953425119758,
+      "learning_rate": 1.621124437899324e-07,
+      "loss": 0.6216,
+      "step": 40083
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.417236822961545,
+      "learning_rate": 1.6197563535003347e-07,
+      "loss": 0.6605,
+      "step": 40084
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2798391695851898,
+      "learning_rate": 1.6183888419022854e-07,
+      "loss": 0.6103,
+      "step": 40085
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3250773443424753,
+      "learning_rate": 1.617021903113114e-07,
+      "loss": 0.5529,
+      "step": 40086
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3345732404334312,
+      "learning_rate": 1.6156555371408033e-07,
+      "loss": 0.5956,
+      "step": 40087
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8363710787836343,
+      "learning_rate": 1.6142897439932804e-07,
+      "loss": 0.6704,
+      "step": 40088
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2962219499057217,
+      "learning_rate": 1.6129245236785162e-07,
+      "loss": 0.6086,
+      "step": 40089
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.608890720550223,
+      "learning_rate": 1.6115598762044493e-07,
+      "loss": 0.6493,
+      "step": 40090
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5388760058971998,
+      "learning_rate": 1.6101958015790397e-07,
+      "loss": 0.6183,
+      "step": 40091
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5829919537149788,
+      "learning_rate": 1.6088322998102147e-07,
+      "loss": 0.7064,
+      "step": 40092
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4700008159918745,
+      "learning_rate": 1.607469370905912e-07,
+      "loss": 0.6975,
+      "step": 40093
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.297627189248817,
+      "learning_rate": 1.606107014874081e-07,
+      "loss": 0.6753,
+      "step": 40094
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3399853859793802,
+      "learning_rate": 1.6047452317226485e-07,
+      "loss": 0.7688,
+      "step": 40095
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3421105802061915,
+      "learning_rate": 1.6033840214595309e-07,
+      "loss": 0.7206,
+      "step": 40096
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8986439938140445,
+      "learning_rate": 1.602023384092677e-07,
+      "loss": 0.6765,
+      "step": 40097
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4467671120022236,
+      "learning_rate": 1.6006633196299804e-07,
+      "loss": 0.6447,
+      "step": 40098
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3591729455555712,
+      "learning_rate": 1.5993038280793905e-07,
+      "loss": 0.7184,
+      "step": 40099
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3731915264331382,
+      "learning_rate": 1.5979449094487898e-07,
+      "loss": 0.6759,
+      "step": 40100
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2747381588725983,
+      "learning_rate": 1.5965865637461163e-07,
+      "loss": 0.7732,
+      "step": 40101
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6242129092974945,
+      "learning_rate": 1.595228790979264e-07,
+      "loss": 0.4847,
+      "step": 40102
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5185153685283093,
+      "learning_rate": 1.5938715911561596e-07,
+      "loss": 0.7121,
+      "step": 40103
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2502793300090438,
+      "learning_rate": 1.592514964284686e-07,
+      "loss": 0.5495,
+      "step": 40104
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4173499302713364,
+      "learning_rate": 1.5911589103727253e-07,
+      "loss": 0.7192,
+      "step": 40105
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3605426675779009,
+      "learning_rate": 1.589803429428205e-07,
+      "loss": 0.6749,
+      "step": 40106
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6096076964973967,
+      "learning_rate": 1.5884485214590072e-07,
+      "loss": 0.4762,
+      "step": 40107
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3290367714383193,
+      "learning_rate": 1.587094186473026e-07,
+      "loss": 0.6927,
+      "step": 40108
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4270690575155682,
+      "learning_rate": 1.5857404244781326e-07,
+      "loss": 0.763,
+      "step": 40109
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3321199540086397,
+      "learning_rate": 1.5843872354822099e-07,
+      "loss": 0.7129,
+      "step": 40110
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3119844475226148,
+      "learning_rate": 1.5830346194931513e-07,
+      "loss": 0.6837,
+      "step": 40111
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2629874669856755,
+      "learning_rate": 1.5816825765188171e-07,
+      "loss": 0.7415,
+      "step": 40112
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6162936926735875,
+      "learning_rate": 1.58033110656709e-07,
+      "loss": 0.5231,
+      "step": 40113
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.221246458287539,
+      "learning_rate": 1.5789802096458306e-07,
+      "loss": 0.6118,
+      "step": 40114
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.304400844004287,
+      "learning_rate": 1.5776298857629103e-07,
+      "loss": 0.6623,
+      "step": 40115
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7156422711093766,
+      "learning_rate": 1.576280134926189e-07,
+      "loss": 0.661,
+      "step": 40116
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3486522293339631,
+      "learning_rate": 1.5749309571435277e-07,
+      "loss": 0.7278,
+      "step": 40117
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.187588194823122,
+      "learning_rate": 1.5735823524227755e-07,
+      "loss": 0.6029,
+      "step": 40118
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3174007880711756,
+      "learning_rate": 1.5722343207717816e-07,
+      "loss": 0.6455,
+      "step": 40119
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5981533860504856,
+      "learning_rate": 1.5708868621984175e-07,
+      "loss": 0.5129,
+      "step": 40120
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3931980204591647,
+      "learning_rate": 1.569539976710499e-07,
+      "loss": 0.7832,
+      "step": 40121
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5697969368274063,
+      "learning_rate": 1.5681936643158867e-07,
+      "loss": 0.6623,
+      "step": 40122
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4076932325897447,
+      "learning_rate": 1.566847925022408e-07,
+      "loss": 0.7318,
+      "step": 40123
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3159900756995793,
+      "learning_rate": 1.5655027588379113e-07,
+      "loss": 0.5897,
+      "step": 40124
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4416920947810565,
+      "learning_rate": 1.5641581657702244e-07,
+      "loss": 0.6739,
+      "step": 40125
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4037269825329026,
+      "learning_rate": 1.5628141458271628e-07,
+      "loss": 0.7079,
+      "step": 40126
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.571743394612663,
+      "learning_rate": 1.561470699016565e-07,
+      "loss": 0.7537,
+      "step": 40127
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5547675672219174,
+      "learning_rate": 1.5601278253462582e-07,
+      "loss": 0.7008,
+      "step": 40128
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.757272515208064,
+      "learning_rate": 1.558785524824047e-07,
+      "loss": 0.7193,
+      "step": 40129
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2120907506994256,
+      "learning_rate": 1.5574437974577473e-07,
+      "loss": 0.5837,
+      "step": 40130
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.426670821948255,
+      "learning_rate": 1.5561026432551752e-07,
+      "loss": 0.773,
+      "step": 40131
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6296580238352532,
+      "learning_rate": 1.5547620622241467e-07,
+      "loss": 0.4863,
+      "step": 40132
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.351872800732231,
+      "learning_rate": 1.5534220543724666e-07,
+      "loss": 0.5946,
+      "step": 40133
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5895067646097171,
+      "learning_rate": 1.5520826197079176e-07,
+      "loss": 0.4788,
+      "step": 40134
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2984781540272625,
+      "learning_rate": 1.5507437582383156e-07,
+      "loss": 0.6821,
+      "step": 40135
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.200247489623327,
+      "learning_rate": 1.5494054699714545e-07,
+      "loss": 0.7867,
+      "step": 40136
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5984464029248157,
+      "learning_rate": 1.548067754915128e-07,
+      "loss": 0.4755,
+      "step": 40137
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3680265810530656,
+      "learning_rate": 1.5467306130771077e-07,
+      "loss": 0.6296,
+      "step": 40138
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6260997758109197,
+      "learning_rate": 1.5453940444652092e-07,
+      "loss": 0.7022,
+      "step": 40139
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.709824744449664,
+      "learning_rate": 1.5440580490871827e-07,
+      "loss": 0.7114,
+      "step": 40140
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3583694441587262,
+      "learning_rate": 1.5427226269508322e-07,
+      "loss": 0.6959,
+      "step": 40141
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4083591826002142,
+      "learning_rate": 1.5413877780639076e-07,
+      "loss": 0.6043,
+      "step": 40142
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8465920396306237,
+      "learning_rate": 1.5400535024342023e-07,
+      "loss": 0.6771,
+      "step": 40143
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3572660014909899,
+      "learning_rate": 1.538719800069477e-07,
+      "loss": 0.6072,
+      "step": 40144
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4731636574009028,
+      "learning_rate": 1.537386670977503e-07,
+      "loss": 0.7686,
+      "step": 40145
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4555519372055519,
+      "learning_rate": 1.5360541151660303e-07,
+      "loss": 0.7666,
+      "step": 40146
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9391997295442989,
+      "learning_rate": 1.5347221326428187e-07,
+      "loss": 0.7422,
+      "step": 40147
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.39797422868103,
+      "learning_rate": 1.5333907234156288e-07,
+      "loss": 0.6806,
+      "step": 40148
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2566835749719127,
+      "learning_rate": 1.532059887492221e-07,
+      "loss": 0.5653,
+      "step": 40149
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3829980213658162,
+      "learning_rate": 1.530729624880345e-07,
+      "loss": 0.7201,
+      "step": 40150
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5077816587493287,
+      "learning_rate": 1.5293999355877164e-07,
+      "loss": 0.8102,
+      "step": 40151
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6265378592592925,
+      "learning_rate": 1.528070819622096e-07,
+      "loss": 0.5411,
+      "step": 40152
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5812862226080364,
+      "learning_rate": 1.526742276991222e-07,
+      "loss": 0.7428,
+      "step": 40153
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.569165787785882,
+      "learning_rate": 1.525414307702855e-07,
+      "loss": 0.5394,
+      "step": 40154
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4507077399804722,
+      "learning_rate": 1.5240869117646774e-07,
+      "loss": 0.5443,
+      "step": 40155
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.291713852405914,
+      "learning_rate": 1.5227600891844495e-07,
+      "loss": 0.6852,
+      "step": 40156
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6240437776995844,
+      "learning_rate": 1.521433839969888e-07,
+      "loss": 0.7085,
+      "step": 40157
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2853564970717586,
+      "learning_rate": 1.5201081641287196e-07,
+      "loss": 0.759,
+      "step": 40158
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6024851374940389,
+      "learning_rate": 1.5187830616686605e-07,
+      "loss": 0.4617,
+      "step": 40159
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5967550761014478,
+      "learning_rate": 1.5174585325974266e-07,
+      "loss": 0.5249,
+      "step": 40160
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4541480375983997,
+      "learning_rate": 1.516134576922723e-07,
+      "loss": 0.6712,
+      "step": 40161
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5854290639611595,
+      "learning_rate": 1.5148111946522659e-07,
+      "loss": 0.4631,
+      "step": 40162
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3835441377797253,
+      "learning_rate": 1.51348838579376e-07,
+      "loss": 0.6846,
+      "step": 40163
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5032667077920834,
+      "learning_rate": 1.51216615035491e-07,
+      "loss": 0.7262,
+      "step": 40164
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4297654708868919,
+      "learning_rate": 1.5108444883433994e-07,
+      "loss": 0.7783,
+      "step": 40165
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3738406277321142,
+      "learning_rate": 1.5095233997669323e-07,
+      "loss": 0.6439,
+      "step": 40166
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.617435333625608,
+      "learning_rate": 1.508202884633203e-07,
+      "loss": 0.649,
+      "step": 40167
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3475764925435516,
+      "learning_rate": 1.5068829429499054e-07,
+      "loss": 0.6149,
+      "step": 40168
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.334967124454331,
+      "learning_rate": 1.505563574724711e-07,
+      "loss": 0.5668,
+      "step": 40169
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.719809194746225,
+      "learning_rate": 1.5042447799653136e-07,
+      "loss": 0.5726,
+      "step": 40170
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2965598223058505,
+      "learning_rate": 1.5029265586793851e-07,
+      "loss": 0.6187,
+      "step": 40171
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7258198826735993,
+      "learning_rate": 1.501608910874608e-07,
+      "loss": 0.6623,
+      "step": 40172
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.433589517860961,
+      "learning_rate": 1.5002918365586317e-07,
+      "loss": 0.6765,
+      "step": 40173
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.27854857080708,
+      "learning_rate": 1.4989753357391502e-07,
+      "loss": 0.653,
+      "step": 40174
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5880106804196287,
+      "learning_rate": 1.4976594084238127e-07,
+      "loss": 0.476,
+      "step": 40175
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8859315779680106,
+      "learning_rate": 1.4963440546203023e-07,
+      "loss": 0.6847,
+      "step": 40176
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5965270466125585,
+      "learning_rate": 1.495029274336246e-07,
+      "loss": 0.4707,
+      "step": 40177
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.459154814699769,
+      "learning_rate": 1.4937150675793266e-07,
+      "loss": 0.7625,
+      "step": 40178
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3243206099807825,
+      "learning_rate": 1.4924014343571824e-07,
+      "loss": 0.6102,
+      "step": 40179
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3242888910729755,
+      "learning_rate": 1.491088374677463e-07,
+      "loss": 0.7287,
+      "step": 40180
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4027936079162788,
+      "learning_rate": 1.4897758885478064e-07,
+      "loss": 0.6265,
+      "step": 40181
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5262491554433217,
+      "learning_rate": 1.4884639759758735e-07,
+      "loss": 0.5916,
+      "step": 40182
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2398490353015044,
+      "learning_rate": 1.48715263696928e-07,
+      "loss": 0.6496,
+      "step": 40183
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0172224118973836,
+      "learning_rate": 1.485841871535687e-07,
+      "loss": 0.6949,
+      "step": 40184
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1639444154217053,
+      "learning_rate": 1.4845316796827103e-07,
+      "loss": 0.6702,
+      "step": 40185
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.466959549009879,
+      "learning_rate": 1.483222061417966e-07,
+      "loss": 0.6551,
+      "step": 40186
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4952638512195124,
+      "learning_rate": 1.4819130167491036e-07,
+      "loss": 0.637,
+      "step": 40187
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5714825192086125,
+      "learning_rate": 1.4806045456837393e-07,
+      "loss": 0.4782,
+      "step": 40188
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.464192310498557,
+      "learning_rate": 1.479296648229478e-07,
+      "loss": 0.6552,
+      "step": 40189
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2582841072547795,
+      "learning_rate": 1.4779893243939358e-07,
+      "loss": 0.6961,
+      "step": 40190
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3480201526710343,
+      "learning_rate": 1.47668257418474e-07,
+      "loss": 0.6444,
+      "step": 40191
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3953400102893072,
+      "learning_rate": 1.4753763976094848e-07,
+      "loss": 0.6332,
+      "step": 40192
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.507296649486419,
+      "learning_rate": 1.474070794675797e-07,
+      "loss": 0.7506,
+      "step": 40193
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.332609844526047,
+      "learning_rate": 1.4727657653912374e-07,
+      "loss": 0.6734,
+      "step": 40194
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6959071129544507,
+      "learning_rate": 1.4714613097634445e-07,
+      "loss": 0.7239,
+      "step": 40195
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7561915176354115,
+      "learning_rate": 1.4701574277999896e-07,
+      "loss": 0.7081,
+      "step": 40196
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3198223917770655,
+      "learning_rate": 1.468854119508478e-07,
+      "loss": 0.7429,
+      "step": 40197
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.305093179409143,
+      "learning_rate": 1.4675513848964928e-07,
+      "loss": 0.6259,
+      "step": 40198
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3332270015642542,
+      "learning_rate": 1.4662492239716164e-07,
+      "loss": 0.6958,
+      "step": 40199
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7324138377914338,
+      "learning_rate": 1.4649476367414318e-07,
+      "loss": 0.7934,
+      "step": 40200
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6676238187142596,
+      "learning_rate": 1.4636466232135217e-07,
+      "loss": 0.7625,
+      "step": 40201
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.4028150568809528,
+      "learning_rate": 1.4623461833954467e-07,
+      "loss": 0.5548,
+      "step": 40202
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3459134033170561,
+      "learning_rate": 1.4610463172948008e-07,
+      "loss": 0.6985,
+      "step": 40203
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4585985349321966,
+      "learning_rate": 1.4597470249191225e-07,
+      "loss": 0.8563,
+      "step": 40204
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4129238437829734,
+      "learning_rate": 1.458448306276017e-07,
+      "loss": 0.7165,
+      "step": 40205
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.9332216872790684,
+      "learning_rate": 1.457150161373011e-07,
+      "loss": 0.6834,
+      "step": 40206
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.33623978839007,
+      "learning_rate": 1.455852590217677e-07,
+      "loss": 0.7136,
+      "step": 40207
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3317156496946412,
+      "learning_rate": 1.4545555928175637e-07,
+      "loss": 0.8176,
+      "step": 40208
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.093976772384204,
+      "learning_rate": 1.4532591691802323e-07,
+      "loss": 0.693,
+      "step": 40209
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3525848386304984,
+      "learning_rate": 1.451963319313221e-07,
+      "loss": 0.6564,
+      "step": 40210
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.368008899620346,
+      "learning_rate": 1.4506680432240794e-07,
+      "loss": 0.6288,
+      "step": 40211
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4428528811181547,
+      "learning_rate": 1.449373340920346e-07,
+      "loss": 0.7345,
+      "step": 40212
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3319250946170915,
+      "learning_rate": 1.4480792124095698e-07,
+      "loss": 0.5868,
+      "step": 40213
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4952404618170807,
+      "learning_rate": 1.446785657699279e-07,
+      "loss": 0.7344,
+      "step": 40214
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7373780783651627,
+      "learning_rate": 1.445492676796989e-07,
+      "loss": 0.6087,
+      "step": 40215
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5234391555276612,
+      "learning_rate": 1.4442002697102497e-07,
+      "loss": 0.6383,
+      "step": 40216
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5844534696099135,
+      "learning_rate": 1.4429084364465885e-07,
+      "loss": 0.576,
+      "step": 40217
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6366226782111961,
+      "learning_rate": 1.44161717701351e-07,
+      "loss": 0.5031,
+      "step": 40218
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.4314700995634344,
+      "learning_rate": 1.440326491418531e-07,
+      "loss": 0.7074,
+      "step": 40219
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3587298382685997,
+      "learning_rate": 1.4390363796691897e-07,
+      "loss": 0.7406,
+      "step": 40220
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3370356446017742,
+      "learning_rate": 1.437746841772969e-07,
+      "loss": 0.6897,
+      "step": 40221
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.411998346287542,
+      "learning_rate": 1.4364578777373962e-07,
+      "loss": 0.8619,
+      "step": 40222
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.351158815847223,
+      "learning_rate": 1.4351694875699763e-07,
+      "loss": 0.744,
+      "step": 40223
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3483007979298027,
+      "learning_rate": 1.4338816712781921e-07,
+      "loss": 0.6683,
+      "step": 40224
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4958701486424184,
+      "learning_rate": 1.4325944288695603e-07,
+      "loss": 0.6091,
+      "step": 40225
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4066976141881578,
+      "learning_rate": 1.4313077603515745e-07,
+      "loss": 0.5881,
+      "step": 40226
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5795860721441402,
+      "learning_rate": 1.4300216657317178e-07,
+      "loss": 0.6079,
+      "step": 40227
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2786666550833905,
+      "learning_rate": 1.4287361450174843e-07,
+      "loss": 0.6646,
+      "step": 40228
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.487885729807141,
+      "learning_rate": 1.4274511982163454e-07,
+      "loss": 0.776,
+      "step": 40229
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4899212417344725,
+      "learning_rate": 1.4261668253358062e-07,
+      "loss": 0.7592,
+      "step": 40230
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4445887603207936,
+      "learning_rate": 1.424883026383328e-07,
+      "loss": 0.5628,
+      "step": 40231
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2866531439784932,
+      "learning_rate": 1.423599801366382e-07,
+      "loss": 0.5785,
+      "step": 40232
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5581999198443531,
+      "learning_rate": 1.4223171502924516e-07,
+      "loss": 0.492,
+      "step": 40233
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.298684289257564,
+      "learning_rate": 1.421035073168997e-07,
+      "loss": 0.7206,
+      "step": 40234
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5867189368592419,
+      "learning_rate": 1.4197535700035014e-07,
+      "loss": 0.4469,
+      "step": 40235
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4392803990483327,
+      "learning_rate": 1.418472640803392e-07,
+      "loss": 0.7821,
+      "step": 40236
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7817787437605717,
+      "learning_rate": 1.417192285576141e-07,
+      "loss": 0.7189,
+      "step": 40237
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4181467146069195,
+      "learning_rate": 1.415912504329231e-07,
+      "loss": 0.7265,
+      "step": 40238
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.327712266298879,
+      "learning_rate": 1.4146332970700782e-07,
+      "loss": 0.6409,
+      "step": 40239
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.401853606482501,
+      "learning_rate": 1.4133546638061323e-07,
+      "loss": 0.5536,
+      "step": 40240
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.645980073102324,
+      "learning_rate": 1.4120766045448653e-07,
+      "loss": 0.4752,
+      "step": 40241
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2597713417655398,
+      "learning_rate": 1.4107991192936821e-07,
+      "loss": 0.6534,
+      "step": 40242
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5134391949230928,
+      "learning_rate": 1.4095222080600546e-07,
+      "loss": 0.6443,
+      "step": 40243
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5244805075347894,
+      "learning_rate": 1.408245870851388e-07,
+      "loss": 0.714,
+      "step": 40244
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3303664904752763,
+      "learning_rate": 1.406970107675143e-07,
+      "loss": 0.714,
+      "step": 40245
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9299290324204956,
+      "learning_rate": 1.4056949185387136e-07,
+      "loss": 0.6871,
+      "step": 40246
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.297716834501123,
+      "learning_rate": 1.4044203034495497e-07,
+      "loss": 0.6817,
+      "step": 40247
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.014813347326067,
+      "learning_rate": 1.4031462624150782e-07,
+      "loss": 0.6075,
+      "step": 40248
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6625587437294043,
+      "learning_rate": 1.4018727954426937e-07,
+      "loss": 0.7553,
+      "step": 40249
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4208869497541055,
+      "learning_rate": 1.4005999025398231e-07,
+      "loss": 0.8157,
+      "step": 40250
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2635098043912898,
+      "learning_rate": 1.399327583713872e-07,
+      "loss": 0.6378,
+      "step": 40251
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6175800247377303,
+      "learning_rate": 1.3980558389722677e-07,
+      "loss": 0.4837,
+      "step": 40252
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6358721917714847,
+      "learning_rate": 1.3967846683223818e-07,
+      "loss": 0.5525,
+      "step": 40253
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3831261663225525,
+      "learning_rate": 1.3955140717716308e-07,
+      "loss": 0.6102,
+      "step": 40254
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6125059918882133,
+      "learning_rate": 1.3942440493274313e-07,
+      "loss": 0.5107,
+      "step": 40255
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.33199428340957,
+      "learning_rate": 1.3929746009971434e-07,
+      "loss": 0.6713,
+      "step": 40256
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.237870490568754,
+      "learning_rate": 1.3917057267881951e-07,
+      "loss": 0.6342,
+      "step": 40257
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4728424898531334,
+      "learning_rate": 1.3904374267079356e-07,
+      "loss": 0.7431,
+      "step": 40258
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.57085424466896,
+      "learning_rate": 1.3891697007637817e-07,
+      "loss": 0.6609,
+      "step": 40259
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2729711427591268,
+      "learning_rate": 1.3879025489630936e-07,
+      "loss": 0.6666,
+      "step": 40260
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2314466862184954,
+      "learning_rate": 1.386635971313266e-07,
+      "loss": 0.5802,
+      "step": 40261
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7143366169246101,
+      "learning_rate": 1.3853699678216592e-07,
+      "loss": 0.7046,
+      "step": 40262
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.482883930749631,
+      "learning_rate": 1.3841045384956453e-07,
+      "loss": 0.7126,
+      "step": 40263
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2287941767887742,
+      "learning_rate": 1.3828396833425962e-07,
+      "loss": 0.5022,
+      "step": 40264
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6017983494984307,
+      "learning_rate": 1.3815754023698834e-07,
+      "loss": 0.6964,
+      "step": 40265
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4665790137450725,
+      "learning_rate": 1.3803116955848573e-07,
+      "loss": 0.6353,
+      "step": 40266
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.444642085912684,
+      "learning_rate": 1.379048562994878e-07,
+      "loss": 0.7384,
+      "step": 40267
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6646979809568092,
+      "learning_rate": 1.3777860046072955e-07,
+      "loss": 0.6596,
+      "step": 40268
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6161212632257711,
+      "learning_rate": 1.3765240204294706e-07,
+      "loss": 0.5018,
+      "step": 40269
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8326419538146812,
+      "learning_rate": 1.3752626104687528e-07,
+      "loss": 0.6762,
+      "step": 40270
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3250897972618145,
+      "learning_rate": 1.3740017747324698e-07,
+      "loss": 0.6313,
+      "step": 40271
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3065138040388693,
+      "learning_rate": 1.372741513227982e-07,
+      "loss": 0.5554,
+      "step": 40272
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3332816716740337,
+      "learning_rate": 1.3714818259626173e-07,
+      "loss": 0.6225,
+      "step": 40273
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6311065791579585,
+      "learning_rate": 1.3702227129437028e-07,
+      "loss": 0.5037,
+      "step": 40274
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3765992767301998,
+      "learning_rate": 1.3689641741785775e-07,
+      "loss": 0.6504,
+      "step": 40275
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3052336569298086,
+      "learning_rate": 1.3677062096745686e-07,
+      "loss": 0.6465,
+      "step": 40276
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.592889570714732,
+      "learning_rate": 1.3664488194390147e-07,
+      "loss": 0.6333,
+      "step": 40277
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3578153909135044,
+      "learning_rate": 1.365192003479221e-07,
+      "loss": 0.5741,
+      "step": 40278
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3556505060109774,
+      "learning_rate": 1.363935761802493e-07,
+      "loss": 0.6492,
+      "step": 40279
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2515815762978515,
+      "learning_rate": 1.362680094416169e-07,
+      "loss": 0.7163,
+      "step": 40280
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4264185903617397,
+      "learning_rate": 1.3614250013275433e-07,
+      "loss": 0.6304,
+      "step": 40281
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3580147223683934,
+      "learning_rate": 1.3601704825439323e-07,
+      "loss": 0.6334,
+      "step": 40282
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5977803794625954,
+      "learning_rate": 1.35891653807263e-07,
+      "loss": 0.4703,
+      "step": 40283
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3586621329561734,
+      "learning_rate": 1.357663167920953e-07,
+      "loss": 0.6571,
+      "step": 40284
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3100678175307539,
+      "learning_rate": 1.356410372096184e-07,
+      "loss": 0.5794,
+      "step": 40285
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3295339133443038,
+      "learning_rate": 1.35515815060564e-07,
+      "loss": 0.6759,
+      "step": 40286
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5234008262916587,
+      "learning_rate": 1.3539065034565813e-07,
+      "loss": 0.6464,
+      "step": 40287
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2239799916697227,
+      "learning_rate": 1.3526554306563134e-07,
+      "loss": 0.5427,
+      "step": 40288
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.30813974169485,
+      "learning_rate": 1.351404932212119e-07,
+      "loss": 0.7188,
+      "step": 40289
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3125232240913323,
+      "learning_rate": 1.3501550081312707e-07,
+      "loss": 0.6662,
+      "step": 40290
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3293096669381448,
+      "learning_rate": 1.3489056584210626e-07,
+      "loss": 0.6643,
+      "step": 40291
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5136109273795821,
+      "learning_rate": 1.3476568830887439e-07,
+      "loss": 0.641,
+      "step": 40292
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6426021523797023,
+      "learning_rate": 1.346408682141609e-07,
+      "loss": 0.6222,
+      "step": 40293
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3062328050020098,
+      "learning_rate": 1.345161055586919e-07,
+      "loss": 0.8229,
+      "step": 40294
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3406590261644116,
+      "learning_rate": 1.3439140034319342e-07,
+      "loss": 0.7117,
+      "step": 40295
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5664463797434446,
+      "learning_rate": 1.342667525683905e-07,
+      "loss": 0.4071,
+      "step": 40296
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6187694229735373,
+      "learning_rate": 1.3414216223501031e-07,
+      "loss": 0.4926,
+      "step": 40297
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5603218526971634,
+      "learning_rate": 1.3401762934377894e-07,
+      "loss": 0.6304,
+      "step": 40298
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2968402020396894,
+      "learning_rate": 1.3389315389541911e-07,
+      "loss": 0.7237,
+      "step": 40299
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6311702703816382,
+      "learning_rate": 1.3376873589065809e-07,
+      "loss": 0.7937,
+      "step": 40300
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3571273112756048,
+      "learning_rate": 1.3364437533021856e-07,
+      "loss": 0.7893,
+      "step": 40301
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5750446122543456,
+      "learning_rate": 1.3352007221482443e-07,
+      "loss": 0.5052,
+      "step": 40302
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2564756731902798,
+      "learning_rate": 1.3339582654520177e-07,
+      "loss": 0.6881,
+      "step": 40303
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3357229166001938,
+      "learning_rate": 1.3327163832207114e-07,
+      "loss": 0.5843,
+      "step": 40304
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.687451849944574,
+      "learning_rate": 1.3314750754615634e-07,
+      "loss": 0.716,
+      "step": 40305
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.531121224328481,
+      "learning_rate": 1.330234342181813e-07,
+      "loss": 0.6762,
+      "step": 40306
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4367427877164811,
+      "learning_rate": 1.3289941833886767e-07,
+      "loss": 0.6024,
+      "step": 40307
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2748249029225411,
+      "learning_rate": 1.3277545990893813e-07,
+      "loss": 0.6377,
+      "step": 40308
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4267277683727182,
+      "learning_rate": 1.326515589291122e-07,
+      "loss": 0.7889,
+      "step": 40309
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.866310207502877,
+      "learning_rate": 1.3252771540011367e-07,
+      "loss": 0.804,
+      "step": 40310
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.333370431984068,
+      "learning_rate": 1.324039293226631e-07,
+      "loss": 0.6871,
+      "step": 40311
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.429944996775903,
+      "learning_rate": 1.3228020069748105e-07,
+      "loss": 0.5982,
+      "step": 40312
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2612819873290195,
+      "learning_rate": 1.321565295252869e-07,
+      "loss": 0.6688,
+      "step": 40313
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2848935822990026,
+      "learning_rate": 1.3203291580680122e-07,
+      "loss": 0.5486,
+      "step": 40314
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.396164176161774,
+      "learning_rate": 1.3190935954274565e-07,
+      "loss": 0.5848,
+      "step": 40315
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3955404913962146,
+      "learning_rate": 1.3178586073383738e-07,
+      "loss": 0.6386,
+      "step": 40316
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4802415829660087,
+      "learning_rate": 1.3166241938079472e-07,
+      "loss": 0.731,
+      "step": 40317
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4125263547274949,
+      "learning_rate": 1.3153903548433822e-07,
+      "loss": 0.7731,
+      "step": 40318
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3200500321147814,
+      "learning_rate": 1.314157090451873e-07,
+      "loss": 0.6305,
+      "step": 40319
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4674723696265437,
+      "learning_rate": 1.3129244006405805e-07,
+      "loss": 0.6539,
+      "step": 40320
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7892598498872276,
+      "learning_rate": 1.311692285416666e-07,
+      "loss": 0.6554,
+      "step": 40321
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3522314197756253,
+      "learning_rate": 1.3104607447873452e-07,
+      "loss": 0.7932,
+      "step": 40322
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.473503975911549,
+      "learning_rate": 1.3092297787597576e-07,
+      "loss": 0.5481,
+      "step": 40323
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.389077432449636,
+      "learning_rate": 1.3079993873410857e-07,
+      "loss": 0.6453,
+      "step": 40324
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4186154729265479,
+      "learning_rate": 1.3067695705384687e-07,
+      "loss": 0.6293,
+      "step": 40325
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.6556623943931124,
+      "learning_rate": 1.3055403283591006e-07,
+      "loss": 0.7217,
+      "step": 40326
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3184091110583276,
+      "learning_rate": 1.3043116608101204e-07,
+      "loss": 0.6188,
+      "step": 40327
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5974614246569443,
+      "learning_rate": 1.3030835678986886e-07,
+      "loss": 0.6631,
+      "step": 40328
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.348851656730498,
+      "learning_rate": 1.3018560496319443e-07,
+      "loss": 0.7199,
+      "step": 40329
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3484574387510468,
+      "learning_rate": 1.3006291060170373e-07,
+      "loss": 0.7214,
+      "step": 40330
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.617486428421509,
+      "learning_rate": 1.2994027370611173e-07,
+      "loss": 0.7467,
+      "step": 40331
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.881667686372712,
+      "learning_rate": 1.298176942771323e-07,
+      "loss": 0.6921,
+      "step": 40332
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4905215001566554,
+      "learning_rate": 1.2969517231548045e-07,
+      "loss": 0.6432,
+      "step": 40333
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3067679586175351,
+      "learning_rate": 1.295727078218667e-07,
+      "loss": 0.66,
+      "step": 40334
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2318075107406659,
+      "learning_rate": 1.2945030079700606e-07,
+      "loss": 0.7269,
+      "step": 40335
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5347340074265332,
+      "learning_rate": 1.2932795124161124e-07,
+      "loss": 0.7391,
+      "step": 40336
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6598740257430478,
+      "learning_rate": 1.2920565915639395e-07,
+      "loss": 0.7424,
+      "step": 40337
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.346631754365924,
+      "learning_rate": 1.290834245420658e-07,
+      "loss": 0.7109,
+      "step": 40338
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2366810680818499,
+      "learning_rate": 1.2896124739933846e-07,
+      "loss": 0.6487,
+      "step": 40339
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2930559922862876,
+      "learning_rate": 1.288391277289247e-07,
+      "loss": 0.5014,
+      "step": 40340
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.483377640385259,
+      "learning_rate": 1.2871706553153395e-07,
+      "loss": 0.6832,
+      "step": 40341
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5697485503407527,
+      "learning_rate": 1.2859506080787898e-07,
+      "loss": 0.7381,
+      "step": 40342
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5143530712407287,
+      "learning_rate": 1.284731135586681e-07,
+      "loss": 0.7005,
+      "step": 40343
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4273171842521435,
+      "learning_rate": 1.2835122378461186e-07,
+      "loss": 0.715,
+      "step": 40344
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3842140965734417,
+      "learning_rate": 1.2822939148641967e-07,
+      "loss": 0.6728,
+      "step": 40345
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6200209903012706,
+      "learning_rate": 1.2810761666480321e-07,
+      "loss": 0.6355,
+      "step": 40346
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.223317752127398,
+      "learning_rate": 1.2798589932046856e-07,
+      "loss": 0.6377,
+      "step": 40347
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3602266906860747,
+      "learning_rate": 1.2786423945412407e-07,
+      "loss": 0.6187,
+      "step": 40348
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4334163024868645,
+      "learning_rate": 1.2774263706648027e-07,
+      "loss": 0.6588,
+      "step": 40349
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6275504487848882,
+      "learning_rate": 1.2762109215824547e-07,
+      "loss": 0.7477,
+      "step": 40350
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2631785677920273,
+      "learning_rate": 1.274996047301258e-07,
+      "loss": 0.6114,
+      "step": 40351
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.301458233399208,
+      "learning_rate": 1.2737817478282845e-07,
+      "loss": 0.6129,
+      "step": 40352
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.315178622206669,
+      "learning_rate": 1.2725680231706061e-07,
+      "loss": 0.6494,
+      "step": 40353
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2265675169710613,
+      "learning_rate": 1.2713548733352955e-07,
+      "loss": 0.641,
+      "step": 40354
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2896175209945446,
+      "learning_rate": 1.2701422983294243e-07,
+      "loss": 0.6757,
+      "step": 40355
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4368040623687517,
+      "learning_rate": 1.2689302981600316e-07,
+      "loss": 0.6693,
+      "step": 40356
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1721440221901331,
+      "learning_rate": 1.2677188728341895e-07,
+      "loss": 0.5623,
+      "step": 40357
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4239153195409326,
+      "learning_rate": 1.2665080223589366e-07,
+      "loss": 0.83,
+      "step": 40358
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3618190943227926,
+      "learning_rate": 1.2652977467413453e-07,
+      "loss": 0.5926,
+      "step": 40359
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.55095306276931,
+      "learning_rate": 1.264088045988443e-07,
+      "loss": 0.6462,
+      "step": 40360
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5796285060402115,
+      "learning_rate": 1.26287892010728e-07,
+      "loss": 0.4851,
+      "step": 40361
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5247899671136795,
+      "learning_rate": 1.2616703691048947e-07,
+      "loss": 0.6768,
+      "step": 40362
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2935637673906974,
+      "learning_rate": 1.2604623929883265e-07,
+      "loss": 0.5739,
+      "step": 40363
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.193684249947466,
+      "learning_rate": 1.2592549917646136e-07,
+      "loss": 0.6316,
+      "step": 40364
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6016493020816644,
+      "learning_rate": 1.2580481654407616e-07,
+      "loss": 0.7522,
+      "step": 40365
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5674834068718912,
+      "learning_rate": 1.2568419140238208e-07,
+      "loss": 0.6498,
+      "step": 40366
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7937736560432533,
+      "learning_rate": 1.2556362375208186e-07,
+      "loss": 0.6132,
+      "step": 40367
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3942380189665764,
+      "learning_rate": 1.2544311359387606e-07,
+      "loss": 0.5965,
+      "step": 40368
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3764871284861122,
+      "learning_rate": 1.2532266092846524e-07,
+      "loss": 0.5893,
+      "step": 40369
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8257172169162785,
+      "learning_rate": 1.2520226575655325e-07,
+      "loss": 0.6222,
+      "step": 40370
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.427959058076378,
+      "learning_rate": 1.2508192807883956e-07,
+      "loss": 0.7269,
+      "step": 40371
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6216939528789659,
+      "learning_rate": 1.2496164789602583e-07,
+      "loss": 0.524,
+      "step": 40372
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7211354789964495,
+      "learning_rate": 1.248414252088115e-07,
+      "loss": 0.6246,
+      "step": 40373
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.295077307842463,
+      "learning_rate": 1.2472126001789708e-07,
+      "loss": 0.7265,
+      "step": 40374
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2449470781254466,
+      "learning_rate": 1.2460115232398207e-07,
+      "loss": 0.6344,
+      "step": 40375
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1792951423910119,
+      "learning_rate": 1.2448110212776477e-07,
+      "loss": 0.6656,
+      "step": 40376
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.339230877643814,
+      "learning_rate": 1.2436110942994573e-07,
+      "loss": 0.6231,
+      "step": 40377
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4934175608982383,
+      "learning_rate": 1.242411742312233e-07,
+      "loss": 0.5592,
+      "step": 40378
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4062879151313452,
+      "learning_rate": 1.2412129653229355e-07,
+      "loss": 0.5278,
+      "step": 40379
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2902973577689303,
+      "learning_rate": 1.2400147633385817e-07,
+      "loss": 0.6537,
+      "step": 40380
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8536255683006087,
+      "learning_rate": 1.2388171363661216e-07,
+      "loss": 0.6917,
+      "step": 40381
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4232221944501282,
+      "learning_rate": 1.2376200844125387e-07,
+      "loss": 0.6888,
+      "step": 40382
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.388258623875273,
+      "learning_rate": 1.2364236074847934e-07,
+      "loss": 0.704,
+      "step": 40383
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.602917839632461,
+      "learning_rate": 1.2352277055898587e-07,
+      "loss": 0.6639,
+      "step": 40384
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.297229149610303,
+      "learning_rate": 1.2340323787347064e-07,
+      "loss": 0.6868,
+      "step": 40385
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3915869435266113,
+      "learning_rate": 1.2328376269262754e-07,
+      "loss": 0.6405,
+      "step": 40386
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3423812032486417,
+      "learning_rate": 1.231643450171527e-07,
+      "loss": 0.5503,
+      "step": 40387
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2347512556375546,
+      "learning_rate": 1.2304498484774329e-07,
+      "loss": 0.7113,
+      "step": 40388
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4344592430800542,
+      "learning_rate": 1.2292568218509325e-07,
+      "loss": 0.7928,
+      "step": 40389
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2944918230642293,
+      "learning_rate": 1.2280643702989647e-07,
+      "loss": 0.6245,
+      "step": 40390
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5851165765451054,
+      "learning_rate": 1.2268724938284682e-07,
+      "loss": 0.7197,
+      "step": 40391
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.597765837824332,
+      "learning_rate": 1.225681192446404e-07,
+      "loss": 0.6708,
+      "step": 40392
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2274237526575127,
+      "learning_rate": 1.224490466159689e-07,
+      "loss": 0.6086,
+      "step": 40393
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3893323292574644,
+      "learning_rate": 1.223300314975262e-07,
+      "loss": 0.7667,
+      "step": 40394
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3948152588540363,
+      "learning_rate": 1.2221107389000508e-07,
+      "loss": 0.6472,
+      "step": 40395
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4403665262402838,
+      "learning_rate": 1.2209217379409834e-07,
+      "loss": 0.6654,
+      "step": 40396
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5000576171629625,
+      "learning_rate": 1.219733312104987e-07,
+      "loss": 0.5911,
+      "step": 40397
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4987082166263268,
+      "learning_rate": 1.2185454613989678e-07,
+      "loss": 0.6687,
+      "step": 40398
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5950892865739051,
+      "learning_rate": 1.2173581858298644e-07,
+      "loss": 0.4725,
+      "step": 40399
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4981837493372068,
+      "learning_rate": 1.2161714854045713e-07,
+      "loss": 0.7868,
+      "step": 40400
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3929858110431252,
+      "learning_rate": 1.2149853601299945e-07,
+      "loss": 0.7281,
+      "step": 40401
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3503601742704157,
+      "learning_rate": 1.21379981001305e-07,
+      "loss": 0.7847,
+      "step": 40402
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3938105771960567,
+      "learning_rate": 1.2126148350606438e-07,
+      "loss": 0.6077,
+      "step": 40403
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2624825780892097,
+      "learning_rate": 1.2114304352796703e-07,
+      "loss": 0.5895,
+      "step": 40404
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3537363223293892,
+      "learning_rate": 1.210246610677024e-07,
+      "loss": 0.7293,
+      "step": 40405
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.376212183097351,
+      "learning_rate": 1.2090633612595881e-07,
+      "loss": 0.7183,
+      "step": 40406
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.377908551418916,
+      "learning_rate": 1.2078806870342684e-07,
+      "loss": 0.6536,
+      "step": 40407
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5946600338048704,
+      "learning_rate": 1.2066985880079484e-07,
+      "loss": 0.5188,
+      "step": 40408
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3936047292954137,
+      "learning_rate": 1.2055170641875003e-07,
+      "loss": 0.5974,
+      "step": 40409
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.431445402592115,
+      "learning_rate": 1.2043361155798184e-07,
+      "loss": 0.5977,
+      "step": 40410
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3585426218769217,
+      "learning_rate": 1.2031557421917639e-07,
+      "loss": 0.7257,
+      "step": 40411
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4368442221795916,
+      "learning_rate": 1.2019759440302093e-07,
+      "loss": 0.7192,
+      "step": 40412
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3768848501815203,
+      "learning_rate": 1.2007967211020378e-07,
+      "loss": 0.6617,
+      "step": 40413
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4022893601427158,
+      "learning_rate": 1.1996180734141104e-07,
+      "loss": 0.6648,
+      "step": 40414
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.549791416320665,
+      "learning_rate": 1.1984400009732776e-07,
+      "loss": 0.8174,
+      "step": 40415
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4110912967901652,
+      "learning_rate": 1.1972625037864117e-07,
+      "loss": 0.5887,
+      "step": 40416
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8960102448192,
+      "learning_rate": 1.1960855818603624e-07,
+      "loss": 0.6017,
+      "step": 40417
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6085439726505637,
+      "learning_rate": 1.194909235201991e-07,
+      "loss": 0.4789,
+      "step": 40418
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3688980340948869,
+      "learning_rate": 1.1937334638181252e-07,
+      "loss": 0.6222,
+      "step": 40419
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5437014802486437,
+      "learning_rate": 1.1925582677156268e-07,
+      "loss": 0.7902,
+      "step": 40420
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3273925370803443,
+      "learning_rate": 1.1913836469013451e-07,
+      "loss": 0.7291,
+      "step": 40421
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3609265569020659,
+      "learning_rate": 1.1902096013821085e-07,
+      "loss": 0.6612,
+      "step": 40422
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.708107787752012,
+      "learning_rate": 1.1890361311647558e-07,
+      "loss": 0.7084,
+      "step": 40423
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5059429871432144,
+      "learning_rate": 1.1878632362561149e-07,
+      "loss": 0.6682,
+      "step": 40424
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4659621679467183,
+      "learning_rate": 1.1866909166630136e-07,
+      "loss": 0.7171,
+      "step": 40425
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3453659697718647,
+      "learning_rate": 1.185519172392291e-07,
+      "loss": 0.7643,
+      "step": 40426
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.36625534489082,
+      "learning_rate": 1.1843480034507415e-07,
+      "loss": 0.778,
+      "step": 40427
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.400278631125353,
+      "learning_rate": 1.1831774098452153e-07,
+      "loss": 0.6135,
+      "step": 40428
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3802153667244128,
+      "learning_rate": 1.1820073915825181e-07,
+      "loss": 0.8163,
+      "step": 40429
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2641929750991345,
+      "learning_rate": 1.1808379486694444e-07,
+      "loss": 0.6919,
+      "step": 40430
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6712523561652801,
+      "learning_rate": 1.1796690811128442e-07,
+      "loss": 0.6435,
+      "step": 40431
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.712737572294401,
+      "learning_rate": 1.1785007889194788e-07,
+      "loss": 0.6136,
+      "step": 40432
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.672353344525798,
+      "learning_rate": 1.1773330720961652e-07,
+      "loss": 0.7034,
+      "step": 40433
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3102366521474857,
+      "learning_rate": 1.1761659306497086e-07,
+      "loss": 0.6759,
+      "step": 40434
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.487328351774409,
+      "learning_rate": 1.1749993645869151e-07,
+      "loss": 0.696,
+      "step": 40435
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2534594612782555,
+      "learning_rate": 1.1738333739145457e-07,
+      "loss": 0.6344,
+      "step": 40436
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6296604962609249,
+      "learning_rate": 1.1726679586394064e-07,
+      "loss": 0.4486,
+      "step": 40437
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2753329490481369,
+      "learning_rate": 1.1715031187682912e-07,
+      "loss": 0.5625,
+      "step": 40438
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6163608452799626,
+      "learning_rate": 1.1703388543079619e-07,
+      "loss": 0.5228,
+      "step": 40439
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4473882921932997,
+      "learning_rate": 1.169175165265224e-07,
+      "loss": 0.6504,
+      "step": 40440
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6132025601174124,
+      "learning_rate": 1.1680120516468274e-07,
+      "loss": 0.5086,
+      "step": 40441
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.345420792181811,
+      "learning_rate": 1.1668495134595558e-07,
+      "loss": 0.6093,
+      "step": 40442
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3180735720901917,
+      "learning_rate": 1.1656875507101706e-07,
+      "loss": 0.5768,
+      "step": 40443
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5709669048123323,
+      "learning_rate": 1.164526163405455e-07,
+      "loss": 0.7452,
+      "step": 40444
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5259251759571049,
+      "learning_rate": 1.1633653515521593e-07,
+      "loss": 0.7013,
+      "step": 40445
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.627102570430971,
+      "learning_rate": 1.1622051151570335e-07,
+      "loss": 0.5594,
+      "step": 40446
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6114157499140066,
+      "learning_rate": 1.161045454226839e-07,
+      "loss": 0.5125,
+      "step": 40447
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5901751626998655,
+      "learning_rate": 1.1598863687683371e-07,
+      "loss": 0.6623,
+      "step": 40448
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4180289839657458,
+      "learning_rate": 1.1587278587882667e-07,
+      "loss": 0.6733,
+      "step": 40449
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1863196529552764,
+      "learning_rate": 1.157569924293367e-07,
+      "loss": 0.5521,
+      "step": 40450
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5857085357470676,
+      "learning_rate": 1.156412565290399e-07,
+      "loss": 0.5742,
+      "step": 40451
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3343200735014906,
+      "learning_rate": 1.155255781786091e-07,
+      "loss": 0.656,
+      "step": 40452
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2824882376749507,
+      "learning_rate": 1.1540995737871708e-07,
+      "loss": 0.6824,
+      "step": 40453
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2718782721795794,
+      "learning_rate": 1.1529439413003773e-07,
+      "loss": 0.5664,
+      "step": 40454
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5972965379921171,
+      "learning_rate": 1.1517888843324387e-07,
+      "loss": 0.4793,
+      "step": 40455
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7651227072582405,
+      "learning_rate": 1.1506344028900829e-07,
+      "loss": 0.7306,
+      "step": 40456
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6826395261555267,
+      "learning_rate": 1.1494804969800377e-07,
+      "loss": 0.7307,
+      "step": 40457
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2418197205564643,
+      "learning_rate": 1.148327166608998e-07,
+      "loss": 0.6545,
+      "step": 40458
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2951171852087582,
+      "learning_rate": 1.1471744117837024e-07,
+      "loss": 0.6635,
+      "step": 40459
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.284986545605839,
+      "learning_rate": 1.1460222325108461e-07,
+      "loss": 0.6806,
+      "step": 40460
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2760299763645275,
+      "learning_rate": 1.1448706287971567e-07,
+      "loss": 0.5201,
+      "step": 40461
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7246935929567087,
+      "learning_rate": 1.1437196006493179e-07,
+      "loss": 0.6903,
+      "step": 40462
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3351538515439094,
+      "learning_rate": 1.1425691480740353e-07,
+      "loss": 0.7503,
+      "step": 40463
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2983113984778574,
+      "learning_rate": 1.1414192710780259e-07,
+      "loss": 0.6163,
+      "step": 40464
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.273656114842902,
+      "learning_rate": 1.140269969667962e-07,
+      "loss": 0.5702,
+      "step": 40465
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7324105836461077,
+      "learning_rate": 1.1391212438505605e-07,
+      "loss": 0.6358,
+      "step": 40466
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3663667062005793,
+      "learning_rate": 1.1379730936324717e-07,
+      "loss": 0.6978,
+      "step": 40467
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6106087702135354,
+      "learning_rate": 1.1368255190204125e-07,
+      "loss": 0.6422,
+      "step": 40468
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3887976175447276,
+      "learning_rate": 1.1356785200210552e-07,
+      "loss": 0.5643,
+      "step": 40469
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4555077472593307,
+      "learning_rate": 1.1345320966410833e-07,
+      "loss": 0.6705,
+      "step": 40470
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3498809471739315,
+      "learning_rate": 1.1333862488871583e-07,
+      "loss": 0.6672,
+      "step": 40471
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3048872097300221,
+      "learning_rate": 1.1322409767659526e-07,
+      "loss": 0.7024,
+      "step": 40472
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6035098591031595,
+      "learning_rate": 1.1310962802841607e-07,
+      "loss": 0.5001,
+      "step": 40473
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4400223599670225,
+      "learning_rate": 1.129952159448422e-07,
+      "loss": 0.6101,
+      "step": 40474
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.369555608542,
+      "learning_rate": 1.1288086142653865e-07,
+      "loss": 0.6414,
+      "step": 40475
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4407195900788463,
+      "learning_rate": 1.1276656447417378e-07,
+      "loss": 0.664,
+      "step": 40476
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.497628994437282,
+      "learning_rate": 1.1265232508841262e-07,
+      "loss": 0.5509,
+      "step": 40477
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5543005027386971,
+      "learning_rate": 1.1253814326992019e-07,
+      "loss": 0.7413,
+      "step": 40478
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3763003455471636,
+      "learning_rate": 1.124240190193604e-07,
+      "loss": 0.625,
+      "step": 40479
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4878339011436728,
+      "learning_rate": 1.1230995233739939e-07,
+      "loss": 0.7655,
+      "step": 40480
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.278773106725421,
+      "learning_rate": 1.1219594322469885e-07,
+      "loss": 0.6484,
+      "step": 40481
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5614912343234646,
+      "learning_rate": 1.1208199168192491e-07,
+      "loss": 0.6142,
+      "step": 40482
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4430316286794955,
+      "learning_rate": 1.1196809770973926e-07,
+      "loss": 0.6151,
+      "step": 40483
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4133587614443777,
+      "learning_rate": 1.1185426130880694e-07,
+      "loss": 0.7237,
+      "step": 40484
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5578429682358567,
+      "learning_rate": 1.1174048247978852e-07,
+      "loss": 0.4686,
+      "step": 40485
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3094695422149887,
+      "learning_rate": 1.1162676122334793e-07,
+      "loss": 0.6588,
+      "step": 40486
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7215013376067825,
+      "learning_rate": 1.1151309754014683e-07,
+      "loss": 0.6825,
+      "step": 40487
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4139533462529714,
+      "learning_rate": 1.1139949143084804e-07,
+      "loss": 0.5951,
+      "step": 40488
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4871282955075287,
+      "learning_rate": 1.1128594289611105e-07,
+      "loss": 0.746,
+      "step": 40489
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7993988317062077,
+      "learning_rate": 1.1117245193659864e-07,
+      "loss": 0.6802,
+      "step": 40490
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4693056931510011,
+      "learning_rate": 1.110590185529703e-07,
+      "loss": 0.797,
+      "step": 40491
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6161431018974592,
+      "learning_rate": 1.109456427458877e-07,
+      "loss": 0.4814,
+      "step": 40492
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.384669085665304,
+      "learning_rate": 1.1083232451600922e-07,
+      "loss": 0.7343,
+      "step": 40493
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.461489570046143,
+      "learning_rate": 1.1071906386399767e-07,
+      "loss": 0.8084,
+      "step": 40494
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4356673473232981,
+      "learning_rate": 1.1060586079050917e-07,
+      "loss": 0.7442,
+      "step": 40495
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8583139766721422,
+      "learning_rate": 1.1049271529620542e-07,
+      "loss": 0.7477,
+      "step": 40496
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.401574086666871,
+      "learning_rate": 1.1037962738174257e-07,
+      "loss": 0.6876,
+      "step": 40497
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5879853903933683,
+      "learning_rate": 1.102665970477823e-07,
+      "loss": 0.489,
+      "step": 40498
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4263919933156957,
+      "learning_rate": 1.1015362429497967e-07,
+      "loss": 0.6469,
+      "step": 40499
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.403009550772012,
+      "learning_rate": 1.1004070912399411e-07,
+      "loss": 0.7022,
+      "step": 40500
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5045065257371484,
+      "learning_rate": 1.099278515354829e-07,
+      "loss": 0.6587,
+      "step": 40501
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3010482533632601,
+      "learning_rate": 1.0981505153010218e-07,
+      "loss": 0.7129,
+      "step": 40502
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3182436409953933,
+      "learning_rate": 1.0970230910851031e-07,
+      "loss": 0.8067,
+      "step": 40503
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.511011310267782,
+      "learning_rate": 1.0958962427136232e-07,
+      "loss": 0.7198,
+      "step": 40504
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4040483292106922,
+      "learning_rate": 1.0947699701931436e-07,
+      "loss": 0.6203,
+      "step": 40505
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3064550106814878,
+      "learning_rate": 1.0936442735302255e-07,
+      "loss": 0.6174,
+      "step": 40506
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.519722235690335,
+      "learning_rate": 1.0925191527314416e-07,
+      "loss": 0.5848,
+      "step": 40507
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.372296562156314,
+      "learning_rate": 1.0913946078033089e-07,
+      "loss": 0.7313,
+      "step": 40508
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3077910810908926,
+      "learning_rate": 1.0902706387523887e-07,
+      "loss": 0.7815,
+      "step": 40509
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4622763548510391,
+      "learning_rate": 1.0891472455852315e-07,
+      "loss": 0.7215,
+      "step": 40510
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.491528380953076,
+      "learning_rate": 1.0880244283083762e-07,
+      "loss": 0.6933,
+      "step": 40511
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.266479395326358,
+      "learning_rate": 1.0869021869283625e-07,
+      "loss": 0.5622,
+      "step": 40512
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4204344797567925,
+      "learning_rate": 1.085780521451707e-07,
+      "loss": 0.6099,
+      "step": 40513
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4526655772291792,
+      "learning_rate": 1.0846594318849602e-07,
+      "loss": 0.6282,
+      "step": 40514
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8390330466337594,
+      "learning_rate": 1.0835389182346389e-07,
+      "loss": 0.7162,
+      "step": 40515
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.315903769785406,
+      "learning_rate": 1.0824189805072827e-07,
+      "loss": 0.6898,
+      "step": 40516
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.626506555937951,
+      "learning_rate": 1.081299618709386e-07,
+      "loss": 0.4933,
+      "step": 40517
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3448982886405798,
+      "learning_rate": 1.0801808328474772e-07,
+      "loss": 0.6589,
+      "step": 40518
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2974363324922125,
+      "learning_rate": 1.0790626229280842e-07,
+      "loss": 0.6258,
+      "step": 40519
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3885524462266046,
+      "learning_rate": 1.077944988957702e-07,
+      "loss": 0.6554,
+      "step": 40520
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5575873214562463,
+      "learning_rate": 1.0768279309428365e-07,
+      "loss": 0.5758,
+      "step": 40521
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2891514597413518,
+      "learning_rate": 1.0757114488900156e-07,
+      "loss": 0.5957,
+      "step": 40522
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3473971272662648,
+      "learning_rate": 1.0745955428057009e-07,
+      "loss": 0.6443,
+      "step": 40523
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6082949546771691,
+      "learning_rate": 1.0734802126964206e-07,
+      "loss": 0.5425,
+      "step": 40524
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6079730234680483,
+      "learning_rate": 1.0723654585686582e-07,
+      "loss": 0.6625,
+      "step": 40525
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3749012551161799,
+      "learning_rate": 1.0712512804288977e-07,
+      "loss": 0.6624,
+      "step": 40526
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.54778362608875,
+      "learning_rate": 1.0701376782836337e-07,
+      "loss": 0.655,
+      "step": 40527
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.543430694936462,
+      "learning_rate": 1.0690246521393499e-07,
+      "loss": 0.6431,
+      "step": 40528
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.602910999544861,
+      "learning_rate": 1.0679122020025301e-07,
+      "loss": 0.7569,
+      "step": 40529
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.2323217335286625,
+      "learning_rate": 1.0668003278796468e-07,
+      "loss": 0.7026,
+      "step": 40530
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2351575489177984,
+      "learning_rate": 1.0656890297771727e-07,
+      "loss": 0.638,
+      "step": 40531
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3725930497215426,
+      "learning_rate": 1.0645783077015692e-07,
+      "loss": 0.6335,
+      "step": 40532
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5804808978858612,
+      "learning_rate": 1.0634681616593312e-07,
+      "loss": 0.4721,
+      "step": 40533
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6229242508885036,
+      "learning_rate": 1.0623585916568979e-07,
+      "loss": 0.5051,
+      "step": 40534
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4936971906129959,
+      "learning_rate": 1.0612495977007197e-07,
+      "loss": 0.6386,
+      "step": 40535
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1979889053740211,
+      "learning_rate": 1.0601411797972804e-07,
+      "loss": 0.7218,
+      "step": 40536
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4509655020461762,
+      "learning_rate": 1.0590333379530304e-07,
+      "loss": 0.6692,
+      "step": 40537
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5109957858279806,
+      "learning_rate": 1.0579260721744089e-07,
+      "loss": 0.6313,
+      "step": 40538
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4398932233147659,
+      "learning_rate": 1.0568193824678663e-07,
+      "loss": 0.6857,
+      "step": 40539
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4009570840054237,
+      "learning_rate": 1.055713268839853e-07,
+      "loss": 0.5862,
+      "step": 40540
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2918615654495027,
+      "learning_rate": 1.0546077312967862e-07,
+      "loss": 0.6644,
+      "step": 40541
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.876235369517646,
+      "learning_rate": 1.0535027698451272e-07,
+      "loss": 0.663,
+      "step": 40542
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4888961162422065,
+      "learning_rate": 1.0523983844913043e-07,
+      "loss": 0.61,
+      "step": 40543
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2634613391869722,
+      "learning_rate": 1.0512945752417347e-07,
+      "loss": 0.7479,
+      "step": 40544
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4863019361840817,
+      "learning_rate": 1.0501913421028575e-07,
+      "loss": 0.8072,
+      "step": 40545
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5068181246451269,
+      "learning_rate": 1.049088685081101e-07,
+      "loss": 0.6504,
+      "step": 40546
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.244761245610734,
+      "learning_rate": 1.0479866041828824e-07,
+      "loss": 0.6753,
+      "step": 40547
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6079481401818461,
+      "learning_rate": 1.0468850994145962e-07,
+      "loss": 0.5456,
+      "step": 40548
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3803076796771576,
+      "learning_rate": 1.0457841707826822e-07,
+      "loss": 0.6661,
+      "step": 40549
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4534229412466693,
+      "learning_rate": 1.0446838182935349e-07,
+      "loss": 0.6395,
+      "step": 40550
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4108792361466076,
+      "learning_rate": 1.0435840419535826e-07,
+      "loss": 0.6324,
+      "step": 40551
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6762004533388686,
+      "learning_rate": 1.042484841769198e-07,
+      "loss": 0.7364,
+      "step": 40552
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4292202993645087,
+      "learning_rate": 1.0413862177467982e-07,
+      "loss": 0.7823,
+      "step": 40553
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3234077475409138,
+      "learning_rate": 1.040288169892778e-07,
+      "loss": 0.6645,
+      "step": 40554
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3796981354927877,
+      "learning_rate": 1.0391906982135324e-07,
+      "loss": 0.5532,
+      "step": 40555
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.450682385707193,
+      "learning_rate": 1.0380938027154452e-07,
+      "loss": 0.6853,
+      "step": 40556
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2038772119411942,
+      "learning_rate": 1.0369974834049112e-07,
+      "loss": 0.6519,
+      "step": 40557
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6827950919317454,
+      "learning_rate": 1.0359017402883143e-07,
+      "loss": 0.6519,
+      "step": 40558
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4084952526630399,
+      "learning_rate": 1.0348065733720269e-07,
+      "loss": 0.5942,
+      "step": 40559
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2960698928763008,
+      "learning_rate": 1.0337119826624109e-07,
+      "loss": 0.6725,
+      "step": 40560
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3893108943987644,
+      "learning_rate": 1.0326179681658721e-07,
+      "loss": 0.7052,
+      "step": 40561
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4970125226888373,
+      "learning_rate": 1.031524529888761e-07,
+      "loss": 0.6549,
+      "step": 40562
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.38070767080134,
+      "learning_rate": 1.0304316678374504e-07,
+      "loss": 0.6847,
+      "step": 40563
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6620826507870239,
+      "learning_rate": 1.0293393820182907e-07,
+      "loss": 0.6766,
+      "step": 40564
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.546521306636501,
+      "learning_rate": 1.0282476724376544e-07,
+      "loss": 0.6638,
+      "step": 40565
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4362927159829908,
+      "learning_rate": 1.0271565391018922e-07,
+      "loss": 0.7262,
+      "step": 40566
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2018038473732064,
+      "learning_rate": 1.0260659820173658e-07,
+      "loss": 0.6876,
+      "step": 40567
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8331688657369023,
+      "learning_rate": 1.0249760011904253e-07,
+      "loss": 0.6286,
+      "step": 40568
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5536728726824294,
+      "learning_rate": 1.0238865966273881e-07,
+      "loss": 0.7524,
+      "step": 40569
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7598937221888138,
+      "learning_rate": 1.0227977683346268e-07,
+      "loss": 0.7472,
+      "step": 40570
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2197468810184902,
+      "learning_rate": 1.021709516318481e-07,
+      "loss": 0.6423,
+      "step": 40571
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.503858130531077,
+      "learning_rate": 1.0206218405852786e-07,
+      "loss": 0.5977,
+      "step": 40572
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3685120137190574,
+      "learning_rate": 1.019534741141337e-07,
+      "loss": 0.7298,
+      "step": 40573
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.293801819631806,
+      "learning_rate": 1.0184482179930066e-07,
+      "loss": 0.5697,
+      "step": 40574
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3188882386650416,
+      "learning_rate": 1.0173622711466157e-07,
+      "loss": 0.6197,
+      "step": 40575
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3126392310166708,
+      "learning_rate": 1.0162769006084705e-07,
+      "loss": 0.6891,
+      "step": 40576
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3851544953716097,
+      "learning_rate": 1.0151921063848991e-07,
+      "loss": 0.7445,
+      "step": 40577
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6275369672512576,
+      "learning_rate": 1.0141078884822187e-07,
+      "loss": 0.6328,
+      "step": 40578
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9442522917408596,
+      "learning_rate": 1.0130242469067464e-07,
+      "loss": 0.7368,
+      "step": 40579
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3137103330943363,
+      "learning_rate": 1.0119411816647884e-07,
+      "loss": 0.6841,
+      "step": 40580
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4772000787980277,
+      "learning_rate": 1.0108586927626285e-07,
+      "loss": 0.5399,
+      "step": 40581
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.021528140553589,
+      "learning_rate": 1.009776780206606e-07,
+      "loss": 0.5163,
+      "step": 40582
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3402122211780352,
+      "learning_rate": 1.008695444002994e-07,
+      "loss": 0.7255,
+      "step": 40583
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6135484161525522,
+      "learning_rate": 1.0076146841581091e-07,
+      "loss": 0.7491,
+      "step": 40584
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.412018060637808,
+      "learning_rate": 1.0065345006782246e-07,
+      "loss": 0.6758,
+      "step": 40585
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2646309691910969,
+      "learning_rate": 1.0054548935696351e-07,
+      "loss": 0.7417,
+      "step": 40586
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4416134759309651,
+      "learning_rate": 1.0043758628386247e-07,
+      "loss": 0.687,
+      "step": 40587
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3585793763238587,
+      "learning_rate": 1.0032974084914882e-07,
+      "loss": 0.5549,
+      "step": 40588
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2136236901037913,
+      "learning_rate": 1.0022195305344984e-07,
+      "loss": 0.5393,
+      "step": 40589
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5933513389210228,
+      "learning_rate": 1.0011422289739169e-07,
+      "loss": 0.4902,
+      "step": 40590
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.450365396210495,
+      "learning_rate": 1.0000655038160278e-07,
+      "loss": 0.7214,
+      "step": 40591
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3434741257278016,
+      "learning_rate": 9.989893550671148e-08,
+      "loss": 0.7051,
+      "step": 40592
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2146453227733003,
+      "learning_rate": 9.979137827334173e-08,
+      "loss": 0.6938,
+      "step": 40593
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.26625441603328,
+      "learning_rate": 9.968387868212192e-08,
+      "loss": 0.6947,
+      "step": 40594
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5411082768890039,
+      "learning_rate": 9.95764367336749e-08,
+      "loss": 0.6421,
+      "step": 40595
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.15954907817503,
+      "learning_rate": 9.946905242863014e-08,
+      "loss": 0.6084,
+      "step": 40596
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7291051045941208,
+      "learning_rate": 9.93617257676116e-08,
+      "loss": 0.5294,
+      "step": 40597
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3818753320281596,
+      "learning_rate": 9.925445675124212e-08,
+      "loss": 0.6152,
+      "step": 40598
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4830127880647355,
+      "learning_rate": 9.914724538014787e-08,
+      "loss": 0.7703,
+      "step": 40599
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8342978312406033,
+      "learning_rate": 9.904009165495276e-08,
+      "loss": 0.7,
+      "step": 40600
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4571313018098764,
+      "learning_rate": 9.893299557628188e-08,
+      "loss": 0.6761,
+      "step": 40601
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7631908492924402,
+      "learning_rate": 9.882595714475696e-08,
+      "loss": 0.6595,
+      "step": 40602
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3773077607122757,
+      "learning_rate": 9.871897636100191e-08,
+      "loss": 0.6204,
+      "step": 40603
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3885981298253725,
+      "learning_rate": 9.86120532256396e-08,
+      "loss": 0.6238,
+      "step": 40604
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6267168337948754,
+      "learning_rate": 9.850518773929286e-08,
+      "loss": 0.6319,
+      "step": 40605
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.247934058153312,
+      "learning_rate": 9.839837990258338e-08,
+      "loss": 0.5535,
+      "step": 40606
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2615964795865886,
+      "learning_rate": 9.829162971613405e-08,
+      "loss": 0.6022,
+      "step": 40607
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.610646476847622,
+      "learning_rate": 9.818493718056543e-08,
+      "loss": 0.7367,
+      "step": 40608
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7070547866480779,
+      "learning_rate": 9.807830229649929e-08,
+      "loss": 0.6164,
+      "step": 40609
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6451599093875027,
+      "learning_rate": 9.79717250645551e-08,
+      "loss": 0.7092,
+      "step": 40610
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2597127580477223,
+      "learning_rate": 9.78652054853546e-08,
+      "loss": 0.6956,
+      "step": 40611
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4427099462464026,
+      "learning_rate": 9.775874355951842e-08,
+      "loss": 0.6115,
+      "step": 40612
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3827302981423357,
+      "learning_rate": 9.765233928766493e-08,
+      "loss": 0.8073,
+      "step": 40613
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3840436347186265,
+      "learning_rate": 9.754599267041697e-08,
+      "loss": 0.5876,
+      "step": 40614
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4424794796309386,
+      "learning_rate": 9.743970370838963e-08,
+      "loss": 0.7255,
+      "step": 40615
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3851484731974542,
+      "learning_rate": 9.733347240220237e-08,
+      "loss": 0.7842,
+      "step": 40616
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.453422346573806,
+      "learning_rate": 9.722729875247694e-08,
+      "loss": 0.6744,
+      "step": 40617
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2383052298492674,
+      "learning_rate": 9.71211827598284e-08,
+      "loss": 0.8384,
+      "step": 40618
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6589526286975544,
+      "learning_rate": 9.701512442487515e-08,
+      "loss": 0.5912,
+      "step": 40619
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.420546659203044,
+      "learning_rate": 9.690912374823446e-08,
+      "loss": 0.7377,
+      "step": 40620
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.264268373370389,
+      "learning_rate": 9.680318073052474e-08,
+      "loss": 0.7439,
+      "step": 40621
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.516968560590806,
+      "learning_rate": 9.669729537236327e-08,
+      "loss": 0.5303,
+      "step": 40622
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6211169008185405,
+      "learning_rate": 9.6591467674364e-08,
+      "loss": 0.4981,
+      "step": 40623
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.504891403570259,
+      "learning_rate": 9.648569763714532e-08,
+      "loss": 0.5702,
+      "step": 40624
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4041513322723638,
+      "learning_rate": 9.637998526132231e-08,
+      "loss": 0.6831,
+      "step": 40625
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4390129162470349,
+      "learning_rate": 9.627433054751001e-08,
+      "loss": 0.826,
+      "step": 40626
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.402264284515057,
+      "learning_rate": 9.616873349632461e-08,
+      "loss": 0.7753,
+      "step": 40627
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3571916575261473,
+      "learning_rate": 9.606319410838006e-08,
+      "loss": 0.7012,
+      "step": 40628
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.335434768936815,
+      "learning_rate": 9.595771238429142e-08,
+      "loss": 0.6269,
+      "step": 40629
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7340631895887326,
+      "learning_rate": 9.585228832467153e-08,
+      "loss": 0.5855,
+      "step": 40630
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.389465185492575,
+      "learning_rate": 9.574692193013769e-08,
+      "loss": 0.6291,
+      "step": 40631
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4524091601346951,
+      "learning_rate": 9.564161320129828e-08,
+      "loss": 0.6765,
+      "step": 40632
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6086061061383651,
+      "learning_rate": 9.553636213876949e-08,
+      "loss": 0.5132,
+      "step": 40633
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.922562502704302,
+      "learning_rate": 9.543116874316416e-08,
+      "loss": 0.6768,
+      "step": 40634
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.580890988969071,
+      "learning_rate": 9.532603301509514e-08,
+      "loss": 0.7091,
+      "step": 40635
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6356132664740788,
+      "learning_rate": 9.522095495517303e-08,
+      "loss": 0.4771,
+      "step": 40636
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2862270205494748,
+      "learning_rate": 9.511593456400958e-08,
+      "loss": 0.6616,
+      "step": 40637
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.351464336758542,
+      "learning_rate": 9.501097184221763e-08,
+      "loss": 0.7316,
+      "step": 40638
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5933450636997188,
+      "learning_rate": 9.490606679040782e-08,
+      "loss": 0.4804,
+      "step": 40639
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6141078204908758,
+      "learning_rate": 9.480121940919074e-08,
+      "loss": 0.4974,
+      "step": 40640
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5619729965285839,
+      "learning_rate": 9.469642969917592e-08,
+      "loss": 0.4624,
+      "step": 40641
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0302585736914,
+      "learning_rate": 9.45916976609762e-08,
+      "loss": 0.6644,
+      "step": 40642
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5577095903215309,
+      "learning_rate": 9.448702329519777e-08,
+      "loss": 0.4516,
+      "step": 40643
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.541161712440458,
+      "learning_rate": 9.438240660245346e-08,
+      "loss": 0.8276,
+      "step": 40644
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.22943392893524,
+      "learning_rate": 9.427784758335056e-08,
+      "loss": 0.7509,
+      "step": 40645
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.08086071996922,
+      "learning_rate": 9.41733462384975e-08,
+      "loss": 0.6743,
+      "step": 40646
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6493684001400015,
+      "learning_rate": 9.406890256850376e-08,
+      "loss": 0.7933,
+      "step": 40647
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5573164212741726,
+      "learning_rate": 9.396451657397666e-08,
+      "loss": 0.708,
+      "step": 40648
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6104177024985218,
+      "learning_rate": 9.386018825552568e-08,
+      "loss": 0.712,
+      "step": 40649
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5861140532846365,
+      "learning_rate": 9.37559176137548e-08,
+      "loss": 0.4907,
+      "step": 40650
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5785639653886103,
+      "learning_rate": 9.365170464927464e-08,
+      "loss": 0.6618,
+      "step": 40651
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.954527517962818,
+      "learning_rate": 9.354754936269029e-08,
+      "loss": 0.7952,
+      "step": 40652
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3416494375438535,
+      "learning_rate": 9.344345175460901e-08,
+      "loss": 0.7856,
+      "step": 40653
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.311180086978061,
+      "learning_rate": 9.333941182563477e-08,
+      "loss": 0.6652,
+      "step": 40654
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2985839198957494,
+      "learning_rate": 9.323542957637599e-08,
+      "loss": 0.6356,
+      "step": 40655
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5655474997438372,
+      "learning_rate": 9.313150500743661e-08,
+      "loss": 0.6501,
+      "step": 40656
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3459478251820014,
+      "learning_rate": 9.302763811942283e-08,
+      "loss": 0.7433,
+      "step": 40657
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5431576091509873,
+      "learning_rate": 9.29238289129375e-08,
+      "loss": 0.6325,
+      "step": 40658
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3842152905968195,
+      "learning_rate": 9.282007738858679e-08,
+      "loss": 0.7717,
+      "step": 40659
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3646678827724106,
+      "learning_rate": 9.271638354697354e-08,
+      "loss": 0.673,
+      "step": 40660
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8948472442861841,
+      "learning_rate": 9.261274738870286e-08,
+      "loss": 0.6745,
+      "step": 40661
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2919549439886926,
+      "learning_rate": 9.250916891437756e-08,
+      "loss": 0.6173,
+      "step": 40662
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5585796904575864,
+      "learning_rate": 9.24056481245994e-08,
+      "loss": 0.5962,
+      "step": 40663
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4153098580978714,
+      "learning_rate": 9.230218501997346e-08,
+      "loss": 0.6114,
+      "step": 40664
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5433804190517864,
+      "learning_rate": 9.219877960110036e-08,
+      "loss": 0.6255,
+      "step": 40665
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1661656468043513,
+      "learning_rate": 9.209543186858294e-08,
+      "loss": 0.5195,
+      "step": 40666
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4805481891830434,
+      "learning_rate": 9.199214182302186e-08,
+      "loss": 0.5489,
+      "step": 40667
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4253239196316847,
+      "learning_rate": 9.188890946501993e-08,
+      "loss": 0.7103,
+      "step": 40668
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.352196826235406,
+      "learning_rate": 9.178573479517783e-08,
+      "loss": 0.6645,
+      "step": 40669
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5338607099617294,
+      "learning_rate": 9.168261781409726e-08,
+      "loss": 0.7165,
+      "step": 40670
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6881832660923368,
+      "learning_rate": 9.157955852237554e-08,
+      "loss": 0.6709,
+      "step": 40671
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.314788016176321,
+      "learning_rate": 9.147655692061442e-08,
+      "loss": 0.737,
+      "step": 40672
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4091385167250505,
+      "learning_rate": 9.13736130094145e-08,
+      "loss": 0.7312,
+      "step": 40673
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9925942356514494,
+      "learning_rate": 9.127072678937532e-08,
+      "loss": 0.5666,
+      "step": 40674
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5177129519952093,
+      "learning_rate": 9.116789826109417e-08,
+      "loss": 0.7329,
+      "step": 40675
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.910708726636,
+      "learning_rate": 9.106512742516948e-08,
+      "loss": 0.7512,
+      "step": 40676
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6095794482012159,
+      "learning_rate": 9.096241428220299e-08,
+      "loss": 0.5134,
+      "step": 40677
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3700168814039697,
+      "learning_rate": 9.085975883278864e-08,
+      "loss": 0.6541,
+      "step": 40678
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1762817563984986,
+      "learning_rate": 9.075716107752708e-08,
+      "loss": 0.5547,
+      "step": 40679
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6026773088154949,
+      "learning_rate": 9.065462101701339e-08,
+      "loss": 0.5071,
+      "step": 40680
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6047572536350498,
+      "learning_rate": 9.055213865184708e-08,
+      "loss": 0.6411,
+      "step": 40681
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.409995857616614,
+      "learning_rate": 9.044971398262325e-08,
+      "loss": 0.6083,
+      "step": 40682
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.261282466600762,
+      "learning_rate": 9.034734700993696e-08,
+      "loss": 0.6342,
+      "step": 40683
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3335510705027815,
+      "learning_rate": 9.024503773438776e-08,
+      "loss": 0.6966,
+      "step": 40684
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5496258290670726,
+      "learning_rate": 9.014278615656735e-08,
+      "loss": 0.6628,
+      "step": 40685
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3233687138748518,
+      "learning_rate": 9.004059227707528e-08,
+      "loss": 0.5611,
+      "step": 40686
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4349204713579904,
+      "learning_rate": 8.99384560965022e-08,
+      "loss": 0.7318,
+      "step": 40687
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4644443421036546,
+      "learning_rate": 8.98363776154465e-08,
+      "loss": 0.6276,
+      "step": 40688
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4737604725356568,
+      "learning_rate": 8.973435683449994e-08,
+      "loss": 0.7065,
+      "step": 40689
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5198223900996157,
+      "learning_rate": 8.96323937542587e-08,
+      "loss": 0.6839,
+      "step": 40690
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4391724781651196,
+      "learning_rate": 8.953048837531453e-08,
+      "loss": 0.4956,
+      "step": 40691
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3472216153943521,
+      "learning_rate": 8.94286406982603e-08,
+      "loss": 0.6492,
+      "step": 40692
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4835272178589964,
+      "learning_rate": 8.932685072369107e-08,
+      "loss": 0.7634,
+      "step": 40693
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2713649291777385,
+      "learning_rate": 8.922511845219972e-08,
+      "loss": 0.6576,
+      "step": 40694
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2967634228118696,
+      "learning_rate": 8.912344388437577e-08,
+      "loss": 0.6761,
+      "step": 40695
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5448316258171608,
+      "learning_rate": 8.902182702081318e-08,
+      "loss": 0.6618,
+      "step": 40696
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6341438322442539,
+      "learning_rate": 8.892026786210372e-08,
+      "loss": 0.5744,
+      "step": 40697
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3824261989249502,
+      "learning_rate": 8.881876640883802e-08,
+      "loss": 0.6694,
+      "step": 40698
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5433147862339323,
+      "learning_rate": 8.871732266160782e-08,
+      "loss": 0.672,
+      "step": 40699
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2933253641659106,
+      "learning_rate": 8.861593662100265e-08,
+      "loss": 0.7122,
+      "step": 40700
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4557032825508733,
+      "learning_rate": 8.851460828761316e-08,
+      "loss": 0.5784,
+      "step": 40701
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.06105941334077,
+      "learning_rate": 8.841333766202997e-08,
+      "loss": 0.6838,
+      "step": 40702
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4188810007690333,
+      "learning_rate": 8.831212474484263e-08,
+      "loss": 0.5055,
+      "step": 40703
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7995484985534467,
+      "learning_rate": 8.821096953663955e-08,
+      "loss": 0.7362,
+      "step": 40704
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3704531197185938,
+      "learning_rate": 8.810987203801136e-08,
+      "loss": 0.6146,
+      "step": 40705
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5351722171178455,
+      "learning_rate": 8.800883224954315e-08,
+      "loss": 0.6808,
+      "step": 40706
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4844954911831278,
+      "learning_rate": 8.790785017182779e-08,
+      "loss": 0.7267,
+      "step": 40707
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2341259241241271,
+      "learning_rate": 8.780692580545036e-08,
+      "loss": 0.6555,
+      "step": 40708
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5623504471453666,
+      "learning_rate": 8.770605915099927e-08,
+      "loss": 0.6508,
+      "step": 40709
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6058453449432171,
+      "learning_rate": 8.760525020906074e-08,
+      "loss": 0.5138,
+      "step": 40710
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1552760923787715,
+      "learning_rate": 8.750449898022317e-08,
+      "loss": 0.6347,
+      "step": 40711
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.1804364880133638,
+      "learning_rate": 8.740380546507388e-08,
+      "loss": 0.6636,
+      "step": 40712
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.60618331545993,
+      "learning_rate": 8.730316966419682e-08,
+      "loss": 0.7641,
+      "step": 40713
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5200347253898279,
+      "learning_rate": 8.720259157817823e-08,
+      "loss": 0.7561,
+      "step": 40714
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.696225209625178,
+      "learning_rate": 8.71020712076054e-08,
+      "loss": 0.6855,
+      "step": 40715
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6271498009089944,
+      "learning_rate": 8.70016085530634e-08,
+      "loss": 0.7442,
+      "step": 40716
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5974871026833866,
+      "learning_rate": 8.69012036151351e-08,
+      "loss": 0.7677,
+      "step": 40717
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.173328172641148,
+      "learning_rate": 8.680085639440672e-08,
+      "loss": 0.7371,
+      "step": 40718
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3226868254070256,
+      "learning_rate": 8.670056689146334e-08,
+      "loss": 0.721,
+      "step": 40719
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2976309073391694,
+      "learning_rate": 8.66003351068867e-08,
+      "loss": 0.6815,
+      "step": 40720
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5069045215418224,
+      "learning_rate": 8.650016104126192e-08,
+      "loss": 0.5653,
+      "step": 40721
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3587453972699561,
+      "learning_rate": 8.640004469517072e-08,
+      "loss": 0.6804,
+      "step": 40722
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.306778628302505,
+      "learning_rate": 8.62999860691982e-08,
+      "loss": 0.7181,
+      "step": 40723
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.270507116417434,
+      "learning_rate": 8.619998516392503e-08,
+      "loss": 0.7731,
+      "step": 40724
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5489623046517103,
+      "learning_rate": 8.610004197993516e-08,
+      "loss": 0.5338,
+      "step": 40725
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4892301151951928,
+      "learning_rate": 8.600015651780924e-08,
+      "loss": 0.7663,
+      "step": 40726
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4079712043180315,
+      "learning_rate": 8.590032877812792e-08,
+      "loss": 0.6463,
+      "step": 40727
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3989801523513317,
+      "learning_rate": 8.580055876147408e-08,
+      "loss": 0.6468,
+      "step": 40728
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2283934373499197,
+      "learning_rate": 8.570084646842946e-08,
+      "loss": 0.7213,
+      "step": 40729
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5315556937498724,
+      "learning_rate": 8.56011918995725e-08,
+      "loss": 0.7556,
+      "step": 40730
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.297585206175817,
+      "learning_rate": 8.550159505548383e-08,
+      "loss": 0.597,
+      "step": 40731
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3938048274464372,
+      "learning_rate": 8.540205593674411e-08,
+      "loss": 0.5708,
+      "step": 40732
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4635751544801836,
+      "learning_rate": 8.530257454393287e-08,
+      "loss": 0.6797,
+      "step": 40733
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4918351802648568,
+      "learning_rate": 8.520315087762964e-08,
+      "loss": 0.6239,
+      "step": 40734
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2556237046960912,
+      "learning_rate": 8.510378493841065e-08,
+      "loss": 0.6079,
+      "step": 40735
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4578702028149555,
+      "learning_rate": 8.500447672685764e-08,
+      "loss": 0.7161,
+      "step": 40736
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.377823136703705,
+      "learning_rate": 8.490522624354792e-08,
+      "loss": 0.7508,
+      "step": 40737
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5141593507701843,
+      "learning_rate": 8.48060334890588e-08,
+      "loss": 0.6557,
+      "step": 40738
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.1196596939168892,
+      "learning_rate": 8.470689846396762e-08,
+      "loss": 0.7001,
+      "step": 40739
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4166105089263585,
+      "learning_rate": 8.460782116885169e-08,
+      "loss": 0.7605,
+      "step": 40740
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4730288428933118,
+      "learning_rate": 8.450880160428831e-08,
+      "loss": 0.6278,
+      "step": 40741
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5894572438741041,
+      "learning_rate": 8.440983977085481e-08,
+      "loss": 0.697,
+      "step": 40742
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6054859270458771,
+      "learning_rate": 8.43109356691263e-08,
+      "loss": 0.4568,
+      "step": 40743
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2986722510971596,
+      "learning_rate": 8.421208929967784e-08,
+      "loss": 0.6082,
+      "step": 40744
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3698328459478568,
+      "learning_rate": 8.411330066308676e-08,
+      "loss": 0.7244,
+      "step": 40745
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.869889258785597,
+      "learning_rate": 8.401456975992706e-08,
+      "loss": 0.8008,
+      "step": 40746
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.672575128617589,
+      "learning_rate": 8.391589659077381e-08,
+      "loss": 0.6525,
+      "step": 40747
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.540294018462722,
+      "learning_rate": 8.381728115620213e-08,
+      "loss": 0.6884,
+      "step": 40748
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.309820699021396,
+      "learning_rate": 8.371872345678378e-08,
+      "loss": 0.7305,
+      "step": 40749
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5894303296886039,
+      "learning_rate": 8.362022349309718e-08,
+      "loss": 0.5366,
+      "step": 40750
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4367704083857213,
+      "learning_rate": 8.352178126571186e-08,
+      "loss": 0.66,
+      "step": 40751
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1573558714505179,
+      "learning_rate": 8.342339677520184e-08,
+      "loss": 0.6399,
+      "step": 40752
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5898118811247114,
+      "learning_rate": 8.332507002213996e-08,
+      "loss": 0.5154,
+      "step": 40753
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2907327258643708,
+      "learning_rate": 8.322680100710023e-08,
+      "loss": 0.7403,
+      "step": 40754
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6742921420297407,
+      "learning_rate": 8.31285897306533e-08,
+      "loss": 0.6575,
+      "step": 40755
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6031020509263458,
+      "learning_rate": 8.303043619337093e-08,
+      "loss": 0.5077,
+      "step": 40756
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5933173782922126,
+      "learning_rate": 8.293234039582487e-08,
+      "loss": 0.4326,
+      "step": 40757
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.332008126594674,
+      "learning_rate": 8.283430233858691e-08,
+      "loss": 0.7218,
+      "step": 40758
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6160585888158365,
+      "learning_rate": 8.27363220222277e-08,
+      "loss": 0.6542,
+      "step": 40759
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3324374798178893,
+      "learning_rate": 8.263839944731677e-08,
+      "loss": 0.7084,
+      "step": 40760
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3619048025403357,
+      "learning_rate": 8.254053461442591e-08,
+      "loss": 0.6982,
+      "step": 40761
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.266373105337659,
+      "learning_rate": 8.244272752412241e-08,
+      "loss": 0.6232,
+      "step": 40762
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6365320223485547,
+      "learning_rate": 8.234497817697806e-08,
+      "loss": 0.6655,
+      "step": 40763
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2600260990858476,
+      "learning_rate": 8.224728657356129e-08,
+      "loss": 0.6406,
+      "step": 40764
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6961378258547883,
+      "learning_rate": 8.214965271444053e-08,
+      "loss": 0.5712,
+      "step": 40765
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3454548087446352,
+      "learning_rate": 8.205207660018311e-08,
+      "loss": 0.8171,
+      "step": 40766
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2126677457666093,
+      "learning_rate": 8.195455823136078e-08,
+      "loss": 0.6188,
+      "step": 40767
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8981959237299093,
+      "learning_rate": 8.185709760853754e-08,
+      "loss": 0.5287,
+      "step": 40768
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.472946394160512,
+      "learning_rate": 8.175969473228296e-08,
+      "loss": 0.6323,
+      "step": 40769
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6037402176845038,
+      "learning_rate": 8.16623496031621e-08,
+      "loss": 0.514,
+      "step": 40770
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.445085498282336,
+      "learning_rate": 8.156506222174455e-08,
+      "loss": 0.6483,
+      "step": 40771
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4426019513572943,
+      "learning_rate": 8.146783258859536e-08,
+      "loss": 0.6337,
+      "step": 40772
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.356234134184227,
+      "learning_rate": 8.137066070427968e-08,
+      "loss": 0.6102,
+      "step": 40773
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3729868459284502,
+      "learning_rate": 8.127354656936481e-08,
+      "loss": 0.7362,
+      "step": 40774
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3436669481181793,
+      "learning_rate": 8.117649018441587e-08,
+      "loss": 0.7074,
+      "step": 40775
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3298687512321365,
+      "learning_rate": 8.107949154999684e-08,
+      "loss": 0.6763,
+      "step": 40776
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2606728830645773,
+      "learning_rate": 8.098255066667393e-08,
+      "loss": 0.5701,
+      "step": 40777
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4024149439797762,
+      "learning_rate": 8.088566753501004e-08,
+      "loss": 0.6245,
+      "step": 40778
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.280881757833662,
+      "learning_rate": 8.078884215557137e-08,
+      "loss": 0.6105,
+      "step": 40779
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4042022377383863,
+      "learning_rate": 8.06920745289197e-08,
+      "loss": 0.7565,
+      "step": 40780
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2659311807230362,
+      "learning_rate": 8.059536465561901e-08,
+      "loss": 0.726,
+      "step": 40781
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3545965307906664,
+      "learning_rate": 8.049871253623332e-08,
+      "loss": 0.6026,
+      "step": 40782
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.440701939778928,
+      "learning_rate": 8.040211817132438e-08,
+      "loss": 0.7243,
+      "step": 40783
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3456022129101466,
+      "learning_rate": 8.030558156145507e-08,
+      "loss": 0.627,
+      "step": 40784
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.332546527304889,
+      "learning_rate": 8.020910270718607e-08,
+      "loss": 0.53,
+      "step": 40785
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4605587606753545,
+      "learning_rate": 8.011268160908137e-08,
+      "loss": 0.7169,
+      "step": 40786
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2380610006866364,
+      "learning_rate": 8.00163182676994e-08,
+      "loss": 0.6191,
+      "step": 40787
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3198228009393085,
+      "learning_rate": 7.992001268360528e-08,
+      "loss": 0.6989,
+      "step": 40788
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3112287778636424,
+      "learning_rate": 7.982376485735632e-08,
+      "loss": 0.673,
+      "step": 40789
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9524361646511075,
+      "learning_rate": 7.972757478951432e-08,
+      "loss": 0.677,
+      "step": 40790
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6145835082547304,
+      "learning_rate": 7.963144248063881e-08,
+      "loss": 0.6069,
+      "step": 40791
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3307710877418024,
+      "learning_rate": 7.953536793129046e-08,
+      "loss": 0.6004,
+      "step": 40792
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3693737195932711,
+      "learning_rate": 7.943935114202772e-08,
+      "loss": 0.6187,
+      "step": 40793
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.472474889912337,
+      "learning_rate": 7.934339211340902e-08,
+      "loss": 0.8198,
+      "step": 40794
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2608584057411467,
+      "learning_rate": 7.924749084599393e-08,
+      "loss": 0.571,
+      "step": 40795
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5284219495981692,
+      "learning_rate": 7.915164734034198e-08,
+      "loss": 0.7365,
+      "step": 40796
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6105840729349772,
+      "learning_rate": 7.90558615970094e-08,
+      "loss": 0.4646,
+      "step": 40797
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3027895608512867,
+      "learning_rate": 7.896013361655464e-08,
+      "loss": 0.557,
+      "step": 40798
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5212170638996882,
+      "learning_rate": 7.886446339953502e-08,
+      "loss": 0.7917,
+      "step": 40799
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4240644820934334,
+      "learning_rate": 7.876885094650677e-08,
+      "loss": 0.6715,
+      "step": 40800
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4357078214988523,
+      "learning_rate": 7.867329625802833e-08,
+      "loss": 0.8673,
+      "step": 40801
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3611504705830613,
+      "learning_rate": 7.857779933465482e-08,
+      "loss": 0.6744,
+      "step": 40802
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6034617768622794,
+      "learning_rate": 7.848236017694133e-08,
+      "loss": 0.6118,
+      "step": 40803
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5234501525043247,
+      "learning_rate": 7.838697878544632e-08,
+      "loss": 0.7134,
+      "step": 40804
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3546098586957482,
+      "learning_rate": 7.829165516072268e-08,
+      "loss": 0.6423,
+      "step": 40805
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3011574439362688,
+      "learning_rate": 7.819638930332552e-08,
+      "loss": 0.5994,
+      "step": 40806
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6593790982982949,
+      "learning_rate": 7.810118121381105e-08,
+      "loss": 0.5196,
+      "step": 40807
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4035427589194736,
+      "learning_rate": 7.800603089273218e-08,
+      "loss": 0.6623,
+      "step": 40808
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.344566953215566,
+      "learning_rate": 7.791093834064289e-08,
+      "loss": 0.6054,
+      "step": 40809
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5720597091871945,
+      "learning_rate": 7.781590355809831e-08,
+      "loss": 0.4724,
+      "step": 40810
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5520085171792795,
+      "learning_rate": 7.772092654565133e-08,
+      "loss": 0.6398,
+      "step": 40811
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.936262071980439,
+      "learning_rate": 7.762600730385373e-08,
+      "loss": 0.7461,
+      "step": 40812
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.752733401303991,
+      "learning_rate": 7.753114583325838e-08,
+      "loss": 0.8329,
+      "step": 40813
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4950354745793308,
+      "learning_rate": 7.74363421344193e-08,
+      "loss": 0.6189,
+      "step": 40814
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4006307427732287,
+      "learning_rate": 7.734159620788717e-08,
+      "loss": 0.5999,
+      "step": 40815
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3493973918246316,
+      "learning_rate": 7.724690805421265e-08,
+      "loss": 0.6785,
+      "step": 40816
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4416240049668094,
+      "learning_rate": 7.715227767394862e-08,
+      "loss": 0.7745,
+      "step": 40817
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2683127323040937,
+      "learning_rate": 7.705770506764576e-08,
+      "loss": 0.6068,
+      "step": 40818
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6751720995382018,
+      "learning_rate": 7.696319023585475e-08,
+      "loss": 0.8209,
+      "step": 40819
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2448710618014498,
+      "learning_rate": 7.686873317912514e-08,
+      "loss": 0.6685,
+      "step": 40820
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6867074720115443,
+      "learning_rate": 7.677433389800759e-08,
+      "loss": 0.8021,
+      "step": 40821
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3917629695236922,
+      "learning_rate": 7.667999239305057e-08,
+      "loss": 0.6038,
+      "step": 40822
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3650391201610323,
+      "learning_rate": 7.658570866480586e-08,
+      "loss": 0.6774,
+      "step": 40823
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.233032114965322,
+      "learning_rate": 7.649148271381967e-08,
+      "loss": 0.7174,
+      "step": 40824
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6644176039868956,
+      "learning_rate": 7.63973145406427e-08,
+      "loss": 0.5636,
+      "step": 40825
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5103359325782397,
+      "learning_rate": 7.630320414582115e-08,
+      "loss": 0.6122,
+      "step": 40826
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.32850236466519,
+      "learning_rate": 7.620915152990571e-08,
+      "loss": 0.6818,
+      "step": 40827
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4639983857386445,
+      "learning_rate": 7.61151566934415e-08,
+      "loss": 0.7769,
+      "step": 40828
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2882105553394967,
+      "learning_rate": 7.602121963697695e-08,
+      "loss": 0.7243,
+      "step": 40829
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2496386983515484,
+      "learning_rate": 7.592734036105831e-08,
+      "loss": 0.6742,
+      "step": 40830
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6495256751614067,
+      "learning_rate": 7.583351886623291e-08,
+      "loss": 0.5658,
+      "step": 40831
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4204740020750535,
+      "learning_rate": 7.5739755153047e-08,
+      "loss": 0.7494,
+      "step": 40832
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2333349193876106,
+      "learning_rate": 7.564604922204566e-08,
+      "loss": 0.6798,
+      "step": 40833
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4538027334397028,
+      "learning_rate": 7.555240107377404e-08,
+      "loss": 0.5321,
+      "step": 40834
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4573743134928216,
+      "learning_rate": 7.545881070878059e-08,
+      "loss": 0.7532,
+      "step": 40835
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.826190150293448,
+      "learning_rate": 7.536527812760708e-08,
+      "loss": 0.5729,
+      "step": 40836
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5017871162698724,
+      "learning_rate": 7.527180333079865e-08,
+      "loss": 0.7204,
+      "step": 40837
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8204851938841955,
+      "learning_rate": 7.517838631889929e-08,
+      "loss": 0.7937,
+      "step": 40838
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2670062944109515,
+      "learning_rate": 7.508502709245524e-08,
+      "loss": 0.6596,
+      "step": 40839
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3218446037995615,
+      "learning_rate": 7.499172565200718e-08,
+      "loss": 0.7185,
+      "step": 40840
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5084191956232282,
+      "learning_rate": 7.489848199809913e-08,
+      "loss": 0.7424,
+      "step": 40841
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.479257474897331,
+      "learning_rate": 7.480529613127507e-08,
+      "loss": 0.6934,
+      "step": 40842
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6543652599148937,
+      "learning_rate": 7.471216805207682e-08,
+      "loss": 0.5858,
+      "step": 40843
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8761524838918393,
+      "learning_rate": 7.461909776104725e-08,
+      "loss": 0.684,
+      "step": 40844
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.549672065080334,
+      "learning_rate": 7.452608525872595e-08,
+      "loss": 0.6494,
+      "step": 40845
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.5801748673622593,
+      "learning_rate": 7.443313054565804e-08,
+      "loss": 0.7191,
+      "step": 40846
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6890756910181028,
+      "learning_rate": 7.434023362238086e-08,
+      "loss": 0.6351,
+      "step": 40847
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5901480708987478,
+      "learning_rate": 7.42473944894384e-08,
+      "loss": 0.5207,
+      "step": 40848
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2021645629479263,
+      "learning_rate": 7.415461314737026e-08,
+      "loss": 0.5088,
+      "step": 40849
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2196376540404603,
+      "learning_rate": 7.406188959671601e-08,
+      "loss": 0.6815,
+      "step": 40850
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.331474712414759,
+      "learning_rate": 7.39692238380163e-08,
+      "loss": 0.726,
+      "step": 40851
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8021959231552336,
+      "learning_rate": 7.387661587180961e-08,
+      "loss": 0.7828,
+      "step": 40852
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5994342492176146,
+      "learning_rate": 7.37840656986366e-08,
+      "loss": 0.4846,
+      "step": 40853
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3921004564200368,
+      "learning_rate": 7.369157331903465e-08,
+      "loss": 0.7506,
+      "step": 40854
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6443852746928702,
+      "learning_rate": 7.35991387335433e-08,
+      "loss": 0.6813,
+      "step": 40855
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.037614166626862,
+      "learning_rate": 7.350676194270101e-08,
+      "loss": 0.6934,
+      "step": 40856
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3789099367660793,
+      "learning_rate": 7.341444294704403e-08,
+      "loss": 0.6985,
+      "step": 40857
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5746590026750202,
+      "learning_rate": 7.332218174711081e-08,
+      "loss": 0.599,
+      "step": 40858
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1907318856300197,
+      "learning_rate": 7.322997834343871e-08,
+      "loss": 0.5574,
+      "step": 40859
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8347064323951066,
+      "learning_rate": 7.313783273656506e-08,
+      "loss": 0.7106,
+      "step": 40860
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.106431540635315,
+      "learning_rate": 7.304574492702609e-08,
+      "loss": 0.6039,
+      "step": 40861
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6183205383628152,
+      "learning_rate": 7.295371491535697e-08,
+      "loss": 0.5261,
+      "step": 40862
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9811361088737216,
+      "learning_rate": 7.28617427020939e-08,
+      "loss": 0.6672,
+      "step": 40863
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4159754719981146,
+      "learning_rate": 7.276982828777312e-08,
+      "loss": 0.6005,
+      "step": 40864
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3482618001772073,
+      "learning_rate": 7.267797167292867e-08,
+      "loss": 0.5634,
+      "step": 40865
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7642332570937436,
+      "learning_rate": 7.258617285809677e-08,
+      "loss": 0.6767,
+      "step": 40866
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4106423001900574,
+      "learning_rate": 7.249443184381033e-08,
+      "loss": 0.5391,
+      "step": 40867
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3736789372914204,
+      "learning_rate": 7.240274863060448e-08,
+      "loss": 0.7647,
+      "step": 40868
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3902714677065575,
+      "learning_rate": 7.231112321901435e-08,
+      "loss": 0.6757,
+      "step": 40869
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3232924542983684,
+      "learning_rate": 7.221955560957062e-08,
+      "loss": 0.7218,
+      "step": 40870
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5355929093675273,
+      "learning_rate": 7.21280458028073e-08,
+      "loss": 0.7758,
+      "step": 40871
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3121854301670686,
+      "learning_rate": 7.203659379925843e-08,
+      "loss": 0.7266,
+      "step": 40872
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6218940805805422,
+      "learning_rate": 7.194519959945467e-08,
+      "loss": 0.4707,
+      "step": 40873
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2696971259850403,
+      "learning_rate": 7.185386320393006e-08,
+      "loss": 0.633,
+      "step": 40874
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.607683930424156,
+      "learning_rate": 7.176258461321528e-08,
+      "loss": 0.5095,
+      "step": 40875
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.709055412868157,
+      "learning_rate": 7.167136382784212e-08,
+      "loss": 0.7021,
+      "step": 40876
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4268287949093557,
+      "learning_rate": 7.158020084834128e-08,
+      "loss": 0.6364,
+      "step": 40877
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.355174240173331,
+      "learning_rate": 7.148909567524343e-08,
+      "loss": 0.6424,
+      "step": 40878
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.67568155440368,
+      "learning_rate": 7.139804830908037e-08,
+      "loss": 0.5848,
+      "step": 40879
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4404843237585137,
+      "learning_rate": 7.130705875037946e-08,
+      "loss": 0.6794,
+      "step": 40880
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.278391097863252,
+      "learning_rate": 7.121612699967362e-08,
+      "loss": 0.5918,
+      "step": 40881
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.403241227413712,
+      "learning_rate": 7.112525305749129e-08,
+      "loss": 0.6566,
+      "step": 40882
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1755070244770922,
+      "learning_rate": 7.103443692435985e-08,
+      "loss": 0.5709,
+      "step": 40883
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3306302247032094,
+      "learning_rate": 7.094367860081108e-08,
+      "loss": 0.6054,
+      "step": 40884
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3984764288978664,
+      "learning_rate": 7.085297808737013e-08,
+      "loss": 0.7489,
+      "step": 40885
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3907079162010556,
+      "learning_rate": 7.076233538456767e-08,
+      "loss": 0.7451,
+      "step": 40886
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7557863228747008,
+      "learning_rate": 7.067175049292884e-08,
+      "loss": 0.6761,
+      "step": 40887
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3700619320410412,
+      "learning_rate": 7.058122341298435e-08,
+      "loss": 0.6178,
+      "step": 40888
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9681299315957825,
+      "learning_rate": 7.049075414525818e-08,
+      "loss": 0.7251,
+      "step": 40889
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4229705910958188,
+      "learning_rate": 7.040034269027995e-08,
+      "loss": 0.6827,
+      "step": 40890
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4745540032881392,
+      "learning_rate": 7.030998904857478e-08,
+      "loss": 0.6957,
+      "step": 40891
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.765824382152911,
+      "learning_rate": 7.021969322066668e-08,
+      "loss": 0.6488,
+      "step": 40892
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.565415540619443,
+      "learning_rate": 7.012945520708414e-08,
+      "loss": 0.657,
+      "step": 40893
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.376930511012871,
+      "learning_rate": 7.003927500835118e-08,
+      "loss": 0.6469,
+      "step": 40894
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4700409023956704,
+      "learning_rate": 6.994915262499513e-08,
+      "loss": 0.6937,
+      "step": 40895
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5097418325351943,
+      "learning_rate": 6.985908805753672e-08,
+      "loss": 0.719,
+      "step": 40896
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6219910294571123,
+      "learning_rate": 6.976908130650329e-08,
+      "loss": 0.4725,
+      "step": 40897
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4123460305496955,
+      "learning_rate": 6.967913237241885e-08,
+      "loss": 0.7511,
+      "step": 40898
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.376654609944393,
+      "learning_rate": 6.958924125580635e-08,
+      "loss": 0.6816,
+      "step": 40899
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.0728413003388986,
+      "learning_rate": 6.949940795718868e-08,
+      "loss": 0.7236,
+      "step": 40900
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.349537109981111,
+      "learning_rate": 6.940963247708877e-08,
+      "loss": 0.679,
+      "step": 40901
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4388694733197254,
+      "learning_rate": 6.931991481603062e-08,
+      "loss": 0.5853,
+      "step": 40902
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.394032962545202,
+      "learning_rate": 6.923025497453606e-08,
+      "loss": 0.6997,
+      "step": 40903
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4340987990324159,
+      "learning_rate": 6.914065295312689e-08,
+      "loss": 0.6815,
+      "step": 40904
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.489407605665985,
+      "learning_rate": 6.90511087523249e-08,
+      "loss": 0.7487,
+      "step": 40905
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.680455947911285,
+      "learning_rate": 6.896162237265192e-08,
+      "loss": 0.8358,
+      "step": 40906
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3141851925941896,
+      "learning_rate": 6.887219381462751e-08,
+      "loss": 0.6455,
+      "step": 40907
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.381734381196193,
+      "learning_rate": 6.878282307877349e-08,
+      "loss": 0.665,
+      "step": 40908
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6007247967564768,
+      "learning_rate": 6.869351016561055e-08,
+      "loss": 0.5094,
+      "step": 40909
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4261153617888973,
+      "learning_rate": 6.860425507565826e-08,
+      "loss": 0.6025,
+      "step": 40910
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2807592762638562,
+      "learning_rate": 6.851505780943512e-08,
+      "loss": 0.7107,
+      "step": 40911
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2751940607069014,
+      "learning_rate": 6.842591836746293e-08,
+      "loss": 0.6108,
+      "step": 40912
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9316792915744039,
+      "learning_rate": 6.833683675025904e-08,
+      "loss": 0.6371,
+      "step": 40913
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.6776722051936717,
+      "learning_rate": 6.824781295834193e-08,
+      "loss": 0.7415,
+      "step": 40914
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4897880642158037,
+      "learning_rate": 6.815884699223007e-08,
+      "loss": 0.6966,
+      "step": 40915
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9879739164968644,
+      "learning_rate": 6.806993885244307e-08,
+      "loss": 0.7477,
+      "step": 40916
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3654739304769836,
+      "learning_rate": 6.798108853949603e-08,
+      "loss": 0.7457,
+      "step": 40917
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4271404951683642,
+      "learning_rate": 6.789229605390856e-08,
+      "loss": 0.6221,
+      "step": 40918
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4542704812686449,
+      "learning_rate": 6.780356139619582e-08,
+      "loss": 0.7156,
+      "step": 40919
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2740406076215836,
+      "learning_rate": 6.771488456687514e-08,
+      "loss": 0.7254,
+      "step": 40920
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3491692832122988,
+      "learning_rate": 6.76262655664639e-08,
+      "loss": 0.7826,
+      "step": 40921
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3588397062135227,
+      "learning_rate": 6.753770439547614e-08,
+      "loss": 0.6721,
+      "step": 40922
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4768101970640348,
+      "learning_rate": 6.74492010544281e-08,
+      "loss": 0.7708,
+      "step": 40923
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.537545663066458,
+      "learning_rate": 6.736075554383604e-08,
+      "loss": 0.7476,
+      "step": 40924
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8440131567208473,
+      "learning_rate": 6.727236786421509e-08,
+      "loss": 0.6462,
+      "step": 40925
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3060593779515013,
+      "learning_rate": 6.718403801607709e-08,
+      "loss": 0.683,
+      "step": 40926
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6578181943678505,
+      "learning_rate": 6.70957659999405e-08,
+      "loss": 0.6129,
+      "step": 40927
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6029504484553216,
+      "learning_rate": 6.700755181631491e-08,
+      "loss": 0.5905,
+      "step": 40928
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2916875078992283,
+      "learning_rate": 6.691939546571769e-08,
+      "loss": 0.7321,
+      "step": 40929
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5998587570506781,
+      "learning_rate": 6.683129694866063e-08,
+      "loss": 0.5022,
+      "step": 40930
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3391316021138244,
+      "learning_rate": 6.674325626565558e-08,
+      "loss": 0.6848,
+      "step": 40931
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.37817137820826,
+      "learning_rate": 6.665527341721544e-08,
+      "loss": 0.713,
+      "step": 40932
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4308677516134125,
+      "learning_rate": 6.656734840385425e-08,
+      "loss": 0.6703,
+      "step": 40933
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4611719226753017,
+      "learning_rate": 6.647948122608383e-08,
+      "loss": 0.6748,
+      "step": 40934
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6225770156420416,
+      "learning_rate": 6.639167188441264e-08,
+      "loss": 0.6959,
+      "step": 40935
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3897486023020464,
+      "learning_rate": 6.630392037935474e-08,
+      "loss": 0.6167,
+      "step": 40936
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5138424671376327,
+      "learning_rate": 6.621622671142191e-08,
+      "loss": 0.7631,
+      "step": 40937
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2855816181060298,
+      "learning_rate": 6.612859088112155e-08,
+      "loss": 0.6894,
+      "step": 40938
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.085469964326424,
+      "learning_rate": 6.604101288896548e-08,
+      "loss": 0.551,
+      "step": 40939
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.8230539023128927,
+      "learning_rate": 6.595349273546437e-08,
+      "loss": 0.6867,
+      "step": 40940
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5274784142413937,
+      "learning_rate": 6.586603042112671e-08,
+      "loss": 0.7467,
+      "step": 40941
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.649224379061799,
+      "learning_rate": 6.577862594646323e-08,
+      "loss": 0.7158,
+      "step": 40942
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2750165961518014,
+      "learning_rate": 6.569127931198016e-08,
+      "loss": 0.6518,
+      "step": 40943
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4361330269133659,
+      "learning_rate": 6.560399051818822e-08,
+      "loss": 0.7026,
+      "step": 40944
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4458854749767236,
+      "learning_rate": 6.551675956559478e-08,
+      "loss": 0.6677,
+      "step": 40945
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.72664068470683,
+      "learning_rate": 6.542958645470832e-08,
+      "loss": 0.6496,
+      "step": 40946
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.734519250741556,
+      "learning_rate": 6.534247118603621e-08,
+      "loss": 0.7294,
+      "step": 40947
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0954166403366803,
+      "learning_rate": 6.525541376008471e-08,
+      "loss": 0.804,
+      "step": 40948
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.59127603003134,
+      "learning_rate": 6.51684141773623e-08,
+      "loss": 0.6229,
+      "step": 40949
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.310939503866309,
+      "learning_rate": 6.508147243837415e-08,
+      "loss": 0.718,
+      "step": 40950
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2046756598388553,
+      "learning_rate": 6.499458854362761e-08,
+      "loss": 0.5347,
+      "step": 40951
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5073442918465314,
+      "learning_rate": 6.490776249362785e-08,
+      "loss": 0.635,
+      "step": 40952
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3961454288796902,
+      "learning_rate": 6.482099428888e-08,
+      "loss": 0.6199,
+      "step": 40953
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6874541151997744,
+      "learning_rate": 6.473428392989034e-08,
+      "loss": 0.6814,
+      "step": 40954
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.599989793454991,
+      "learning_rate": 6.46476314171629e-08,
+      "loss": 0.4848,
+      "step": 40955
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5621313448416432,
+      "learning_rate": 6.456103675120173e-08,
+      "loss": 0.8485,
+      "step": 40956
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0713891814530703,
+      "learning_rate": 6.447449993251197e-08,
+      "loss": 0.6879,
+      "step": 40957
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0090242367382225,
+      "learning_rate": 6.438802096159769e-08,
+      "loss": 0.6894,
+      "step": 40958
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6077444406619675,
+      "learning_rate": 6.430159983896179e-08,
+      "loss": 0.518,
+      "step": 40959
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3150363754294832,
+      "learning_rate": 6.42152365651072e-08,
+      "loss": 0.6848,
+      "step": 40960
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5052360065988024,
+      "learning_rate": 6.412893114053798e-08,
+      "loss": 0.5902,
+      "step": 40961
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4251834600801714,
+      "learning_rate": 6.404268356575483e-08,
+      "loss": 0.6697,
+      "step": 40962
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3076731766522152,
+      "learning_rate": 6.395649384126179e-08,
+      "loss": 0.6449,
+      "step": 40963
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.32156215615219,
+      "learning_rate": 6.387036196755847e-08,
+      "loss": 0.7947,
+      "step": 40964
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.612642888324432,
+      "learning_rate": 6.378428794514891e-08,
+      "loss": 0.6077,
+      "step": 40965
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.392163891938235,
+      "learning_rate": 6.369827177453158e-08,
+      "loss": 0.7181,
+      "step": 40966
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2721565696398216,
+      "learning_rate": 6.361231345621055e-08,
+      "loss": 0.6,
+      "step": 40967
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4753003053434761,
+      "learning_rate": 6.352641299068318e-08,
+      "loss": 0.6123,
+      "step": 40968
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4490269440396606,
+      "learning_rate": 6.344057037845131e-08,
+      "loss": 0.723,
+      "step": 40969
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5378823677046878,
+      "learning_rate": 6.33547856200134e-08,
+      "loss": 0.6859,
+      "step": 40970
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3080468005782429,
+      "learning_rate": 6.32690587158713e-08,
+      "loss": 0.5952,
+      "step": 40971
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6718200658100786,
+      "learning_rate": 6.318338966652127e-08,
+      "loss": 0.7005,
+      "step": 40972
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4509730270199919,
+      "learning_rate": 6.309777847246401e-08,
+      "loss": 0.7612,
+      "step": 40973
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.472166584715659,
+      "learning_rate": 6.301222513419692e-08,
+      "loss": 0.7366,
+      "step": 40974
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.347818704793774,
+      "learning_rate": 6.292672965221958e-08,
+      "loss": 0.7899,
+      "step": 40975
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.667519612467245,
+      "learning_rate": 6.284129202702938e-08,
+      "loss": 0.7147,
+      "step": 40976
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6337940523488609,
+      "learning_rate": 6.275591225912148e-08,
+      "loss": 0.5154,
+      "step": 40977
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2946884519718602,
+      "learning_rate": 6.267059034899547e-08,
+      "loss": 0.6804,
+      "step": 40978
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4450998446644752,
+      "learning_rate": 6.258532629714765e-08,
+      "loss": 0.6547,
+      "step": 40979
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5078448133622577,
+      "learning_rate": 6.250012010407425e-08,
+      "loss": 0.7014,
+      "step": 40980
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2254197017426005,
+      "learning_rate": 6.241497177027156e-08,
+      "loss": 0.5966,
+      "step": 40981
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.382468372435955,
+      "learning_rate": 6.232988129623475e-08,
+      "loss": 0.6697,
+      "step": 40982
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6320797211212248,
+      "learning_rate": 6.224484868245895e-08,
+      "loss": 0.6378,
+      "step": 40983
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.614089876732504,
+      "learning_rate": 6.215987392944045e-08,
+      "loss": 0.5497,
+      "step": 40984
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.529095230344931,
+      "learning_rate": 6.20749570376733e-08,
+      "loss": 0.6236,
+      "step": 40985
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4395040480189998,
+      "learning_rate": 6.199009800765265e-08,
+      "loss": 0.6069,
+      "step": 40986
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.631966260387594,
+      "learning_rate": 6.190529683987145e-08,
+      "loss": 0.6726,
+      "step": 40987
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4090221643552505,
+      "learning_rate": 6.182055353482375e-08,
+      "loss": 0.7746,
+      "step": 40988
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4196203539405323,
+      "learning_rate": 6.173586809300358e-08,
+      "loss": 0.6171,
+      "step": 40989
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8669511160378234,
+      "learning_rate": 6.16512405149039e-08,
+      "loss": 0.7164,
+      "step": 40990
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6035987027838146,
+      "learning_rate": 6.156667080101652e-08,
+      "loss": 0.6511,
+      "step": 40991
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.280430281959538,
+      "learning_rate": 6.14821589518344e-08,
+      "loss": 0.6272,
+      "step": 40992
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4973002781151137,
+      "learning_rate": 6.139770496785047e-08,
+      "loss": 0.7893,
+      "step": 40993
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.263412591347037,
+      "learning_rate": 6.131330884955433e-08,
+      "loss": 0.6877,
+      "step": 40994
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.458635300339612,
+      "learning_rate": 6.122897059743894e-08,
+      "loss": 0.6503,
+      "step": 40995
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3136166815354857,
+      "learning_rate": 6.114469021199498e-08,
+      "loss": 0.6385,
+      "step": 40996
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3627687596739007,
+      "learning_rate": 6.106046769371432e-08,
+      "loss": 0.7262,
+      "step": 40997
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4901626500010527,
+      "learning_rate": 6.097630304308543e-08,
+      "loss": 0.6248,
+      "step": 40998
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4958441694222442,
+      "learning_rate": 6.089219626059905e-08,
+      "loss": 0.7281,
+      "step": 40999
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0755812817836587,
+      "learning_rate": 6.080814734674478e-08,
+      "loss": 0.6583,
+      "step": 41000
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.27115480456392,
+      "learning_rate": 6.072415630201333e-08,
+      "loss": 0.5756,
+      "step": 41001
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.318180203558888,
+      "learning_rate": 6.0640223126891e-08,
+      "loss": 0.7375,
+      "step": 41002
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.209271897045659,
+      "learning_rate": 6.055634782186848e-08,
+      "loss": 0.5717,
+      "step": 41003
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.413390329713024,
+      "learning_rate": 6.047253038743428e-08,
+      "loss": 0.5623,
+      "step": 41004
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6053236127913638,
+      "learning_rate": 6.03887708240758e-08,
+      "loss": 0.5239,
+      "step": 41005
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.363984977703026,
+      "learning_rate": 6.03050691322804e-08,
+      "loss": 0.6669,
+      "step": 41006
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2985216564795115,
+      "learning_rate": 6.02214253125366e-08,
+      "loss": 0.6606,
+      "step": 41007
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4462705344626252,
+      "learning_rate": 6.013783936532846e-08,
+      "loss": 0.7348,
+      "step": 41008
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3370497503665097,
+      "learning_rate": 6.00543112911467e-08,
+      "loss": 0.7594,
+      "step": 41009
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.517326404640066,
+      "learning_rate": 5.997084109047424e-08,
+      "loss": 0.7163,
+      "step": 41010
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4201488999922711,
+      "learning_rate": 5.98874287637996e-08,
+      "loss": 0.5875,
+      "step": 41011
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.437986423122067,
+      "learning_rate": 5.980407431160684e-08,
+      "loss": 0.6415,
+      "step": 41012
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0514246698252063,
+      "learning_rate": 5.972077773438223e-08,
+      "loss": 0.748,
+      "step": 41013
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4385513618351657,
+      "learning_rate": 5.963753903260983e-08,
+      "loss": 0.619,
+      "step": 41014
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3636670458661557,
+      "learning_rate": 5.95543582067748e-08,
+      "loss": 0.6652,
+      "step": 41015
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.426960456002484,
+      "learning_rate": 5.94712352573612e-08,
+      "loss": 0.5689,
+      "step": 41016
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2319971426280725,
+      "learning_rate": 5.938817018485421e-08,
+      "loss": 0.7727,
+      "step": 41017
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2919374952851266,
+      "learning_rate": 5.930516298973565e-08,
+      "loss": 0.6407,
+      "step": 41018
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5029487745172627,
+      "learning_rate": 5.922221367248959e-08,
+      "loss": 0.7,
+      "step": 41019
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3943984454565073,
+      "learning_rate": 5.913932223359897e-08,
+      "loss": 0.6371,
+      "step": 41020
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.379328913446627,
+      "learning_rate": 5.905648867354563e-08,
+      "loss": 0.6663,
+      "step": 41021
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6192215560021661,
+      "learning_rate": 5.897371299281473e-08,
+      "loss": 0.525,
+      "step": 41022
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4160457052538165,
+      "learning_rate": 5.8890995191884795e-08,
+      "loss": 0.6955,
+      "step": 41023
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.342080457033897,
+      "learning_rate": 5.880833527123875e-08,
+      "loss": 0.5838,
+      "step": 41024
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5678953387542175,
+      "learning_rate": 5.872573323135844e-08,
+      "loss": 0.4781,
+      "step": 41025
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4231717153003167,
+      "learning_rate": 5.8643189072723485e-08,
+      "loss": 0.72,
+      "step": 41026
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2581033055695987,
+      "learning_rate": 5.856070279581572e-08,
+      "loss": 0.7126,
+      "step": 41027
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2918248254018088,
+      "learning_rate": 5.847827440111475e-08,
+      "loss": 0.6829,
+      "step": 41028
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.312059432956962,
+      "learning_rate": 5.839590388910132e-08,
+      "loss": 0.7711,
+      "step": 41029
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4272285317440967,
+      "learning_rate": 5.831359126025393e-08,
+      "loss": 0.698,
+      "step": 41030
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.288603826051597,
+      "learning_rate": 5.8231336515052196e-08,
+      "loss": 0.6734,
+      "step": 41031
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4085571538091095,
+      "learning_rate": 5.814913965397573e-08,
+      "loss": 0.7557,
+      "step": 41032
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5094181873794548,
+      "learning_rate": 5.806700067750193e-08,
+      "loss": 0.6676,
+      "step": 41033
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.349826755342944,
+      "learning_rate": 5.79849195861093e-08,
+      "loss": 0.7285,
+      "step": 41034
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5716726856835143,
+      "learning_rate": 5.7902896380276354e-08,
+      "loss": 0.5076,
+      "step": 41035
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.278924570617986,
+      "learning_rate": 5.782093106048159e-08,
+      "loss": 0.5777,
+      "step": 41036
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7084110334792277,
+      "learning_rate": 5.7739023627199074e-08,
+      "loss": 0.6613,
+      "step": 41037
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5095623007426606,
+      "learning_rate": 5.765717408090843e-08,
+      "loss": 0.6093,
+      "step": 41038
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2547969633128147,
+      "learning_rate": 5.7575382422085935e-08,
+      "loss": 0.6115,
+      "step": 41039
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.580875688129798,
+      "learning_rate": 5.749364865120677e-08,
+      "loss": 0.5147,
+      "step": 41040
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4254404530375349,
+      "learning_rate": 5.7411972768747213e-08,
+      "loss": 0.6153,
+      "step": 41041
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4244640621053581,
+      "learning_rate": 5.733035477518245e-08,
+      "loss": 0.7126,
+      "step": 41042
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6936674190705772,
+      "learning_rate": 5.7248794670989874e-08,
+      "loss": 0.7259,
+      "step": 41043
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2340235905717596,
+      "learning_rate": 5.7167292456641324e-08,
+      "loss": 0.6508,
+      "step": 41044
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2491016668804094,
+      "learning_rate": 5.708584813261308e-08,
+      "loss": 0.6182,
+      "step": 41045
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3695752893244677,
+      "learning_rate": 5.700446169937812e-08,
+      "loss": 0.6627,
+      "step": 41046
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5333893661495812,
+      "learning_rate": 5.69231331574116e-08,
+      "loss": 0.6642,
+      "step": 41047
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3415953450100349,
+      "learning_rate": 5.6841862507186484e-08,
+      "loss": 0.6776,
+      "step": 41048
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1241220857327174,
+      "learning_rate": 5.6760649749175726e-08,
+      "loss": 0.6652,
+      "step": 41049
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.059195941027738,
+      "learning_rate": 5.667949488385338e-08,
+      "loss": 0.6801,
+      "step": 41050
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3198783209513898,
+      "learning_rate": 5.6598397911690194e-08,
+      "loss": 0.6728,
+      "step": 41051
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3197915071810598,
+      "learning_rate": 5.651735883316023e-08,
+      "loss": 0.5914,
+      "step": 41052
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2219612455260174,
+      "learning_rate": 5.643637764873311e-08,
+      "loss": 0.6359,
+      "step": 41053
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.425887191993288,
+      "learning_rate": 5.635545435888179e-08,
+      "loss": 0.6746,
+      "step": 41054
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3191068717004075,
+      "learning_rate": 5.6274588964077006e-08,
+      "loss": 0.6522,
+      "step": 41055
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.694594142366939,
+      "learning_rate": 5.619378146478949e-08,
+      "loss": 0.6542,
+      "step": 41056
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3830797355959221,
+      "learning_rate": 5.6113031861491086e-08,
+      "loss": 0.6936,
+      "step": 41057
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3065509877656296,
+      "learning_rate": 5.60323401546492e-08,
+      "loss": 0.6249,
+      "step": 41058
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.425698281758012,
+      "learning_rate": 5.595170634473568e-08,
+      "loss": 0.5961,
+      "step": 41059
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6606059355058647,
+      "learning_rate": 5.587113043222015e-08,
+      "loss": 0.7001,
+      "step": 41060
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3715148053982955,
+      "learning_rate": 5.579061241757111e-08,
+      "loss": 0.6712,
+      "step": 41061
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3624240630988513,
+      "learning_rate": 5.571015230125709e-08,
+      "loss": 0.6387,
+      "step": 41062
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6030789608162571,
+      "learning_rate": 5.5629750083746606e-08,
+      "loss": 0.4754,
+      "step": 41063
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5852474197518067,
+      "learning_rate": 5.5549405765509265e-08,
+      "loss": 0.7185,
+      "step": 41064
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3089577717448369,
+      "learning_rate": 5.546911934701027e-08,
+      "loss": 0.6802,
+      "step": 41065
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3000316714108395,
+      "learning_rate": 5.538889082871923e-08,
+      "loss": 0.5584,
+      "step": 41066
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3118293841396607,
+      "learning_rate": 5.5308720211102454e-08,
+      "loss": 0.7389,
+      "step": 41067
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3961590027416704,
+      "learning_rate": 5.522860749462622e-08,
+      "loss": 0.6471,
+      "step": 41068
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4763811541103717,
+      "learning_rate": 5.514855267975794e-08,
+      "loss": 0.743,
+      "step": 41069
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4986837821991505,
+      "learning_rate": 5.506855576696279e-08,
+      "loss": 0.6355,
+      "step": 41070
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.675817739785081,
+      "learning_rate": 5.498861675670819e-08,
+      "loss": 0.6019,
+      "step": 41071
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.782276009672224,
+      "learning_rate": 5.490873564945709e-08,
+      "loss": 0.6542,
+      "step": 41072
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5964952581028526,
+      "learning_rate": 5.482891244567689e-08,
+      "loss": 0.4842,
+      "step": 41073
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2114693182071659,
+      "learning_rate": 5.474914714583057e-08,
+      "loss": 0.6971,
+      "step": 41074
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6033755197744057,
+      "learning_rate": 5.46694397503833e-08,
+      "loss": 0.5075,
+      "step": 41075
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.599095434927562,
+      "learning_rate": 5.458979025979916e-08,
+      "loss": 0.7755,
+      "step": 41076
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.370720447981113,
+      "learning_rate": 5.451019867454221e-08,
+      "loss": 0.7854,
+      "step": 41077
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.6956704656066393,
+      "learning_rate": 5.4430664995076543e-08,
+      "loss": 0.5132,
+      "step": 41078
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.487053539364202,
+      "learning_rate": 5.435118922186289e-08,
+      "loss": 0.5487,
+      "step": 41079
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.42362829809805,
+      "learning_rate": 5.427177135536532e-08,
+      "loss": 0.5953,
+      "step": 41080
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4325949560037547,
+      "learning_rate": 5.4192411396046806e-08,
+      "loss": 0.7545,
+      "step": 41081
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2750664918620678,
+      "learning_rate": 5.41131093443692e-08,
+      "loss": 0.5714,
+      "step": 41082
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3234855830392427,
+      "learning_rate": 5.4033865200793236e-08,
+      "loss": 0.6433,
+      "step": 41083
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1804698318879085,
+      "learning_rate": 5.395467896578077e-08,
+      "loss": 0.6192,
+      "step": 41084
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3385772360896544,
+      "learning_rate": 5.387555063979366e-08,
+      "loss": 0.6808,
+      "step": 41085
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.631585071939967,
+      "learning_rate": 5.3796480223292643e-08,
+      "loss": 0.4986,
+      "step": 41086
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3269063021777177,
+      "learning_rate": 5.371746771673625e-08,
+      "loss": 0.6961,
+      "step": 41087
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.384761475535311,
+      "learning_rate": 5.363851312058632e-08,
+      "loss": 0.65,
+      "step": 41088
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2769942670554628,
+      "learning_rate": 5.355961643530139e-08,
+      "loss": 0.7783,
+      "step": 41089
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.624920771710859,
+      "learning_rate": 5.3480777661341077e-08,
+      "loss": 0.7738,
+      "step": 41090
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3123344805474897,
+      "learning_rate": 5.340199679916502e-08,
+      "loss": 0.6023,
+      "step": 41091
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.192694468671203,
+      "learning_rate": 5.332327384923175e-08,
+      "loss": 0.6173,
+      "step": 41092
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6636845527882647,
+      "learning_rate": 5.3244608811999776e-08,
+      "loss": 0.6032,
+      "step": 41093
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4741638058709943,
+      "learning_rate": 5.3166001687926515e-08,
+      "loss": 0.6869,
+      "step": 41094
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.378300162747052,
+      "learning_rate": 5.308745247747049e-08,
+      "loss": 0.7107,
+      "step": 41095
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5850285405975488,
+      "learning_rate": 5.3008961181086895e-08,
+      "loss": 0.6344,
+      "step": 41096
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4405768083242982,
+      "learning_rate": 5.293052779923535e-08,
+      "loss": 0.6829,
+      "step": 41097
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2590002041072061,
+      "learning_rate": 5.285215233237107e-08,
+      "loss": 0.5991,
+      "step": 41098
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3598068050294572,
+      "learning_rate": 5.277383478095144e-08,
+      "loss": 0.8405,
+      "step": 41099
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.271160787600719,
+      "learning_rate": 5.2695575145431664e-08,
+      "loss": 0.5907,
+      "step": 41100
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2598418657548527,
+      "learning_rate": 5.261737342626694e-08,
+      "loss": 0.6941,
+      "step": 41101
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4409131709619714,
+      "learning_rate": 5.253922962391356e-08,
+      "loss": 0.5991,
+      "step": 41102
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5189966228643335,
+      "learning_rate": 5.2461143738825605e-08,
+      "loss": 0.6167,
+      "step": 41103
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6511207049827856,
+      "learning_rate": 5.238311577145827e-08,
+      "loss": 0.6792,
+      "step": 41104
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7462703254490684,
+      "learning_rate": 5.230514572226675e-08,
+      "loss": 0.7463,
+      "step": 41105
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.375571360358334,
+      "learning_rate": 5.22272335917029e-08,
+      "loss": 0.5291,
+      "step": 41106
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4208442389664875,
+      "learning_rate": 5.214937938022191e-08,
+      "loss": 0.814,
+      "step": 41107
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2700035422778693,
+      "learning_rate": 5.2071583088276755e-08,
+      "loss": 0.6437,
+      "step": 41108
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5211480682621301,
+      "learning_rate": 5.1993844716321515e-08,
+      "loss": 0.6997,
+      "step": 41109
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.307315482323232,
+      "learning_rate": 5.191616426480584e-08,
+      "loss": 0.5383,
+      "step": 41110
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.8740976743794504,
+      "learning_rate": 5.183854173418379e-08,
+      "loss": 0.6769,
+      "step": 41111
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4163458784127598,
+      "learning_rate": 5.1760977124908354e-08,
+      "loss": 0.6457,
+      "step": 41112
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.7587859684264298,
+      "learning_rate": 5.168347043742916e-08,
+      "loss": 0.6179,
+      "step": 41113
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6202391879785165,
+      "learning_rate": 5.160602167219919e-08,
+      "loss": 0.5326,
+      "step": 41114
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3761612144581987,
+      "learning_rate": 5.152863082966808e-08,
+      "loss": 0.6771,
+      "step": 41115
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7194947607792523,
+      "learning_rate": 5.145129791028658e-08,
+      "loss": 0.7475,
+      "step": 41116
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4009752332788132,
+      "learning_rate": 5.137402291450655e-08,
+      "loss": 0.6671,
+      "step": 41117
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2863061354616994,
+      "learning_rate": 5.129680584277541e-08,
+      "loss": 0.7159,
+      "step": 41118
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5352380712710563,
+      "learning_rate": 5.1219646695543915e-08,
+      "loss": 0.7014,
+      "step": 41119
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7567365461673439,
+      "learning_rate": 5.1142545473261694e-08,
+      "loss": 0.6337,
+      "step": 41120
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.467727450586604,
+      "learning_rate": 5.10655021763784e-08,
+      "loss": 0.6481,
+      "step": 41121
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3303401049491408,
+      "learning_rate": 5.0988516805340337e-08,
+      "loss": 0.8158,
+      "step": 41122
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4851010877582977,
+      "learning_rate": 5.091158936059715e-08,
+      "loss": 0.716,
+      "step": 41123
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5476908181809005,
+      "learning_rate": 5.0834719842596246e-08,
+      "loss": 0.4967,
+      "step": 41124
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.406709444150739,
+      "learning_rate": 5.075790825178617e-08,
+      "loss": 0.6839,
+      "step": 41125
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8797569760557897,
+      "learning_rate": 5.0681154588612114e-08,
+      "loss": 0.8192,
+      "step": 41126
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4087142966958477,
+      "learning_rate": 5.0604458853523716e-08,
+      "loss": 0.8395,
+      "step": 41127
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.963306313143388,
+      "learning_rate": 5.052782104696508e-08,
+      "loss": 0.7066,
+      "step": 41128
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4744198540569582,
+      "learning_rate": 5.045124116938471e-08,
+      "loss": 0.7032,
+      "step": 41129
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4660649731096749,
+      "learning_rate": 5.037471922122561e-08,
+      "loss": 0.6518,
+      "step": 41130
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4161349687968476,
+      "learning_rate": 5.02982552029363e-08,
+      "loss": 0.6511,
+      "step": 41131
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5806536873480683,
+      "learning_rate": 5.022184911495864e-08,
+      "loss": 0.6308,
+      "step": 41132
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4529728501973762,
+      "learning_rate": 5.0145500957741176e-08,
+      "loss": 0.6984,
+      "step": 41133
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6312846545696196,
+      "learning_rate": 5.006921073172688e-08,
+      "loss": 0.4623,
+      "step": 41134
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4258460067101821,
+      "learning_rate": 4.999297843735873e-08,
+      "loss": 0.7388,
+      "step": 41135
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.329836050729755,
+      "learning_rate": 4.991680407508193e-08,
+      "loss": 0.6509,
+      "step": 41136
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.670550817951338,
+      "learning_rate": 4.9840687645339446e-08,
+      "loss": 0.5908,
+      "step": 41137
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2349306046640613,
+      "learning_rate": 4.976462914857538e-08,
+      "loss": 0.6017,
+      "step": 41138
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6276692566052307,
+      "learning_rate": 4.968862858523049e-08,
+      "loss": 0.5062,
+      "step": 41139
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6494856444664776,
+      "learning_rate": 4.961268595574886e-08,
+      "loss": 0.6001,
+      "step": 41140
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2766273971636661,
+      "learning_rate": 4.9536801260573475e-08,
+      "loss": 0.737,
+      "step": 41141
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4502190288738148,
+      "learning_rate": 4.946097450014509e-08,
+      "loss": 0.516,
+      "step": 41142
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6451245623386406,
+      "learning_rate": 4.9385205674904455e-08,
+      "loss": 0.7716,
+      "step": 41143
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.653895346158309,
+      "learning_rate": 4.930949478529346e-08,
+      "loss": 0.6551,
+      "step": 41144
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2317723418478927,
+      "learning_rate": 4.923384183175284e-08,
+      "loss": 0.5337,
+      "step": 41145
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6185856732091546,
+      "learning_rate": 4.9158246814723367e-08,
+      "loss": 0.482,
+      "step": 41146
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.339853251074849,
+      "learning_rate": 4.90827097346458e-08,
+      "loss": 0.6727,
+      "step": 41147
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9775485392773162,
+      "learning_rate": 4.9007230591957555e-08,
+      "loss": 0.654,
+      "step": 41148
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3276052151621898,
+      "learning_rate": 4.8931809387100515e-08,
+      "loss": 0.7138,
+      "step": 41149
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7904666464190686,
+      "learning_rate": 4.8856446120513215e-08,
+      "loss": 0.647,
+      "step": 41150
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5825049653639354,
+      "learning_rate": 4.8781140792633074e-08,
+      "loss": 0.4571,
+      "step": 41151
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5978711927520652,
+      "learning_rate": 4.870589340390086e-08,
+      "loss": 0.6132,
+      "step": 41152
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.433557940022118,
+      "learning_rate": 4.863070395475289e-08,
+      "loss": 0.5714,
+      "step": 41153
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.339337122904654,
+      "learning_rate": 4.85555724456277e-08,
+      "loss": 0.6664,
+      "step": 41154
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4631744591046008,
+      "learning_rate": 4.848049887696271e-08,
+      "loss": 0.5777,
+      "step": 41155
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6157711166415032,
+      "learning_rate": 4.8405483249194255e-08,
+      "loss": 0.4892,
+      "step": 41156
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6311346652120606,
+      "learning_rate": 4.8330525562759745e-08,
+      "loss": 0.516,
+      "step": 41157
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.5449129317945127,
+      "learning_rate": 4.8255625818096616e-08,
+      "loss": 0.6404,
+      "step": 41158
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.37644678854104,
+      "learning_rate": 4.818078401563897e-08,
+      "loss": 0.5722,
+      "step": 41159
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3230923777671966,
+      "learning_rate": 4.810600015582312e-08,
+      "loss": 0.6821,
+      "step": 41160
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4057321519183583,
+      "learning_rate": 4.803127423908538e-08,
+      "loss": 0.7214,
+      "step": 41161
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5553774029313177,
+      "learning_rate": 4.795660626586096e-08,
+      "loss": 0.4677,
+      "step": 41162
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3853076944793168,
+      "learning_rate": 4.788199623658285e-08,
+      "loss": 0.5892,
+      "step": 41163
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6381136499090447,
+      "learning_rate": 4.7807444151687363e-08,
+      "loss": 0.5684,
+      "step": 41164
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3710915703001352,
+      "learning_rate": 4.773295001160638e-08,
+      "loss": 0.8547,
+      "step": 41165
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5553640624661556,
+      "learning_rate": 4.765851381677622e-08,
+      "loss": 0.4457,
+      "step": 41166
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.014197656354625,
+      "learning_rate": 4.758413556762986e-08,
+      "loss": 0.5787,
+      "step": 41167
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.500937520823884,
+      "learning_rate": 4.750981526459808e-08,
+      "loss": 0.7898,
+      "step": 41168
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3314785003399263,
+      "learning_rate": 4.7435552908114966e-08,
+      "loss": 0.5873,
+      "step": 41169
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3110015345369979,
+      "learning_rate": 4.7361348498613516e-08,
+      "loss": 0.6194,
+      "step": 41170
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.291838796822663,
+      "learning_rate": 4.728720203652559e-08,
+      "loss": 0.7282,
+      "step": 41171
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4796910883046237,
+      "learning_rate": 4.721311352228197e-08,
+      "loss": 0.7536,
+      "step": 41172
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.453945143923415,
+      "learning_rate": 4.713908295631564e-08,
+      "loss": 0.6797,
+      "step": 41173
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.320396972201018,
+      "learning_rate": 4.706511033905514e-08,
+      "loss": 0.7157,
+      "step": 41174
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5912335147346208,
+      "learning_rate": 4.6991195670932355e-08,
+      "loss": 0.676,
+      "step": 41175
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4203873283686725,
+      "learning_rate": 4.6917338952380266e-08,
+      "loss": 0.7531,
+      "step": 41176
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5348955604311239,
+      "learning_rate": 4.684354018382409e-08,
+      "loss": 0.6295,
+      "step": 41177
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.476840506159206,
+      "learning_rate": 4.6769799365696814e-08,
+      "loss": 0.5917,
+      "step": 41178
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.38959187117432,
+      "learning_rate": 4.669611649842698e-08,
+      "loss": 0.7586,
+      "step": 41179
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3817896261159095,
+      "learning_rate": 4.662249158244425e-08,
+      "loss": 0.6285,
+      "step": 41180
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2891570617840777,
+      "learning_rate": 4.6548924618176064e-08,
+      "loss": 0.6338,
+      "step": 41181
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5573195016229708,
+      "learning_rate": 4.6475415606049845e-08,
+      "loss": 0.8362,
+      "step": 41182
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7696698457906486,
+      "learning_rate": 4.640196454649637e-08,
+      "loss": 0.7495,
+      "step": 41183
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.236146649906886,
+      "learning_rate": 4.632857143994196e-08,
+      "loss": 0.657,
+      "step": 41184
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2345392912387576,
+      "learning_rate": 4.625523628681405e-08,
+      "loss": 0.5585,
+      "step": 41185
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2996721783368224,
+      "learning_rate": 4.6181959087538975e-08,
+      "loss": 0.7016,
+      "step": 41186
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.532646920464141,
+      "learning_rate": 4.610873984254416e-08,
+      "loss": 0.6675,
+      "step": 41187
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4641721704612425,
+      "learning_rate": 4.603557855225482e-08,
+      "loss": 0.5985,
+      "step": 41188
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4487000618254937,
+      "learning_rate": 4.59624752170984e-08,
+      "loss": 0.7054,
+      "step": 41189
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5584314259955104,
+      "learning_rate": 4.588942983750011e-08,
+      "loss": 0.4503,
+      "step": 41190
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5970658356498677,
+      "learning_rate": 4.5816442413884055e-08,
+      "loss": 0.6557,
+      "step": 41191
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4819062299476244,
+      "learning_rate": 4.574351294667656e-08,
+      "loss": 0.7423,
+      "step": 41192
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3414625067758246,
+      "learning_rate": 4.5670641436301735e-08,
+      "loss": 0.5687,
+      "step": 41193
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4134092980453903,
+      "learning_rate": 4.559782788318479e-08,
+      "loss": 0.6912,
+      "step": 41194
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5549424648900168,
+      "learning_rate": 4.552507228774761e-08,
+      "loss": 0.5577,
+      "step": 41195
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7072132101624848,
+      "learning_rate": 4.545237465041541e-08,
+      "loss": 0.7462,
+      "step": 41196
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.445500254950745,
+      "learning_rate": 4.5379734971611186e-08,
+      "loss": 0.6523,
+      "step": 41197
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5674724606882535,
+      "learning_rate": 4.530715325175794e-08,
+      "loss": 0.6022,
+      "step": 41198
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4643046758177536,
+      "learning_rate": 4.5234629491277546e-08,
+      "loss": 0.5725,
+      "step": 41199
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3607487918715355,
+      "learning_rate": 4.516216369059301e-08,
+      "loss": 0.699,
+      "step": 41200
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5734888215855387,
+      "learning_rate": 4.5089755850126206e-08,
+      "loss": 0.8367,
+      "step": 41201
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5269019310550074,
+      "learning_rate": 4.501740597029791e-08,
+      "loss": 0.7527,
+      "step": 41202
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7871612561898425,
+      "learning_rate": 4.494511405153002e-08,
+      "loss": 0.6334,
+      "step": 41203
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3565864539121013,
+      "learning_rate": 4.4872880094244395e-08,
+      "loss": 0.7152,
+      "step": 41204
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4173874739841728,
+      "learning_rate": 4.4800704098859614e-08,
+      "loss": 0.5901,
+      "step": 41205
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6566784208122176,
+      "learning_rate": 4.472858606579755e-08,
+      "loss": 0.5423,
+      "step": 41206
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.557066439404057,
+      "learning_rate": 4.465652599547676e-08,
+      "loss": 0.7242,
+      "step": 41207
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8273822029376028,
+      "learning_rate": 4.4584523888318023e-08,
+      "loss": 0.5476,
+      "step": 41208
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6131941979450608,
+      "learning_rate": 4.451257974473988e-08,
+      "loss": 0.8271,
+      "step": 41209
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.614086207988787,
+      "learning_rate": 4.4440693565160895e-08,
+      "loss": 0.7036,
+      "step": 41210
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3184502959909359,
+      "learning_rate": 4.4368865350000736e-08,
+      "loss": 0.6995,
+      "step": 41211
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.362336011303442,
+      "learning_rate": 4.4297095099676836e-08,
+      "loss": 0.6655,
+      "step": 41212
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3963581754066647,
+      "learning_rate": 4.4225382814606643e-08,
+      "loss": 0.6757,
+      "step": 41213
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4761630737060165,
+      "learning_rate": 4.415372849520871e-08,
+      "loss": 0.4802,
+      "step": 41214
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3805522079125387,
+      "learning_rate": 4.408213214189938e-08,
+      "loss": 0.6654,
+      "step": 41215
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.411910007616743,
+      "learning_rate": 4.4010593755096086e-08,
+      "loss": 0.713,
+      "step": 41216
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3200992045120035,
+      "learning_rate": 4.393911333521517e-08,
+      "loss": 0.5657,
+      "step": 41217
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3573803353836094,
+      "learning_rate": 4.386769088267295e-08,
+      "loss": 0.7567,
+      "step": 41218
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3830311734564313,
+      "learning_rate": 4.379632639788467e-08,
+      "loss": 0.7312,
+      "step": 41219
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6565387801865021,
+      "learning_rate": 4.372501988126665e-08,
+      "loss": 0.7108,
+      "step": 41220
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.352879665366106,
+      "learning_rate": 4.365377133323301e-08,
+      "loss": 0.6372,
+      "step": 41221
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6266672729707365,
+      "learning_rate": 4.3582580754200074e-08,
+      "loss": 0.4883,
+      "step": 41222
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4473465411108883,
+      "learning_rate": 4.3511448144580857e-08,
+      "loss": 0.5555,
+      "step": 41223
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.538856616797437,
+      "learning_rate": 4.3440373504790576e-08,
+      "loss": 0.6871,
+      "step": 41224
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4255137057799696,
+      "learning_rate": 4.336935683524224e-08,
+      "loss": 0.6881,
+      "step": 41225
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.151897600949847,
+      "learning_rate": 4.3298398136349954e-08,
+      "loss": 0.7195,
+      "step": 41226
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3140004333487068,
+      "learning_rate": 4.322749740852672e-08,
+      "loss": 0.6105,
+      "step": 41227
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4875155885137488,
+      "learning_rate": 4.315665465218555e-08,
+      "loss": 0.6482,
+      "step": 41228
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3895479305490637,
+      "learning_rate": 4.308586986773833e-08,
+      "loss": 0.7662,
+      "step": 41229
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5921853473992588,
+      "learning_rate": 4.3015143055596954e-08,
+      "loss": 0.5052,
+      "step": 41230
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4229570716957078,
+      "learning_rate": 4.2944474216174425e-08,
+      "loss": 0.7658,
+      "step": 41231
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2890068681251354,
+      "learning_rate": 4.287386334988153e-08,
+      "loss": 0.6571,
+      "step": 41232
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3706363638440633,
+      "learning_rate": 4.2803310457129046e-08,
+      "loss": 0.667,
+      "step": 41233
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5207830757894842,
+      "learning_rate": 4.273281553832775e-08,
+      "loss": 0.601,
+      "step": 41234
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7613521639774405,
+      "learning_rate": 4.266237859388955e-08,
+      "loss": 0.6637,
+      "step": 41235
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.252846804465183,
+      "learning_rate": 4.2591999624223e-08,
+      "loss": 0.584,
+      "step": 41236
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.420116644947613,
+      "learning_rate": 4.2521678629737774e-08,
+      "loss": 0.59,
+      "step": 41237
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.594184502711359,
+      "learning_rate": 4.245141561084354e-08,
+      "loss": 0.4946,
+      "step": 41238
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5737083634919125,
+      "learning_rate": 4.238121056794997e-08,
+      "loss": 0.5638,
+      "step": 41239
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.358571749552112,
+      "learning_rate": 4.231106350146563e-08,
+      "loss": 0.6822,
+      "step": 41240
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3246709552067988,
+      "learning_rate": 4.224097441179908e-08,
+      "loss": 0.6365,
+      "step": 41241
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1958948960871485,
+      "learning_rate": 4.217094329935778e-08,
+      "loss": 0.6323,
+      "step": 41242
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5638043072411412,
+      "learning_rate": 4.210097016455028e-08,
+      "loss": 0.6678,
+      "step": 41243
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5834487223372926,
+      "learning_rate": 4.2031055007784036e-08,
+      "loss": 0.5481,
+      "step": 41244
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4976283644726036,
+      "learning_rate": 4.196119782946429e-08,
+      "loss": 0.6409,
+      "step": 41245
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2967664084254247,
+      "learning_rate": 4.1891398629999583e-08,
+      "loss": 0.6309,
+      "step": 41246
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6762473183980264,
+      "learning_rate": 4.182165740979627e-08,
+      "loss": 0.7461,
+      "step": 41247
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4468781903186,
+      "learning_rate": 4.175197416925958e-08,
+      "loss": 0.6497,
+      "step": 41248
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4734268677745428,
+      "learning_rate": 4.168234890879586e-08,
+      "loss": 0.7829,
+      "step": 41249
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2880917144266075,
+      "learning_rate": 4.161278162881033e-08,
+      "loss": 0.5936,
+      "step": 41250
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3128732086316144,
+      "learning_rate": 4.1543272329707116e-08,
+      "loss": 0.7169,
+      "step": 41251
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.453945490580287,
+      "learning_rate": 4.1473821011892565e-08,
+      "loss": 0.7921,
+      "step": 41252
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5482885789109501,
+      "learning_rate": 4.14044276757708e-08,
+      "loss": 0.4561,
+      "step": 41253
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4923468721248474,
+      "learning_rate": 4.133509232174482e-08,
+      "loss": 0.6848,
+      "step": 41254
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.353269716052535,
+      "learning_rate": 4.126581495021764e-08,
+      "loss": 0.7502,
+      "step": 41255
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.332965667976691,
+      "learning_rate": 4.119659556159561e-08,
+      "loss": 0.5591,
+      "step": 41256
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3043274407716985,
+      "learning_rate": 4.11274341562784e-08,
+      "loss": 0.5844,
+      "step": 41257
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4124024171593639,
+      "learning_rate": 4.105833073467125e-08,
+      "loss": 0.6643,
+      "step": 41258
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4713784410755968,
+      "learning_rate": 4.0989285297174944e-08,
+      "loss": 0.7846,
+      "step": 41259
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3860485626703334,
+      "learning_rate": 4.092029784419138e-08,
+      "loss": 0.633,
+      "step": 41260
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.422418006121018,
+      "learning_rate": 4.085136837612358e-08,
+      "loss": 0.6206,
+      "step": 41261
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4039650542101432,
+      "learning_rate": 4.0782496893372327e-08,
+      "loss": 0.6963,
+      "step": 41262
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2921885272098774,
+      "learning_rate": 4.071368339633841e-08,
+      "loss": 0.6917,
+      "step": 41263
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.521213554495566,
+      "learning_rate": 4.064492788542152e-08,
+      "loss": 0.6089,
+      "step": 41264
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.690366861039752,
+      "learning_rate": 4.057623036102465e-08,
+      "loss": 0.5886,
+      "step": 41265
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0878338473887132,
+      "learning_rate": 4.0507590823544164e-08,
+      "loss": 0.6759,
+      "step": 41266
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3630963855298714,
+      "learning_rate": 4.043900927338307e-08,
+      "loss": 0.6607,
+      "step": 41267
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9948245489381702,
+      "learning_rate": 4.037048571093771e-08,
+      "loss": 0.7767,
+      "step": 41268
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2111492050875263,
+      "learning_rate": 4.0302020136609995e-08,
+      "loss": 0.6862,
+      "step": 41269
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4157571172803418,
+      "learning_rate": 4.0233612550796273e-08,
+      "loss": 0.586,
+      "step": 41270
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.380273468579263,
+      "learning_rate": 4.0165262953896225e-08,
+      "loss": 0.7367,
+      "step": 41271
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6054690100011729,
+      "learning_rate": 4.00969713463073e-08,
+      "loss": 0.4525,
+      "step": 41272
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2663067938061316,
+      "learning_rate": 4.002873772842586e-08,
+      "loss": 0.8019,
+      "step": 41273
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2865189979814906,
+      "learning_rate": 3.996056210065158e-08,
+      "loss": 0.6065,
+      "step": 41274
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6336375340264482,
+      "learning_rate": 3.989244446338081e-08,
+      "loss": 0.7914,
+      "step": 41275
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6359843100871356,
+      "learning_rate": 3.982438481700768e-08,
+      "loss": 0.7753,
+      "step": 41276
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6210494508457105,
+      "learning_rate": 3.975638316193076e-08,
+      "loss": 0.6377,
+      "step": 41277
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2621098057410274,
+      "learning_rate": 3.96884394985475e-08,
+      "loss": 0.6625,
+      "step": 41278
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7090038532185885,
+      "learning_rate": 3.962055382724983e-08,
+      "loss": 0.7624,
+      "step": 41279
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2559256963937937,
+      "learning_rate": 3.955272614843408e-08,
+      "loss": 0.6571,
+      "step": 41280
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2798438269680517,
+      "learning_rate": 3.948495646249662e-08,
+      "loss": 0.573,
+      "step": 41281
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3580944337313046,
+      "learning_rate": 3.941724476983155e-08,
+      "loss": 0.6188,
+      "step": 41282
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2498364646463844,
+      "learning_rate": 3.934959107083303e-08,
+      "loss": 0.6701,
+      "step": 41283
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5886489036104008,
+      "learning_rate": 3.9281995365892944e-08,
+      "loss": 0.5126,
+      "step": 41284
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2895819418252183,
+      "learning_rate": 3.921445765540877e-08,
+      "loss": 0.5803,
+      "step": 41285
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4416670950403094,
+      "learning_rate": 3.914697793977018e-08,
+      "loss": 0.6131,
+      "step": 41286
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.572882526814829,
+      "learning_rate": 3.9079556219371317e-08,
+      "loss": 0.7103,
+      "step": 41287
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.646821736117765,
+      "learning_rate": 3.9012192494605197e-08,
+      "loss": 0.6943,
+      "step": 41288
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3960739303885041,
+      "learning_rate": 3.894488676586372e-08,
+      "loss": 0.6031,
+      "step": 41289
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5083135910808505,
+      "learning_rate": 3.8877639033538805e-08,
+      "loss": 0.6701,
+      "step": 41290
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9925515649427348,
+      "learning_rate": 3.881044929802236e-08,
+      "loss": 0.646,
+      "step": 41291
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4075299859434065,
+      "learning_rate": 3.8743317559705174e-08,
+      "loss": 0.7351,
+      "step": 41292
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2409623835260695,
+      "learning_rate": 3.867624381897694e-08,
+      "loss": 0.5809,
+      "step": 41293
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.747382754382346,
+      "learning_rate": 3.8609228076229576e-08,
+      "loss": 0.6471,
+      "step": 41294
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4438868628850325,
+      "learning_rate": 3.854227033185387e-08,
+      "loss": 0.7074,
+      "step": 41295
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3794540568289855,
+      "learning_rate": 3.84753705862384e-08,
+      "loss": 0.5019,
+      "step": 41296
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.132106036864744,
+      "learning_rate": 3.840852883977286e-08,
+      "loss": 0.6652,
+      "step": 41297
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.476336440387976,
+      "learning_rate": 3.8341745092846936e-08,
+      "loss": 0.7951,
+      "step": 41298
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6929174382799534,
+      "learning_rate": 3.827501934584921e-08,
+      "loss": 0.6586,
+      "step": 41299
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3581575413759797,
+      "learning_rate": 3.8208351599167135e-08,
+      "loss": 0.661,
+      "step": 41300
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3373985009205238,
+      "learning_rate": 3.814174185319153e-08,
+      "loss": 0.6756,
+      "step": 41301
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6794836676695553,
+      "learning_rate": 3.8075190108307624e-08,
+      "loss": 0.647,
+      "step": 41302
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4542763236647736,
+      "learning_rate": 3.8008696364904006e-08,
+      "loss": 0.6316,
+      "step": 41303
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2758181989235313,
+      "learning_rate": 3.7942260623367035e-08,
+      "loss": 0.6582,
+      "step": 41304
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6463777307145584,
+      "learning_rate": 3.787588288408528e-08,
+      "loss": 0.4626,
+      "step": 41305
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2782211145010933,
+      "learning_rate": 3.7809563147443993e-08,
+      "loss": 0.6377,
+      "step": 41306
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4402447861563539,
+      "learning_rate": 3.7743301413828424e-08,
+      "loss": 0.6198,
+      "step": 41307
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6258622400690756,
+      "learning_rate": 3.767709768362604e-08,
+      "loss": 0.5155,
+      "step": 41308
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3747334578066186,
+      "learning_rate": 3.7610951957220975e-08,
+      "loss": 0.725,
+      "step": 41309
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.31292357888037,
+      "learning_rate": 3.754486423499959e-08,
+      "loss": 0.7217,
+      "step": 41310
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6118994735023643,
+      "learning_rate": 3.747883451734602e-08,
+      "loss": 0.5158,
+      "step": 41311
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6827128419162123,
+      "learning_rate": 3.7412862804645514e-08,
+      "loss": 0.7644,
+      "step": 41312
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2928752645907842,
+      "learning_rate": 3.7346949097279984e-08,
+      "loss": 0.6575,
+      "step": 41313
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4659244167040297,
+      "learning_rate": 3.728109339563579e-08,
+      "loss": 0.7135,
+      "step": 41314
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6242755303309335,
+      "learning_rate": 3.721529570009486e-08,
+      "loss": 0.7406,
+      "step": 41315
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.3826083378002116,
+      "learning_rate": 3.71495560110402e-08,
+      "loss": 0.7403,
+      "step": 41316
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4391888885817974,
+      "learning_rate": 3.7083874328855965e-08,
+      "loss": 0.7066,
+      "step": 41317
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2097420672888974,
+      "learning_rate": 3.701825065392184e-08,
+      "loss": 0.8527,
+      "step": 41318
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.962860582279537,
+      "learning_rate": 3.6952684986623076e-08,
+      "loss": 0.6831,
+      "step": 41319
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.939325283297011,
+      "learning_rate": 3.6887177327340486e-08,
+      "loss": 0.557,
+      "step": 41320
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3857685749999913,
+      "learning_rate": 3.682172767645375e-08,
+      "loss": 0.5992,
+      "step": 41321
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5080115934596625,
+      "learning_rate": 3.675633603434592e-08,
+      "loss": 0.6422,
+      "step": 41322
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.7033420349848782,
+      "learning_rate": 3.6691002401396666e-08,
+      "loss": 0.6353,
+      "step": 41323
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2829947460736393,
+      "learning_rate": 3.662572677798682e-08,
+      "loss": 0.5922,
+      "step": 41324
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.357874260126691,
+      "learning_rate": 3.656050916449716e-08,
+      "loss": 0.8926,
+      "step": 41325
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3848599653044265,
+      "learning_rate": 3.649534956130518e-08,
+      "loss": 0.6762,
+      "step": 41326
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3775939694436692,
+      "learning_rate": 3.64302479687928e-08,
+      "loss": 0.6397,
+      "step": 41327
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6157423587440487,
+      "learning_rate": 3.636520438733748e-08,
+      "loss": 0.4407,
+      "step": 41328
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2599864125258853,
+      "learning_rate": 3.6300218817317825e-08,
+      "loss": 0.6754,
+      "step": 41329
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.578034665779675,
+      "learning_rate": 3.6235291259113516e-08,
+      "loss": 0.5179,
+      "step": 41330
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3246050006573773,
+      "learning_rate": 3.6170421713102034e-08,
+      "loss": 0.636,
+      "step": 41331
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6579604224165216,
+      "learning_rate": 3.6105610179659745e-08,
+      "loss": 0.7421,
+      "step": 41332
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4483140146613087,
+      "learning_rate": 3.6040856659166345e-08,
+      "loss": 0.6089,
+      "step": 41333
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3379501506931248,
+      "learning_rate": 3.597616115199709e-08,
+      "loss": 0.7066,
+      "step": 41334
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3672452402268873,
+      "learning_rate": 3.5911523658528347e-08,
+      "loss": 0.6235,
+      "step": 41335
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3001990678996809,
+      "learning_rate": 3.58469441791387e-08,
+      "loss": 0.5707,
+      "step": 41336
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.54187252534797,
+      "learning_rate": 3.578242271420229e-08,
+      "loss": 0.5807,
+      "step": 41337
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4791194034989807,
+      "learning_rate": 3.571795926409549e-08,
+      "loss": 0.6867,
+      "step": 41338
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6069322451994078,
+      "learning_rate": 3.5653553829192444e-08,
+      "loss": 0.5041,
+      "step": 41339
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3965963266984058,
+      "learning_rate": 3.558920640986952e-08,
+      "loss": 0.5958,
+      "step": 41340
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.633050311765,
+      "learning_rate": 3.5524917006500844e-08,
+      "loss": 0.7058,
+      "step": 41341
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3244767742484065,
+      "learning_rate": 3.546068561946059e-08,
+      "loss": 0.6035,
+      "step": 41342
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.338522121883851,
+      "learning_rate": 3.5396512249123996e-08,
+      "loss": 0.6578,
+      "step": 41343
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4161462533339715,
+      "learning_rate": 3.533239689586188e-08,
+      "loss": 0.7384,
+      "step": 41344
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.725712661387112,
+      "learning_rate": 3.52683395600506e-08,
+      "loss": 0.7307,
+      "step": 41345
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.416445435783723,
+      "learning_rate": 3.5204340242062096e-08,
+      "loss": 0.7307,
+      "step": 41346
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3341928266675773,
+      "learning_rate": 3.5140398942268286e-08,
+      "loss": 0.6391,
+      "step": 41347
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3489206537119596,
+      "learning_rate": 3.50765156610422e-08,
+      "loss": 0.6565,
+      "step": 41348
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.349510210204941,
+      "learning_rate": 3.5012690398754654e-08,
+      "loss": 0.7711,
+      "step": 41349
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2985948616896552,
+      "learning_rate": 3.494892315577869e-08,
+      "loss": 0.5525,
+      "step": 41350
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6225531301222034,
+      "learning_rate": 3.488521393248401e-08,
+      "loss": 0.664,
+      "step": 41351
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.639492557598591,
+      "learning_rate": 3.4821562729243644e-08,
+      "loss": 0.5873,
+      "step": 41352
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.306844022036194,
+      "learning_rate": 3.475796954642729e-08,
+      "loss": 0.7146,
+      "step": 41353
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2637310879483934,
+      "learning_rate": 3.469443438440467e-08,
+      "loss": 0.603,
+      "step": 41354
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3240088635992342,
+      "learning_rate": 3.463095724354548e-08,
+      "loss": 0.712,
+      "step": 41355
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.320777315860763,
+      "learning_rate": 3.456753812421943e-08,
+      "loss": 0.6954,
+      "step": 41356
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.127449525962458,
+      "learning_rate": 3.450417702679731e-08,
+      "loss": 0.5889,
+      "step": 41357
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.311032429773099,
+      "learning_rate": 3.444087395164664e-08,
+      "loss": 0.4831,
+      "step": 41358
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.537230653644474,
+      "learning_rate": 3.437762889913598e-08,
+      "loss": 0.688,
+      "step": 41359
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4056220322481108,
+      "learning_rate": 3.4314441869633954e-08,
+      "loss": 0.6697,
+      "step": 41360
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2750931866776114,
+      "learning_rate": 3.425131286350802e-08,
+      "loss": 0.794,
+      "step": 41361
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2311589899156614,
+      "learning_rate": 3.418824188112679e-08,
+      "loss": 0.6236,
+      "step": 41362
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6153852277826336,
+      "learning_rate": 3.412522892285663e-08,
+      "loss": 0.4843,
+      "step": 41363
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6411224262759428,
+      "learning_rate": 3.4062273989063923e-08,
+      "loss": 0.5199,
+      "step": 41364
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6097200100219233,
+      "learning_rate": 3.3999377080115026e-08,
+      "loss": 0.6447,
+      "step": 41365
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6176147516172381,
+      "learning_rate": 3.393653819637854e-08,
+      "loss": 0.7951,
+      "step": 41366
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.321428663564778,
+      "learning_rate": 3.387375733821863e-08,
+      "loss": 0.6466,
+      "step": 41367
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.287405754910756,
+      "learning_rate": 3.381103450600054e-08,
+      "loss": 0.5044,
+      "step": 41368
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.234837002810925,
+      "learning_rate": 3.3748369700089547e-08,
+      "loss": 0.6636,
+      "step": 41369
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6162847017507451,
+      "learning_rate": 3.368576292085091e-08,
+      "loss": 0.7166,
+      "step": 41370
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.511916109225175,
+      "learning_rate": 3.362321416864878e-08,
+      "loss": 0.7362,
+      "step": 41371
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.305089557484737,
+      "learning_rate": 3.356072344384842e-08,
+      "loss": 0.8726,
+      "step": 41372
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.640759028146463,
+      "learning_rate": 3.3498290746812876e-08,
+      "loss": 0.6839,
+      "step": 41373
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.281373220249513,
+      "learning_rate": 3.3435916077905196e-08,
+      "loss": 0.7017,
+      "step": 41374
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6508152501356992,
+      "learning_rate": 3.3373599437488414e-08,
+      "loss": 0.5968,
+      "step": 41375
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4065220300249741,
+      "learning_rate": 3.331134082592669e-08,
+      "loss": 0.6848,
+      "step": 41376
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.355010797680975,
+      "learning_rate": 3.3249140243581947e-08,
+      "loss": 0.6403,
+      "step": 41377
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5674861847782222,
+      "learning_rate": 3.318699769081501e-08,
+      "loss": 0.6779,
+      "step": 41378
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7565630194173683,
+      "learning_rate": 3.312491316798894e-08,
+      "loss": 0.7189,
+      "step": 41379
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.0628725739351736,
+      "learning_rate": 3.3062886675466754e-08,
+      "loss": 0.713,
+      "step": 41380
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.317082708686471,
+      "learning_rate": 3.300091821360596e-08,
+      "loss": 0.6956,
+      "step": 41381
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.303435215609403,
+      "learning_rate": 3.29390077827707e-08,
+      "loss": 0.737,
+      "step": 41382
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5576108253451464,
+      "learning_rate": 3.287715538331959e-08,
+      "loss": 0.7023,
+      "step": 41383
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6309594271875774,
+      "learning_rate": 3.281536101561233e-08,
+      "loss": 0.7201,
+      "step": 41384
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5688796140581842,
+      "learning_rate": 3.2753624680010864e-08,
+      "loss": 0.492,
+      "step": 41385
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2370861109453657,
+      "learning_rate": 3.2691946376872674e-08,
+      "loss": 0.6276,
+      "step": 41386
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3610947595691538,
+      "learning_rate": 3.2630326106558584e-08,
+      "loss": 0.7094,
+      "step": 41387
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6461386159358672,
+      "learning_rate": 3.256876386942498e-08,
+      "loss": 0.5279,
+      "step": 41388
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5259478728444256,
+      "learning_rate": 3.250725966583268e-08,
+      "loss": 0.5575,
+      "step": 41389
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3517421753458707,
+      "learning_rate": 3.244581349613807e-08,
+      "loss": 0.7116,
+      "step": 41390
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3222865444250462,
+      "learning_rate": 3.238442536069974e-08,
+      "loss": 0.6366,
+      "step": 41391
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7451833156075385,
+      "learning_rate": 3.2323095259875205e-08,
+      "loss": 0.6981,
+      "step": 41392
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5930082675683708,
+      "learning_rate": 3.226182319402194e-08,
+      "loss": 0.4944,
+      "step": 41393
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.20718045860676,
+      "learning_rate": 3.220060916349521e-08,
+      "loss": 0.7597,
+      "step": 41394
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.450762000598406,
+      "learning_rate": 3.213945316865252e-08,
+      "loss": 0.5963,
+      "step": 41395
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2991320853891886,
+      "learning_rate": 3.2078355209849145e-08,
+      "loss": 0.554,
+      "step": 41396
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9683020303291694,
+      "learning_rate": 3.201731528744145e-08,
+      "loss": 0.69,
+      "step": 41397
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3133623418946574,
+      "learning_rate": 3.195633340178583e-08,
+      "loss": 0.6038,
+      "step": 41398
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.002852183241616,
+      "learning_rate": 3.189540955323534e-08,
+      "loss": 0.7678,
+      "step": 41399
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1605577663625908,
+      "learning_rate": 3.183454374214634e-08,
+      "loss": 0.6109,
+      "step": 41400
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4182945649417087,
+      "learning_rate": 3.17737359688719e-08,
+      "loss": 0.5352,
+      "step": 41401
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5417590782583723,
+      "learning_rate": 3.1712986233767286e-08,
+      "loss": 0.6757,
+      "step": 41402
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4107277494544093,
+      "learning_rate": 3.1652294537184436e-08,
+      "loss": 0.7217,
+      "step": 41403
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4977145397841896,
+      "learning_rate": 3.1591660879478624e-08,
+      "loss": 0.7283,
+      "step": 41404
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3608095195330665,
+      "learning_rate": 3.153108526100179e-08,
+      "loss": 0.7169,
+      "step": 41405
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.62896776106778,
+      "learning_rate": 3.147056768210699e-08,
+      "loss": 0.515,
+      "step": 41406
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.273732653387456,
+      "learning_rate": 3.141010814314727e-08,
+      "loss": 0.6182,
+      "step": 41407
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2744868577545474,
+      "learning_rate": 3.1349706644473454e-08,
+      "loss": 0.6062,
+      "step": 41408
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3351580490467578,
+      "learning_rate": 3.1289363186437494e-08,
+      "loss": 0.6074,
+      "step": 41409
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.078953571624037,
+      "learning_rate": 3.122907776939022e-08,
+      "loss": 0.5903,
+      "step": 41410
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4247276425223043,
+      "learning_rate": 3.1168850393684666e-08,
+      "loss": 0.6704,
+      "step": 41411
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6129992898802503,
+      "learning_rate": 3.1108681059669466e-08,
+      "loss": 0.456,
+      "step": 41412
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6767752821242197,
+      "learning_rate": 3.1048569767695435e-08,
+      "loss": 0.8456,
+      "step": 41413
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5223315679692384,
+      "learning_rate": 3.098851651811341e-08,
+      "loss": 0.7863,
+      "step": 41414
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5916536536337471,
+      "learning_rate": 3.0928521311271997e-08,
+      "loss": 0.6961,
+      "step": 41415
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4548301779747947,
+      "learning_rate": 3.086858414752092e-08,
+      "loss": 0.6662,
+      "step": 41416
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.539345322371742,
+      "learning_rate": 3.0808705027208786e-08,
+      "loss": 0.7338,
+      "step": 41417
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2117305300021592,
+      "learning_rate": 3.07488839506842e-08,
+      "loss": 0.6634,
+      "step": 41418
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4929925101265071,
+      "learning_rate": 3.06891209182969e-08,
+      "loss": 0.6751,
+      "step": 41419
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.706727832118963,
+      "learning_rate": 3.062941593039326e-08,
+      "loss": 0.7224,
+      "step": 41420
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5475806845796148,
+      "learning_rate": 3.056976898732189e-08,
+      "loss": 0.482,
+      "step": 41421
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.731508535871663,
+      "learning_rate": 3.051018008942919e-08,
+      "loss": 0.7222,
+      "step": 41422
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2793628681861704,
+      "learning_rate": 3.045064923706265e-08,
+      "loss": 0.5897,
+      "step": 41423
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6062887726708381,
+      "learning_rate": 3.039117643056755e-08,
+      "loss": 0.5117,
+      "step": 41424
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6375339282486572,
+      "learning_rate": 3.0331761670292505e-08,
+      "loss": 0.5307,
+      "step": 41425
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4796538135834345,
+      "learning_rate": 3.027240495658279e-08,
+      "loss": 0.7277,
+      "step": 41426
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.318812296512122,
+      "learning_rate": 3.021310628978258e-08,
+      "loss": 0.6819,
+      "step": 41427
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4880926836688941,
+      "learning_rate": 3.015386567023826e-08,
+      "loss": 0.6174,
+      "step": 41428
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5170382149089525,
+      "learning_rate": 3.009468309829511e-08,
+      "loss": 0.6841,
+      "step": 41429
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5363254602423175,
+      "learning_rate": 3.003555857429619e-08,
+      "loss": 0.7124,
+      "step": 41430
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5078417653746032,
+      "learning_rate": 2.9976492098586775e-08,
+      "loss": 0.7461,
+      "step": 41431
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5443417331185292,
+      "learning_rate": 2.991748367151104e-08,
+      "loss": 0.703,
+      "step": 41432
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7139068575829823,
+      "learning_rate": 2.985853329341315e-08,
+      "loss": 0.6048,
+      "step": 41433
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.612046014956989,
+      "learning_rate": 2.9799640964633946e-08,
+      "loss": 0.7254,
+      "step": 41434
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5433276609800892,
+      "learning_rate": 2.9740806685517597e-08,
+      "loss": 0.7427,
+      "step": 41435
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2279923164920794,
+      "learning_rate": 2.9682030456408272e-08,
+      "loss": 0.8012,
+      "step": 41436
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6070452427007909,
+      "learning_rate": 2.96233122776457e-08,
+      "loss": 0.517,
+      "step": 41437
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3547166570207885,
+      "learning_rate": 2.9564652149572938e-08,
+      "loss": 0.6645,
+      "step": 41438
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3669483126301725,
+      "learning_rate": 2.950605007253082e-08,
+      "loss": 0.6693,
+      "step": 41439
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3679959881800543,
+      "learning_rate": 2.9447506046862417e-08,
+      "loss": 0.6975,
+      "step": 41440
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3072855675816841,
+      "learning_rate": 2.938902007290634e-08,
+      "loss": 0.5989,
+      "step": 41441
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.394869860735506,
+      "learning_rate": 2.9330592151003424e-08,
+      "loss": 0.7473,
+      "step": 41442
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2804114044868131,
+      "learning_rate": 2.927222228149562e-08,
+      "loss": 0.6603,
+      "step": 41443
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.342281409087871,
+      "learning_rate": 2.921391046472155e-08,
+      "loss": 0.7233,
+      "step": 41444
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3170465469403305,
+      "learning_rate": 2.915565670101983e-08,
+      "loss": 0.5964,
+      "step": 41445
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3462363882720871,
+      "learning_rate": 2.9097460990731296e-08,
+      "loss": 0.4629,
+      "step": 41446
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6744952300858025,
+      "learning_rate": 2.9039323334193458e-08,
+      "loss": 0.6321,
+      "step": 41447
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.495488568896746,
+      "learning_rate": 2.8981243731746044e-08,
+      "loss": 0.6713,
+      "step": 41448
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.424219840141725,
+      "learning_rate": 2.8923222183726563e-08,
+      "loss": 0.8062,
+      "step": 41449
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.222285932198546,
+      "learning_rate": 2.886525869047363e-08,
+      "loss": 0.5885,
+      "step": 41450
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.362871088157658,
+      "learning_rate": 2.8807353252322533e-08,
+      "loss": 0.7036,
+      "step": 41451
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4890883683033265,
+      "learning_rate": 2.8749505869613003e-08,
+      "loss": 0.7702,
+      "step": 41452
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3046244228860144,
+      "learning_rate": 2.8691716542679215e-08,
+      "loss": 0.5144,
+      "step": 41453
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3978265791560296,
+      "learning_rate": 2.8633985271860898e-08,
+      "loss": 0.6187,
+      "step": 41454
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.419834010566787,
+      "learning_rate": 2.8576312057491118e-08,
+      "loss": 0.7154,
+      "step": 41455
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4431304142600823,
+      "learning_rate": 2.8518696899907382e-08,
+      "loss": 0.6526,
+      "step": 41456
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.791652441292285,
+      "learning_rate": 2.8461139799444982e-08,
+      "loss": 0.8435,
+      "step": 41457
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.94658051776808,
+      "learning_rate": 2.840364075643809e-08,
+      "loss": 0.6932,
+      "step": 41458
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7396708616684475,
+      "learning_rate": 2.8346199771221995e-08,
+      "loss": 0.7424,
+      "step": 41459
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6034959418605867,
+      "learning_rate": 2.8288816844131982e-08,
+      "loss": 0.4911,
+      "step": 41460
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6736892766457945,
+      "learning_rate": 2.8231491975502233e-08,
+      "loss": 0.5813,
+      "step": 41461
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6199215294029767,
+      "learning_rate": 2.8174225165664703e-08,
+      "loss": 0.6375,
+      "step": 41462
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.362132081104451,
+      "learning_rate": 2.8117016414953567e-08,
+      "loss": 0.6678,
+      "step": 41463
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3710870308440546,
+      "learning_rate": 2.8059865723703006e-08,
+      "loss": 0.6366,
+      "step": 41464
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0123124205522998,
+      "learning_rate": 2.800277309224497e-08,
+      "loss": 0.6353,
+      "step": 41465
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3693918013672493,
+      "learning_rate": 2.7945738520911426e-08,
+      "loss": 0.6134,
+      "step": 41466
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8732409309428086,
+      "learning_rate": 2.788876201003432e-08,
+      "loss": 0.7395,
+      "step": 41467
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8699554520098822,
+      "learning_rate": 2.7831843559946724e-08,
+      "loss": 0.796,
+      "step": 41468
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4351749292294607,
+      "learning_rate": 2.7774983170979487e-08,
+      "loss": 0.6052,
+      "step": 41469
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4289738340856102,
+      "learning_rate": 2.771818084346345e-08,
+      "loss": 0.5941,
+      "step": 41470
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.9275369475841213,
+      "learning_rate": 2.766143657772835e-08,
+      "loss": 0.6213,
+      "step": 41471
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8671601091189776,
+      "learning_rate": 2.7604750374106148e-08,
+      "loss": 0.7401,
+      "step": 41472
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.258867234558446,
+      "learning_rate": 2.7548122232926578e-08,
+      "loss": 0.6335,
+      "step": 41473
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5026760063538653,
+      "learning_rate": 2.7491552154519375e-08,
+      "loss": 0.7365,
+      "step": 41474
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3000657842094492,
+      "learning_rate": 2.7435040139213164e-08,
+      "loss": 0.5687,
+      "step": 41475
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.506811955261544,
+      "learning_rate": 2.7378586187336576e-08,
+      "loss": 0.8093,
+      "step": 41476
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2733573266121443,
+      "learning_rate": 2.732219029921934e-08,
+      "loss": 0.6449,
+      "step": 41477
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.589464901069775,
+      "learning_rate": 2.72658524751912e-08,
+      "loss": 0.8005,
+      "step": 41478
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2743738215586249,
+      "learning_rate": 2.7209572715577447e-08,
+      "loss": 0.5961,
+      "step": 41479
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2071685073044571,
+      "learning_rate": 2.7153351020706707e-08,
+      "loss": 0.5257,
+      "step": 41480
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4127588665513502,
+      "learning_rate": 2.7097187390906497e-08,
+      "loss": 0.5919,
+      "step": 41481
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5687222586673593,
+      "learning_rate": 2.7041081826504333e-08,
+      "loss": 0.621,
+      "step": 41482
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4427823884297124,
+      "learning_rate": 2.698503432782662e-08,
+      "loss": 0.6489,
+      "step": 41483
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.455541959952463,
+      "learning_rate": 2.6929044895198652e-08,
+      "loss": 0.725,
+      "step": 41484
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.523625596119304,
+      "learning_rate": 2.6873113528947946e-08,
+      "loss": 0.6176,
+      "step": 41485
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5970741701787299,
+      "learning_rate": 2.681724022939869e-08,
+      "loss": 0.7285,
+      "step": 41486
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6282281506725591,
+      "learning_rate": 2.6761424996877284e-08,
+      "loss": 0.5305,
+      "step": 41487
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5561199295761947,
+      "learning_rate": 2.6705667831707915e-08,
+      "loss": 0.7062,
+      "step": 41488
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7217192453263732,
+      "learning_rate": 2.664996873421588e-08,
+      "loss": 0.8673,
+      "step": 41489
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4276353100806352,
+      "learning_rate": 2.6594327704725365e-08,
+      "loss": 0.6944,
+      "step": 41490
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.622736560836694,
+      "learning_rate": 2.6538744743560552e-08,
+      "loss": 0.5192,
+      "step": 41491
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5343820286818213,
+      "learning_rate": 2.648321985104452e-08,
+      "loss": 0.6806,
+      "step": 41492
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4866784856962094,
+      "learning_rate": 2.6427753027500335e-08,
+      "loss": 0.6789,
+      "step": 41493
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.40894421297742,
+      "learning_rate": 2.6372344273251082e-08,
+      "loss": 0.6612,
+      "step": 41494
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5779616339066946,
+      "learning_rate": 2.6316993588620944e-08,
+      "loss": 0.4933,
+      "step": 41495
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.677125024850764,
+      "learning_rate": 2.626170097392966e-08,
+      "loss": 0.6425,
+      "step": 41496
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6023430935466823,
+      "learning_rate": 2.620646642950142e-08,
+      "loss": 0.5304,
+      "step": 41497
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4230308306669086,
+      "learning_rate": 2.6151289955655966e-08,
+      "loss": 0.6918,
+      "step": 41498
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.470029295006009,
+      "learning_rate": 2.6096171552715265e-08,
+      "loss": 0.798,
+      "step": 41499
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6494999987604415,
+      "learning_rate": 2.604111122100128e-08,
+      "loss": 0.673,
+      "step": 41500
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5509080067721477,
+      "learning_rate": 2.5986108960832644e-08,
+      "loss": 0.457,
+      "step": 41501
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.748129056543453,
+      "learning_rate": 2.5931164772530214e-08,
+      "loss": 0.7262,
+      "step": 41502
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9767876408446436,
+      "learning_rate": 2.5876278656415954e-08,
+      "loss": 0.7663,
+      "step": 41503
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.363901463825637,
+      "learning_rate": 2.5821450612806276e-08,
+      "loss": 0.6008,
+      "step": 41504
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4541303420717429,
+      "learning_rate": 2.5766680642023147e-08,
+      "loss": 0.702,
+      "step": 41505
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3511344532107585,
+      "learning_rate": 2.5711968744382975e-08,
+      "loss": 0.6999,
+      "step": 41506
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2337740730711864,
+      "learning_rate": 2.565731492020662e-08,
+      "loss": 0.6345,
+      "step": 41507
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.327903266526959,
+      "learning_rate": 2.5602719169810496e-08,
+      "loss": 0.8086,
+      "step": 41508
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4581381377604836,
+      "learning_rate": 2.5548181493513235e-08,
+      "loss": 0.6931,
+      "step": 41509
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4040505345670076,
+      "learning_rate": 2.5493701891632362e-08,
+      "loss": 0.7661,
+      "step": 41510
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5045581961408727,
+      "learning_rate": 2.5439280364485398e-08,
+      "loss": 0.6905,
+      "step": 41511
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7080833395339257,
+      "learning_rate": 2.538491691238765e-08,
+      "loss": 0.7676,
+      "step": 41512
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4113760493013643,
+      "learning_rate": 2.5330611535658857e-08,
+      "loss": 0.7249,
+      "step": 41513
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3953442358385164,
+      "learning_rate": 2.5276364234610995e-08,
+      "loss": 0.685,
+      "step": 41514
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2301252317263254,
+      "learning_rate": 2.5222175009563808e-08,
+      "loss": 0.6557,
+      "step": 41515
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2807697930178743,
+      "learning_rate": 2.516804386083038e-08,
+      "loss": 0.5274,
+      "step": 41516
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5022990110084384,
+      "learning_rate": 2.5113970788726018e-08,
+      "loss": 0.8367,
+      "step": 41517
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2801537983408293,
+      "learning_rate": 2.5059955793567125e-08,
+      "loss": 0.6364,
+      "step": 41518
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3774445276039498,
+      "learning_rate": 2.5005998875665683e-08,
+      "loss": 0.6692,
+      "step": 41519
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4279100861106204,
+      "learning_rate": 2.495210003533921e-08,
+      "loss": 0.7408,
+      "step": 41520
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3599082148548807,
+      "learning_rate": 2.4898259272898573e-08,
+      "loss": 0.647,
+      "step": 41521
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4995166502838566,
+      "learning_rate": 2.4844476588659073e-08,
+      "loss": 0.6518,
+      "step": 41522
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.7764217584025763,
+      "learning_rate": 2.4790751982931572e-08,
+      "loss": 0.8816,
+      "step": 41523
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2821529253952637,
+      "learning_rate": 2.4737085456032483e-08,
+      "loss": 0.5857,
+      "step": 41524
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4832888628162388,
+      "learning_rate": 2.4683477008270453e-08,
+      "loss": 0.8559,
+      "step": 41525
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4214127546435065,
+      "learning_rate": 2.4629926639960777e-08,
+      "loss": 0.6854,
+      "step": 41526
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5160587905395604,
+      "learning_rate": 2.4576434351412103e-08,
+      "loss": 0.6388,
+      "step": 41527
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6748869894133078,
+      "learning_rate": 2.4523000142939735e-08,
+      "loss": 0.6231,
+      "step": 41528
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3740612081943113,
+      "learning_rate": 2.44696240148512e-08,
+      "loss": 0.6357,
+      "step": 41529
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4087740084591855,
+      "learning_rate": 2.4416305967458476e-08,
+      "loss": 0.6702,
+      "step": 41530
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4032966282887658,
+      "learning_rate": 2.436304600107353e-08,
+      "loss": 0.5744,
+      "step": 41531
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6446673599269082,
+      "learning_rate": 2.43098441160039e-08,
+      "loss": 0.7493,
+      "step": 41532
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3381600066109698,
+      "learning_rate": 2.4256700312561554e-08,
+      "loss": 0.6825,
+      "step": 41533
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.44655655006086,
+      "learning_rate": 2.4203614591054024e-08,
+      "loss": 0.6707,
+      "step": 41534
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3017397816682574,
+      "learning_rate": 2.4150586951792175e-08,
+      "loss": 0.6195,
+      "step": 41535
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4708934241052904,
+      "learning_rate": 2.409761739508354e-08,
+      "loss": 0.6608,
+      "step": 41536
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7707266127188517,
+      "learning_rate": 2.4044705921236756e-08,
+      "loss": 0.7816,
+      "step": 41537
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3595216402628765,
+      "learning_rate": 2.3991852530560467e-08,
+      "loss": 0.6574,
+      "step": 41538
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.243902024705058,
+      "learning_rate": 2.3939057223362206e-08,
+      "loss": 0.6092,
+      "step": 41539
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.365358924309258,
+      "learning_rate": 2.3886319999948393e-08,
+      "loss": 0.6261,
+      "step": 41540
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2648064836197224,
+      "learning_rate": 2.383364086062656e-08,
+      "loss": 0.6424,
+      "step": 41541
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3593256660059885,
+      "learning_rate": 2.3781019805704243e-08,
+      "loss": 0.6331,
+      "step": 41542
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3643907495006868,
+      "learning_rate": 2.3728456835487857e-08,
+      "loss": 0.7052,
+      "step": 41543
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4174465728124637,
+      "learning_rate": 2.3675951950281605e-08,
+      "loss": 0.6224,
+      "step": 41544
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5166387754051076,
+      "learning_rate": 2.3623505150393024e-08,
+      "loss": 0.6526,
+      "step": 41545
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.381786634701717,
+      "learning_rate": 2.3571116436127418e-08,
+      "loss": 0.5204,
+      "step": 41546
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6218510158443055,
+      "learning_rate": 2.351878580778788e-08,
+      "loss": 0.4999,
+      "step": 41547
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5696203517593932,
+      "learning_rate": 2.3466513265680835e-08,
+      "loss": 0.6841,
+      "step": 41548
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.693698615697561,
+      "learning_rate": 2.341429881011159e-08,
+      "loss": 0.6281,
+      "step": 41549
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6512581439786473,
+      "learning_rate": 2.336214244138102e-08,
+      "loss": 0.6304,
+      "step": 41550
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1806272887332538,
+      "learning_rate": 2.331004415979554e-08,
+      "loss": 0.7159,
+      "step": 41551
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4520623531047374,
+      "learning_rate": 2.325800396565825e-08,
+      "loss": 0.7202,
+      "step": 41552
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4264199225112,
+      "learning_rate": 2.3206021859270012e-08,
+      "loss": 0.5857,
+      "step": 41553
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3644232386713044,
+      "learning_rate": 2.315409784093503e-08,
+      "loss": 0.6554,
+      "step": 41554
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.283612041425049,
+      "learning_rate": 2.3102231910956398e-08,
+      "loss": 0.697,
+      "step": 41555
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3919472072525183,
+      "learning_rate": 2.305042406963498e-08,
+      "loss": 0.6408,
+      "step": 41556
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2105125719047694,
+      "learning_rate": 2.2998674317272762e-08,
+      "loss": 0.4753,
+      "step": 41557
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.37699684135388,
+      "learning_rate": 2.2946982654170614e-08,
+      "loss": 0.7304,
+      "step": 41558
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4635488626945834,
+      "learning_rate": 2.2895349080630512e-08,
+      "loss": 0.6527,
+      "step": 41559
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3367517099658939,
+      "learning_rate": 2.2843773596951115e-08,
+      "loss": 0.7816,
+      "step": 41560
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8859217037005167,
+      "learning_rate": 2.2792256203434393e-08,
+      "loss": 0.5612,
+      "step": 41561
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9931285221470947,
+      "learning_rate": 2.2740796900380114e-08,
+      "loss": 0.8563,
+      "step": 41562
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5318545922249671,
+      "learning_rate": 2.2689395688088034e-08,
+      "loss": 0.7433,
+      "step": 41563
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8423361321683804,
+      "learning_rate": 2.2638052566856806e-08,
+      "loss": 0.5924,
+      "step": 41564
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3824173383649783,
+      "learning_rate": 2.2586767536985078e-08,
+      "loss": 0.5714,
+      "step": 41565
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3196627715584648,
+      "learning_rate": 2.253554059877261e-08,
+      "loss": 0.6068,
+      "step": 41566
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.450865566120746,
+      "learning_rate": 2.2484371752516942e-08,
+      "loss": 0.71,
+      "step": 41567
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5709152797118098,
+      "learning_rate": 2.2433260998515615e-08,
+      "loss": 0.6685,
+      "step": 41568
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5407222246789898,
+      "learning_rate": 2.238220833706728e-08,
+      "loss": 0.6282,
+      "step": 41569
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.380280057472268,
+      "learning_rate": 2.2331213768468363e-08,
+      "loss": 0.7735,
+      "step": 41570
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5501413754815696,
+      "learning_rate": 2.2280277293016404e-08,
+      "loss": 0.7078,
+      "step": 41571
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3082159076457291,
+      "learning_rate": 2.2229398911006726e-08,
+      "loss": 0.5354,
+      "step": 41572
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6991111165314543,
+      "learning_rate": 2.2178578622736867e-08,
+      "loss": 0.5634,
+      "step": 41573
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4804305811271836,
+      "learning_rate": 2.2127816428502148e-08,
+      "loss": 0.6157,
+      "step": 41574
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5426710276701914,
+      "learning_rate": 2.207711232859788e-08,
+      "loss": 0.8156,
+      "step": 41575
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2807461786818053,
+      "learning_rate": 2.2026466323320504e-08,
+      "loss": 0.6724,
+      "step": 41576
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6333410946621455,
+      "learning_rate": 2.1975878412964225e-08,
+      "loss": 0.5524,
+      "step": 41577
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.575368065168396,
+      "learning_rate": 2.1925348597822137e-08,
+      "loss": 0.5643,
+      "step": 41578
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.290693927890414,
+      "learning_rate": 2.1874876878189564e-08,
+      "loss": 0.6313,
+      "step": 41579
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4047081678219286,
+      "learning_rate": 2.1824463254361826e-08,
+      "loss": 0.6508,
+      "step": 41580
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.279582511214537,
+      "learning_rate": 2.177410772663091e-08,
+      "loss": 0.6776,
+      "step": 41581
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4915416151386818,
+      "learning_rate": 2.1723810295288807e-08,
+      "loss": 0.6421,
+      "step": 41582
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.657023475863656,
+      "learning_rate": 2.167357096063083e-08,
+      "loss": 0.6539,
+      "step": 41583
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.66654764336531,
+      "learning_rate": 2.1623389722947864e-08,
+      "loss": 0.7449,
+      "step": 41584
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5032631704122563,
+      "learning_rate": 2.1573266582533002e-08,
+      "loss": 0.6267,
+      "step": 41585
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2267314701275651,
+      "learning_rate": 2.152320153967824e-08,
+      "loss": 0.6813,
+      "step": 41586
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2541823425310756,
+      "learning_rate": 2.147319459467334e-08,
+      "loss": 0.7373,
+      "step": 41587
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.504015764486762,
+      "learning_rate": 2.1423245747811407e-08,
+      "loss": 0.5948,
+      "step": 41588
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.440778318767711,
+      "learning_rate": 2.13733549993822e-08,
+      "loss": 0.5159,
+      "step": 41589
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5170107454994537,
+      "learning_rate": 2.1323522349676607e-08,
+      "loss": 0.6218,
+      "step": 41590
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.410446905685927,
+      "learning_rate": 2.12737477989855e-08,
+      "loss": 0.648,
+      "step": 41591
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2490800696811948,
+      "learning_rate": 2.122403134759643e-08,
+      "loss": 0.7231,
+      "step": 41592
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.53381633158675,
+      "learning_rate": 2.1174372995801385e-08,
+      "loss": 0.7451,
+      "step": 41593
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3416482047108897,
+      "learning_rate": 2.112477274388791e-08,
+      "loss": 0.7649,
+      "step": 41594
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3729002325998645,
+      "learning_rate": 2.107523059214467e-08,
+      "loss": 0.6611,
+      "step": 41595
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7177725561795325,
+      "learning_rate": 2.1025746540861425e-08,
+      "loss": 0.6994,
+      "step": 41596
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3179479211910379,
+      "learning_rate": 2.0976320590325727e-08,
+      "loss": 0.6263,
+      "step": 41597
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4507847020733609,
+      "learning_rate": 2.0926952740825124e-08,
+      "loss": 0.6882,
+      "step": 41598
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6165499336258252,
+      "learning_rate": 2.0877642992647164e-08,
+      "loss": 0.5174,
+      "step": 41599
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2114615617850557,
+      "learning_rate": 2.082839134607828e-08,
+      "loss": 0.5238,
+      "step": 41600
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.588911861655119,
+      "learning_rate": 2.0779197801406026e-08,
+      "loss": 0.5135,
+      "step": 41601
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5964509198406365,
+      "learning_rate": 2.0730062358916835e-08,
+      "loss": 0.6928,
+      "step": 41602
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2907741154433214,
+      "learning_rate": 2.068098501889715e-08,
+      "loss": 0.6126,
+      "step": 41603
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.049079379819534,
+      "learning_rate": 2.063196578163118e-08,
+      "loss": 0.7248,
+      "step": 41604
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4123936326800715,
+      "learning_rate": 2.058300464740648e-08,
+      "loss": 0.7793,
+      "step": 41605
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4755059997665403,
+      "learning_rate": 2.0534101616506154e-08,
+      "loss": 0.697,
+      "step": 41606
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1934420422120746,
+      "learning_rate": 2.0485256689215525e-08,
+      "loss": 0.6367,
+      "step": 41607
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3813446903205346,
+      "learning_rate": 2.043646986581993e-08,
+      "loss": 0.7457,
+      "step": 41608
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1761483187429846,
+      "learning_rate": 2.038774114660136e-08,
+      "loss": 0.6504,
+      "step": 41609
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3973836132040722,
+      "learning_rate": 2.033907053184625e-08,
+      "loss": 0.6924,
+      "step": 41610
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9457961786217501,
+      "learning_rate": 2.029045802183549e-08,
+      "loss": 0.6999,
+      "step": 41611
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.312516663633277,
+      "learning_rate": 2.024190361685441e-08,
+      "loss": 0.6465,
+      "step": 41612
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.321745155594829,
+      "learning_rate": 2.019340731718389e-08,
+      "loss": 0.7485,
+      "step": 41613
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6019679649410137,
+      "learning_rate": 2.014496912310704e-08,
+      "loss": 0.5692,
+      "step": 41614
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6784250519849906,
+      "learning_rate": 2.0096589034905856e-08,
+      "loss": 0.7988,
+      "step": 41615
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.368997389650624,
+      "learning_rate": 2.0048267052861226e-08,
+      "loss": 0.6993,
+      "step": 41616
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3765839502227084,
+      "learning_rate": 2.0000003177255145e-08,
+      "loss": 0.5818,
+      "step": 41617
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5070049588161474,
+      "learning_rate": 1.9951797408369612e-08,
+      "loss": 0.709,
+      "step": 41618
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6049626730381872,
+      "learning_rate": 1.99036497464844e-08,
+      "loss": 0.6231,
+      "step": 41619
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2975605668327075,
+      "learning_rate": 1.9855560191879285e-08,
+      "loss": 0.5455,
+      "step": 41620
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3651343075197604,
+      "learning_rate": 1.9807528744834048e-08,
+      "loss": 0.6699,
+      "step": 41621
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8259783498836082,
+      "learning_rate": 1.975955540563068e-08,
+      "loss": 0.711,
+      "step": 41622
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.599674290991497,
+      "learning_rate": 1.9711640174545633e-08,
+      "loss": 0.4703,
+      "step": 41623
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4168134487625303,
+      "learning_rate": 1.9663783051858676e-08,
+      "loss": 0.627,
+      "step": 41624
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.26576721492409,
+      "learning_rate": 1.9615984037849588e-08,
+      "loss": 0.5915,
+      "step": 41625
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4217200446360412,
+      "learning_rate": 1.9568243132797036e-08,
+      "loss": 0.5839,
+      "step": 41626
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.125205418892112,
+      "learning_rate": 1.9520560336976357e-08,
+      "loss": 0.5651,
+      "step": 41627
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8257798817732087,
+      "learning_rate": 1.9472935650667325e-08,
+      "loss": 0.6683,
+      "step": 41628
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5524025826659271,
+      "learning_rate": 1.9425369074146384e-08,
+      "loss": 0.4868,
+      "step": 41629
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4377322233409482,
+      "learning_rate": 1.937786060769109e-08,
+      "loss": 0.7416,
+      "step": 41630
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3744467955094328,
+      "learning_rate": 1.933041025157789e-08,
+      "loss": 0.6921,
+      "step": 41631
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3257057660643388,
+      "learning_rate": 1.928301800608212e-08,
+      "loss": 0.6498,
+      "step": 41632
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5653869704081725,
+      "learning_rate": 1.9235683871481338e-08,
+      "loss": 0.4944,
+      "step": 41633
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.235760716251074,
+      "learning_rate": 1.918840784804976e-08,
+      "loss": 0.5537,
+      "step": 41634
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4912007131800606,
+      "learning_rate": 1.9141189936063842e-08,
+      "loss": 0.618,
+      "step": 41635
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3667346528367212,
+      "learning_rate": 1.9094030135796694e-08,
+      "loss": 0.6073,
+      "step": 41636
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.376949003248099,
+      "learning_rate": 1.904692844752476e-08,
+      "loss": 0.7441,
+      "step": 41637
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3566075077462845,
+      "learning_rate": 1.8999884871521157e-08,
+      "loss": 0.6926,
+      "step": 41638
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3207784031135241,
+      "learning_rate": 1.895289940806011e-08,
+      "loss": 0.7215,
+      "step": 41639
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5587138120795299,
+      "learning_rate": 1.8905972057415845e-08,
+      "loss": 0.8675,
+      "step": 41640
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.542258865555003,
+      "learning_rate": 1.8859102819860365e-08,
+      "loss": 0.6735,
+      "step": 41641
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4910188489164173,
+      "learning_rate": 1.8812291695666783e-08,
+      "loss": 0.6071,
+      "step": 41642
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5701108440382125,
+      "learning_rate": 1.876553868510822e-08,
+      "loss": 0.6455,
+      "step": 41643
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4327520730940626,
+      "learning_rate": 1.8718843788457787e-08,
+      "loss": 0.8742,
+      "step": 41644
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.402134683256793,
+      "learning_rate": 1.8672207005985267e-08,
+      "loss": 0.603,
+      "step": 41645
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.331430767191199,
+      "learning_rate": 1.8625628337963775e-08,
+      "loss": 0.6355,
+      "step": 41646
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.311340033100814,
+      "learning_rate": 1.8579107784663097e-08,
+      "loss": 0.762,
+      "step": 41647
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3502946499633317,
+      "learning_rate": 1.853264534635635e-08,
+      "loss": 0.678,
+      "step": 41648
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4214604360207244,
+      "learning_rate": 1.8486241023311093e-08,
+      "loss": 0.6685,
+      "step": 41649
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5946875535285627,
+      "learning_rate": 1.843989481579933e-08,
+      "loss": 0.5067,
+      "step": 41650
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.359959087504706,
+      "learning_rate": 1.8393606724090852e-08,
+      "loss": 0.738,
+      "step": 41651
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4099994718435995,
+      "learning_rate": 1.8347376748455438e-08,
+      "loss": 0.6868,
+      "step": 41652
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.794986187535192,
+      "learning_rate": 1.8301204889160652e-08,
+      "loss": 0.5979,
+      "step": 41653
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4222241045727133,
+      "learning_rate": 1.8255091146477387e-08,
+      "loss": 0.6622,
+      "step": 41654
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5059702490961417,
+      "learning_rate": 1.8209035520673213e-08,
+      "loss": 0.6021,
+      "step": 41655
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2999402707068612,
+      "learning_rate": 1.816303801201569e-08,
+      "loss": 0.7177,
+      "step": 41656
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4956552957976084,
+      "learning_rate": 1.8117098620773487e-08,
+      "loss": 0.6765,
+      "step": 41657
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.266582558494109,
+      "learning_rate": 1.8071217347213066e-08,
+      "loss": 0.685,
+      "step": 41658
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3499765214158794,
+      "learning_rate": 1.8025394191603095e-08,
+      "loss": 0.6585,
+      "step": 41659
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1655431570061165,
+      "learning_rate": 1.7979629154208923e-08,
+      "loss": 0.6397,
+      "step": 41660
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3552956320908505,
+      "learning_rate": 1.7933922235298108e-08,
+      "loss": 0.6356,
+      "step": 41661
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3383465806907313,
+      "learning_rate": 1.7888273435135994e-08,
+      "loss": 0.6524,
+      "step": 41662
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3496747258135067,
+      "learning_rate": 1.7842682753987927e-08,
+      "loss": 0.6846,
+      "step": 41663
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.290313148196163,
+      "learning_rate": 1.779715019212036e-08,
+      "loss": 0.6578,
+      "step": 41664
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3680720957068044,
+      "learning_rate": 1.7751675749798635e-08,
+      "loss": 0.6265,
+      "step": 41665
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5121675320717114,
+      "learning_rate": 1.7706259427286988e-08,
+      "loss": 0.5382,
+      "step": 41666
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.481875768982512,
+      "learning_rate": 1.7660901224849647e-08,
+      "loss": 0.7086,
+      "step": 41667
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.359276044926957,
+      "learning_rate": 1.761560114275085e-08,
+      "loss": 0.5422,
+      "step": 41668
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3738443980740862,
+      "learning_rate": 1.7570359181254827e-08,
+      "loss": 0.6845,
+      "step": 41669
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.4606833552954726,
+      "learning_rate": 1.752517534062359e-08,
+      "loss": 0.5985,
+      "step": 41670
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2368827642057552,
+      "learning_rate": 1.7480049621122485e-08,
+      "loss": 0.6735,
+      "step": 41671
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5892584125262883,
+      "learning_rate": 1.7434982023012414e-08,
+      "loss": 0.4829,
+      "step": 41672
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3575200622646364,
+      "learning_rate": 1.738997254655539e-08,
+      "loss": 0.5806,
+      "step": 41673
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7491141335995175,
+      "learning_rate": 1.7345021192015643e-08,
+      "loss": 0.6196,
+      "step": 41674
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2739444246137381,
+      "learning_rate": 1.730012795965297e-08,
+      "loss": 0.6848,
+      "step": 41675
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2562714155304882,
+      "learning_rate": 1.7255292849729376e-08,
+      "loss": 0.6565,
+      "step": 41676
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6533716661643764,
+      "learning_rate": 1.7210515862505773e-08,
+      "loss": 0.5505,
+      "step": 41677
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5470354236928237,
+      "learning_rate": 1.7165796998244168e-08,
+      "loss": 0.7332,
+      "step": 41678
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.611581615201782,
+      "learning_rate": 1.712113625720324e-08,
+      "loss": 0.6728,
+      "step": 41679
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5678855535796932,
+      "learning_rate": 1.7076533639642788e-08,
+      "loss": 0.4866,
+      "step": 41680
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.491558944065948,
+      "learning_rate": 1.7031989145823713e-08,
+      "loss": 0.7988,
+      "step": 41681
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6231044837759556,
+      "learning_rate": 1.6987502776005803e-08,
+      "loss": 0.4906,
+      "step": 41682
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3742143341460311,
+      "learning_rate": 1.6943074530446636e-08,
+      "loss": 0.7606,
+      "step": 41683
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4861845637555544,
+      "learning_rate": 1.6898704409404888e-08,
+      "loss": 0.7184,
+      "step": 41684
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4744668809619659,
+      "learning_rate": 1.6854392413139244e-08,
+      "loss": 0.7205,
+      "step": 41685
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5070803312491912,
+      "learning_rate": 1.6810138541909492e-08,
+      "loss": 0.6461,
+      "step": 41686
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4199146201553947,
+      "learning_rate": 1.676594279597099e-08,
+      "loss": 0.7279,
+      "step": 41687
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5615961045042028,
+      "learning_rate": 1.6721805175581307e-08,
+      "loss": 0.8404,
+      "step": 41688
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9614632259192746,
+      "learning_rate": 1.6677725680999124e-08,
+      "loss": 0.7482,
+      "step": 41689
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3413915741713496,
+      "learning_rate": 1.6633704312478683e-08,
+      "loss": 0.637,
+      "step": 41690
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3677554196210944,
+      "learning_rate": 1.6589741070277555e-08,
+      "loss": 0.6123,
+      "step": 41691
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3154241699435438,
+      "learning_rate": 1.65458359546522e-08,
+      "loss": 0.5877,
+      "step": 41692
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.324659794616615,
+      "learning_rate": 1.6501988965857972e-08,
+      "loss": 0.6249,
+      "step": 41693
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3150271452231925,
+      "learning_rate": 1.6458200104149115e-08,
+      "loss": 0.681,
+      "step": 41694
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6690652541236497,
+      "learning_rate": 1.641446936978208e-08,
+      "loss": 0.7086,
+      "step": 41695
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9470000382889534,
+      "learning_rate": 1.637079676301001e-08,
+      "loss": 0.7146,
+      "step": 41696
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.291227120225091,
+      "learning_rate": 1.632718228408825e-08,
+      "loss": 0.7626,
+      "step": 41697
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5882274031101686,
+      "learning_rate": 1.6283625933269932e-08,
+      "loss": 0.6833,
+      "step": 41698
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3756980240328622,
+      "learning_rate": 1.6240127710810405e-08,
+      "loss": 0.6914,
+      "step": 41699
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5087724710411776,
+      "learning_rate": 1.6196687616960583e-08,
+      "loss": 0.6481,
+      "step": 41700
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4496375689547274,
+      "learning_rate": 1.6153305651974703e-08,
+      "loss": 0.747,
+      "step": 41701
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5511662039731762,
+      "learning_rate": 1.610998181610479e-08,
+      "loss": 0.6554,
+      "step": 41702
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3898852230672198,
+      "learning_rate": 1.6066716109603975e-08,
+      "loss": 0.496,
+      "step": 41703
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.333238038876883,
+      "learning_rate": 1.6023508532724273e-08,
+      "loss": 0.6424,
+      "step": 41704
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.453895021810377,
+      "learning_rate": 1.5980359085715493e-08,
+      "loss": 0.7406,
+      "step": 41705
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5023937089728208,
+      "learning_rate": 1.5937267768829646e-08,
+      "loss": 0.7389,
+      "step": 41706
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4296811632976947,
+      "learning_rate": 1.5894234582318758e-08,
+      "loss": 0.6685,
+      "step": 41707
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.45466915183042,
+      "learning_rate": 1.585125952643263e-08,
+      "loss": 0.6121,
+      "step": 41708
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6066366821663215,
+      "learning_rate": 1.5808342601421055e-08,
+      "loss": 0.4736,
+      "step": 41709
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3207738333516148,
+      "learning_rate": 1.5765483807533843e-08,
+      "loss": 0.712,
+      "step": 41710
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.211528172405223,
+      "learning_rate": 1.5722683145020788e-08,
+      "loss": 0.5661,
+      "step": 41711
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4098215114500185,
+      "learning_rate": 1.5679940614131696e-08,
+      "loss": 0.6738,
+      "step": 41712
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9113778554701322,
+      "learning_rate": 1.563725621511525e-08,
+      "loss": 0.6229,
+      "step": 41713
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3617226310557617,
+      "learning_rate": 1.559462994822014e-08,
+      "loss": 0.8001,
+      "step": 41714
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4442585440242361,
+      "learning_rate": 1.5552061813692843e-08,
+      "loss": 0.6832,
+      "step": 41715
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.282521171970617,
+      "learning_rate": 1.5509551811783152e-08,
+      "loss": 0.6465,
+      "step": 41716
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5508687486045536,
+      "learning_rate": 1.546709994273865e-08,
+      "loss": 0.7274,
+      "step": 41717
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5935696006556311,
+      "learning_rate": 1.5424706206804695e-08,
+      "loss": 0.5564,
+      "step": 41718
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5927234194458073,
+      "learning_rate": 1.538237060422887e-08,
+      "loss": 0.4864,
+      "step": 41719
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.446472306971634,
+      "learning_rate": 1.5340093135258748e-08,
+      "loss": 0.623,
+      "step": 41720
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.597690935701258,
+      "learning_rate": 1.5297873800139696e-08,
+      "loss": 0.66,
+      "step": 41721
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2321032575554476,
+      "learning_rate": 1.5255712599117066e-08,
+      "loss": 0.6034,
+      "step": 41722
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.541781086569559,
+      "learning_rate": 1.5213609532436223e-08,
+      "loss": 0.7723,
+      "step": 41723
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5698961784650376,
+      "learning_rate": 1.5171564600343636e-08,
+      "loss": 0.6111,
+      "step": 41724
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7566281312752852,
+      "learning_rate": 1.512957780308355e-08,
+      "loss": 0.687,
+      "step": 41725
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1103590615987218,
+      "learning_rate": 1.5087649140900218e-08,
+      "loss": 0.5874,
+      "step": 41726
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0842764949999886,
+      "learning_rate": 1.5045778614036775e-08,
+      "loss": 0.5916,
+      "step": 41727
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1260145503746366,
+      "learning_rate": 1.5003966222739697e-08,
+      "loss": 0.6236,
+      "step": 41728
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.451983641268037,
+      "learning_rate": 1.4962211967249895e-08,
+      "loss": 0.6186,
+      "step": 41729
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4426059859469034,
+      "learning_rate": 1.4920515847810513e-08,
+      "loss": 0.6738,
+      "step": 41730
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2177224530895634,
+      "learning_rate": 1.4878877864665797e-08,
+      "loss": 0.6687,
+      "step": 41731
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4402198853895822,
+      "learning_rate": 1.4837298018057778e-08,
+      "loss": 0.6007,
+      "step": 41732
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3521389450860577,
+      "learning_rate": 1.4795776308228482e-08,
+      "loss": 0.7192,
+      "step": 41733
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3651632755196108,
+      "learning_rate": 1.475431273541883e-08,
+      "loss": 0.5153,
+      "step": 41734
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5004542459911463,
+      "learning_rate": 1.4712907299870848e-08,
+      "loss": 0.7135,
+      "step": 41735
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.613061164031673,
+      "learning_rate": 1.4671560001825458e-08,
+      "loss": 0.6477,
+      "step": 41736
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3696997500948476,
+      "learning_rate": 1.4630270841523575e-08,
+      "loss": 0.7101,
+      "step": 41737
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8956025794395931,
+      "learning_rate": 1.4589039819206118e-08,
+      "loss": 0.5006,
+      "step": 41738
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.406802439252757,
+      "learning_rate": 1.4547866935111787e-08,
+      "loss": 0.7013,
+      "step": 41739
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4751161939277868,
+      "learning_rate": 1.45067521894815e-08,
+      "loss": 0.5932,
+      "step": 41740
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.719563083195126,
+      "learning_rate": 1.4465695582553951e-08,
+      "loss": 0.5952,
+      "step": 41741
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.230631896480381,
+      "learning_rate": 1.4424697114570063e-08,
+      "loss": 0.7311,
+      "step": 41742
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0856001543777705,
+      "learning_rate": 1.4383756785765201e-08,
+      "loss": 0.5993,
+      "step": 41743
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.460656646945474,
+      "learning_rate": 1.4342874596380285e-08,
+      "loss": 0.6357,
+      "step": 41744
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4962252457724015,
+      "learning_rate": 1.4302050546651791e-08,
+      "loss": 0.6407,
+      "step": 41745
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4762591972726387,
+      "learning_rate": 1.4261284636819527e-08,
+      "loss": 0.7068,
+      "step": 41746
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4622338200468774,
+      "learning_rate": 1.4220576867118862e-08,
+      "loss": 0.5588,
+      "step": 41747
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.349499535774762,
+      "learning_rate": 1.4179927237786272e-08,
+      "loss": 0.6793,
+      "step": 41748
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4662471268538941,
+      "learning_rate": 1.4139335749061567e-08,
+      "loss": 0.6431,
+      "step": 41749
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.2706967353599996,
+      "learning_rate": 1.4098802401177891e-08,
+      "loss": 0.6783,
+      "step": 41750
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4724244227477201,
+      "learning_rate": 1.4058327194372834e-08,
+      "loss": 0.666,
+      "step": 41751
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5864527159381564,
+      "learning_rate": 1.4017910128881762e-08,
+      "loss": 0.5085,
+      "step": 41752
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.314825044645596,
+      "learning_rate": 1.3977551204938932e-08,
+      "loss": 0.7632,
+      "step": 41753
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3996874410552103,
+      "learning_rate": 1.3937250422781934e-08,
+      "loss": 0.541,
+      "step": 41754
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4992174693584066,
+      "learning_rate": 1.3897007782642802e-08,
+      "loss": 0.7165,
+      "step": 41755
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3161972394592338,
+      "learning_rate": 1.3856823284756904e-08,
+      "loss": 0.6654,
+      "step": 41756
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3862837261461607,
+      "learning_rate": 1.3816696929358498e-08,
+      "loss": 0.5775,
+      "step": 41757
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4301486515285629,
+      "learning_rate": 1.377662871668073e-08,
+      "loss": 0.7538,
+      "step": 41758
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4268795643159196,
+      "learning_rate": 1.3736618646956746e-08,
+      "loss": 0.7339,
+      "step": 41759
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5776826670022276,
+      "learning_rate": 1.3696666720419693e-08,
+      "loss": 0.4927,
+      "step": 41760
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3209380311880763,
+      "learning_rate": 1.365677293730272e-08,
+      "loss": 0.6894,
+      "step": 41761
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5924199537676327,
+      "learning_rate": 1.3616937297837863e-08,
+      "loss": 0.5285,
+      "step": 41762
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4642126222824412,
+      "learning_rate": 1.3577159802256045e-08,
+      "loss": 0.6073,
+      "step": 41763
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4885847182950127,
+      "learning_rate": 1.3537440450790417e-08,
+      "loss": 0.826,
+      "step": 41764
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3486926963045274,
+      "learning_rate": 1.3497779243670794e-08,
+      "loss": 0.7408,
+      "step": 41765
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3145325479152608,
+      "learning_rate": 1.3458176181129211e-08,
+      "loss": 0.6765,
+      "step": 41766
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3039780473944829,
+      "learning_rate": 1.3418631263395487e-08,
+      "loss": 0.6598,
+      "step": 41767
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.160316244754289,
+      "learning_rate": 1.3379144490700546e-08,
+      "loss": 0.6598,
+      "step": 41768
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.484590172769281,
+      "learning_rate": 1.3339715863273094e-08,
+      "loss": 0.6703,
+      "step": 41769
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2828355186018132,
+      "learning_rate": 1.330034538134517e-08,
+      "loss": 0.7575,
+      "step": 41770
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3918481013081319,
+      "learning_rate": 1.3261033045143257e-08,
+      "loss": 0.6797,
+      "step": 41771
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.605810373432521,
+      "learning_rate": 1.3221778854897171e-08,
+      "loss": 0.4739,
+      "step": 41772
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4944751910325969,
+      "learning_rate": 1.318258281083562e-08,
+      "loss": 0.6106,
+      "step": 41773
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2777459335511367,
+      "learning_rate": 1.3143444913187309e-08,
+      "loss": 0.648,
+      "step": 41774
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2795600699541962,
+      "learning_rate": 1.3104365162179833e-08,
+      "loss": 0.5973,
+      "step": 41775
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.345825813432798,
+      "learning_rate": 1.3065343558040789e-08,
+      "loss": 0.6891,
+      "step": 41776
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.551861885518633,
+      "learning_rate": 1.302638010099666e-08,
+      "loss": 0.7217,
+      "step": 41777
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.429806962137474,
+      "learning_rate": 1.2987474791273936e-08,
+      "loss": 0.6842,
+      "step": 41778
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6036720734067302,
+      "learning_rate": 1.2948627629101318e-08,
+      "loss": 0.527,
+      "step": 41779
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6046153182519692,
+      "learning_rate": 1.2909838614703073e-08,
+      "loss": 0.5003,
+      "step": 41780
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.628952635814586,
+      "learning_rate": 1.287110774830569e-08,
+      "loss": 0.6836,
+      "step": 41781
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4654962476117783,
+      "learning_rate": 1.2832435030133427e-08,
+      "loss": 0.5724,
+      "step": 41782
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4566580293136369,
+      "learning_rate": 1.2793820460413886e-08,
+      "loss": 0.7965,
+      "step": 41783
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5491957375360252,
+      "learning_rate": 1.275526403937022e-08,
+      "loss": 0.6802,
+      "step": 41784
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4944928632640138,
+      "learning_rate": 1.2716765767227801e-08,
+      "loss": 0.7202,
+      "step": 41785
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3306296635823496,
+      "learning_rate": 1.2678325644209789e-08,
+      "loss": 0.5555,
+      "step": 41786
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.474337184087324,
+      "learning_rate": 1.2639943670540445e-08,
+      "loss": 0.7405,
+      "step": 41787
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.667223382310731,
+      "learning_rate": 1.2601619846444035e-08,
+      "loss": 0.7392,
+      "step": 41788
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2341247514107594,
+      "learning_rate": 1.2563354172142605e-08,
+      "loss": 0.6303,
+      "step": 41789
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4758391798100028,
+      "learning_rate": 1.252514664785931e-08,
+      "loss": 0.5342,
+      "step": 41790
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8992476828978408,
+      "learning_rate": 1.2486997273816192e-08,
+      "loss": 0.5974,
+      "step": 41791
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7520427342383629,
+      "learning_rate": 1.2448906050236408e-08,
+      "loss": 0.6974,
+      "step": 41792
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2356366072277565,
+      "learning_rate": 1.2410872977340893e-08,
+      "loss": 0.5857,
+      "step": 41793
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5918299384867135,
+      "learning_rate": 1.2372898055350579e-08,
+      "loss": 0.4769,
+      "step": 41794
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.642375295386068,
+      "learning_rate": 1.2334981284487513e-08,
+      "loss": 0.5038,
+      "step": 41795
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2658994008574838,
+      "learning_rate": 1.2297122664972627e-08,
+      "loss": 0.6923,
+      "step": 41796
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2310070691716135,
+      "learning_rate": 1.2259322197025747e-08,
+      "loss": 0.6894,
+      "step": 41797
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6738520887472226,
+      "learning_rate": 1.2221579880867807e-08,
+      "loss": 0.6247,
+      "step": 41798
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4622845345140563,
+      "learning_rate": 1.218389571671752e-08,
+      "loss": 0.6999,
+      "step": 41799
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4349136331343708,
+      "learning_rate": 1.2146269704793601e-08,
+      "loss": 0.6192,
+      "step": 41800
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2960921404319754,
+      "learning_rate": 1.2108701845318094e-08,
+      "loss": 0.8082,
+      "step": 41801
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2922539383679097,
+      "learning_rate": 1.2071192138506382e-08,
+      "loss": 0.6597,
+      "step": 41802
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4973945017822536,
+      "learning_rate": 1.203374058457829e-08,
+      "loss": 0.7485,
+      "step": 41803
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5658272043741852,
+      "learning_rate": 1.1996347183752532e-08,
+      "loss": 0.5055,
+      "step": 41804
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.590264243338696,
+      "learning_rate": 1.1959011936246712e-08,
+      "loss": 0.5984,
+      "step": 41805
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4403493391879396,
+      "learning_rate": 1.1921734842277321e-08,
+      "loss": 0.5786,
+      "step": 41806
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9569505330207757,
+      "learning_rate": 1.1884515902060856e-08,
+      "loss": 0.7619,
+      "step": 41807
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3359123832031394,
+      "learning_rate": 1.1847355115814918e-08,
+      "loss": 0.7362,
+      "step": 41808
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.25978521208913,
+      "learning_rate": 1.1810252483757112e-08,
+      "loss": 0.7059,
+      "step": 41809
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3750842790303957,
+      "learning_rate": 1.177320800610171e-08,
+      "loss": 0.7199,
+      "step": 41810
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5891916967768794,
+      "learning_rate": 1.1736221683064097e-08,
+      "loss": 0.4949,
+      "step": 41811
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6209639144301813,
+      "learning_rate": 1.1699293514861877e-08,
+      "loss": 0.6284,
+      "step": 41812
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2515858600603273,
+      "learning_rate": 1.1662423501707099e-08,
+      "loss": 0.6531,
+      "step": 41813
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.556579787146961,
+      "learning_rate": 1.162561164381737e-08,
+      "loss": 0.7729,
+      "step": 41814
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2696840686474684,
+      "learning_rate": 1.158885794140474e-08,
+      "loss": 0.7472,
+      "step": 41815
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.244406136776329,
+      "learning_rate": 1.1552162394684596e-08,
+      "loss": 0.5912,
+      "step": 41816
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5039158821571585,
+      "learning_rate": 1.1515525003868988e-08,
+      "loss": 0.746,
+      "step": 41817
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2614943932899945,
+      "learning_rate": 1.14789457691733e-08,
+      "loss": 0.5349,
+      "step": 41818
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4444401905731927,
+      "learning_rate": 1.1442424690809584e-08,
+      "loss": 0.7399,
+      "step": 41819
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4398553915863879,
+      "learning_rate": 1.1405961768989893e-08,
+      "loss": 0.6857,
+      "step": 41820
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4270388749153637,
+      "learning_rate": 1.1369557003927389e-08,
+      "loss": 0.7536,
+      "step": 41821
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.400934001340128,
+      "learning_rate": 1.1333210395833017e-08,
+      "loss": 0.6437,
+      "step": 41822
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5866636985289027,
+      "learning_rate": 1.1296921944919936e-08,
+      "loss": 0.6308,
+      "step": 41823
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8981664342979188,
+      "learning_rate": 1.126069165139798e-08,
+      "loss": 0.7851,
+      "step": 41824
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.301385073509684,
+      "learning_rate": 1.1224519515478095e-08,
+      "loss": 0.6964,
+      "step": 41825
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.010634591405603,
+      "learning_rate": 1.1188405537372327e-08,
+      "loss": 0.666,
+      "step": 41826
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.488224471640396,
+      "learning_rate": 1.1152349717290512e-08,
+      "loss": 0.6959,
+      "step": 41827
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4950971888253721,
+      "learning_rate": 1.111635205544137e-08,
+      "loss": 0.6669,
+      "step": 41828
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5183994546007944,
+      "learning_rate": 1.1080412552034736e-08,
+      "loss": 0.6569,
+      "step": 41829
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3336264987571629,
+      "learning_rate": 1.104453120728155e-08,
+      "loss": 0.6629,
+      "step": 41830
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6705499494098217,
+      "learning_rate": 1.1008708021389426e-08,
+      "loss": 0.697,
+      "step": 41831
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3653809706583147,
+      "learning_rate": 1.0972942994565971e-08,
+      "loss": 0.66,
+      "step": 41832
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4849620482468002,
+      "learning_rate": 1.0937236127022133e-08,
+      "loss": 0.6944,
+      "step": 41833
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3284052096051098,
+      "learning_rate": 1.0901587418962189e-08,
+      "loss": 0.7255,
+      "step": 41834
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6054087557938991,
+      "learning_rate": 1.0865996870597084e-08,
+      "loss": 0.5001,
+      "step": 41835
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2174099147212418,
+      "learning_rate": 1.0830464482132208e-08,
+      "loss": 0.638,
+      "step": 41836
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.451299052602878,
+      "learning_rate": 1.0794990253775172e-08,
+      "loss": 0.6278,
+      "step": 41837
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5636773830492066,
+      "learning_rate": 1.075957418573248e-08,
+      "loss": 0.4674,
+      "step": 41838
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2576789615927988,
+      "learning_rate": 1.072421627820952e-08,
+      "loss": 0.6249,
+      "step": 41839
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2856989600271944,
+      "learning_rate": 1.0688916531412796e-08,
+      "loss": 0.6986,
+      "step": 41840
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4201801213522536,
+      "learning_rate": 1.0653674945547699e-08,
+      "loss": 0.7266,
+      "step": 41841
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.121164950189274,
+      "learning_rate": 1.0618491520819619e-08,
+      "loss": 0.6865,
+      "step": 41842
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5566870153909,
+      "learning_rate": 1.0583366257432836e-08,
+      "loss": 0.773,
+      "step": 41843
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4745745700654722,
+      "learning_rate": 1.0548299155593855e-08,
+      "loss": 0.707,
+      "step": 41844
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2551209149199645,
+      "learning_rate": 1.0513290215503624e-08,
+      "loss": 0.7181,
+      "step": 41845
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7729302692717182,
+      "learning_rate": 1.0478339437368645e-08,
+      "loss": 0.695,
+      "step": 41846
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5143956375147396,
+      "learning_rate": 1.044344682139098e-08,
+      "loss": 0.7473,
+      "step": 41847
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3343671340385646,
+      "learning_rate": 1.0408612367773796e-08,
+      "loss": 0.6102,
+      "step": 41848
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2734062400540715,
+      "learning_rate": 1.0373836076720267e-08,
+      "loss": 0.5735,
+      "step": 41849
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4436145098831417,
+      "learning_rate": 1.0339117948433564e-08,
+      "loss": 0.6519,
+      "step": 41850
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4070824714832832,
+      "learning_rate": 1.0304457983114636e-08,
+      "loss": 0.6935,
+      "step": 41851
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3360422798107916,
+      "learning_rate": 1.0269856180966653e-08,
+      "loss": 0.5891,
+      "step": 41852
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.448492459936085,
+      "learning_rate": 1.0235312542189458e-08,
+      "loss": 0.6332,
+      "step": 41853
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3272036021956457,
+      "learning_rate": 1.0200827066986219e-08,
+      "loss": 0.71,
+      "step": 41854
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7777012475299259,
+      "learning_rate": 1.0166399755554557e-08,
+      "loss": 0.6215,
+      "step": 41855
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5874020991132628,
+      "learning_rate": 1.0132030608098752e-08,
+      "loss": 0.5113,
+      "step": 41856
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3840444689983735,
+      "learning_rate": 1.0097719624815316e-08,
+      "loss": 0.7094,
+      "step": 41857
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9091727593316352,
+      "learning_rate": 1.0063466805907418e-08,
+      "loss": 0.7001,
+      "step": 41858
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5919139548867405,
+      "learning_rate": 1.0029272151571567e-08,
+      "loss": 0.4827,
+      "step": 41859
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5610896970801964,
+      "learning_rate": 9.995135662007604e-09,
+      "loss": 0.6231,
+      "step": 41860
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5877597024176213,
+      "learning_rate": 9.96105733741648e-09,
+      "loss": 0.7116,
+      "step": 41861
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5944920925826407,
+      "learning_rate": 9.927037177993593e-09,
+      "loss": 0.526,
+      "step": 41862
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6543510516034545,
+      "learning_rate": 9.893075183938783e-09,
+      "loss": 0.7404,
+      "step": 41863
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3331424764214388,
+      "learning_rate": 9.859171355448561e-09,
+      "loss": 0.5963,
+      "step": 41864
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4210250532827828,
+      "learning_rate": 9.825325692721654e-09,
+      "loss": 0.6474,
+      "step": 41865
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4691834709241154,
+      "learning_rate": 9.791538195954575e-09,
+      "loss": 0.7861,
+      "step": 41866
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.454910819124733,
+      "learning_rate": 9.757808865343831e-09,
+      "loss": 0.7664,
+      "step": 41867
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.7727993885143665,
+      "learning_rate": 9.724137701085933e-09,
+      "loss": 0.632,
+      "step": 41868
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7974169559775601,
+      "learning_rate": 9.690524703377391e-09,
+      "loss": 0.8672,
+      "step": 41869
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.098092385165718,
+      "learning_rate": 9.656969872413601e-09,
+      "loss": 0.6921,
+      "step": 41870
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3464479090485377,
+      "learning_rate": 9.623473208388855e-09,
+      "loss": 0.6855,
+      "step": 41871
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4122830133575404,
+      "learning_rate": 9.590034711499663e-09,
+      "loss": 0.7782,
+      "step": 41872
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6625631714894684,
+      "learning_rate": 9.556654381940311e-09,
+      "loss": 0.6017,
+      "step": 41873
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7204958467505072,
+      "learning_rate": 9.523332219905091e-09,
+      "loss": 0.5082,
+      "step": 41874
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4760683515650879,
+      "learning_rate": 9.490068225588289e-09,
+      "loss": 0.6392,
+      "step": 41875
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5539478634552937,
+      "learning_rate": 9.456862399183087e-09,
+      "loss": 0.7108,
+      "step": 41876
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7877693336581026,
+      "learning_rate": 9.423714740882661e-09,
+      "loss": 0.7466,
+      "step": 41877
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5433008363330125,
+      "learning_rate": 9.390625250880191e-09,
+      "loss": 0.7509,
+      "step": 41878
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5853100184005444,
+      "learning_rate": 9.357593929368857e-09,
+      "loss": 0.7331,
+      "step": 41879
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6027542395727565,
+      "learning_rate": 9.324620776540727e-09,
+      "loss": 0.7973,
+      "step": 41880
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7692168051961843,
+      "learning_rate": 9.291705792586758e-09,
+      "loss": 0.6861,
+      "step": 41881
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3221707127688564,
+      "learning_rate": 9.25884897770013e-09,
+      "loss": 0.5701,
+      "step": 41882
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3822636367963697,
+      "learning_rate": 9.226050332071801e-09,
+      "loss": 0.653,
+      "step": 41883
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6391432712468763,
+      "learning_rate": 9.193309855891619e-09,
+      "loss": 0.5176,
+      "step": 41884
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4646648759805037,
+      "learning_rate": 9.160627549351652e-09,
+      "loss": 0.6309,
+      "step": 41885
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8750098427266448,
+      "learning_rate": 9.12800341264175e-09,
+      "loss": 0.7544,
+      "step": 41886
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6955386722139911,
+      "learning_rate": 9.095437445950651e-09,
+      "loss": 0.6658,
+      "step": 41887
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5775326652279513,
+      "learning_rate": 9.06292964947042e-09,
+      "loss": 0.6952,
+      "step": 41888
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.538796190580732,
+      "learning_rate": 9.030480023387579e-09,
+      "loss": 0.8358,
+      "step": 41889
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6167328857077108,
+      "learning_rate": 8.998088567893081e-09,
+      "loss": 0.7176,
+      "step": 41890
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4151356567466915,
+      "learning_rate": 8.96575528317456e-09,
+      "loss": 0.5661,
+      "step": 41891
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2457439105898613,
+      "learning_rate": 8.93348016942075e-09,
+      "loss": 0.6092,
+      "step": 41892
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.516769543553518,
+      "learning_rate": 8.90126322681928e-09,
+      "loss": 0.6122,
+      "step": 41893
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5305438742057351,
+      "learning_rate": 8.869104455557776e-09,
+      "loss": 0.6864,
+      "step": 41894
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7007817277159654,
+      "learning_rate": 8.837003855822756e-09,
+      "loss": 0.7259,
+      "step": 41895
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3567787218905867,
+      "learning_rate": 8.804961427802961e-09,
+      "loss": 0.6483,
+      "step": 41896
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5669891236687428,
+      "learning_rate": 8.772977171682684e-09,
+      "loss": 0.4955,
+      "step": 41897
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4173506872048507,
+      "learning_rate": 8.741051087650665e-09,
+      "loss": 0.7014,
+      "step": 41898
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2479110206218498,
+      "learning_rate": 8.70918317589009e-09,
+      "loss": 0.5057,
+      "step": 41899
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5101651284480553,
+      "learning_rate": 8.677373436588588e-09,
+      "loss": 0.7227,
+      "step": 41900
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5919613827162464,
+      "learning_rate": 8.645621869930453e-09,
+      "loss": 0.5063,
+      "step": 41901
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4828234253541388,
+      "learning_rate": 8.613928476101097e-09,
+      "loss": 0.6789,
+      "step": 41902
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3091036088873305,
+      "learning_rate": 8.582293255284813e-09,
+      "loss": 0.6745,
+      "step": 41903
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2612137669818164,
+      "learning_rate": 8.550716207664788e-09,
+      "loss": 0.7594,
+      "step": 41904
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3418175467098017,
+      "learning_rate": 8.519197333426431e-09,
+      "loss": 0.7495,
+      "step": 41905
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4992368255885427,
+      "learning_rate": 8.48773663275182e-09,
+      "loss": 0.7991,
+      "step": 41906
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.315704806149029,
+      "learning_rate": 8.456334105825248e-09,
+      "loss": 0.5769,
+      "step": 41907
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6090595199285962,
+      "learning_rate": 8.424989752829904e-09,
+      "loss": 0.5005,
+      "step": 41908
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4556876002814896,
+      "learning_rate": 8.393703573946754e-09,
+      "loss": 0.6894,
+      "step": 41909
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3619737609866707,
+      "learning_rate": 8.362475569358985e-09,
+      "loss": 0.6063,
+      "step": 41910
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.998177815741326,
+      "learning_rate": 8.331305739247565e-09,
+      "loss": 0.6633,
+      "step": 41911
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3732556309061994,
+      "learning_rate": 8.300194083795676e-09,
+      "loss": 0.6123,
+      "step": 41912
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5255874460713708,
+      "learning_rate": 8.269140603182069e-09,
+      "loss": 0.6354,
+      "step": 41913
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.106540177285123,
+      "learning_rate": 8.238145297589927e-09,
+      "loss": 0.5409,
+      "step": 41914
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3932559638472035,
+      "learning_rate": 8.207208167198e-09,
+      "loss": 0.5637,
+      "step": 41915
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4066996345653406,
+      "learning_rate": 8.176329212187251e-09,
+      "loss": 0.7221,
+      "step": 41916
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5033319610849964,
+      "learning_rate": 8.145508432737537e-09,
+      "loss": 0.7558,
+      "step": 41917
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5194941337803638,
+      "learning_rate": 8.114745829027604e-09,
+      "loss": 0.6254,
+      "step": 41918
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.350597181246667,
+      "learning_rate": 8.084041401237308e-09,
+      "loss": 0.753,
+      "step": 41919
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5410213413694553,
+      "learning_rate": 8.053395149545395e-09,
+      "loss": 0.6005,
+      "step": 41920
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.442231714206358,
+      "learning_rate": 8.022807074129501e-09,
+      "loss": 0.7652,
+      "step": 41921
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.592759574983519,
+      "learning_rate": 7.992277175168373e-09,
+      "loss": 0.5429,
+      "step": 41922
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8379421170860102,
+      "learning_rate": 7.961805452839644e-09,
+      "loss": 0.7017,
+      "step": 41923
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6720428048004874,
+      "learning_rate": 7.931391907320951e-09,
+      "loss": 0.6583,
+      "step": 41924
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0718162787170824,
+      "learning_rate": 7.901036538788819e-09,
+      "loss": 0.7091,
+      "step": 41925
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3883887708893163,
+      "learning_rate": 7.870739347420886e-09,
+      "loss": 0.6347,
+      "step": 41926
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5773996994017663,
+      "learning_rate": 7.840500333392565e-09,
+      "loss": 0.5748,
+      "step": 41927
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6819931019962473,
+      "learning_rate": 7.81031949688038e-09,
+      "loss": 0.7426,
+      "step": 41928
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.581699777024273,
+      "learning_rate": 7.78019683805975e-09,
+      "loss": 0.6475,
+      "step": 41929
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5784311839639582,
+      "learning_rate": 7.750132357106089e-09,
+      "loss": 0.7437,
+      "step": 41930
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4519271015713098,
+      "learning_rate": 7.72012605419481e-09,
+      "loss": 0.615,
+      "step": 41931
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5811035740246273,
+      "learning_rate": 7.69017792950022e-09,
+      "loss": 0.4597,
+      "step": 41932
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.353660230657997,
+      "learning_rate": 7.660287983197735e-09,
+      "loss": 0.7463,
+      "step": 41933
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4908459661219051,
+      "learning_rate": 7.630456215460546e-09,
+      "loss": 0.6052,
+      "step": 41934
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.2615128562185105,
+      "learning_rate": 7.600682626461853e-09,
+      "loss": 0.6546,
+      "step": 41935
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.448779968969193,
+      "learning_rate": 7.570967216374847e-09,
+      "loss": 0.5323,
+      "step": 41936
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8326608164778617,
+      "learning_rate": 7.541309985373834e-09,
+      "loss": 0.6055,
+      "step": 41937
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4398707753178202,
+      "learning_rate": 7.51171093363201e-09,
+      "loss": 0.6769,
+      "step": 41938
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8099041038323218,
+      "learning_rate": 7.482170061319238e-09,
+      "loss": 0.8417,
+      "step": 41939
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4254243879313022,
+      "learning_rate": 7.452687368608713e-09,
+      "loss": 0.7054,
+      "step": 41940
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4508123997478573,
+      "learning_rate": 7.42326285567363e-09,
+      "loss": 0.7277,
+      "step": 41941
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6571601154500633,
+      "learning_rate": 7.393896522682742e-09,
+      "loss": 0.7277,
+      "step": 41942
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4176787266048985,
+      "learning_rate": 7.364588369808135e-09,
+      "loss": 0.7603,
+      "step": 41943
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3711055961384804,
+      "learning_rate": 7.335338397220782e-09,
+      "loss": 0.7049,
+      "step": 41944
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6226772382758257,
+      "learning_rate": 7.306146605090547e-09,
+      "loss": 0.5333,
+      "step": 41945
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3315908071861802,
+      "learning_rate": 7.277012993588406e-09,
+      "loss": 0.6205,
+      "step": 41946
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.298708995665526,
+      "learning_rate": 7.247937562882001e-09,
+      "loss": 0.6751,
+      "step": 41947
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.425329660296711,
+      "learning_rate": 7.218920313142308e-09,
+      "loss": 0.7098,
+      "step": 41948
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3218761972776596,
+      "learning_rate": 7.1899612445380795e-09,
+      "loss": 0.7824,
+      "step": 41949
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.318647239646912,
+      "learning_rate": 7.16106035723696e-09,
+      "loss": 0.7481,
+      "step": 41950
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5749608518945766,
+      "learning_rate": 7.132217651407702e-09,
+      "loss": 0.6089,
+      "step": 41951
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.427876902996519,
+      "learning_rate": 7.103433127219061e-09,
+      "loss": 0.5883,
+      "step": 41952
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.365780701905294,
+      "learning_rate": 7.074706784837571e-09,
+      "loss": 0.7177,
+      "step": 41953
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6255627509999906,
+      "learning_rate": 7.0460386244308735e-09,
+      "loss": 0.7801,
+      "step": 41954
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4427150531003274,
+      "learning_rate": 7.017428646165503e-09,
+      "loss": 0.6799,
+      "step": 41955
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4773426356322756,
+      "learning_rate": 6.988876850209103e-09,
+      "loss": 0.699,
+      "step": 41956
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6697604309871354,
+      "learning_rate": 6.960383236727098e-09,
+      "loss": 0.7657,
+      "step": 41957
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4028926172021325,
+      "learning_rate": 6.931947805884909e-09,
+      "loss": 0.5989,
+      "step": 41958
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5952623273826911,
+      "learning_rate": 6.903570557849071e-09,
+      "loss": 0.4678,
+      "step": 41959
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.290710540890448,
+      "learning_rate": 6.875251492783897e-09,
+      "loss": 0.6768,
+      "step": 41960
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9619151896549578,
+      "learning_rate": 6.8469906108559195e-09,
+      "loss": 0.6665,
+      "step": 41961
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5777725404051572,
+      "learning_rate": 6.8187879122272315e-09,
+      "loss": 0.6177,
+      "step": 41962
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.386083488899832,
+      "learning_rate": 6.790643397064367e-09,
+      "loss": 0.6874,
+      "step": 41963
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5934308159258784,
+      "learning_rate": 6.762557065529418e-09,
+      "loss": 0.5611,
+      "step": 41964
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3612440769487704,
+      "learning_rate": 6.734528917786698e-09,
+      "loss": 0.5621,
+      "step": 41965
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2765357837677767,
+      "learning_rate": 6.706558953999409e-09,
+      "loss": 0.5731,
+      "step": 41966
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3333943324946431,
+      "learning_rate": 6.678647174330755e-09,
+      "loss": 0.6961,
+      "step": 41967
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5835096652642806,
+      "learning_rate": 6.650793578942827e-09,
+      "loss": 0.4696,
+      "step": 41968
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6648158954277357,
+      "learning_rate": 6.622998167997718e-09,
+      "loss": 0.5907,
+      "step": 41969
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6010150674820716,
+      "learning_rate": 6.595260941657522e-09,
+      "loss": 0.6231,
+      "step": 41970
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5187035853861293,
+      "learning_rate": 6.56758190008322e-09,
+      "loss": 0.7467,
+      "step": 41971
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5358226871873102,
+      "learning_rate": 6.539961043436904e-09,
+      "loss": 0.693,
+      "step": 41972
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.430984800257069,
+      "learning_rate": 6.512398371877338e-09,
+      "loss": 0.7455,
+      "step": 41973
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3946101724974314,
+      "learning_rate": 6.484893885567722e-09,
+      "loss": 0.5709,
+      "step": 41974
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3417920301183677,
+      "learning_rate": 6.457447584666821e-09,
+      "loss": 0.6554,
+      "step": 41975
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2488148140727386,
+      "learning_rate": 6.430059469334504e-09,
+      "loss": 0.7088,
+      "step": 41976
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.615898218866742,
+      "learning_rate": 6.4027295397306455e-09,
+      "loss": 0.6137,
+      "step": 41977
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2837073261234027,
+      "learning_rate": 6.375457796014006e-09,
+      "loss": 0.6391,
+      "step": 41978
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4308969517964174,
+      "learning_rate": 6.3482442383422385e-09,
+      "loss": 0.7157,
+      "step": 41979
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6716382690420488,
+      "learning_rate": 6.3210888668763235e-09,
+      "loss": 0.585,
+      "step": 41980
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3892166855067225,
+      "learning_rate": 6.293991681771694e-09,
+      "loss": 0.6235,
+      "step": 41981
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.342513363642713,
+      "learning_rate": 6.266952683188221e-09,
+      "loss": 0.6491,
+      "step": 41982
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1873693935142944,
+      "learning_rate": 6.2399718712824465e-09,
+      "loss": 0.6091,
+      "step": 41983
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4319728816583641,
+      "learning_rate": 6.2130492462120214e-09,
+      "loss": 0.8001,
+      "step": 41984
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.184125521109437,
+      "learning_rate": 6.186184808132378e-09,
+      "loss": 0.5493,
+      "step": 41985
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6257633822502455,
+      "learning_rate": 6.159378557201168e-09,
+      "loss": 0.7189,
+      "step": 41986
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5736277984889956,
+      "learning_rate": 6.1326304935738215e-09,
+      "loss": 0.4701,
+      "step": 41987
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5980786869480228,
+      "learning_rate": 6.10594061740577e-09,
+      "loss": 0.6574,
+      "step": 41988
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5771833661264376,
+      "learning_rate": 6.0793089288535555e-09,
+      "loss": 0.4549,
+      "step": 41989
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7316905356888754,
+      "learning_rate": 6.052735428071499e-09,
+      "loss": 0.6299,
+      "step": 41990
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.033503686789355,
+      "learning_rate": 6.0262201152150315e-09,
+      "loss": 0.6125,
+      "step": 41991
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6237914748948686,
+      "learning_rate": 5.999762990437363e-09,
+      "loss": 0.5379,
+      "step": 41992
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4469387308298545,
+      "learning_rate": 5.973364053892816e-09,
+      "loss": 0.6754,
+      "step": 41993
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.631537536766716,
+      "learning_rate": 5.947023305735711e-09,
+      "loss": 0.648,
+      "step": 41994
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.420461781898453,
+      "learning_rate": 5.920740746119258e-09,
+      "loss": 0.5669,
+      "step": 41995
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4260742073303003,
+      "learning_rate": 5.894516375196668e-09,
+      "loss": 0.6742,
+      "step": 41996
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4499100640191536,
+      "learning_rate": 5.868350193120043e-09,
+      "loss": 0.7865,
+      "step": 41997
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.387079220195242,
+      "learning_rate": 5.842242200041481e-09,
+      "loss": 0.6707,
+      "step": 41998
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4306372212188376,
+      "learning_rate": 5.816192396114195e-09,
+      "loss": 0.5582,
+      "step": 41999
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3316517615685757,
+      "learning_rate": 5.790200781488065e-09,
+      "loss": 0.6474,
+      "step": 42000
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3502017250099136,
+      "learning_rate": 5.7642673563163e-09,
+      "loss": 0.6944,
+      "step": 42001
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6558405123330373,
+      "learning_rate": 5.738392120749892e-09,
+      "loss": 0.5075,
+      "step": 42002
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.338236950271016,
+      "learning_rate": 5.712575074937609e-09,
+      "loss": 0.6418,
+      "step": 42003
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3682883125678846,
+      "learning_rate": 5.686816219030444e-09,
+      "loss": 0.6699,
+      "step": 42004
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4356275928885487,
+      "learning_rate": 5.661115553179386e-09,
+      "loss": 0.6388,
+      "step": 42005
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7162018221124367,
+      "learning_rate": 5.6354730775332046e-09,
+      "loss": 0.7141,
+      "step": 42006
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5701839521974068,
+      "learning_rate": 5.60988879224178e-09,
+      "loss": 0.6588,
+      "step": 42007
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7184342531357697,
+      "learning_rate": 5.584362697453882e-09,
+      "loss": 0.7786,
+      "step": 42008
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7109122911490675,
+      "learning_rate": 5.558894793318281e-09,
+      "loss": 0.6441,
+      "step": 42009
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2627050643446234,
+      "learning_rate": 5.533485079982637e-09,
+      "loss": 0.7209,
+      "step": 42010
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2572358492008542,
+      "learning_rate": 5.508133557594608e-09,
+      "loss": 0.7517,
+      "step": 42011
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4875790317799973,
+      "learning_rate": 5.482840226302966e-09,
+      "loss": 0.7522,
+      "step": 42012
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3693700908381945,
+      "learning_rate": 5.457605086254258e-09,
+      "loss": 0.5838,
+      "step": 42013
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8878203306366272,
+      "learning_rate": 5.4324281375961466e-09,
+      "loss": 0.753,
+      "step": 42014
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4075047151301296,
+      "learning_rate": 5.4073093804740685e-09,
+      "loss": 0.7441,
+      "step": 42015
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1597305542408145,
+      "learning_rate": 5.382248815034574e-09,
+      "loss": 0.5733,
+      "step": 42016
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3060873135528297,
+      "learning_rate": 5.357246441424213e-09,
+      "loss": 0.8192,
+      "step": 42017
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5474166492440826,
+      "learning_rate": 5.332302259788424e-09,
+      "loss": 0.5055,
+      "step": 42018
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.294054198721937,
+      "learning_rate": 5.307416270272647e-09,
+      "loss": 0.7256,
+      "step": 42019
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3911229004628611,
+      "learning_rate": 5.282588473021211e-09,
+      "loss": 0.6302,
+      "step": 42020
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5355322212464304,
+      "learning_rate": 5.257818868178444e-09,
+      "loss": 0.7612,
+      "step": 42021
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4487856123328329,
+      "learning_rate": 5.2331074558886756e-09,
+      "loss": 0.6542,
+      "step": 42022
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1841896572063786,
+      "learning_rate": 5.208454236296234e-09,
+      "loss": 0.6686,
+      "step": 42023
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3661481682131085,
+      "learning_rate": 5.18385920954434e-09,
+      "loss": 0.7266,
+      "step": 42024
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4371763524163517,
+      "learning_rate": 5.1593223757762104e-09,
+      "loss": 0.6554,
+      "step": 42025
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5077124618856697,
+      "learning_rate": 5.134843735135065e-09,
+      "loss": 0.7505,
+      "step": 42026
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3346731627052992,
+      "learning_rate": 5.110423287764121e-09,
+      "loss": 0.6437,
+      "step": 42027
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5856092155669632,
+      "learning_rate": 5.086061033803269e-09,
+      "loss": 0.4827,
+      "step": 42028
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.894039644908435,
+      "learning_rate": 5.061756973396836e-09,
+      "loss": 0.7734,
+      "step": 42029
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.366624156478617,
+      "learning_rate": 5.037511106684712e-09,
+      "loss": 0.7103,
+      "step": 42030
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.253611018388241,
+      "learning_rate": 5.013323433807893e-09,
+      "loss": 0.6731,
+      "step": 42031
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.363190240596634,
+      "learning_rate": 4.989193954908489e-09,
+      "loss": 0.6298,
+      "step": 42032
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4245355057272466,
+      "learning_rate": 4.9651226701263874e-09,
+      "loss": 0.7614,
+      "step": 42033
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3562415348745496,
+      "learning_rate": 4.9411095796014775e-09,
+      "loss": 0.6318,
+      "step": 42034
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7632512089711079,
+      "learning_rate": 4.917154683473646e-09,
+      "loss": 0.7628,
+      "step": 42035
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4025629941093556,
+      "learning_rate": 4.8932579818827816e-09,
+      "loss": 0.6645,
+      "step": 42036
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.538715030161854,
+      "learning_rate": 4.869419474967663e-09,
+      "loss": 0.7348,
+      "step": 42037
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6384595210967083,
+      "learning_rate": 4.845639162868177e-09,
+      "loss": 0.6052,
+      "step": 42038
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.289552063279596,
+      "learning_rate": 4.821917045720881e-09,
+      "loss": 0.6697,
+      "step": 42039
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6510566938415525,
+      "learning_rate": 4.7982531236645535e-09,
+      "loss": 0.7289,
+      "step": 42040
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3451371457374561,
+      "learning_rate": 4.774647396837973e-09,
+      "loss": 0.5342,
+      "step": 42041
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.359784663265331,
+      "learning_rate": 4.751099865376585e-09,
+      "loss": 0.6844,
+      "step": 42042
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.408259889085475,
+      "learning_rate": 4.727610529419169e-09,
+      "loss": 0.5967,
+      "step": 42043
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5155078657602221,
+      "learning_rate": 4.704179389103391e-09,
+      "loss": 0.8019,
+      "step": 42044
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3680844288663128,
+      "learning_rate": 4.6808064445635905e-09,
+      "loss": 0.6571,
+      "step": 42045
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.309017113590339,
+      "learning_rate": 4.6574916959363226e-09,
+      "loss": 0.7184,
+      "step": 42046
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5833113859561232,
+      "learning_rate": 4.634235143357036e-09,
+      "loss": 0.6511,
+      "step": 42047
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3001507184571455,
+      "learning_rate": 4.611036786962286e-09,
+      "loss": 0.6126,
+      "step": 42048
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4581629551361979,
+      "learning_rate": 4.587896626886412e-09,
+      "loss": 0.7487,
+      "step": 42049
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4778323966398583,
+      "learning_rate": 4.5648146632648605e-09,
+      "loss": 0.5242,
+      "step": 42050
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9449239577939508,
+      "learning_rate": 4.541790896230858e-09,
+      "loss": 0.713,
+      "step": 42051
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5581754557888736,
+      "learning_rate": 4.518825325919851e-09,
+      "loss": 0.7551,
+      "step": 42052
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4473833097270958,
+      "learning_rate": 4.495917952463957e-09,
+      "loss": 0.6166,
+      "step": 42053
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4758738019262498,
+      "learning_rate": 4.473068775997513e-09,
+      "loss": 0.6273,
+      "step": 42054
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.630840154135469,
+      "learning_rate": 4.450277796653746e-09,
+      "loss": 0.6795,
+      "step": 42055
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3658965787958588,
+      "learning_rate": 4.427545014564771e-09,
+      "loss": 0.7408,
+      "step": 42056
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2870618404807912,
+      "learning_rate": 4.404870429862706e-09,
+      "loss": 0.6651,
+      "step": 42057
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.558160378939757,
+      "learning_rate": 4.382254042680778e-09,
+      "loss": 0.515,
+      "step": 42058
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3459572742596189,
+      "learning_rate": 4.359695853149992e-09,
+      "loss": 0.5989,
+      "step": 42059
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6212871972072234,
+      "learning_rate": 4.337195861400245e-09,
+      "loss": 0.5112,
+      "step": 42060
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4274067089630789,
+      "learning_rate": 4.314754067565874e-09,
+      "loss": 0.6676,
+      "step": 42061
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4414326409465439,
+      "learning_rate": 4.292370471774554e-09,
+      "loss": 0.8078,
+      "step": 42062
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5027428443522588,
+      "learning_rate": 4.270045074157292e-09,
+      "loss": 0.6837,
+      "step": 42063
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2717431338175496,
+      "learning_rate": 4.247777874845094e-09,
+      "loss": 0.6146,
+      "step": 42064
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5190416969729599,
+      "learning_rate": 4.225568873966746e-09,
+      "loss": 0.6274,
+      "step": 42065
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5121537464051011,
+      "learning_rate": 4.203418071652143e-09,
+      "loss": 0.7214,
+      "step": 42066
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.167446083509144,
+      "learning_rate": 4.181325468028963e-09,
+      "loss": 0.6534,
+      "step": 42067
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3521027223960276,
+      "learning_rate": 4.1592910632271e-09,
+      "loss": 0.6384,
+      "step": 42068
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3549552677040526,
+      "learning_rate": 4.13731485737423e-09,
+      "loss": 0.7519,
+      "step": 42069
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3714175869244507,
+      "learning_rate": 4.11539685059914e-09,
+      "loss": 0.5505,
+      "step": 42070
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3813365384957714,
+      "learning_rate": 4.093537043028395e-09,
+      "loss": 0.5013,
+      "step": 42071
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5951347376870513,
+      "learning_rate": 4.07173543478967e-09,
+      "loss": 0.5251,
+      "step": 42072
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2659322140797125,
+      "learning_rate": 4.04999202601064e-09,
+      "loss": 0.5863,
+      "step": 42073
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7341098318188888,
+      "learning_rate": 4.028306816816763e-09,
+      "loss": 0.6712,
+      "step": 42074
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2497915645971693,
+      "learning_rate": 4.006679807334601e-09,
+      "loss": 0.6051,
+      "step": 42075
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4415971735156776,
+      "learning_rate": 3.985110997690722e-09,
+      "loss": 0.7089,
+      "step": 42076
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.345206652986753,
+      "learning_rate": 3.963600388009469e-09,
+      "loss": 0.6092,
+      "step": 42077
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3378229995916597,
+      "learning_rate": 3.94214797841741e-09,
+      "loss": 0.686,
+      "step": 42078
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2995969444572755,
+      "learning_rate": 3.920753769038888e-09,
+      "loss": 0.7082,
+      "step": 42079
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7602955984529818,
+      "learning_rate": 3.899417759998248e-09,
+      "loss": 0.7298,
+      "step": 42080
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3761879162357822,
+      "learning_rate": 3.878139951419835e-09,
+      "loss": 0.5456,
+      "step": 42081
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5931372628980744,
+      "learning_rate": 3.856920343427994e-09,
+      "loss": 0.6668,
+      "step": 42082
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3343918585193912,
+      "learning_rate": 3.835758936145961e-09,
+      "loss": 0.6931,
+      "step": 42083
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1234439055298315,
+      "learning_rate": 3.81465572969586e-09,
+      "loss": 0.644,
+      "step": 42084
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5509601925158325,
+      "learning_rate": 3.793610724202035e-09,
+      "loss": 0.7536,
+      "step": 42085
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3508972787783144,
+      "learning_rate": 3.772623919786611e-09,
+      "loss": 0.5703,
+      "step": 42086
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2726727613294548,
+      "learning_rate": 3.751695316572823e-09,
+      "loss": 0.6864,
+      "step": 42087
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5495354352074466,
+      "learning_rate": 3.730824914679465e-09,
+      "loss": 0.6746,
+      "step": 42088
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5330660837349004,
+      "learning_rate": 3.7100127142308816e-09,
+      "loss": 0.5327,
+      "step": 42089
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.906686281875011,
+      "learning_rate": 3.6892587153480874e-09,
+      "loss": 0.654,
+      "step": 42090
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.475792580553723,
+      "learning_rate": 3.6685629181509864e-09,
+      "loss": 0.669,
+      "step": 42091
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2064739956654706,
+      "learning_rate": 3.6479253227594825e-09,
+      "loss": 0.533,
+      "step": 42092
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.7547705921576884,
+      "learning_rate": 3.62734592929459e-09,
+      "loss": 0.5719,
+      "step": 42093
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3287159341274746,
+      "learning_rate": 3.6068247378773236e-09,
+      "loss": 0.6762,
+      "step": 42094
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5338499848054554,
+      "learning_rate": 3.5863617486242565e-09,
+      "loss": 0.6079,
+      "step": 42095
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5566102118068326,
+      "learning_rate": 3.565956961657513e-09,
+      "loss": 0.6026,
+      "step": 42096
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6169953954791146,
+      "learning_rate": 3.5456103770936668e-09,
+      "loss": 0.4838,
+      "step": 42097
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6180064591010386,
+      "learning_rate": 3.525321995052622e-09,
+      "loss": 0.5688,
+      "step": 42098
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3692926272402264,
+      "learning_rate": 3.5050918156520618e-09,
+      "loss": 0.7488,
+      "step": 42099
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3671835195747402,
+      "learning_rate": 3.4849198390096705e-09,
+      "loss": 0.8063,
+      "step": 42100
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2443532053547226,
+      "learning_rate": 3.4648060652431314e-09,
+      "loss": 0.6749,
+      "step": 42101
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3163238919668185,
+      "learning_rate": 3.444750494469018e-09,
+      "loss": 0.6254,
+      "step": 42102
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5003361999436065,
+      "learning_rate": 3.4247531268039034e-09,
+      "loss": 0.7204,
+      "step": 42103
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3545550808330664,
+      "learning_rate": 3.404813962365472e-09,
+      "loss": 0.6905,
+      "step": 42104
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.051277186913893,
+      "learning_rate": 3.384933001269186e-09,
+      "loss": 0.6992,
+      "step": 42105
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3962533479606016,
+      "learning_rate": 3.3651102436293993e-09,
+      "loss": 0.6299,
+      "step": 42106
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3633490535821509,
+      "learning_rate": 3.345345689563795e-09,
+      "loss": 0.726,
+      "step": 42107
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3493319251801514,
+      "learning_rate": 3.325639339186726e-09,
+      "loss": 0.632,
+      "step": 42108
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5867455753331551,
+      "learning_rate": 3.3059911926125455e-09,
+      "loss": 0.4658,
+      "step": 42109
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.534828180668833,
+      "learning_rate": 3.2864012499556066e-09,
+      "loss": 0.7704,
+      "step": 42110
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4680969447434369,
+      "learning_rate": 3.266869511330262e-09,
+      "loss": 0.7472,
+      "step": 42111
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.550736646485965,
+      "learning_rate": 3.2473959768497542e-09,
+      "loss": 0.7061,
+      "step": 42112
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6208450285548118,
+      "learning_rate": 3.227980646628437e-09,
+      "loss": 0.52,
+      "step": 42113
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4453963082954593,
+      "learning_rate": 3.2086235207784423e-09,
+      "loss": 0.6403,
+      "step": 42114
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.516365620941477,
+      "learning_rate": 3.189324599413013e-09,
+      "loss": 0.6944,
+      "step": 42115
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3542167714291453,
+      "learning_rate": 3.170083882643171e-09,
+      "loss": 0.5421,
+      "step": 42116
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8969981538947307,
+      "learning_rate": 3.1509013705832703e-09,
+      "loss": 0.7402,
+      "step": 42117
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4809572033036218,
+      "learning_rate": 3.131777063343222e-09,
+      "loss": 0.7222,
+      "step": 42118
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3592630507807941,
+      "learning_rate": 3.1127109610351593e-09,
+      "loss": 0.5394,
+      "step": 42119
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6066098218021697,
+      "learning_rate": 3.0937030637701037e-09,
+      "loss": 0.5029,
+      "step": 42120
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4113664085898803,
+      "learning_rate": 3.074753371657968e-09,
+      "loss": 0.7255,
+      "step": 42121
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2928929069932813,
+      "learning_rate": 3.0558618848097744e-09,
+      "loss": 0.5511,
+      "step": 42122
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6016027021346975,
+      "learning_rate": 3.0370286033354348e-09,
+      "loss": 0.4925,
+      "step": 42123
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4662160301840197,
+      "learning_rate": 3.018253527343751e-09,
+      "loss": 0.6049,
+      "step": 42124
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.765220397714158,
+      "learning_rate": 2.999536656945745e-09,
+      "loss": 0.7491,
+      "step": 42125
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7780257443209861,
+      "learning_rate": 2.980877992249109e-09,
+      "loss": 0.7365,
+      "step": 42126
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.194180948690256,
+      "learning_rate": 2.9622775333626453e-09,
+      "loss": 0.7426,
+      "step": 42127
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3553066993319989,
+      "learning_rate": 2.9437352803951546e-09,
+      "loss": 0.581,
+      "step": 42128
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3827440311208492,
+      "learning_rate": 2.9252512334543293e-09,
+      "loss": 0.5702,
+      "step": 42129
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9239604556115235,
+      "learning_rate": 2.9068253926478605e-09,
+      "loss": 0.8438,
+      "step": 42130
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3518520267709262,
+      "learning_rate": 2.8884577580823304e-09,
+      "loss": 0.6743,
+      "step": 42131
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4649073541760245,
+      "learning_rate": 2.87014832986543e-09,
+      "loss": 0.6929,
+      "step": 42132
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5287104301248011,
+      "learning_rate": 2.8518971081037403e-09,
+      "loss": 0.6633,
+      "step": 42133
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4233012078015919,
+      "learning_rate": 2.8337040929027336e-09,
+      "loss": 0.7181,
+      "step": 42134
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3318108109975257,
+      "learning_rate": 2.8155692843689907e-09,
+      "loss": 0.6385,
+      "step": 42135
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.606490569578698,
+      "learning_rate": 2.797492682609093e-09,
+      "loss": 0.8394,
+      "step": 42136
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4962656245548296,
+      "learning_rate": 2.7794742877262914e-09,
+      "loss": 0.7328,
+      "step": 42137
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7155172083388825,
+      "learning_rate": 2.761514099826057e-09,
+      "loss": 0.6515,
+      "step": 42138
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.565316672632872,
+      "learning_rate": 2.743612119014971e-09,
+      "loss": 0.688,
+      "step": 42139
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4307795731540012,
+      "learning_rate": 2.725768345394064e-09,
+      "loss": 0.6449,
+      "step": 42140
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3432094210477754,
+      "learning_rate": 2.7079827790699177e-09,
+      "loss": 0.7969,
+      "step": 42141
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5441381184711156,
+      "learning_rate": 2.690255420144672e-09,
+      "loss": 0.7336,
+      "step": 42142
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6577843429504833,
+      "learning_rate": 2.6725862687215774e-09,
+      "loss": 0.6995,
+      "step": 42143
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.089857189620445,
+      "learning_rate": 2.6549753249038858e-09,
+      "loss": 0.7816,
+      "step": 42144
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3197503047928316,
+      "learning_rate": 2.6374225887937365e-09,
+      "loss": 0.73,
+      "step": 42145
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.530305074477787,
+      "learning_rate": 2.619928060494381e-09,
+      "loss": 0.6295,
+      "step": 42146
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3481607471347086,
+      "learning_rate": 2.6024917401068495e-09,
+      "loss": 0.6614,
+      "step": 42147
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2905370540216148,
+      "learning_rate": 2.585113627732172e-09,
+      "loss": 0.6851,
+      "step": 42148
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8747429403904885,
+      "learning_rate": 2.567793723471379e-09,
+      "loss": 0.6362,
+      "step": 42149
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6053889105031367,
+      "learning_rate": 2.550532027427721e-09,
+      "loss": 0.5115,
+      "step": 42150
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6635882461462819,
+      "learning_rate": 2.533328539698898e-09,
+      "loss": 0.7018,
+      "step": 42151
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7362467161371093,
+      "learning_rate": 2.5161832603870506e-09,
+      "loss": 0.5587,
+      "step": 42152
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6372094857462749,
+      "learning_rate": 2.499096189589878e-09,
+      "loss": 0.5148,
+      "step": 42153
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4119859771836878,
+      "learning_rate": 2.4820673274095207e-09,
+      "loss": 0.5725,
+      "step": 42154
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4762505059856228,
+      "learning_rate": 2.4650966739425686e-09,
+      "loss": 0.5873,
+      "step": 42155
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.685046391757633,
+      "learning_rate": 2.4481842292900517e-09,
+      "loss": 0.6435,
+      "step": 42156
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.646463858341839,
+      "learning_rate": 2.431329993548559e-09,
+      "loss": 0.588,
+      "step": 42157
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.224670744001627,
+      "learning_rate": 2.414533966816901e-09,
+      "loss": 0.6528,
+      "step": 42158
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4691567851087075,
+      "learning_rate": 2.3977961491938874e-09,
+      "loss": 0.5776,
+      "step": 42159
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.407113226596183,
+      "learning_rate": 2.381116540774997e-09,
+      "loss": 0.672,
+      "step": 42160
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5216321284574483,
+      "learning_rate": 2.36449514165904e-09,
+      "loss": 0.6676,
+      "step": 42161
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6024148885807115,
+      "learning_rate": 2.347931951942606e-09,
+      "loss": 0.5164,
+      "step": 42162
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5263742712839459,
+      "learning_rate": 2.3314269717211735e-09,
+      "loss": 0.573,
+      "step": 42163
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5786706115255902,
+      "learning_rate": 2.3149802010913323e-09,
+      "loss": 0.758,
+      "step": 42164
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4648252523466287,
+      "learning_rate": 2.298591640149672e-09,
+      "loss": 0.7639,
+      "step": 42165
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.407915441593997,
+      "learning_rate": 2.2822612889894513e-09,
+      "loss": 0.6644,
+      "step": 42166
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4069396768246343,
+      "learning_rate": 2.2659891477083694e-09,
+      "loss": 0.5785,
+      "step": 42167
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3222052449196273,
+      "learning_rate": 2.2497752163996855e-09,
+      "loss": 0.6622,
+      "step": 42168
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.700570184781972,
+      "learning_rate": 2.2336194951577685e-09,
+      "loss": 0.7438,
+      "step": 42169
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4147826048583931,
+      "learning_rate": 2.217521984076987e-09,
+      "loss": 0.5749,
+      "step": 42170
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7252558182165505,
+      "learning_rate": 2.2014826832506e-09,
+      "loss": 0.6981,
+      "step": 42171
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3370264066356161,
+      "learning_rate": 2.185501592772976e-09,
+      "loss": 0.632,
+      "step": 42172
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.511115100875623,
+      "learning_rate": 2.1695787127362644e-09,
+      "loss": 0.815,
+      "step": 42173
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.478303434604903,
+      "learning_rate": 2.1537140432348335e-09,
+      "loss": 0.6692,
+      "step": 42174
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6473443905302114,
+      "learning_rate": 2.1379075843586116e-09,
+      "loss": 0.6696,
+      "step": 42175
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6012509896474532,
+      "learning_rate": 2.1221593362008575e-09,
+      "loss": 0.5137,
+      "step": 42176
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.374713137592779,
+      "learning_rate": 2.1064692988526093e-09,
+      "loss": 0.6696,
+      "step": 42177
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5151515284149333,
+      "learning_rate": 2.0908374724060156e-09,
+      "loss": 0.6579,
+      "step": 42178
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2627420907462863,
+      "learning_rate": 2.0752638569521144e-09,
+      "loss": 0.6969,
+      "step": 42179
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2185385345471027,
+      "learning_rate": 2.0597484525819446e-09,
+      "loss": 0.6366,
+      "step": 42180
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2187386958825164,
+      "learning_rate": 2.0442912593843233e-09,
+      "loss": 0.6359,
+      "step": 42181
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.573569310344007,
+      "learning_rate": 2.0288922774502896e-09,
+      "loss": 0.5848,
+      "step": 42182
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4945457820971304,
+      "learning_rate": 2.013551506869771e-09,
+      "loss": 0.766,
+      "step": 42183
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4338234834107149,
+      "learning_rate": 1.9982689477315854e-09,
+      "loss": 0.667,
+      "step": 42184
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6130149258149477,
+      "learning_rate": 1.9830446001245507e-09,
+      "loss": 0.4903,
+      "step": 42185
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3822240271222639,
+      "learning_rate": 1.967878464137485e-09,
+      "loss": 0.5815,
+      "step": 42186
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3351423193566232,
+      "learning_rate": 1.9527705398592056e-09,
+      "loss": 0.6529,
+      "step": 42187
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3457954803092786,
+      "learning_rate": 1.9377208273763105e-09,
+      "loss": 0.6299,
+      "step": 42188
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5977607238716292,
+      "learning_rate": 1.9227293267787272e-09,
+      "loss": 0.7464,
+      "step": 42189
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.395745258121738,
+      "learning_rate": 1.9077960381508333e-09,
+      "loss": 0.5828,
+      "step": 42190
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4078977016321914,
+      "learning_rate": 1.892920961582556e-09,
+      "loss": 0.6294,
+      "step": 42191
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4583077315178004,
+      "learning_rate": 1.878104097158273e-09,
+      "loss": 0.7178,
+      "step": 42192
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4892912086959587,
+      "learning_rate": 1.8633454449645814e-09,
+      "loss": 0.6313,
+      "step": 42193
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3381700646482213,
+      "learning_rate": 1.8486450050880788e-09,
+      "loss": 0.6193,
+      "step": 42194
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.614862672517805,
+      "learning_rate": 1.8340027776142521e-09,
+      "loss": 0.5109,
+      "step": 42195
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.277898408658872,
+      "learning_rate": 1.8194187626285887e-09,
+      "loss": 0.7371,
+      "step": 42196
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4478034883539606,
+      "learning_rate": 1.8048929602154652e-09,
+      "loss": 0.683,
+      "step": 42197
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4915592375610356,
+      "learning_rate": 1.790425370459259e-09,
+      "loss": 0.7444,
+      "step": 42198
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.37274185576872,
+      "learning_rate": 1.7760159934454568e-09,
+      "loss": 0.7169,
+      "step": 42199
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7764048740783815,
+      "learning_rate": 1.7616648292562154e-09,
+      "loss": 0.7263,
+      "step": 42200
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6430728556901775,
+      "learning_rate": 1.7473718779759118e-09,
+      "loss": 0.7825,
+      "step": 42201
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3247383204409147,
+      "learning_rate": 1.733137139688923e-09,
+      "loss": 0.6431,
+      "step": 42202
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0315482128580404,
+      "learning_rate": 1.718960614476295e-09,
+      "loss": 0.7048,
+      "step": 42203
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3455432285605151,
+      "learning_rate": 1.7048423024201843e-09,
+      "loss": 0.5802,
+      "step": 42204
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3837686134969682,
+      "learning_rate": 1.6907822036049682e-09,
+      "loss": 0.5399,
+      "step": 42205
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.57351310443617,
+      "learning_rate": 1.6767803181116927e-09,
+      "loss": 0.5005,
+      "step": 42206
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1842582571253226,
+      "learning_rate": 1.6628366460214041e-09,
+      "loss": 0.7162,
+      "step": 42207
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3103154909396255,
+      "learning_rate": 1.6489511874151488e-09,
+      "loss": 0.5123,
+      "step": 42208
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3382459252146803,
+      "learning_rate": 1.6351239423739728e-09,
+      "loss": 0.572,
+      "step": 42209
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5435138050159147,
+      "learning_rate": 1.621354910978923e-09,
+      "loss": 0.6857,
+      "step": 42210
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2516863095954969,
+      "learning_rate": 1.6076440933099347e-09,
+      "loss": 0.6191,
+      "step": 42211
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3516024064647458,
+      "learning_rate": 1.5939914894469444e-09,
+      "loss": 0.7466,
+      "step": 42212
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.955301717319953,
+      "learning_rate": 1.580397099468778e-09,
+      "loss": 0.7254,
+      "step": 42213
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8859930138330565,
+      "learning_rate": 1.5668609234542608e-09,
+      "loss": 0.7513,
+      "step": 42214
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8231812967948655,
+      "learning_rate": 1.5533829614844397e-09,
+      "loss": 0.6758,
+      "step": 42215
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.543559099141698,
+      "learning_rate": 1.5399632136348096e-09,
+      "loss": 0.7406,
+      "step": 42216
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5613026969270506,
+      "learning_rate": 1.526601679986417e-09,
+      "loss": 0.4841,
+      "step": 42217
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2841130696040302,
+      "learning_rate": 1.5132983606147566e-09,
+      "loss": 0.6479,
+      "step": 42218
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3720921499620415,
+      "learning_rate": 1.5000532555975444e-09,
+      "loss": 0.6306,
+      "step": 42219
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2899942101726272,
+      "learning_rate": 1.4868663650136062e-09,
+      "loss": 0.6234,
+      "step": 42220
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.231345144214907,
+      "learning_rate": 1.473737688938437e-09,
+      "loss": 0.6896,
+      "step": 42221
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4079831204111726,
+      "learning_rate": 1.460667227447532e-09,
+      "loss": 0.6612,
+      "step": 42222
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3849531550801357,
+      "learning_rate": 1.447654980619717e-09,
+      "loss": 0.6598,
+      "step": 42223
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5806093832050111,
+      "learning_rate": 1.434700948528267e-09,
+      "loss": 0.4566,
+      "step": 42224
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5364146188017878,
+      "learning_rate": 1.421805131248677e-09,
+      "loss": 0.6769,
+      "step": 42225
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.459389944515019,
+      "learning_rate": 1.4089675288586623e-09,
+      "loss": 0.6716,
+      "step": 42226
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5554635548661278,
+      "learning_rate": 1.3961881414292776e-09,
+      "loss": 0.6684,
+      "step": 42227
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2364695714661436,
+      "learning_rate": 1.3834669690382385e-09,
+      "loss": 0.7485,
+      "step": 42228
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4290283235026564,
+      "learning_rate": 1.3708040117577093e-09,
+      "loss": 0.8124,
+      "step": 42229
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4610759581962205,
+      "learning_rate": 1.358199269662075e-09,
+      "loss": 0.7715,
+      "step": 42230
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.340333247010491,
+      "learning_rate": 1.3456527428235e-09,
+      "loss": 0.7024,
+      "step": 42231
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6797134431707565,
+      "learning_rate": 1.3331644313174796e-09,
+      "loss": 0.6627,
+      "step": 42232
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3845542627999003,
+      "learning_rate": 1.320734335213958e-09,
+      "loss": 0.7505,
+      "step": 42233
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4324991076303724,
+      "learning_rate": 1.3083624545873197e-09,
+      "loss": 0.7622,
+      "step": 42234
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.438909923003465,
+      "learning_rate": 1.2960487895097295e-09,
+      "loss": 0.7683,
+      "step": 42235
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3140750675181256,
+      "learning_rate": 1.2837933400511315e-09,
+      "loss": 0.6399,
+      "step": 42236
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6606856154916134,
+      "learning_rate": 1.2715961062836902e-09,
+      "loss": 0.5143,
+      "step": 42237
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3949482702828904,
+      "learning_rate": 1.2594570882784595e-09,
+      "loss": 0.69,
+      "step": 42238
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6952131252090845,
+      "learning_rate": 1.2473762861064943e-09,
+      "loss": 0.573,
+      "step": 42239
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5964949611584027,
+      "learning_rate": 1.2353536998388483e-09,
+      "loss": 0.4712,
+      "step": 42240
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4387284717142932,
+      "learning_rate": 1.2233893295432453e-09,
+      "loss": 0.7234,
+      "step": 42241
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5087599985419726,
+      "learning_rate": 1.2114831752907397e-09,
+      "loss": 0.5982,
+      "step": 42242
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5117224728984482,
+      "learning_rate": 1.1996352371512755e-09,
+      "loss": 0.7507,
+      "step": 42243
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.724815109862047,
+      "learning_rate": 1.187845515192576e-09,
+      "loss": 0.6345,
+      "step": 42244
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4613578304548867,
+      "learning_rate": 1.1761140094834756e-09,
+      "loss": 0.6662,
+      "step": 42245
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.500720485771972,
+      "learning_rate": 1.1644407200939177e-09,
+      "loss": 0.7596,
+      "step": 42246
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4783981683977372,
+      "learning_rate": 1.152825647089406e-09,
+      "loss": 0.7515,
+      "step": 42247
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3962897852580316,
+      "learning_rate": 1.1412687905387743e-09,
+      "loss": 0.5626,
+      "step": 42248
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2843991516669855,
+      "learning_rate": 1.1297701505097459e-09,
+      "loss": 0.5628,
+      "step": 42249
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.706828467327101,
+      "learning_rate": 1.1183297270678239e-09,
+      "loss": 0.8159,
+      "step": 42250
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3190111622663314,
+      "learning_rate": 1.1069475202818425e-09,
+      "loss": 0.9008,
+      "step": 42251
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5085010974322048,
+      "learning_rate": 1.0956235302161944e-09,
+      "loss": 0.7281,
+      "step": 42252
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3454113127608873,
+      "learning_rate": 1.084357756937493e-09,
+      "loss": 0.7368,
+      "step": 42253
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4170461787699276,
+      "learning_rate": 1.0731502005112415e-09,
+      "loss": 0.6802,
+      "step": 42254
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8062283487079551,
+      "learning_rate": 1.062000861002943e-09,
+      "loss": 0.6021,
+      "step": 42255
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4305901010699866,
+      "learning_rate": 1.0509097384769906e-09,
+      "loss": 0.744,
+      "step": 42256
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4187510565366552,
+      "learning_rate": 1.039876832997777e-09,
+      "loss": 0.5736,
+      "step": 42257
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.424742215284564,
+      "learning_rate": 1.0289021446308057e-09,
+      "loss": 0.8381,
+      "step": 42258
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1663270926616485,
+      "learning_rate": 1.017985673439359e-09,
+      "loss": 0.57,
+      "step": 42259
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.437625596583135,
+      "learning_rate": 1.0071274194856096e-09,
+      "loss": 0.7295,
+      "step": 42260
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3430588794022134,
+      "learning_rate": 9.963273828350606e-10,
+      "loss": 0.8245,
+      "step": 42261
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2707905109965842,
+      "learning_rate": 9.855855635498845e-10,
+      "loss": 0.6453,
+      "step": 42262
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4893064512455485,
+      "learning_rate": 9.749019616911437e-10,
+      "loss": 0.7076,
+      "step": 42263
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1795922243943167,
+      "learning_rate": 9.642765773232309e-10,
+      "loss": 0.6437,
+      "step": 42264
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5322702283393328,
+      "learning_rate": 9.537094105060985e-10,
+      "loss": 0.7081,
+      "step": 42265
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.40463649139546,
+      "learning_rate": 9.432004613019186e-10,
+      "loss": 0.7357,
+      "step": 42266
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4931921374661141,
+      "learning_rate": 9.327497297728638e-10,
+      "loss": 0.7274,
+      "step": 42267
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.318682444203194,
+      "learning_rate": 9.223572159788863e-10,
+      "loss": 0.7238,
+      "step": 42268
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2632173705962693,
+      "learning_rate": 9.12022919979938e-10,
+      "loss": 0.753,
+      "step": 42269
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5809180542964565,
+      "learning_rate": 9.017468418370812e-10,
+      "loss": 0.4828,
+      "step": 42270
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3912909104439644,
+      "learning_rate": 8.915289816102679e-10,
+      "loss": 0.5976,
+      "step": 42271
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.357698567838284,
+      "learning_rate": 8.813693393583401e-10,
+      "loss": 0.6493,
+      "step": 42272
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5993434129236848,
+      "learning_rate": 8.712679151412495e-10,
+      "loss": 0.4871,
+      "step": 42273
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3127250206304097,
+      "learning_rate": 8.612247090167281e-10,
+      "loss": 0.6547,
+      "step": 42274
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9155826049062539,
+      "learning_rate": 8.512397210447276e-10,
+      "loss": 0.7725,
+      "step": 42275
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4122311342389726,
+      "learning_rate": 8.413129512818697e-10,
+      "loss": 0.6141,
+      "step": 42276
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2852094026336416,
+      "learning_rate": 8.314443997858857e-10,
+      "loss": 0.6472,
+      "step": 42277
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5920591867745473,
+      "learning_rate": 8.216340666156175e-10,
+      "loss": 0.505,
+      "step": 42278
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7804091654094465,
+      "learning_rate": 8.118819518276866e-10,
+      "loss": 0.6718,
+      "step": 42279
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.408804740728697,
+      "learning_rate": 8.021880554787143e-10,
+      "loss": 0.6312,
+      "step": 42280
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3842481955103487,
+      "learning_rate": 7.925523776242116e-10,
+      "loss": 0.7824,
+      "step": 42281
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3913893284434202,
+      "learning_rate": 7.829749183219104e-10,
+      "loss": 0.7221,
+      "step": 42282
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4039051119028292,
+      "learning_rate": 7.734556776273217e-10,
+      "loss": 0.7288,
+      "step": 42283
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6910854875075714,
+      "learning_rate": 7.639946555937361e-10,
+      "loss": 0.7282,
+      "step": 42284
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.222123730229755,
+      "learning_rate": 7.545918522788853e-10,
+      "loss": 0.6362,
+      "step": 42285
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3062422541729741,
+      "learning_rate": 7.452472677371702e-10,
+      "loss": 0.8445,
+      "step": 42286
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.471272211826234,
+      "learning_rate": 7.359609020218816e-10,
+      "loss": 0.712,
+      "step": 42287
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7594266070851141,
+      "learning_rate": 7.267327551863101e-10,
+      "loss": 0.5714,
+      "step": 42288
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.041852508854727,
+      "learning_rate": 7.175628272870772e-10,
+      "loss": 0.6477,
+      "step": 42289
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.392499366133736,
+      "learning_rate": 7.08451118375253e-10,
+      "loss": 0.5772,
+      "step": 42290
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.400289888113863,
+      "learning_rate": 6.993976285041282e-10,
+      "loss": 0.6276,
+      "step": 42291
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5913019624007684,
+      "learning_rate": 6.904023577269936e-10,
+      "loss": 0.5678,
+      "step": 42292
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5873412174873593,
+      "learning_rate": 6.814653060971399e-10,
+      "loss": 0.6522,
+      "step": 42293
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3592577895423192,
+      "learning_rate": 6.725864736645271e-10,
+      "loss": 0.7103,
+      "step": 42294
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5706343420395736,
+      "learning_rate": 6.637658604813357e-10,
+      "loss": 0.482,
+      "step": 42295
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5687106476186705,
+      "learning_rate": 6.550034666008564e-10,
+      "loss": 0.6899,
+      "step": 42296
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2721087978595293,
+      "learning_rate": 6.462992920719391e-10,
+      "loss": 0.5165,
+      "step": 42297
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.406732485270816,
+      "learning_rate": 6.376533369456539e-10,
+      "loss": 0.7718,
+      "step": 42298
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2680385829005187,
+      "learning_rate": 6.290656012730711e-10,
+      "loss": 0.6335,
+      "step": 42299
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2593605910010306,
+      "learning_rate": 6.205360851041509e-10,
+      "loss": 0.7402,
+      "step": 42300
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4256861743291067,
+      "learning_rate": 6.120647884877429e-10,
+      "loss": 0.7957,
+      "step": 42301
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4129906387038045,
+      "learning_rate": 6.036517114738072e-10,
+      "loss": 0.7057,
+      "step": 42302
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5829315340631597,
+      "learning_rate": 5.952968541111937e-10,
+      "loss": 0.4718,
+      "step": 42303
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4858085089659798,
+      "learning_rate": 5.87000216448752e-10,
+      "loss": 0.7704,
+      "step": 42304
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3501988565092384,
+      "learning_rate": 5.78761798534222e-10,
+      "loss": 0.6731,
+      "step": 42305
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4148740535860873,
+      "learning_rate": 5.70581600415343e-10,
+      "loss": 0.6764,
+      "step": 42306
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3766518399232865,
+      "learning_rate": 5.62459622140965e-10,
+      "loss": 0.801,
+      "step": 42307
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5490305667157933,
+      "learning_rate": 5.543958637577174e-10,
+      "loss": 0.5653,
+      "step": 42308
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5227448047380343,
+      "learning_rate": 5.463903253122294e-10,
+      "loss": 0.6822,
+      "step": 42309
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7319692802980533,
+      "learning_rate": 5.384430068522406e-10,
+      "loss": 0.5974,
+      "step": 42310
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5416822137674542,
+      "learning_rate": 5.305539084221601e-10,
+      "loss": 0.6299,
+      "step": 42311
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.374738021846897,
+      "learning_rate": 5.227230300697273e-10,
+      "loss": 0.6044,
+      "step": 42312
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.340160313491122,
+      "learning_rate": 5.149503718393511e-10,
+      "loss": 0.7063,
+      "step": 42313
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2895876063853342,
+      "learning_rate": 5.072359337776611e-10,
+      "loss": 0.7085,
+      "step": 42314
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.385633309110983,
+      "learning_rate": 4.995797159279558e-10,
+      "loss": 0.6901,
+      "step": 42315
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.678208011791385,
+      "learning_rate": 4.919817183346443e-10,
+      "loss": 0.6151,
+      "step": 42316
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.477173364888389,
+      "learning_rate": 4.844419410443557e-10,
+      "loss": 0.7115,
+      "step": 42317
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4446109208976292,
+      "learning_rate": 4.769603840981684e-10,
+      "loss": 0.7161,
+      "step": 42318
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4663133797391208,
+      "learning_rate": 4.695370475416016e-10,
+      "loss": 0.5759,
+      "step": 42319
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3313222476912105,
+      "learning_rate": 4.621719314168438e-10,
+      "loss": 0.6617,
+      "step": 42320
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5879006720468357,
+      "learning_rate": 4.5486503576719354e-10,
+      "loss": 0.4675,
+      "step": 42321
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.207299694912544,
+      "learning_rate": 4.476163606348394e-10,
+      "loss": 0.6367,
+      "step": 42322
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5788215520785989,
+      "learning_rate": 4.4042590606308e-10,
+      "loss": 0.4669,
+      "step": 42323
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3070197195067614,
+      "learning_rate": 4.3329367209188344e-10,
+      "loss": 0.5468,
+      "step": 42324
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3259130741578764,
+      "learning_rate": 4.262196587645484e-10,
+      "loss": 0.7187,
+      "step": 42325
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6386064631285594,
+      "learning_rate": 4.1920386612215313e-10,
+      "loss": 0.496,
+      "step": 42326
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4010198254529245,
+      "learning_rate": 4.1224629420466567e-10,
+      "loss": 0.6794,
+      "step": 42327
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.523420044307244,
+      "learning_rate": 4.05346943052054e-10,
+      "loss": 0.5867,
+      "step": 42328
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3786841587163168,
+      "learning_rate": 3.9850581270539645e-10,
+      "loss": 0.6379,
+      "step": 42329
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5908970314058828,
+      "learning_rate": 3.917229032057712e-10,
+      "loss": 0.4866,
+      "step": 42330
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2927855131258095,
+      "learning_rate": 3.849982145898157e-10,
+      "loss": 0.6529,
+      "step": 42331
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5351100957178117,
+      "learning_rate": 3.7833174689860806e-10,
+      "loss": 0.6004,
+      "step": 42332
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5773181833716199,
+      "learning_rate": 3.717235001710062e-10,
+      "loss": 0.7652,
+      "step": 42333
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.41952542268393,
+      "learning_rate": 3.651734744447577e-10,
+      "loss": 0.7515,
+      "step": 42334
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7310634293074414,
+      "learning_rate": 3.586816697587203e-10,
+      "loss": 0.6343,
+      "step": 42335
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.61510744460638,
+      "learning_rate": 3.522480861506417e-10,
+      "loss": 0.6666,
+      "step": 42336
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8929346128101978,
+      "learning_rate": 3.458727236571591e-10,
+      "loss": 0.604,
+      "step": 42337
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2539246863841769,
+      "learning_rate": 3.3955558231490994e-10,
+      "loss": 0.5597,
+      "step": 42338
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3612120852893996,
+      "learning_rate": 3.3329666216275205e-10,
+      "loss": 0.6853,
+      "step": 42339
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4046535287635358,
+      "learning_rate": 3.2709596323510227e-10,
+      "loss": 0.5624,
+      "step": 42340
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6118637828861393,
+      "learning_rate": 3.2095348556970827e-10,
+      "loss": 0.6792,
+      "step": 42341
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3040990570194297,
+      "learning_rate": 3.1486922920098696e-10,
+      "loss": 0.7691,
+      "step": 42342
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4479138513740557,
+      "learning_rate": 3.088431941655756e-10,
+      "loss": 0.6764,
+      "step": 42343
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5130938096139113,
+      "learning_rate": 3.028753804978912e-10,
+      "loss": 0.7117,
+      "step": 42344
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6052989327064329,
+      "learning_rate": 2.9696578823235067e-10,
+      "loss": 0.486,
+      "step": 42345
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3058694061428162,
+      "learning_rate": 2.9111441740337085e-10,
+      "loss": 0.7069,
+      "step": 42346
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.1702717761961496,
+      "learning_rate": 2.8532126804647895e-10,
+      "loss": 0.7,
+      "step": 42347
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6021310279969528,
+      "learning_rate": 2.795863401927612e-10,
+      "loss": 0.4869,
+      "step": 42348
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6654394005238695,
+      "learning_rate": 2.739096338788549e-10,
+      "loss": 0.6081,
+      "step": 42349
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3526250248254204,
+      "learning_rate": 2.682911491347362e-10,
+      "loss": 0.7932,
+      "step": 42350
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3883739677224178,
+      "learning_rate": 2.627308859948219e-10,
+      "loss": 0.6675,
+      "step": 42351
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.322120406772599,
+      "learning_rate": 2.5722884449130845e-10,
+      "loss": 0.7633,
+      "step": 42352
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2387118574493345,
+      "learning_rate": 2.517850246563924e-10,
+      "loss": 0.5397,
+      "step": 42353
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4165131710436405,
+      "learning_rate": 2.4639942652116e-10,
+      "loss": 0.7575,
+      "step": 42354
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4084159103708962,
+      "learning_rate": 2.4107205011780764e-10,
+      "loss": 0.7761,
+      "step": 42355
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9367805225230548,
+      "learning_rate": 2.358028954763114e-10,
+      "loss": 0.8118,
+      "step": 42356
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7406913778255606,
+      "learning_rate": 2.305919626277575e-10,
+      "loss": 0.6646,
+      "step": 42357
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5749779639815068,
+      "learning_rate": 2.25439251602122e-10,
+      "loss": 0.4838,
+      "step": 42358
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4367083210391027,
+      "learning_rate": 2.203447624304911e-10,
+      "loss": 0.6638,
+      "step": 42359
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.342144405457038,
+      "learning_rate": 2.1530849514173057e-10,
+      "loss": 0.7494,
+      "step": 42360
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6384668136929603,
+      "learning_rate": 2.103304497658165e-10,
+      "loss": 0.6511,
+      "step": 42361
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4335076119908998,
+      "learning_rate": 2.0541062633050446e-10,
+      "loss": 0.6185,
+      "step": 42362
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4058542046945557,
+      "learning_rate": 2.0054902486466022e-10,
+      "loss": 0.6697,
+      "step": 42363
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5354022353919126,
+      "learning_rate": 1.957456453982598e-10,
+      "loss": 0.6283,
+      "step": 42364
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5197946948813215,
+      "learning_rate": 1.9100048795794858e-10,
+      "loss": 0.533,
+      "step": 42365
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.57569469525607,
+      "learning_rate": 1.8631355257148209e-10,
+      "loss": 0.4118,
+      "step": 42366
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.710847939278678,
+      "learning_rate": 1.8168483926550574e-10,
+      "loss": 0.6145,
+      "step": 42367
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.032779143337543,
+      "learning_rate": 1.7711434806888527e-10,
+      "loss": 0.6598,
+      "step": 42368
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8298342881320147,
+      "learning_rate": 1.7260207900604565e-10,
+      "loss": 0.681,
+      "step": 42369
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5920299359686887,
+      "learning_rate": 1.6814803210363217e-10,
+      "loss": 0.6445,
+      "step": 42370
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5504895136927643,
+      "learning_rate": 1.6375220738940046e-10,
+      "loss": 0.6934,
+      "step": 42371
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4548222680591518,
+      "learning_rate": 1.5941460488777538e-10,
+      "loss": 0.7354,
+      "step": 42372
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3167882882930169,
+      "learning_rate": 1.5513522462318188e-10,
+      "loss": 0.6973,
+      "step": 42373
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4733291171572291,
+      "learning_rate": 1.5091406662115505e-10,
+      "loss": 0.6922,
+      "step": 42374
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4004607255400776,
+      "learning_rate": 1.4675113090723004e-10,
+      "loss": 0.6885,
+      "step": 42375
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4132544720639324,
+      "learning_rate": 1.426464175036113e-10,
+      "loss": 0.6675,
+      "step": 42376
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3303896708418435,
+      "learning_rate": 1.385999264369442e-10,
+      "loss": 0.6254,
+      "step": 42377
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8021746552711269,
+      "learning_rate": 1.3461165772832297e-10,
+      "loss": 0.7411,
+      "step": 42378
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7212469066217604,
+      "learning_rate": 1.3068161140217251e-10,
+      "loss": 0.6019,
+      "step": 42379
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.330617930711768,
+      "learning_rate": 1.2680978748069728e-10,
+      "loss": 0.6359,
+      "step": 42380
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.250602114400364,
+      "learning_rate": 1.2299618598721197e-10,
+      "loss": 0.7625,
+      "step": 42381
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5744861766261011,
+      "learning_rate": 1.1924080694281083e-10,
+      "loss": 0.707,
+      "step": 42382
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3090174656386513,
+      "learning_rate": 1.1554365037080851e-10,
+      "loss": 0.6829,
+      "step": 42383
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3861590507622035,
+      "learning_rate": 1.1190471629229927e-10,
+      "loss": 0.4997,
+      "step": 42384
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.697624340222206,
+      "learning_rate": 1.0832400472726711e-10,
+      "loss": 0.6441,
+      "step": 42385
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.278725616761078,
+      "learning_rate": 1.0480151569791652e-10,
+      "loss": 0.6286,
+      "step": 42386
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7592432826082238,
+      "learning_rate": 1.0133724922534171e-10,
+      "loss": 0.6211,
+      "step": 42387
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4436714015721215,
+      "learning_rate": 9.793120532730627e-11,
+      "loss": 0.6943,
+      "step": 42388
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5864894412378271,
+      "learning_rate": 9.458338402601464e-11,
+      "loss": 0.519,
+      "step": 42389
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.565627810432323,
+      "learning_rate": 9.129378533923038e-11,
+      "loss": 0.5441,
+      "step": 42390
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3787795854519802,
+      "learning_rate": 8.806240928693755e-11,
+      "loss": 0.6128,
+      "step": 42391
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3132206363693435,
+      "learning_rate": 8.488925588800989e-11,
+      "loss": 0.6546,
+      "step": 42392
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.589873610566551,
+      "learning_rate": 8.177432516132122e-11,
+      "loss": 0.6677,
+      "step": 42393
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3986705841849967,
+      "learning_rate": 7.871761712352489e-11,
+      "loss": 0.6963,
+      "step": 42394
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.294922735165311,
+      "learning_rate": 7.571913179349466e-11,
+      "loss": 0.7494,
+      "step": 42395
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.622019680220968,
+      "learning_rate": 7.277886918899413e-11,
+      "loss": 0.641,
+      "step": 42396
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4846308809390805,
+      "learning_rate": 6.989682932667663e-11,
+      "loss": 0.6444,
+      "step": 42397
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4813465156070773,
+      "learning_rate": 6.707301222319551e-11,
+      "loss": 0.5952,
+      "step": 42398
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.435164079609251,
+      "learning_rate": 6.430741789520412e-11,
+      "loss": 0.7,
+      "step": 42399
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6419882437194555,
+      "learning_rate": 6.160004635824556e-11,
+      "loss": 0.515,
+      "step": 42400
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5479840288410265,
+      "learning_rate": 5.895089762897322e-11,
+      "loss": 0.6962,
+      "step": 42401
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1809992341588211,
+      "learning_rate": 5.6359971722930176e-11,
+      "loss": 0.5998,
+      "step": 42402
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.16135777473882,
+      "learning_rate": 5.3827268654549346e-11,
+      "loss": 0.6045,
+      "step": 42403
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3062562329652865,
+      "learning_rate": 5.135278843826363e-11,
+      "loss": 0.6513,
+      "step": 42404
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6026942050217547,
+      "learning_rate": 4.8936531088505936e-11,
+      "loss": 0.5125,
+      "step": 42405
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3619934685342459,
+      "learning_rate": 4.657849662081937e-11,
+      "loss": 0.6932,
+      "step": 42406
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2842309487264016,
+      "learning_rate": 4.4278685047416395e-11,
+      "loss": 0.5914,
+      "step": 42407
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3979982639244926,
+      "learning_rate": 4.203709638161968e-11,
+      "loss": 0.6829,
+      "step": 42408
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5444983314702687,
+      "learning_rate": 3.985373063786213e-11,
+      "loss": 0.6234,
+      "step": 42409
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.429268274764494,
+      "learning_rate": 3.772858782724598e-11,
+      "loss": 0.6567,
+      "step": 42410
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4302586235377606,
+      "learning_rate": 3.566166796420412e-11,
+      "loss": 0.6335,
+      "step": 42411
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3933516626270164,
+      "learning_rate": 3.3652971057618336e-11,
+      "loss": 0.6009,
+      "step": 42412
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5775420905463449,
+      "learning_rate": 3.170249712303175e-11,
+      "loss": 0.4922,
+      "step": 42413
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2178718195719511,
+      "learning_rate": 2.9810246168215926e-11,
+      "loss": 0.5607,
+      "step": 42414
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4420536208545573,
+      "learning_rate": 2.7976218206493543e-11,
+      "loss": 0.6784,
+      "step": 42415
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8624466662043888,
+      "learning_rate": 2.6200413247856605e-11,
+      "loss": 0.83,
+      "step": 42416
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2569030685088696,
+      "learning_rate": 2.4482831302297118e-11,
+      "loss": 0.7813,
+      "step": 42417
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.313545123806084,
+      "learning_rate": 2.2823472379807086e-11,
+      "loss": 0.7786,
+      "step": 42418
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3828769743376093,
+      "learning_rate": 2.1222336491488748e-11,
+      "loss": 0.8028,
+      "step": 42419
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6629928569311712,
+      "learning_rate": 1.967942364511366e-11,
+      "loss": 0.6809,
+      "step": 42420
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7404110530548635,
+      "learning_rate": 1.8194733849563605e-11,
+      "loss": 0.61,
+      "step": 42421
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.883681185981101,
+      "learning_rate": 1.6768267114830594e-11,
+      "loss": 0.6955,
+      "step": 42422
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5631371253951796,
+      "learning_rate": 1.5400023447575962e-11,
+      "loss": 0.7282,
+      "step": 42423
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.627884438557777,
+      "learning_rate": 1.4090002856681495e-11,
+      "loss": 0.6821,
+      "step": 42424
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.53811700059162,
+      "learning_rate": 1.2838205351028976e-11,
+      "loss": 0.6662,
+      "step": 42425
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3164796846215177,
+      "learning_rate": 1.1644630935059298e-11,
+      "loss": 0.7132,
+      "step": 42426
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1953561793811358,
+      "learning_rate": 1.0509279617654245e-11,
+      "loss": 0.6292,
+      "step": 42427
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3175306870468515,
+      "learning_rate": 9.432151405475153e-12,
+      "loss": 0.683,
+      "step": 42428
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7762685695822598,
+      "learning_rate": 8.413246304073142e-12,
+      "loss": 0.597,
+      "step": 42429
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5016564106075851,
+      "learning_rate": 7.452564318999322e-12,
+      "loss": 0.7078,
+      "step": 42430
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1936609442879196,
+      "learning_rate": 6.550105456915035e-12,
+      "loss": 0.6199,
+      "step": 42431
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.6257908022048637,
+      "learning_rate": 5.705869723371394e-12,
+      "loss": 0.6576,
+      "step": 42432
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.600463301654748,
+      "learning_rate": 4.919857121699068e-12,
+      "loss": 0.7157,
+      "step": 42433
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2711650759961495,
+      "learning_rate": 4.192067657449173e-12,
+      "loss": 0.6495,
+      "step": 42434
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7214577987993989,
+      "learning_rate": 3.5225013350626004e-12,
+      "loss": 0.6364,
+      "step": 42435
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.754419586090059,
+      "learning_rate": 2.9111581578700197e-12,
+      "loss": 0.749,
+      "step": 42436
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.195744702399388,
+      "learning_rate": 2.3580381292021006e-12,
+      "loss": 0.6395,
+      "step": 42437
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.619247880545094,
+      "learning_rate": 1.863141253499734e-12,
+      "loss": 0.7042,
+      "step": 42438
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.144205778235928,
+      "learning_rate": 1.4264675329833665e-12,
+      "loss": 0.6874,
+      "step": 42439
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.582533468377906,
+      "learning_rate": 1.0480169687632214e-12,
+      "loss": 0.4914,
+      "step": 42440
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1891139561822501,
+      "learning_rate": 7.277895663904133e-13,
+      "loss": 0.7061,
+      "step": 42441
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.386346704407859,
+      "learning_rate": 4.657853236444964e-13,
+      "loss": 0.5547,
+      "step": 42442
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2176662397790952,
+      "learning_rate": 2.620042460765859e-13,
+      "loss": 0.7391,
+      "step": 42443
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5101198019666489,
+      "learning_rate": 1.1644633146623564e-13,
+      "loss": 0.6882,
+      "step": 42444
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2126592635339049,
+      "learning_rate": 2.9111583144114664e-14,
+      "loss": 0.5484,
+      "step": 42445
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0714081235241186,
+      "learning_rate": 0.0,
+      "loss": 0.504,
+      "step": 42446
+    },
+    {
+      "epoch": 1.0,
+      "step": 42446,
+      "total_flos": 8.203217535467572e+19,
+      "train_loss": 0.7028406760385685,
+      "train_runtime": 175184.6397,
+      "train_samples_per_second": 31.014,
+      "train_steps_per_second": 0.242
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 42446,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 10000,
+  "total_flos": 8.203217535467572e+19,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}