diff --git "a/checkpoint-5000/trainer_state.json" "b/checkpoint-5000/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-5000/trainer_state.json"
@@ -0,0 +1,35034 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.279808,
+  "eval_steps": 500,
+  "global_step": 5000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.000256,
+      "grad_norm": 13.9375,
+      "learning_rate": 5e-05,
+      "loss": 5.0668,
+      "step": 1
+    },
+    {
+      "epoch": 0.000512,
+      "grad_norm": 4.34375,
+      "learning_rate": 4.999573415237608e-05,
+      "loss": 3.8496,
+      "step": 2
+    },
+    {
+      "epoch": 0.000768,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.999146830475216e-05,
+      "loss": 3.6141,
+      "step": 3
+    },
+    {
+      "epoch": 0.001024,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.998720245712823e-05,
+      "loss": 3.3289,
+      "step": 4
+    },
+    {
+      "epoch": 0.00128,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.998293660950431e-05,
+      "loss": 3.3289,
+      "step": 5
+    },
+    {
+      "epoch": 0.001536,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.997867076188039e-05,
+      "loss": 3.2139,
+      "step": 6
+    },
+    {
+      "epoch": 0.001792,
+      "grad_norm": 1.984375,
+      "learning_rate": 4.997440491425647e-05,
+      "loss": 3.1277,
+      "step": 7
+    },
+    {
+      "epoch": 0.002048,
+      "grad_norm": 1.4765625,
+      "learning_rate": 4.997013906663254e-05,
+      "loss": 2.9207,
+      "step": 8
+    },
+    {
+      "epoch": 0.002304,
+      "grad_norm": 1.46875,
+      "learning_rate": 4.996587321900862e-05,
+      "loss": 2.853,
+      "step": 9
+    },
+    {
+      "epoch": 0.00256,
+      "grad_norm": 1.3359375,
+      "learning_rate": 4.996160737138469e-05,
+      "loss": 2.8234,
+      "step": 10
+    },
+    {
+      "epoch": 0.002816,
+      "grad_norm": 1.2265625,
+      "learning_rate": 4.995734152376078e-05,
+      "loss": 2.8791,
+      "step": 11
+    },
+    {
+      "epoch": 0.003072,
+      "grad_norm": 1.0859375,
+      "learning_rate": 4.995307567613685e-05,
+      "loss": 2.8151,
+      "step": 12
+    },
+    {
+      "epoch": 0.003328,
+      "grad_norm": 1.078125,
+      "learning_rate": 4.994880982851293e-05,
+      "loss": 2.8136,
+      "step": 13
+    },
+    {
+      "epoch": 0.003584,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.9944543980889e-05,
+      "loss": 2.733,
+      "step": 14
+    },
+    {
+      "epoch": 0.00384,
+      "grad_norm": 0.984375,
+      "learning_rate": 4.994027813326508e-05,
+      "loss": 2.7711,
+      "step": 15
+    },
+    {
+      "epoch": 0.004096,
+      "grad_norm": 0.9296875,
+      "learning_rate": 4.993601228564116e-05,
+      "loss": 2.6806,
+      "step": 16
+    },
+    {
+      "epoch": 0.004352,
+      "grad_norm": 0.90625,
+      "learning_rate": 4.993174643801724e-05,
+      "loss": 2.6814,
+      "step": 17
+    },
+    {
+      "epoch": 0.004608,
+      "grad_norm": 0.94140625,
+      "learning_rate": 4.992748059039332e-05,
+      "loss": 2.7762,
+      "step": 18
+    },
+    {
+      "epoch": 0.004864,
+      "grad_norm": 0.90625,
+      "learning_rate": 4.9923214742769393e-05,
+      "loss": 2.7266,
+      "step": 19
+    },
+    {
+      "epoch": 0.00512,
+      "grad_norm": 0.9296875,
+      "learning_rate": 4.991894889514547e-05,
+      "loss": 2.6561,
+      "step": 20
+    },
+    {
+      "epoch": 0.005376,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.991468304752154e-05,
+      "loss": 2.655,
+      "step": 21
+    },
+    {
+      "epoch": 0.005632,
+      "grad_norm": 0.921875,
+      "learning_rate": 4.991041719989762e-05,
+      "loss": 2.6547,
+      "step": 22
+    },
+    {
+      "epoch": 0.005888,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.99061513522737e-05,
+      "loss": 2.5858,
+      "step": 23
+    },
+    {
+      "epoch": 0.006144,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.990188550464978e-05,
+      "loss": 2.6207,
+      "step": 24
+    },
+    {
+      "epoch": 0.0064,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.989761965702585e-05,
+      "loss": 2.6117,
+      "step": 25
+    },
+    {
+      "epoch": 0.006656,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.9893353809401934e-05,
+      "loss": 2.5768,
+      "step": 26
+    },
+    {
+      "epoch": 0.006912,
+      "grad_norm": 0.93359375,
+      "learning_rate": 4.988908796177801e-05,
+      "loss": 2.5712,
+      "step": 27
+    },
+    {
+      "epoch": 0.007168,
+      "grad_norm": 0.96484375,
+      "learning_rate": 4.988482211415408e-05,
+      "loss": 2.6222,
+      "step": 28
+    },
+    {
+      "epoch": 0.007424,
+      "grad_norm": 0.9140625,
+      "learning_rate": 4.9880556266530156e-05,
+      "loss": 2.6012,
+      "step": 29
+    },
+    {
+      "epoch": 0.00768,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.987629041890624e-05,
+      "loss": 2.5622,
+      "step": 30
+    },
+    {
+      "epoch": 0.007936,
+      "grad_norm": 0.8671875,
+      "learning_rate": 4.987202457128232e-05,
+      "loss": 2.5489,
+      "step": 31
+    },
+    {
+      "epoch": 0.008192,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.986775872365839e-05,
+      "loss": 2.5326,
+      "step": 32
+    },
+    {
+      "epoch": 0.008448,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.986349287603447e-05,
+      "loss": 2.4663,
+      "step": 33
+    },
+    {
+      "epoch": 0.008704,
+      "grad_norm": 0.7890625,
+      "learning_rate": 4.985922702841055e-05,
+      "loss": 2.4393,
+      "step": 34
+    },
+    {
+      "epoch": 0.00896,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.985496118078662e-05,
+      "loss": 2.5233,
+      "step": 35
+    },
+    {
+      "epoch": 0.009216,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.98506953331627e-05,
+      "loss": 2.5162,
+      "step": 36
+    },
+    {
+      "epoch": 0.009472,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.9846429485538784e-05,
+      "loss": 2.4974,
+      "step": 37
+    },
+    {
+      "epoch": 0.009728,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.984216363791486e-05,
+      "loss": 2.452,
+      "step": 38
+    },
+    {
+      "epoch": 0.009984,
+      "grad_norm": 0.9296875,
+      "learning_rate": 4.983789779029093e-05,
+      "loss": 2.5105,
+      "step": 39
+    },
+    {
+      "epoch": 0.01024,
+      "grad_norm": 0.87109375,
+      "learning_rate": 4.983363194266701e-05,
+      "loss": 2.4345,
+      "step": 40
+    },
+    {
+      "epoch": 0.010496,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.982936609504309e-05,
+      "loss": 2.3587,
+      "step": 41
+    },
+    {
+      "epoch": 0.010752,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.982510024741916e-05,
+      "loss": 2.5015,
+      "step": 42
+    },
+    {
+      "epoch": 0.011008,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.9820834399795244e-05,
+      "loss": 2.4501,
+      "step": 43
+    },
+    {
+      "epoch": 0.011264,
+      "grad_norm": 0.9296875,
+      "learning_rate": 4.981656855217132e-05,
+      "loss": 2.4863,
+      "step": 44
+    },
+    {
+      "epoch": 0.01152,
+      "grad_norm": 0.91015625,
+      "learning_rate": 4.98123027045474e-05,
+      "loss": 2.4312,
+      "step": 45
+    },
+    {
+      "epoch": 0.011776,
+      "grad_norm": 0.9296875,
+      "learning_rate": 4.980803685692347e-05,
+      "loss": 2.4501,
+      "step": 46
+    },
+    {
+      "epoch": 0.012032,
+      "grad_norm": 0.94921875,
+      "learning_rate": 4.980377100929955e-05,
+      "loss": 2.4296,
+      "step": 47
+    },
+    {
+      "epoch": 0.012288,
+      "grad_norm": 0.9140625,
+      "learning_rate": 4.979950516167562e-05,
+      "loss": 2.3717,
+      "step": 48
+    },
+    {
+      "epoch": 0.012544,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.97952393140517e-05,
+      "loss": 2.4149,
+      "step": 49
+    },
+    {
+      "epoch": 0.0128,
+      "grad_norm": 0.94921875,
+      "learning_rate": 4.9790973466427784e-05,
+      "loss": 2.4097,
+      "step": 50
+    },
+    {
+      "epoch": 0.013056,
+      "grad_norm": 0.95703125,
+      "learning_rate": 4.978670761880386e-05,
+      "loss": 2.4458,
+      "step": 51
+    },
+    {
+      "epoch": 0.013312,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.978244177117994e-05,
+      "loss": 2.3199,
+      "step": 52
+    },
+    {
+      "epoch": 0.013568,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.977817592355601e-05,
+      "loss": 2.3236,
+      "step": 53
+    },
+    {
+      "epoch": 0.013824,
+      "grad_norm": 0.9375,
+      "learning_rate": 4.977391007593209e-05,
+      "loss": 2.2662,
+      "step": 54
+    },
+    {
+      "epoch": 0.01408,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.976964422830817e-05,
+      "loss": 2.3522,
+      "step": 55
+    },
+    {
+      "epoch": 0.014336,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.976537838068425e-05,
+      "loss": 2.2466,
+      "step": 56
+    },
+    {
+      "epoch": 0.014592,
+      "grad_norm": 1.1171875,
+      "learning_rate": 4.9761112533060324e-05,
+      "loss": 2.3267,
+      "step": 57
+    },
+    {
+      "epoch": 0.014848,
+      "grad_norm": 1.25,
+      "learning_rate": 4.97568466854364e-05,
+      "loss": 2.2987,
+      "step": 58
+    },
+    {
+      "epoch": 0.015104,
+      "grad_norm": 1.140625,
+      "learning_rate": 4.975258083781247e-05,
+      "loss": 2.2966,
+      "step": 59
+    },
+    {
+      "epoch": 0.01536,
+      "grad_norm": 1.3828125,
+      "learning_rate": 4.9748314990188554e-05,
+      "loss": 2.3059,
+      "step": 60
+    },
+    {
+      "epoch": 0.015616,
+      "grad_norm": 1.328125,
+      "learning_rate": 4.974404914256463e-05,
+      "loss": 2.2253,
+      "step": 61
+    },
+    {
+      "epoch": 0.015872,
+      "grad_norm": 1.2109375,
+      "learning_rate": 4.973978329494071e-05,
+      "loss": 2.2493,
+      "step": 62
+    },
+    {
+      "epoch": 0.016128,
+      "grad_norm": 1.21875,
+      "learning_rate": 4.973551744731678e-05,
+      "loss": 2.2106,
+      "step": 63
+    },
+    {
+      "epoch": 0.016384,
+      "grad_norm": 1.2734375,
+      "learning_rate": 4.9731251599692864e-05,
+      "loss": 2.243,
+      "step": 64
+    },
+    {
+      "epoch": 0.01664,
+      "grad_norm": 1.1328125,
+      "learning_rate": 4.972698575206894e-05,
+      "loss": 2.1977,
+      "step": 65
+    },
+    {
+      "epoch": 0.016896,
+      "grad_norm": 1.2421875,
+      "learning_rate": 4.972271990444501e-05,
+      "loss": 2.2525,
+      "step": 66
+    },
+    {
+      "epoch": 0.017152,
+      "grad_norm": 1.28125,
+      "learning_rate": 4.9718454056821094e-05,
+      "loss": 2.1998,
+      "step": 67
+    },
+    {
+      "epoch": 0.017408,
+      "grad_norm": 1.375,
+      "learning_rate": 4.971418820919717e-05,
+      "loss": 2.2602,
+      "step": 68
+    },
+    {
+      "epoch": 0.017664,
+      "grad_norm": 1.296875,
+      "learning_rate": 4.970992236157325e-05,
+      "loss": 2.1674,
+      "step": 69
+    },
+    {
+      "epoch": 0.01792,
+      "grad_norm": 1.4765625,
+      "learning_rate": 4.970565651394932e-05,
+      "loss": 2.1507,
+      "step": 70
+    },
+    {
+      "epoch": 0.018176,
+      "grad_norm": 1.09375,
+      "learning_rate": 4.9701390666325404e-05,
+      "loss": 2.1094,
+      "step": 71
+    },
+    {
+      "epoch": 0.018432,
+      "grad_norm": 1.53125,
+      "learning_rate": 4.969712481870148e-05,
+      "loss": 2.1872,
+      "step": 72
+    },
+    {
+      "epoch": 0.018688,
+      "grad_norm": 1.3125,
+      "learning_rate": 4.969285897107755e-05,
+      "loss": 2.1455,
+      "step": 73
+    },
+    {
+      "epoch": 0.018944,
+      "grad_norm": 1.375,
+      "learning_rate": 4.9688593123453634e-05,
+      "loss": 2.1714,
+      "step": 74
+    },
+    {
+      "epoch": 0.0192,
+      "grad_norm": 1.4296875,
+      "learning_rate": 4.9684327275829715e-05,
+      "loss": 2.1809,
+      "step": 75
+    },
+    {
+      "epoch": 0.019456,
+      "grad_norm": 1.3125,
+      "learning_rate": 4.968006142820579e-05,
+      "loss": 2.0828,
+      "step": 76
+    },
+    {
+      "epoch": 0.019712,
+      "grad_norm": 1.4375,
+      "learning_rate": 4.9675795580581863e-05,
+      "loss": 2.1553,
+      "step": 77
+    },
+    {
+      "epoch": 0.019968,
+      "grad_norm": 1.3984375,
+      "learning_rate": 4.967152973295794e-05,
+      "loss": 2.0594,
+      "step": 78
+    },
+    {
+      "epoch": 0.020224,
+      "grad_norm": 1.4375,
+      "learning_rate": 4.966726388533402e-05,
+      "loss": 2.105,
+      "step": 79
+    },
+    {
+      "epoch": 0.02048,
+      "grad_norm": 1.3671875,
+      "learning_rate": 4.966299803771009e-05,
+      "loss": 2.0872,
+      "step": 80
+    },
+    {
+      "epoch": 0.020736,
+      "grad_norm": 1.6953125,
+      "learning_rate": 4.9658732190086174e-05,
+      "loss": 2.0694,
+      "step": 81
+    },
+    {
+      "epoch": 0.020992,
+      "grad_norm": 1.546875,
+      "learning_rate": 4.965446634246225e-05,
+      "loss": 2.0909,
+      "step": 82
+    },
+    {
+      "epoch": 0.021248,
+      "grad_norm": 1.4765625,
+      "learning_rate": 4.965020049483833e-05,
+      "loss": 2.1063,
+      "step": 83
+    },
+    {
+      "epoch": 0.021504,
+      "grad_norm": 1.5546875,
+      "learning_rate": 4.9645934647214404e-05,
+      "loss": 2.0504,
+      "step": 84
+    },
+    {
+      "epoch": 0.02176,
+      "grad_norm": 1.484375,
+      "learning_rate": 4.964166879959048e-05,
+      "loss": 2.122,
+      "step": 85
+    },
+    {
+      "epoch": 0.022016,
+      "grad_norm": 1.5546875,
+      "learning_rate": 4.963740295196656e-05,
+      "loss": 2.0383,
+      "step": 86
+    },
+    {
+      "epoch": 0.022272,
+      "grad_norm": 1.65625,
+      "learning_rate": 4.963313710434263e-05,
+      "loss": 2.0272,
+      "step": 87
+    },
+    {
+      "epoch": 0.022528,
+      "grad_norm": 1.6015625,
+      "learning_rate": 4.9628871256718714e-05,
+      "loss": 2.0743,
+      "step": 88
+    },
+    {
+      "epoch": 0.022784,
+      "grad_norm": 1.578125,
+      "learning_rate": 4.962460540909479e-05,
+      "loss": 2.0283,
+      "step": 89
+    },
+    {
+      "epoch": 0.02304,
+      "grad_norm": 1.4375,
+      "learning_rate": 4.962033956147087e-05,
+      "loss": 2.0151,
+      "step": 90
+    },
+    {
+      "epoch": 0.023296,
+      "grad_norm": 1.734375,
+      "learning_rate": 4.9616073713846944e-05,
+      "loss": 2.0023,
+      "step": 91
+    },
+    {
+      "epoch": 0.023552,
+      "grad_norm": 1.5390625,
+      "learning_rate": 4.961180786622302e-05,
+      "loss": 1.9912,
+      "step": 92
+    },
+    {
+      "epoch": 0.023808,
+      "grad_norm": 1.625,
+      "learning_rate": 4.960754201859909e-05,
+      "loss": 1.9845,
+      "step": 93
+    },
+    {
+      "epoch": 0.024064,
+      "grad_norm": 1.59375,
+      "learning_rate": 4.960327617097518e-05,
+      "loss": 2.0745,
+      "step": 94
+    },
+    {
+      "epoch": 0.02432,
+      "grad_norm": 1.5546875,
+      "learning_rate": 4.9599010323351254e-05,
+      "loss": 1.9677,
+      "step": 95
+    },
+    {
+      "epoch": 0.024576,
+      "grad_norm": 1.6171875,
+      "learning_rate": 4.959474447572733e-05,
+      "loss": 2.0309,
+      "step": 96
+    },
+    {
+      "epoch": 0.024832,
+      "grad_norm": 1.546875,
+      "learning_rate": 4.95904786281034e-05,
+      "loss": 1.9489,
+      "step": 97
+    },
+    {
+      "epoch": 0.025088,
+      "grad_norm": 1.6328125,
+      "learning_rate": 4.9586212780479484e-05,
+      "loss": 2.0103,
+      "step": 98
+    },
+    {
+      "epoch": 0.025344,
+      "grad_norm": 1.5078125,
+      "learning_rate": 4.958194693285556e-05,
+      "loss": 1.9483,
+      "step": 99
+    },
+    {
+      "epoch": 0.0256,
+      "grad_norm": 1.625,
+      "learning_rate": 4.957768108523164e-05,
+      "loss": 1.9969,
+      "step": 100
+    },
+    {
+      "epoch": 0.025856,
+      "grad_norm": 1.59375,
+      "learning_rate": 4.957341523760772e-05,
+      "loss": 1.955,
+      "step": 101
+    },
+    {
+      "epoch": 0.026112,
+      "grad_norm": 1.4140625,
+      "learning_rate": 4.9569149389983795e-05,
+      "loss": 1.8897,
+      "step": 102
+    },
+    {
+      "epoch": 0.026368,
+      "grad_norm": 1.703125,
+      "learning_rate": 4.956488354235987e-05,
+      "loss": 1.9777,
+      "step": 103
+    },
+    {
+      "epoch": 0.026624,
+      "grad_norm": 1.6875,
+      "learning_rate": 4.956061769473594e-05,
+      "loss": 1.9827,
+      "step": 104
+    },
+    {
+      "epoch": 0.02688,
+      "grad_norm": 1.4921875,
+      "learning_rate": 4.9556351847112024e-05,
+      "loss": 1.9131,
+      "step": 105
+    },
+    {
+      "epoch": 0.027136,
+      "grad_norm": 1.7578125,
+      "learning_rate": 4.95520859994881e-05,
+      "loss": 1.9134,
+      "step": 106
+    },
+    {
+      "epoch": 0.027392,
+      "grad_norm": 1.671875,
+      "learning_rate": 4.954782015186418e-05,
+      "loss": 1.9172,
+      "step": 107
+    },
+    {
+      "epoch": 0.027648,
+      "grad_norm": 1.640625,
+      "learning_rate": 4.9543554304240254e-05,
+      "loss": 1.9225,
+      "step": 108
+    },
+    {
+      "epoch": 0.027904,
+      "grad_norm": 1.5703125,
+      "learning_rate": 4.9539288456616335e-05,
+      "loss": 1.9454,
+      "step": 109
+    },
+    {
+      "epoch": 0.02816,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.953502260899241e-05,
+      "loss": 1.9331,
+      "step": 110
+    },
+    {
+      "epoch": 0.028416,
+      "grad_norm": 1.6796875,
+      "learning_rate": 4.9530756761368483e-05,
+      "loss": 1.933,
+      "step": 111
+    },
+    {
+      "epoch": 0.028672,
+      "grad_norm": 1.8515625,
+      "learning_rate": 4.952649091374456e-05,
+      "loss": 1.9129,
+      "step": 112
+    },
+    {
+      "epoch": 0.028928,
+      "grad_norm": 1.8203125,
+      "learning_rate": 4.9522225066120645e-05,
+      "loss": 1.8792,
+      "step": 113
+    },
+    {
+      "epoch": 0.029184,
+      "grad_norm": 2.015625,
+      "learning_rate": 4.951795921849672e-05,
+      "loss": 1.9328,
+      "step": 114
+    },
+    {
+      "epoch": 0.02944,
+      "grad_norm": 1.6796875,
+      "learning_rate": 4.9513693370872794e-05,
+      "loss": 1.8749,
+      "step": 115
+    },
+    {
+      "epoch": 0.029696,
+      "grad_norm": 1.703125,
+      "learning_rate": 4.9509427523248875e-05,
+      "loss": 1.9127,
+      "step": 116
+    },
+    {
+      "epoch": 0.029952,
+      "grad_norm": 1.96875,
+      "learning_rate": 4.950516167562495e-05,
+      "loss": 1.8625,
+      "step": 117
+    },
+    {
+      "epoch": 0.030208,
+      "grad_norm": 1.6015625,
+      "learning_rate": 4.9500895828001024e-05,
+      "loss": 1.8334,
+      "step": 118
+    },
+    {
+      "epoch": 0.030464,
+      "grad_norm": 1.8984375,
+      "learning_rate": 4.9496629980377105e-05,
+      "loss": 1.8872,
+      "step": 119
+    },
+    {
+      "epoch": 0.03072,
+      "grad_norm": 1.90625,
+      "learning_rate": 4.9492364132753186e-05,
+      "loss": 1.8097,
+      "step": 120
+    },
+    {
+      "epoch": 0.030976,
+      "grad_norm": 1.734375,
+      "learning_rate": 4.948809828512926e-05,
+      "loss": 1.9105,
+      "step": 121
+    },
+    {
+      "epoch": 0.031232,
+      "grad_norm": 1.765625,
+      "learning_rate": 4.9483832437505334e-05,
+      "loss": 1.7726,
+      "step": 122
+    },
+    {
+      "epoch": 0.031488,
+      "grad_norm": 1.6875,
+      "learning_rate": 4.947956658988141e-05,
+      "loss": 1.8824,
+      "step": 123
+    },
+    {
+      "epoch": 0.031744,
+      "grad_norm": 1.84375,
+      "learning_rate": 4.947530074225749e-05,
+      "loss": 1.8696,
+      "step": 124
+    },
+    {
+      "epoch": 0.032,
+      "grad_norm": 1.6328125,
+      "learning_rate": 4.9471034894633564e-05,
+      "loss": 1.8297,
+      "step": 125
+    },
+    {
+      "epoch": 0.032256,
+      "grad_norm": 1.7109375,
+      "learning_rate": 4.9466769047009645e-05,
+      "loss": 1.7575,
+      "step": 126
+    },
+    {
+      "epoch": 0.032512,
+      "grad_norm": 1.7890625,
+      "learning_rate": 4.946250319938572e-05,
+      "loss": 1.7896,
+      "step": 127
+    },
+    {
+      "epoch": 0.032768,
+      "grad_norm": 1.828125,
+      "learning_rate": 4.94582373517618e-05,
+      "loss": 1.8516,
+      "step": 128
+    },
+    {
+      "epoch": 0.033024,
+      "grad_norm": 1.890625,
+      "learning_rate": 4.9453971504137874e-05,
+      "loss": 1.7899,
+      "step": 129
+    },
+    {
+      "epoch": 0.03328,
+      "grad_norm": 1.9296875,
+      "learning_rate": 4.944970565651395e-05,
+      "loss": 1.8454,
+      "step": 130
+    },
+    {
+      "epoch": 0.033536,
+      "grad_norm": 1.84375,
+      "learning_rate": 4.944543980889002e-05,
+      "loss": 1.8143,
+      "step": 131
+    },
+    {
+      "epoch": 0.033792,
+      "grad_norm": 1.9140625,
+      "learning_rate": 4.9441173961266104e-05,
+      "loss": 1.7879,
+      "step": 132
+    },
+    {
+      "epoch": 0.034048,
+      "grad_norm": 1.953125,
+      "learning_rate": 4.9436908113642185e-05,
+      "loss": 1.8338,
+      "step": 133
+    },
+    {
+      "epoch": 0.034304,
+      "grad_norm": 1.84375,
+      "learning_rate": 4.943264226601826e-05,
+      "loss": 1.7488,
+      "step": 134
+    },
+    {
+      "epoch": 0.03456,
+      "grad_norm": 1.953125,
+      "learning_rate": 4.942837641839434e-05,
+      "loss": 1.7858,
+      "step": 135
+    },
+    {
+      "epoch": 0.034816,
+      "grad_norm": 1.7421875,
+      "learning_rate": 4.9424110570770415e-05,
+      "loss": 1.7527,
+      "step": 136
+    },
+    {
+      "epoch": 0.035072,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.941984472314649e-05,
+      "loss": 1.8653,
+      "step": 137
+    },
+    {
+      "epoch": 0.035328,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.941557887552257e-05,
+      "loss": 1.7514,
+      "step": 138
+    },
+    {
+      "epoch": 0.035584,
+      "grad_norm": 1.8203125,
+      "learning_rate": 4.941131302789865e-05,
+      "loss": 1.775,
+      "step": 139
+    },
+    {
+      "epoch": 0.03584,
+      "grad_norm": 1.765625,
+      "learning_rate": 4.9407047180274725e-05,
+      "loss": 1.7787,
+      "step": 140
+    },
+    {
+      "epoch": 0.036096,
+      "grad_norm": 1.7890625,
+      "learning_rate": 4.94027813326508e-05,
+      "loss": 1.7926,
+      "step": 141
+    },
+    {
+      "epoch": 0.036352,
+      "grad_norm": 1.9140625,
+      "learning_rate": 4.9398515485026874e-05,
+      "loss": 1.7329,
+      "step": 142
+    },
+    {
+      "epoch": 0.036608,
+      "grad_norm": 1.8359375,
+      "learning_rate": 4.9394249637402955e-05,
+      "loss": 1.7812,
+      "step": 143
+    },
+    {
+      "epoch": 0.036864,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.938998378977903e-05,
+      "loss": 1.7115,
+      "step": 144
+    },
+    {
+      "epoch": 0.03712,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.938571794215511e-05,
+      "loss": 1.7321,
+      "step": 145
+    },
+    {
+      "epoch": 0.037376,
+      "grad_norm": 1.84375,
+      "learning_rate": 4.9381452094531184e-05,
+      "loss": 1.8258,
+      "step": 146
+    },
+    {
+      "epoch": 0.037632,
+      "grad_norm": 2.015625,
+      "learning_rate": 4.9377186246907265e-05,
+      "loss": 1.7198,
+      "step": 147
+    },
+    {
+      "epoch": 0.037888,
+      "grad_norm": 1.9140625,
+      "learning_rate": 4.937292039928334e-05,
+      "loss": 1.7018,
+      "step": 148
+    },
+    {
+      "epoch": 0.038144,
+      "grad_norm": 1.921875,
+      "learning_rate": 4.9368654551659414e-05,
+      "loss": 1.7487,
+      "step": 149
+    },
+    {
+      "epoch": 0.0384,
+      "grad_norm": 1.90625,
+      "learning_rate": 4.9364388704035495e-05,
+      "loss": 1.7039,
+      "step": 150
+    },
+    {
+      "epoch": 0.038656,
+      "grad_norm": 1.78125,
+      "learning_rate": 4.936012285641157e-05,
+      "loss": 1.7421,
+      "step": 151
+    },
+    {
+      "epoch": 0.038912,
+      "grad_norm": 1.84375,
+      "learning_rate": 4.935585700878765e-05,
+      "loss": 1.6556,
+      "step": 152
+    },
+    {
+      "epoch": 0.039168,
+      "grad_norm": 1.90625,
+      "learning_rate": 4.9351591161163725e-05,
+      "loss": 1.7277,
+      "step": 153
+    },
+    {
+      "epoch": 0.039424,
+      "grad_norm": 1.90625,
+      "learning_rate": 4.9347325313539806e-05,
+      "loss": 1.7489,
+      "step": 154
+    },
+    {
+      "epoch": 0.03968,
+      "grad_norm": 2.0,
+      "learning_rate": 4.934305946591588e-05,
+      "loss": 1.7052,
+      "step": 155
+    },
+    {
+      "epoch": 0.039936,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.9338793618291954e-05,
+      "loss": 1.7708,
+      "step": 156
+    },
+    {
+      "epoch": 0.040192,
+      "grad_norm": 1.9609375,
+      "learning_rate": 4.9334527770668035e-05,
+      "loss": 1.6825,
+      "step": 157
+    },
+    {
+      "epoch": 0.040448,
+      "grad_norm": 1.84375,
+      "learning_rate": 4.9330261923044116e-05,
+      "loss": 1.7026,
+      "step": 158
+    },
+    {
+      "epoch": 0.040704,
+      "grad_norm": 1.921875,
+      "learning_rate": 4.932599607542019e-05,
+      "loss": 1.7233,
+      "step": 159
+    },
+    {
+      "epoch": 0.04096,
+      "grad_norm": 1.984375,
+      "learning_rate": 4.9321730227796265e-05,
+      "loss": 1.7464,
+      "step": 160
+    },
+    {
+      "epoch": 0.041216,
+      "grad_norm": 1.8984375,
+      "learning_rate": 4.931746438017234e-05,
+      "loss": 1.6769,
+      "step": 161
+    },
+    {
+      "epoch": 0.041472,
+      "grad_norm": 1.7734375,
+      "learning_rate": 4.931319853254842e-05,
+      "loss": 1.7106,
+      "step": 162
+    },
+    {
+      "epoch": 0.041728,
+      "grad_norm": 1.9296875,
+      "learning_rate": 4.9308932684924494e-05,
+      "loss": 1.6307,
+      "step": 163
+    },
+    {
+      "epoch": 0.041984,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.9304666837300575e-05,
+      "loss": 1.7086,
+      "step": 164
+    },
+    {
+      "epoch": 0.04224,
+      "grad_norm": 2.03125,
+      "learning_rate": 4.9300400989676656e-05,
+      "loss": 1.6809,
+      "step": 165
+    },
+    {
+      "epoch": 0.042496,
+      "grad_norm": 1.875,
+      "learning_rate": 4.929613514205273e-05,
+      "loss": 1.6646,
+      "step": 166
+    },
+    {
+      "epoch": 0.042752,
+      "grad_norm": 1.890625,
+      "learning_rate": 4.9291869294428805e-05,
+      "loss": 1.6178,
+      "step": 167
+    },
+    {
+      "epoch": 0.043008,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.928760344680488e-05,
+      "loss": 1.7098,
+      "step": 168
+    },
+    {
+      "epoch": 0.043264,
+      "grad_norm": 1.796875,
+      "learning_rate": 4.928333759918096e-05,
+      "loss": 1.6471,
+      "step": 169
+    },
+    {
+      "epoch": 0.04352,
+      "grad_norm": 2.0,
+      "learning_rate": 4.9279071751557035e-05,
+      "loss": 1.7198,
+      "step": 170
+    },
+    {
+      "epoch": 0.043776,
+      "grad_norm": 1.9921875,
+      "learning_rate": 4.9274805903933116e-05,
+      "loss": 1.6987,
+      "step": 171
+    },
+    {
+      "epoch": 0.044032,
+      "grad_norm": 1.9609375,
+      "learning_rate": 4.927054005630919e-05,
+      "loss": 1.6829,
+      "step": 172
+    },
+    {
+      "epoch": 0.044288,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.926627420868527e-05,
+      "loss": 1.7319,
+      "step": 173
+    },
+    {
+      "epoch": 0.044544,
+      "grad_norm": 2.046875,
+      "learning_rate": 4.9262008361061345e-05,
+      "loss": 1.6764,
+      "step": 174
+    },
+    {
+      "epoch": 0.0448,
+      "grad_norm": 1.9140625,
+      "learning_rate": 4.925774251343742e-05,
+      "loss": 1.6083,
+      "step": 175
+    },
+    {
+      "epoch": 0.045056,
+      "grad_norm": 2.03125,
+      "learning_rate": 4.92534766658135e-05,
+      "loss": 1.6746,
+      "step": 176
+    },
+    {
+      "epoch": 0.045312,
+      "grad_norm": 2.0,
+      "learning_rate": 4.924921081818958e-05,
+      "loss": 1.6424,
+      "step": 177
+    },
+    {
+      "epoch": 0.045568,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.9244944970565656e-05,
+      "loss": 1.6847,
+      "step": 178
+    },
+    {
+      "epoch": 0.045824,
+      "grad_norm": 2.046875,
+      "learning_rate": 4.924067912294173e-05,
+      "loss": 1.669,
+      "step": 179
+    },
+    {
+      "epoch": 0.04608,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.9236413275317804e-05,
+      "loss": 1.6319,
+      "step": 180
+    },
+    {
+      "epoch": 0.046336,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.9232147427693885e-05,
+      "loss": 1.6182,
+      "step": 181
+    },
+    {
+      "epoch": 0.046592,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.922788158006996e-05,
+      "loss": 1.6482,
+      "step": 182
+    },
+    {
+      "epoch": 0.046848,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.922361573244604e-05,
+      "loss": 1.5928,
+      "step": 183
+    },
+    {
+      "epoch": 0.047104,
+      "grad_norm": 1.96875,
+      "learning_rate": 4.921934988482212e-05,
+      "loss": 1.6464,
+      "step": 184
+    },
+    {
+      "epoch": 0.04736,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.9215084037198196e-05,
+      "loss": 1.6083,
+      "step": 185
+    },
+    {
+      "epoch": 0.047616,
+      "grad_norm": 1.9296875,
+      "learning_rate": 4.921081818957427e-05,
+      "loss": 1.5948,
+      "step": 186
+    },
+    {
+      "epoch": 0.047872,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.9206552341950344e-05,
+      "loss": 1.6405,
+      "step": 187
+    },
+    {
+      "epoch": 0.048128,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.9202286494326426e-05,
+      "loss": 1.6457,
+      "step": 188
+    },
+    {
+      "epoch": 0.048384,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.91980206467025e-05,
+      "loss": 1.6023,
+      "step": 189
+    },
+    {
+      "epoch": 0.04864,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.919375479907858e-05,
+      "loss": 1.6197,
+      "step": 190
+    },
+    {
+      "epoch": 0.048896,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.9189488951454655e-05,
+      "loss": 1.6202,
+      "step": 191
+    },
+    {
+      "epoch": 0.049152,
+      "grad_norm": 2.015625,
+      "learning_rate": 4.9185223103830736e-05,
+      "loss": 1.5307,
+      "step": 192
+    },
+    {
+      "epoch": 0.049408,
+      "grad_norm": 2.125,
+      "learning_rate": 4.918095725620681e-05,
+      "loss": 1.6035,
+      "step": 193
+    },
+    {
+      "epoch": 0.049664,
+      "grad_norm": 2.0,
+      "learning_rate": 4.9176691408582885e-05,
+      "loss": 1.611,
+      "step": 194
+    },
+    {
+      "epoch": 0.04992,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.9172425560958966e-05,
+      "loss": 1.6442,
+      "step": 195
+    },
+    {
+      "epoch": 0.050176,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.916815971333505e-05,
+      "loss": 1.5583,
+      "step": 196
+    },
+    {
+      "epoch": 0.050432,
+      "grad_norm": 1.7734375,
+      "learning_rate": 4.916389386571112e-05,
+      "loss": 1.5812,
+      "step": 197
+    },
+    {
+      "epoch": 0.050688,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.9159628018087195e-05,
+      "loss": 1.6093,
+      "step": 198
+    },
+    {
+      "epoch": 0.050944,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.9155362170463276e-05,
+      "loss": 1.5823,
+      "step": 199
+    },
+    {
+      "epoch": 0.0512,
+      "grad_norm": 1.984375,
+      "learning_rate": 4.915109632283935e-05,
+      "loss": 1.6072,
+      "step": 200
+    },
+    {
+      "epoch": 0.051456,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.9146830475215425e-05,
+      "loss": 1.6181,
+      "step": 201
+    },
+    {
+      "epoch": 0.051712,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.9142564627591506e-05,
+      "loss": 1.5957,
+      "step": 202
+    },
+    {
+      "epoch": 0.051968,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.913829877996759e-05,
+      "loss": 1.5871,
+      "step": 203
+    },
+    {
+      "epoch": 0.052224,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.913403293234366e-05,
+      "loss": 1.6018,
+      "step": 204
+    },
+    {
+      "epoch": 0.05248,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.9129767084719735e-05,
+      "loss": 1.6011,
+      "step": 205
+    },
+    {
+      "epoch": 0.052736,
+      "grad_norm": 1.984375,
+      "learning_rate": 4.912550123709581e-05,
+      "loss": 1.5924,
+      "step": 206
+    },
+    {
+      "epoch": 0.052992,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.912123538947189e-05,
+      "loss": 1.6036,
+      "step": 207
+    },
+    {
+      "epoch": 0.053248,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.9116969541847965e-05,
+      "loss": 1.5979,
+      "step": 208
+    },
+    {
+      "epoch": 0.053504,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.9112703694224046e-05,
+      "loss": 1.5598,
+      "step": 209
+    },
+    {
+      "epoch": 0.05376,
+      "grad_norm": 2.03125,
+      "learning_rate": 4.910843784660012e-05,
+      "loss": 1.5844,
+      "step": 210
+    },
+    {
+      "epoch": 0.054016,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.91041719989762e-05,
+      "loss": 1.5775,
+      "step": 211
+    },
+    {
+      "epoch": 0.054272,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.9099906151352276e-05,
+      "loss": 1.5797,
+      "step": 212
+    },
+    {
+      "epoch": 0.054528,
+      "grad_norm": 2.03125,
+      "learning_rate": 4.909564030372835e-05,
+      "loss": 1.5664,
+      "step": 213
+    },
+    {
+      "epoch": 0.054784,
+      "grad_norm": 2.015625,
+      "learning_rate": 4.9091374456104424e-05,
+      "loss": 1.5471,
+      "step": 214
+    },
+    {
+      "epoch": 0.05504,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.908710860848051e-05,
+      "loss": 1.5752,
+      "step": 215
+    },
+    {
+      "epoch": 0.055296,
+      "grad_norm": 1.84375,
+      "learning_rate": 4.9082842760856586e-05,
+      "loss": 1.5398,
+      "step": 216
+    },
+    {
+      "epoch": 0.055552,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.907857691323266e-05,
+      "loss": 1.5433,
+      "step": 217
+    },
+    {
+      "epoch": 0.055808,
+      "grad_norm": 2.125,
+      "learning_rate": 4.907431106560874e-05,
+      "loss": 1.5691,
+      "step": 218
+    },
+    {
+      "epoch": 0.056064,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.9070045217984816e-05,
+      "loss": 1.4747,
+      "step": 219
+    },
+    {
+      "epoch": 0.05632,
+      "grad_norm": 2.25,
+      "learning_rate": 4.906577937036089e-05,
+      "loss": 1.5545,
+      "step": 220
+    },
+    {
+      "epoch": 0.056576,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.906151352273697e-05,
+      "loss": 1.5337,
+      "step": 221
+    },
+    {
+      "epoch": 0.056832,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.905724767511305e-05,
+      "loss": 1.5217,
+      "step": 222
+    },
+    {
+      "epoch": 0.057088,
+      "grad_norm": 1.9921875,
+      "learning_rate": 4.9052981827489126e-05,
+      "loss": 1.5326,
+      "step": 223
+    },
+    {
+      "epoch": 0.057344,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.90487159798652e-05,
+      "loss": 1.5819,
+      "step": 224
+    },
+    {
+      "epoch": 0.0576,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.9044450132241275e-05,
+      "loss": 1.5518,
+      "step": 225
+    },
+    {
+      "epoch": 0.057856,
+      "grad_norm": 1.9609375,
+      "learning_rate": 4.9040184284617356e-05,
+      "loss": 1.5557,
+      "step": 226
+    },
+    {
+      "epoch": 0.058112,
+      "grad_norm": 2.015625,
+      "learning_rate": 4.903591843699343e-05,
+      "loss": 1.5525,
+      "step": 227
+    },
+    {
+      "epoch": 0.058368,
+      "grad_norm": 2.375,
+      "learning_rate": 4.903165258936951e-05,
+      "loss": 1.5862,
+      "step": 228
+    },
+    {
+      "epoch": 0.058624,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.9027386741745586e-05,
+      "loss": 1.551,
+      "step": 229
+    },
+    {
+      "epoch": 0.05888,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.902312089412167e-05,
+      "loss": 1.5245,
+      "step": 230
+    },
+    {
+      "epoch": 0.059136,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.901885504649774e-05,
+      "loss": 1.555,
+      "step": 231
+    },
+    {
+      "epoch": 0.059392,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.9014589198873815e-05,
+      "loss": 1.5364,
+      "step": 232
+    },
+    {
+      "epoch": 0.059648,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.9010323351249896e-05,
+      "loss": 1.5301,
+      "step": 233
+    },
+    {
+      "epoch": 0.059904,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.900605750362598e-05,
+      "loss": 1.5062,
+      "step": 234
+    },
+    {
+      "epoch": 0.06016,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.900179165600205e-05,
+      "loss": 1.488,
+      "step": 235
+    },
+    {
+      "epoch": 0.060416,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.8997525808378126e-05,
+      "loss": 1.4925,
+      "step": 236
+    },
+    {
+      "epoch": 0.060672,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.899325996075421e-05,
+      "loss": 1.5577,
+      "step": 237
+    },
+    {
+      "epoch": 0.060928,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.898899411313028e-05,
+      "loss": 1.5758,
+      "step": 238
+    },
+    {
+      "epoch": 0.061184,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.8984728265506355e-05,
+      "loss": 1.5494,
+      "step": 239
+    },
+    {
+      "epoch": 0.06144,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.8980462417882436e-05,
+      "loss": 1.4806,
+      "step": 240
+    },
+    {
+      "epoch": 0.061696,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.897619657025852e-05,
+      "loss": 1.5353,
+      "step": 241
+    },
+    {
+      "epoch": 0.061952,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.897193072263459e-05,
+      "loss": 1.4716,
+      "step": 242
+    },
+    {
+      "epoch": 0.062208,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.8967664875010666e-05,
+      "loss": 1.4682,
+      "step": 243
+    },
+    {
+      "epoch": 0.062464,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.896339902738674e-05,
+      "loss": 1.5145,
+      "step": 244
+    },
+    {
+      "epoch": 0.06272,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.895913317976282e-05,
+      "loss": 1.5411,
+      "step": 245
+    },
+    {
+      "epoch": 0.062976,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.8954867332138896e-05,
+      "loss": 1.5289,
+      "step": 246
+    },
+    {
+      "epoch": 0.063232,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.895060148451498e-05,
+      "loss": 1.496,
+      "step": 247
+    },
+    {
+      "epoch": 0.063488,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.894633563689106e-05,
+      "loss": 1.5183,
+      "step": 248
+    },
+    {
+      "epoch": 0.063744,
+      "grad_norm": 2.375,
+      "learning_rate": 4.894206978926713e-05,
+      "loss": 1.476,
+      "step": 249
+    },
+    {
+      "epoch": 0.064,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.8937803941643206e-05,
+      "loss": 1.5199,
+      "step": 250
+    },
+    {
+      "epoch": 0.064256,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.893353809401928e-05,
+      "loss": 1.5078,
+      "step": 251
+    },
+    {
+      "epoch": 0.064512,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.892927224639536e-05,
+      "loss": 1.4749,
+      "step": 252
+    },
+    {
+      "epoch": 0.064768,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.8925006398771436e-05,
+      "loss": 1.4371,
+      "step": 253
+    },
+    {
+      "epoch": 0.065024,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.892074055114752e-05,
+      "loss": 1.4957,
+      "step": 254
+    },
+    {
+      "epoch": 0.06528,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.891647470352359e-05,
+      "loss": 1.511,
+      "step": 255
+    },
+    {
+      "epoch": 0.065536,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.891220885589967e-05,
+      "loss": 1.4942,
+      "step": 256
+    },
+    {
+      "epoch": 0.065792,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.8907943008275746e-05,
+      "loss": 1.4961,
+      "step": 257
+    },
+    {
+      "epoch": 0.066048,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.890367716065182e-05,
+      "loss": 1.4534,
+      "step": 258
+    },
+    {
+      "epoch": 0.066304,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.88994113130279e-05,
+      "loss": 1.4739,
+      "step": 259
+    },
+    {
+      "epoch": 0.06656,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.889514546540398e-05,
+      "loss": 1.4574,
+      "step": 260
+    },
+    {
+      "epoch": 0.066816,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.889087961778006e-05,
+      "loss": 1.474,
+      "step": 261
+    },
+    {
+      "epoch": 0.067072,
+      "grad_norm": 2.375,
+      "learning_rate": 4.888661377015613e-05,
+      "loss": 1.48,
+      "step": 262
+    },
+    {
+      "epoch": 0.067328,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.8882347922532206e-05,
+      "loss": 1.4356,
+      "step": 263
+    },
+    {
+      "epoch": 0.067584,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.8878082074908287e-05,
+      "loss": 1.4966,
+      "step": 264
+    },
+    {
+      "epoch": 0.06784,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.887381622728436e-05,
+      "loss": 1.4575,
+      "step": 265
+    },
+    {
+      "epoch": 0.068096,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.886955037966044e-05,
+      "loss": 1.4602,
+      "step": 266
+    },
+    {
+      "epoch": 0.068352,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.886528453203652e-05,
+      "loss": 1.4882,
+      "step": 267
+    },
+    {
+      "epoch": 0.068608,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.88610186844126e-05,
+      "loss": 1.4301,
+      "step": 268
+    },
+    {
+      "epoch": 0.068864,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.885675283678867e-05,
+      "loss": 1.4798,
+      "step": 269
+    },
+    {
+      "epoch": 0.06912,
+      "grad_norm": 2.625,
+      "learning_rate": 4.8852486989164746e-05,
+      "loss": 1.4329,
+      "step": 270
+    },
+    {
+      "epoch": 0.069376,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.884822114154083e-05,
+      "loss": 1.4347,
+      "step": 271
+    },
+    {
+      "epoch": 0.069632,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.88439552939169e-05,
+      "loss": 1.4584,
+      "step": 272
+    },
+    {
+      "epoch": 0.069888,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.883968944629298e-05,
+      "loss": 1.4704,
+      "step": 273
+    },
+    {
+      "epoch": 0.070144,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.8835423598669056e-05,
+      "loss": 1.4863,
+      "step": 274
+    },
+    {
+      "epoch": 0.0704,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.883115775104514e-05,
+      "loss": 1.4406,
+      "step": 275
+    },
+    {
+      "epoch": 0.070656,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.882689190342121e-05,
+      "loss": 1.4104,
+      "step": 276
+    },
+    {
+      "epoch": 0.070912,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.8822626055797286e-05,
+      "loss": 1.4123,
+      "step": 277
+    },
+    {
+      "epoch": 0.071168,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.881836020817337e-05,
+      "loss": 1.4639,
+      "step": 278
+    },
+    {
+      "epoch": 0.071424,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.881409436054945e-05,
+      "loss": 1.445,
+      "step": 279
+    },
+    {
+      "epoch": 0.07168,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.880982851292552e-05,
+      "loss": 1.5113,
+      "step": 280
+    },
+    {
+      "epoch": 0.071936,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.8805562665301597e-05,
+      "loss": 1.4277,
+      "step": 281
+    },
+    {
+      "epoch": 0.072192,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.880129681767768e-05,
+      "loss": 1.4616,
+      "step": 282
+    },
+    {
+      "epoch": 0.072448,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.879703097005375e-05,
+      "loss": 1.4176,
+      "step": 283
+    },
+    {
+      "epoch": 0.072704,
+      "grad_norm": 2.5,
+      "learning_rate": 4.8792765122429826e-05,
+      "loss": 1.4402,
+      "step": 284
+    },
+    {
+      "epoch": 0.07296,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.878849927480591e-05,
+      "loss": 1.4465,
+      "step": 285
+    },
+    {
+      "epoch": 0.073216,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.878423342718199e-05,
+      "loss": 1.491,
+      "step": 286
+    },
+    {
+      "epoch": 0.073472,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.877996757955806e-05,
+      "loss": 1.4301,
+      "step": 287
+    },
+    {
+      "epoch": 0.073728,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.877570173193414e-05,
+      "loss": 1.4456,
+      "step": 288
+    },
+    {
+      "epoch": 0.073984,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.877143588431021e-05,
+      "loss": 1.4613,
+      "step": 289
+    },
+    {
+      "epoch": 0.07424,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.876717003668629e-05,
+      "loss": 1.4438,
+      "step": 290
+    },
+    {
+      "epoch": 0.074496,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.8762904189062366e-05,
+      "loss": 1.4258,
+      "step": 291
+    },
+    {
+      "epoch": 0.074752,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.875863834143845e-05,
+      "loss": 1.4418,
+      "step": 292
+    },
+    {
+      "epoch": 0.075008,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.875437249381452e-05,
+      "loss": 1.4312,
+      "step": 293
+    },
+    {
+      "epoch": 0.075264,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.87501066461906e-05,
+      "loss": 1.4274,
+      "step": 294
+    },
+    {
+      "epoch": 0.07552,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.874584079856668e-05,
+      "loss": 1.4523,
+      "step": 295
+    },
+    {
+      "epoch": 0.075776,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.874157495094275e-05,
+      "loss": 1.4219,
+      "step": 296
+    },
+    {
+      "epoch": 0.076032,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.873730910331883e-05,
+      "loss": 1.4023,
+      "step": 297
+    },
+    {
+      "epoch": 0.076288,
+      "grad_norm": 2.375,
+      "learning_rate": 4.873304325569491e-05,
+      "loss": 1.3625,
+      "step": 298
+    },
+    {
+      "epoch": 0.076544,
+      "grad_norm": 2.25,
+      "learning_rate": 4.872877740807099e-05,
+      "loss": 1.3942,
+      "step": 299
+    },
+    {
+      "epoch": 0.0768,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.872451156044706e-05,
+      "loss": 1.4187,
+      "step": 300
+    },
+    {
+      "epoch": 0.077056,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.872024571282314e-05,
+      "loss": 1.438,
+      "step": 301
+    },
+    {
+      "epoch": 0.077312,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.871597986519922e-05,
+      "loss": 1.3752,
+      "step": 302
+    },
+    {
+      "epoch": 0.077568,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.871171401757529e-05,
+      "loss": 1.3774,
+      "step": 303
+    },
+    {
+      "epoch": 0.077824,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.870744816995137e-05,
+      "loss": 1.393,
+      "step": 304
+    },
+    {
+      "epoch": 0.07808,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.8703182322327453e-05,
+      "loss": 1.4618,
+      "step": 305
+    },
+    {
+      "epoch": 0.078336,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.869891647470353e-05,
+      "loss": 1.4248,
+      "step": 306
+    },
+    {
+      "epoch": 0.078592,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.86946506270796e-05,
+      "loss": 1.3912,
+      "step": 307
+    },
+    {
+      "epoch": 0.078848,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.8690384779455676e-05,
+      "loss": 1.407,
+      "step": 308
+    },
+    {
+      "epoch": 0.079104,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.868611893183176e-05,
+      "loss": 1.4209,
+      "step": 309
+    },
+    {
+      "epoch": 0.07936,
+      "grad_norm": 2.25,
+      "learning_rate": 4.868185308420783e-05,
+      "loss": 1.3726,
+      "step": 310
+    },
+    {
+      "epoch": 0.079616,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.867758723658391e-05,
+      "loss": 1.3994,
+      "step": 311
+    },
+    {
+      "epoch": 0.079872,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.867332138895999e-05,
+      "loss": 1.3895,
+      "step": 312
+    },
+    {
+      "epoch": 0.080128,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.866905554133607e-05,
+      "loss": 1.3558,
+      "step": 313
+    },
+    {
+      "epoch": 0.080384,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.866478969371214e-05,
+      "loss": 1.394,
+      "step": 314
+    },
+    {
+      "epoch": 0.08064,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.8660523846088216e-05,
+      "loss": 1.4176,
+      "step": 315
+    },
+    {
+      "epoch": 0.080896,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.86562579984643e-05,
+      "loss": 1.4403,
+      "step": 316
+    },
+    {
+      "epoch": 0.081152,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.865199215084038e-05,
+      "loss": 1.3934,
+      "step": 317
+    },
+    {
+      "epoch": 0.081408,
+      "grad_norm": 2.625,
+      "learning_rate": 4.864772630321645e-05,
+      "loss": 1.4171,
+      "step": 318
+    },
+    {
+      "epoch": 0.081664,
+      "grad_norm": 2.375,
+      "learning_rate": 4.864346045559253e-05,
+      "loss": 1.3782,
+      "step": 319
+    },
+    {
+      "epoch": 0.08192,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.863919460796861e-05,
+      "loss": 1.3604,
+      "step": 320
+    },
+    {
+      "epoch": 0.082176,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.863492876034468e-05,
+      "loss": 1.3604,
+      "step": 321
+    },
+    {
+      "epoch": 0.082432,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.863066291272076e-05,
+      "loss": 1.3712,
+      "step": 322
+    },
+    {
+      "epoch": 0.082688,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.862639706509684e-05,
+      "loss": 1.3733,
+      "step": 323
+    },
+    {
+      "epoch": 0.082944,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.862213121747292e-05,
+      "loss": 1.3858,
+      "step": 324
+    },
+    {
+      "epoch": 0.0832,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.861786536984899e-05,
+      "loss": 1.3398,
+      "step": 325
+    },
+    {
+      "epoch": 0.083456,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.861359952222507e-05,
+      "loss": 1.3975,
+      "step": 326
+    },
+    {
+      "epoch": 0.083712,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.860933367460114e-05,
+      "loss": 1.3495,
+      "step": 327
+    },
+    {
+      "epoch": 0.083968,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.860506782697722e-05,
+      "loss": 1.3591,
+      "step": 328
+    },
+    {
+      "epoch": 0.084224,
+      "grad_norm": 2.375,
+      "learning_rate": 4.86008019793533e-05,
+      "loss": 1.3565,
+      "step": 329
+    },
+    {
+      "epoch": 0.08448,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.859653613172938e-05,
+      "loss": 1.3631,
+      "step": 330
+    },
+    {
+      "epoch": 0.084736,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.859227028410546e-05,
+      "loss": 1.3668,
+      "step": 331
+    },
+    {
+      "epoch": 0.084992,
+      "grad_norm": 2.25,
+      "learning_rate": 4.858800443648153e-05,
+      "loss": 1.3649,
+      "step": 332
+    },
+    {
+      "epoch": 0.085248,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.858373858885761e-05,
+      "loss": 1.3302,
+      "step": 333
+    },
+    {
+      "epoch": 0.085504,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.857947274123368e-05,
+      "loss": 1.3837,
+      "step": 334
+    },
+    {
+      "epoch": 0.08576,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.857520689360976e-05,
+      "loss": 1.3522,
+      "step": 335
+    },
+    {
+      "epoch": 0.086016,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.8570941045985844e-05,
+      "loss": 1.3364,
+      "step": 336
+    },
+    {
+      "epoch": 0.086272,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.856667519836192e-05,
+      "loss": 1.3699,
+      "step": 337
+    },
+    {
+      "epoch": 0.086528,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.856240935073799e-05,
+      "loss": 1.3882,
+      "step": 338
+    },
+    {
+      "epoch": 0.086784,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.8558143503114073e-05,
+      "loss": 1.3487,
+      "step": 339
+    },
+    {
+      "epoch": 0.08704,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.855387765549015e-05,
+      "loss": 1.3729,
+      "step": 340
+    },
+    {
+      "epoch": 0.087296,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.854961180786622e-05,
+      "loss": 1.361,
+      "step": 341
+    },
+    {
+      "epoch": 0.087552,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.85453459602423e-05,
+      "loss": 1.3687,
+      "step": 342
+    },
+    {
+      "epoch": 0.087808,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.8541080112618384e-05,
+      "loss": 1.3309,
+      "step": 343
+    },
+    {
+      "epoch": 0.088064,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.853681426499446e-05,
+      "loss": 1.3244,
+      "step": 344
+    },
+    {
+      "epoch": 0.08832,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.853254841737053e-05,
+      "loss": 1.3232,
+      "step": 345
+    },
+    {
+      "epoch": 0.088576,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.8528282569746614e-05,
+      "loss": 1.3778,
+      "step": 346
+    },
+    {
+      "epoch": 0.088832,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.852401672212269e-05,
+      "loss": 1.3067,
+      "step": 347
+    },
+    {
+      "epoch": 0.089088,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.851975087449876e-05,
+      "loss": 1.3457,
+      "step": 348
+    },
+    {
+      "epoch": 0.089344,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.851548502687484e-05,
+      "loss": 1.3158,
+      "step": 349
+    },
+    {
+      "epoch": 0.0896,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.8511219179250924e-05,
+      "loss": 1.3724,
+      "step": 350
+    },
+    {
+      "epoch": 0.089856,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.8506953331627e-05,
+      "loss": 1.3133,
+      "step": 351
+    },
+    {
+      "epoch": 0.090112,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.850268748400307e-05,
+      "loss": 1.2902,
+      "step": 352
+    },
+    {
+      "epoch": 0.090368,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.849842163637915e-05,
+      "loss": 1.3383,
+      "step": 353
+    },
+    {
+      "epoch": 0.090624,
+      "grad_norm": 2.375,
+      "learning_rate": 4.849415578875523e-05,
+      "loss": 1.3155,
+      "step": 354
+    },
+    {
+      "epoch": 0.09088,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.84898899411313e-05,
+      "loss": 1.3205,
+      "step": 355
+    },
+    {
+      "epoch": 0.091136,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.848562409350738e-05,
+      "loss": 1.3488,
+      "step": 356
+    },
+    {
+      "epoch": 0.091392,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.848135824588346e-05,
+      "loss": 1.326,
+      "step": 357
+    },
+    {
+      "epoch": 0.091648,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.847709239825954e-05,
+      "loss": 1.356,
+      "step": 358
+    },
+    {
+      "epoch": 0.091904,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.847282655063561e-05,
+      "loss": 1.3191,
+      "step": 359
+    },
+    {
+      "epoch": 0.09216,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.846856070301169e-05,
+      "loss": 1.3219,
+      "step": 360
+    },
+    {
+      "epoch": 0.092416,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.846429485538777e-05,
+      "loss": 1.3356,
+      "step": 361
+    },
+    {
+      "epoch": 0.092672,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.846002900776385e-05,
+      "loss": 1.2743,
+      "step": 362
+    },
+    {
+      "epoch": 0.092928,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.8455763160139924e-05,
+      "loss": 1.3086,
+      "step": 363
+    },
+    {
+      "epoch": 0.093184,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.8451497312516e-05,
+      "loss": 1.3146,
+      "step": 364
+    },
+    {
+      "epoch": 0.09344,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.844723146489208e-05,
+      "loss": 1.3004,
+      "step": 365
+    },
+    {
+      "epoch": 0.093696,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.844296561726815e-05,
+      "loss": 1.3229,
+      "step": 366
+    },
+    {
+      "epoch": 0.093952,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.843869976964423e-05,
+      "loss": 1.3104,
+      "step": 367
+    },
+    {
+      "epoch": 0.094208,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.843443392202031e-05,
+      "loss": 1.2928,
+      "step": 368
+    },
+    {
+      "epoch": 0.094464,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.843016807439639e-05,
+      "loss": 1.3326,
+      "step": 369
+    },
+    {
+      "epoch": 0.09472,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.8425902226772464e-05,
+      "loss": 1.3077,
+      "step": 370
+    },
+    {
+      "epoch": 0.094976,
+      "grad_norm": 2.5,
+      "learning_rate": 4.842163637914854e-05,
+      "loss": 1.3201,
+      "step": 371
+    },
+    {
+      "epoch": 0.095232,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.841737053152461e-05,
+      "loss": 1.3025,
+      "step": 372
+    },
+    {
+      "epoch": 0.095488,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.841310468390069e-05,
+      "loss": 1.2899,
+      "step": 373
+    },
+    {
+      "epoch": 0.095744,
+      "grad_norm": 2.375,
+      "learning_rate": 4.840883883627677e-05,
+      "loss": 1.314,
+      "step": 374
+    },
+    {
+      "epoch": 0.096,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.840457298865285e-05,
+      "loss": 1.3321,
+      "step": 375
+    },
+    {
+      "epoch": 0.096256,
+      "grad_norm": 2.625,
+      "learning_rate": 4.840030714102892e-05,
+      "loss": 1.3422,
+      "step": 376
+    },
+    {
+      "epoch": 0.096512,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.8396041293405004e-05,
+      "loss": 1.3137,
+      "step": 377
+    },
+    {
+      "epoch": 0.096768,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.839177544578108e-05,
+      "loss": 1.3039,
+      "step": 378
+    },
+    {
+      "epoch": 0.097024,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.838750959815715e-05,
+      "loss": 1.3006,
+      "step": 379
+    },
+    {
+      "epoch": 0.09728,
+      "grad_norm": 2.375,
+      "learning_rate": 4.8383243750533234e-05,
+      "loss": 1.3158,
+      "step": 380
+    },
+    {
+      "epoch": 0.097536,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.8378977902909315e-05,
+      "loss": 1.2508,
+      "step": 381
+    },
+    {
+      "epoch": 0.097792,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.837471205528539e-05,
+      "loss": 1.2765,
+      "step": 382
+    },
+    {
+      "epoch": 0.098048,
+      "grad_norm": 2.625,
+      "learning_rate": 4.837044620766146e-05,
+      "loss": 1.3461,
+      "step": 383
+    },
+    {
+      "epoch": 0.098304,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.8366180360037544e-05,
+      "loss": 1.3068,
+      "step": 384
+    },
+    {
+      "epoch": 0.09856,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.836191451241362e-05,
+      "loss": 1.2956,
+      "step": 385
+    },
+    {
+      "epoch": 0.098816,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.835764866478969e-05,
+      "loss": 1.3028,
+      "step": 386
+    },
+    {
+      "epoch": 0.099072,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.8353382817165774e-05,
+      "loss": 1.317,
+      "step": 387
+    },
+    {
+      "epoch": 0.099328,
+      "grad_norm": 3.0,
+      "learning_rate": 4.8349116969541855e-05,
+      "loss": 1.2737,
+      "step": 388
+    },
+    {
+      "epoch": 0.099584,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.834485112191793e-05,
+      "loss": 1.2997,
+      "step": 389
+    },
+    {
+      "epoch": 0.09984,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.8340585274294e-05,
+      "loss": 1.2789,
+      "step": 390
+    },
+    {
+      "epoch": 0.100096,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.833631942667008e-05,
+      "loss": 1.2491,
+      "step": 391
+    },
+    {
+      "epoch": 0.100352,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.833205357904616e-05,
+      "loss": 1.2935,
+      "step": 392
+    },
+    {
+      "epoch": 0.100608,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.832778773142223e-05,
+      "loss": 1.3159,
+      "step": 393
+    },
+    {
+      "epoch": 0.100864,
+      "grad_norm": 2.625,
+      "learning_rate": 4.8323521883798314e-05,
+      "loss": 1.3087,
+      "step": 394
+    },
+    {
+      "epoch": 0.10112,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.8319256036174395e-05,
+      "loss": 1.2751,
+      "step": 395
+    },
+    {
+      "epoch": 0.101376,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.831499018855047e-05,
+      "loss": 1.2759,
+      "step": 396
+    },
+    {
+      "epoch": 0.101632,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.8310724340926543e-05,
+      "loss": 1.2697,
+      "step": 397
+    },
+    {
+      "epoch": 0.101888,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.830645849330262e-05,
+      "loss": 1.2734,
+      "step": 398
+    },
+    {
+      "epoch": 0.102144,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.83021926456787e-05,
+      "loss": 1.2737,
+      "step": 399
+    },
+    {
+      "epoch": 0.1024,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.829792679805478e-05,
+      "loss": 1.2237,
+      "step": 400
+    },
+    {
+      "epoch": 0.102656,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.8293660950430854e-05,
+      "loss": 1.2734,
+      "step": 401
+    },
+    {
+      "epoch": 0.102912,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.828939510280693e-05,
+      "loss": 1.2318,
+      "step": 402
+    },
+    {
+      "epoch": 0.103168,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.828512925518301e-05,
+      "loss": 1.2986,
+      "step": 403
+    },
+    {
+      "epoch": 0.103424,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.8280863407559084e-05,
+      "loss": 1.2212,
+      "step": 404
+    },
+    {
+      "epoch": 0.10368,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.827659755993516e-05,
+      "loss": 1.2677,
+      "step": 405
+    },
+    {
+      "epoch": 0.103936,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.827233171231124e-05,
+      "loss": 1.2345,
+      "step": 406
+    },
+    {
+      "epoch": 0.104192,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.826806586468732e-05,
+      "loss": 1.2529,
+      "step": 407
+    },
+    {
+      "epoch": 0.104448,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.8263800017063394e-05,
+      "loss": 1.2423,
+      "step": 408
+    },
+    {
+      "epoch": 0.104704,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.825953416943947e-05,
+      "loss": 1.2637,
+      "step": 409
+    },
+    {
+      "epoch": 0.10496,
+      "grad_norm": 2.5,
+      "learning_rate": 4.825526832181554e-05,
+      "loss": 1.2396,
+      "step": 410
+    },
+    {
+      "epoch": 0.105216,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.8251002474191624e-05,
+      "loss": 1.23,
+      "step": 411
+    },
+    {
+      "epoch": 0.105472,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.82467366265677e-05,
+      "loss": 1.2238,
+      "step": 412
+    },
+    {
+      "epoch": 0.105728,
+      "grad_norm": 2.625,
+      "learning_rate": 4.824247077894378e-05,
+      "loss": 1.2604,
+      "step": 413
+    },
+    {
+      "epoch": 0.105984,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.823820493131986e-05,
+      "loss": 1.2625,
+      "step": 414
+    },
+    {
+      "epoch": 0.10624,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.8233939083695934e-05,
+      "loss": 1.2954,
+      "step": 415
+    },
+    {
+      "epoch": 0.106496,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.822967323607201e-05,
+      "loss": 1.2746,
+      "step": 416
+    },
+    {
+      "epoch": 0.106752,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.822540738844808e-05,
+      "loss": 1.2788,
+      "step": 417
+    },
+    {
+      "epoch": 0.107008,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.8221141540824164e-05,
+      "loss": 1.2453,
+      "step": 418
+    },
+    {
+      "epoch": 0.107264,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.8216875693200245e-05,
+      "loss": 1.2855,
+      "step": 419
+    },
+    {
+      "epoch": 0.10752,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.821260984557632e-05,
+      "loss": 1.239,
+      "step": 420
+    },
+    {
+      "epoch": 0.107776,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.8208343997952394e-05,
+      "loss": 1.2436,
+      "step": 421
+    },
+    {
+      "epoch": 0.108032,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.8204078150328475e-05,
+      "loss": 1.2382,
+      "step": 422
+    },
+    {
+      "epoch": 0.108288,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.819981230270455e-05,
+      "loss": 1.2441,
+      "step": 423
+    },
+    {
+      "epoch": 0.108544,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.819554645508062e-05,
+      "loss": 1.245,
+      "step": 424
+    },
+    {
+      "epoch": 0.1088,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.8191280607456704e-05,
+      "loss": 1.2207,
+      "step": 425
+    },
+    {
+      "epoch": 0.109056,
+      "grad_norm": 2.625,
+      "learning_rate": 4.8187014759832785e-05,
+      "loss": 1.2481,
+      "step": 426
+    },
+    {
+      "epoch": 0.109312,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.818274891220886e-05,
+      "loss": 1.2372,
+      "step": 427
+    },
+    {
+      "epoch": 0.109568,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.8178483064584934e-05,
+      "loss": 1.2303,
+      "step": 428
+    },
+    {
+      "epoch": 0.109824,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.8174217216961015e-05,
+      "loss": 1.2388,
+      "step": 429
+    },
+    {
+      "epoch": 0.11008,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.816995136933709e-05,
+      "loss": 1.2365,
+      "step": 430
+    },
+    {
+      "epoch": 0.110336,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.8165685521713163e-05,
+      "loss": 1.2435,
+      "step": 431
+    },
+    {
+      "epoch": 0.110592,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.8161419674089244e-05,
+      "loss": 1.225,
+      "step": 432
+    },
+    {
+      "epoch": 0.110848,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.8157153826465325e-05,
+      "loss": 1.1945,
+      "step": 433
+    },
+    {
+      "epoch": 0.111104,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.81528879788414e-05,
+      "loss": 1.2339,
+      "step": 434
+    },
+    {
+      "epoch": 0.11136,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.8148622131217474e-05,
+      "loss": 1.2488,
+      "step": 435
+    },
+    {
+      "epoch": 0.111616,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.814435628359355e-05,
+      "loss": 1.2166,
+      "step": 436
+    },
+    {
+      "epoch": 0.111872,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.814009043596963e-05,
+      "loss": 1.1974,
+      "step": 437
+    },
+    {
+      "epoch": 0.112128,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.813582458834571e-05,
+      "loss": 1.2465,
+      "step": 438
+    },
+    {
+      "epoch": 0.112384,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.8131558740721785e-05,
+      "loss": 1.2608,
+      "step": 439
+    },
+    {
+      "epoch": 0.11264,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.812729289309786e-05,
+      "loss": 1.2065,
+      "step": 440
+    },
+    {
+      "epoch": 0.112896,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.812302704547394e-05,
+      "loss": 1.2481,
+      "step": 441
+    },
+    {
+      "epoch": 0.113152,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.8118761197850014e-05,
+      "loss": 1.2284,
+      "step": 442
+    },
+    {
+      "epoch": 0.113408,
+      "grad_norm": 2.375,
+      "learning_rate": 4.811449535022609e-05,
+      "loss": 1.2119,
+      "step": 443
+    },
+    {
+      "epoch": 0.113664,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.811022950260217e-05,
+      "loss": 1.1955,
+      "step": 444
+    },
+    {
+      "epoch": 0.11392,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.810596365497825e-05,
+      "loss": 1.2275,
+      "step": 445
+    },
+    {
+      "epoch": 0.114176,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.8101697807354325e-05,
+      "loss": 1.2163,
+      "step": 446
+    },
+    {
+      "epoch": 0.114432,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.80974319597304e-05,
+      "loss": 1.2016,
+      "step": 447
+    },
+    {
+      "epoch": 0.114688,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.809316611210648e-05,
+      "loss": 1.2061,
+      "step": 448
+    },
+    {
+      "epoch": 0.114944,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.8088900264482554e-05,
+      "loss": 1.2003,
+      "step": 449
+    },
+    {
+      "epoch": 0.1152,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.808463441685863e-05,
+      "loss": 1.2788,
+      "step": 450
+    },
+    {
+      "epoch": 0.115456,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.808036856923471e-05,
+      "loss": 1.2559,
+      "step": 451
+    },
+    {
+      "epoch": 0.115712,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.807610272161079e-05,
+      "loss": 1.2246,
+      "step": 452
+    },
+    {
+      "epoch": 0.115968,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.8071836873986865e-05,
+      "loss": 1.2219,
+      "step": 453
+    },
+    {
+      "epoch": 0.116224,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.806757102636294e-05,
+      "loss": 1.1989,
+      "step": 454
+    },
+    {
+      "epoch": 0.11648,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.8063305178739014e-05,
+      "loss": 1.2343,
+      "step": 455
+    },
+    {
+      "epoch": 0.116736,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.8059039331115095e-05,
+      "loss": 1.1503,
+      "step": 456
+    },
+    {
+      "epoch": 0.116992,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.805477348349117e-05,
+      "loss": 1.2004,
+      "step": 457
+    },
+    {
+      "epoch": 0.117248,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.805050763586725e-05,
+      "loss": 1.1581,
+      "step": 458
+    },
+    {
+      "epoch": 0.117504,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.8046241788243324e-05,
+      "loss": 1.2024,
+      "step": 459
+    },
+    {
+      "epoch": 0.11776,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.8041975940619405e-05,
+      "loss": 1.2146,
+      "step": 460
+    },
+    {
+      "epoch": 0.118016,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.803771009299548e-05,
+      "loss": 1.203,
+      "step": 461
+    },
+    {
+      "epoch": 0.118272,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.8033444245371554e-05,
+      "loss": 1.1742,
+      "step": 462
+    },
+    {
+      "epoch": 0.118528,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.8029178397747635e-05,
+      "loss": 1.2186,
+      "step": 463
+    },
+    {
+      "epoch": 0.118784,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.8024912550123716e-05,
+      "loss": 1.2669,
+      "step": 464
+    },
+    {
+      "epoch": 0.11904,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.802064670249979e-05,
+      "loss": 1.1652,
+      "step": 465
+    },
+    {
+      "epoch": 0.119296,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.8016380854875864e-05,
+      "loss": 1.2108,
+      "step": 466
+    },
+    {
+      "epoch": 0.119552,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.8012115007251945e-05,
+      "loss": 1.1808,
+      "step": 467
+    },
+    {
+      "epoch": 0.119808,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.800784915962802e-05,
+      "loss": 1.1963,
+      "step": 468
+    },
+    {
+      "epoch": 0.120064,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.8003583312004094e-05,
+      "loss": 1.2297,
+      "step": 469
+    },
+    {
+      "epoch": 0.12032,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.7999317464380175e-05,
+      "loss": 1.2167,
+      "step": 470
+    },
+    {
+      "epoch": 0.120576,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.7995051616756256e-05,
+      "loss": 1.1687,
+      "step": 471
+    },
+    {
+      "epoch": 0.120832,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.799078576913233e-05,
+      "loss": 1.1696,
+      "step": 472
+    },
+    {
+      "epoch": 0.121088,
+      "grad_norm": 3.0,
+      "learning_rate": 4.7986519921508405e-05,
+      "loss": 1.2201,
+      "step": 473
+    },
+    {
+      "epoch": 0.121344,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.798225407388448e-05,
+      "loss": 1.1975,
+      "step": 474
+    },
+    {
+      "epoch": 0.1216,
+      "grad_norm": 2.75,
+      "learning_rate": 4.797798822626056e-05,
+      "loss": 1.1767,
+      "step": 475
+    },
+    {
+      "epoch": 0.121856,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.7973722378636634e-05,
+      "loss": 1.2076,
+      "step": 476
+    },
+    {
+      "epoch": 0.122112,
+      "grad_norm": 2.875,
+      "learning_rate": 4.7969456531012715e-05,
+      "loss": 1.212,
+      "step": 477
+    },
+    {
+      "epoch": 0.122368,
+      "grad_norm": 2.5,
+      "learning_rate": 4.7965190683388796e-05,
+      "loss": 1.1181,
+      "step": 478
+    },
+    {
+      "epoch": 0.122624,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.796092483576487e-05,
+      "loss": 1.2206,
+      "step": 479
+    },
+    {
+      "epoch": 0.12288,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.7956658988140945e-05,
+      "loss": 1.1415,
+      "step": 480
+    },
+    {
+      "epoch": 0.123136,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.795239314051702e-05,
+      "loss": 1.1652,
+      "step": 481
+    },
+    {
+      "epoch": 0.123392,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.79481272928931e-05,
+      "loss": 1.1567,
+      "step": 482
+    },
+    {
+      "epoch": 0.123648,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.794386144526918e-05,
+      "loss": 1.2082,
+      "step": 483
+    },
+    {
+      "epoch": 0.123904,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.7939595597645255e-05,
+      "loss": 1.1374,
+      "step": 484
+    },
+    {
+      "epoch": 0.12416,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.793532975002133e-05,
+      "loss": 1.1723,
+      "step": 485
+    },
+    {
+      "epoch": 0.124416,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.793106390239741e-05,
+      "loss": 1.1584,
+      "step": 486
+    },
+    {
+      "epoch": 0.124672,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.7926798054773485e-05,
+      "loss": 1.1791,
+      "step": 487
+    },
+    {
+      "epoch": 0.124928,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.792253220714956e-05,
+      "loss": 1.1432,
+      "step": 488
+    },
+    {
+      "epoch": 0.125184,
+      "grad_norm": 2.625,
+      "learning_rate": 4.791826635952564e-05,
+      "loss": 1.17,
+      "step": 489
+    },
+    {
+      "epoch": 0.12544,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.791400051190172e-05,
+      "loss": 1.1851,
+      "step": 490
+    },
+    {
+      "epoch": 0.125696,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.7909734664277796e-05,
+      "loss": 1.1523,
+      "step": 491
+    },
+    {
+      "epoch": 0.125952,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.790546881665387e-05,
+      "loss": 1.1821,
+      "step": 492
+    },
+    {
+      "epoch": 0.126208,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.790120296902995e-05,
+      "loss": 1.2014,
+      "step": 493
+    },
+    {
+      "epoch": 0.126464,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.7896937121406025e-05,
+      "loss": 1.2193,
+      "step": 494
+    },
+    {
+      "epoch": 0.12672,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.78926712737821e-05,
+      "loss": 1.1866,
+      "step": 495
+    },
+    {
+      "epoch": 0.126976,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.788840542615818e-05,
+      "loss": 1.1713,
+      "step": 496
+    },
+    {
+      "epoch": 0.127232,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.788413957853426e-05,
+      "loss": 1.1778,
+      "step": 497
+    },
+    {
+      "epoch": 0.127488,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.7879873730910336e-05,
+      "loss": 1.1478,
+      "step": 498
+    },
+    {
+      "epoch": 0.127744,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.787560788328641e-05,
+      "loss": 1.1695,
+      "step": 499
+    },
+    {
+      "epoch": 0.128,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.7871342035662484e-05,
+      "loss": 1.1644,
+      "step": 500
+    },
+    {
+      "epoch": 0.128256,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.7867076188038565e-05,
+      "loss": 1.1625,
+      "step": 501
+    },
+    {
+      "epoch": 0.128512,
+      "grad_norm": 2.875,
+      "learning_rate": 4.7862810340414646e-05,
+      "loss": 1.1323,
+      "step": 502
+    },
+    {
+      "epoch": 0.128768,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.785854449279072e-05,
+      "loss": 1.2053,
+      "step": 503
+    },
+    {
+      "epoch": 0.129024,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.7854278645166795e-05,
+      "loss": 1.1687,
+      "step": 504
+    },
+    {
+      "epoch": 0.12928,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.7850012797542876e-05,
+      "loss": 1.1676,
+      "step": 505
+    },
+    {
+      "epoch": 0.129536,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.784574694991895e-05,
+      "loss": 1.1573,
+      "step": 506
+    },
+    {
+      "epoch": 0.129792,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.7841481102295024e-05,
+      "loss": 1.1496,
+      "step": 507
+    },
+    {
+      "epoch": 0.130048,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.7837215254671106e-05,
+      "loss": 1.1821,
+      "step": 508
+    },
+    {
+      "epoch": 0.130304,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.7832949407047187e-05,
+      "loss": 1.1726,
+      "step": 509
+    },
+    {
+      "epoch": 0.13056,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.782868355942326e-05,
+      "loss": 1.1828,
+      "step": 510
+    },
+    {
+      "epoch": 0.130816,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.7824417711799335e-05,
+      "loss": 1.1455,
+      "step": 511
+    },
+    {
+      "epoch": 0.131072,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.7820151864175416e-05,
+      "loss": 1.1556,
+      "step": 512
+    },
+    {
+      "epoch": 0.131328,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.781588601655149e-05,
+      "loss": 1.1594,
+      "step": 513
+    },
+    {
+      "epoch": 0.131584,
+      "grad_norm": 3.234375,
+      "learning_rate": 4.7811620168927565e-05,
+      "loss": 1.1747,
+      "step": 514
+    },
+    {
+      "epoch": 0.13184,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.7807354321303646e-05,
+      "loss": 1.1451,
+      "step": 515
+    },
+    {
+      "epoch": 0.132096,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.780308847367973e-05,
+      "loss": 1.1487,
+      "step": 516
+    },
+    {
+      "epoch": 0.132352,
+      "grad_norm": 3.21875,
+      "learning_rate": 4.77988226260558e-05,
+      "loss": 1.2026,
+      "step": 517
+    },
+    {
+      "epoch": 0.132608,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.7794556778431875e-05,
+      "loss": 1.1329,
+      "step": 518
+    },
+    {
+      "epoch": 0.132864,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.779029093080795e-05,
+      "loss": 1.1144,
+      "step": 519
+    },
+    {
+      "epoch": 0.13312,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.778602508318403e-05,
+      "loss": 1.1052,
+      "step": 520
+    },
+    {
+      "epoch": 0.133376,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.778175923556011e-05,
+      "loss": 1.1138,
+      "step": 521
+    },
+    {
+      "epoch": 0.133632,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.7777493387936186e-05,
+      "loss": 1.156,
+      "step": 522
+    },
+    {
+      "epoch": 0.133888,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.777322754031226e-05,
+      "loss": 1.1496,
+      "step": 523
+    },
+    {
+      "epoch": 0.134144,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.776896169268834e-05,
+      "loss": 1.1275,
+      "step": 524
+    },
+    {
+      "epoch": 0.1344,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.7764695845064415e-05,
+      "loss": 1.1242,
+      "step": 525
+    },
+    {
+      "epoch": 0.134656,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.776042999744049e-05,
+      "loss": 1.1725,
+      "step": 526
+    },
+    {
+      "epoch": 0.134912,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.775616414981657e-05,
+      "loss": 1.1613,
+      "step": 527
+    },
+    {
+      "epoch": 0.135168,
+      "grad_norm": 2.875,
+      "learning_rate": 4.775189830219265e-05,
+      "loss": 1.1712,
+      "step": 528
+    },
+    {
+      "epoch": 0.135424,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.7747632454568726e-05,
+      "loss": 1.1165,
+      "step": 529
+    },
+    {
+      "epoch": 0.13568,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.77433666069448e-05,
+      "loss": 1.1593,
+      "step": 530
+    },
+    {
+      "epoch": 0.135936,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.773910075932088e-05,
+      "loss": 1.1183,
+      "step": 531
+    },
+    {
+      "epoch": 0.136192,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.7734834911696956e-05,
+      "loss": 1.094,
+      "step": 532
+    },
+    {
+      "epoch": 0.136448,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.773056906407303e-05,
+      "loss": 1.1586,
+      "step": 533
+    },
+    {
+      "epoch": 0.136704,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.772630321644911e-05,
+      "loss": 1.1353,
+      "step": 534
+    },
+    {
+      "epoch": 0.13696,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.772203736882519e-05,
+      "loss": 1.1287,
+      "step": 535
+    },
+    {
+      "epoch": 0.137216,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.7717771521201266e-05,
+      "loss": 1.1246,
+      "step": 536
+    },
+    {
+      "epoch": 0.137472,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.771350567357734e-05,
+      "loss": 1.1722,
+      "step": 537
+    },
+    {
+      "epoch": 0.137728,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.7709239825953415e-05,
+      "loss": 1.1104,
+      "step": 538
+    },
+    {
+      "epoch": 0.137984,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.7704973978329496e-05,
+      "loss": 1.1529,
+      "step": 539
+    },
+    {
+      "epoch": 0.13824,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.770070813070558e-05,
+      "loss": 1.1282,
+      "step": 540
+    },
+    {
+      "epoch": 0.138496,
+      "grad_norm": 2.75,
+      "learning_rate": 4.769644228308165e-05,
+      "loss": 1.1246,
+      "step": 541
+    },
+    {
+      "epoch": 0.138752,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.769217643545773e-05,
+      "loss": 1.1092,
+      "step": 542
+    },
+    {
+      "epoch": 0.139008,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.7687910587833807e-05,
+      "loss": 1.0936,
+      "step": 543
+    },
+    {
+      "epoch": 0.139264,
+      "grad_norm": 3.5,
+      "learning_rate": 4.768364474020988e-05,
+      "loss": 1.0865,
+      "step": 544
+    },
+    {
+      "epoch": 0.13952,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.7679378892585955e-05,
+      "loss": 1.11,
+      "step": 545
+    },
+    {
+      "epoch": 0.139776,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.7675113044962036e-05,
+      "loss": 1.1452,
+      "step": 546
+    },
+    {
+      "epoch": 0.140032,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.767084719733812e-05,
+      "loss": 1.1198,
+      "step": 547
+    },
+    {
+      "epoch": 0.140288,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.766658134971419e-05,
+      "loss": 1.0723,
+      "step": 548
+    },
+    {
+      "epoch": 0.140544,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.7662315502090266e-05,
+      "loss": 1.0745,
+      "step": 549
+    },
+    {
+      "epoch": 0.1408,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.765804965446635e-05,
+      "loss": 1.1535,
+      "step": 550
+    },
+    {
+      "epoch": 0.141056,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.765378380684242e-05,
+      "loss": 1.114,
+      "step": 551
+    },
+    {
+      "epoch": 0.141312,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.7649517959218495e-05,
+      "loss": 1.1597,
+      "step": 552
+    },
+    {
+      "epoch": 0.141568,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.7645252111594576e-05,
+      "loss": 1.1561,
+      "step": 553
+    },
+    {
+      "epoch": 0.141824,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.764098626397066e-05,
+      "loss": 1.0907,
+      "step": 554
+    },
+    {
+      "epoch": 0.14208,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.763672041634673e-05,
+      "loss": 1.1042,
+      "step": 555
+    },
+    {
+      "epoch": 0.142336,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.7632454568722806e-05,
+      "loss": 1.1243,
+      "step": 556
+    },
+    {
+      "epoch": 0.142592,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.762818872109888e-05,
+      "loss": 1.1019,
+      "step": 557
+    },
+    {
+      "epoch": 0.142848,
+      "grad_norm": 3.25,
+      "learning_rate": 4.762392287347496e-05,
+      "loss": 1.0956,
+      "step": 558
+    },
+    {
+      "epoch": 0.143104,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.761965702585104e-05,
+      "loss": 1.0961,
+      "step": 559
+    },
+    {
+      "epoch": 0.14336,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.7615391178227116e-05,
+      "loss": 1.1031,
+      "step": 560
+    },
+    {
+      "epoch": 0.143616,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.76111253306032e-05,
+      "loss": 1.1259,
+      "step": 561
+    },
+    {
+      "epoch": 0.143872,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.760685948297927e-05,
+      "loss": 1.1205,
+      "step": 562
+    },
+    {
+      "epoch": 0.144128,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.7602593635355346e-05,
+      "loss": 1.0901,
+      "step": 563
+    },
+    {
+      "epoch": 0.144384,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.759832778773142e-05,
+      "loss": 1.1116,
+      "step": 564
+    },
+    {
+      "epoch": 0.14464,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.75940619401075e-05,
+      "loss": 1.0722,
+      "step": 565
+    },
+    {
+      "epoch": 0.144896,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.758979609248358e-05,
+      "loss": 1.0947,
+      "step": 566
+    },
+    {
+      "epoch": 0.145152,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.758553024485966e-05,
+      "loss": 1.1324,
+      "step": 567
+    },
+    {
+      "epoch": 0.145408,
+      "grad_norm": 2.625,
+      "learning_rate": 4.758126439723573e-05,
+      "loss": 1.0917,
+      "step": 568
+    },
+    {
+      "epoch": 0.145664,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.757699854961181e-05,
+      "loss": 1.0916,
+      "step": 569
+    },
+    {
+      "epoch": 0.14592,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.7572732701987886e-05,
+      "loss": 1.1307,
+      "step": 570
+    },
+    {
+      "epoch": 0.146176,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.756846685436396e-05,
+      "loss": 1.1125,
+      "step": 571
+    },
+    {
+      "epoch": 0.146432,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.756420100674004e-05,
+      "loss": 1.0905,
+      "step": 572
+    },
+    {
+      "epoch": 0.146688,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.755993515911612e-05,
+      "loss": 1.0652,
+      "step": 573
+    },
+    {
+      "epoch": 0.146944,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.75556693114922e-05,
+      "loss": 1.1072,
+      "step": 574
+    },
+    {
+      "epoch": 0.1472,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.755140346386827e-05,
+      "loss": 1.0555,
+      "step": 575
+    },
+    {
+      "epoch": 0.147456,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.754713761624435e-05,
+      "loss": 1.0561,
+      "step": 576
+    },
+    {
+      "epoch": 0.147712,
+      "grad_norm": 3.125,
+      "learning_rate": 4.7542871768620426e-05,
+      "loss": 1.0716,
+      "step": 577
+    },
+    {
+      "epoch": 0.147968,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.75386059209965e-05,
+      "loss": 1.0848,
+      "step": 578
+    },
+    {
+      "epoch": 0.148224,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.753434007337258e-05,
+      "loss": 1.0966,
+      "step": 579
+    },
+    {
+      "epoch": 0.14848,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.753007422574866e-05,
+      "loss": 1.1057,
+      "step": 580
+    },
+    {
+      "epoch": 0.148736,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.752580837812474e-05,
+      "loss": 1.0805,
+      "step": 581
+    },
+    {
+      "epoch": 0.148992,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.752154253050081e-05,
+      "loss": 1.0888,
+      "step": 582
+    },
+    {
+      "epoch": 0.149248,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.7517276682876886e-05,
+      "loss": 1.0826,
+      "step": 583
+    },
+    {
+      "epoch": 0.149504,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.751301083525297e-05,
+      "loss": 1.1176,
+      "step": 584
+    },
+    {
+      "epoch": 0.14976,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.750874498762905e-05,
+      "loss": 1.0937,
+      "step": 585
+    },
+    {
+      "epoch": 0.150016,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.750447914000512e-05,
+      "loss": 1.0707,
+      "step": 586
+    },
+    {
+      "epoch": 0.150272,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.7500213292381196e-05,
+      "loss": 1.0484,
+      "step": 587
+    },
+    {
+      "epoch": 0.150528,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.749594744475728e-05,
+      "loss": 1.0829,
+      "step": 588
+    },
+    {
+      "epoch": 0.150784,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.749168159713335e-05,
+      "loss": 1.0793,
+      "step": 589
+    },
+    {
+      "epoch": 0.15104,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.7487415749509426e-05,
+      "loss": 1.0851,
+      "step": 590
+    },
+    {
+      "epoch": 0.151296,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.748314990188551e-05,
+      "loss": 1.0708,
+      "step": 591
+    },
+    {
+      "epoch": 0.151552,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.747888405426159e-05,
+      "loss": 1.0972,
+      "step": 592
+    },
+    {
+      "epoch": 0.151808,
+      "grad_norm": 2.625,
+      "learning_rate": 4.747461820663766e-05,
+      "loss": 1.0678,
+      "step": 593
+    },
+    {
+      "epoch": 0.152064,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.7470352359013736e-05,
+      "loss": 1.0392,
+      "step": 594
+    },
+    {
+      "epoch": 0.15232,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.746608651138982e-05,
+      "loss": 1.0581,
+      "step": 595
+    },
+    {
+      "epoch": 0.152576,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.746182066376589e-05,
+      "loss": 1.0928,
+      "step": 596
+    },
+    {
+      "epoch": 0.152832,
+      "grad_norm": 2.75,
+      "learning_rate": 4.7457554816141966e-05,
+      "loss": 1.022,
+      "step": 597
+    },
+    {
+      "epoch": 0.153088,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.745328896851805e-05,
+      "loss": 1.0889,
+      "step": 598
+    },
+    {
+      "epoch": 0.153344,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.744902312089413e-05,
+      "loss": 1.0376,
+      "step": 599
+    },
+    {
+      "epoch": 0.1536,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.74447572732702e-05,
+      "loss": 1.0898,
+      "step": 600
+    },
+    {
+      "epoch": 0.153856,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.7440491425646277e-05,
+      "loss": 1.0821,
+      "step": 601
+    },
+    {
+      "epoch": 0.154112,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.743622557802235e-05,
+      "loss": 1.0557,
+      "step": 602
+    },
+    {
+      "epoch": 0.154368,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.743195973039843e-05,
+      "loss": 1.0831,
+      "step": 603
+    },
+    {
+      "epoch": 0.154624,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.742769388277451e-05,
+      "loss": 1.0406,
+      "step": 604
+    },
+    {
+      "epoch": 0.15488,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.742342803515059e-05,
+      "loss": 1.0452,
+      "step": 605
+    },
+    {
+      "epoch": 0.155136,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.741916218752666e-05,
+      "loss": 1.0851,
+      "step": 606
+    },
+    {
+      "epoch": 0.155392,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.741489633990274e-05,
+      "loss": 1.0612,
+      "step": 607
+    },
+    {
+      "epoch": 0.155648,
+      "grad_norm": 2.875,
+      "learning_rate": 4.741063049227882e-05,
+      "loss": 1.0837,
+      "step": 608
+    },
+    {
+      "epoch": 0.155904,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.740636464465489e-05,
+      "loss": 1.0559,
+      "step": 609
+    },
+    {
+      "epoch": 0.15616,
+      "grad_norm": 2.75,
+      "learning_rate": 4.740209879703097e-05,
+      "loss": 1.0843,
+      "step": 610
+    },
+    {
+      "epoch": 0.156416,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.739783294940705e-05,
+      "loss": 1.104,
+      "step": 611
+    },
+    {
+      "epoch": 0.156672,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.739356710178313e-05,
+      "loss": 1.0977,
+      "step": 612
+    },
+    {
+      "epoch": 0.156928,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.73893012541592e-05,
+      "loss": 1.099,
+      "step": 613
+    },
+    {
+      "epoch": 0.157184,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.738503540653528e-05,
+      "loss": 1.0176,
+      "step": 614
+    },
+    {
+      "epoch": 0.15744,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.738076955891136e-05,
+      "loss": 1.0127,
+      "step": 615
+    },
+    {
+      "epoch": 0.157696,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.737650371128743e-05,
+      "loss": 1.0515,
+      "step": 616
+    },
+    {
+      "epoch": 0.157952,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.737223786366351e-05,
+      "loss": 1.1335,
+      "step": 617
+    },
+    {
+      "epoch": 0.158208,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.736797201603959e-05,
+      "loss": 1.0904,
+      "step": 618
+    },
+    {
+      "epoch": 0.158464,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.736370616841567e-05,
+      "loss": 1.061,
+      "step": 619
+    },
+    {
+      "epoch": 0.15872,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.735944032079174e-05,
+      "loss": 1.091,
+      "step": 620
+    },
+    {
+      "epoch": 0.158976,
+      "grad_norm": 3.0,
+      "learning_rate": 4.7355174473167816e-05,
+      "loss": 1.0191,
+      "step": 621
+    },
+    {
+      "epoch": 0.159232,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.73509086255439e-05,
+      "loss": 1.0931,
+      "step": 622
+    },
+    {
+      "epoch": 0.159488,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.734664277791998e-05,
+      "loss": 1.0844,
+      "step": 623
+    },
+    {
+      "epoch": 0.159744,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.734237693029605e-05,
+      "loss": 1.073,
+      "step": 624
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.7338111082672134e-05,
+      "loss": 1.0952,
+      "step": 625
+    },
+    {
+      "epoch": 0.160256,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.733384523504821e-05,
+      "loss": 1.0517,
+      "step": 626
+    },
+    {
+      "epoch": 0.160512,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.732957938742428e-05,
+      "loss": 1.0576,
+      "step": 627
+    },
+    {
+      "epoch": 0.160768,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.7325313539800356e-05,
+      "loss": 1.0528,
+      "step": 628
+    },
+    {
+      "epoch": 0.161024,
+      "grad_norm": 3.0,
+      "learning_rate": 4.732104769217644e-05,
+      "loss": 1.0443,
+      "step": 629
+    },
+    {
+      "epoch": 0.16128,
+      "grad_norm": 2.625,
+      "learning_rate": 4.731678184455252e-05,
+      "loss": 1.0563,
+      "step": 630
+    },
+    {
+      "epoch": 0.161536,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.731251599692859e-05,
+      "loss": 1.033,
+      "step": 631
+    },
+    {
+      "epoch": 0.161792,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.730825014930467e-05,
+      "loss": 1.032,
+      "step": 632
+    },
+    {
+      "epoch": 0.162048,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.730398430168075e-05,
+      "loss": 1.0593,
+      "step": 633
+    },
+    {
+      "epoch": 0.162304,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.729971845405682e-05,
+      "loss": 1.0544,
+      "step": 634
+    },
+    {
+      "epoch": 0.16256,
+      "grad_norm": 2.75,
+      "learning_rate": 4.7295452606432896e-05,
+      "loss": 1.0359,
+      "step": 635
+    },
+    {
+      "epoch": 0.162816,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.729118675880898e-05,
+      "loss": 1.0893,
+      "step": 636
+    },
+    {
+      "epoch": 0.163072,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.728692091118506e-05,
+      "loss": 1.0746,
+      "step": 637
+    },
+    {
+      "epoch": 0.163328,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.728265506356113e-05,
+      "loss": 1.0457,
+      "step": 638
+    },
+    {
+      "epoch": 0.163584,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.727838921593721e-05,
+      "loss": 1.092,
+      "step": 639
+    },
+    {
+      "epoch": 0.16384,
+      "grad_norm": 2.75,
+      "learning_rate": 4.727412336831328e-05,
+      "loss": 1.0607,
+      "step": 640
+    },
+    {
+      "epoch": 0.164096,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.726985752068936e-05,
+      "loss": 1.0339,
+      "step": 641
+    },
+    {
+      "epoch": 0.164352,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.7265591673065443e-05,
+      "loss": 1.0298,
+      "step": 642
+    },
+    {
+      "epoch": 0.164608,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.726132582544152e-05,
+      "loss": 1.0336,
+      "step": 643
+    },
+    {
+      "epoch": 0.164864,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.72570599778176e-05,
+      "loss": 1.0617,
+      "step": 644
+    },
+    {
+      "epoch": 0.16512,
+      "grad_norm": 3.0,
+      "learning_rate": 4.725279413019367e-05,
+      "loss": 1.0389,
+      "step": 645
+    },
+    {
+      "epoch": 0.165376,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.724852828256975e-05,
+      "loss": 1.0635,
+      "step": 646
+    },
+    {
+      "epoch": 0.165632,
+      "grad_norm": 2.75,
+      "learning_rate": 4.724426243494582e-05,
+      "loss": 1.0387,
+      "step": 647
+    },
+    {
+      "epoch": 0.165888,
+      "grad_norm": 3.125,
+      "learning_rate": 4.72399965873219e-05,
+      "loss": 1.0478,
+      "step": 648
+    },
+    {
+      "epoch": 0.166144,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.7235730739697984e-05,
+      "loss": 1.0675,
+      "step": 649
+    },
+    {
+      "epoch": 0.1664,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.723146489207406e-05,
+      "loss": 1.0358,
+      "step": 650
+    },
+    {
+      "epoch": 0.166656,
+      "grad_norm": 3.234375,
+      "learning_rate": 4.722719904445013e-05,
+      "loss": 1.0367,
+      "step": 651
+    },
+    {
+      "epoch": 0.166912,
+      "grad_norm": 3.5625,
+      "learning_rate": 4.722293319682621e-05,
+      "loss": 1.0396,
+      "step": 652
+    },
+    {
+      "epoch": 0.167168,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.721866734920229e-05,
+      "loss": 0.9982,
+      "step": 653
+    },
+    {
+      "epoch": 0.167424,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.721440150157836e-05,
+      "loss": 1.0896,
+      "step": 654
+    },
+    {
+      "epoch": 0.16768,
+      "grad_norm": 3.671875,
+      "learning_rate": 4.721013565395444e-05,
+      "loss": 1.0533,
+      "step": 655
+    },
+    {
+      "epoch": 0.167936,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.7205869806330524e-05,
+      "loss": 1.0219,
+      "step": 656
+    },
+    {
+      "epoch": 0.168192,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.72016039587066e-05,
+      "loss": 1.0104,
+      "step": 657
+    },
+    {
+      "epoch": 0.168448,
+      "grad_norm": 3.296875,
+      "learning_rate": 4.719733811108267e-05,
+      "loss": 1.0534,
+      "step": 658
+    },
+    {
+      "epoch": 0.168704,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.7193072263458753e-05,
+      "loss": 1.0153,
+      "step": 659
+    },
+    {
+      "epoch": 0.16896,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.718880641583483e-05,
+      "loss": 1.0059,
+      "step": 660
+    },
+    {
+      "epoch": 0.169216,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.718454056821091e-05,
+      "loss": 1.0332,
+      "step": 661
+    },
+    {
+      "epoch": 0.169472,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.718027472058698e-05,
+      "loss": 1.0622,
+      "step": 662
+    },
+    {
+      "epoch": 0.169728,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.7176008872963064e-05,
+      "loss": 1.0039,
+      "step": 663
+    },
+    {
+      "epoch": 0.169984,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.717174302533914e-05,
+      "loss": 1.036,
+      "step": 664
+    },
+    {
+      "epoch": 0.17024,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.716747717771521e-05,
+      "loss": 1.038,
+      "step": 665
+    },
+    {
+      "epoch": 0.170496,
+      "grad_norm": 3.234375,
+      "learning_rate": 4.716321133009129e-05,
+      "loss": 1.0608,
+      "step": 666
+    },
+    {
+      "epoch": 0.170752,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.715894548246737e-05,
+      "loss": 1.0772,
+      "step": 667
+    },
+    {
+      "epoch": 0.171008,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.715467963484345e-05,
+      "loss": 1.0124,
+      "step": 668
+    },
+    {
+      "epoch": 0.171264,
+      "grad_norm": 2.875,
+      "learning_rate": 4.715041378721952e-05,
+      "loss": 1.0043,
+      "step": 669
+    },
+    {
+      "epoch": 0.17152,
+      "grad_norm": 3.453125,
+      "learning_rate": 4.71461479395956e-05,
+      "loss": 1.0233,
+      "step": 670
+    },
+    {
+      "epoch": 0.171776,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.714188209197168e-05,
+      "loss": 1.0378,
+      "step": 671
+    },
+    {
+      "epoch": 0.172032,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.713761624434775e-05,
+      "loss": 1.0391,
+      "step": 672
+    },
+    {
+      "epoch": 0.172288,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.713335039672383e-05,
+      "loss": 1.0243,
+      "step": 673
+    },
+    {
+      "epoch": 0.172544,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.712908454909991e-05,
+      "loss": 1.0418,
+      "step": 674
+    },
+    {
+      "epoch": 0.1728,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.712481870147599e-05,
+      "loss": 0.9967,
+      "step": 675
+    },
+    {
+      "epoch": 0.173056,
+      "grad_norm": 3.125,
+      "learning_rate": 4.7120552853852063e-05,
+      "loss": 1.029,
+      "step": 676
+    },
+    {
+      "epoch": 0.173312,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.711628700622814e-05,
+      "loss": 1.0238,
+      "step": 677
+    },
+    {
+      "epoch": 0.173568,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.711202115860422e-05,
+      "loss": 0.9907,
+      "step": 678
+    },
+    {
+      "epoch": 0.173824,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.710775531098029e-05,
+      "loss": 0.9974,
+      "step": 679
+    },
+    {
+      "epoch": 0.17408,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.710348946335637e-05,
+      "loss": 1.0259,
+      "step": 680
+    },
+    {
+      "epoch": 0.174336,
+      "grad_norm": 3.25,
+      "learning_rate": 4.709922361573245e-05,
+      "loss": 1.0392,
+      "step": 681
+    },
+    {
+      "epoch": 0.174592,
+      "grad_norm": 2.875,
+      "learning_rate": 4.709495776810853e-05,
+      "loss": 1.0154,
+      "step": 682
+    },
+    {
+      "epoch": 0.174848,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.7090691920484604e-05,
+      "loss": 1.0285,
+      "step": 683
+    },
+    {
+      "epoch": 0.175104,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.708642607286068e-05,
+      "loss": 1.0462,
+      "step": 684
+    },
+    {
+      "epoch": 0.17536,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.708216022523675e-05,
+      "loss": 1.0417,
+      "step": 685
+    },
+    {
+      "epoch": 0.175616,
+      "grad_norm": 2.5,
+      "learning_rate": 4.707789437761283e-05,
+      "loss": 1.0221,
+      "step": 686
+    },
+    {
+      "epoch": 0.175872,
+      "grad_norm": 3.0,
+      "learning_rate": 4.7073628529988914e-05,
+      "loss": 0.9965,
+      "step": 687
+    },
+    {
+      "epoch": 0.176128,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.706936268236499e-05,
+      "loss": 1.0167,
+      "step": 688
+    },
+    {
+      "epoch": 0.176384,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.706509683474106e-05,
+      "loss": 1.0239,
+      "step": 689
+    },
+    {
+      "epoch": 0.17664,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.7060830987117144e-05,
+      "loss": 1.0033,
+      "step": 690
+    },
+    {
+      "epoch": 0.176896,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.705656513949322e-05,
+      "loss": 1.0442,
+      "step": 691
+    },
+    {
+      "epoch": 0.177152,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.705229929186929e-05,
+      "loss": 1.0192,
+      "step": 692
+    },
+    {
+      "epoch": 0.177408,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.704803344424537e-05,
+      "loss": 1.0329,
+      "step": 693
+    },
+    {
+      "epoch": 0.177664,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.7043767596621454e-05,
+      "loss": 1.0115,
+      "step": 694
+    },
+    {
+      "epoch": 0.17792,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.703950174899753e-05,
+      "loss": 0.9813,
+      "step": 695
+    },
+    {
+      "epoch": 0.178176,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.70352359013736e-05,
+      "loss": 0.9962,
+      "step": 696
+    },
+    {
+      "epoch": 0.178432,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.7030970053749684e-05,
+      "loss": 1.0117,
+      "step": 697
+    },
+    {
+      "epoch": 0.178688,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.702670420612576e-05,
+      "loss": 1.0092,
+      "step": 698
+    },
+    {
+      "epoch": 0.178944,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.702243835850183e-05,
+      "loss": 1.028,
+      "step": 699
+    },
+    {
+      "epoch": 0.1792,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.7018172510877914e-05,
+      "loss": 1.0395,
+      "step": 700
+    },
+    {
+      "epoch": 0.179456,
+      "grad_norm": 2.875,
+      "learning_rate": 4.7013906663253995e-05,
+      "loss": 1.016,
+      "step": 701
+    },
+    {
+      "epoch": 0.179712,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.700964081563007e-05,
+      "loss": 1.0156,
+      "step": 702
+    },
+    {
+      "epoch": 0.179968,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.700537496800614e-05,
+      "loss": 1.0528,
+      "step": 703
+    },
+    {
+      "epoch": 0.180224,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.700110912038222e-05,
+      "loss": 1.0129,
+      "step": 704
+    },
+    {
+      "epoch": 0.18048,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.69968432727583e-05,
+      "loss": 1.0087,
+      "step": 705
+    },
+    {
+      "epoch": 0.180736,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.699257742513438e-05,
+      "loss": 1.0265,
+      "step": 706
+    },
+    {
+      "epoch": 0.180992,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.6988311577510454e-05,
+      "loss": 0.9866,
+      "step": 707
+    },
+    {
+      "epoch": 0.181248,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.6984045729886535e-05,
+      "loss": 0.9846,
+      "step": 708
+    },
+    {
+      "epoch": 0.181504,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.697977988226261e-05,
+      "loss": 0.9859,
+      "step": 709
+    },
+    {
+      "epoch": 0.18176,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.697551403463868e-05,
+      "loss": 0.9765,
+      "step": 710
+    },
+    {
+      "epoch": 0.182016,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.697124818701476e-05,
+      "loss": 1.0432,
+      "step": 711
+    },
+    {
+      "epoch": 0.182272,
+      "grad_norm": 3.5625,
+      "learning_rate": 4.696698233939084e-05,
+      "loss": 0.9856,
+      "step": 712
+    },
+    {
+      "epoch": 0.182528,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.696271649176692e-05,
+      "loss": 1.0017,
+      "step": 713
+    },
+    {
+      "epoch": 0.182784,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.6958450644142994e-05,
+      "loss": 1.0054,
+      "step": 714
+    },
+    {
+      "epoch": 0.18304,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.695418479651907e-05,
+      "loss": 0.9733,
+      "step": 715
+    },
+    {
+      "epoch": 0.183296,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.694991894889515e-05,
+      "loss": 0.9473,
+      "step": 716
+    },
+    {
+      "epoch": 0.183552,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.6945653101271224e-05,
+      "loss": 1.0048,
+      "step": 717
+    },
+    {
+      "epoch": 0.183808,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.69413872536473e-05,
+      "loss": 1.0121,
+      "step": 718
+    },
+    {
+      "epoch": 0.184064,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.693712140602338e-05,
+      "loss": 0.9624,
+      "step": 719
+    },
+    {
+      "epoch": 0.18432,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.693285555839946e-05,
+      "loss": 0.9657,
+      "step": 720
+    },
+    {
+      "epoch": 0.184576,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.6928589710775534e-05,
+      "loss": 0.9657,
+      "step": 721
+    },
+    {
+      "epoch": 0.184832,
+      "grad_norm": 3.0,
+      "learning_rate": 4.692432386315161e-05,
+      "loss": 1.0121,
+      "step": 722
+    },
+    {
+      "epoch": 0.185088,
+      "grad_norm": 2.625,
+      "learning_rate": 4.692005801552769e-05,
+      "loss": 0.9519,
+      "step": 723
+    },
+    {
+      "epoch": 0.185344,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.6915792167903764e-05,
+      "loss": 1.0066,
+      "step": 724
+    },
+    {
+      "epoch": 0.1856,
+      "grad_norm": 3.34375,
+      "learning_rate": 4.6911526320279845e-05,
+      "loss": 1.0124,
+      "step": 725
+    },
+    {
+      "epoch": 0.185856,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.690726047265592e-05,
+      "loss": 0.9502,
+      "step": 726
+    },
+    {
+      "epoch": 0.186112,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.6902994625032e-05,
+      "loss": 1.0253,
+      "step": 727
+    },
+    {
+      "epoch": 0.186368,
+      "grad_norm": 3.3125,
+      "learning_rate": 4.6898728777408074e-05,
+      "loss": 0.9477,
+      "step": 728
+    },
+    {
+      "epoch": 0.186624,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.689446292978415e-05,
+      "loss": 0.9844,
+      "step": 729
+    },
+    {
+      "epoch": 0.18688,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.689019708216022e-05,
+      "loss": 1.0033,
+      "step": 730
+    },
+    {
+      "epoch": 0.187136,
+      "grad_norm": 3.25,
+      "learning_rate": 4.6885931234536304e-05,
+      "loss": 0.9581,
+      "step": 731
+    },
+    {
+      "epoch": 0.187392,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.6881665386912385e-05,
+      "loss": 1.0039,
+      "step": 732
+    },
+    {
+      "epoch": 0.187648,
+      "grad_norm": 2.5,
+      "learning_rate": 4.687739953928846e-05,
+      "loss": 0.9984,
+      "step": 733
+    },
+    {
+      "epoch": 0.187904,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.6873133691664533e-05,
+      "loss": 0.9676,
+      "step": 734
+    },
+    {
+      "epoch": 0.18816,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.6868867844040615e-05,
+      "loss": 1.0037,
+      "step": 735
+    },
+    {
+      "epoch": 0.188416,
+      "grad_norm": 3.296875,
+      "learning_rate": 4.686460199641669e-05,
+      "loss": 1.01,
+      "step": 736
+    },
+    {
+      "epoch": 0.188672,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.686033614879276e-05,
+      "loss": 0.9934,
+      "step": 737
+    },
+    {
+      "epoch": 0.188928,
+      "grad_norm": 2.75,
+      "learning_rate": 4.6856070301168844e-05,
+      "loss": 0.9992,
+      "step": 738
+    },
+    {
+      "epoch": 0.189184,
+      "grad_norm": 3.375,
+      "learning_rate": 4.6851804453544925e-05,
+      "loss": 0.9767,
+      "step": 739
+    },
+    {
+      "epoch": 0.18944,
+      "grad_norm": 2.625,
+      "learning_rate": 4.6847538605921e-05,
+      "loss": 1.003,
+      "step": 740
+    },
+    {
+      "epoch": 0.189696,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.6843272758297074e-05,
+      "loss": 0.9985,
+      "step": 741
+    },
+    {
+      "epoch": 0.189952,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.6839006910673155e-05,
+      "loss": 0.982,
+      "step": 742
+    },
+    {
+      "epoch": 0.190208,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.683474106304923e-05,
+      "loss": 1.0125,
+      "step": 743
+    },
+    {
+      "epoch": 0.190464,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.683047521542531e-05,
+      "loss": 1.0059,
+      "step": 744
+    },
+    {
+      "epoch": 0.19072,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.6826209367801384e-05,
+      "loss": 0.9404,
+      "step": 745
+    },
+    {
+      "epoch": 0.190976,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.6821943520177465e-05,
+      "loss": 0.9485,
+      "step": 746
+    },
+    {
+      "epoch": 0.191232,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.681767767255354e-05,
+      "loss": 1.0116,
+      "step": 747
+    },
+    {
+      "epoch": 0.191488,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.6813411824929614e-05,
+      "loss": 0.9691,
+      "step": 748
+    },
+    {
+      "epoch": 0.191744,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.680914597730569e-05,
+      "loss": 0.9841,
+      "step": 749
+    },
+    {
+      "epoch": 0.192,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.680488012968177e-05,
+      "loss": 0.976,
+      "step": 750
+    },
+    {
+      "epoch": 0.192256,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.680061428205785e-05,
+      "loss": 0.9733,
+      "step": 751
+    },
+    {
+      "epoch": 0.192512,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.6796348434433924e-05,
+      "loss": 0.9303,
+      "step": 752
+    },
+    {
+      "epoch": 0.192768,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.679208258681e-05,
+      "loss": 0.8962,
+      "step": 753
+    },
+    {
+      "epoch": 0.193024,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.678781673918608e-05,
+      "loss": 0.9606,
+      "step": 754
+    },
+    {
+      "epoch": 0.19328,
+      "grad_norm": 2.625,
+      "learning_rate": 4.6783550891562154e-05,
+      "loss": 0.9567,
+      "step": 755
+    },
+    {
+      "epoch": 0.193536,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.677928504393823e-05,
+      "loss": 1.009,
+      "step": 756
+    },
+    {
+      "epoch": 0.193792,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.677501919631431e-05,
+      "loss": 0.9203,
+      "step": 757
+    },
+    {
+      "epoch": 0.194048,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.677075334869039e-05,
+      "loss": 0.9471,
+      "step": 758
+    },
+    {
+      "epoch": 0.194304,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.6766487501066465e-05,
+      "loss": 0.9793,
+      "step": 759
+    },
+    {
+      "epoch": 0.19456,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.676222165344254e-05,
+      "loss": 0.9236,
+      "step": 760
+    },
+    {
+      "epoch": 0.194816,
+      "grad_norm": 3.21875,
+      "learning_rate": 4.675795580581862e-05,
+      "loss": 0.9696,
+      "step": 761
+    },
+    {
+      "epoch": 0.195072,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.6753689958194694e-05,
+      "loss": 0.9607,
+      "step": 762
+    },
+    {
+      "epoch": 0.195328,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.6749424110570775e-05,
+      "loss": 0.9592,
+      "step": 763
+    },
+    {
+      "epoch": 0.195584,
+      "grad_norm": 2.875,
+      "learning_rate": 4.674515826294685e-05,
+      "loss": 0.9702,
+      "step": 764
+    },
+    {
+      "epoch": 0.19584,
+      "grad_norm": 2.625,
+      "learning_rate": 4.674089241532293e-05,
+      "loss": 0.9534,
+      "step": 765
+    },
+    {
+      "epoch": 0.196096,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.6736626567699005e-05,
+      "loss": 0.9675,
+      "step": 766
+    },
+    {
+      "epoch": 0.196352,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.673236072007508e-05,
+      "loss": 0.9604,
+      "step": 767
+    },
+    {
+      "epoch": 0.196608,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.6728094872451153e-05,
+      "loss": 0.9541,
+      "step": 768
+    },
+    {
+      "epoch": 0.196864,
+      "grad_norm": 3.0,
+      "learning_rate": 4.6723829024827234e-05,
+      "loss": 0.9788,
+      "step": 769
+    },
+    {
+      "epoch": 0.19712,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.6719563177203315e-05,
+      "loss": 0.9604,
+      "step": 770
+    },
+    {
+      "epoch": 0.197376,
+      "grad_norm": 2.625,
+      "learning_rate": 4.671529732957939e-05,
+      "loss": 0.9715,
+      "step": 771
+    },
+    {
+      "epoch": 0.197632,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.671103148195547e-05,
+      "loss": 0.9651,
+      "step": 772
+    },
+    {
+      "epoch": 0.197888,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.6706765634331545e-05,
+      "loss": 0.9836,
+      "step": 773
+    },
+    {
+      "epoch": 0.198144,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.670249978670762e-05,
+      "loss": 0.9686,
+      "step": 774
+    },
+    {
+      "epoch": 0.1984,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.6698233939083694e-05,
+      "loss": 0.9677,
+      "step": 775
+    },
+    {
+      "epoch": 0.198656,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.6693968091459775e-05,
+      "loss": 0.967,
+      "step": 776
+    },
+    {
+      "epoch": 0.198912,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.6689702243835856e-05,
+      "loss": 0.9454,
+      "step": 777
+    },
+    {
+      "epoch": 0.199168,
+      "grad_norm": 3.125,
+      "learning_rate": 4.668543639621193e-05,
+      "loss": 0.9217,
+      "step": 778
+    },
+    {
+      "epoch": 0.199424,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.6681170548588004e-05,
+      "loss": 0.946,
+      "step": 779
+    },
+    {
+      "epoch": 0.19968,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.6676904700964085e-05,
+      "loss": 0.9581,
+      "step": 780
+    },
+    {
+      "epoch": 0.199936,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.667263885334016e-05,
+      "loss": 0.9734,
+      "step": 781
+    },
+    {
+      "epoch": 0.200192,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.6668373005716234e-05,
+      "loss": 0.9528,
+      "step": 782
+    },
+    {
+      "epoch": 0.200448,
+      "grad_norm": 2.625,
+      "learning_rate": 4.6664107158092315e-05,
+      "loss": 0.954,
+      "step": 783
+    },
+    {
+      "epoch": 0.200704,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.6659841310468396e-05,
+      "loss": 0.9757,
+      "step": 784
+    },
+    {
+      "epoch": 0.20096,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.665557546284447e-05,
+      "loss": 0.9172,
+      "step": 785
+    },
+    {
+      "epoch": 0.201216,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.6651309615220544e-05,
+      "loss": 0.9396,
+      "step": 786
+    },
+    {
+      "epoch": 0.201472,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.664704376759662e-05,
+      "loss": 0.9069,
+      "step": 787
+    },
+    {
+      "epoch": 0.201728,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.66427779199727e-05,
+      "loss": 0.9471,
+      "step": 788
+    },
+    {
+      "epoch": 0.201984,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.663851207234878e-05,
+      "loss": 0.9714,
+      "step": 789
+    },
+    {
+      "epoch": 0.20224,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.6634246224724855e-05,
+      "loss": 0.9605,
+      "step": 790
+    },
+    {
+      "epoch": 0.202496,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.6629980377100936e-05,
+      "loss": 0.9515,
+      "step": 791
+    },
+    {
+      "epoch": 0.202752,
+      "grad_norm": 3.296875,
+      "learning_rate": 4.662571452947701e-05,
+      "loss": 0.9792,
+      "step": 792
+    },
+    {
+      "epoch": 0.203008,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.6621448681853085e-05,
+      "loss": 0.9637,
+      "step": 793
+    },
+    {
+      "epoch": 0.203264,
+      "grad_norm": 3.0,
+      "learning_rate": 4.661718283422916e-05,
+      "loss": 0.9465,
+      "step": 794
+    },
+    {
+      "epoch": 0.20352,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.661291698660524e-05,
+      "loss": 0.9713,
+      "step": 795
+    },
+    {
+      "epoch": 0.203776,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.660865113898132e-05,
+      "loss": 0.938,
+      "step": 796
+    },
+    {
+      "epoch": 0.204032,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.6604385291357395e-05,
+      "loss": 0.9608,
+      "step": 797
+    },
+    {
+      "epoch": 0.204288,
+      "grad_norm": 2.75,
+      "learning_rate": 4.660011944373347e-05,
+      "loss": 0.9672,
+      "step": 798
+    },
+    {
+      "epoch": 0.204544,
+      "grad_norm": 2.625,
+      "learning_rate": 4.659585359610955e-05,
+      "loss": 0.8968,
+      "step": 799
+    },
+    {
+      "epoch": 0.2048,
+      "grad_norm": 2.625,
+      "learning_rate": 4.6591587748485625e-05,
+      "loss": 0.9664,
+      "step": 800
+    },
+    {
+      "epoch": 0.205056,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.65873219008617e-05,
+      "loss": 0.9547,
+      "step": 801
+    },
+    {
+      "epoch": 0.205312,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.658305605323778e-05,
+      "loss": 0.9586,
+      "step": 802
+    },
+    {
+      "epoch": 0.205568,
+      "grad_norm": 2.875,
+      "learning_rate": 4.657879020561386e-05,
+      "loss": 0.9285,
+      "step": 803
+    },
+    {
+      "epoch": 0.205824,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.6574524357989935e-05,
+      "loss": 0.8998,
+      "step": 804
+    },
+    {
+      "epoch": 0.20608,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.657025851036601e-05,
+      "loss": 0.9609,
+      "step": 805
+    },
+    {
+      "epoch": 0.206336,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.656599266274209e-05,
+      "loss": 0.9251,
+      "step": 806
+    },
+    {
+      "epoch": 0.206592,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.6561726815118165e-05,
+      "loss": 1.0036,
+      "step": 807
+    },
+    {
+      "epoch": 0.206848,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.6557460967494246e-05,
+      "loss": 0.918,
+      "step": 808
+    },
+    {
+      "epoch": 0.207104,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.655319511987032e-05,
+      "loss": 0.9285,
+      "step": 809
+    },
+    {
+      "epoch": 0.20736,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.65489292722464e-05,
+      "loss": 0.9383,
+      "step": 810
+    },
+    {
+      "epoch": 0.207616,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.6544663424622476e-05,
+      "loss": 0.962,
+      "step": 811
+    },
+    {
+      "epoch": 0.207872,
+      "grad_norm": 2.75,
+      "learning_rate": 4.654039757699855e-05,
+      "loss": 0.9247,
+      "step": 812
+    },
+    {
+      "epoch": 0.208128,
+      "grad_norm": 2.875,
+      "learning_rate": 4.6536131729374624e-05,
+      "loss": 0.893,
+      "step": 813
+    },
+    {
+      "epoch": 0.208384,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.6531865881750705e-05,
+      "loss": 0.909,
+      "step": 814
+    },
+    {
+      "epoch": 0.20864,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.6527600034126786e-05,
+      "loss": 0.9071,
+      "step": 815
+    },
+    {
+      "epoch": 0.208896,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.652333418650286e-05,
+      "loss": 0.9485,
+      "step": 816
+    },
+    {
+      "epoch": 0.209152,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.6519068338878935e-05,
+      "loss": 0.9729,
+      "step": 817
+    },
+    {
+      "epoch": 0.209408,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.6514802491255016e-05,
+      "loss": 0.932,
+      "step": 818
+    },
+    {
+      "epoch": 0.209664,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.651053664363109e-05,
+      "loss": 0.9597,
+      "step": 819
+    },
+    {
+      "epoch": 0.20992,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.6506270796007164e-05,
+      "loss": 0.9527,
+      "step": 820
+    },
+    {
+      "epoch": 0.210176,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.6502004948383245e-05,
+      "loss": 0.9685,
+      "step": 821
+    },
+    {
+      "epoch": 0.210432,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.6497739100759326e-05,
+      "loss": 0.9665,
+      "step": 822
+    },
+    {
+      "epoch": 0.210688,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.64934732531354e-05,
+      "loss": 0.9012,
+      "step": 823
+    },
+    {
+      "epoch": 0.210944,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.6489207405511475e-05,
+      "loss": 0.9344,
+      "step": 824
+    },
+    {
+      "epoch": 0.2112,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.6484941557887556e-05,
+      "loss": 0.8912,
+      "step": 825
+    },
+    {
+      "epoch": 0.211456,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.648067571026363e-05,
+      "loss": 0.9521,
+      "step": 826
+    },
+    {
+      "epoch": 0.211712,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.647640986263971e-05,
+      "loss": 0.9134,
+      "step": 827
+    },
+    {
+      "epoch": 0.211968,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.6472144015015786e-05,
+      "loss": 0.9404,
+      "step": 828
+    },
+    {
+      "epoch": 0.212224,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.6467878167391867e-05,
+      "loss": 0.9372,
+      "step": 829
+    },
+    {
+      "epoch": 0.21248,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.646361231976794e-05,
+      "loss": 0.9403,
+      "step": 830
+    },
+    {
+      "epoch": 0.212736,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.6459346472144015e-05,
+      "loss": 0.9587,
+      "step": 831
+    },
+    {
+      "epoch": 0.212992,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.645508062452009e-05,
+      "loss": 0.8967,
+      "step": 832
+    },
+    {
+      "epoch": 0.213248,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.645081477689617e-05,
+      "loss": 0.9115,
+      "step": 833
+    },
+    {
+      "epoch": 0.213504,
+      "grad_norm": 2.375,
+      "learning_rate": 4.644654892927225e-05,
+      "loss": 0.9156,
+      "step": 834
+    },
+    {
+      "epoch": 0.21376,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.6442283081648326e-05,
+      "loss": 0.9111,
+      "step": 835
+    },
+    {
+      "epoch": 0.214016,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.64380172340244e-05,
+      "loss": 0.9633,
+      "step": 836
+    },
+    {
+      "epoch": 0.214272,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.643375138640048e-05,
+      "loss": 0.9434,
+      "step": 837
+    },
+    {
+      "epoch": 0.214528,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.6429485538776555e-05,
+      "loss": 0.9391,
+      "step": 838
+    },
+    {
+      "epoch": 0.214784,
+      "grad_norm": 2.625,
+      "learning_rate": 4.642521969115263e-05,
+      "loss": 0.953,
+      "step": 839
+    },
+    {
+      "epoch": 0.21504,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.642095384352871e-05,
+      "loss": 0.9327,
+      "step": 840
+    },
+    {
+      "epoch": 0.215296,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.641668799590479e-05,
+      "loss": 0.9473,
+      "step": 841
+    },
+    {
+      "epoch": 0.215552,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.6412422148280866e-05,
+      "loss": 0.9374,
+      "step": 842
+    },
+    {
+      "epoch": 0.215808,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.640815630065694e-05,
+      "loss": 0.8839,
+      "step": 843
+    },
+    {
+      "epoch": 0.216064,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.640389045303302e-05,
+      "loss": 0.9283,
+      "step": 844
+    },
+    {
+      "epoch": 0.21632,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.6399624605409096e-05,
+      "loss": 0.9211,
+      "step": 845
+    },
+    {
+      "epoch": 0.216576,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.6395358757785177e-05,
+      "loss": 0.9196,
+      "step": 846
+    },
+    {
+      "epoch": 0.216832,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.639109291016125e-05,
+      "loss": 0.8719,
+      "step": 847
+    },
+    {
+      "epoch": 0.217088,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.638682706253733e-05,
+      "loss": 0.916,
+      "step": 848
+    },
+    {
+      "epoch": 0.217344,
+      "grad_norm": 2.75,
+      "learning_rate": 4.6382561214913406e-05,
+      "loss": 0.9521,
+      "step": 849
+    },
+    {
+      "epoch": 0.2176,
+      "grad_norm": 3.0,
+      "learning_rate": 4.637829536728948e-05,
+      "loss": 0.896,
+      "step": 850
+    },
+    {
+      "epoch": 0.217856,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.6374029519665555e-05,
+      "loss": 0.9073,
+      "step": 851
+    },
+    {
+      "epoch": 0.218112,
+      "grad_norm": 3.0,
+      "learning_rate": 4.6369763672041636e-05,
+      "loss": 0.9213,
+      "step": 852
+    },
+    {
+      "epoch": 0.218368,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.636549782441772e-05,
+      "loss": 0.8887,
+      "step": 853
+    },
+    {
+      "epoch": 0.218624,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.636123197679379e-05,
+      "loss": 0.9401,
+      "step": 854
+    },
+    {
+      "epoch": 0.21888,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.635696612916987e-05,
+      "loss": 0.9456,
+      "step": 855
+    },
+    {
+      "epoch": 0.219136,
+      "grad_norm": 2.875,
+      "learning_rate": 4.6352700281545946e-05,
+      "loss": 0.93,
+      "step": 856
+    },
+    {
+      "epoch": 0.219392,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.634843443392202e-05,
+      "loss": 0.8592,
+      "step": 857
+    },
+    {
+      "epoch": 0.219648,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.6344168586298095e-05,
+      "loss": 0.9317,
+      "step": 858
+    },
+    {
+      "epoch": 0.219904,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.6339902738674176e-05,
+      "loss": 0.9288,
+      "step": 859
+    },
+    {
+      "epoch": 0.22016,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.633563689105026e-05,
+      "loss": 0.862,
+      "step": 860
+    },
+    {
+      "epoch": 0.220416,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.633137104342633e-05,
+      "loss": 0.9135,
+      "step": 861
+    },
+    {
+      "epoch": 0.220672,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.6327105195802405e-05,
+      "loss": 0.9251,
+      "step": 862
+    },
+    {
+      "epoch": 0.220928,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.6322839348178487e-05,
+      "loss": 0.9043,
+      "step": 863
+    },
+    {
+      "epoch": 0.221184,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.631857350055456e-05,
+      "loss": 0.883,
+      "step": 864
+    },
+    {
+      "epoch": 0.22144,
+      "grad_norm": 3.0,
+      "learning_rate": 4.631430765293064e-05,
+      "loss": 0.9097,
+      "step": 865
+    },
+    {
+      "epoch": 0.221696,
+      "grad_norm": 2.875,
+      "learning_rate": 4.6310041805306716e-05,
+      "loss": 0.9406,
+      "step": 866
+    },
+    {
+      "epoch": 0.221952,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.63057759576828e-05,
+      "loss": 0.9144,
+      "step": 867
+    },
+    {
+      "epoch": 0.222208,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.630151011005887e-05,
+      "loss": 0.8875,
+      "step": 868
+    },
+    {
+      "epoch": 0.222464,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.6297244262434946e-05,
+      "loss": 0.925,
+      "step": 869
+    },
+    {
+      "epoch": 0.22272,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.629297841481103e-05,
+      "loss": 0.8806,
+      "step": 870
+    },
+    {
+      "epoch": 0.222976,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.62887125671871e-05,
+      "loss": 0.8936,
+      "step": 871
+    },
+    {
+      "epoch": 0.223232,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.628444671956318e-05,
+      "loss": 0.892,
+      "step": 872
+    },
+    {
+      "epoch": 0.223488,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.6280180871939256e-05,
+      "loss": 0.9039,
+      "step": 873
+    },
+    {
+      "epoch": 0.223744,
+      "grad_norm": 2.75,
+      "learning_rate": 4.627591502431534e-05,
+      "loss": 0.857,
+      "step": 874
+    },
+    {
+      "epoch": 0.224,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.627164917669141e-05,
+      "loss": 0.9158,
+      "step": 875
+    },
+    {
+      "epoch": 0.224256,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.6267383329067486e-05,
+      "loss": 0.9105,
+      "step": 876
+    },
+    {
+      "epoch": 0.224512,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.626311748144356e-05,
+      "loss": 0.9084,
+      "step": 877
+    },
+    {
+      "epoch": 0.224768,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.625885163381964e-05,
+      "loss": 0.9195,
+      "step": 878
+    },
+    {
+      "epoch": 0.225024,
+      "grad_norm": 2.75,
+      "learning_rate": 4.625458578619572e-05,
+      "loss": 0.9086,
+      "step": 879
+    },
+    {
+      "epoch": 0.22528,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.6250319938571796e-05,
+      "loss": 0.9117,
+      "step": 880
+    },
+    {
+      "epoch": 0.225536,
+      "grad_norm": 3.125,
+      "learning_rate": 4.624605409094787e-05,
+      "loss": 0.8896,
+      "step": 881
+    },
+    {
+      "epoch": 0.225792,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.624178824332395e-05,
+      "loss": 0.9216,
+      "step": 882
+    },
+    {
+      "epoch": 0.226048,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.6237522395700026e-05,
+      "loss": 0.9018,
+      "step": 883
+    },
+    {
+      "epoch": 0.226304,
+      "grad_norm": 3.25,
+      "learning_rate": 4.623325654807611e-05,
+      "loss": 0.9083,
+      "step": 884
+    },
+    {
+      "epoch": 0.22656,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.622899070045218e-05,
+      "loss": 0.8984,
+      "step": 885
+    },
+    {
+      "epoch": 0.226816,
+      "grad_norm": 2.5,
+      "learning_rate": 4.622472485282826e-05,
+      "loss": 0.9448,
+      "step": 886
+    },
+    {
+      "epoch": 0.227072,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.622045900520434e-05,
+      "loss": 0.9051,
+      "step": 887
+    },
+    {
+      "epoch": 0.227328,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.621619315758041e-05,
+      "loss": 0.9039,
+      "step": 888
+    },
+    {
+      "epoch": 0.227584,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.621192730995649e-05,
+      "loss": 0.8937,
+      "step": 889
+    },
+    {
+      "epoch": 0.22784,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.6207661462332566e-05,
+      "loss": 0.8839,
+      "step": 890
+    },
+    {
+      "epoch": 0.228096,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.620339561470865e-05,
+      "loss": 0.9112,
+      "step": 891
+    },
+    {
+      "epoch": 0.228352,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.619912976708472e-05,
+      "loss": 0.9193,
+      "step": 892
+    },
+    {
+      "epoch": 0.228608,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.61948639194608e-05,
+      "loss": 0.9017,
+      "step": 893
+    },
+    {
+      "epoch": 0.228864,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.619059807183688e-05,
+      "loss": 0.9122,
+      "step": 894
+    },
+    {
+      "epoch": 0.22912,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.618633222421295e-05,
+      "loss": 0.8612,
+      "step": 895
+    },
+    {
+      "epoch": 0.229376,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.6182066376589025e-05,
+      "loss": 0.8832,
+      "step": 896
+    },
+    {
+      "epoch": 0.229632,
+      "grad_norm": 2.5,
+      "learning_rate": 4.6177800528965106e-05,
+      "loss": 0.9129,
+      "step": 897
+    },
+    {
+      "epoch": 0.229888,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.617353468134119e-05,
+      "loss": 0.8786,
+      "step": 898
+    },
+    {
+      "epoch": 0.230144,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.616926883371726e-05,
+      "loss": 0.8972,
+      "step": 899
+    },
+    {
+      "epoch": 0.2304,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.6165002986093336e-05,
+      "loss": 0.9036,
+      "step": 900
+    },
+    {
+      "epoch": 0.230656,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.616073713846942e-05,
+      "loss": 0.9004,
+      "step": 901
+    },
+    {
+      "epoch": 0.230912,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.615647129084549e-05,
+      "loss": 0.9321,
+      "step": 902
+    },
+    {
+      "epoch": 0.231168,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.6152205443221566e-05,
+      "loss": 0.9085,
+      "step": 903
+    },
+    {
+      "epoch": 0.231424,
+      "grad_norm": 2.875,
+      "learning_rate": 4.6147939595597653e-05,
+      "loss": 0.901,
+      "step": 904
+    },
+    {
+      "epoch": 0.23168,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.614367374797373e-05,
+      "loss": 0.8535,
+      "step": 905
+    },
+    {
+      "epoch": 0.231936,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.61394079003498e-05,
+      "loss": 0.8907,
+      "step": 906
+    },
+    {
+      "epoch": 0.232192,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.6135142052725876e-05,
+      "loss": 0.8898,
+      "step": 907
+    },
+    {
+      "epoch": 0.232448,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.613087620510196e-05,
+      "loss": 0.8634,
+      "step": 908
+    },
+    {
+      "epoch": 0.232704,
+      "grad_norm": 3.0,
+      "learning_rate": 4.612661035747803e-05,
+      "loss": 0.889,
+      "step": 909
+    },
+    {
+      "epoch": 0.23296,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.612234450985411e-05,
+      "loss": 0.8987,
+      "step": 910
+    },
+    {
+      "epoch": 0.233216,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.611807866223019e-05,
+      "loss": 0.9151,
+      "step": 911
+    },
+    {
+      "epoch": 0.233472,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.611381281460627e-05,
+      "loss": 0.9043,
+      "step": 912
+    },
+    {
+      "epoch": 0.233728,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.610954696698234e-05,
+      "loss": 0.8916,
+      "step": 913
+    },
+    {
+      "epoch": 0.233984,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.6105281119358416e-05,
+      "loss": 0.8897,
+      "step": 914
+    },
+    {
+      "epoch": 0.23424,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.610101527173449e-05,
+      "loss": 0.9014,
+      "step": 915
+    },
+    {
+      "epoch": 0.234496,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.609674942411057e-05,
+      "loss": 0.86,
+      "step": 916
+    },
+    {
+      "epoch": 0.234752,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.609248357648665e-05,
+      "loss": 0.8805,
+      "step": 917
+    },
+    {
+      "epoch": 0.235008,
+      "grad_norm": 2.875,
+      "learning_rate": 4.608821772886273e-05,
+      "loss": 0.8773,
+      "step": 918
+    },
+    {
+      "epoch": 0.235264,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.60839518812388e-05,
+      "loss": 0.8947,
+      "step": 919
+    },
+    {
+      "epoch": 0.23552,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.607968603361488e-05,
+      "loss": 0.8729,
+      "step": 920
+    },
+    {
+      "epoch": 0.235776,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.6075420185990957e-05,
+      "loss": 0.8939,
+      "step": 921
+    },
+    {
+      "epoch": 0.236032,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.607115433836703e-05,
+      "loss": 0.905,
+      "step": 922
+    },
+    {
+      "epoch": 0.236288,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.606688849074312e-05,
+      "loss": 0.8786,
+      "step": 923
+    },
+    {
+      "epoch": 0.236544,
+      "grad_norm": 2.5,
+      "learning_rate": 4.606262264311919e-05,
+      "loss": 0.8644,
+      "step": 924
+    },
+    {
+      "epoch": 0.2368,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.605835679549527e-05,
+      "loss": 0.9417,
+      "step": 925
+    },
+    {
+      "epoch": 0.237056,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.605409094787134e-05,
+      "loss": 0.892,
+      "step": 926
+    },
+    {
+      "epoch": 0.237312,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.604982510024742e-05,
+      "loss": 0.8649,
+      "step": 927
+    },
+    {
+      "epoch": 0.237568,
+      "grad_norm": 2.625,
+      "learning_rate": 4.60455592526235e-05,
+      "loss": 0.8885,
+      "step": 928
+    },
+    {
+      "epoch": 0.237824,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.604129340499958e-05,
+      "loss": 0.8721,
+      "step": 929
+    },
+    {
+      "epoch": 0.23808,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.603702755737565e-05,
+      "loss": 0.89,
+      "step": 930
+    },
+    {
+      "epoch": 0.238336,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.603276170975173e-05,
+      "loss": 0.9133,
+      "step": 931
+    },
+    {
+      "epoch": 0.238592,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.602849586212781e-05,
+      "loss": 0.8898,
+      "step": 932
+    },
+    {
+      "epoch": 0.238848,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.602423001450388e-05,
+      "loss": 0.8714,
+      "step": 933
+    },
+    {
+      "epoch": 0.239104,
+      "grad_norm": 2.875,
+      "learning_rate": 4.6019964166879956e-05,
+      "loss": 0.8845,
+      "step": 934
+    },
+    {
+      "epoch": 0.23936,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.601569831925604e-05,
+      "loss": 0.8654,
+      "step": 935
+    },
+    {
+      "epoch": 0.239616,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.601143247163212e-05,
+      "loss": 0.8652,
+      "step": 936
+    },
+    {
+      "epoch": 0.239872,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.600716662400819e-05,
+      "loss": 0.8301,
+      "step": 937
+    },
+    {
+      "epoch": 0.240128,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.600290077638427e-05,
+      "loss": 0.9076,
+      "step": 938
+    },
+    {
+      "epoch": 0.240384,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.599863492876035e-05,
+      "loss": 0.8616,
+      "step": 939
+    },
+    {
+      "epoch": 0.24064,
+      "grad_norm": 2.625,
+      "learning_rate": 4.599436908113642e-05,
+      "loss": 0.8695,
+      "step": 940
+    },
+    {
+      "epoch": 0.240896,
+      "grad_norm": 2.625,
+      "learning_rate": 4.5990103233512496e-05,
+      "loss": 0.8602,
+      "step": 941
+    },
+    {
+      "epoch": 0.241152,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.598583738588858e-05,
+      "loss": 0.8731,
+      "step": 942
+    },
+    {
+      "epoch": 0.241408,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.598157153826466e-05,
+      "loss": 0.8515,
+      "step": 943
+    },
+    {
+      "epoch": 0.241664,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.597730569064073e-05,
+      "loss": 0.8323,
+      "step": 944
+    },
+    {
+      "epoch": 0.24192,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.597303984301681e-05,
+      "loss": 0.901,
+      "step": 945
+    },
+    {
+      "epoch": 0.242176,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.596877399539289e-05,
+      "loss": 0.8403,
+      "step": 946
+    },
+    {
+      "epoch": 0.242432,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.596450814776896e-05,
+      "loss": 0.8844,
+      "step": 947
+    },
+    {
+      "epoch": 0.242688,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.596024230014504e-05,
+      "loss": 0.8582,
+      "step": 948
+    },
+    {
+      "epoch": 0.242944,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.595597645252112e-05,
+      "loss": 0.8715,
+      "step": 949
+    },
+    {
+      "epoch": 0.2432,
+      "grad_norm": 2.875,
+      "learning_rate": 4.59517106048972e-05,
+      "loss": 0.8891,
+      "step": 950
+    },
+    {
+      "epoch": 0.243456,
+      "grad_norm": 3.125,
+      "learning_rate": 4.594744475727327e-05,
+      "loss": 0.8585,
+      "step": 951
+    },
+    {
+      "epoch": 0.243712,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.594317890964935e-05,
+      "loss": 0.8741,
+      "step": 952
+    },
+    {
+      "epoch": 0.243968,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.593891306202543e-05,
+      "loss": 0.8668,
+      "step": 953
+    },
+    {
+      "epoch": 0.244224,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.59346472144015e-05,
+      "loss": 0.8835,
+      "step": 954
+    },
+    {
+      "epoch": 0.24448,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.593038136677758e-05,
+      "loss": 0.8872,
+      "step": 955
+    },
+    {
+      "epoch": 0.244736,
+      "grad_norm": 2.875,
+      "learning_rate": 4.592611551915366e-05,
+      "loss": 0.9271,
+      "step": 956
+    },
+    {
+      "epoch": 0.244992,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.592184967152974e-05,
+      "loss": 0.8419,
+      "step": 957
+    },
+    {
+      "epoch": 0.245248,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.591758382390581e-05,
+      "loss": 0.8864,
+      "step": 958
+    },
+    {
+      "epoch": 0.245504,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.591331797628189e-05,
+      "loss": 0.8904,
+      "step": 959
+    },
+    {
+      "epoch": 0.24576,
+      "grad_norm": 3.0,
+      "learning_rate": 4.590905212865796e-05,
+      "loss": 0.8517,
+      "step": 960
+    },
+    {
+      "epoch": 0.246016,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.590478628103404e-05,
+      "loss": 0.8183,
+      "step": 961
+    },
+    {
+      "epoch": 0.246272,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.5900520433410123e-05,
+      "loss": 0.8743,
+      "step": 962
+    },
+    {
+      "epoch": 0.246528,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.58962545857862e-05,
+      "loss": 0.8363,
+      "step": 963
+    },
+    {
+      "epoch": 0.246784,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.589198873816227e-05,
+      "loss": 0.8421,
+      "step": 964
+    },
+    {
+      "epoch": 0.24704,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.588772289053835e-05,
+      "loss": 0.871,
+      "step": 965
+    },
+    {
+      "epoch": 0.247296,
+      "grad_norm": 3.3125,
+      "learning_rate": 4.588345704291443e-05,
+      "loss": 0.8446,
+      "step": 966
+    },
+    {
+      "epoch": 0.247552,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.587919119529051e-05,
+      "loss": 0.8753,
+      "step": 967
+    },
+    {
+      "epoch": 0.247808,
+      "grad_norm": 3.0,
+      "learning_rate": 4.587492534766658e-05,
+      "loss": 0.8317,
+      "step": 968
+    },
+    {
+      "epoch": 0.248064,
+      "grad_norm": 3.125,
+      "learning_rate": 4.5870659500042664e-05,
+      "loss": 0.8805,
+      "step": 969
+    },
+    {
+      "epoch": 0.24832,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.586639365241874e-05,
+      "loss": 0.8603,
+      "step": 970
+    },
+    {
+      "epoch": 0.248576,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.586212780479481e-05,
+      "loss": 0.8581,
+      "step": 971
+    },
+    {
+      "epoch": 0.248832,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.585786195717089e-05,
+      "loss": 0.8124,
+      "step": 972
+    },
+    {
+      "epoch": 0.249088,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.585359610954697e-05,
+      "loss": 0.8497,
+      "step": 973
+    },
+    {
+      "epoch": 0.249344,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.584933026192305e-05,
+      "loss": 0.8796,
+      "step": 974
+    },
+    {
+      "epoch": 0.2496,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.584506441429912e-05,
+      "loss": 0.8587,
+      "step": 975
+    },
+    {
+      "epoch": 0.249856,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.5840798566675204e-05,
+      "loss": 0.8785,
+      "step": 976
+    },
+    {
+      "epoch": 0.250112,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.583653271905128e-05,
+      "loss": 0.82,
+      "step": 977
+    },
+    {
+      "epoch": 0.250368,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.583226687142735e-05,
+      "loss": 0.8628,
+      "step": 978
+    },
+    {
+      "epoch": 0.250624,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.582800102380343e-05,
+      "loss": 0.8211,
+      "step": 979
+    },
+    {
+      "epoch": 0.25088,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.582373517617951e-05,
+      "loss": 0.8486,
+      "step": 980
+    },
+    {
+      "epoch": 0.251136,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.581946932855559e-05,
+      "loss": 0.832,
+      "step": 981
+    },
+    {
+      "epoch": 0.251392,
+      "grad_norm": 2.75,
+      "learning_rate": 4.581520348093166e-05,
+      "loss": 0.8369,
+      "step": 982
+    },
+    {
+      "epoch": 0.251648,
+      "grad_norm": 2.875,
+      "learning_rate": 4.581093763330774e-05,
+      "loss": 0.8263,
+      "step": 983
+    },
+    {
+      "epoch": 0.251904,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.580667178568382e-05,
+      "loss": 0.8426,
+      "step": 984
+    },
+    {
+      "epoch": 0.25216,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.580240593805989e-05,
+      "loss": 0.8527,
+      "step": 985
+    },
+    {
+      "epoch": 0.252416,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.5798140090435974e-05,
+      "loss": 0.858,
+      "step": 986
+    },
+    {
+      "epoch": 0.252672,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.5793874242812055e-05,
+      "loss": 0.8523,
+      "step": 987
+    },
+    {
+      "epoch": 0.252928,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.578960839518813e-05,
+      "loss": 0.8854,
+      "step": 988
+    },
+    {
+      "epoch": 0.253184,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.57853425475642e-05,
+      "loss": 0.8431,
+      "step": 989
+    },
+    {
+      "epoch": 0.25344,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.578107669994028e-05,
+      "loss": 0.8519,
+      "step": 990
+    },
+    {
+      "epoch": 0.253696,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.577681085231636e-05,
+      "loss": 0.8495,
+      "step": 991
+    },
+    {
+      "epoch": 0.253952,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.577254500469243e-05,
+      "loss": 0.8938,
+      "step": 992
+    },
+    {
+      "epoch": 0.254208,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.5768279157068514e-05,
+      "loss": 0.8255,
+      "step": 993
+    },
+    {
+      "epoch": 0.254464,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.576401330944459e-05,
+      "loss": 0.8685,
+      "step": 994
+    },
+    {
+      "epoch": 0.25472,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.575974746182067e-05,
+      "loss": 0.8755,
+      "step": 995
+    },
+    {
+      "epoch": 0.254976,
+      "grad_norm": 2.75,
+      "learning_rate": 4.5755481614196743e-05,
+      "loss": 0.8152,
+      "step": 996
+    },
+    {
+      "epoch": 0.255232,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.575121576657282e-05,
+      "loss": 0.837,
+      "step": 997
+    },
+    {
+      "epoch": 0.255488,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.574694991894889e-05,
+      "loss": 0.8594,
+      "step": 998
+    },
+    {
+      "epoch": 0.255744,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.574268407132497e-05,
+      "loss": 0.8701,
+      "step": 999
+    },
+    {
+      "epoch": 0.256,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.5738418223701054e-05,
+      "loss": 0.8559,
+      "step": 1000
+    },
+    {
+      "epoch": 0.256256,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.573415237607713e-05,
+      "loss": 0.8478,
+      "step": 1001
+    },
+    {
+      "epoch": 0.256512,
+      "grad_norm": 2.75,
+      "learning_rate": 4.572988652845321e-05,
+      "loss": 0.8233,
+      "step": 1002
+    },
+    {
+      "epoch": 0.256768,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.5725620680829284e-05,
+      "loss": 0.7952,
+      "step": 1003
+    },
+    {
+      "epoch": 0.257024,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.572135483320536e-05,
+      "loss": 0.7962,
+      "step": 1004
+    },
+    {
+      "epoch": 0.25728,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.571708898558143e-05,
+      "loss": 0.8512,
+      "step": 1005
+    },
+    {
+      "epoch": 0.257536,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.571282313795752e-05,
+      "loss": 0.8246,
+      "step": 1006
+    },
+    {
+      "epoch": 0.257792,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.5708557290333594e-05,
+      "loss": 0.8635,
+      "step": 1007
+    },
+    {
+      "epoch": 0.258048,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.570429144270967e-05,
+      "loss": 0.827,
+      "step": 1008
+    },
+    {
+      "epoch": 0.258304,
+      "grad_norm": 2.875,
+      "learning_rate": 4.570002559508574e-05,
+      "loss": 0.8616,
+      "step": 1009
+    },
+    {
+      "epoch": 0.25856,
+      "grad_norm": 2.75,
+      "learning_rate": 4.5695759747461824e-05,
+      "loss": 0.8313,
+      "step": 1010
+    },
+    {
+      "epoch": 0.258816,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.56914938998379e-05,
+      "loss": 0.8951,
+      "step": 1011
+    },
+    {
+      "epoch": 0.259072,
+      "grad_norm": 3.40625,
+      "learning_rate": 4.568722805221398e-05,
+      "loss": 0.8052,
+      "step": 1012
+    },
+    {
+      "epoch": 0.259328,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.568296220459005e-05,
+      "loss": 0.8491,
+      "step": 1013
+    },
+    {
+      "epoch": 0.259584,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.5678696356966134e-05,
+      "loss": 0.8619,
+      "step": 1014
+    },
+    {
+      "epoch": 0.25984,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.567443050934221e-05,
+      "loss": 0.9084,
+      "step": 1015
+    },
+    {
+      "epoch": 0.260096,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.567016466171828e-05,
+      "loss": 0.8961,
+      "step": 1016
+    },
+    {
+      "epoch": 0.260352,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.566589881409436e-05,
+      "loss": 0.8228,
+      "step": 1017
+    },
+    {
+      "epoch": 0.260608,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.566163296647044e-05,
+      "loss": 0.8517,
+      "step": 1018
+    },
+    {
+      "epoch": 0.260864,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.565736711884652e-05,
+      "loss": 0.8452,
+      "step": 1019
+    },
+    {
+      "epoch": 0.26112,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.5653101271222594e-05,
+      "loss": 0.8579,
+      "step": 1020
+    },
+    {
+      "epoch": 0.261376,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.5648835423598675e-05,
+      "loss": 0.8825,
+      "step": 1021
+    },
+    {
+      "epoch": 0.261632,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.564456957597475e-05,
+      "loss": 0.8377,
+      "step": 1022
+    },
+    {
+      "epoch": 0.261888,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.564030372835082e-05,
+      "loss": 0.811,
+      "step": 1023
+    },
+    {
+      "epoch": 0.262144,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.56360378807269e-05,
+      "loss": 0.8531,
+      "step": 1024
+    },
+    {
+      "epoch": 0.2624,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.5631772033102985e-05,
+      "loss": 0.8744,
+      "step": 1025
+    },
+    {
+      "epoch": 0.262656,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.562750618547906e-05,
+      "loss": 0.8267,
+      "step": 1026
+    },
+    {
+      "epoch": 0.262912,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.5623240337855134e-05,
+      "loss": 0.8348,
+      "step": 1027
+    },
+    {
+      "epoch": 0.263168,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.561897449023121e-05,
+      "loss": 0.8625,
+      "step": 1028
+    },
+    {
+      "epoch": 0.263424,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.561470864260729e-05,
+      "loss": 0.9029,
+      "step": 1029
+    },
+    {
+      "epoch": 0.26368,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.561044279498336e-05,
+      "loss": 0.8264,
+      "step": 1030
+    },
+    {
+      "epoch": 0.263936,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.5606176947359444e-05,
+      "loss": 0.8292,
+      "step": 1031
+    },
+    {
+      "epoch": 0.264192,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.560191109973552e-05,
+      "loss": 0.8178,
+      "step": 1032
+    },
+    {
+      "epoch": 0.264448,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.55976452521116e-05,
+      "loss": 0.8403,
+      "step": 1033
+    },
+    {
+      "epoch": 0.264704,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.5593379404487674e-05,
+      "loss": 0.8668,
+      "step": 1034
+    },
+    {
+      "epoch": 0.26496,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.558911355686375e-05,
+      "loss": 0.824,
+      "step": 1035
+    },
+    {
+      "epoch": 0.265216,
+      "grad_norm": 2.625,
+      "learning_rate": 4.558484770923983e-05,
+      "loss": 0.808,
+      "step": 1036
+    },
+    {
+      "epoch": 0.265472,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.5580581861615904e-05,
+      "loss": 0.8612,
+      "step": 1037
+    },
+    {
+      "epoch": 0.265728,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.5576316013991985e-05,
+      "loss": 0.8551,
+      "step": 1038
+    },
+    {
+      "epoch": 0.265984,
+      "grad_norm": 2.75,
+      "learning_rate": 4.557205016636806e-05,
+      "loss": 0.8335,
+      "step": 1039
+    },
+    {
+      "epoch": 0.26624,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.556778431874414e-05,
+      "loss": 0.854,
+      "step": 1040
+    },
+    {
+      "epoch": 0.266496,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.5563518471120214e-05,
+      "loss": 0.8513,
+      "step": 1041
+    },
+    {
+      "epoch": 0.266752,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.555925262349629e-05,
+      "loss": 0.8691,
+      "step": 1042
+    },
+    {
+      "epoch": 0.267008,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.555498677587236e-05,
+      "loss": 0.8174,
+      "step": 1043
+    },
+    {
+      "epoch": 0.267264,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.5550720928248444e-05,
+      "loss": 0.8657,
+      "step": 1044
+    },
+    {
+      "epoch": 0.26752,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.5546455080624525e-05,
+      "loss": 0.8336,
+      "step": 1045
+    },
+    {
+      "epoch": 0.267776,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.55421892330006e-05,
+      "loss": 0.8728,
+      "step": 1046
+    },
+    {
+      "epoch": 0.268032,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.553792338537667e-05,
+      "loss": 0.8139,
+      "step": 1047
+    },
+    {
+      "epoch": 0.268288,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.5533657537752754e-05,
+      "loss": 0.8269,
+      "step": 1048
+    },
+    {
+      "epoch": 0.268544,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.552939169012883e-05,
+      "loss": 0.8432,
+      "step": 1049
+    },
+    {
+      "epoch": 0.2688,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.552512584250491e-05,
+      "loss": 0.8187,
+      "step": 1050
+    },
+    {
+      "epoch": 0.269056,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.552085999488099e-05,
+      "loss": 0.8349,
+      "step": 1051
+    },
+    {
+      "epoch": 0.269312,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.5516594147257065e-05,
+      "loss": 0.8355,
+      "step": 1052
+    },
+    {
+      "epoch": 0.269568,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.551232829963314e-05,
+      "loss": 0.8225,
+      "step": 1053
+    },
+    {
+      "epoch": 0.269824,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.5508062452009213e-05,
+      "loss": 0.8456,
+      "step": 1054
+    },
+    {
+      "epoch": 0.27008,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.5503796604385295e-05,
+      "loss": 0.8386,
+      "step": 1055
+    },
+    {
+      "epoch": 0.270336,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.549953075676137e-05,
+      "loss": 0.8291,
+      "step": 1056
+    },
+    {
+      "epoch": 0.270592,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.549526490913745e-05,
+      "loss": 0.8471,
+      "step": 1057
+    },
+    {
+      "epoch": 0.270848,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.5490999061513524e-05,
+      "loss": 0.8195,
+      "step": 1058
+    },
+    {
+      "epoch": 0.271104,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.5486733213889605e-05,
+      "loss": 0.8255,
+      "step": 1059
+    },
+    {
+      "epoch": 0.27136,
+      "grad_norm": 2.875,
+      "learning_rate": 4.548246736626568e-05,
+      "loss": 0.8035,
+      "step": 1060
+    },
+    {
+      "epoch": 0.271616,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.5478201518641754e-05,
+      "loss": 0.8109,
+      "step": 1061
+    },
+    {
+      "epoch": 0.271872,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.547393567101783e-05,
+      "loss": 0.8188,
+      "step": 1062
+    },
+    {
+      "epoch": 0.272128,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.546966982339391e-05,
+      "loss": 0.8182,
+      "step": 1063
+    },
+    {
+      "epoch": 0.272384,
+      "grad_norm": 3.125,
+      "learning_rate": 4.546540397576999e-05,
+      "loss": 0.8484,
+      "step": 1064
+    },
+    {
+      "epoch": 0.27264,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.5461138128146064e-05,
+      "loss": 0.8189,
+      "step": 1065
+    },
+    {
+      "epoch": 0.272896,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.545687228052214e-05,
+      "loss": 0.8047,
+      "step": 1066
+    },
+    {
+      "epoch": 0.273152,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.545260643289822e-05,
+      "loss": 0.8643,
+      "step": 1067
+    },
+    {
+      "epoch": 0.273408,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.5448340585274294e-05,
+      "loss": 0.7722,
+      "step": 1068
+    },
+    {
+      "epoch": 0.273664,
+      "grad_norm": 2.375,
+      "learning_rate": 4.5444074737650375e-05,
+      "loss": 0.84,
+      "step": 1069
+    },
+    {
+      "epoch": 0.27392,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.5439808890026456e-05,
+      "loss": 0.8644,
+      "step": 1070
+    },
+    {
+      "epoch": 0.274176,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.543554304240253e-05,
+      "loss": 0.8523,
+      "step": 1071
+    },
+    {
+      "epoch": 0.274432,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.5431277194778604e-05,
+      "loss": 0.814,
+      "step": 1072
+    },
+    {
+      "epoch": 0.274688,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.542701134715468e-05,
+      "loss": 0.8274,
+      "step": 1073
+    },
+    {
+      "epoch": 0.274944,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.542274549953076e-05,
+      "loss": 0.833,
+      "step": 1074
+    },
+    {
+      "epoch": 0.2752,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.5418479651906834e-05,
+      "loss": 0.8145,
+      "step": 1075
+    },
+    {
+      "epoch": 0.275456,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.5414213804282915e-05,
+      "loss": 0.8203,
+      "step": 1076
+    },
+    {
+      "epoch": 0.275712,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.540994795665899e-05,
+      "loss": 0.8462,
+      "step": 1077
+    },
+    {
+      "epoch": 0.275968,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.540568210903507e-05,
+      "loss": 0.8075,
+      "step": 1078
+    },
+    {
+      "epoch": 0.276224,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.5401416261411145e-05,
+      "loss": 0.8172,
+      "step": 1079
+    },
+    {
+      "epoch": 0.27648,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.539715041378722e-05,
+      "loss": 0.8009,
+      "step": 1080
+    },
+    {
+      "epoch": 0.276736,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.539288456616329e-05,
+      "loss": 0.836,
+      "step": 1081
+    },
+    {
+      "epoch": 0.276992,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.5388618718539374e-05,
+      "loss": 0.8407,
+      "step": 1082
+    },
+    {
+      "epoch": 0.277248,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.5384352870915455e-05,
+      "loss": 0.8634,
+      "step": 1083
+    },
+    {
+      "epoch": 0.277504,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.538008702329153e-05,
+      "loss": 0.8502,
+      "step": 1084
+    },
+    {
+      "epoch": 0.27776,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.537582117566761e-05,
+      "loss": 0.7966,
+      "step": 1085
+    },
+    {
+      "epoch": 0.278016,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.5371555328043685e-05,
+      "loss": 0.8427,
+      "step": 1086
+    },
+    {
+      "epoch": 0.278272,
+      "grad_norm": 2.875,
+      "learning_rate": 4.536728948041976e-05,
+      "loss": 0.8427,
+      "step": 1087
+    },
+    {
+      "epoch": 0.278528,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.536302363279584e-05,
+      "loss": 0.8346,
+      "step": 1088
+    },
+    {
+      "epoch": 0.278784,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.535875778517192e-05,
+      "loss": 0.8081,
+      "step": 1089
+    },
+    {
+      "epoch": 0.27904,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.5354491937547995e-05,
+      "loss": 0.8027,
+      "step": 1090
+    },
+    {
+      "epoch": 0.279296,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.535022608992407e-05,
+      "loss": 0.8278,
+      "step": 1091
+    },
+    {
+      "epoch": 0.279552,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.5345960242300144e-05,
+      "loss": 0.8215,
+      "step": 1092
+    },
+    {
+      "epoch": 0.279808,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.5341694394676225e-05,
+      "loss": 0.751,
+      "step": 1093
+    },
+    {
+      "epoch": 0.280064,
+      "grad_norm": 2.625,
+      "learning_rate": 4.53374285470523e-05,
+      "loss": 0.8307,
+      "step": 1094
+    },
+    {
+      "epoch": 0.28032,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.533316269942838e-05,
+      "loss": 0.8501,
+      "step": 1095
+    },
+    {
+      "epoch": 0.280576,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.5328896851804455e-05,
+      "loss": 0.8042,
+      "step": 1096
+    },
+    {
+      "epoch": 0.280832,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.5324631004180536e-05,
+      "loss": 0.8271,
+      "step": 1097
+    },
+    {
+      "epoch": 0.281088,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.532036515655661e-05,
+      "loss": 0.8089,
+      "step": 1098
+    },
+    {
+      "epoch": 0.281344,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.5316099308932684e-05,
+      "loss": 0.7992,
+      "step": 1099
+    },
+    {
+      "epoch": 0.2816,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.5311833461308765e-05,
+      "loss": 0.8397,
+      "step": 1100
+    },
+    {
+      "epoch": 0.281856,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.530756761368484e-05,
+      "loss": 0.7935,
+      "step": 1101
+    },
+    {
+      "epoch": 0.282112,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.530330176606092e-05,
+      "loss": 0.8092,
+      "step": 1102
+    },
+    {
+      "epoch": 0.282368,
+      "grad_norm": 2.625,
+      "learning_rate": 4.5299035918436995e-05,
+      "loss": 0.8255,
+      "step": 1103
+    },
+    {
+      "epoch": 0.282624,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.5294770070813076e-05,
+      "loss": 0.8117,
+      "step": 1104
+    },
+    {
+      "epoch": 0.28288,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.529050422318915e-05,
+      "loss": 0.7812,
+      "step": 1105
+    },
+    {
+      "epoch": 0.283136,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.5286238375565224e-05,
+      "loss": 0.8643,
+      "step": 1106
+    },
+    {
+      "epoch": 0.283392,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.5281972527941305e-05,
+      "loss": 0.8064,
+      "step": 1107
+    },
+    {
+      "epoch": 0.283648,
+      "grad_norm": 2.75,
+      "learning_rate": 4.5277706680317386e-05,
+      "loss": 0.8675,
+      "step": 1108
+    },
+    {
+      "epoch": 0.283904,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.527344083269346e-05,
+      "loss": 0.7941,
+      "step": 1109
+    },
+    {
+      "epoch": 0.28416,
+      "grad_norm": 2.5,
+      "learning_rate": 4.5269174985069535e-05,
+      "loss": 0.7958,
+      "step": 1110
+    },
+    {
+      "epoch": 0.284416,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.526490913744561e-05,
+      "loss": 0.8118,
+      "step": 1111
+    },
+    {
+      "epoch": 0.284672,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.526064328982169e-05,
+      "loss": 0.831,
+      "step": 1112
+    },
+    {
+      "epoch": 0.284928,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.5256377442197765e-05,
+      "loss": 0.8063,
+      "step": 1113
+    },
+    {
+      "epoch": 0.285184,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.5252111594573846e-05,
+      "loss": 0.7885,
+      "step": 1114
+    },
+    {
+      "epoch": 0.28544,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.524784574694992e-05,
+      "loss": 0.8095,
+      "step": 1115
+    },
+    {
+      "epoch": 0.285696,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.5243579899326e-05,
+      "loss": 0.8346,
+      "step": 1116
+    },
+    {
+      "epoch": 0.285952,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.5239314051702075e-05,
+      "loss": 0.7993,
+      "step": 1117
+    },
+    {
+      "epoch": 0.286208,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.523504820407815e-05,
+      "loss": 0.7902,
+      "step": 1118
+    },
+    {
+      "epoch": 0.286464,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.523078235645423e-05,
+      "loss": 0.8257,
+      "step": 1119
+    },
+    {
+      "epoch": 0.28672,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.5226516508830305e-05,
+      "loss": 0.8014,
+      "step": 1120
+    },
+    {
+      "epoch": 0.286976,
+      "grad_norm": 2.75,
+      "learning_rate": 4.5222250661206386e-05,
+      "loss": 0.8312,
+      "step": 1121
+    },
+    {
+      "epoch": 0.287232,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.521798481358246e-05,
+      "loss": 0.8522,
+      "step": 1122
+    },
+    {
+      "epoch": 0.287488,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.521371896595854e-05,
+      "loss": 0.8137,
+      "step": 1123
+    },
+    {
+      "epoch": 0.287744,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.5209453118334615e-05,
+      "loss": 0.7556,
+      "step": 1124
+    },
+    {
+      "epoch": 0.288,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.520518727071069e-05,
+      "loss": 0.7643,
+      "step": 1125
+    },
+    {
+      "epoch": 0.288256,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.5200921423086764e-05,
+      "loss": 0.8147,
+      "step": 1126
+    },
+    {
+      "epoch": 0.288512,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.519665557546285e-05,
+      "loss": 0.8248,
+      "step": 1127
+    },
+    {
+      "epoch": 0.288768,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.5192389727838926e-05,
+      "loss": 0.8276,
+      "step": 1128
+    },
+    {
+      "epoch": 0.289024,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.5188123880215e-05,
+      "loss": 0.7618,
+      "step": 1129
+    },
+    {
+      "epoch": 0.28928,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.5183858032591075e-05,
+      "loss": 0.8175,
+      "step": 1130
+    },
+    {
+      "epoch": 0.289536,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.5179592184967156e-05,
+      "loss": 0.8162,
+      "step": 1131
+    },
+    {
+      "epoch": 0.289792,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.517532633734323e-05,
+      "loss": 0.7685,
+      "step": 1132
+    },
+    {
+      "epoch": 0.290048,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.517106048971931e-05,
+      "loss": 0.7879,
+      "step": 1133
+    },
+    {
+      "epoch": 0.290304,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.516679464209539e-05,
+      "loss": 0.7564,
+      "step": 1134
+    },
+    {
+      "epoch": 0.29056,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.5162528794471466e-05,
+      "loss": 0.7658,
+      "step": 1135
+    },
+    {
+      "epoch": 0.290816,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.515826294684754e-05,
+      "loss": 0.8071,
+      "step": 1136
+    },
+    {
+      "epoch": 0.291072,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.5153997099223615e-05,
+      "loss": 0.8038,
+      "step": 1137
+    },
+    {
+      "epoch": 0.291328,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.5149731251599696e-05,
+      "loss": 0.7565,
+      "step": 1138
+    },
+    {
+      "epoch": 0.291584,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.514546540397577e-05,
+      "loss": 0.809,
+      "step": 1139
+    },
+    {
+      "epoch": 0.29184,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.514119955635185e-05,
+      "loss": 0.7834,
+      "step": 1140
+    },
+    {
+      "epoch": 0.292096,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.5136933708727925e-05,
+      "loss": 0.8345,
+      "step": 1141
+    },
+    {
+      "epoch": 0.292352,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.5132667861104006e-05,
+      "loss": 0.7937,
+      "step": 1142
+    },
+    {
+      "epoch": 0.292608,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.512840201348008e-05,
+      "loss": 0.8117,
+      "step": 1143
+    },
+    {
+      "epoch": 0.292864,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.5124136165856155e-05,
+      "loss": 0.8303,
+      "step": 1144
+    },
+    {
+      "epoch": 0.29312,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.511987031823223e-05,
+      "loss": 0.7534,
+      "step": 1145
+    },
+    {
+      "epoch": 0.293376,
+      "grad_norm": 2.625,
+      "learning_rate": 4.511560447060831e-05,
+      "loss": 0.78,
+      "step": 1146
+    },
+    {
+      "epoch": 0.293632,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.511133862298439e-05,
+      "loss": 0.7983,
+      "step": 1147
+    },
+    {
+      "epoch": 0.293888,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.5107072775360466e-05,
+      "loss": 0.7987,
+      "step": 1148
+    },
+    {
+      "epoch": 0.294144,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.5102806927736547e-05,
+      "loss": 0.7651,
+      "step": 1149
+    },
+    {
+      "epoch": 0.2944,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.509854108011262e-05,
+      "loss": 0.819,
+      "step": 1150
+    },
+    {
+      "epoch": 0.294656,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.5094275232488695e-05,
+      "loss": 0.7819,
+      "step": 1151
+    },
+    {
+      "epoch": 0.294912,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.5090009384864776e-05,
+      "loss": 0.7781,
+      "step": 1152
+    },
+    {
+      "epoch": 0.295168,
+      "grad_norm": 2.875,
+      "learning_rate": 4.508574353724086e-05,
+      "loss": 0.8109,
+      "step": 1153
+    },
+    {
+      "epoch": 0.295424,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.508147768961693e-05,
+      "loss": 0.8212,
+      "step": 1154
+    },
+    {
+      "epoch": 0.29568,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.5077211841993006e-05,
+      "loss": 0.8135,
+      "step": 1155
+    },
+    {
+      "epoch": 0.295936,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.507294599436908e-05,
+      "loss": 0.7826,
+      "step": 1156
+    },
+    {
+      "epoch": 0.296192,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.506868014674516e-05,
+      "loss": 0.7762,
+      "step": 1157
+    },
+    {
+      "epoch": 0.296448,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.5064414299121235e-05,
+      "loss": 0.7764,
+      "step": 1158
+    },
+    {
+      "epoch": 0.296704,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.5060148451497316e-05,
+      "loss": 0.8111,
+      "step": 1159
+    },
+    {
+      "epoch": 0.29696,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.505588260387339e-05,
+      "loss": 0.8042,
+      "step": 1160
+    },
+    {
+      "epoch": 0.297216,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.505161675624947e-05,
+      "loss": 0.7658,
+      "step": 1161
+    },
+    {
+      "epoch": 0.297472,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.5047350908625546e-05,
+      "loss": 0.7637,
+      "step": 1162
+    },
+    {
+      "epoch": 0.297728,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.504308506100162e-05,
+      "loss": 0.7803,
+      "step": 1163
+    },
+    {
+      "epoch": 0.297984,
+      "grad_norm": 2.5,
+      "learning_rate": 4.5038819213377694e-05,
+      "loss": 0.7908,
+      "step": 1164
+    },
+    {
+      "epoch": 0.29824,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.5034553365753776e-05,
+      "loss": 0.8145,
+      "step": 1165
+    },
+    {
+      "epoch": 0.298496,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.5030287518129857e-05,
+      "loss": 0.739,
+      "step": 1166
+    },
+    {
+      "epoch": 0.298752,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.502602167050593e-05,
+      "loss": 0.7992,
+      "step": 1167
+    },
+    {
+      "epoch": 0.299008,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.502175582288201e-05,
+      "loss": 0.7755,
+      "step": 1168
+    },
+    {
+      "epoch": 0.299264,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.5017489975258086e-05,
+      "loss": 0.7774,
+      "step": 1169
+    },
+    {
+      "epoch": 0.29952,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.501322412763416e-05,
+      "loss": 0.8134,
+      "step": 1170
+    },
+    {
+      "epoch": 0.299776,
+      "grad_norm": 2.625,
+      "learning_rate": 4.500895828001024e-05,
+      "loss": 0.7781,
+      "step": 1171
+    },
+    {
+      "epoch": 0.300032,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.500469243238632e-05,
+      "loss": 0.8036,
+      "step": 1172
+    },
+    {
+      "epoch": 0.300288,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.50004265847624e-05,
+      "loss": 0.7942,
+      "step": 1173
+    },
+    {
+      "epoch": 0.300544,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.499616073713847e-05,
+      "loss": 0.7831,
+      "step": 1174
+    },
+    {
+      "epoch": 0.3008,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.4991894889514545e-05,
+      "loss": 0.783,
+      "step": 1175
+    },
+    {
+      "epoch": 0.301056,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.4987629041890626e-05,
+      "loss": 0.7434,
+      "step": 1176
+    },
+    {
+      "epoch": 0.301312,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.49833631942667e-05,
+      "loss": 0.7529,
+      "step": 1177
+    },
+    {
+      "epoch": 0.301568,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.497909734664278e-05,
+      "loss": 0.7916,
+      "step": 1178
+    },
+    {
+      "epoch": 0.301824,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.4974831499018856e-05,
+      "loss": 0.7802,
+      "step": 1179
+    },
+    {
+      "epoch": 0.30208,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.497056565139494e-05,
+      "loss": 0.8156,
+      "step": 1180
+    },
+    {
+      "epoch": 0.302336,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.496629980377101e-05,
+      "loss": 0.7526,
+      "step": 1181
+    },
+    {
+      "epoch": 0.302592,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.4962033956147085e-05,
+      "loss": 0.7666,
+      "step": 1182
+    },
+    {
+      "epoch": 0.302848,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.4957768108523167e-05,
+      "loss": 0.7779,
+      "step": 1183
+    },
+    {
+      "epoch": 0.303104,
+      "grad_norm": 2.75,
+      "learning_rate": 4.495350226089924e-05,
+      "loss": 0.7864,
+      "step": 1184
+    },
+    {
+      "epoch": 0.30336,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.494923641327532e-05,
+      "loss": 0.7937,
+      "step": 1185
+    },
+    {
+      "epoch": 0.303616,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.4944970565651396e-05,
+      "loss": 0.7577,
+      "step": 1186
+    },
+    {
+      "epoch": 0.303872,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.494070471802748e-05,
+      "loss": 0.7519,
+      "step": 1187
+    },
+    {
+      "epoch": 0.304128,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.493643887040355e-05,
+      "loss": 0.8264,
+      "step": 1188
+    },
+    {
+      "epoch": 0.304384,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.4932173022779626e-05,
+      "loss": 0.7985,
+      "step": 1189
+    },
+    {
+      "epoch": 0.30464,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.492790717515571e-05,
+      "loss": 0.7868,
+      "step": 1190
+    },
+    {
+      "epoch": 0.304896,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.492364132753179e-05,
+      "loss": 0.7843,
+      "step": 1191
+    },
+    {
+      "epoch": 0.305152,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.491937547990786e-05,
+      "loss": 0.8093,
+      "step": 1192
+    },
+    {
+      "epoch": 0.305408,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.4915109632283936e-05,
+      "loss": 0.7585,
+      "step": 1193
+    },
+    {
+      "epoch": 0.305664,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.491084378466001e-05,
+      "loss": 0.7964,
+      "step": 1194
+    },
+    {
+      "epoch": 0.30592,
+      "grad_norm": 2.875,
+      "learning_rate": 4.490657793703609e-05,
+      "loss": 0.7582,
+      "step": 1195
+    },
+    {
+      "epoch": 0.306176,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.4902312089412166e-05,
+      "loss": 0.7972,
+      "step": 1196
+    },
+    {
+      "epoch": 0.306432,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.489804624178825e-05,
+      "loss": 0.8068,
+      "step": 1197
+    },
+    {
+      "epoch": 0.306688,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.489378039416433e-05,
+      "loss": 0.7922,
+      "step": 1198
+    },
+    {
+      "epoch": 0.306944,
+      "grad_norm": 3.25,
+      "learning_rate": 4.48895145465404e-05,
+      "loss": 0.7744,
+      "step": 1199
+    },
+    {
+      "epoch": 0.3072,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.4885248698916476e-05,
+      "loss": 0.782,
+      "step": 1200
+    },
+    {
+      "epoch": 0.307456,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.488098285129255e-05,
+      "loss": 0.7731,
+      "step": 1201
+    },
+    {
+      "epoch": 0.307712,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.487671700366863e-05,
+      "loss": 0.7652,
+      "step": 1202
+    },
+    {
+      "epoch": 0.307968,
+      "grad_norm": 2.75,
+      "learning_rate": 4.4872451156044706e-05,
+      "loss": 0.7526,
+      "step": 1203
+    },
+    {
+      "epoch": 0.308224,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.486818530842079e-05,
+      "loss": 0.798,
+      "step": 1204
+    },
+    {
+      "epoch": 0.30848,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.486391946079686e-05,
+      "loss": 0.7915,
+      "step": 1205
+    },
+    {
+      "epoch": 0.308736,
+      "grad_norm": 2.875,
+      "learning_rate": 4.485965361317294e-05,
+      "loss": 0.8119,
+      "step": 1206
+    },
+    {
+      "epoch": 0.308992,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.485538776554902e-05,
+      "loss": 0.8037,
+      "step": 1207
+    },
+    {
+      "epoch": 0.309248,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.485112191792509e-05,
+      "loss": 0.8007,
+      "step": 1208
+    },
+    {
+      "epoch": 0.309504,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.484685607030117e-05,
+      "loss": 0.7698,
+      "step": 1209
+    },
+    {
+      "epoch": 0.30976,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.484259022267725e-05,
+      "loss": 0.8264,
+      "step": 1210
+    },
+    {
+      "epoch": 0.310016,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.483832437505333e-05,
+      "loss": 0.766,
+      "step": 1211
+    },
+    {
+      "epoch": 0.310272,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.48340585274294e-05,
+      "loss": 0.7625,
+      "step": 1212
+    },
+    {
+      "epoch": 0.310528,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.4829792679805476e-05,
+      "loss": 0.7748,
+      "step": 1213
+    },
+    {
+      "epoch": 0.310784,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.482552683218156e-05,
+      "loss": 0.7963,
+      "step": 1214
+    },
+    {
+      "epoch": 0.31104,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.482126098455763e-05,
+      "loss": 0.8146,
+      "step": 1215
+    },
+    {
+      "epoch": 0.311296,
+      "grad_norm": 2.625,
+      "learning_rate": 4.481699513693371e-05,
+      "loss": 0.7709,
+      "step": 1216
+    },
+    {
+      "epoch": 0.311552,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.481272928930979e-05,
+      "loss": 0.7521,
+      "step": 1217
+    },
+    {
+      "epoch": 0.311808,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.480846344168587e-05,
+      "loss": 0.7296,
+      "step": 1218
+    },
+    {
+      "epoch": 0.312064,
+      "grad_norm": 2.75,
+      "learning_rate": 4.480419759406194e-05,
+      "loss": 0.7567,
+      "step": 1219
+    },
+    {
+      "epoch": 0.31232,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.4799931746438016e-05,
+      "loss": 0.7689,
+      "step": 1220
+    },
+    {
+      "epoch": 0.312576,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.47956658988141e-05,
+      "loss": 0.7995,
+      "step": 1221
+    },
+    {
+      "epoch": 0.312832,
+      "grad_norm": 3.0,
+      "learning_rate": 4.479140005119017e-05,
+      "loss": 0.7726,
+      "step": 1222
+    },
+    {
+      "epoch": 0.313088,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.478713420356625e-05,
+      "loss": 0.791,
+      "step": 1223
+    },
+    {
+      "epoch": 0.313344,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.478286835594233e-05,
+      "loss": 0.7585,
+      "step": 1224
+    },
+    {
+      "epoch": 0.3136,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.477860250831841e-05,
+      "loss": 0.7318,
+      "step": 1225
+    },
+    {
+      "epoch": 0.313856,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.477433666069448e-05,
+      "loss": 0.7569,
+      "step": 1226
+    },
+    {
+      "epoch": 0.314112,
+      "grad_norm": 2.875,
+      "learning_rate": 4.4770070813070556e-05,
+      "loss": 0.7694,
+      "step": 1227
+    },
+    {
+      "epoch": 0.314368,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.476580496544663e-05,
+      "loss": 0.7652,
+      "step": 1228
+    },
+    {
+      "epoch": 0.314624,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.476153911782272e-05,
+      "loss": 0.7894,
+      "step": 1229
+    },
+    {
+      "epoch": 0.31488,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.475727327019879e-05,
+      "loss": 0.7435,
+      "step": 1230
+    },
+    {
+      "epoch": 0.315136,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.475300742257487e-05,
+      "loss": 0.7579,
+      "step": 1231
+    },
+    {
+      "epoch": 0.315392,
+      "grad_norm": 3.0,
+      "learning_rate": 4.474874157495095e-05,
+      "loss": 0.7762,
+      "step": 1232
+    },
+    {
+      "epoch": 0.315648,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.474447572732702e-05,
+      "loss": 0.7172,
+      "step": 1233
+    },
+    {
+      "epoch": 0.315904,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.4740209879703096e-05,
+      "loss": 0.7579,
+      "step": 1234
+    },
+    {
+      "epoch": 0.31616,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.473594403207918e-05,
+      "loss": 0.7967,
+      "step": 1235
+    },
+    {
+      "epoch": 0.316416,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.473167818445526e-05,
+      "loss": 0.7317,
+      "step": 1236
+    },
+    {
+      "epoch": 0.316672,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.472741233683133e-05,
+      "loss": 0.7871,
+      "step": 1237
+    },
+    {
+      "epoch": 0.316928,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.472314648920741e-05,
+      "loss": 0.7584,
+      "step": 1238
+    },
+    {
+      "epoch": 0.317184,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.471888064158348e-05,
+      "loss": 0.7058,
+      "step": 1239
+    },
+    {
+      "epoch": 0.31744,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.471461479395956e-05,
+      "loss": 0.7599,
+      "step": 1240
+    },
+    {
+      "epoch": 0.317696,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.4710348946335637e-05,
+      "loss": 0.716,
+      "step": 1241
+    },
+    {
+      "epoch": 0.317952,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.470608309871172e-05,
+      "loss": 0.7761,
+      "step": 1242
+    },
+    {
+      "epoch": 0.318208,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.470181725108779e-05,
+      "loss": 0.7554,
+      "step": 1243
+    },
+    {
+      "epoch": 0.318464,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.469755140346387e-05,
+      "loss": 0.7571,
+      "step": 1244
+    },
+    {
+      "epoch": 0.31872,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.469328555583995e-05,
+      "loss": 0.7495,
+      "step": 1245
+    },
+    {
+      "epoch": 0.318976,
+      "grad_norm": 2.75,
+      "learning_rate": 4.468901970821602e-05,
+      "loss": 0.7912,
+      "step": 1246
+    },
+    {
+      "epoch": 0.319232,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.46847538605921e-05,
+      "loss": 0.773,
+      "step": 1247
+    },
+    {
+      "epoch": 0.319488,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.4680488012968184e-05,
+      "loss": 0.7521,
+      "step": 1248
+    },
+    {
+      "epoch": 0.319744,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.467622216534426e-05,
+      "loss": 0.7812,
+      "step": 1249
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.467195631772033e-05,
+      "loss": 0.7802,
+      "step": 1250
+    },
+    {
+      "epoch": 0.320256,
+      "grad_norm": 3.0,
+      "learning_rate": 4.466769047009641e-05,
+      "loss": 0.769,
+      "step": 1251
+    },
+    {
+      "epoch": 0.320512,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.466342462247249e-05,
+      "loss": 0.7397,
+      "step": 1252
+    },
+    {
+      "epoch": 0.320768,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.465915877484856e-05,
+      "loss": 0.7564,
+      "step": 1253
+    },
+    {
+      "epoch": 0.321024,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.465489292722464e-05,
+      "loss": 0.7778,
+      "step": 1254
+    },
+    {
+      "epoch": 0.32128,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.4650627079600724e-05,
+      "loss": 0.7266,
+      "step": 1255
+    },
+    {
+      "epoch": 0.321536,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.46463612319768e-05,
+      "loss": 0.7554,
+      "step": 1256
+    },
+    {
+      "epoch": 0.321792,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.464209538435287e-05,
+      "loss": 0.7747,
+      "step": 1257
+    },
+    {
+      "epoch": 0.322048,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.4637829536728947e-05,
+      "loss": 0.8037,
+      "step": 1258
+    },
+    {
+      "epoch": 0.322304,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.463356368910503e-05,
+      "loss": 0.7513,
+      "step": 1259
+    },
+    {
+      "epoch": 0.32256,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.46292978414811e-05,
+      "loss": 0.7528,
+      "step": 1260
+    },
+    {
+      "epoch": 0.322816,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.462503199385718e-05,
+      "loss": 0.78,
+      "step": 1261
+    },
+    {
+      "epoch": 0.323072,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.462076614623326e-05,
+      "loss": 0.7496,
+      "step": 1262
+    },
+    {
+      "epoch": 0.323328,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.461650029860934e-05,
+      "loss": 0.7667,
+      "step": 1263
+    },
+    {
+      "epoch": 0.323584,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.461223445098541e-05,
+      "loss": 0.7539,
+      "step": 1264
+    },
+    {
+      "epoch": 0.32384,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.460796860336149e-05,
+      "loss": 0.7361,
+      "step": 1265
+    },
+    {
+      "epoch": 0.324096,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.460370275573757e-05,
+      "loss": 0.7816,
+      "step": 1266
+    },
+    {
+      "epoch": 0.324352,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.459943690811364e-05,
+      "loss": 0.7395,
+      "step": 1267
+    },
+    {
+      "epoch": 0.324608,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.459517106048972e-05,
+      "loss": 0.7854,
+      "step": 1268
+    },
+    {
+      "epoch": 0.324864,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.45909052128658e-05,
+      "loss": 0.7649,
+      "step": 1269
+    },
+    {
+      "epoch": 0.32512,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.458663936524188e-05,
+      "loss": 0.7341,
+      "step": 1270
+    },
+    {
+      "epoch": 0.325376,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.458237351761795e-05,
+      "loss": 0.7988,
+      "step": 1271
+    },
+    {
+      "epoch": 0.325632,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.457810766999403e-05,
+      "loss": 0.7385,
+      "step": 1272
+    },
+    {
+      "epoch": 0.325888,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.457384182237011e-05,
+      "loss": 0.7625,
+      "step": 1273
+    },
+    {
+      "epoch": 0.326144,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.456957597474619e-05,
+      "loss": 0.7715,
+      "step": 1274
+    },
+    {
+      "epoch": 0.3264,
+      "grad_norm": 3.125,
+      "learning_rate": 4.456531012712226e-05,
+      "loss": 0.7755,
+      "step": 1275
+    },
+    {
+      "epoch": 0.326656,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.456104427949834e-05,
+      "loss": 0.761,
+      "step": 1276
+    },
+    {
+      "epoch": 0.326912,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.455677843187441e-05,
+      "loss": 0.7674,
+      "step": 1277
+    },
+    {
+      "epoch": 0.327168,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.455251258425049e-05,
+      "loss": 0.7695,
+      "step": 1278
+    },
+    {
+      "epoch": 0.327424,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.454824673662657e-05,
+      "loss": 0.7739,
+      "step": 1279
+    },
+    {
+      "epoch": 0.32768,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.454398088900265e-05,
+      "loss": 0.7745,
+      "step": 1280
+    },
+    {
+      "epoch": 0.327936,
+      "grad_norm": 2.625,
+      "learning_rate": 4.453971504137873e-05,
+      "loss": 0.7611,
+      "step": 1281
+    },
+    {
+      "epoch": 0.328192,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.4535449193754803e-05,
+      "loss": 0.7827,
+      "step": 1282
+    },
+    {
+      "epoch": 0.328448,
+      "grad_norm": 3.0,
+      "learning_rate": 4.453118334613088e-05,
+      "loss": 0.7564,
+      "step": 1283
+    },
+    {
+      "epoch": 0.328704,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.452691749850695e-05,
+      "loss": 0.7324,
+      "step": 1284
+    },
+    {
+      "epoch": 0.32896,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.452265165088303e-05,
+      "loss": 0.7941,
+      "step": 1285
+    },
+    {
+      "epoch": 0.329216,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.451838580325911e-05,
+      "loss": 0.7366,
+      "step": 1286
+    },
+    {
+      "epoch": 0.329472,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.451411995563519e-05,
+      "loss": 0.7273,
+      "step": 1287
+    },
+    {
+      "epoch": 0.329728,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.450985410801126e-05,
+      "loss": 0.7764,
+      "step": 1288
+    },
+    {
+      "epoch": 0.329984,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.4505588260387344e-05,
+      "loss": 0.7357,
+      "step": 1289
+    },
+    {
+      "epoch": 0.33024,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.450132241276342e-05,
+      "loss": 0.7382,
+      "step": 1290
+    },
+    {
+      "epoch": 0.330496,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.449705656513949e-05,
+      "loss": 0.6898,
+      "step": 1291
+    },
+    {
+      "epoch": 0.330752,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.449279071751557e-05,
+      "loss": 0.7602,
+      "step": 1292
+    },
+    {
+      "epoch": 0.331008,
+      "grad_norm": 2.625,
+      "learning_rate": 4.4488524869891654e-05,
+      "loss": 0.7507,
+      "step": 1293
+    },
+    {
+      "epoch": 0.331264,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.448425902226773e-05,
+      "loss": 0.7626,
+      "step": 1294
+    },
+    {
+      "epoch": 0.33152,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.44799931746438e-05,
+      "loss": 0.7591,
+      "step": 1295
+    },
+    {
+      "epoch": 0.331776,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.447572732701988e-05,
+      "loss": 0.7912,
+      "step": 1296
+    },
+    {
+      "epoch": 0.332032,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.447146147939596e-05,
+      "loss": 0.7366,
+      "step": 1297
+    },
+    {
+      "epoch": 0.332288,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.446719563177203e-05,
+      "loss": 0.7182,
+      "step": 1298
+    },
+    {
+      "epoch": 0.332544,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.4462929784148113e-05,
+      "loss": 0.7713,
+      "step": 1299
+    },
+    {
+      "epoch": 0.3328,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.4458663936524195e-05,
+      "loss": 0.7403,
+      "step": 1300
+    },
+    {
+      "epoch": 0.333056,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.445439808890027e-05,
+      "loss": 0.7832,
+      "step": 1301
+    },
+    {
+      "epoch": 0.333312,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.445013224127634e-05,
+      "loss": 0.7712,
+      "step": 1302
+    },
+    {
+      "epoch": 0.333568,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.444586639365242e-05,
+      "loss": 0.7555,
+      "step": 1303
+    },
+    {
+      "epoch": 0.333824,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.44416005460285e-05,
+      "loss": 0.7243,
+      "step": 1304
+    },
+    {
+      "epoch": 0.33408,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.443733469840457e-05,
+      "loss": 0.7555,
+      "step": 1305
+    },
+    {
+      "epoch": 0.334336,
+      "grad_norm": 2.875,
+      "learning_rate": 4.4433068850780654e-05,
+      "loss": 0.7878,
+      "step": 1306
+    },
+    {
+      "epoch": 0.334592,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.442880300315673e-05,
+      "loss": 0.7824,
+      "step": 1307
+    },
+    {
+      "epoch": 0.334848,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.442453715553281e-05,
+      "loss": 0.7251,
+      "step": 1308
+    },
+    {
+      "epoch": 0.335104,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.442027130790888e-05,
+      "loss": 0.7572,
+      "step": 1309
+    },
+    {
+      "epoch": 0.33536,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.441600546028496e-05,
+      "loss": 0.7666,
+      "step": 1310
+    },
+    {
+      "epoch": 0.335616,
+      "grad_norm": 2.5,
+      "learning_rate": 4.441173961266104e-05,
+      "loss": 0.7847,
+      "step": 1311
+    },
+    {
+      "epoch": 0.335872,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.440747376503712e-05,
+      "loss": 0.7296,
+      "step": 1312
+    },
+    {
+      "epoch": 0.336128,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.4403207917413194e-05,
+      "loss": 0.7597,
+      "step": 1313
+    },
+    {
+      "epoch": 0.336384,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.439894206978927e-05,
+      "loss": 0.7571,
+      "step": 1314
+    },
+    {
+      "epoch": 0.33664,
+      "grad_norm": 2.5,
+      "learning_rate": 4.439467622216535e-05,
+      "loss": 0.7494,
+      "step": 1315
+    },
+    {
+      "epoch": 0.336896,
+      "grad_norm": 2.5,
+      "learning_rate": 4.4390410374541423e-05,
+      "loss": 0.7551,
+      "step": 1316
+    },
+    {
+      "epoch": 0.337152,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.43861445269175e-05,
+      "loss": 0.7201,
+      "step": 1317
+    },
+    {
+      "epoch": 0.337408,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.438187867929358e-05,
+      "loss": 0.7692,
+      "step": 1318
+    },
+    {
+      "epoch": 0.337664,
+      "grad_norm": 2.5,
+      "learning_rate": 4.437761283166966e-05,
+      "loss": 0.725,
+      "step": 1319
+    },
+    {
+      "epoch": 0.33792,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.4373346984045734e-05,
+      "loss": 0.7193,
+      "step": 1320
+    },
+    {
+      "epoch": 0.338176,
+      "grad_norm": 2.875,
+      "learning_rate": 4.436908113642181e-05,
+      "loss": 0.7611,
+      "step": 1321
+    },
+    {
+      "epoch": 0.338432,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.436481528879788e-05,
+      "loss": 0.765,
+      "step": 1322
+    },
+    {
+      "epoch": 0.338688,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.4360549441173964e-05,
+      "loss": 0.6874,
+      "step": 1323
+    },
+    {
+      "epoch": 0.338944,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.435628359355004e-05,
+      "loss": 0.7775,
+      "step": 1324
+    },
+    {
+      "epoch": 0.3392,
+      "grad_norm": 2.75,
+      "learning_rate": 4.435201774592612e-05,
+      "loss": 0.8337,
+      "step": 1325
+    },
+    {
+      "epoch": 0.339456,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.434775189830219e-05,
+      "loss": 0.7324,
+      "step": 1326
+    },
+    {
+      "epoch": 0.339712,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.4343486050678274e-05,
+      "loss": 0.714,
+      "step": 1327
+    },
+    {
+      "epoch": 0.339968,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.433922020305435e-05,
+      "loss": 0.78,
+      "step": 1328
+    },
+    {
+      "epoch": 0.340224,
+      "grad_norm": 2.625,
+      "learning_rate": 4.433495435543042e-05,
+      "loss": 0.7633,
+      "step": 1329
+    },
+    {
+      "epoch": 0.34048,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.4330688507806504e-05,
+      "loss": 0.7647,
+      "step": 1330
+    },
+    {
+      "epoch": 0.340736,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.4326422660182585e-05,
+      "loss": 0.7499,
+      "step": 1331
+    },
+    {
+      "epoch": 0.340992,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.432215681255866e-05,
+      "loss": 0.742,
+      "step": 1332
+    },
+    {
+      "epoch": 0.341248,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.431789096493473e-05,
+      "loss": 0.7037,
+      "step": 1333
+    },
+    {
+      "epoch": 0.341504,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.4313625117310814e-05,
+      "loss": 0.7,
+      "step": 1334
+    },
+    {
+      "epoch": 0.34176,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.430935926968689e-05,
+      "loss": 0.766,
+      "step": 1335
+    },
+    {
+      "epoch": 0.342016,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.430509342206296e-05,
+      "loss": 0.739,
+      "step": 1336
+    },
+    {
+      "epoch": 0.342272,
+      "grad_norm": 2.75,
+      "learning_rate": 4.4300827574439044e-05,
+      "loss": 0.7634,
+      "step": 1337
+    },
+    {
+      "epoch": 0.342528,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.4296561726815125e-05,
+      "loss": 0.7449,
+      "step": 1338
+    },
+    {
+      "epoch": 0.342784,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.42922958791912e-05,
+      "loss": 0.7351,
+      "step": 1339
+    },
+    {
+      "epoch": 0.34304,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.4288030031567274e-05,
+      "loss": 0.7351,
+      "step": 1340
+    },
+    {
+      "epoch": 0.343296,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.428376418394335e-05,
+      "loss": 0.7678,
+      "step": 1341
+    },
+    {
+      "epoch": 0.343552,
+      "grad_norm": 2.75,
+      "learning_rate": 4.427949833631943e-05,
+      "loss": 0.7384,
+      "step": 1342
+    },
+    {
+      "epoch": 0.343808,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.42752324886955e-05,
+      "loss": 0.7344,
+      "step": 1343
+    },
+    {
+      "epoch": 0.344064,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.4270966641071584e-05,
+      "loss": 0.7815,
+      "step": 1344
+    },
+    {
+      "epoch": 0.34432,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.426670079344766e-05,
+      "loss": 0.7102,
+      "step": 1345
+    },
+    {
+      "epoch": 0.344576,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.426243494582374e-05,
+      "loss": 0.6879,
+      "step": 1346
+    },
+    {
+      "epoch": 0.344832,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.4258169098199814e-05,
+      "loss": 0.7223,
+      "step": 1347
+    },
+    {
+      "epoch": 0.345088,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.425390325057589e-05,
+      "loss": 0.7489,
+      "step": 1348
+    },
+    {
+      "epoch": 0.345344,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.424963740295197e-05,
+      "loss": 0.6848,
+      "step": 1349
+    },
+    {
+      "epoch": 0.3456,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.424537155532805e-05,
+      "loss": 0.7441,
+      "step": 1350
+    },
+    {
+      "epoch": 0.345856,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.4241105707704124e-05,
+      "loss": 0.7328,
+      "step": 1351
+    },
+    {
+      "epoch": 0.346112,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.42368398600802e-05,
+      "loss": 0.7452,
+      "step": 1352
+    },
+    {
+      "epoch": 0.346368,
+      "grad_norm": 2.75,
+      "learning_rate": 4.423257401245628e-05,
+      "loss": 0.761,
+      "step": 1353
+    },
+    {
+      "epoch": 0.346624,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.4228308164832354e-05,
+      "loss": 0.7479,
+      "step": 1354
+    },
+    {
+      "epoch": 0.34688,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.422404231720843e-05,
+      "loss": 0.7443,
+      "step": 1355
+    },
+    {
+      "epoch": 0.347136,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.421977646958451e-05,
+      "loss": 0.7635,
+      "step": 1356
+    },
+    {
+      "epoch": 0.347392,
+      "grad_norm": 2.5,
+      "learning_rate": 4.421551062196059e-05,
+      "loss": 0.7619,
+      "step": 1357
+    },
+    {
+      "epoch": 0.347648,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.4211244774336665e-05,
+      "loss": 0.758,
+      "step": 1358
+    },
+    {
+      "epoch": 0.347904,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.420697892671274e-05,
+      "loss": 0.7369,
+      "step": 1359
+    },
+    {
+      "epoch": 0.34816,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.420271307908881e-05,
+      "loss": 0.7649,
+      "step": 1360
+    },
+    {
+      "epoch": 0.348416,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.4198447231464894e-05,
+      "loss": 0.7169,
+      "step": 1361
+    },
+    {
+      "epoch": 0.348672,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.419418138384097e-05,
+      "loss": 0.7654,
+      "step": 1362
+    },
+    {
+      "epoch": 0.348928,
+      "grad_norm": 2.625,
+      "learning_rate": 4.418991553621705e-05,
+      "loss": 0.7959,
+      "step": 1363
+    },
+    {
+      "epoch": 0.349184,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.418564968859313e-05,
+      "loss": 0.7101,
+      "step": 1364
+    },
+    {
+      "epoch": 0.34944,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.4181383840969205e-05,
+      "loss": 0.753,
+      "step": 1365
+    },
+    {
+      "epoch": 0.349696,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.417711799334528e-05,
+      "loss": 0.7101,
+      "step": 1366
+    },
+    {
+      "epoch": 0.349952,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.417285214572135e-05,
+      "loss": 0.7377,
+      "step": 1367
+    },
+    {
+      "epoch": 0.350208,
+      "grad_norm": 2.875,
+      "learning_rate": 4.4168586298097434e-05,
+      "loss": 0.7182,
+      "step": 1368
+    },
+    {
+      "epoch": 0.350464,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.416432045047351e-05,
+      "loss": 0.7057,
+      "step": 1369
+    },
+    {
+      "epoch": 0.35072,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.416005460284959e-05,
+      "loss": 0.7937,
+      "step": 1370
+    },
+    {
+      "epoch": 0.350976,
+      "grad_norm": 2.5,
+      "learning_rate": 4.4155788755225664e-05,
+      "loss": 0.7224,
+      "step": 1371
+    },
+    {
+      "epoch": 0.351232,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.4151522907601745e-05,
+      "loss": 0.706,
+      "step": 1372
+    },
+    {
+      "epoch": 0.351488,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.414725705997782e-05,
+      "loss": 0.7347,
+      "step": 1373
+    },
+    {
+      "epoch": 0.351744,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.4142991212353893e-05,
+      "loss": 0.7248,
+      "step": 1374
+    },
+    {
+      "epoch": 0.352,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.4138725364729975e-05,
+      "loss": 0.7525,
+      "step": 1375
+    },
+    {
+      "epoch": 0.352256,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.4134459517106056e-05,
+      "loss": 0.7782,
+      "step": 1376
+    },
+    {
+      "epoch": 0.352512,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.413019366948213e-05,
+      "loss": 0.7651,
+      "step": 1377
+    },
+    {
+      "epoch": 0.352768,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.4125927821858204e-05,
+      "loss": 0.7412,
+      "step": 1378
+    },
+    {
+      "epoch": 0.353024,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.4121661974234285e-05,
+      "loss": 0.721,
+      "step": 1379
+    },
+    {
+      "epoch": 0.35328,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.411739612661036e-05,
+      "loss": 0.7084,
+      "step": 1380
+    },
+    {
+      "epoch": 0.353536,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.4113130278986434e-05,
+      "loss": 0.7524,
+      "step": 1381
+    },
+    {
+      "epoch": 0.353792,
+      "grad_norm": 2.625,
+      "learning_rate": 4.4108864431362515e-05,
+      "loss": 0.733,
+      "step": 1382
+    },
+    {
+      "epoch": 0.354048,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.4104598583738596e-05,
+      "loss": 0.7392,
+      "step": 1383
+    },
+    {
+      "epoch": 0.354304,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.410033273611467e-05,
+      "loss": 0.7431,
+      "step": 1384
+    },
+    {
+      "epoch": 0.35456,
+      "grad_norm": 2.875,
+      "learning_rate": 4.4096066888490744e-05,
+      "loss": 0.7349,
+      "step": 1385
+    },
+    {
+      "epoch": 0.354816,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.409180104086682e-05,
+      "loss": 0.7701,
+      "step": 1386
+    },
+    {
+      "epoch": 0.355072,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.40875351932429e-05,
+      "loss": 0.7292,
+      "step": 1387
+    },
+    {
+      "epoch": 0.355328,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.4083269345618974e-05,
+      "loss": 0.7494,
+      "step": 1388
+    },
+    {
+      "epoch": 0.355584,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.4079003497995055e-05,
+      "loss": 0.7234,
+      "step": 1389
+    },
+    {
+      "epoch": 0.35584,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.407473765037113e-05,
+      "loss": 0.7233,
+      "step": 1390
+    },
+    {
+      "epoch": 0.356096,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.407047180274721e-05,
+      "loss": 0.7373,
+      "step": 1391
+    },
+    {
+      "epoch": 0.356352,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.4066205955123285e-05,
+      "loss": 0.782,
+      "step": 1392
+    },
+    {
+      "epoch": 0.356608,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.406194010749936e-05,
+      "loss": 0.7088,
+      "step": 1393
+    },
+    {
+      "epoch": 0.356864,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.405767425987544e-05,
+      "loss": 0.7147,
+      "step": 1394
+    },
+    {
+      "epoch": 0.35712,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.405340841225152e-05,
+      "loss": 0.7284,
+      "step": 1395
+    },
+    {
+      "epoch": 0.357376,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.4049142564627595e-05,
+      "loss": 0.7593,
+      "step": 1396
+    },
+    {
+      "epoch": 0.357632,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.404487671700367e-05,
+      "loss": 0.7317,
+      "step": 1397
+    },
+    {
+      "epoch": 0.357888,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.404061086937975e-05,
+      "loss": 0.7471,
+      "step": 1398
+    },
+    {
+      "epoch": 0.358144,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.4036345021755825e-05,
+      "loss": 0.7439,
+      "step": 1399
+    },
+    {
+      "epoch": 0.3584,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.40320791741319e-05,
+      "loss": 0.7315,
+      "step": 1400
+    },
+    {
+      "epoch": 0.358656,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.402781332650798e-05,
+      "loss": 0.7378,
+      "step": 1401
+    },
+    {
+      "epoch": 0.358912,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.402354747888406e-05,
+      "loss": 0.7316,
+      "step": 1402
+    },
+    {
+      "epoch": 0.359168,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.4019281631260135e-05,
+      "loss": 0.7004,
+      "step": 1403
+    },
+    {
+      "epoch": 0.359424,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.401501578363621e-05,
+      "loss": 0.7653,
+      "step": 1404
+    },
+    {
+      "epoch": 0.35968,
+      "grad_norm": 2.25,
+      "learning_rate": 4.4010749936012284e-05,
+      "loss": 0.7179,
+      "step": 1405
+    },
+    {
+      "epoch": 0.359936,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.4006484088388365e-05,
+      "loss": 0.7254,
+      "step": 1406
+    },
+    {
+      "epoch": 0.360192,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.400221824076444e-05,
+      "loss": 0.7393,
+      "step": 1407
+    },
+    {
+      "epoch": 0.360448,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.399795239314052e-05,
+      "loss": 0.749,
+      "step": 1408
+    },
+    {
+      "epoch": 0.360704,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.3993686545516594e-05,
+      "loss": 0.7727,
+      "step": 1409
+    },
+    {
+      "epoch": 0.36096,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.3989420697892676e-05,
+      "loss": 0.6885,
+      "step": 1410
+    },
+    {
+      "epoch": 0.361216,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.398515485026875e-05,
+      "loss": 0.6928,
+      "step": 1411
+    },
+    {
+      "epoch": 0.361472,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.3980889002644824e-05,
+      "loss": 0.7181,
+      "step": 1412
+    },
+    {
+      "epoch": 0.361728,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.3976623155020905e-05,
+      "loss": 0.7296,
+      "step": 1413
+    },
+    {
+      "epoch": 0.361984,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.3972357307396986e-05,
+      "loss": 0.7322,
+      "step": 1414
+    },
+    {
+      "epoch": 0.36224,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.396809145977306e-05,
+      "loss": 0.6991,
+      "step": 1415
+    },
+    {
+      "epoch": 0.362496,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.3963825612149135e-05,
+      "loss": 0.719,
+      "step": 1416
+    },
+    {
+      "epoch": 0.362752,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.3959559764525216e-05,
+      "loss": 0.7441,
+      "step": 1417
+    },
+    {
+      "epoch": 0.363008,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.395529391690129e-05,
+      "loss": 0.7237,
+      "step": 1418
+    },
+    {
+      "epoch": 0.363264,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.3951028069277364e-05,
+      "loss": 0.6989,
+      "step": 1419
+    },
+    {
+      "epoch": 0.36352,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.3946762221653445e-05,
+      "loss": 0.7339,
+      "step": 1420
+    },
+    {
+      "epoch": 0.363776,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.3942496374029526e-05,
+      "loss": 0.7161,
+      "step": 1421
+    },
+    {
+      "epoch": 0.364032,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.39382305264056e-05,
+      "loss": 0.7219,
+      "step": 1422
+    },
+    {
+      "epoch": 0.364288,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.3933964678781675e-05,
+      "loss": 0.7418,
+      "step": 1423
+    },
+    {
+      "epoch": 0.364544,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.392969883115775e-05,
+      "loss": 0.7482,
+      "step": 1424
+    },
+    {
+      "epoch": 0.3648,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.392543298353383e-05,
+      "loss": 0.7427,
+      "step": 1425
+    },
+    {
+      "epoch": 0.365056,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.3921167135909904e-05,
+      "loss": 0.73,
+      "step": 1426
+    },
+    {
+      "epoch": 0.365312,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.3916901288285985e-05,
+      "loss": 0.722,
+      "step": 1427
+    },
+    {
+      "epoch": 0.365568,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.3912635440662067e-05,
+      "loss": 0.7153,
+      "step": 1428
+    },
+    {
+      "epoch": 0.365824,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.390836959303814e-05,
+      "loss": 0.711,
+      "step": 1429
+    },
+    {
+      "epoch": 0.36608,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.3904103745414215e-05,
+      "loss": 0.724,
+      "step": 1430
+    },
+    {
+      "epoch": 0.366336,
+      "grad_norm": 2.875,
+      "learning_rate": 4.389983789779029e-05,
+      "loss": 0.7646,
+      "step": 1431
+    },
+    {
+      "epoch": 0.366592,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.389557205016637e-05,
+      "loss": 0.7068,
+      "step": 1432
+    },
+    {
+      "epoch": 0.366848,
+      "grad_norm": 2.375,
+      "learning_rate": 4.389130620254245e-05,
+      "loss": 0.7501,
+      "step": 1433
+    },
+    {
+      "epoch": 0.367104,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.3887040354918526e-05,
+      "loss": 0.7264,
+      "step": 1434
+    },
+    {
+      "epoch": 0.36736,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.38827745072946e-05,
+      "loss": 0.6941,
+      "step": 1435
+    },
+    {
+      "epoch": 0.367616,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.387850865967068e-05,
+      "loss": 0.7235,
+      "step": 1436
+    },
+    {
+      "epoch": 0.367872,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.3874242812046755e-05,
+      "loss": 0.6939,
+      "step": 1437
+    },
+    {
+      "epoch": 0.368128,
+      "grad_norm": 2.875,
+      "learning_rate": 4.386997696442283e-05,
+      "loss": 0.7425,
+      "step": 1438
+    },
+    {
+      "epoch": 0.368384,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.386571111679891e-05,
+      "loss": 0.6986,
+      "step": 1439
+    },
+    {
+      "epoch": 0.36864,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.386144526917499e-05,
+      "loss": 0.7733,
+      "step": 1440
+    },
+    {
+      "epoch": 0.368896,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.3857179421551066e-05,
+      "loss": 0.6908,
+      "step": 1441
+    },
+    {
+      "epoch": 0.369152,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.385291357392714e-05,
+      "loss": 0.7518,
+      "step": 1442
+    },
+    {
+      "epoch": 0.369408,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.3848647726303214e-05,
+      "loss": 0.6895,
+      "step": 1443
+    },
+    {
+      "epoch": 0.369664,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.3844381878679295e-05,
+      "loss": 0.7039,
+      "step": 1444
+    },
+    {
+      "epoch": 0.36992,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.384011603105537e-05,
+      "loss": 0.716,
+      "step": 1445
+    },
+    {
+      "epoch": 0.370176,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.383585018343145e-05,
+      "loss": 0.6873,
+      "step": 1446
+    },
+    {
+      "epoch": 0.370432,
+      "grad_norm": 2.75,
+      "learning_rate": 4.383158433580753e-05,
+      "loss": 0.6972,
+      "step": 1447
+    },
+    {
+      "epoch": 0.370688,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.3827318488183606e-05,
+      "loss": 0.7302,
+      "step": 1448
+    },
+    {
+      "epoch": 0.370944,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.382305264055968e-05,
+      "loss": 0.7397,
+      "step": 1449
+    },
+    {
+      "epoch": 0.3712,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.3818786792935755e-05,
+      "loss": 0.6869,
+      "step": 1450
+    },
+    {
+      "epoch": 0.371456,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.3814520945311836e-05,
+      "loss": 0.7252,
+      "step": 1451
+    },
+    {
+      "epoch": 0.371712,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.381025509768792e-05,
+      "loss": 0.7391,
+      "step": 1452
+    },
+    {
+      "epoch": 0.371968,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.380598925006399e-05,
+      "loss": 0.7236,
+      "step": 1453
+    },
+    {
+      "epoch": 0.372224,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.3801723402440065e-05,
+      "loss": 0.743,
+      "step": 1454
+    },
+    {
+      "epoch": 0.37248,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.3797457554816146e-05,
+      "loss": 0.7525,
+      "step": 1455
+    },
+    {
+      "epoch": 0.372736,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.379319170719222e-05,
+      "loss": 0.7408,
+      "step": 1456
+    },
+    {
+      "epoch": 0.372992,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.3788925859568295e-05,
+      "loss": 0.7015,
+      "step": 1457
+    },
+    {
+      "epoch": 0.373248,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.3784660011944376e-05,
+      "loss": 0.6763,
+      "step": 1458
+    },
+    {
+      "epoch": 0.373504,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.378039416432046e-05,
+      "loss": 0.7126,
+      "step": 1459
+    },
+    {
+      "epoch": 0.37376,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.377612831669653e-05,
+      "loss": 0.6998,
+      "step": 1460
+    },
+    {
+      "epoch": 0.374016,
+      "grad_norm": 2.75,
+      "learning_rate": 4.3771862469072605e-05,
+      "loss": 0.7113,
+      "step": 1461
+    },
+    {
+      "epoch": 0.374272,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.3767596621448686e-05,
+      "loss": 0.6934,
+      "step": 1462
+    },
+    {
+      "epoch": 0.374528,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.376333077382476e-05,
+      "loss": 0.7347,
+      "step": 1463
+    },
+    {
+      "epoch": 0.374784,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.3759064926200835e-05,
+      "loss": 0.686,
+      "step": 1464
+    },
+    {
+      "epoch": 0.37504,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.3754799078576916e-05,
+      "loss": 0.6733,
+      "step": 1465
+    },
+    {
+      "epoch": 0.375296,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.3750533230953e-05,
+      "loss": 0.7125,
+      "step": 1466
+    },
+    {
+      "epoch": 0.375552,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.374626738332907e-05,
+      "loss": 0.7097,
+      "step": 1467
+    },
+    {
+      "epoch": 0.375808,
+      "grad_norm": 2.5,
+      "learning_rate": 4.3742001535705146e-05,
+      "loss": 0.7072,
+      "step": 1468
+    },
+    {
+      "epoch": 0.376064,
+      "grad_norm": 2.75,
+      "learning_rate": 4.373773568808122e-05,
+      "loss": 0.7305,
+      "step": 1469
+    },
+    {
+      "epoch": 0.37632,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.37334698404573e-05,
+      "loss": 0.7499,
+      "step": 1470
+    },
+    {
+      "epoch": 0.376576,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.3729203992833375e-05,
+      "loss": 0.7101,
+      "step": 1471
+    },
+    {
+      "epoch": 0.376832,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.3724938145209456e-05,
+      "loss": 0.6967,
+      "step": 1472
+    },
+    {
+      "epoch": 0.377088,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.372067229758553e-05,
+      "loss": 0.7119,
+      "step": 1473
+    },
+    {
+      "epoch": 0.377344,
+      "grad_norm": 2.5,
+      "learning_rate": 4.371640644996161e-05,
+      "loss": 0.7216,
+      "step": 1474
+    },
+    {
+      "epoch": 0.3776,
+      "grad_norm": 2.75,
+      "learning_rate": 4.3712140602337686e-05,
+      "loss": 0.7246,
+      "step": 1475
+    },
+    {
+      "epoch": 0.377856,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.370787475471376e-05,
+      "loss": 0.732,
+      "step": 1476
+    },
+    {
+      "epoch": 0.378112,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.370360890708984e-05,
+      "loss": 0.6912,
+      "step": 1477
+    },
+    {
+      "epoch": 0.378368,
+      "grad_norm": 3.0,
+      "learning_rate": 4.369934305946592e-05,
+      "loss": 0.7288,
+      "step": 1478
+    },
+    {
+      "epoch": 0.378624,
+      "grad_norm": 3.125,
+      "learning_rate": 4.3695077211841996e-05,
+      "loss": 0.6926,
+      "step": 1479
+    },
+    {
+      "epoch": 0.37888,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.369081136421807e-05,
+      "loss": 0.6937,
+      "step": 1480
+    },
+    {
+      "epoch": 0.379136,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.368654551659415e-05,
+      "loss": 0.7046,
+      "step": 1481
+    },
+    {
+      "epoch": 0.379392,
+      "grad_norm": 2.875,
+      "learning_rate": 4.3682279668970226e-05,
+      "loss": 0.7332,
+      "step": 1482
+    },
+    {
+      "epoch": 0.379648,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.36780138213463e-05,
+      "loss": 0.676,
+      "step": 1483
+    },
+    {
+      "epoch": 0.379904,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.367374797372238e-05,
+      "loss": 0.7306,
+      "step": 1484
+    },
+    {
+      "epoch": 0.38016,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.366948212609846e-05,
+      "loss": 0.6826,
+      "step": 1485
+    },
+    {
+      "epoch": 0.380416,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.3665216278474537e-05,
+      "loss": 0.7017,
+      "step": 1486
+    },
+    {
+      "epoch": 0.380672,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.366095043085061e-05,
+      "loss": 0.7476,
+      "step": 1487
+    },
+    {
+      "epoch": 0.380928,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.3656684583226685e-05,
+      "loss": 0.67,
+      "step": 1488
+    },
+    {
+      "epoch": 0.381184,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.3652418735602766e-05,
+      "loss": 0.7496,
+      "step": 1489
+    },
+    {
+      "epoch": 0.38144,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.364815288797884e-05,
+      "loss": 0.6717,
+      "step": 1490
+    },
+    {
+      "epoch": 0.381696,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.364388704035492e-05,
+      "loss": 0.6875,
+      "step": 1491
+    },
+    {
+      "epoch": 0.381952,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.3639621192730996e-05,
+      "loss": 0.6955,
+      "step": 1492
+    },
+    {
+      "epoch": 0.382208,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.363535534510708e-05,
+      "loss": 0.689,
+      "step": 1493
+    },
+    {
+      "epoch": 0.382464,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.363108949748315e-05,
+      "loss": 0.6798,
+      "step": 1494
+    },
+    {
+      "epoch": 0.38272,
+      "grad_norm": 2.75,
+      "learning_rate": 4.3626823649859225e-05,
+      "loss": 0.6981,
+      "step": 1495
+    },
+    {
+      "epoch": 0.382976,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.3622557802235306e-05,
+      "loss": 0.749,
+      "step": 1496
+    },
+    {
+      "epoch": 0.383232,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.361829195461139e-05,
+      "loss": 0.6926,
+      "step": 1497
+    },
+    {
+      "epoch": 0.383488,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.361402610698746e-05,
+      "loss": 0.686,
+      "step": 1498
+    },
+    {
+      "epoch": 0.383744,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.3609760259363536e-05,
+      "loss": 0.7196,
+      "step": 1499
+    },
+    {
+      "epoch": 0.384,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.360549441173962e-05,
+      "loss": 0.7208,
+      "step": 1500
+    },
+    {
+      "epoch": 0.384256,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.360122856411569e-05,
+      "loss": 0.6823,
+      "step": 1501
+    },
+    {
+      "epoch": 0.384512,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.3596962716491766e-05,
+      "loss": 0.725,
+      "step": 1502
+    },
+    {
+      "epoch": 0.384768,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.3592696868867847e-05,
+      "loss": 0.7014,
+      "step": 1503
+    },
+    {
+      "epoch": 0.385024,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.358843102124393e-05,
+      "loss": 0.6872,
+      "step": 1504
+    },
+    {
+      "epoch": 0.38528,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.358416517362e-05,
+      "loss": 0.7275,
+      "step": 1505
+    },
+    {
+      "epoch": 0.385536,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.3579899325996076e-05,
+      "loss": 0.7212,
+      "step": 1506
+    },
+    {
+      "epoch": 0.385792,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.357563347837215e-05,
+      "loss": 0.7378,
+      "step": 1507
+    },
+    {
+      "epoch": 0.386048,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.357136763074823e-05,
+      "loss": 0.6993,
+      "step": 1508
+    },
+    {
+      "epoch": 0.386304,
+      "grad_norm": 2.625,
+      "learning_rate": 4.3567101783124306e-05,
+      "loss": 0.6595,
+      "step": 1509
+    },
+    {
+      "epoch": 0.38656,
+      "grad_norm": 3.0,
+      "learning_rate": 4.356283593550039e-05,
+      "loss": 0.7302,
+      "step": 1510
+    },
+    {
+      "epoch": 0.386816,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.355857008787647e-05,
+      "loss": 0.6922,
+      "step": 1511
+    },
+    {
+      "epoch": 0.387072,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.355430424025254e-05,
+      "loss": 0.6858,
+      "step": 1512
+    },
+    {
+      "epoch": 0.387328,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.3550038392628616e-05,
+      "loss": 0.773,
+      "step": 1513
+    },
+    {
+      "epoch": 0.387584,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.354577254500469e-05,
+      "loss": 0.7197,
+      "step": 1514
+    },
+    {
+      "epoch": 0.38784,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.354150669738077e-05,
+      "loss": 0.7421,
+      "step": 1515
+    },
+    {
+      "epoch": 0.388096,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.353724084975685e-05,
+      "loss": 0.687,
+      "step": 1516
+    },
+    {
+      "epoch": 0.388352,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.353297500213293e-05,
+      "loss": 0.7107,
+      "step": 1517
+    },
+    {
+      "epoch": 0.388608,
+      "grad_norm": 2.625,
+      "learning_rate": 4.3528709154509e-05,
+      "loss": 0.6864,
+      "step": 1518
+    },
+    {
+      "epoch": 0.388864,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.352444330688508e-05,
+      "loss": 0.7218,
+      "step": 1519
+    },
+    {
+      "epoch": 0.38912,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.3520177459261157e-05,
+      "loss": 0.6927,
+      "step": 1520
+    },
+    {
+      "epoch": 0.389376,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.351591161163723e-05,
+      "loss": 0.6829,
+      "step": 1521
+    },
+    {
+      "epoch": 0.389632,
+      "grad_norm": 2.75,
+      "learning_rate": 4.351164576401331e-05,
+      "loss": 0.6996,
+      "step": 1522
+    },
+    {
+      "epoch": 0.389888,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.350737991638939e-05,
+      "loss": 0.6763,
+      "step": 1523
+    },
+    {
+      "epoch": 0.390144,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.350311406876547e-05,
+      "loss": 0.7229,
+      "step": 1524
+    },
+    {
+      "epoch": 0.3904,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.349884822114154e-05,
+      "loss": 0.6973,
+      "step": 1525
+    },
+    {
+      "epoch": 0.390656,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.349458237351762e-05,
+      "loss": 0.7336,
+      "step": 1526
+    },
+    {
+      "epoch": 0.390912,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.34903165258937e-05,
+      "loss": 0.6805,
+      "step": 1527
+    },
+    {
+      "epoch": 0.391168,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.348605067826977e-05,
+      "loss": 0.7329,
+      "step": 1528
+    },
+    {
+      "epoch": 0.391424,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.348178483064585e-05,
+      "loss": 0.6951,
+      "step": 1529
+    },
+    {
+      "epoch": 0.39168,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.347751898302193e-05,
+      "loss": 0.7335,
+      "step": 1530
+    },
+    {
+      "epoch": 0.391936,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.347325313539801e-05,
+      "loss": 0.6941,
+      "step": 1531
+    },
+    {
+      "epoch": 0.392192,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.346898728777408e-05,
+      "loss": 0.7124,
+      "step": 1532
+    },
+    {
+      "epoch": 0.392448,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.3464721440150156e-05,
+      "loss": 0.7248,
+      "step": 1533
+    },
+    {
+      "epoch": 0.392704,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.346045559252624e-05,
+      "loss": 0.662,
+      "step": 1534
+    },
+    {
+      "epoch": 0.39296,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.345618974490232e-05,
+      "loss": 0.6914,
+      "step": 1535
+    },
+    {
+      "epoch": 0.393216,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.345192389727839e-05,
+      "loss": 0.712,
+      "step": 1536
+    },
+    {
+      "epoch": 0.393472,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.3447658049654466e-05,
+      "loss": 0.7049,
+      "step": 1537
+    },
+    {
+      "epoch": 0.393728,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.344339220203055e-05,
+      "loss": 0.6976,
+      "step": 1538
+    },
+    {
+      "epoch": 0.393984,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.343912635440662e-05,
+      "loss": 0.7509,
+      "step": 1539
+    },
+    {
+      "epoch": 0.39424,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.3434860506782696e-05,
+      "loss": 0.667,
+      "step": 1540
+    },
+    {
+      "epoch": 0.394496,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.343059465915878e-05,
+      "loss": 0.716,
+      "step": 1541
+    },
+    {
+      "epoch": 0.394752,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.342632881153486e-05,
+      "loss": 0.6706,
+      "step": 1542
+    },
+    {
+      "epoch": 0.395008,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.342206296391093e-05,
+      "loss": 0.6881,
+      "step": 1543
+    },
+    {
+      "epoch": 0.395264,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.341779711628701e-05,
+      "loss": 0.741,
+      "step": 1544
+    },
+    {
+      "epoch": 0.39552,
+      "grad_norm": 2.625,
+      "learning_rate": 4.341353126866309e-05,
+      "loss": 0.6485,
+      "step": 1545
+    },
+    {
+      "epoch": 0.395776,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.340926542103916e-05,
+      "loss": 0.7342,
+      "step": 1546
+    },
+    {
+      "epoch": 0.396032,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.3404999573415236e-05,
+      "loss": 0.6714,
+      "step": 1547
+    },
+    {
+      "epoch": 0.396288,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.340073372579132e-05,
+      "loss": 0.7346,
+      "step": 1548
+    },
+    {
+      "epoch": 0.396544,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.33964678781674e-05,
+      "loss": 0.7321,
+      "step": 1549
+    },
+    {
+      "epoch": 0.3968,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.339220203054347e-05,
+      "loss": 0.6918,
+      "step": 1550
+    },
+    {
+      "epoch": 0.397056,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.338793618291955e-05,
+      "loss": 0.6977,
+      "step": 1551
+    },
+    {
+      "epoch": 0.397312,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.338367033529562e-05,
+      "loss": 0.7053,
+      "step": 1552
+    },
+    {
+      "epoch": 0.397568,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.33794044876717e-05,
+      "loss": 0.7162,
+      "step": 1553
+    },
+    {
+      "epoch": 0.397824,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.337513864004778e-05,
+      "loss": 0.6474,
+      "step": 1554
+    },
+    {
+      "epoch": 0.39808,
+      "grad_norm": 2.625,
+      "learning_rate": 4.337087279242386e-05,
+      "loss": 0.6437,
+      "step": 1555
+    },
+    {
+      "epoch": 0.398336,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.336660694479993e-05,
+      "loss": 0.6861,
+      "step": 1556
+    },
+    {
+      "epoch": 0.398592,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.336234109717601e-05,
+      "loss": 0.6858,
+      "step": 1557
+    },
+    {
+      "epoch": 0.398848,
+      "grad_norm": 2.875,
+      "learning_rate": 4.335807524955209e-05,
+      "loss": 0.6819,
+      "step": 1558
+    },
+    {
+      "epoch": 0.399104,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.335380940192816e-05,
+      "loss": 0.7111,
+      "step": 1559
+    },
+    {
+      "epoch": 0.39936,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.334954355430424e-05,
+      "loss": 0.7201,
+      "step": 1560
+    },
+    {
+      "epoch": 0.399616,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.3345277706680323e-05,
+      "loss": 0.6794,
+      "step": 1561
+    },
+    {
+      "epoch": 0.399872,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.33410118590564e-05,
+      "loss": 0.7045,
+      "step": 1562
+    },
+    {
+      "epoch": 0.400128,
+      "grad_norm": 2.875,
+      "learning_rate": 4.333674601143247e-05,
+      "loss": 0.6884,
+      "step": 1563
+    },
+    {
+      "epoch": 0.400384,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.333248016380855e-05,
+      "loss": 0.6928,
+      "step": 1564
+    },
+    {
+      "epoch": 0.40064,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.332821431618463e-05,
+      "loss": 0.6734,
+      "step": 1565
+    },
+    {
+      "epoch": 0.400896,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.33239484685607e-05,
+      "loss": 0.7,
+      "step": 1566
+    },
+    {
+      "epoch": 0.401152,
+      "grad_norm": 2.625,
+      "learning_rate": 4.331968262093678e-05,
+      "loss": 0.6719,
+      "step": 1567
+    },
+    {
+      "epoch": 0.401408,
+      "grad_norm": 2.75,
+      "learning_rate": 4.3315416773312864e-05,
+      "loss": 0.686,
+      "step": 1568
+    },
+    {
+      "epoch": 0.401664,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.331115092568894e-05,
+      "loss": 0.682,
+      "step": 1569
+    },
+    {
+      "epoch": 0.40192,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.330688507806501e-05,
+      "loss": 0.6865,
+      "step": 1570
+    },
+    {
+      "epoch": 0.402176,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.3302619230441086e-05,
+      "loss": 0.6534,
+      "step": 1571
+    },
+    {
+      "epoch": 0.402432,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.329835338281717e-05,
+      "loss": 0.7075,
+      "step": 1572
+    },
+    {
+      "epoch": 0.402688,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.329408753519325e-05,
+      "loss": 0.6402,
+      "step": 1573
+    },
+    {
+      "epoch": 0.402944,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.328982168756932e-05,
+      "loss": 0.7562,
+      "step": 1574
+    },
+    {
+      "epoch": 0.4032,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.3285555839945404e-05,
+      "loss": 0.7184,
+      "step": 1575
+    },
+    {
+      "epoch": 0.403456,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.328128999232148e-05,
+      "loss": 0.7241,
+      "step": 1576
+    },
+    {
+      "epoch": 0.403712,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.327702414469755e-05,
+      "loss": 0.7397,
+      "step": 1577
+    },
+    {
+      "epoch": 0.403968,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.3272758297073627e-05,
+      "loss": 0.7104,
+      "step": 1578
+    },
+    {
+      "epoch": 0.404224,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.326849244944971e-05,
+      "loss": 0.68,
+      "step": 1579
+    },
+    {
+      "epoch": 0.40448,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.326422660182579e-05,
+      "loss": 0.6881,
+      "step": 1580
+    },
+    {
+      "epoch": 0.404736,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.325996075420186e-05,
+      "loss": 0.6996,
+      "step": 1581
+    },
+    {
+      "epoch": 0.404992,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.325569490657794e-05,
+      "loss": 0.6954,
+      "step": 1582
+    },
+    {
+      "epoch": 0.405248,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.325142905895402e-05,
+      "loss": 0.6963,
+      "step": 1583
+    },
+    {
+      "epoch": 0.405504,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.324716321133009e-05,
+      "loss": 0.6358,
+      "step": 1584
+    },
+    {
+      "epoch": 0.40576,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.324289736370617e-05,
+      "loss": 0.6712,
+      "step": 1585
+    },
+    {
+      "epoch": 0.406016,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.323863151608225e-05,
+      "loss": 0.6361,
+      "step": 1586
+    },
+    {
+      "epoch": 0.406272,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.323436566845833e-05,
+      "loss": 0.7294,
+      "step": 1587
+    },
+    {
+      "epoch": 0.406528,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.32300998208344e-05,
+      "loss": 0.6942,
+      "step": 1588
+    },
+    {
+      "epoch": 0.406784,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.322583397321048e-05,
+      "loss": 0.6606,
+      "step": 1589
+    },
+    {
+      "epoch": 0.40704,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.322156812558655e-05,
+      "loss": 0.6515,
+      "step": 1590
+    },
+    {
+      "epoch": 0.407296,
+      "grad_norm": 2.625,
+      "learning_rate": 4.321730227796263e-05,
+      "loss": 0.6884,
+      "step": 1591
+    },
+    {
+      "epoch": 0.407552,
+      "grad_norm": 2.625,
+      "learning_rate": 4.321303643033871e-05,
+      "loss": 0.6873,
+      "step": 1592
+    },
+    {
+      "epoch": 0.407808,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.320877058271479e-05,
+      "loss": 0.6755,
+      "step": 1593
+    },
+    {
+      "epoch": 0.408064,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.320450473509087e-05,
+      "loss": 0.655,
+      "step": 1594
+    },
+    {
+      "epoch": 0.40832,
+      "grad_norm": 2.625,
+      "learning_rate": 4.320023888746694e-05,
+      "loss": 0.653,
+      "step": 1595
+    },
+    {
+      "epoch": 0.408576,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.319597303984302e-05,
+      "loss": 0.6652,
+      "step": 1596
+    },
+    {
+      "epoch": 0.408832,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.319170719221909e-05,
+      "loss": 0.6622,
+      "step": 1597
+    },
+    {
+      "epoch": 0.409088,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.318744134459517e-05,
+      "loss": 0.6843,
+      "step": 1598
+    },
+    {
+      "epoch": 0.409344,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.3183175496971254e-05,
+      "loss": 0.6802,
+      "step": 1599
+    },
+    {
+      "epoch": 0.4096,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.317890964934733e-05,
+      "loss": 0.6934,
+      "step": 1600
+    },
+    {
+      "epoch": 0.409856,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.31746438017234e-05,
+      "loss": 0.7021,
+      "step": 1601
+    },
+    {
+      "epoch": 0.410112,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.3170377954099484e-05,
+      "loss": 0.6795,
+      "step": 1602
+    },
+    {
+      "epoch": 0.410368,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.316611210647556e-05,
+      "loss": 0.6865,
+      "step": 1603
+    },
+    {
+      "epoch": 0.410624,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.316184625885163e-05,
+      "loss": 0.6653,
+      "step": 1604
+    },
+    {
+      "epoch": 0.41088,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.315758041122771e-05,
+      "loss": 0.7059,
+      "step": 1605
+    },
+    {
+      "epoch": 0.411136,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.3153314563603794e-05,
+      "loss": 0.7013,
+      "step": 1606
+    },
+    {
+      "epoch": 0.411392,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.314904871597987e-05,
+      "loss": 0.6859,
+      "step": 1607
+    },
+    {
+      "epoch": 0.411648,
+      "grad_norm": 2.625,
+      "learning_rate": 4.314478286835594e-05,
+      "loss": 0.6986,
+      "step": 1608
+    },
+    {
+      "epoch": 0.411904,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.3140517020732024e-05,
+      "loss": 0.7364,
+      "step": 1609
+    },
+    {
+      "epoch": 0.41216,
+      "grad_norm": 2.75,
+      "learning_rate": 4.31362511731081e-05,
+      "loss": 0.6612,
+      "step": 1610
+    },
+    {
+      "epoch": 0.412416,
+      "grad_norm": 2.625,
+      "learning_rate": 4.313198532548417e-05,
+      "loss": 0.6716,
+      "step": 1611
+    },
+    {
+      "epoch": 0.412672,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.312771947786025e-05,
+      "loss": 0.6446,
+      "step": 1612
+    },
+    {
+      "epoch": 0.412928,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.3123453630236334e-05,
+      "loss": 0.7022,
+      "step": 1613
+    },
+    {
+      "epoch": 0.413184,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.311918778261241e-05,
+      "loss": 0.6779,
+      "step": 1614
+    },
+    {
+      "epoch": 0.41344,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.311492193498848e-05,
+      "loss": 0.7446,
+      "step": 1615
+    },
+    {
+      "epoch": 0.413696,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.311065608736456e-05,
+      "loss": 0.7091,
+      "step": 1616
+    },
+    {
+      "epoch": 0.413952,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.310639023974064e-05,
+      "loss": 0.7044,
+      "step": 1617
+    },
+    {
+      "epoch": 0.414208,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.310212439211672e-05,
+      "loss": 0.719,
+      "step": 1618
+    },
+    {
+      "epoch": 0.414464,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.3097858544492793e-05,
+      "loss": 0.6608,
+      "step": 1619
+    },
+    {
+      "epoch": 0.41472,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.309359269686887e-05,
+      "loss": 0.6919,
+      "step": 1620
+    },
+    {
+      "epoch": 0.414976,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.308932684924495e-05,
+      "loss": 0.6573,
+      "step": 1621
+    },
+    {
+      "epoch": 0.415232,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.308506100162102e-05,
+      "loss": 0.7054,
+      "step": 1622
+    },
+    {
+      "epoch": 0.415488,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.30807951539971e-05,
+      "loss": 0.6442,
+      "step": 1623
+    },
+    {
+      "epoch": 0.415744,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.307652930637318e-05,
+      "loss": 0.6434,
+      "step": 1624
+    },
+    {
+      "epoch": 0.416,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.307226345874926e-05,
+      "loss": 0.653,
+      "step": 1625
+    },
+    {
+      "epoch": 0.416256,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.3067997611125334e-05,
+      "loss": 0.721,
+      "step": 1626
+    },
+    {
+      "epoch": 0.416512,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.306373176350141e-05,
+      "loss": 0.6668,
+      "step": 1627
+    },
+    {
+      "epoch": 0.416768,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.305946591587749e-05,
+      "loss": 0.6397,
+      "step": 1628
+    },
+    {
+      "epoch": 0.417024,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.305520006825356e-05,
+      "loss": 0.7215,
+      "step": 1629
+    },
+    {
+      "epoch": 0.41728,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.305093422062964e-05,
+      "loss": 0.6886,
+      "step": 1630
+    },
+    {
+      "epoch": 0.417536,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.304666837300572e-05,
+      "loss": 0.6788,
+      "step": 1631
+    },
+    {
+      "epoch": 0.417792,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.30424025253818e-05,
+      "loss": 0.6776,
+      "step": 1632
+    },
+    {
+      "epoch": 0.418048,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.3038136677757874e-05,
+      "loss": 0.7035,
+      "step": 1633
+    },
+    {
+      "epoch": 0.418304,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.303387083013395e-05,
+      "loss": 0.6852,
+      "step": 1634
+    },
+    {
+      "epoch": 0.41856,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.302960498251002e-05,
+      "loss": 0.6557,
+      "step": 1635
+    },
+    {
+      "epoch": 0.418816,
+      "grad_norm": 2.75,
+      "learning_rate": 4.3025339134886103e-05,
+      "loss": 0.6456,
+      "step": 1636
+    },
+    {
+      "epoch": 0.419072,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.3021073287262184e-05,
+      "loss": 0.6875,
+      "step": 1637
+    },
+    {
+      "epoch": 0.419328,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.301680743963826e-05,
+      "loss": 0.6684,
+      "step": 1638
+    },
+    {
+      "epoch": 0.419584,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.301254159201433e-05,
+      "loss": 0.6853,
+      "step": 1639
+    },
+    {
+      "epoch": 0.41984,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.3008275744390414e-05,
+      "loss": 0.6893,
+      "step": 1640
+    },
+    {
+      "epoch": 0.420096,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.300400989676649e-05,
+      "loss": 0.6915,
+      "step": 1641
+    },
+    {
+      "epoch": 0.420352,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.299974404914256e-05,
+      "loss": 0.6289,
+      "step": 1642
+    },
+    {
+      "epoch": 0.420608,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.2995478201518644e-05,
+      "loss": 0.6672,
+      "step": 1643
+    },
+    {
+      "epoch": 0.420864,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.2991212353894725e-05,
+      "loss": 0.6774,
+      "step": 1644
+    },
+    {
+      "epoch": 0.42112,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.29869465062708e-05,
+      "loss": 0.7069,
+      "step": 1645
+    },
+    {
+      "epoch": 0.421376,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.298268065864687e-05,
+      "loss": 0.641,
+      "step": 1646
+    },
+    {
+      "epoch": 0.421632,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.2978414811022954e-05,
+      "loss": 0.705,
+      "step": 1647
+    },
+    {
+      "epoch": 0.421888,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.297414896339903e-05,
+      "loss": 0.6572,
+      "step": 1648
+    },
+    {
+      "epoch": 0.422144,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.29698831157751e-05,
+      "loss": 0.6688,
+      "step": 1649
+    },
+    {
+      "epoch": 0.4224,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.2965617268151184e-05,
+      "loss": 0.6664,
+      "step": 1650
+    },
+    {
+      "epoch": 0.422656,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.2961351420527265e-05,
+      "loss": 0.6858,
+      "step": 1651
+    },
+    {
+      "epoch": 0.422912,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.295708557290334e-05,
+      "loss": 0.6795,
+      "step": 1652
+    },
+    {
+      "epoch": 0.423168,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.2952819725279413e-05,
+      "loss": 0.6663,
+      "step": 1653
+    },
+    {
+      "epoch": 0.423424,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.294855387765549e-05,
+      "loss": 0.6603,
+      "step": 1654
+    },
+    {
+      "epoch": 0.42368,
+      "grad_norm": 2.75,
+      "learning_rate": 4.294428803003157e-05,
+      "loss": 0.6794,
+      "step": 1655
+    },
+    {
+      "epoch": 0.423936,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.294002218240765e-05,
+      "loss": 0.6808,
+      "step": 1656
+    },
+    {
+      "epoch": 0.424192,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.2935756334783724e-05,
+      "loss": 0.6749,
+      "step": 1657
+    },
+    {
+      "epoch": 0.424448,
+      "grad_norm": 2.75,
+      "learning_rate": 4.2931490487159805e-05,
+      "loss": 0.7026,
+      "step": 1658
+    },
+    {
+      "epoch": 0.424704,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.292722463953588e-05,
+      "loss": 0.6983,
+      "step": 1659
+    },
+    {
+      "epoch": 0.42496,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.2922958791911954e-05,
+      "loss": 0.668,
+      "step": 1660
+    },
+    {
+      "epoch": 0.425216,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.291869294428803e-05,
+      "loss": 0.668,
+      "step": 1661
+    },
+    {
+      "epoch": 0.425472,
+      "grad_norm": 3.0,
+      "learning_rate": 4.291442709666411e-05,
+      "loss": 0.6747,
+      "step": 1662
+    },
+    {
+      "epoch": 0.425728,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.291016124904019e-05,
+      "loss": 0.6593,
+      "step": 1663
+    },
+    {
+      "epoch": 0.425984,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.2905895401416264e-05,
+      "loss": 0.6869,
+      "step": 1664
+    },
+    {
+      "epoch": 0.42624,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.290162955379234e-05,
+      "loss": 0.6998,
+      "step": 1665
+    },
+    {
+      "epoch": 0.426496,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.289736370616842e-05,
+      "loss": 0.6784,
+      "step": 1666
+    },
+    {
+      "epoch": 0.426752,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.2893097858544494e-05,
+      "loss": 0.6515,
+      "step": 1667
+    },
+    {
+      "epoch": 0.427008,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.288883201092057e-05,
+      "loss": 0.6657,
+      "step": 1668
+    },
+    {
+      "epoch": 0.427264,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.288456616329665e-05,
+      "loss": 0.678,
+      "step": 1669
+    },
+    {
+      "epoch": 0.42752,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.288030031567273e-05,
+      "loss": 0.6749,
+      "step": 1670
+    },
+    {
+      "epoch": 0.427776,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.2876034468048804e-05,
+      "loss": 0.6828,
+      "step": 1671
+    },
+    {
+      "epoch": 0.428032,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.287176862042488e-05,
+      "loss": 0.7051,
+      "step": 1672
+    },
+    {
+      "epoch": 0.428288,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.286750277280095e-05,
+      "loss": 0.6756,
+      "step": 1673
+    },
+    {
+      "epoch": 0.428544,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.2863236925177034e-05,
+      "loss": 0.6683,
+      "step": 1674
+    },
+    {
+      "epoch": 0.4288,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.2858971077553115e-05,
+      "loss": 0.6891,
+      "step": 1675
+    },
+    {
+      "epoch": 0.429056,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.285470522992919e-05,
+      "loss": 0.6523,
+      "step": 1676
+    },
+    {
+      "epoch": 0.429312,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.285043938230527e-05,
+      "loss": 0.7094,
+      "step": 1677
+    },
+    {
+      "epoch": 0.429568,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.2846173534681345e-05,
+      "loss": 0.6845,
+      "step": 1678
+    },
+    {
+      "epoch": 0.429824,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.284190768705742e-05,
+      "loss": 0.6445,
+      "step": 1679
+    },
+    {
+      "epoch": 0.43008,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.283764183943349e-05,
+      "loss": 0.6545,
+      "step": 1680
+    },
+    {
+      "epoch": 0.430336,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.2833375991809574e-05,
+      "loss": 0.668,
+      "step": 1681
+    },
+    {
+      "epoch": 0.430592,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.2829110144185655e-05,
+      "loss": 0.6679,
+      "step": 1682
+    },
+    {
+      "epoch": 0.430848,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.282484429656173e-05,
+      "loss": 0.6429,
+      "step": 1683
+    },
+    {
+      "epoch": 0.431104,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.2820578448937804e-05,
+      "loss": 0.6682,
+      "step": 1684
+    },
+    {
+      "epoch": 0.43136,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.2816312601313885e-05,
+      "loss": 0.6497,
+      "step": 1685
+    },
+    {
+      "epoch": 0.431616,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.281204675368996e-05,
+      "loss": 0.6997,
+      "step": 1686
+    },
+    {
+      "epoch": 0.431872,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.280778090606603e-05,
+      "loss": 0.6654,
+      "step": 1687
+    },
+    {
+      "epoch": 0.432128,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.2803515058442114e-05,
+      "loss": 0.6671,
+      "step": 1688
+    },
+    {
+      "epoch": 0.432384,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.2799249210818195e-05,
+      "loss": 0.6216,
+      "step": 1689
+    },
+    {
+      "epoch": 0.43264,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.279498336319427e-05,
+      "loss": 0.6793,
+      "step": 1690
+    },
+    {
+      "epoch": 0.432896,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.2790717515570344e-05,
+      "loss": 0.6716,
+      "step": 1691
+    },
+    {
+      "epoch": 0.433152,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.2786451667946425e-05,
+      "loss": 0.6806,
+      "step": 1692
+    },
+    {
+      "epoch": 0.433408,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.27821858203225e-05,
+      "loss": 0.6492,
+      "step": 1693
+    },
+    {
+      "epoch": 0.433664,
+      "grad_norm": 2.5,
+      "learning_rate": 4.2777919972698574e-05,
+      "loss": 0.6305,
+      "step": 1694
+    },
+    {
+      "epoch": 0.43392,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.2773654125074655e-05,
+      "loss": 0.6189,
+      "step": 1695
+    },
+    {
+      "epoch": 0.434176,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.2769388277450736e-05,
+      "loss": 0.6458,
+      "step": 1696
+    },
+    {
+      "epoch": 0.434432,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.276512242982681e-05,
+      "loss": 0.7145,
+      "step": 1697
+    },
+    {
+      "epoch": 0.434688,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.2760856582202884e-05,
+      "loss": 0.6771,
+      "step": 1698
+    },
+    {
+      "epoch": 0.434944,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.275659073457896e-05,
+      "loss": 0.6921,
+      "step": 1699
+    },
+    {
+      "epoch": 0.4352,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.275232488695504e-05,
+      "loss": 0.7022,
+      "step": 1700
+    },
+    {
+      "epoch": 0.435456,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.274805903933112e-05,
+      "loss": 0.7176,
+      "step": 1701
+    },
+    {
+      "epoch": 0.435712,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.2743793191707195e-05,
+      "loss": 0.7063,
+      "step": 1702
+    },
+    {
+      "epoch": 0.435968,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.273952734408327e-05,
+      "loss": 0.6535,
+      "step": 1703
+    },
+    {
+      "epoch": 0.436224,
+      "grad_norm": 2.75,
+      "learning_rate": 4.273526149645935e-05,
+      "loss": 0.6626,
+      "step": 1704
+    },
+    {
+      "epoch": 0.43648,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.2730995648835424e-05,
+      "loss": 0.6531,
+      "step": 1705
+    },
+    {
+      "epoch": 0.436736,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.27267298012115e-05,
+      "loss": 0.6626,
+      "step": 1706
+    },
+    {
+      "epoch": 0.436992,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.272246395358758e-05,
+      "loss": 0.6455,
+      "step": 1707
+    },
+    {
+      "epoch": 0.437248,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.271819810596366e-05,
+      "loss": 0.6952,
+      "step": 1708
+    },
+    {
+      "epoch": 0.437504,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.2713932258339735e-05,
+      "loss": 0.6768,
+      "step": 1709
+    },
+    {
+      "epoch": 0.43776,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.270966641071581e-05,
+      "loss": 0.6416,
+      "step": 1710
+    },
+    {
+      "epoch": 0.438016,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.270540056309189e-05,
+      "loss": 0.6503,
+      "step": 1711
+    },
+    {
+      "epoch": 0.438272,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.2701134715467965e-05,
+      "loss": 0.6459,
+      "step": 1712
+    },
+    {
+      "epoch": 0.438528,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.269686886784404e-05,
+      "loss": 0.6319,
+      "step": 1713
+    },
+    {
+      "epoch": 0.438784,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.269260302022012e-05,
+      "loss": 0.6884,
+      "step": 1714
+    },
+    {
+      "epoch": 0.43904,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.26883371725962e-05,
+      "loss": 0.6711,
+      "step": 1715
+    },
+    {
+      "epoch": 0.439296,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.2684071324972275e-05,
+      "loss": 0.6924,
+      "step": 1716
+    },
+    {
+      "epoch": 0.439552,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.267980547734835e-05,
+      "loss": 0.6262,
+      "step": 1717
+    },
+    {
+      "epoch": 0.439808,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.2675539629724424e-05,
+      "loss": 0.707,
+      "step": 1718
+    },
+    {
+      "epoch": 0.440064,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.2671273782100505e-05,
+      "loss": 0.6398,
+      "step": 1719
+    },
+    {
+      "epoch": 0.44032,
+      "grad_norm": 2.875,
+      "learning_rate": 4.2667007934476586e-05,
+      "loss": 0.6084,
+      "step": 1720
+    },
+    {
+      "epoch": 0.440576,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.266274208685266e-05,
+      "loss": 0.6449,
+      "step": 1721
+    },
+    {
+      "epoch": 0.440832,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.2658476239228734e-05,
+      "loss": 0.6449,
+      "step": 1722
+    },
+    {
+      "epoch": 0.441088,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.2654210391604815e-05,
+      "loss": 0.71,
+      "step": 1723
+    },
+    {
+      "epoch": 0.441344,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.264994454398089e-05,
+      "loss": 0.6591,
+      "step": 1724
+    },
+    {
+      "epoch": 0.4416,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.2645678696356964e-05,
+      "loss": 0.6671,
+      "step": 1725
+    },
+    {
+      "epoch": 0.441856,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.2641412848733045e-05,
+      "loss": 0.6446,
+      "step": 1726
+    },
+    {
+      "epoch": 0.442112,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.2637147001109126e-05,
+      "loss": 0.7014,
+      "step": 1727
+    },
+    {
+      "epoch": 0.442368,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.26328811534852e-05,
+      "loss": 0.6533,
+      "step": 1728
+    },
+    {
+      "epoch": 0.442624,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.2628615305861274e-05,
+      "loss": 0.6599,
+      "step": 1729
+    },
+    {
+      "epoch": 0.44288,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.2624349458237356e-05,
+      "loss": 0.6511,
+      "step": 1730
+    },
+    {
+      "epoch": 0.443136,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.262008361061343e-05,
+      "loss": 0.6769,
+      "step": 1731
+    },
+    {
+      "epoch": 0.443392,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.2615817762989504e-05,
+      "loss": 0.6765,
+      "step": 1732
+    },
+    {
+      "epoch": 0.443648,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.2611551915365585e-05,
+      "loss": 0.6817,
+      "step": 1733
+    },
+    {
+      "epoch": 0.443904,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.2607286067741666e-05,
+      "loss": 0.687,
+      "step": 1734
+    },
+    {
+      "epoch": 0.44416,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.260302022011774e-05,
+      "loss": 0.6892,
+      "step": 1735
+    },
+    {
+      "epoch": 0.444416,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.2598754372493815e-05,
+      "loss": 0.6673,
+      "step": 1736
+    },
+    {
+      "epoch": 0.444672,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.259448852486989e-05,
+      "loss": 0.666,
+      "step": 1737
+    },
+    {
+      "epoch": 0.444928,
+      "grad_norm": 2.625,
+      "learning_rate": 4.259022267724597e-05,
+      "loss": 0.6602,
+      "step": 1738
+    },
+    {
+      "epoch": 0.445184,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.258595682962205e-05,
+      "loss": 0.6843,
+      "step": 1739
+    },
+    {
+      "epoch": 0.44544,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.2581690981998125e-05,
+      "loss": 0.6518,
+      "step": 1740
+    },
+    {
+      "epoch": 0.445696,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.2577425134374206e-05,
+      "loss": 0.6141,
+      "step": 1741
+    },
+    {
+      "epoch": 0.445952,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.257315928675028e-05,
+      "loss": 0.6496,
+      "step": 1742
+    },
+    {
+      "epoch": 0.446208,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.2568893439126355e-05,
+      "loss": 0.6545,
+      "step": 1743
+    },
+    {
+      "epoch": 0.446464,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.256462759150243e-05,
+      "loss": 0.6182,
+      "step": 1744
+    },
+    {
+      "epoch": 0.44672,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.256036174387851e-05,
+      "loss": 0.6887,
+      "step": 1745
+    },
+    {
+      "epoch": 0.446976,
+      "grad_norm": 2.5,
+      "learning_rate": 4.255609589625459e-05,
+      "loss": 0.607,
+      "step": 1746
+    },
+    {
+      "epoch": 0.447232,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.2551830048630665e-05,
+      "loss": 0.6434,
+      "step": 1747
+    },
+    {
+      "epoch": 0.447488,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.254756420100674e-05,
+      "loss": 0.6455,
+      "step": 1748
+    },
+    {
+      "epoch": 0.447744,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.254329835338282e-05,
+      "loss": 0.6661,
+      "step": 1749
+    },
+    {
+      "epoch": 0.448,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.2539032505758895e-05,
+      "loss": 0.6591,
+      "step": 1750
+    },
+    {
+      "epoch": 0.448256,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.253476665813497e-05,
+      "loss": 0.6577,
+      "step": 1751
+    },
+    {
+      "epoch": 0.448512,
+      "grad_norm": 2.625,
+      "learning_rate": 4.253050081051105e-05,
+      "loss": 0.665,
+      "step": 1752
+    },
+    {
+      "epoch": 0.448768,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.252623496288713e-05,
+      "loss": 0.6509,
+      "step": 1753
+    },
+    {
+      "epoch": 0.449024,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.2521969115263206e-05,
+      "loss": 0.6288,
+      "step": 1754
+    },
+    {
+      "epoch": 0.44928,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.251770326763928e-05,
+      "loss": 0.6683,
+      "step": 1755
+    },
+    {
+      "epoch": 0.449536,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.251343742001536e-05,
+      "loss": 0.6236,
+      "step": 1756
+    },
+    {
+      "epoch": 0.449792,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.2509171572391435e-05,
+      "loss": 0.6265,
+      "step": 1757
+    },
+    {
+      "epoch": 0.450048,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.2504905724767516e-05,
+      "loss": 0.6567,
+      "step": 1758
+    },
+    {
+      "epoch": 0.450304,
+      "grad_norm": 2.375,
+      "learning_rate": 4.250063987714359e-05,
+      "loss": 0.6785,
+      "step": 1759
+    },
+    {
+      "epoch": 0.45056,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.249637402951967e-05,
+      "loss": 0.7115,
+      "step": 1760
+    },
+    {
+      "epoch": 0.450816,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.2492108181895746e-05,
+      "loss": 0.6502,
+      "step": 1761
+    },
+    {
+      "epoch": 0.451072,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.248784233427182e-05,
+      "loss": 0.69,
+      "step": 1762
+    },
+    {
+      "epoch": 0.451328,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.2483576486647894e-05,
+      "loss": 0.639,
+      "step": 1763
+    },
+    {
+      "epoch": 0.451584,
+      "grad_norm": 2.375,
+      "learning_rate": 4.2479310639023975e-05,
+      "loss": 0.6573,
+      "step": 1764
+    },
+    {
+      "epoch": 0.45184,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.2475044791400056e-05,
+      "loss": 0.663,
+      "step": 1765
+    },
+    {
+      "epoch": 0.452096,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.247077894377613e-05,
+      "loss": 0.7069,
+      "step": 1766
+    },
+    {
+      "epoch": 0.452352,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.2466513096152205e-05,
+      "loss": 0.6517,
+      "step": 1767
+    },
+    {
+      "epoch": 0.452608,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.2462247248528286e-05,
+      "loss": 0.6615,
+      "step": 1768
+    },
+    {
+      "epoch": 0.452864,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.245798140090436e-05,
+      "loss": 0.6422,
+      "step": 1769
+    },
+    {
+      "epoch": 0.45312,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.2453715553280435e-05,
+      "loss": 0.6583,
+      "step": 1770
+    },
+    {
+      "epoch": 0.453376,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.2449449705656516e-05,
+      "loss": 0.6745,
+      "step": 1771
+    },
+    {
+      "epoch": 0.453632,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.24451838580326e-05,
+      "loss": 0.6283,
+      "step": 1772
+    },
+    {
+      "epoch": 0.453888,
+      "grad_norm": 2.375,
+      "learning_rate": 4.244091801040867e-05,
+      "loss": 0.7095,
+      "step": 1773
+    },
+    {
+      "epoch": 0.454144,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.2436652162784745e-05,
+      "loss": 0.6626,
+      "step": 1774
+    },
+    {
+      "epoch": 0.4544,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.2432386315160826e-05,
+      "loss": 0.6296,
+      "step": 1775
+    },
+    {
+      "epoch": 0.454656,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.24281204675369e-05,
+      "loss": 0.6355,
+      "step": 1776
+    },
+    {
+      "epoch": 0.454912,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.242385461991298e-05,
+      "loss": 0.6664,
+      "step": 1777
+    },
+    {
+      "epoch": 0.455168,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.2419588772289056e-05,
+      "loss": 0.662,
+      "step": 1778
+    },
+    {
+      "epoch": 0.455424,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.241532292466514e-05,
+      "loss": 0.6682,
+      "step": 1779
+    },
+    {
+      "epoch": 0.45568,
+      "grad_norm": 2.625,
+      "learning_rate": 4.241105707704121e-05,
+      "loss": 0.6223,
+      "step": 1780
+    },
+    {
+      "epoch": 0.455936,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.2406791229417285e-05,
+      "loss": 0.6432,
+      "step": 1781
+    },
+    {
+      "epoch": 0.456192,
+      "grad_norm": 2.75,
+      "learning_rate": 4.240252538179336e-05,
+      "loss": 0.6455,
+      "step": 1782
+    },
+    {
+      "epoch": 0.456448,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.239825953416944e-05,
+      "loss": 0.6833,
+      "step": 1783
+    },
+    {
+      "epoch": 0.456704,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.239399368654552e-05,
+      "loss": 0.6876,
+      "step": 1784
+    },
+    {
+      "epoch": 0.45696,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.2389727838921596e-05,
+      "loss": 0.6515,
+      "step": 1785
+    },
+    {
+      "epoch": 0.457216,
+      "grad_norm": 2.625,
+      "learning_rate": 4.238546199129767e-05,
+      "loss": 0.6611,
+      "step": 1786
+    },
+    {
+      "epoch": 0.457472,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.238119614367375e-05,
+      "loss": 0.6822,
+      "step": 1787
+    },
+    {
+      "epoch": 0.457728,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.2376930296049826e-05,
+      "loss": 0.6289,
+      "step": 1788
+    },
+    {
+      "epoch": 0.457984,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.23726644484259e-05,
+      "loss": 0.678,
+      "step": 1789
+    },
+    {
+      "epoch": 0.45824,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.236839860080198e-05,
+      "loss": 0.6509,
+      "step": 1790
+    },
+    {
+      "epoch": 0.458496,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.236413275317806e-05,
+      "loss": 0.6449,
+      "step": 1791
+    },
+    {
+      "epoch": 0.458752,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.2359866905554136e-05,
+      "loss": 0.6897,
+      "step": 1792
+    },
+    {
+      "epoch": 0.459008,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.235560105793021e-05,
+      "loss": 0.6823,
+      "step": 1793
+    },
+    {
+      "epoch": 0.459264,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.235133521030629e-05,
+      "loss": 0.6281,
+      "step": 1794
+    },
+    {
+      "epoch": 0.45952,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.2347069362682366e-05,
+      "loss": 0.6489,
+      "step": 1795
+    },
+    {
+      "epoch": 0.459776,
+      "grad_norm": 2.75,
+      "learning_rate": 4.234280351505845e-05,
+      "loss": 0.6665,
+      "step": 1796
+    },
+    {
+      "epoch": 0.460032,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.233853766743452e-05,
+      "loss": 0.6476,
+      "step": 1797
+    },
+    {
+      "epoch": 0.460288,
+      "grad_norm": 2.75,
+      "learning_rate": 4.23342718198106e-05,
+      "loss": 0.6473,
+      "step": 1798
+    },
+    {
+      "epoch": 0.460544,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.2330005972186676e-05,
+      "loss": 0.6885,
+      "step": 1799
+    },
+    {
+      "epoch": 0.4608,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.232574012456275e-05,
+      "loss": 0.6785,
+      "step": 1800
+    },
+    {
+      "epoch": 0.461056,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.2321474276938825e-05,
+      "loss": 0.6454,
+      "step": 1801
+    },
+    {
+      "epoch": 0.461312,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.2317208429314906e-05,
+      "loss": 0.6069,
+      "step": 1802
+    },
+    {
+      "epoch": 0.461568,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.231294258169099e-05,
+      "loss": 0.6729,
+      "step": 1803
+    },
+    {
+      "epoch": 0.461824,
+      "grad_norm": 2.625,
+      "learning_rate": 4.230867673406706e-05,
+      "loss": 0.6489,
+      "step": 1804
+    },
+    {
+      "epoch": 0.46208,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.230441088644314e-05,
+      "loss": 0.6541,
+      "step": 1805
+    },
+    {
+      "epoch": 0.462336,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.2300145038819217e-05,
+      "loss": 0.6377,
+      "step": 1806
+    },
+    {
+      "epoch": 0.462592,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.229587919119529e-05,
+      "loss": 0.6658,
+      "step": 1807
+    },
+    {
+      "epoch": 0.462848,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.2291613343571365e-05,
+      "loss": 0.6164,
+      "step": 1808
+    },
+    {
+      "epoch": 0.463104,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.2287347495947446e-05,
+      "loss": 0.6218,
+      "step": 1809
+    },
+    {
+      "epoch": 0.46336,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.228308164832353e-05,
+      "loss": 0.6867,
+      "step": 1810
+    },
+    {
+      "epoch": 0.463616,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.22788158006996e-05,
+      "loss": 0.6605,
+      "step": 1811
+    },
+    {
+      "epoch": 0.463872,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.2274549953075676e-05,
+      "loss": 0.7173,
+      "step": 1812
+    },
+    {
+      "epoch": 0.464128,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.227028410545176e-05,
+      "loss": 0.6626,
+      "step": 1813
+    },
+    {
+      "epoch": 0.464384,
+      "grad_norm": 2.5,
+      "learning_rate": 4.226601825782783e-05,
+      "loss": 0.6404,
+      "step": 1814
+    },
+    {
+      "epoch": 0.46464,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.2261752410203905e-05,
+      "loss": 0.6281,
+      "step": 1815
+    },
+    {
+      "epoch": 0.464896,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.2257486562579986e-05,
+      "loss": 0.6481,
+      "step": 1816
+    },
+    {
+      "epoch": 0.465152,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.225322071495607e-05,
+      "loss": 0.6234,
+      "step": 1817
+    },
+    {
+      "epoch": 0.465408,
+      "grad_norm": 2.625,
+      "learning_rate": 4.224895486733214e-05,
+      "loss": 0.6176,
+      "step": 1818
+    },
+    {
+      "epoch": 0.465664,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.2244689019708216e-05,
+      "loss": 0.6274,
+      "step": 1819
+    },
+    {
+      "epoch": 0.46592,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.224042317208429e-05,
+      "loss": 0.6459,
+      "step": 1820
+    },
+    {
+      "epoch": 0.466176,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.223615732446037e-05,
+      "loss": 0.629,
+      "step": 1821
+    },
+    {
+      "epoch": 0.466432,
+      "grad_norm": 2.5,
+      "learning_rate": 4.223189147683645e-05,
+      "loss": 0.6072,
+      "step": 1822
+    },
+    {
+      "epoch": 0.466688,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.2227625629212527e-05,
+      "loss": 0.6563,
+      "step": 1823
+    },
+    {
+      "epoch": 0.466944,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.222335978158861e-05,
+      "loss": 0.652,
+      "step": 1824
+    },
+    {
+      "epoch": 0.4672,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.221909393396468e-05,
+      "loss": 0.648,
+      "step": 1825
+    },
+    {
+      "epoch": 0.467456,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.2214828086340756e-05,
+      "loss": 0.6394,
+      "step": 1826
+    },
+    {
+      "epoch": 0.467712,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.221056223871683e-05,
+      "loss": 0.6349,
+      "step": 1827
+    },
+    {
+      "epoch": 0.467968,
+      "grad_norm": 2.5,
+      "learning_rate": 4.220629639109291e-05,
+      "loss": 0.6666,
+      "step": 1828
+    },
+    {
+      "epoch": 0.468224,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.220203054346899e-05,
+      "loss": 0.6625,
+      "step": 1829
+    },
+    {
+      "epoch": 0.46848,
+      "grad_norm": 2.5,
+      "learning_rate": 4.219776469584507e-05,
+      "loss": 0.6824,
+      "step": 1830
+    },
+    {
+      "epoch": 0.468736,
+      "grad_norm": 2.75,
+      "learning_rate": 4.219349884822114e-05,
+      "loss": 0.6135,
+      "step": 1831
+    },
+    {
+      "epoch": 0.468992,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.218923300059722e-05,
+      "loss": 0.6562,
+      "step": 1832
+    },
+    {
+      "epoch": 0.469248,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.2184967152973296e-05,
+      "loss": 0.6252,
+      "step": 1833
+    },
+    {
+      "epoch": 0.469504,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.218070130534937e-05,
+      "loss": 0.5786,
+      "step": 1834
+    },
+    {
+      "epoch": 0.46976,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.217643545772545e-05,
+      "loss": 0.657,
+      "step": 1835
+    },
+    {
+      "epoch": 0.470016,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.217216961010153e-05,
+      "loss": 0.6396,
+      "step": 1836
+    },
+    {
+      "epoch": 0.470272,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.216790376247761e-05,
+      "loss": 0.655,
+      "step": 1837
+    },
+    {
+      "epoch": 0.470528,
+      "grad_norm": 2.625,
+      "learning_rate": 4.216363791485368e-05,
+      "loss": 0.6632,
+      "step": 1838
+    },
+    {
+      "epoch": 0.470784,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.215937206722976e-05,
+      "loss": 0.6167,
+      "step": 1839
+    },
+    {
+      "epoch": 0.47104,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.2155106219605837e-05,
+      "loss": 0.6731,
+      "step": 1840
+    },
+    {
+      "epoch": 0.471296,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.215084037198192e-05,
+      "loss": 0.6291,
+      "step": 1841
+    },
+    {
+      "epoch": 0.471552,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.214657452435799e-05,
+      "loss": 0.6683,
+      "step": 1842
+    },
+    {
+      "epoch": 0.471808,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.214230867673407e-05,
+      "loss": 0.6298,
+      "step": 1843
+    },
+    {
+      "epoch": 0.472064,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.213804282911015e-05,
+      "loss": 0.6108,
+      "step": 1844
+    },
+    {
+      "epoch": 0.47232,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.213377698148622e-05,
+      "loss": 0.6656,
+      "step": 1845
+    },
+    {
+      "epoch": 0.472576,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.2129511133862296e-05,
+      "loss": 0.6727,
+      "step": 1846
+    },
+    {
+      "epoch": 0.472832,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.212524528623838e-05,
+      "loss": 0.6548,
+      "step": 1847
+    },
+    {
+      "epoch": 0.473088,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.212097943861446e-05,
+      "loss": 0.6594,
+      "step": 1848
+    },
+    {
+      "epoch": 0.473344,
+      "grad_norm": 2.75,
+      "learning_rate": 4.211671359099053e-05,
+      "loss": 0.6119,
+      "step": 1849
+    },
+    {
+      "epoch": 0.4736,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.2112447743366606e-05,
+      "loss": 0.719,
+      "step": 1850
+    },
+    {
+      "epoch": 0.473856,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.210818189574269e-05,
+      "loss": 0.671,
+      "step": 1851
+    },
+    {
+      "epoch": 0.474112,
+      "grad_norm": 2.5,
+      "learning_rate": 4.210391604811876e-05,
+      "loss": 0.6524,
+      "step": 1852
+    },
+    {
+      "epoch": 0.474368,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.2099650200494836e-05,
+      "loss": 0.6421,
+      "step": 1853
+    },
+    {
+      "epoch": 0.474624,
+      "grad_norm": 2.875,
+      "learning_rate": 4.209538435287092e-05,
+      "loss": 0.6152,
+      "step": 1854
+    },
+    {
+      "epoch": 0.47488,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.2091118505247e-05,
+      "loss": 0.632,
+      "step": 1855
+    },
+    {
+      "epoch": 0.475136,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.208685265762307e-05,
+      "loss": 0.6648,
+      "step": 1856
+    },
+    {
+      "epoch": 0.475392,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.2082586809999146e-05,
+      "loss": 0.6195,
+      "step": 1857
+    },
+    {
+      "epoch": 0.475648,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.207832096237523e-05,
+      "loss": 0.6692,
+      "step": 1858
+    },
+    {
+      "epoch": 0.475904,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.20740551147513e-05,
+      "loss": 0.6011,
+      "step": 1859
+    },
+    {
+      "epoch": 0.47616,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.206978926712738e-05,
+      "loss": 0.6025,
+      "step": 1860
+    },
+    {
+      "epoch": 0.476416,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.206552341950346e-05,
+      "loss": 0.6238,
+      "step": 1861
+    },
+    {
+      "epoch": 0.476672,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.206125757187954e-05,
+      "loss": 0.6198,
+      "step": 1862
+    },
+    {
+      "epoch": 0.476928,
+      "grad_norm": 2.5,
+      "learning_rate": 4.205699172425561e-05,
+      "loss": 0.6375,
+      "step": 1863
+    },
+    {
+      "epoch": 0.477184,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.205272587663169e-05,
+      "loss": 0.6306,
+      "step": 1864
+    },
+    {
+      "epoch": 0.47744,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.204846002900776e-05,
+      "loss": 0.6506,
+      "step": 1865
+    },
+    {
+      "epoch": 0.477696,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.204419418138384e-05,
+      "loss": 0.6598,
+      "step": 1866
+    },
+    {
+      "epoch": 0.477952,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.203992833375992e-05,
+      "loss": 0.6409,
+      "step": 1867
+    },
+    {
+      "epoch": 0.478208,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.2035662486136e-05,
+      "loss": 0.6153,
+      "step": 1868
+    },
+    {
+      "epoch": 0.478464,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.203139663851207e-05,
+      "loss": 0.6835,
+      "step": 1869
+    },
+    {
+      "epoch": 0.47872,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.202713079088815e-05,
+      "loss": 0.6422,
+      "step": 1870
+    },
+    {
+      "epoch": 0.478976,
+      "grad_norm": 2.625,
+      "learning_rate": 4.202286494326423e-05,
+      "loss": 0.6575,
+      "step": 1871
+    },
+    {
+      "epoch": 0.479232,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.20185990956403e-05,
+      "loss": 0.6337,
+      "step": 1872
+    },
+    {
+      "epoch": 0.479488,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.201433324801638e-05,
+      "loss": 0.6952,
+      "step": 1873
+    },
+    {
+      "epoch": 0.479744,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.201006740039246e-05,
+      "loss": 0.6342,
+      "step": 1874
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.200580155276854e-05,
+      "loss": 0.6252,
+      "step": 1875
+    },
+    {
+      "epoch": 0.480256,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.200153570514461e-05,
+      "loss": 0.636,
+      "step": 1876
+    },
+    {
+      "epoch": 0.480512,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.199726985752069e-05,
+      "loss": 0.6343,
+      "step": 1877
+    },
+    {
+      "epoch": 0.480768,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.199300400989677e-05,
+      "loss": 0.6251,
+      "step": 1878
+    },
+    {
+      "epoch": 0.481024,
+      "grad_norm": 2.375,
+      "learning_rate": 4.198873816227285e-05,
+      "loss": 0.6107,
+      "step": 1879
+    },
+    {
+      "epoch": 0.48128,
+      "grad_norm": 2.25,
+      "learning_rate": 4.198447231464892e-05,
+      "loss": 0.6271,
+      "step": 1880
+    },
+    {
+      "epoch": 0.481536,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.1980206467025003e-05,
+      "loss": 0.6511,
+      "step": 1881
+    },
+    {
+      "epoch": 0.481792,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.197594061940108e-05,
+      "loss": 0.6352,
+      "step": 1882
+    },
+    {
+      "epoch": 0.482048,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.197167477177715e-05,
+      "loss": 0.645,
+      "step": 1883
+    },
+    {
+      "epoch": 0.482304,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.1967408924153226e-05,
+      "loss": 0.6623,
+      "step": 1884
+    },
+    {
+      "epoch": 0.48256,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.196314307652931e-05,
+      "loss": 0.6052,
+      "step": 1885
+    },
+    {
+      "epoch": 0.482816,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.195887722890539e-05,
+      "loss": 0.6342,
+      "step": 1886
+    },
+    {
+      "epoch": 0.483072,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.195461138128146e-05,
+      "loss": 0.664,
+      "step": 1887
+    },
+    {
+      "epoch": 0.483328,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.1950345533657544e-05,
+      "loss": 0.6498,
+      "step": 1888
+    },
+    {
+      "epoch": 0.483584,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.194607968603362e-05,
+      "loss": 0.6219,
+      "step": 1889
+    },
+    {
+      "epoch": 0.48384,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.194181383840969e-05,
+      "loss": 0.6404,
+      "step": 1890
+    },
+    {
+      "epoch": 0.484096,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.1937547990785766e-05,
+      "loss": 0.6275,
+      "step": 1891
+    },
+    {
+      "epoch": 0.484352,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.193328214316185e-05,
+      "loss": 0.6555,
+      "step": 1892
+    },
+    {
+      "epoch": 0.484608,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.192901629553793e-05,
+      "loss": 0.6343,
+      "step": 1893
+    },
+    {
+      "epoch": 0.484864,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.1924750447914e-05,
+      "loss": 0.6536,
+      "step": 1894
+    },
+    {
+      "epoch": 0.48512,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.192048460029008e-05,
+      "loss": 0.6235,
+      "step": 1895
+    },
+    {
+      "epoch": 0.485376,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.191621875266616e-05,
+      "loss": 0.6733,
+      "step": 1896
+    },
+    {
+      "epoch": 0.485632,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.191195290504223e-05,
+      "loss": 0.6115,
+      "step": 1897
+    },
+    {
+      "epoch": 0.485888,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.190768705741831e-05,
+      "loss": 0.6595,
+      "step": 1898
+    },
+    {
+      "epoch": 0.486144,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.190342120979439e-05,
+      "loss": 0.6289,
+      "step": 1899
+    },
+    {
+      "epoch": 0.4864,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.189915536217047e-05,
+      "loss": 0.6683,
+      "step": 1900
+    },
+    {
+      "epoch": 0.486656,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.189488951454654e-05,
+      "loss": 0.6408,
+      "step": 1901
+    },
+    {
+      "epoch": 0.486912,
+      "grad_norm": 2.5,
+      "learning_rate": 4.189062366692262e-05,
+      "loss": 0.6502,
+      "step": 1902
+    },
+    {
+      "epoch": 0.487168,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.18863578192987e-05,
+      "loss": 0.6109,
+      "step": 1903
+    },
+    {
+      "epoch": 0.487424,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.188209197167477e-05,
+      "loss": 0.6807,
+      "step": 1904
+    },
+    {
+      "epoch": 0.48768,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.1877826124050854e-05,
+      "loss": 0.6413,
+      "step": 1905
+    },
+    {
+      "epoch": 0.487936,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.187356027642693e-05,
+      "loss": 0.6531,
+      "step": 1906
+    },
+    {
+      "epoch": 0.488192,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.186929442880301e-05,
+      "loss": 0.6387,
+      "step": 1907
+    },
+    {
+      "epoch": 0.488448,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.186502858117908e-05,
+      "loss": 0.6595,
+      "step": 1908
+    },
+    {
+      "epoch": 0.488704,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.186076273355516e-05,
+      "loss": 0.6183,
+      "step": 1909
+    },
+    {
+      "epoch": 0.48896,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.185649688593123e-05,
+      "loss": 0.6168,
+      "step": 1910
+    },
+    {
+      "epoch": 0.489216,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.185223103830731e-05,
+      "loss": 0.674,
+      "step": 1911
+    },
+    {
+      "epoch": 0.489472,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.1847965190683394e-05,
+      "loss": 0.6295,
+      "step": 1912
+    },
+    {
+      "epoch": 0.489728,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.184369934305947e-05,
+      "loss": 0.6612,
+      "step": 1913
+    },
+    {
+      "epoch": 0.489984,
+      "grad_norm": 2.5,
+      "learning_rate": 4.183943349543554e-05,
+      "loss": 0.6207,
+      "step": 1914
+    },
+    {
+      "epoch": 0.49024,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.183516764781162e-05,
+      "loss": 0.6354,
+      "step": 1915
+    },
+    {
+      "epoch": 0.490496,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.18309018001877e-05,
+      "loss": 0.6517,
+      "step": 1916
+    },
+    {
+      "epoch": 0.490752,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.182663595256377e-05,
+      "loss": 0.6184,
+      "step": 1917
+    },
+    {
+      "epoch": 0.491008,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.182237010493985e-05,
+      "loss": 0.5959,
+      "step": 1918
+    },
+    {
+      "epoch": 0.491264,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.1818104257315934e-05,
+      "loss": 0.6375,
+      "step": 1919
+    },
+    {
+      "epoch": 0.49152,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.181383840969201e-05,
+      "loss": 0.6422,
+      "step": 1920
+    },
+    {
+      "epoch": 0.491776,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.180957256206808e-05,
+      "loss": 0.6417,
+      "step": 1921
+    },
+    {
+      "epoch": 0.492032,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.1805306714444164e-05,
+      "loss": 0.6708,
+      "step": 1922
+    },
+    {
+      "epoch": 0.492288,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.180104086682024e-05,
+      "loss": 0.6688,
+      "step": 1923
+    },
+    {
+      "epoch": 0.492544,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.179677501919632e-05,
+      "loss": 0.6093,
+      "step": 1924
+    },
+    {
+      "epoch": 0.4928,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.179250917157239e-05,
+      "loss": 0.6,
+      "step": 1925
+    },
+    {
+      "epoch": 0.493056,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.1788243323948474e-05,
+      "loss": 0.6448,
+      "step": 1926
+    },
+    {
+      "epoch": 0.493312,
+      "grad_norm": 2.75,
+      "learning_rate": 4.178397747632455e-05,
+      "loss": 0.5946,
+      "step": 1927
+    },
+    {
+      "epoch": 0.493568,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.177971162870062e-05,
+      "loss": 0.6115,
+      "step": 1928
+    },
+    {
+      "epoch": 0.493824,
+      "grad_norm": 2.75,
+      "learning_rate": 4.17754457810767e-05,
+      "loss": 0.6288,
+      "step": 1929
+    },
+    {
+      "epoch": 0.49408,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.177117993345278e-05,
+      "loss": 0.6687,
+      "step": 1930
+    },
+    {
+      "epoch": 0.494336,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.176691408582886e-05,
+      "loss": 0.6401,
+      "step": 1931
+    },
+    {
+      "epoch": 0.494592,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.176264823820493e-05,
+      "loss": 0.6322,
+      "step": 1932
+    },
+    {
+      "epoch": 0.494848,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.175838239058101e-05,
+      "loss": 0.6074,
+      "step": 1933
+    },
+    {
+      "epoch": 0.495104,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.175411654295709e-05,
+      "loss": 0.6243,
+      "step": 1934
+    },
+    {
+      "epoch": 0.49536,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.174985069533316e-05,
+      "loss": 0.6853,
+      "step": 1935
+    },
+    {
+      "epoch": 0.495616,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.174558484770924e-05,
+      "loss": 0.6298,
+      "step": 1936
+    },
+    {
+      "epoch": 0.495872,
+      "grad_norm": 2.625,
+      "learning_rate": 4.1741319000085325e-05,
+      "loss": 0.5976,
+      "step": 1937
+    },
+    {
+      "epoch": 0.496128,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.17370531524614e-05,
+      "loss": 0.6356,
+      "step": 1938
+    },
+    {
+      "epoch": 0.496384,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.1732787304837473e-05,
+      "loss": 0.6111,
+      "step": 1939
+    },
+    {
+      "epoch": 0.49664,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.172852145721355e-05,
+      "loss": 0.6472,
+      "step": 1940
+    },
+    {
+      "epoch": 0.496896,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.172425560958963e-05,
+      "loss": 0.6054,
+      "step": 1941
+    },
+    {
+      "epoch": 0.497152,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.17199897619657e-05,
+      "loss": 0.6193,
+      "step": 1942
+    },
+    {
+      "epoch": 0.497408,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.1715723914341784e-05,
+      "loss": 0.644,
+      "step": 1943
+    },
+    {
+      "epoch": 0.497664,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.171145806671786e-05,
+      "loss": 0.6214,
+      "step": 1944
+    },
+    {
+      "epoch": 0.49792,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.170719221909394e-05,
+      "loss": 0.6042,
+      "step": 1945
+    },
+    {
+      "epoch": 0.498176,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.1702926371470014e-05,
+      "loss": 0.5859,
+      "step": 1946
+    },
+    {
+      "epoch": 0.498432,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.169866052384609e-05,
+      "loss": 0.5836,
+      "step": 1947
+    },
+    {
+      "epoch": 0.498688,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.169439467622216e-05,
+      "loss": 0.6206,
+      "step": 1948
+    },
+    {
+      "epoch": 0.498944,
+      "grad_norm": 2.5,
+      "learning_rate": 4.169012882859824e-05,
+      "loss": 0.64,
+      "step": 1949
+    },
+    {
+      "epoch": 0.4992,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.1685862980974324e-05,
+      "loss": 0.6233,
+      "step": 1950
+    },
+    {
+      "epoch": 0.499456,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.16815971333504e-05,
+      "loss": 0.5805,
+      "step": 1951
+    },
+    {
+      "epoch": 0.499712,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.167733128572648e-05,
+      "loss": 0.6215,
+      "step": 1952
+    },
+    {
+      "epoch": 0.499968,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.1673065438102554e-05,
+      "loss": 0.6343,
+      "step": 1953
+    },
+    {
+      "epoch": 0.500224,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.166879959047863e-05,
+      "loss": 0.6292,
+      "step": 1954
+    },
+    {
+      "epoch": 0.50048,
+      "grad_norm": 2.375,
+      "learning_rate": 4.16645337428547e-05,
+      "loss": 0.624,
+      "step": 1955
+    },
+    {
+      "epoch": 0.500736,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.1660267895230783e-05,
+      "loss": 0.6571,
+      "step": 1956
+    },
+    {
+      "epoch": 0.500992,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.1656002047606864e-05,
+      "loss": 0.6199,
+      "step": 1957
+    },
+    {
+      "epoch": 0.501248,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.165173619998294e-05,
+      "loss": 0.6246,
+      "step": 1958
+    },
+    {
+      "epoch": 0.501504,
+      "grad_norm": 2.75,
+      "learning_rate": 4.164747035235901e-05,
+      "loss": 0.647,
+      "step": 1959
+    },
+    {
+      "epoch": 0.50176,
+      "grad_norm": 2.875,
+      "learning_rate": 4.1643204504735094e-05,
+      "loss": 0.5968,
+      "step": 1960
+    },
+    {
+      "epoch": 0.502016,
+      "grad_norm": 2.5,
+      "learning_rate": 4.163893865711117e-05,
+      "loss": 0.6433,
+      "step": 1961
+    },
+    {
+      "epoch": 0.502272,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.163467280948725e-05,
+      "loss": 0.6462,
+      "step": 1962
+    },
+    {
+      "epoch": 0.502528,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.1630406961863324e-05,
+      "loss": 0.6536,
+      "step": 1963
+    },
+    {
+      "epoch": 0.502784,
+      "grad_norm": 2.5,
+      "learning_rate": 4.1626141114239405e-05,
+      "loss": 0.6222,
+      "step": 1964
+    },
+    {
+      "epoch": 0.50304,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.162187526661548e-05,
+      "loss": 0.6431,
+      "step": 1965
+    },
+    {
+      "epoch": 0.503296,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.161760941899155e-05,
+      "loss": 0.6221,
+      "step": 1966
+    },
+    {
+      "epoch": 0.503552,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.161334357136763e-05,
+      "loss": 0.6208,
+      "step": 1967
+    },
+    {
+      "epoch": 0.503808,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.160907772374371e-05,
+      "loss": 0.6164,
+      "step": 1968
+    },
+    {
+      "epoch": 0.504064,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.160481187611979e-05,
+      "loss": 0.6133,
+      "step": 1969
+    },
+    {
+      "epoch": 0.50432,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.1600546028495864e-05,
+      "loss": 0.6011,
+      "step": 1970
+    },
+    {
+      "epoch": 0.504576,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.1596280180871945e-05,
+      "loss": 0.6766,
+      "step": 1971
+    },
+    {
+      "epoch": 0.504832,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.159201433324802e-05,
+      "loss": 0.6414,
+      "step": 1972
+    },
+    {
+      "epoch": 0.505088,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.1587748485624093e-05,
+      "loss": 0.6311,
+      "step": 1973
+    },
+    {
+      "epoch": 0.505344,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.158348263800017e-05,
+      "loss": 0.6374,
+      "step": 1974
+    },
+    {
+      "epoch": 0.5056,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.157921679037625e-05,
+      "loss": 0.6343,
+      "step": 1975
+    },
+    {
+      "epoch": 0.505856,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.157495094275233e-05,
+      "loss": 0.6041,
+      "step": 1976
+    },
+    {
+      "epoch": 0.506112,
+      "grad_norm": 2.625,
+      "learning_rate": 4.1570685095128404e-05,
+      "loss": 0.6253,
+      "step": 1977
+    },
+    {
+      "epoch": 0.506368,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.156641924750448e-05,
+      "loss": 0.6448,
+      "step": 1978
+    },
+    {
+      "epoch": 0.506624,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.156215339988056e-05,
+      "loss": 0.645,
+      "step": 1979
+    },
+    {
+      "epoch": 0.50688,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.1557887552256634e-05,
+      "loss": 0.6325,
+      "step": 1980
+    },
+    {
+      "epoch": 0.507136,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.1553621704632715e-05,
+      "loss": 0.6484,
+      "step": 1981
+    },
+    {
+      "epoch": 0.507392,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.154935585700879e-05,
+      "loss": 0.642,
+      "step": 1982
+    },
+    {
+      "epoch": 0.507648,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.154509000938487e-05,
+      "loss": 0.634,
+      "step": 1983
+    },
+    {
+      "epoch": 0.507904,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.1540824161760944e-05,
+      "loss": 0.6276,
+      "step": 1984
+    },
+    {
+      "epoch": 0.50816,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.153655831413702e-05,
+      "loss": 0.6227,
+      "step": 1985
+    },
+    {
+      "epoch": 0.508416,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.15322924665131e-05,
+      "loss": 0.6949,
+      "step": 1986
+    },
+    {
+      "epoch": 0.508672,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.1528026618889174e-05,
+      "loss": 0.6146,
+      "step": 1987
+    },
+    {
+      "epoch": 0.508928,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.1523760771265255e-05,
+      "loss": 0.6309,
+      "step": 1988
+    },
+    {
+      "epoch": 0.509184,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.151949492364133e-05,
+      "loss": 0.5969,
+      "step": 1989
+    },
+    {
+      "epoch": 0.50944,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.151522907601741e-05,
+      "loss": 0.6229,
+      "step": 1990
+    },
+    {
+      "epoch": 0.509696,
+      "grad_norm": 2.625,
+      "learning_rate": 4.1510963228393484e-05,
+      "loss": 0.6222,
+      "step": 1991
+    },
+    {
+      "epoch": 0.509952,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.150669738076956e-05,
+      "loss": 0.6194,
+      "step": 1992
+    },
+    {
+      "epoch": 0.510208,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.150243153314563e-05,
+      "loss": 0.6092,
+      "step": 1993
+    },
+    {
+      "epoch": 0.510464,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.1498165685521714e-05,
+      "loss": 0.6541,
+      "step": 1994
+    },
+    {
+      "epoch": 0.51072,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.1493899837897795e-05,
+      "loss": 0.643,
+      "step": 1995
+    },
+    {
+      "epoch": 0.510976,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.148963399027387e-05,
+      "loss": 0.6467,
+      "step": 1996
+    },
+    {
+      "epoch": 0.511232,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.1485368142649944e-05,
+      "loss": 0.6128,
+      "step": 1997
+    },
+    {
+      "epoch": 0.511488,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.1481102295026025e-05,
+      "loss": 0.5968,
+      "step": 1998
+    },
+    {
+      "epoch": 0.511744,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.14768364474021e-05,
+      "loss": 0.6561,
+      "step": 1999
+    },
+    {
+      "epoch": 0.512,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.147257059977818e-05,
+      "loss": 0.6535,
+      "step": 2000
+    },
+    {
+      "epoch": 0.512256,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.146830475215426e-05,
+      "loss": 0.5952,
+      "step": 2001
+    },
+    {
+      "epoch": 0.512512,
+      "grad_norm": 2.375,
+      "learning_rate": 4.1464038904530335e-05,
+      "loss": 0.657,
+      "step": 2002
+    },
+    {
+      "epoch": 0.512768,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.145977305690641e-05,
+      "loss": 0.646,
+      "step": 2003
+    },
+    {
+      "epoch": 0.513024,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.1455507209282484e-05,
+      "loss": 0.6377,
+      "step": 2004
+    },
+    {
+      "epoch": 0.51328,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.1451241361658565e-05,
+      "loss": 0.596,
+      "step": 2005
+    },
+    {
+      "epoch": 0.513536,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.144697551403464e-05,
+      "loss": 0.6233,
+      "step": 2006
+    },
+    {
+      "epoch": 0.513792,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.144270966641072e-05,
+      "loss": 0.6182,
+      "step": 2007
+    },
+    {
+      "epoch": 0.514048,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.1438443818786794e-05,
+      "loss": 0.615,
+      "step": 2008
+    },
+    {
+      "epoch": 0.514304,
+      "grad_norm": 2.5,
+      "learning_rate": 4.1434177971162875e-05,
+      "loss": 0.5712,
+      "step": 2009
+    },
+    {
+      "epoch": 0.51456,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.142991212353895e-05,
+      "loss": 0.647,
+      "step": 2010
+    },
+    {
+      "epoch": 0.514816,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.1425646275915024e-05,
+      "loss": 0.5841,
+      "step": 2011
+    },
+    {
+      "epoch": 0.515072,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.14213804282911e-05,
+      "loss": 0.627,
+      "step": 2012
+    },
+    {
+      "epoch": 0.515328,
+      "grad_norm": 2.875,
+      "learning_rate": 4.141711458066718e-05,
+      "loss": 0.6504,
+      "step": 2013
+    },
+    {
+      "epoch": 0.515584,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.141284873304326e-05,
+      "loss": 0.6294,
+      "step": 2014
+    },
+    {
+      "epoch": 0.51584,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.1408582885419335e-05,
+      "loss": 0.6124,
+      "step": 2015
+    },
+    {
+      "epoch": 0.516096,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.140431703779541e-05,
+      "loss": 0.6154,
+      "step": 2016
+    },
+    {
+      "epoch": 0.516352,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.140005119017149e-05,
+      "loss": 0.6285,
+      "step": 2017
+    },
+    {
+      "epoch": 0.516608,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.1395785342547564e-05,
+      "loss": 0.6201,
+      "step": 2018
+    },
+    {
+      "epoch": 0.516864,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.139151949492364e-05,
+      "loss": 0.5843,
+      "step": 2019
+    },
+    {
+      "epoch": 0.51712,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.1387253647299726e-05,
+      "loss": 0.6199,
+      "step": 2020
+    },
+    {
+      "epoch": 0.517376,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.13829877996758e-05,
+      "loss": 0.6278,
+      "step": 2021
+    },
+    {
+      "epoch": 0.517632,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.1378721952051875e-05,
+      "loss": 0.6232,
+      "step": 2022
+    },
+    {
+      "epoch": 0.517888,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.137445610442795e-05,
+      "loss": 0.6425,
+      "step": 2023
+    },
+    {
+      "epoch": 0.518144,
+      "grad_norm": 2.75,
+      "learning_rate": 4.137019025680403e-05,
+      "loss": 0.6068,
+      "step": 2024
+    },
+    {
+      "epoch": 0.5184,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.1365924409180104e-05,
+      "loss": 0.6475,
+      "step": 2025
+    },
+    {
+      "epoch": 0.518656,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.1361658561556185e-05,
+      "loss": 0.6231,
+      "step": 2026
+    },
+    {
+      "epoch": 0.518912,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.135739271393226e-05,
+      "loss": 0.6212,
+      "step": 2027
+    },
+    {
+      "epoch": 0.519168,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.135312686630834e-05,
+      "loss": 0.6159,
+      "step": 2028
+    },
+    {
+      "epoch": 0.519424,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.1348861018684415e-05,
+      "loss": 0.6388,
+      "step": 2029
+    },
+    {
+      "epoch": 0.51968,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.134459517106049e-05,
+      "loss": 0.6157,
+      "step": 2030
+    },
+    {
+      "epoch": 0.519936,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.1340329323436563e-05,
+      "loss": 0.6173,
+      "step": 2031
+    },
+    {
+      "epoch": 0.520192,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.1336063475812645e-05,
+      "loss": 0.6197,
+      "step": 2032
+    },
+    {
+      "epoch": 0.520448,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.1331797628188726e-05,
+      "loss": 0.6093,
+      "step": 2033
+    },
+    {
+      "epoch": 0.520704,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.13275317805648e-05,
+      "loss": 0.6096,
+      "step": 2034
+    },
+    {
+      "epoch": 0.52096,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.132326593294088e-05,
+      "loss": 0.6041,
+      "step": 2035
+    },
+    {
+      "epoch": 0.521216,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.1319000085316955e-05,
+      "loss": 0.6592,
+      "step": 2036
+    },
+    {
+      "epoch": 0.521472,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.131473423769303e-05,
+      "loss": 0.6518,
+      "step": 2037
+    },
+    {
+      "epoch": 0.521728,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.1310468390069104e-05,
+      "loss": 0.6254,
+      "step": 2038
+    },
+    {
+      "epoch": 0.521984,
+      "grad_norm": 2.75,
+      "learning_rate": 4.130620254244519e-05,
+      "loss": 0.6475,
+      "step": 2039
+    },
+    {
+      "epoch": 0.52224,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.1301936694821266e-05,
+      "loss": 0.6154,
+      "step": 2040
+    },
+    {
+      "epoch": 0.522496,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.129767084719734e-05,
+      "loss": 0.6439,
+      "step": 2041
+    },
+    {
+      "epoch": 0.522752,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.1293404999573414e-05,
+      "loss": 0.6417,
+      "step": 2042
+    },
+    {
+      "epoch": 0.523008,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.1289139151949495e-05,
+      "loss": 0.5916,
+      "step": 2043
+    },
+    {
+      "epoch": 0.523264,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.128487330432557e-05,
+      "loss": 0.6169,
+      "step": 2044
+    },
+    {
+      "epoch": 0.52352,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.128060745670165e-05,
+      "loss": 0.6321,
+      "step": 2045
+    },
+    {
+      "epoch": 0.523776,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.1276341609077725e-05,
+      "loss": 0.586,
+      "step": 2046
+    },
+    {
+      "epoch": 0.524032,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.1272075761453806e-05,
+      "loss": 0.587,
+      "step": 2047
+    },
+    {
+      "epoch": 0.524288,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.126780991382988e-05,
+      "loss": 0.6062,
+      "step": 2048
+    },
+    {
+      "epoch": 0.524544,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.1263544066205954e-05,
+      "loss": 0.6487,
+      "step": 2049
+    },
+    {
+      "epoch": 0.5248,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.125927821858203e-05,
+      "loss": 0.631,
+      "step": 2050
+    },
+    {
+      "epoch": 0.525056,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.125501237095811e-05,
+      "loss": 0.6168,
+      "step": 2051
+    },
+    {
+      "epoch": 0.525312,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.125074652333419e-05,
+      "loss": 0.6255,
+      "step": 2052
+    },
+    {
+      "epoch": 0.525568,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.1246480675710265e-05,
+      "loss": 0.6164,
+      "step": 2053
+    },
+    {
+      "epoch": 0.525824,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.1242214828086346e-05,
+      "loss": 0.638,
+      "step": 2054
+    },
+    {
+      "epoch": 0.52608,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.123794898046242e-05,
+      "loss": 0.6321,
+      "step": 2055
+    },
+    {
+      "epoch": 0.526336,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.1233683132838495e-05,
+      "loss": 0.6168,
+      "step": 2056
+    },
+    {
+      "epoch": 0.526592,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.122941728521457e-05,
+      "loss": 0.581,
+      "step": 2057
+    },
+    {
+      "epoch": 0.526848,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.122515143759065e-05,
+      "loss": 0.6176,
+      "step": 2058
+    },
+    {
+      "epoch": 0.527104,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.122088558996673e-05,
+      "loss": 0.63,
+      "step": 2059
+    },
+    {
+      "epoch": 0.52736,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.1216619742342805e-05,
+      "loss": 0.6103,
+      "step": 2060
+    },
+    {
+      "epoch": 0.527616,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.121235389471888e-05,
+      "loss": 0.6408,
+      "step": 2061
+    },
+    {
+      "epoch": 0.527872,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.120808804709496e-05,
+      "loss": 0.5923,
+      "step": 2062
+    },
+    {
+      "epoch": 0.528128,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.1203822199471035e-05,
+      "loss": 0.5801,
+      "step": 2063
+    },
+    {
+      "epoch": 0.528384,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.1199556351847116e-05,
+      "loss": 0.6259,
+      "step": 2064
+    },
+    {
+      "epoch": 0.52864,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.119529050422319e-05,
+      "loss": 0.6131,
+      "step": 2065
+    },
+    {
+      "epoch": 0.528896,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.119102465659927e-05,
+      "loss": 0.6104,
+      "step": 2066
+    },
+    {
+      "epoch": 0.529152,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.1186758808975345e-05,
+      "loss": 0.5971,
+      "step": 2067
+    },
+    {
+      "epoch": 0.529408,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.118249296135142e-05,
+      "loss": 0.6235,
+      "step": 2068
+    },
+    {
+      "epoch": 0.529664,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.11782271137275e-05,
+      "loss": 0.6444,
+      "step": 2069
+    },
+    {
+      "epoch": 0.52992,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.1173961266103575e-05,
+      "loss": 0.599,
+      "step": 2070
+    },
+    {
+      "epoch": 0.530176,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.1169695418479656e-05,
+      "loss": 0.5807,
+      "step": 2071
+    },
+    {
+      "epoch": 0.530432,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.116542957085573e-05,
+      "loss": 0.6088,
+      "step": 2072
+    },
+    {
+      "epoch": 0.530688,
+      "grad_norm": 2.875,
+      "learning_rate": 4.116116372323181e-05,
+      "loss": 0.6359,
+      "step": 2073
+    },
+    {
+      "epoch": 0.530944,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.1156897875607886e-05,
+      "loss": 0.6069,
+      "step": 2074
+    },
+    {
+      "epoch": 0.5312,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.115263202798396e-05,
+      "loss": 0.6442,
+      "step": 2075
+    },
+    {
+      "epoch": 0.531456,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.1148366180360034e-05,
+      "loss": 0.6293,
+      "step": 2076
+    },
+    {
+      "epoch": 0.531712,
+      "grad_norm": 2.875,
+      "learning_rate": 4.1144100332736115e-05,
+      "loss": 0.6592,
+      "step": 2077
+    },
+    {
+      "epoch": 0.531968,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.1139834485112196e-05,
+      "loss": 0.6319,
+      "step": 2078
+    },
+    {
+      "epoch": 0.532224,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.113556863748827e-05,
+      "loss": 0.6452,
+      "step": 2079
+    },
+    {
+      "epoch": 0.53248,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.1131302789864345e-05,
+      "loss": 0.5807,
+      "step": 2080
+    },
+    {
+      "epoch": 0.532736,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.1127036942240426e-05,
+      "loss": 0.5957,
+      "step": 2081
+    },
+    {
+      "epoch": 0.532992,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.11227710946165e-05,
+      "loss": 0.6092,
+      "step": 2082
+    },
+    {
+      "epoch": 0.533248,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.111850524699258e-05,
+      "loss": 0.5887,
+      "step": 2083
+    },
+    {
+      "epoch": 0.533504,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.111423939936866e-05,
+      "loss": 0.6026,
+      "step": 2084
+    },
+    {
+      "epoch": 0.53376,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.1109973551744736e-05,
+      "loss": 0.6247,
+      "step": 2085
+    },
+    {
+      "epoch": 0.534016,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.110570770412081e-05,
+      "loss": 0.6224,
+      "step": 2086
+    },
+    {
+      "epoch": 0.534272,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.1101441856496885e-05,
+      "loss": 0.5983,
+      "step": 2087
+    },
+    {
+      "epoch": 0.534528,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.1097176008872966e-05,
+      "loss": 0.6103,
+      "step": 2088
+    },
+    {
+      "epoch": 0.534784,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.109291016124904e-05,
+      "loss": 0.6172,
+      "step": 2089
+    },
+    {
+      "epoch": 0.53504,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.108864431362512e-05,
+      "loss": 0.5837,
+      "step": 2090
+    },
+    {
+      "epoch": 0.535296,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.1084378466001196e-05,
+      "loss": 0.6076,
+      "step": 2091
+    },
+    {
+      "epoch": 0.535552,
+      "grad_norm": 2.5,
+      "learning_rate": 4.108011261837728e-05,
+      "loss": 0.6325,
+      "step": 2092
+    },
+    {
+      "epoch": 0.535808,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.107584677075335e-05,
+      "loss": 0.6113,
+      "step": 2093
+    },
+    {
+      "epoch": 0.536064,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.1071580923129425e-05,
+      "loss": 0.6219,
+      "step": 2094
+    },
+    {
+      "epoch": 0.53632,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.10673150755055e-05,
+      "loss": 0.6122,
+      "step": 2095
+    },
+    {
+      "epoch": 0.536576,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.106304922788158e-05,
+      "loss": 0.567,
+      "step": 2096
+    },
+    {
+      "epoch": 0.536832,
+      "grad_norm": 2.75,
+      "learning_rate": 4.105878338025766e-05,
+      "loss": 0.6095,
+      "step": 2097
+    },
+    {
+      "epoch": 0.537088,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.1054517532633736e-05,
+      "loss": 0.5946,
+      "step": 2098
+    },
+    {
+      "epoch": 0.537344,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.105025168500981e-05,
+      "loss": 0.6205,
+      "step": 2099
+    },
+    {
+      "epoch": 0.5376,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.104598583738589e-05,
+      "loss": 0.6124,
+      "step": 2100
+    },
+    {
+      "epoch": 0.537856,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.1041719989761965e-05,
+      "loss": 0.6101,
+      "step": 2101
+    },
+    {
+      "epoch": 0.538112,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.1037454142138046e-05,
+      "loss": 0.5629,
+      "step": 2102
+    },
+    {
+      "epoch": 0.538368,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.103318829451413e-05,
+      "loss": 0.6245,
+      "step": 2103
+    },
+    {
+      "epoch": 0.538624,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.10289224468902e-05,
+      "loss": 0.6179,
+      "step": 2104
+    },
+    {
+      "epoch": 0.53888,
+      "grad_norm": 2.5,
+      "learning_rate": 4.1024656599266276e-05,
+      "loss": 0.5607,
+      "step": 2105
+    },
+    {
+      "epoch": 0.539136,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.102039075164235e-05,
+      "loss": 0.6018,
+      "step": 2106
+    },
+    {
+      "epoch": 0.539392,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.101612490401843e-05,
+      "loss": 0.5782,
+      "step": 2107
+    },
+    {
+      "epoch": 0.539648,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.1011859056394506e-05,
+      "loss": 0.6034,
+      "step": 2108
+    },
+    {
+      "epoch": 0.539904,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.100759320877059e-05,
+      "loss": 0.6055,
+      "step": 2109
+    },
+    {
+      "epoch": 0.54016,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.100332736114666e-05,
+      "loss": 0.6333,
+      "step": 2110
+    },
+    {
+      "epoch": 0.540416,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.099906151352274e-05,
+      "loss": 0.559,
+      "step": 2111
+    },
+    {
+      "epoch": 0.540672,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.0994795665898816e-05,
+      "loss": 0.5789,
+      "step": 2112
+    },
+    {
+      "epoch": 0.540928,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.099052981827489e-05,
+      "loss": 0.6115,
+      "step": 2113
+    },
+    {
+      "epoch": 0.541184,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.0986263970650965e-05,
+      "loss": 0.5696,
+      "step": 2114
+    },
+    {
+      "epoch": 0.54144,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.0981998123027046e-05,
+      "loss": 0.6158,
+      "step": 2115
+    },
+    {
+      "epoch": 0.541696,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.097773227540313e-05,
+      "loss": 0.5661,
+      "step": 2116
+    },
+    {
+      "epoch": 0.541952,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.09734664277792e-05,
+      "loss": 0.61,
+      "step": 2117
+    },
+    {
+      "epoch": 0.542208,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.096920058015528e-05,
+      "loss": 0.5833,
+      "step": 2118
+    },
+    {
+      "epoch": 0.542464,
+      "grad_norm": 2.625,
+      "learning_rate": 4.0964934732531356e-05,
+      "loss": 0.6136,
+      "step": 2119
+    },
+    {
+      "epoch": 0.54272,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.096066888490743e-05,
+      "loss": 0.5829,
+      "step": 2120
+    },
+    {
+      "epoch": 0.542976,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.095640303728351e-05,
+      "loss": 0.5866,
+      "step": 2121
+    },
+    {
+      "epoch": 0.543232,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.095213718965959e-05,
+      "loss": 0.614,
+      "step": 2122
+    },
+    {
+      "epoch": 0.543488,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.094787134203567e-05,
+      "loss": 0.6034,
+      "step": 2123
+    },
+    {
+      "epoch": 0.543744,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.094360549441174e-05,
+      "loss": 0.5657,
+      "step": 2124
+    },
+    {
+      "epoch": 0.544,
+      "grad_norm": 2.5,
+      "learning_rate": 4.0939339646787816e-05,
+      "loss": 0.6144,
+      "step": 2125
+    },
+    {
+      "epoch": 0.544256,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.09350737991639e-05,
+      "loss": 0.6208,
+      "step": 2126
+    },
+    {
+      "epoch": 0.544512,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.093080795153997e-05,
+      "loss": 0.6374,
+      "step": 2127
+    },
+    {
+      "epoch": 0.544768,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.092654210391605e-05,
+      "loss": 0.6011,
+      "step": 2128
+    },
+    {
+      "epoch": 0.545024,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.0922276256292126e-05,
+      "loss": 0.6346,
+      "step": 2129
+    },
+    {
+      "epoch": 0.54528,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.091801040866821e-05,
+      "loss": 0.6084,
+      "step": 2130
+    },
+    {
+      "epoch": 0.545536,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.091374456104428e-05,
+      "loss": 0.6208,
+      "step": 2131
+    },
+    {
+      "epoch": 0.545792,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.0909478713420356e-05,
+      "loss": 0.5819,
+      "step": 2132
+    },
+    {
+      "epoch": 0.546048,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.090521286579644e-05,
+      "loss": 0.6541,
+      "step": 2133
+    },
+    {
+      "epoch": 0.546304,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.090094701817251e-05,
+      "loss": 0.6092,
+      "step": 2134
+    },
+    {
+      "epoch": 0.54656,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.089668117054859e-05,
+      "loss": 0.5951,
+      "step": 2135
+    },
+    {
+      "epoch": 0.546816,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.0892415322924666e-05,
+      "loss": 0.5777,
+      "step": 2136
+    },
+    {
+      "epoch": 0.547072,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.088814947530075e-05,
+      "loss": 0.5995,
+      "step": 2137
+    },
+    {
+      "epoch": 0.547328,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.088388362767682e-05,
+      "loss": 0.5921,
+      "step": 2138
+    },
+    {
+      "epoch": 0.547584,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.0879617780052896e-05,
+      "loss": 0.6081,
+      "step": 2139
+    },
+    {
+      "epoch": 0.54784,
+      "grad_norm": 2.625,
+      "learning_rate": 4.087535193242897e-05,
+      "loss": 0.6259,
+      "step": 2140
+    },
+    {
+      "epoch": 0.548096,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.087108608480506e-05,
+      "loss": 0.6083,
+      "step": 2141
+    },
+    {
+      "epoch": 0.548352,
+      "grad_norm": 2.625,
+      "learning_rate": 4.086682023718113e-05,
+      "loss": 0.6002,
+      "step": 2142
+    },
+    {
+      "epoch": 0.548608,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.0862554389557207e-05,
+      "loss": 0.595,
+      "step": 2143
+    },
+    {
+      "epoch": 0.548864,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.085828854193328e-05,
+      "loss": 0.6173,
+      "step": 2144
+    },
+    {
+      "epoch": 0.54912,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.085402269430936e-05,
+      "loss": 0.6197,
+      "step": 2145
+    },
+    {
+      "epoch": 0.549376,
+      "grad_norm": 2.5,
+      "learning_rate": 4.0849756846685436e-05,
+      "loss": 0.5852,
+      "step": 2146
+    },
+    {
+      "epoch": 0.549632,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.084549099906152e-05,
+      "loss": 0.6228,
+      "step": 2147
+    },
+    {
+      "epoch": 0.549888,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.084122515143759e-05,
+      "loss": 0.5999,
+      "step": 2148
+    },
+    {
+      "epoch": 0.550144,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.083695930381367e-05,
+      "loss": 0.6362,
+      "step": 2149
+    },
+    {
+      "epoch": 0.5504,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.083269345618975e-05,
+      "loss": 0.6257,
+      "step": 2150
+    },
+    {
+      "epoch": 0.550656,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.082842760856582e-05,
+      "loss": 0.6205,
+      "step": 2151
+    },
+    {
+      "epoch": 0.550912,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.08241617609419e-05,
+      "loss": 0.5644,
+      "step": 2152
+    },
+    {
+      "epoch": 0.551168,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.0819895913317976e-05,
+      "loss": 0.593,
+      "step": 2153
+    },
+    {
+      "epoch": 0.551424,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.081563006569406e-05,
+      "loss": 0.5956,
+      "step": 2154
+    },
+    {
+      "epoch": 0.55168,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.081136421807013e-05,
+      "loss": 0.634,
+      "step": 2155
+    },
+    {
+      "epoch": 0.551936,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.080709837044621e-05,
+      "loss": 0.6315,
+      "step": 2156
+    },
+    {
+      "epoch": 0.552192,
+      "grad_norm": 3.0,
+      "learning_rate": 4.080283252282229e-05,
+      "loss": 0.6356,
+      "step": 2157
+    },
+    {
+      "epoch": 0.552448,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.079856667519836e-05,
+      "loss": 0.6057,
+      "step": 2158
+    },
+    {
+      "epoch": 0.552704,
+      "grad_norm": 2.375,
+      "learning_rate": 4.0794300827574435e-05,
+      "loss": 0.5939,
+      "step": 2159
+    },
+    {
+      "epoch": 0.55296,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.0790034979950517e-05,
+      "loss": 0.5972,
+      "step": 2160
+    },
+    {
+      "epoch": 0.553216,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.07857691323266e-05,
+      "loss": 0.6201,
+      "step": 2161
+    },
+    {
+      "epoch": 0.553472,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.078150328470267e-05,
+      "loss": 0.609,
+      "step": 2162
+    },
+    {
+      "epoch": 0.553728,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.0777237437078746e-05,
+      "loss": 0.6141,
+      "step": 2163
+    },
+    {
+      "epoch": 0.553984,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.077297158945483e-05,
+      "loss": 0.6009,
+      "step": 2164
+    },
+    {
+      "epoch": 0.55424,
+      "grad_norm": 2.75,
+      "learning_rate": 4.07687057418309e-05,
+      "loss": 0.5762,
+      "step": 2165
+    },
+    {
+      "epoch": 0.554496,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.076443989420698e-05,
+      "loss": 0.5983,
+      "step": 2166
+    },
+    {
+      "epoch": 0.554752,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.0760174046583064e-05,
+      "loss": 0.6128,
+      "step": 2167
+    },
+    {
+      "epoch": 0.555008,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.075590819895914e-05,
+      "loss": 0.5909,
+      "step": 2168
+    },
+    {
+      "epoch": 0.555264,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.075164235133521e-05,
+      "loss": 0.5565,
+      "step": 2169
+    },
+    {
+      "epoch": 0.55552,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.0747376503711286e-05,
+      "loss": 0.5919,
+      "step": 2170
+    },
+    {
+      "epoch": 0.555776,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.074311065608737e-05,
+      "loss": 0.6091,
+      "step": 2171
+    },
+    {
+      "epoch": 0.556032,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.073884480846344e-05,
+      "loss": 0.5756,
+      "step": 2172
+    },
+    {
+      "epoch": 0.556288,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.073457896083952e-05,
+      "loss": 0.5885,
+      "step": 2173
+    },
+    {
+      "epoch": 0.556544,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.07303131132156e-05,
+      "loss": 0.6218,
+      "step": 2174
+    },
+    {
+      "epoch": 0.5568,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.072604726559168e-05,
+      "loss": 0.5941,
+      "step": 2175
+    },
+    {
+      "epoch": 0.557056,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.072178141796775e-05,
+      "loss": 0.5611,
+      "step": 2176
+    },
+    {
+      "epoch": 0.557312,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.0717515570343826e-05,
+      "loss": 0.6128,
+      "step": 2177
+    },
+    {
+      "epoch": 0.557568,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.07132497227199e-05,
+      "loss": 0.5754,
+      "step": 2178
+    },
+    {
+      "epoch": 0.557824,
+      "grad_norm": 2.25,
+      "learning_rate": 4.070898387509598e-05,
+      "loss": 0.6212,
+      "step": 2179
+    },
+    {
+      "epoch": 0.55808,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.070471802747206e-05,
+      "loss": 0.6069,
+      "step": 2180
+    },
+    {
+      "epoch": 0.558336,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.070045217984814e-05,
+      "loss": 0.5849,
+      "step": 2181
+    },
+    {
+      "epoch": 0.558592,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.069618633222422e-05,
+      "loss": 0.6265,
+      "step": 2182
+    },
+    {
+      "epoch": 0.558848,
+      "grad_norm": 2.5,
+      "learning_rate": 4.069192048460029e-05,
+      "loss": 0.5488,
+      "step": 2183
+    },
+    {
+      "epoch": 0.559104,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.068765463697637e-05,
+      "loss": 0.5929,
+      "step": 2184
+    },
+    {
+      "epoch": 0.55936,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.068338878935245e-05,
+      "loss": 0.605,
+      "step": 2185
+    },
+    {
+      "epoch": 0.559616,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.067912294172853e-05,
+      "loss": 0.5917,
+      "step": 2186
+    },
+    {
+      "epoch": 0.559872,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.06748570941046e-05,
+      "loss": 0.5963,
+      "step": 2187
+    },
+    {
+      "epoch": 0.560128,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.067059124648068e-05,
+      "loss": 0.6183,
+      "step": 2188
+    },
+    {
+      "epoch": 0.560384,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.066632539885675e-05,
+      "loss": 0.6276,
+      "step": 2189
+    },
+    {
+      "epoch": 0.56064,
+      "grad_norm": 2.5,
+      "learning_rate": 4.066205955123283e-05,
+      "loss": 0.5925,
+      "step": 2190
+    },
+    {
+      "epoch": 0.560896,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.065779370360891e-05,
+      "loss": 0.5758,
+      "step": 2191
+    },
+    {
+      "epoch": 0.561152,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.065352785598499e-05,
+      "loss": 0.5661,
+      "step": 2192
+    },
+    {
+      "epoch": 0.561408,
+      "grad_norm": 2.375,
+      "learning_rate": 4.064926200836106e-05,
+      "loss": 0.5846,
+      "step": 2193
+    },
+    {
+      "epoch": 0.561664,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.064499616073714e-05,
+      "loss": 0.5813,
+      "step": 2194
+    },
+    {
+      "epoch": 0.56192,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.064073031311322e-05,
+      "loss": 0.6121,
+      "step": 2195
+    },
+    {
+      "epoch": 0.562176,
+      "grad_norm": 2.375,
+      "learning_rate": 4.063646446548929e-05,
+      "loss": 0.635,
+      "step": 2196
+    },
+    {
+      "epoch": 0.562432,
+      "grad_norm": 2.25,
+      "learning_rate": 4.0632198617865366e-05,
+      "loss": 0.5764,
+      "step": 2197
+    },
+    {
+      "epoch": 0.562688,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.062793277024145e-05,
+      "loss": 0.6104,
+      "step": 2198
+    },
+    {
+      "epoch": 0.562944,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.062366692261753e-05,
+      "loss": 0.6141,
+      "step": 2199
+    },
+    {
+      "epoch": 0.5632,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.06194010749936e-05,
+      "loss": 0.6257,
+      "step": 2200
+    },
+    {
+      "epoch": 0.563456,
+      "grad_norm": 2.25,
+      "learning_rate": 4.0615135227369683e-05,
+      "loss": 0.6254,
+      "step": 2201
+    },
+    {
+      "epoch": 0.563712,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.061086937974576e-05,
+      "loss": 0.5639,
+      "step": 2202
+    },
+    {
+      "epoch": 0.563968,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.060660353212183e-05,
+      "loss": 0.5829,
+      "step": 2203
+    },
+    {
+      "epoch": 0.564224,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.060233768449791e-05,
+      "loss": 0.639,
+      "step": 2204
+    },
+    {
+      "epoch": 0.56448,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.0598071836873994e-05,
+      "loss": 0.5741,
+      "step": 2205
+    },
+    {
+      "epoch": 0.564736,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.059380598925007e-05,
+      "loss": 0.5762,
+      "step": 2206
+    },
+    {
+      "epoch": 0.564992,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.058954014162614e-05,
+      "loss": 0.6176,
+      "step": 2207
+    },
+    {
+      "epoch": 0.565248,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.058527429400222e-05,
+      "loss": 0.5757,
+      "step": 2208
+    },
+    {
+      "epoch": 0.565504,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.05810084463783e-05,
+      "loss": 0.5458,
+      "step": 2209
+    },
+    {
+      "epoch": 0.56576,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.057674259875437e-05,
+      "loss": 0.5714,
+      "step": 2210
+    },
+    {
+      "epoch": 0.566016,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.057247675113045e-05,
+      "loss": 0.5816,
+      "step": 2211
+    },
+    {
+      "epoch": 0.566272,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.056821090350653e-05,
+      "loss": 0.6452,
+      "step": 2212
+    },
+    {
+      "epoch": 0.566528,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.056394505588261e-05,
+      "loss": 0.5882,
+      "step": 2213
+    },
+    {
+      "epoch": 0.566784,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.055967920825868e-05,
+      "loss": 0.6298,
+      "step": 2214
+    },
+    {
+      "epoch": 0.56704,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.055541336063476e-05,
+      "loss": 0.614,
+      "step": 2215
+    },
+    {
+      "epoch": 0.567296,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.055114751301084e-05,
+      "loss": 0.5986,
+      "step": 2216
+    },
+    {
+      "epoch": 0.567552,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.054688166538691e-05,
+      "loss": 0.5876,
+      "step": 2217
+    },
+    {
+      "epoch": 0.567808,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.0542615817762993e-05,
+      "loss": 0.5728,
+      "step": 2218
+    },
+    {
+      "epoch": 0.568064,
+      "grad_norm": 2.75,
+      "learning_rate": 4.053834997013907e-05,
+      "loss": 0.5979,
+      "step": 2219
+    },
+    {
+      "epoch": 0.56832,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.053408412251515e-05,
+      "loss": 0.5913,
+      "step": 2220
+    },
+    {
+      "epoch": 0.568576,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.052981827489122e-05,
+      "loss": 0.5902,
+      "step": 2221
+    },
+    {
+      "epoch": 0.568832,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.05255524272673e-05,
+      "loss": 0.5861,
+      "step": 2222
+    },
+    {
+      "epoch": 0.569088,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.052128657964338e-05,
+      "loss": 0.6083,
+      "step": 2223
+    },
+    {
+      "epoch": 0.569344,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.051702073201946e-05,
+      "loss": 0.6329,
+      "step": 2224
+    },
+    {
+      "epoch": 0.5696,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.0512754884395534e-05,
+      "loss": 0.5515,
+      "step": 2225
+    },
+    {
+      "epoch": 0.569856,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.050848903677161e-05,
+      "loss": 0.613,
+      "step": 2226
+    },
+    {
+      "epoch": 0.570112,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.050422318914768e-05,
+      "loss": 0.6245,
+      "step": 2227
+    },
+    {
+      "epoch": 0.570368,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.049995734152376e-05,
+      "loss": 0.6026,
+      "step": 2228
+    },
+    {
+      "epoch": 0.570624,
+      "grad_norm": 2.375,
+      "learning_rate": 4.049569149389984e-05,
+      "loss": 0.6102,
+      "step": 2229
+    },
+    {
+      "epoch": 0.57088,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.049142564627592e-05,
+      "loss": 0.5679,
+      "step": 2230
+    },
+    {
+      "epoch": 0.571136,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.0487159798652e-05,
+      "loss": 0.6142,
+      "step": 2231
+    },
+    {
+      "epoch": 0.571392,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.0482893951028074e-05,
+      "loss": 0.5921,
+      "step": 2232
+    },
+    {
+      "epoch": 0.571648,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.047862810340415e-05,
+      "loss": 0.5555,
+      "step": 2233
+    },
+    {
+      "epoch": 0.571904,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.047436225578022e-05,
+      "loss": 0.5706,
+      "step": 2234
+    },
+    {
+      "epoch": 0.57216,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.04700964081563e-05,
+      "loss": 0.6195,
+      "step": 2235
+    },
+    {
+      "epoch": 0.572416,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.046583056053238e-05,
+      "loss": 0.5959,
+      "step": 2236
+    },
+    {
+      "epoch": 0.572672,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.046156471290846e-05,
+      "loss": 0.5811,
+      "step": 2237
+    },
+    {
+      "epoch": 0.572928,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.045729886528453e-05,
+      "loss": 0.5581,
+      "step": 2238
+    },
+    {
+      "epoch": 0.573184,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.0453033017660614e-05,
+      "loss": 0.624,
+      "step": 2239
+    },
+    {
+      "epoch": 0.57344,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.044876717003669e-05,
+      "loss": 0.5999,
+      "step": 2240
+    },
+    {
+      "epoch": 0.573696,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.044450132241276e-05,
+      "loss": 0.6287,
+      "step": 2241
+    },
+    {
+      "epoch": 0.573952,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.044023547478884e-05,
+      "loss": 0.597,
+      "step": 2242
+    },
+    {
+      "epoch": 0.574208,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.0435969627164925e-05,
+      "loss": 0.6172,
+      "step": 2243
+    },
+    {
+      "epoch": 0.574464,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.0431703779541e-05,
+      "loss": 0.6463,
+      "step": 2244
+    },
+    {
+      "epoch": 0.57472,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.042743793191707e-05,
+      "loss": 0.5702,
+      "step": 2245
+    },
+    {
+      "epoch": 0.574976,
+      "grad_norm": 2.875,
+      "learning_rate": 4.042317208429315e-05,
+      "loss": 0.5661,
+      "step": 2246
+    },
+    {
+      "epoch": 0.575232,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.041890623666923e-05,
+      "loss": 0.5685,
+      "step": 2247
+    },
+    {
+      "epoch": 0.575488,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.04146403890453e-05,
+      "loss": 0.5572,
+      "step": 2248
+    },
+    {
+      "epoch": 0.575744,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.0410374541421384e-05,
+      "loss": 0.5662,
+      "step": 2249
+    },
+    {
+      "epoch": 0.576,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.0406108693797465e-05,
+      "loss": 0.5978,
+      "step": 2250
+    },
+    {
+      "epoch": 0.576256,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.040184284617354e-05,
+      "loss": 0.5285,
+      "step": 2251
+    },
+    {
+      "epoch": 0.576512,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.039757699854961e-05,
+      "loss": 0.5612,
+      "step": 2252
+    },
+    {
+      "epoch": 0.576768,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.039331115092569e-05,
+      "loss": 0.6134,
+      "step": 2253
+    },
+    {
+      "epoch": 0.577024,
+      "grad_norm": 2.25,
+      "learning_rate": 4.038904530330177e-05,
+      "loss": 0.5591,
+      "step": 2254
+    },
+    {
+      "epoch": 0.57728,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.038477945567784e-05,
+      "loss": 0.5823,
+      "step": 2255
+    },
+    {
+      "epoch": 0.577536,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.0380513608053924e-05,
+      "loss": 0.5999,
+      "step": 2256
+    },
+    {
+      "epoch": 0.577792,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.037624776043e-05,
+      "loss": 0.601,
+      "step": 2257
+    },
+    {
+      "epoch": 0.578048,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.037198191280608e-05,
+      "loss": 0.6257,
+      "step": 2258
+    },
+    {
+      "epoch": 0.578304,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.0367716065182154e-05,
+      "loss": 0.5794,
+      "step": 2259
+    },
+    {
+      "epoch": 0.57856,
+      "grad_norm": 2.5,
+      "learning_rate": 4.036345021755823e-05,
+      "loss": 0.603,
+      "step": 2260
+    },
+    {
+      "epoch": 0.578816,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.03591843699343e-05,
+      "loss": 0.6427,
+      "step": 2261
+    },
+    {
+      "epoch": 0.579072,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.035491852231039e-05,
+      "loss": 0.5892,
+      "step": 2262
+    },
+    {
+      "epoch": 0.579328,
+      "grad_norm": 2.625,
+      "learning_rate": 4.0350652674686464e-05,
+      "loss": 0.5939,
+      "step": 2263
+    },
+    {
+      "epoch": 0.579584,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.034638682706254e-05,
+      "loss": 0.5895,
+      "step": 2264
+    },
+    {
+      "epoch": 0.57984,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.034212097943862e-05,
+      "loss": 0.5917,
+      "step": 2265
+    },
+    {
+      "epoch": 0.580096,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.0337855131814694e-05,
+      "loss": 0.5591,
+      "step": 2266
+    },
+    {
+      "epoch": 0.580352,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.033358928419077e-05,
+      "loss": 0.5995,
+      "step": 2267
+    },
+    {
+      "epoch": 0.580608,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.032932343656685e-05,
+      "loss": 0.5721,
+      "step": 2268
+    },
+    {
+      "epoch": 0.580864,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.032505758894293e-05,
+      "loss": 0.5934,
+      "step": 2269
+    },
+    {
+      "epoch": 0.58112,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.0320791741319004e-05,
+      "loss": 0.5933,
+      "step": 2270
+    },
+    {
+      "epoch": 0.581376,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.031652589369508e-05,
+      "loss": 0.5547,
+      "step": 2271
+    },
+    {
+      "epoch": 0.581632,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.031226004607115e-05,
+      "loss": 0.5558,
+      "step": 2272
+    },
+    {
+      "epoch": 0.581888,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.0307994198447234e-05,
+      "loss": 0.6223,
+      "step": 2273
+    },
+    {
+      "epoch": 0.582144,
+      "grad_norm": 2.75,
+      "learning_rate": 4.030372835082331e-05,
+      "loss": 0.615,
+      "step": 2274
+    },
+    {
+      "epoch": 0.5824,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.029946250319939e-05,
+      "loss": 0.5837,
+      "step": 2275
+    },
+    {
+      "epoch": 0.582656,
+      "grad_norm": 2.375,
+      "learning_rate": 4.0295196655575463e-05,
+      "loss": 0.6035,
+      "step": 2276
+    },
+    {
+      "epoch": 0.582912,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.0290930807951545e-05,
+      "loss": 0.5751,
+      "step": 2277
+    },
+    {
+      "epoch": 0.583168,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.028666496032762e-05,
+      "loss": 0.6039,
+      "step": 2278
+    },
+    {
+      "epoch": 0.583424,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.028239911270369e-05,
+      "loss": 0.5806,
+      "step": 2279
+    },
+    {
+      "epoch": 0.58368,
+      "grad_norm": 2.375,
+      "learning_rate": 4.0278133265079774e-05,
+      "loss": 0.596,
+      "step": 2280
+    },
+    {
+      "epoch": 0.583936,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.027386741745585e-05,
+      "loss": 0.5517,
+      "step": 2281
+    },
+    {
+      "epoch": 0.584192,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.026960156983193e-05,
+      "loss": 0.5773,
+      "step": 2282
+    },
+    {
+      "epoch": 0.584448,
+      "grad_norm": 2.625,
+      "learning_rate": 4.0265335722208004e-05,
+      "loss": 0.554,
+      "step": 2283
+    },
+    {
+      "epoch": 0.584704,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.0261069874584085e-05,
+      "loss": 0.5729,
+      "step": 2284
+    },
+    {
+      "epoch": 0.58496,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.025680402696016e-05,
+      "loss": 0.5615,
+      "step": 2285
+    },
+    {
+      "epoch": 0.585216,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.025253817933623e-05,
+      "loss": 0.5729,
+      "step": 2286
+    },
+    {
+      "epoch": 0.585472,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.0248272331712314e-05,
+      "loss": 0.5828,
+      "step": 2287
+    },
+    {
+      "epoch": 0.585728,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.0244006484088395e-05,
+      "loss": 0.5852,
+      "step": 2288
+    },
+    {
+      "epoch": 0.585984,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.023974063646447e-05,
+      "loss": 0.6054,
+      "step": 2289
+    },
+    {
+      "epoch": 0.58624,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.0235474788840544e-05,
+      "loss": 0.5738,
+      "step": 2290
+    },
+    {
+      "epoch": 0.586496,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.023120894121662e-05,
+      "loss": 0.6111,
+      "step": 2291
+    },
+    {
+      "epoch": 0.586752,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.02269430935927e-05,
+      "loss": 0.6044,
+      "step": 2292
+    },
+    {
+      "epoch": 0.587008,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.0222677245968773e-05,
+      "loss": 0.5646,
+      "step": 2293
+    },
+    {
+      "epoch": 0.587264,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.0218411398344854e-05,
+      "loss": 0.6399,
+      "step": 2294
+    },
+    {
+      "epoch": 0.58752,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.021414555072093e-05,
+      "loss": 0.5913,
+      "step": 2295
+    },
+    {
+      "epoch": 0.587776,
+      "grad_norm": 2.5,
+      "learning_rate": 4.020987970309701e-05,
+      "loss": 0.5603,
+      "step": 2296
+    },
+    {
+      "epoch": 0.588032,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.0205613855473084e-05,
+      "loss": 0.5816,
+      "step": 2297
+    },
+    {
+      "epoch": 0.588288,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.020134800784916e-05,
+      "loss": 0.5914,
+      "step": 2298
+    },
+    {
+      "epoch": 0.588544,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.019708216022524e-05,
+      "loss": 0.5752,
+      "step": 2299
+    },
+    {
+      "epoch": 0.5888,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.0192816312601314e-05,
+      "loss": 0.5806,
+      "step": 2300
+    },
+    {
+      "epoch": 0.589056,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.0188550464977395e-05,
+      "loss": 0.6106,
+      "step": 2301
+    },
+    {
+      "epoch": 0.589312,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.018428461735347e-05,
+      "loss": 0.6221,
+      "step": 2302
+    },
+    {
+      "epoch": 0.589568,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.018001876972955e-05,
+      "loss": 0.5618,
+      "step": 2303
+    },
+    {
+      "epoch": 0.589824,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.0175752922105624e-05,
+      "loss": 0.5765,
+      "step": 2304
+    },
+    {
+      "epoch": 0.59008,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.01714870744817e-05,
+      "loss": 0.61,
+      "step": 2305
+    },
+    {
+      "epoch": 0.590336,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.016722122685778e-05,
+      "loss": 0.6112,
+      "step": 2306
+    },
+    {
+      "epoch": 0.590592,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.016295537923386e-05,
+      "loss": 0.5508,
+      "step": 2307
+    },
+    {
+      "epoch": 0.590848,
+      "grad_norm": 2.375,
+      "learning_rate": 4.0158689531609935e-05,
+      "loss": 0.5938,
+      "step": 2308
+    },
+    {
+      "epoch": 0.591104,
+      "grad_norm": 2.5,
+      "learning_rate": 4.015442368398601e-05,
+      "loss": 0.5666,
+      "step": 2309
+    },
+    {
+      "epoch": 0.59136,
+      "grad_norm": 2.5,
+      "learning_rate": 4.0150157836362083e-05,
+      "loss": 0.5925,
+      "step": 2310
+    },
+    {
+      "epoch": 0.591616,
+      "grad_norm": 2.375,
+      "learning_rate": 4.0145891988738164e-05,
+      "loss": 0.5963,
+      "step": 2311
+    },
+    {
+      "epoch": 0.591872,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.014162614111424e-05,
+      "loss": 0.5818,
+      "step": 2312
+    },
+    {
+      "epoch": 0.592128,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.013736029349032e-05,
+      "loss": 0.6328,
+      "step": 2313
+    },
+    {
+      "epoch": 0.592384,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.01330944458664e-05,
+      "loss": 0.5508,
+      "step": 2314
+    },
+    {
+      "epoch": 0.59264,
+      "grad_norm": 2.5,
+      "learning_rate": 4.0128828598242475e-05,
+      "loss": 0.5696,
+      "step": 2315
+    },
+    {
+      "epoch": 0.592896,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.012456275061855e-05,
+      "loss": 0.5356,
+      "step": 2316
+    },
+    {
+      "epoch": 0.593152,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.0120296902994624e-05,
+      "loss": 0.5619,
+      "step": 2317
+    },
+    {
+      "epoch": 0.593408,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.0116031055370705e-05,
+      "loss": 0.5945,
+      "step": 2318
+    },
+    {
+      "epoch": 0.593664,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.011176520774678e-05,
+      "loss": 0.5711,
+      "step": 2319
+    },
+    {
+      "epoch": 0.59392,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.010749936012286e-05,
+      "loss": 0.6119,
+      "step": 2320
+    },
+    {
+      "epoch": 0.594176,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.0103233512498934e-05,
+      "loss": 0.6085,
+      "step": 2321
+    },
+    {
+      "epoch": 0.594432,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.0098967664875015e-05,
+      "loss": 0.626,
+      "step": 2322
+    },
+    {
+      "epoch": 0.594688,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.009470181725109e-05,
+      "loss": 0.6014,
+      "step": 2323
+    },
+    {
+      "epoch": 0.594944,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.0090435969627164e-05,
+      "loss": 0.562,
+      "step": 2324
+    },
+    {
+      "epoch": 0.5952,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.0086170122003245e-05,
+      "loss": 0.6355,
+      "step": 2325
+    },
+    {
+      "epoch": 0.595456,
+      "grad_norm": 2.375,
+      "learning_rate": 4.0081904274379326e-05,
+      "loss": 0.5481,
+      "step": 2326
+    },
+    {
+      "epoch": 0.595712,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.00776384267554e-05,
+      "loss": 0.5617,
+      "step": 2327
+    },
+    {
+      "epoch": 0.595968,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.0073372579131474e-05,
+      "loss": 0.5375,
+      "step": 2328
+    },
+    {
+      "epoch": 0.596224,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.0069106731507555e-05,
+      "loss": 0.5933,
+      "step": 2329
+    },
+    {
+      "epoch": 0.59648,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.006484088388363e-05,
+      "loss": 0.6097,
+      "step": 2330
+    },
+    {
+      "epoch": 0.596736,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.0060575036259704e-05,
+      "loss": 0.5946,
+      "step": 2331
+    },
+    {
+      "epoch": 0.596992,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.0056309188635785e-05,
+      "loss": 0.5686,
+      "step": 2332
+    },
+    {
+      "epoch": 0.597248,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.0052043341011866e-05,
+      "loss": 0.5616,
+      "step": 2333
+    },
+    {
+      "epoch": 0.597504,
+      "grad_norm": 2.5,
+      "learning_rate": 4.004777749338794e-05,
+      "loss": 0.5715,
+      "step": 2334
+    },
+    {
+      "epoch": 0.59776,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.0043511645764015e-05,
+      "loss": 0.581,
+      "step": 2335
+    },
+    {
+      "epoch": 0.598016,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.003924579814009e-05,
+      "loss": 0.5688,
+      "step": 2336
+    },
+    {
+      "epoch": 0.598272,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.003497995051617e-05,
+      "loss": 0.5826,
+      "step": 2337
+    },
+    {
+      "epoch": 0.598528,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.0030714102892244e-05,
+      "loss": 0.5665,
+      "step": 2338
+    },
+    {
+      "epoch": 0.598784,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.0026448255268325e-05,
+      "loss": 0.6086,
+      "step": 2339
+    },
+    {
+      "epoch": 0.59904,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.00221824076444e-05,
+      "loss": 0.5762,
+      "step": 2340
+    },
+    {
+      "epoch": 0.599296,
+      "grad_norm": 2.25,
+      "learning_rate": 4.001791656002048e-05,
+      "loss": 0.6006,
+      "step": 2341
+    },
+    {
+      "epoch": 0.599552,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.0013650712396555e-05,
+      "loss": 0.5684,
+      "step": 2342
+    },
+    {
+      "epoch": 0.599808,
+      "grad_norm": 2.5,
+      "learning_rate": 4.000938486477263e-05,
+      "loss": 0.5838,
+      "step": 2343
+    },
+    {
+      "epoch": 0.600064,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.00051190171487e-05,
+      "loss": 0.6006,
+      "step": 2344
+    },
+    {
+      "epoch": 0.60032,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.000085316952479e-05,
+      "loss": 0.5767,
+      "step": 2345
+    },
+    {
+      "epoch": 0.600576,
+      "grad_norm": 2.5,
+      "learning_rate": 3.9996587321900865e-05,
+      "loss": 0.5795,
+      "step": 2346
+    },
+    {
+      "epoch": 0.600832,
+      "grad_norm": 2.8125,
+      "learning_rate": 3.999232147427694e-05,
+      "loss": 0.5873,
+      "step": 2347
+    },
+    {
+      "epoch": 0.601088,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.998805562665302e-05,
+      "loss": 0.5657,
+      "step": 2348
+    },
+    {
+      "epoch": 0.601344,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.9983789779029095e-05,
+      "loss": 0.5574,
+      "step": 2349
+    },
+    {
+      "epoch": 0.6016,
+      "grad_norm": 2.625,
+      "learning_rate": 3.997952393140517e-05,
+      "loss": 0.5694,
+      "step": 2350
+    },
+    {
+      "epoch": 0.601856,
+      "grad_norm": 2.375,
+      "learning_rate": 3.997525808378125e-05,
+      "loss": 0.5947,
+      "step": 2351
+    },
+    {
+      "epoch": 0.602112,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.997099223615733e-05,
+      "loss": 0.5917,
+      "step": 2352
+    },
+    {
+      "epoch": 0.602368,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.9966726388533406e-05,
+      "loss": 0.5688,
+      "step": 2353
+    },
+    {
+      "epoch": 0.602624,
+      "grad_norm": 2.703125,
+      "learning_rate": 3.996246054090948e-05,
+      "loss": 0.5343,
+      "step": 2354
+    },
+    {
+      "epoch": 0.60288,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.9958194693285554e-05,
+      "loss": 0.6082,
+      "step": 2355
+    },
+    {
+      "epoch": 0.603136,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.9953928845661635e-05,
+      "loss": 0.5757,
+      "step": 2356
+    },
+    {
+      "epoch": 0.603392,
+      "grad_norm": 2.828125,
+      "learning_rate": 3.994966299803771e-05,
+      "loss": 0.5885,
+      "step": 2357
+    },
+    {
+      "epoch": 0.603648,
+      "grad_norm": 2.8125,
+      "learning_rate": 3.994539715041379e-05,
+      "loss": 0.6081,
+      "step": 2358
+    },
+    {
+      "epoch": 0.603904,
+      "grad_norm": 2.5,
+      "learning_rate": 3.9941131302789865e-05,
+      "loss": 0.552,
+      "step": 2359
+    },
+    {
+      "epoch": 0.60416,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.9936865455165946e-05,
+      "loss": 0.5642,
+      "step": 2360
+    },
+    {
+      "epoch": 0.604416,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.993259960754202e-05,
+      "loss": 0.5694,
+      "step": 2361
+    },
+    {
+      "epoch": 0.604672,
+      "grad_norm": 2.78125,
+      "learning_rate": 3.9928333759918094e-05,
+      "loss": 0.6015,
+      "step": 2362
+    },
+    {
+      "epoch": 0.604928,
+      "grad_norm": 2.609375,
+      "learning_rate": 3.9924067912294175e-05,
+      "loss": 0.6011,
+      "step": 2363
+    },
+    {
+      "epoch": 0.605184,
+      "grad_norm": 2.375,
+      "learning_rate": 3.9919802064670256e-05,
+      "loss": 0.5708,
+      "step": 2364
+    },
+    {
+      "epoch": 0.60544,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.991553621704633e-05,
+      "loss": 0.5928,
+      "step": 2365
+    },
+    {
+      "epoch": 0.605696,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.9911270369422405e-05,
+      "loss": 0.6214,
+      "step": 2366
+    },
+    {
+      "epoch": 0.605952,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.9907004521798486e-05,
+      "loss": 0.5621,
+      "step": 2367
+    },
+    {
+      "epoch": 0.606208,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.990273867417456e-05,
+      "loss": 0.5855,
+      "step": 2368
+    },
+    {
+      "epoch": 0.606464,
+      "grad_norm": 2.59375,
+      "learning_rate": 3.9898472826550635e-05,
+      "loss": 0.6297,
+      "step": 2369
+    },
+    {
+      "epoch": 0.60672,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.9894206978926716e-05,
+      "loss": 0.5471,
+      "step": 2370
+    },
+    {
+      "epoch": 0.606976,
+      "grad_norm": 2.6875,
+      "learning_rate": 3.9889941131302797e-05,
+      "loss": 0.5847,
+      "step": 2371
+    },
+    {
+      "epoch": 0.607232,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.988567528367887e-05,
+      "loss": 0.5798,
+      "step": 2372
+    },
+    {
+      "epoch": 0.607488,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.9881409436054945e-05,
+      "loss": 0.5847,
+      "step": 2373
+    },
+    {
+      "epoch": 0.607744,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.987714358843102e-05,
+      "loss": 0.5759,
+      "step": 2374
+    },
+    {
+      "epoch": 0.608,
+      "grad_norm": 2.5,
+      "learning_rate": 3.98728777408071e-05,
+      "loss": 0.5776,
+      "step": 2375
+    },
+    {
+      "epoch": 0.608256,
+      "grad_norm": 2.640625,
+      "learning_rate": 3.9868611893183175e-05,
+      "loss": 0.5692,
+      "step": 2376
+    },
+    {
+      "epoch": 0.608512,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.9864346045559256e-05,
+      "loss": 0.6022,
+      "step": 2377
+    },
+    {
+      "epoch": 0.608768,
+      "grad_norm": 2.640625,
+      "learning_rate": 3.986008019793534e-05,
+      "loss": 0.6451,
+      "step": 2378
+    },
+    {
+      "epoch": 0.609024,
+      "grad_norm": 2.59375,
+      "learning_rate": 3.985581435031141e-05,
+      "loss": 0.5759,
+      "step": 2379
+    },
+    {
+      "epoch": 0.60928,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.9851548502687485e-05,
+      "loss": 0.5924,
+      "step": 2380
+    },
+    {
+      "epoch": 0.609536,
+      "grad_norm": 2.890625,
+      "learning_rate": 3.984728265506356e-05,
+      "loss": 0.57,
+      "step": 2381
+    },
+    {
+      "epoch": 0.609792,
+      "grad_norm": 2.640625,
+      "learning_rate": 3.984301680743964e-05,
+      "loss": 0.5794,
+      "step": 2382
+    },
+    {
+      "epoch": 0.610048,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.9838750959815715e-05,
+      "loss": 0.5611,
+      "step": 2383
+    },
+    {
+      "epoch": 0.610304,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.9834485112191796e-05,
+      "loss": 0.5401,
+      "step": 2384
+    },
+    {
+      "epoch": 0.61056,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.983021926456787e-05,
+      "loss": 0.5577,
+      "step": 2385
+    },
+    {
+      "epoch": 0.610816,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.982595341694395e-05,
+      "loss": 0.6039,
+      "step": 2386
+    },
+    {
+      "epoch": 0.611072,
+      "grad_norm": 2.5,
+      "learning_rate": 3.9821687569320026e-05,
+      "loss": 0.5836,
+      "step": 2387
+    },
+    {
+      "epoch": 0.611328,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.98174217216961e-05,
+      "loss": 0.5927,
+      "step": 2388
+    },
+    {
+      "epoch": 0.611584,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.981315587407218e-05,
+      "loss": 0.5889,
+      "step": 2389
+    },
+    {
+      "epoch": 0.61184,
+      "grad_norm": 2.578125,
+      "learning_rate": 3.980889002644826e-05,
+      "loss": 0.5498,
+      "step": 2390
+    },
+    {
+      "epoch": 0.612096,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.9804624178824336e-05,
+      "loss": 0.5792,
+      "step": 2391
+    },
+    {
+      "epoch": 0.612352,
+      "grad_norm": 2.65625,
+      "learning_rate": 3.980035833120041e-05,
+      "loss": 0.5445,
+      "step": 2392
+    },
+    {
+      "epoch": 0.612608,
+      "grad_norm": 2.625,
+      "learning_rate": 3.9796092483576485e-05,
+      "loss": 0.5562,
+      "step": 2393
+    },
+    {
+      "epoch": 0.612864,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.9791826635952566e-05,
+      "loss": 0.5643,
+      "step": 2394
+    },
+    {
+      "epoch": 0.61312,
+      "grad_norm": 2.65625,
+      "learning_rate": 3.978756078832864e-05,
+      "loss": 0.5344,
+      "step": 2395
+    },
+    {
+      "epoch": 0.613376,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.978329494070472e-05,
+      "loss": 0.5647,
+      "step": 2396
+    },
+    {
+      "epoch": 0.613632,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.97790290930808e-05,
+      "loss": 0.571,
+      "step": 2397
+    },
+    {
+      "epoch": 0.613888,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.9774763245456876e-05,
+      "loss": 0.5826,
+      "step": 2398
+    },
+    {
+      "epoch": 0.614144,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.977049739783295e-05,
+      "loss": 0.5704,
+      "step": 2399
+    },
+    {
+      "epoch": 0.6144,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.9766231550209025e-05,
+      "loss": 0.5865,
+      "step": 2400
+    },
+    {
+      "epoch": 0.614656,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.9761965702585106e-05,
+      "loss": 0.5936,
+      "step": 2401
+    },
+    {
+      "epoch": 0.614912,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.975769985496118e-05,
+      "loss": 0.5568,
+      "step": 2402
+    },
+    {
+      "epoch": 0.615168,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.975343400733726e-05,
+      "loss": 0.5584,
+      "step": 2403
+    },
+    {
+      "epoch": 0.615424,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.9749168159713335e-05,
+      "loss": 0.5422,
+      "step": 2404
+    },
+    {
+      "epoch": 0.61568,
+      "grad_norm": 2.6875,
+      "learning_rate": 3.9744902312089417e-05,
+      "loss": 0.6044,
+      "step": 2405
+    },
+    {
+      "epoch": 0.615936,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.974063646446549e-05,
+      "loss": 0.5782,
+      "step": 2406
+    },
+    {
+      "epoch": 0.616192,
+      "grad_norm": 2.375,
+      "learning_rate": 3.9736370616841565e-05,
+      "loss": 0.581,
+      "step": 2407
+    },
+    {
+      "epoch": 0.616448,
+      "grad_norm": 2.625,
+      "learning_rate": 3.9732104769217646e-05,
+      "loss": 0.6102,
+      "step": 2408
+    },
+    {
+      "epoch": 0.616704,
+      "grad_norm": 2.796875,
+      "learning_rate": 3.972783892159373e-05,
+      "loss": 0.5806,
+      "step": 2409
+    },
+    {
+      "epoch": 0.61696,
+      "grad_norm": 2.625,
+      "learning_rate": 3.97235730739698e-05,
+      "loss": 0.5727,
+      "step": 2410
+    },
+    {
+      "epoch": 0.617216,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.9719307226345876e-05,
+      "loss": 0.5724,
+      "step": 2411
+    },
+    {
+      "epoch": 0.617472,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.971504137872196e-05,
+      "loss": 0.6031,
+      "step": 2412
+    },
+    {
+      "epoch": 0.617728,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.971077553109803e-05,
+      "loss": 0.5594,
+      "step": 2413
+    },
+    {
+      "epoch": 0.617984,
+      "grad_norm": 2.375,
+      "learning_rate": 3.9706509683474105e-05,
+      "loss": 0.5802,
+      "step": 2414
+    },
+    {
+      "epoch": 0.61824,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.9702243835850186e-05,
+      "loss": 0.6218,
+      "step": 2415
+    },
+    {
+      "epoch": 0.618496,
+      "grad_norm": 2.6875,
+      "learning_rate": 3.969797798822627e-05,
+      "loss": 0.6074,
+      "step": 2416
+    },
+    {
+      "epoch": 0.618752,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.969371214060234e-05,
+      "loss": 0.586,
+      "step": 2417
+    },
+    {
+      "epoch": 0.619008,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.9689446292978416e-05,
+      "loss": 0.5603,
+      "step": 2418
+    },
+    {
+      "epoch": 0.619264,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.968518044535449e-05,
+      "loss": 0.5669,
+      "step": 2419
+    },
+    {
+      "epoch": 0.61952,
+      "grad_norm": 2.578125,
+      "learning_rate": 3.968091459773057e-05,
+      "loss": 0.5967,
+      "step": 2420
+    },
+    {
+      "epoch": 0.619776,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.9676648750106645e-05,
+      "loss": 0.574,
+      "step": 2421
+    },
+    {
+      "epoch": 0.620032,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.9672382902482726e-05,
+      "loss": 0.5563,
+      "step": 2422
+    },
+    {
+      "epoch": 0.620288,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.96681170548588e-05,
+      "loss": 0.5759,
+      "step": 2423
+    },
+    {
+      "epoch": 0.620544,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.966385120723488e-05,
+      "loss": 0.5873,
+      "step": 2424
+    },
+    {
+      "epoch": 0.6208,
+      "grad_norm": 2.640625,
+      "learning_rate": 3.9659585359610956e-05,
+      "loss": 0.575,
+      "step": 2425
+    },
+    {
+      "epoch": 0.621056,
+      "grad_norm": 2.671875,
+      "learning_rate": 3.965531951198703e-05,
+      "loss": 0.5576,
+      "step": 2426
+    },
+    {
+      "epoch": 0.621312,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.965105366436311e-05,
+      "loss": 0.5777,
+      "step": 2427
+    },
+    {
+      "epoch": 0.621568,
+      "grad_norm": 2.703125,
+      "learning_rate": 3.964678781673919e-05,
+      "loss": 0.5717,
+      "step": 2428
+    },
+    {
+      "epoch": 0.621824,
+      "grad_norm": 2.734375,
+      "learning_rate": 3.964252196911527e-05,
+      "loss": 0.5631,
+      "step": 2429
+    },
+    {
+      "epoch": 0.62208,
+      "grad_norm": 2.5,
+      "learning_rate": 3.963825612149134e-05,
+      "loss": 0.5922,
+      "step": 2430
+    },
+    {
+      "epoch": 0.622336,
+      "grad_norm": 2.71875,
+      "learning_rate": 3.963399027386742e-05,
+      "loss": 0.5934,
+      "step": 2431
+    },
+    {
+      "epoch": 0.622592,
+      "grad_norm": 2.90625,
+      "learning_rate": 3.9629724426243496e-05,
+      "loss": 0.5898,
+      "step": 2432
+    },
+    {
+      "epoch": 0.622848,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.962545857861957e-05,
+      "loss": 0.6145,
+      "step": 2433
+    },
+    {
+      "epoch": 0.623104,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.962119273099565e-05,
+      "loss": 0.5681,
+      "step": 2434
+    },
+    {
+      "epoch": 0.62336,
+      "grad_norm": 2.8125,
+      "learning_rate": 3.961692688337173e-05,
+      "loss": 0.57,
+      "step": 2435
+    },
+    {
+      "epoch": 0.623616,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.961266103574781e-05,
+      "loss": 0.6208,
+      "step": 2436
+    },
+    {
+      "epoch": 0.623872,
+      "grad_norm": 2.796875,
+      "learning_rate": 3.960839518812388e-05,
+      "loss": 0.5651,
+      "step": 2437
+    },
+    {
+      "epoch": 0.624128,
+      "grad_norm": 2.921875,
+      "learning_rate": 3.9604129340499955e-05,
+      "loss": 0.5914,
+      "step": 2438
+    },
+    {
+      "epoch": 0.624384,
+      "grad_norm": 2.5,
+      "learning_rate": 3.9599863492876036e-05,
+      "loss": 0.574,
+      "step": 2439
+    },
+    {
+      "epoch": 0.62464,
+      "grad_norm": 2.671875,
+      "learning_rate": 3.959559764525211e-05,
+      "loss": 0.5993,
+      "step": 2440
+    },
+    {
+      "epoch": 0.624896,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.959133179762819e-05,
+      "loss": 0.5833,
+      "step": 2441
+    },
+    {
+      "epoch": 0.625152,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.9587065950004266e-05,
+      "loss": 0.5714,
+      "step": 2442
+    },
+    {
+      "epoch": 0.625408,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.958280010238035e-05,
+      "loss": 0.5977,
+      "step": 2443
+    },
+    {
+      "epoch": 0.625664,
+      "grad_norm": 2.578125,
+      "learning_rate": 3.957853425475642e-05,
+      "loss": 0.5508,
+      "step": 2444
+    },
+    {
+      "epoch": 0.62592,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.9574268407132496e-05,
+      "loss": 0.5622,
+      "step": 2445
+    },
+    {
+      "epoch": 0.626176,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.957000255950858e-05,
+      "loss": 0.5667,
+      "step": 2446
+    },
+    {
+      "epoch": 0.626432,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.956573671188466e-05,
+      "loss": 0.6362,
+      "step": 2447
+    },
+    {
+      "epoch": 0.626688,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.956147086426073e-05,
+      "loss": 0.5748,
+      "step": 2448
+    },
+    {
+      "epoch": 0.626944,
+      "grad_norm": 2.5,
+      "learning_rate": 3.9557205016636806e-05,
+      "loss": 0.5494,
+      "step": 2449
+    },
+    {
+      "epoch": 0.6272,
+      "grad_norm": 2.734375,
+      "learning_rate": 3.955293916901289e-05,
+      "loss": 0.569,
+      "step": 2450
+    },
+    {
+      "epoch": 0.627456,
+      "grad_norm": 2.703125,
+      "learning_rate": 3.954867332138896e-05,
+      "loss": 0.6106,
+      "step": 2451
+    },
+    {
+      "epoch": 0.627712,
+      "grad_norm": 2.59375,
+      "learning_rate": 3.9544407473765036e-05,
+      "loss": 0.6061,
+      "step": 2452
+    },
+    {
+      "epoch": 0.627968,
+      "grad_norm": 2.5625,
+      "learning_rate": 3.954014162614112e-05,
+      "loss": 0.5779,
+      "step": 2453
+    },
+    {
+      "epoch": 0.628224,
+      "grad_norm": 2.375,
+      "learning_rate": 3.95358757785172e-05,
+      "loss": 0.5283,
+      "step": 2454
+    },
+    {
+      "epoch": 0.62848,
+      "grad_norm": 3.0,
+      "learning_rate": 3.953160993089327e-05,
+      "loss": 0.6046,
+      "step": 2455
+    },
+    {
+      "epoch": 0.628736,
+      "grad_norm": 3.09375,
+      "learning_rate": 3.9527344083269346e-05,
+      "loss": 0.5385,
+      "step": 2456
+    },
+    {
+      "epoch": 0.628992,
+      "grad_norm": 2.609375,
+      "learning_rate": 3.952307823564542e-05,
+      "loss": 0.596,
+      "step": 2457
+    },
+    {
+      "epoch": 0.629248,
+      "grad_norm": 2.65625,
+      "learning_rate": 3.95188123880215e-05,
+      "loss": 0.5824,
+      "step": 2458
+    },
+    {
+      "epoch": 0.629504,
+      "grad_norm": 2.578125,
+      "learning_rate": 3.9514546540397576e-05,
+      "loss": 0.5541,
+      "step": 2459
+    },
+    {
+      "epoch": 0.62976,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.951028069277366e-05,
+      "loss": 0.5384,
+      "step": 2460
+    },
+    {
+      "epoch": 0.630016,
+      "grad_norm": 2.96875,
+      "learning_rate": 3.950601484514974e-05,
+      "loss": 0.563,
+      "step": 2461
+    },
+    {
+      "epoch": 0.630272,
+      "grad_norm": 2.875,
+      "learning_rate": 3.950174899752581e-05,
+      "loss": 0.5972,
+      "step": 2462
+    },
+    {
+      "epoch": 0.630528,
+      "grad_norm": 2.59375,
+      "learning_rate": 3.9497483149901887e-05,
+      "loss": 0.5996,
+      "step": 2463
+    },
+    {
+      "epoch": 0.630784,
+      "grad_norm": 2.640625,
+      "learning_rate": 3.949321730227796e-05,
+      "loss": 0.574,
+      "step": 2464
+    },
+    {
+      "epoch": 0.63104,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.948895145465404e-05,
+      "loss": 0.5738,
+      "step": 2465
+    },
+    {
+      "epoch": 0.631296,
+      "grad_norm": 2.59375,
+      "learning_rate": 3.948468560703012e-05,
+      "loss": 0.6405,
+      "step": 2466
+    },
+    {
+      "epoch": 0.631552,
+      "grad_norm": 2.5625,
+      "learning_rate": 3.94804197594062e-05,
+      "loss": 0.5655,
+      "step": 2467
+    },
+    {
+      "epoch": 0.631808,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.947615391178227e-05,
+      "loss": 0.5661,
+      "step": 2468
+    },
+    {
+      "epoch": 0.632064,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.947188806415835e-05,
+      "loss": 0.5936,
+      "step": 2469
+    },
+    {
+      "epoch": 0.63232,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.946762221653443e-05,
+      "loss": 0.5533,
+      "step": 2470
+    },
+    {
+      "epoch": 0.632576,
+      "grad_norm": 2.71875,
+      "learning_rate": 3.94633563689105e-05,
+      "loss": 0.5561,
+      "step": 2471
+    },
+    {
+      "epoch": 0.632832,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.945909052128658e-05,
+      "loss": 0.5744,
+      "step": 2472
+    },
+    {
+      "epoch": 0.633088,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.945482467366266e-05,
+      "loss": 0.5441,
+      "step": 2473
+    },
+    {
+      "epoch": 0.633344,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.945055882603874e-05,
+      "loss": 0.5422,
+      "step": 2474
+    },
+    {
+      "epoch": 0.6336,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.944629297841481e-05,
+      "loss": 0.5187,
+      "step": 2475
+    },
+    {
+      "epoch": 0.633856,
+      "grad_norm": 2.578125,
+      "learning_rate": 3.9442027130790886e-05,
+      "loss": 0.5746,
+      "step": 2476
+    },
+    {
+      "epoch": 0.634112,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.943776128316697e-05,
+      "loss": 0.5723,
+      "step": 2477
+    },
+    {
+      "epoch": 0.634368,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.943349543554304e-05,
+      "loss": 0.5742,
+      "step": 2478
+    },
+    {
+      "epoch": 0.634624,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.942922958791912e-05,
+      "loss": 0.5684,
+      "step": 2479
+    },
+    {
+      "epoch": 0.63488,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.94249637402952e-05,
+      "loss": 0.5674,
+      "step": 2480
+    },
+    {
+      "epoch": 0.635136,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.942069789267128e-05,
+      "loss": 0.5992,
+      "step": 2481
+    },
+    {
+      "epoch": 0.635392,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.941643204504735e-05,
+      "loss": 0.5814,
+      "step": 2482
+    },
+    {
+      "epoch": 0.635648,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.9412166197423426e-05,
+      "loss": 0.5649,
+      "step": 2483
+    },
+    {
+      "epoch": 0.635904,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.940790034979951e-05,
+      "loss": 0.561,
+      "step": 2484
+    },
+    {
+      "epoch": 0.63616,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.940363450217559e-05,
+      "loss": 0.5557,
+      "step": 2485
+    },
+    {
+      "epoch": 0.636416,
+      "grad_norm": 2.671875,
+      "learning_rate": 3.939936865455166e-05,
+      "loss": 0.5812,
+      "step": 2486
+    },
+    {
+      "epoch": 0.636672,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.939510280692774e-05,
+      "loss": 0.5124,
+      "step": 2487
+    },
+    {
+      "epoch": 0.636928,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.939083695930382e-05,
+      "loss": 0.5492,
+      "step": 2488
+    },
+    {
+      "epoch": 0.637184,
+      "grad_norm": 2.375,
+      "learning_rate": 3.938657111167989e-05,
+      "loss": 0.5629,
+      "step": 2489
+    },
+    {
+      "epoch": 0.63744,
+      "grad_norm": 2.75,
+      "learning_rate": 3.9382305264055966e-05,
+      "loss": 0.6036,
+      "step": 2490
+    },
+    {
+      "epoch": 0.637696,
+      "grad_norm": 2.8125,
+      "learning_rate": 3.937803941643205e-05,
+      "loss": 0.5864,
+      "step": 2491
+    },
+    {
+      "epoch": 0.637952,
+      "grad_norm": 2.6875,
+      "learning_rate": 3.937377356880813e-05,
+      "loss": 0.599,
+      "step": 2492
+    },
+    {
+      "epoch": 0.638208,
+      "grad_norm": 2.84375,
+      "learning_rate": 3.93695077211842e-05,
+      "loss": 0.5736,
+      "step": 2493
+    },
+    {
+      "epoch": 0.638464,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.936524187356028e-05,
+      "loss": 0.6025,
+      "step": 2494
+    },
+    {
+      "epoch": 0.63872,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.936097602593636e-05,
+      "loss": 0.6074,
+      "step": 2495
+    },
+    {
+      "epoch": 0.638976,
+      "grad_norm": 2.640625,
+      "learning_rate": 3.935671017831243e-05,
+      "loss": 0.5234,
+      "step": 2496
+    },
+    {
+      "epoch": 0.639232,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.9352444330688507e-05,
+      "loss": 0.5397,
+      "step": 2497
+    },
+    {
+      "epoch": 0.639488,
+      "grad_norm": 2.765625,
+      "learning_rate": 3.934817848306459e-05,
+      "loss": 0.5515,
+      "step": 2498
+    },
+    {
+      "epoch": 0.639744,
+      "grad_norm": 2.59375,
+      "learning_rate": 3.934391263544067e-05,
+      "loss": 0.5917,
+      "step": 2499
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.933964678781674e-05,
+      "loss": 0.5729,
+      "step": 2500
+    },
+    {
+      "epoch": 0.640256,
+      "grad_norm": 2.578125,
+      "learning_rate": 3.933538094019282e-05,
+      "loss": 0.556,
+      "step": 2501
+    },
+    {
+      "epoch": 0.640512,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.933111509256889e-05,
+      "loss": 0.5665,
+      "step": 2502
+    },
+    {
+      "epoch": 0.640768,
+      "grad_norm": 2.5625,
+      "learning_rate": 3.932684924494497e-05,
+      "loss": 0.5613,
+      "step": 2503
+    },
+    {
+      "epoch": 0.641024,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.932258339732105e-05,
+      "loss": 0.5638,
+      "step": 2504
+    },
+    {
+      "epoch": 0.64128,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.931831754969713e-05,
+      "loss": 0.5378,
+      "step": 2505
+    },
+    {
+      "epoch": 0.641536,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.93140517020732e-05,
+      "loss": 0.5552,
+      "step": 2506
+    },
+    {
+      "epoch": 0.641792,
+      "grad_norm": 2.5,
+      "learning_rate": 3.930978585444928e-05,
+      "loss": 0.5535,
+      "step": 2507
+    },
+    {
+      "epoch": 0.642048,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.930552000682536e-05,
+      "loss": 0.5851,
+      "step": 2508
+    },
+    {
+      "epoch": 0.642304,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.930125415920143e-05,
+      "loss": 0.5435,
+      "step": 2509
+    },
+    {
+      "epoch": 0.64256,
+      "grad_norm": 2.640625,
+      "learning_rate": 3.929698831157751e-05,
+      "loss": 0.5443,
+      "step": 2510
+    },
+    {
+      "epoch": 0.642816,
+      "grad_norm": 2.375,
+      "learning_rate": 3.9292722463953594e-05,
+      "loss": 0.5486,
+      "step": 2511
+    },
+    {
+      "epoch": 0.643072,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.928845661632967e-05,
+      "loss": 0.5696,
+      "step": 2512
+    },
+    {
+      "epoch": 0.643328,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.928419076870574e-05,
+      "loss": 0.5887,
+      "step": 2513
+    },
+    {
+      "epoch": 0.643584,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.927992492108182e-05,
+      "loss": 0.5377,
+      "step": 2514
+    },
+    {
+      "epoch": 0.64384,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.92756590734579e-05,
+      "loss": 0.5508,
+      "step": 2515
+    },
+    {
+      "epoch": 0.644096,
+      "grad_norm": 2.375,
+      "learning_rate": 3.927139322583397e-05,
+      "loss": 0.5959,
+      "step": 2516
+    },
+    {
+      "epoch": 0.644352,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.926712737821005e-05,
+      "loss": 0.5793,
+      "step": 2517
+    },
+    {
+      "epoch": 0.644608,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.9262861530586134e-05,
+      "loss": 0.5632,
+      "step": 2518
+    },
+    {
+      "epoch": 0.644864,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.925859568296221e-05,
+      "loss": 0.5366,
+      "step": 2519
+    },
+    {
+      "epoch": 0.64512,
+      "grad_norm": 2.25,
+      "learning_rate": 3.925432983533828e-05,
+      "loss": 0.5618,
+      "step": 2520
+    },
+    {
+      "epoch": 0.645376,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.925006398771436e-05,
+      "loss": 0.5721,
+      "step": 2521
+    },
+    {
+      "epoch": 0.645632,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.924579814009044e-05,
+      "loss": 0.5779,
+      "step": 2522
+    },
+    {
+      "epoch": 0.645888,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.924153229246651e-05,
+      "loss": 0.577,
+      "step": 2523
+    },
+    {
+      "epoch": 0.646144,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.923726644484259e-05,
+      "loss": 0.578,
+      "step": 2524
+    },
+    {
+      "epoch": 0.6464,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.923300059721867e-05,
+      "loss": 0.577,
+      "step": 2525
+    },
+    {
+      "epoch": 0.646656,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.922873474959475e-05,
+      "loss": 0.5472,
+      "step": 2526
+    },
+    {
+      "epoch": 0.646912,
+      "grad_norm": 2.609375,
+      "learning_rate": 3.922446890197082e-05,
+      "loss": 0.5806,
+      "step": 2527
+    },
+    {
+      "epoch": 0.647168,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.92202030543469e-05,
+      "loss": 0.5408,
+      "step": 2528
+    },
+    {
+      "epoch": 0.647424,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.921593720672298e-05,
+      "loss": 0.5714,
+      "step": 2529
+    },
+    {
+      "epoch": 0.64768,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.921167135909906e-05,
+      "loss": 0.5686,
+      "step": 2530
+    },
+    {
+      "epoch": 0.647936,
+      "grad_norm": 2.640625,
+      "learning_rate": 3.920740551147513e-05,
+      "loss": 0.5288,
+      "step": 2531
+    },
+    {
+      "epoch": 0.648192,
+      "grad_norm": 2.6875,
+      "learning_rate": 3.920313966385121e-05,
+      "loss": 0.593,
+      "step": 2532
+    },
+    {
+      "epoch": 0.648448,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.919887381622729e-05,
+      "loss": 0.603,
+      "step": 2533
+    },
+    {
+      "epoch": 0.648704,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.919460796860336e-05,
+      "loss": 0.5233,
+      "step": 2534
+    },
+    {
+      "epoch": 0.64896,
+      "grad_norm": 2.625,
+      "learning_rate": 3.919034212097944e-05,
+      "loss": 0.5996,
+      "step": 2535
+    },
+    {
+      "epoch": 0.649216,
+      "grad_norm": 2.5,
+      "learning_rate": 3.918607627335552e-05,
+      "loss": 0.5458,
+      "step": 2536
+    },
+    {
+      "epoch": 0.649472,
+      "grad_norm": 2.5625,
+      "learning_rate": 3.91818104257316e-05,
+      "loss": 0.5562,
+      "step": 2537
+    },
+    {
+      "epoch": 0.649728,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.9177544578107673e-05,
+      "loss": 0.5483,
+      "step": 2538
+    },
+    {
+      "epoch": 0.649984,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.917327873048375e-05,
+      "loss": 0.5707,
+      "step": 2539
+    },
+    {
+      "epoch": 0.65024,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.916901288285982e-05,
+      "loss": 0.5466,
+      "step": 2540
+    },
+    {
+      "epoch": 0.650496,
+      "grad_norm": 2.375,
+      "learning_rate": 3.91647470352359e-05,
+      "loss": 0.5501,
+      "step": 2541
+    },
+    {
+      "epoch": 0.650752,
+      "grad_norm": 2.78125,
+      "learning_rate": 3.916048118761198e-05,
+      "loss": 0.5799,
+      "step": 2542
+    },
+    {
+      "epoch": 0.651008,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.915621533998806e-05,
+      "loss": 0.589,
+      "step": 2543
+    },
+    {
+      "epoch": 0.651264,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.915194949236414e-05,
+      "loss": 0.5815,
+      "step": 2544
+    },
+    {
+      "epoch": 0.65152,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.9147683644740214e-05,
+      "loss": 0.5385,
+      "step": 2545
+    },
+    {
+      "epoch": 0.651776,
+      "grad_norm": 2.640625,
+      "learning_rate": 3.914341779711629e-05,
+      "loss": 0.5281,
+      "step": 2546
+    },
+    {
+      "epoch": 0.652032,
+      "grad_norm": 2.734375,
+      "learning_rate": 3.913915194949236e-05,
+      "loss": 0.5655,
+      "step": 2547
+    },
+    {
+      "epoch": 0.652288,
+      "grad_norm": 2.71875,
+      "learning_rate": 3.913488610186844e-05,
+      "loss": 0.5724,
+      "step": 2548
+    },
+    {
+      "epoch": 0.652544,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.9130620254244524e-05,
+      "loss": 0.5775,
+      "step": 2549
+    },
+    {
+      "epoch": 0.6528,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.91263544066206e-05,
+      "loss": 0.6236,
+      "step": 2550
+    },
+    {
+      "epoch": 0.653056,
+      "grad_norm": 2.796875,
+      "learning_rate": 3.912208855899667e-05,
+      "loss": 0.5606,
+      "step": 2551
+    },
+    {
+      "epoch": 0.653312,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.9117822711372754e-05,
+      "loss": 0.542,
+      "step": 2552
+    },
+    {
+      "epoch": 0.653568,
+      "grad_norm": 2.59375,
+      "learning_rate": 3.911355686374883e-05,
+      "loss": 0.5457,
+      "step": 2553
+    },
+    {
+      "epoch": 0.653824,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.91092910161249e-05,
+      "loss": 0.5807,
+      "step": 2554
+    },
+    {
+      "epoch": 0.65408,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.910502516850098e-05,
+      "loss": 0.5548,
+      "step": 2555
+    },
+    {
+      "epoch": 0.654336,
+      "grad_norm": 2.71875,
+      "learning_rate": 3.9100759320877064e-05,
+      "loss": 0.5465,
+      "step": 2556
+    },
+    {
+      "epoch": 0.654592,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.909649347325314e-05,
+      "loss": 0.5595,
+      "step": 2557
+    },
+    {
+      "epoch": 0.654848,
+      "grad_norm": 2.59375,
+      "learning_rate": 3.909222762562921e-05,
+      "loss": 0.5827,
+      "step": 2558
+    },
+    {
+      "epoch": 0.655104,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.9087961778005294e-05,
+      "loss": 0.5524,
+      "step": 2559
+    },
+    {
+      "epoch": 0.65536,
+      "grad_norm": 3.078125,
+      "learning_rate": 3.908369593038137e-05,
+      "loss": 0.5642,
+      "step": 2560
+    },
+    {
+      "epoch": 0.655616,
+      "grad_norm": 2.375,
+      "learning_rate": 3.907943008275744e-05,
+      "loss": 0.5488,
+      "step": 2561
+    },
+    {
+      "epoch": 0.655872,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.9075164235133524e-05,
+      "loss": 0.5848,
+      "step": 2562
+    },
+    {
+      "epoch": 0.656128,
+      "grad_norm": 2.6875,
+      "learning_rate": 3.9070898387509605e-05,
+      "loss": 0.5255,
+      "step": 2563
+    },
+    {
+      "epoch": 0.656384,
+      "grad_norm": 2.984375,
+      "learning_rate": 3.906663253988568e-05,
+      "loss": 0.5487,
+      "step": 2564
+    },
+    {
+      "epoch": 0.65664,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.906236669226175e-05,
+      "loss": 0.5677,
+      "step": 2565
+    },
+    {
+      "epoch": 0.656896,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.905810084463783e-05,
+      "loss": 0.6169,
+      "step": 2566
+    },
+    {
+      "epoch": 0.657152,
+      "grad_norm": 2.375,
+      "learning_rate": 3.905383499701391e-05,
+      "loss": 0.5405,
+      "step": 2567
+    },
+    {
+      "epoch": 0.657408,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.904956914938999e-05,
+      "loss": 0.5692,
+      "step": 2568
+    },
+    {
+      "epoch": 0.657664,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.9045303301766064e-05,
+      "loss": 0.5147,
+      "step": 2569
+    },
+    {
+      "epoch": 0.65792,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.904103745414214e-05,
+      "loss": 0.5858,
+      "step": 2570
+    },
+    {
+      "epoch": 0.658176,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.903677160651822e-05,
+      "loss": 0.5729,
+      "step": 2571
+    },
+    {
+      "epoch": 0.658432,
+      "grad_norm": 2.59375,
+      "learning_rate": 3.903250575889429e-05,
+      "loss": 0.5534,
+      "step": 2572
+    },
+    {
+      "epoch": 0.658688,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.902823991127037e-05,
+      "loss": 0.5665,
+      "step": 2573
+    },
+    {
+      "epoch": 0.658944,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.902397406364645e-05,
+      "loss": 0.5439,
+      "step": 2574
+    },
+    {
+      "epoch": 0.6592,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.901970821602253e-05,
+      "loss": 0.533,
+      "step": 2575
+    },
+    {
+      "epoch": 0.659456,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.9015442368398604e-05,
+      "loss": 0.5325,
+      "step": 2576
+    },
+    {
+      "epoch": 0.659712,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.901117652077468e-05,
+      "loss": 0.5634,
+      "step": 2577
+    },
+    {
+      "epoch": 0.659968,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.900691067315076e-05,
+      "loss": 0.5412,
+      "step": 2578
+    },
+    {
+      "epoch": 0.660224,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.9002644825526834e-05,
+      "loss": 0.5088,
+      "step": 2579
+    },
+    {
+      "epoch": 0.66048,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.899837897790291e-05,
+      "loss": 0.5547,
+      "step": 2580
+    },
+    {
+      "epoch": 0.660736,
+      "grad_norm": 2.578125,
+      "learning_rate": 3.899411313027899e-05,
+      "loss": 0.5129,
+      "step": 2581
+    },
+    {
+      "epoch": 0.660992,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.898984728265507e-05,
+      "loss": 0.5353,
+      "step": 2582
+    },
+    {
+      "epoch": 0.661248,
+      "grad_norm": 2.5,
+      "learning_rate": 3.8985581435031144e-05,
+      "loss": 0.5628,
+      "step": 2583
+    },
+    {
+      "epoch": 0.661504,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.898131558740722e-05,
+      "loss": 0.53,
+      "step": 2584
+    },
+    {
+      "epoch": 0.66176,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.897704973978329e-05,
+      "loss": 0.5547,
+      "step": 2585
+    },
+    {
+      "epoch": 0.662016,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.8972783892159374e-05,
+      "loss": 0.5442,
+      "step": 2586
+    },
+    {
+      "epoch": 0.662272,
+      "grad_norm": 2.65625,
+      "learning_rate": 3.8968518044535455e-05,
+      "loss": 0.5847,
+      "step": 2587
+    },
+    {
+      "epoch": 0.662528,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.896425219691153e-05,
+      "loss": 0.5681,
+      "step": 2588
+    },
+    {
+      "epoch": 0.662784,
+      "grad_norm": 2.578125,
+      "learning_rate": 3.89599863492876e-05,
+      "loss": 0.6057,
+      "step": 2589
+    },
+    {
+      "epoch": 0.66304,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.8955720501663684e-05,
+      "loss": 0.5928,
+      "step": 2590
+    },
+    {
+      "epoch": 0.663296,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.895145465403976e-05,
+      "loss": 0.535,
+      "step": 2591
+    },
+    {
+      "epoch": 0.663552,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.894718880641583e-05,
+      "loss": 0.5843,
+      "step": 2592
+    },
+    {
+      "epoch": 0.663808,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.8942922958791914e-05,
+      "loss": 0.5506,
+      "step": 2593
+    },
+    {
+      "epoch": 0.664064,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.8938657111167995e-05,
+      "loss": 0.606,
+      "step": 2594
+    },
+    {
+      "epoch": 0.66432,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.893439126354407e-05,
+      "loss": 0.5543,
+      "step": 2595
+    },
+    {
+      "epoch": 0.664576,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.8930125415920143e-05,
+      "loss": 0.5644,
+      "step": 2596
+    },
+    {
+      "epoch": 0.664832,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.8925859568296225e-05,
+      "loss": 0.5451,
+      "step": 2597
+    },
+    {
+      "epoch": 0.665088,
+      "grad_norm": 2.25,
+      "learning_rate": 3.89215937206723e-05,
+      "loss": 0.5558,
+      "step": 2598
+    },
+    {
+      "epoch": 0.665344,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.891732787304837e-05,
+      "loss": 0.6026,
+      "step": 2599
+    },
+    {
+      "epoch": 0.6656,
+      "grad_norm": 2.765625,
+      "learning_rate": 3.8913062025424454e-05,
+      "loss": 0.5327,
+      "step": 2600
+    },
+    {
+      "epoch": 0.665856,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.8908796177800535e-05,
+      "loss": 0.5594,
+      "step": 2601
+    },
+    {
+      "epoch": 0.666112,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.890453033017661e-05,
+      "loss": 0.5844,
+      "step": 2602
+    },
+    {
+      "epoch": 0.666368,
+      "grad_norm": 2.59375,
+      "learning_rate": 3.8900264482552684e-05,
+      "loss": 0.5608,
+      "step": 2603
+    },
+    {
+      "epoch": 0.666624,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.889599863492876e-05,
+      "loss": 0.5766,
+      "step": 2604
+    },
+    {
+      "epoch": 0.66688,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.889173278730484e-05,
+      "loss": 0.5404,
+      "step": 2605
+    },
+    {
+      "epoch": 0.667136,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.888746693968091e-05,
+      "loss": 0.5551,
+      "step": 2606
+    },
+    {
+      "epoch": 0.667392,
+      "grad_norm": 2.5,
+      "learning_rate": 3.8883201092056994e-05,
+      "loss": 0.5576,
+      "step": 2607
+    },
+    {
+      "epoch": 0.667648,
+      "grad_norm": 2.5,
+      "learning_rate": 3.8878935244433075e-05,
+      "loss": 0.5799,
+      "step": 2608
+    },
+    {
+      "epoch": 0.667904,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.887466939680915e-05,
+      "loss": 0.5641,
+      "step": 2609
+    },
+    {
+      "epoch": 0.66816,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.8870403549185224e-05,
+      "loss": 0.5781,
+      "step": 2610
+    },
+    {
+      "epoch": 0.668416,
+      "grad_norm": 2.75,
+      "learning_rate": 3.88661377015613e-05,
+      "loss": 0.5892,
+      "step": 2611
+    },
+    {
+      "epoch": 0.668672,
+      "grad_norm": 2.640625,
+      "learning_rate": 3.886187185393738e-05,
+      "loss": 0.5648,
+      "step": 2612
+    },
+    {
+      "epoch": 0.668928,
+      "grad_norm": 2.609375,
+      "learning_rate": 3.885760600631346e-05,
+      "loss": 0.536,
+      "step": 2613
+    },
+    {
+      "epoch": 0.669184,
+      "grad_norm": 2.625,
+      "learning_rate": 3.8853340158689534e-05,
+      "loss": 0.5798,
+      "step": 2614
+    },
+    {
+      "epoch": 0.66944,
+      "grad_norm": 2.375,
+      "learning_rate": 3.884907431106561e-05,
+      "loss": 0.5591,
+      "step": 2615
+    },
+    {
+      "epoch": 0.669696,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.884480846344169e-05,
+      "loss": 0.5611,
+      "step": 2616
+    },
+    {
+      "epoch": 0.669952,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.8840542615817764e-05,
+      "loss": 0.5655,
+      "step": 2617
+    },
+    {
+      "epoch": 0.670208,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.883627676819384e-05,
+      "loss": 0.5326,
+      "step": 2618
+    },
+    {
+      "epoch": 0.670464,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.883201092056992e-05,
+      "loss": 0.5154,
+      "step": 2619
+    },
+    {
+      "epoch": 0.67072,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.8827745072946e-05,
+      "loss": 0.5414,
+      "step": 2620
+    },
+    {
+      "epoch": 0.670976,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.8823479225322075e-05,
+      "loss": 0.5545,
+      "step": 2621
+    },
+    {
+      "epoch": 0.671232,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.881921337769815e-05,
+      "loss": 0.5588,
+      "step": 2622
+    },
+    {
+      "epoch": 0.671488,
+      "grad_norm": 2.5625,
+      "learning_rate": 3.881494753007422e-05,
+      "loss": 0.5443,
+      "step": 2623
+    },
+    {
+      "epoch": 0.671744,
+      "grad_norm": 2.65625,
+      "learning_rate": 3.8810681682450304e-05,
+      "loss": 0.569,
+      "step": 2624
+    },
+    {
+      "epoch": 0.672,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.880641583482638e-05,
+      "loss": 0.5328,
+      "step": 2625
+    },
+    {
+      "epoch": 0.672256,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.880214998720246e-05,
+      "loss": 0.5455,
+      "step": 2626
+    },
+    {
+      "epoch": 0.672512,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.879788413957854e-05,
+      "loss": 0.56,
+      "step": 2627
+    },
+    {
+      "epoch": 0.672768,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.8793618291954615e-05,
+      "loss": 0.5394,
+      "step": 2628
+    },
+    {
+      "epoch": 0.673024,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.878935244433069e-05,
+      "loss": 0.5655,
+      "step": 2629
+    },
+    {
+      "epoch": 0.67328,
+      "grad_norm": 2.5625,
+      "learning_rate": 3.8785086596706763e-05,
+      "loss": 0.5963,
+      "step": 2630
+    },
+    {
+      "epoch": 0.673536,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.8780820749082844e-05,
+      "loss": 0.5763,
+      "step": 2631
+    },
+    {
+      "epoch": 0.673792,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.8776554901458925e-05,
+      "loss": 0.5294,
+      "step": 2632
+    },
+    {
+      "epoch": 0.674048,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.8772289053835e-05,
+      "loss": 0.5542,
+      "step": 2633
+    },
+    {
+      "epoch": 0.674304,
+      "grad_norm": 2.65625,
+      "learning_rate": 3.8768023206211074e-05,
+      "loss": 0.5539,
+      "step": 2634
+    },
+    {
+      "epoch": 0.67456,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.8763757358587155e-05,
+      "loss": 0.5406,
+      "step": 2635
+    },
+    {
+      "epoch": 0.674816,
+      "grad_norm": 2.375,
+      "learning_rate": 3.875949151096323e-05,
+      "loss": 0.5697,
+      "step": 2636
+    },
+    {
+      "epoch": 0.675072,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.8755225663339304e-05,
+      "loss": 0.5518,
+      "step": 2637
+    },
+    {
+      "epoch": 0.675328,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.8750959815715385e-05,
+      "loss": 0.5667,
+      "step": 2638
+    },
+    {
+      "epoch": 0.675584,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.8746693968091466e-05,
+      "loss": 0.5595,
+      "step": 2639
+    },
+    {
+      "epoch": 0.67584,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.874242812046754e-05,
+      "loss": 0.5575,
+      "step": 2640
+    },
+    {
+      "epoch": 0.676096,
+      "grad_norm": 2.125,
+      "learning_rate": 3.8738162272843614e-05,
+      "loss": 0.5624,
+      "step": 2641
+    },
+    {
+      "epoch": 0.676352,
+      "grad_norm": 2.65625,
+      "learning_rate": 3.8733896425219695e-05,
+      "loss": 0.5955,
+      "step": 2642
+    },
+    {
+      "epoch": 0.676608,
+      "grad_norm": 2.609375,
+      "learning_rate": 3.872963057759577e-05,
+      "loss": 0.5803,
+      "step": 2643
+    },
+    {
+      "epoch": 0.676864,
+      "grad_norm": 2.5,
+      "learning_rate": 3.8725364729971844e-05,
+      "loss": 0.5784,
+      "step": 2644
+    },
+    {
+      "epoch": 0.67712,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.8721098882347925e-05,
+      "loss": 0.5552,
+      "step": 2645
+    },
+    {
+      "epoch": 0.677376,
+      "grad_norm": 2.671875,
+      "learning_rate": 3.8716833034724006e-05,
+      "loss": 0.5275,
+      "step": 2646
+    },
+    {
+      "epoch": 0.677632,
+      "grad_norm": 2.671875,
+      "learning_rate": 3.871256718710008e-05,
+      "loss": 0.5531,
+      "step": 2647
+    },
+    {
+      "epoch": 0.677888,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.8708301339476154e-05,
+      "loss": 0.5427,
+      "step": 2648
+    },
+    {
+      "epoch": 0.678144,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.870403549185223e-05,
+      "loss": 0.5728,
+      "step": 2649
+    },
+    {
+      "epoch": 0.6784,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.869976964422831e-05,
+      "loss": 0.5677,
+      "step": 2650
+    },
+    {
+      "epoch": 0.678656,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.869550379660439e-05,
+      "loss": 0.5584,
+      "step": 2651
+    },
+    {
+      "epoch": 0.678912,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.8691237948980465e-05,
+      "loss": 0.5871,
+      "step": 2652
+    },
+    {
+      "epoch": 0.679168,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.868697210135654e-05,
+      "loss": 0.5631,
+      "step": 2653
+    },
+    {
+      "epoch": 0.679424,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.868270625373262e-05,
+      "loss": 0.5554,
+      "step": 2654
+    },
+    {
+      "epoch": 0.67968,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.8678440406108695e-05,
+      "loss": 0.5674,
+      "step": 2655
+    },
+    {
+      "epoch": 0.679936,
+      "grad_norm": 2.6875,
+      "learning_rate": 3.867417455848477e-05,
+      "loss": 0.5771,
+      "step": 2656
+    },
+    {
+      "epoch": 0.680192,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.866990871086085e-05,
+      "loss": 0.5537,
+      "step": 2657
+    },
+    {
+      "epoch": 0.680448,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.866564286323693e-05,
+      "loss": 0.5359,
+      "step": 2658
+    },
+    {
+      "epoch": 0.680704,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.8661377015613005e-05,
+      "loss": 0.5726,
+      "step": 2659
+    },
+    {
+      "epoch": 0.68096,
+      "grad_norm": 2.59375,
+      "learning_rate": 3.865711116798908e-05,
+      "loss": 0.5424,
+      "step": 2660
+    },
+    {
+      "epoch": 0.681216,
+      "grad_norm": 2.625,
+      "learning_rate": 3.865284532036516e-05,
+      "loss": 0.6067,
+      "step": 2661
+    },
+    {
+      "epoch": 0.681472,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.8648579472741235e-05,
+      "loss": 0.5758,
+      "step": 2662
+    },
+    {
+      "epoch": 0.681728,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.864431362511731e-05,
+      "loss": 0.5715,
+      "step": 2663
+    },
+    {
+      "epoch": 0.681984,
+      "grad_norm": 2.578125,
+      "learning_rate": 3.864004777749339e-05,
+      "loss": 0.5991,
+      "step": 2664
+    },
+    {
+      "epoch": 0.68224,
+      "grad_norm": 2.65625,
+      "learning_rate": 3.863578192986947e-05,
+      "loss": 0.5697,
+      "step": 2665
+    },
+    {
+      "epoch": 0.682496,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.8631516082245545e-05,
+      "loss": 0.5528,
+      "step": 2666
+    },
+    {
+      "epoch": 0.682752,
+      "grad_norm": 2.375,
+      "learning_rate": 3.862725023462162e-05,
+      "loss": 0.5627,
+      "step": 2667
+    },
+    {
+      "epoch": 0.683008,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.8622984386997694e-05,
+      "loss": 0.5808,
+      "step": 2668
+    },
+    {
+      "epoch": 0.683264,
+      "grad_norm": 2.703125,
+      "learning_rate": 3.8618718539373775e-05,
+      "loss": 0.5769,
+      "step": 2669
+    },
+    {
+      "epoch": 0.68352,
+      "grad_norm": 2.5,
+      "learning_rate": 3.8614452691749856e-05,
+      "loss": 0.5453,
+      "step": 2670
+    },
+    {
+      "epoch": 0.683776,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.861018684412593e-05,
+      "loss": 0.5572,
+      "step": 2671
+    },
+    {
+      "epoch": 0.684032,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.8605920996502005e-05,
+      "loss": 0.5214,
+      "step": 2672
+    },
+    {
+      "epoch": 0.684288,
+      "grad_norm": 2.703125,
+      "learning_rate": 3.8601655148878086e-05,
+      "loss": 0.5201,
+      "step": 2673
+    },
+    {
+      "epoch": 0.684544,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.859738930125416e-05,
+      "loss": 0.5596,
+      "step": 2674
+    },
+    {
+      "epoch": 0.6848,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.8593123453630234e-05,
+      "loss": 0.5734,
+      "step": 2675
+    },
+    {
+      "epoch": 0.685056,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.8588857606006315e-05,
+      "loss": 0.517,
+      "step": 2676
+    },
+    {
+      "epoch": 0.685312,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.8584591758382396e-05,
+      "loss": 0.523,
+      "step": 2677
+    },
+    {
+      "epoch": 0.685568,
+      "grad_norm": 2.375,
+      "learning_rate": 3.858032591075847e-05,
+      "loss": 0.545,
+      "step": 2678
+    },
+    {
+      "epoch": 0.685824,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.8576060063134545e-05,
+      "loss": 0.5957,
+      "step": 2679
+    },
+    {
+      "epoch": 0.68608,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.8571794215510626e-05,
+      "loss": 0.557,
+      "step": 2680
+    },
+    {
+      "epoch": 0.686336,
+      "grad_norm": 2.25,
+      "learning_rate": 3.85675283678867e-05,
+      "loss": 0.5574,
+      "step": 2681
+    },
+    {
+      "epoch": 0.686592,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.8563262520262774e-05,
+      "loss": 0.5065,
+      "step": 2682
+    },
+    {
+      "epoch": 0.686848,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.8558996672638855e-05,
+      "loss": 0.5535,
+      "step": 2683
+    },
+    {
+      "epoch": 0.687104,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.8554730825014936e-05,
+      "loss": 0.5562,
+      "step": 2684
+    },
+    {
+      "epoch": 0.68736,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.855046497739101e-05,
+      "loss": 0.5681,
+      "step": 2685
+    },
+    {
+      "epoch": 0.687616,
+      "grad_norm": 2.375,
+      "learning_rate": 3.8546199129767085e-05,
+      "loss": 0.5194,
+      "step": 2686
+    },
+    {
+      "epoch": 0.687872,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.854193328214316e-05,
+      "loss": 0.5282,
+      "step": 2687
+    },
+    {
+      "epoch": 0.688128,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.853766743451924e-05,
+      "loss": 0.539,
+      "step": 2688
+    },
+    {
+      "epoch": 0.688384,
+      "grad_norm": 2.578125,
+      "learning_rate": 3.853340158689532e-05,
+      "loss": 0.5907,
+      "step": 2689
+    },
+    {
+      "epoch": 0.68864,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.8529135739271396e-05,
+      "loss": 0.5695,
+      "step": 2690
+    },
+    {
+      "epoch": 0.688896,
+      "grad_norm": 2.578125,
+      "learning_rate": 3.8524869891647477e-05,
+      "loss": 0.5716,
+      "step": 2691
+    },
+    {
+      "epoch": 0.689152,
+      "grad_norm": 2.921875,
+      "learning_rate": 3.852060404402355e-05,
+      "loss": 0.5135,
+      "step": 2692
+    },
+    {
+      "epoch": 0.689408,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.8516338196399625e-05,
+      "loss": 0.5088,
+      "step": 2693
+    },
+    {
+      "epoch": 0.689664,
+      "grad_norm": 2.5,
+      "learning_rate": 3.85120723487757e-05,
+      "loss": 0.557,
+      "step": 2694
+    },
+    {
+      "epoch": 0.68992,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.850780650115178e-05,
+      "loss": 0.5859,
+      "step": 2695
+    },
+    {
+      "epoch": 0.690176,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.850354065352786e-05,
+      "loss": 0.5363,
+      "step": 2696
+    },
+    {
+      "epoch": 0.690432,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.8499274805903936e-05,
+      "loss": 0.5442,
+      "step": 2697
+    },
+    {
+      "epoch": 0.690688,
+      "grad_norm": 2.796875,
+      "learning_rate": 3.849500895828001e-05,
+      "loss": 0.6078,
+      "step": 2698
+    },
+    {
+      "epoch": 0.690944,
+      "grad_norm": 2.375,
+      "learning_rate": 3.849074311065609e-05,
+      "loss": 0.5701,
+      "step": 2699
+    },
+    {
+      "epoch": 0.6912,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.8486477263032165e-05,
+      "loss": 0.5565,
+      "step": 2700
+    },
+    {
+      "epoch": 0.691456,
+      "grad_norm": 2.703125,
+      "learning_rate": 3.848221141540824e-05,
+      "loss": 0.5664,
+      "step": 2701
+    },
+    {
+      "epoch": 0.691712,
+      "grad_norm": 2.84375,
+      "learning_rate": 3.847794556778432e-05,
+      "loss": 0.5009,
+      "step": 2702
+    },
+    {
+      "epoch": 0.691968,
+      "grad_norm": 2.59375,
+      "learning_rate": 3.84736797201604e-05,
+      "loss": 0.5416,
+      "step": 2703
+    },
+    {
+      "epoch": 0.692224,
+      "grad_norm": 2.703125,
+      "learning_rate": 3.8469413872536476e-05,
+      "loss": 0.5186,
+      "step": 2704
+    },
+    {
+      "epoch": 0.69248,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.846514802491255e-05,
+      "loss": 0.5438,
+      "step": 2705
+    },
+    {
+      "epoch": 0.692736,
+      "grad_norm": 2.75,
+      "learning_rate": 3.846088217728863e-05,
+      "loss": 0.5389,
+      "step": 2706
+    },
+    {
+      "epoch": 0.692992,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.8456616329664706e-05,
+      "loss": 0.5605,
+      "step": 2707
+    },
+    {
+      "epoch": 0.693248,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.845235048204078e-05,
+      "loss": 0.5169,
+      "step": 2708
+    },
+    {
+      "epoch": 0.693504,
+      "grad_norm": 2.5,
+      "learning_rate": 3.844808463441686e-05,
+      "loss": 0.5211,
+      "step": 2709
+    },
+    {
+      "epoch": 0.69376,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.844381878679294e-05,
+      "loss": 0.5696,
+      "step": 2710
+    },
+    {
+      "epoch": 0.694016,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.8439552939169016e-05,
+      "loss": 0.5604,
+      "step": 2711
+    },
+    {
+      "epoch": 0.694272,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.843528709154509e-05,
+      "loss": 0.5564,
+      "step": 2712
+    },
+    {
+      "epoch": 0.694528,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.8431021243921165e-05,
+      "loss": 0.53,
+      "step": 2713
+    },
+    {
+      "epoch": 0.694784,
+      "grad_norm": 2.625,
+      "learning_rate": 3.8426755396297246e-05,
+      "loss": 0.5495,
+      "step": 2714
+    },
+    {
+      "epoch": 0.69504,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.842248954867333e-05,
+      "loss": 0.5704,
+      "step": 2715
+    },
+    {
+      "epoch": 0.695296,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.84182237010494e-05,
+      "loss": 0.5324,
+      "step": 2716
+    },
+    {
+      "epoch": 0.695552,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.8413957853425475e-05,
+      "loss": 0.5733,
+      "step": 2717
+    },
+    {
+      "epoch": 0.695808,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.8409692005801556e-05,
+      "loss": 0.5803,
+      "step": 2718
+    },
+    {
+      "epoch": 0.696064,
+      "grad_norm": 2.625,
+      "learning_rate": 3.840542615817763e-05,
+      "loss": 0.5608,
+      "step": 2719
+    },
+    {
+      "epoch": 0.69632,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.8401160310553705e-05,
+      "loss": 0.5616,
+      "step": 2720
+    },
+    {
+      "epoch": 0.696576,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.8396894462929786e-05,
+      "loss": 0.5861,
+      "step": 2721
+    },
+    {
+      "epoch": 0.696832,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.839262861530587e-05,
+      "loss": 0.5762,
+      "step": 2722
+    },
+    {
+      "epoch": 0.697088,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.838836276768194e-05,
+      "loss": 0.563,
+      "step": 2723
+    },
+    {
+      "epoch": 0.697344,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.8384096920058015e-05,
+      "loss": 0.5496,
+      "step": 2724
+    },
+    {
+      "epoch": 0.6976,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.8379831072434097e-05,
+      "loss": 0.5398,
+      "step": 2725
+    },
+    {
+      "epoch": 0.697856,
+      "grad_norm": 2.609375,
+      "learning_rate": 3.837556522481017e-05,
+      "loss": 0.5281,
+      "step": 2726
+    },
+    {
+      "epoch": 0.698112,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.8371299377186245e-05,
+      "loss": 0.5518,
+      "step": 2727
+    },
+    {
+      "epoch": 0.698368,
+      "grad_norm": 2.5,
+      "learning_rate": 3.8367033529562326e-05,
+      "loss": 0.5216,
+      "step": 2728
+    },
+    {
+      "epoch": 0.698624,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.836276768193841e-05,
+      "loss": 0.5389,
+      "step": 2729
+    },
+    {
+      "epoch": 0.69888,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.835850183431448e-05,
+      "loss": 0.5334,
+      "step": 2730
+    },
+    {
+      "epoch": 0.699136,
+      "grad_norm": 2.671875,
+      "learning_rate": 3.8354235986690556e-05,
+      "loss": 0.5516,
+      "step": 2731
+    },
+    {
+      "epoch": 0.699392,
+      "grad_norm": 2.5,
+      "learning_rate": 3.834997013906663e-05,
+      "loss": 0.5393,
+      "step": 2732
+    },
+    {
+      "epoch": 0.699648,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.834570429144271e-05,
+      "loss": 0.5938,
+      "step": 2733
+    },
+    {
+      "epoch": 0.699904,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.834143844381879e-05,
+      "loss": 0.5479,
+      "step": 2734
+    },
+    {
+      "epoch": 0.70016,
+      "grad_norm": 2.5,
+      "learning_rate": 3.8337172596194866e-05,
+      "loss": 0.5631,
+      "step": 2735
+    },
+    {
+      "epoch": 0.700416,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.833290674857094e-05,
+      "loss": 0.518,
+      "step": 2736
+    },
+    {
+      "epoch": 0.700672,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.832864090094702e-05,
+      "loss": 0.5305,
+      "step": 2737
+    },
+    {
+      "epoch": 0.700928,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.8324375053323096e-05,
+      "loss": 0.5547,
+      "step": 2738
+    },
+    {
+      "epoch": 0.701184,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.832010920569917e-05,
+      "loss": 0.5561,
+      "step": 2739
+    },
+    {
+      "epoch": 0.70144,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.831584335807525e-05,
+      "loss": 0.5881,
+      "step": 2740
+    },
+    {
+      "epoch": 0.701696,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.831157751045133e-05,
+      "loss": 0.5357,
+      "step": 2741
+    },
+    {
+      "epoch": 0.701952,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.8307311662827406e-05,
+      "loss": 0.509,
+      "step": 2742
+    },
+    {
+      "epoch": 0.702208,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.830304581520348e-05,
+      "loss": 0.561,
+      "step": 2743
+    },
+    {
+      "epoch": 0.702464,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.829877996757956e-05,
+      "loss": 0.5567,
+      "step": 2744
+    },
+    {
+      "epoch": 0.70272,
+      "grad_norm": 2.6875,
+      "learning_rate": 3.8294514119955636e-05,
+      "loss": 0.5638,
+      "step": 2745
+    },
+    {
+      "epoch": 0.702976,
+      "grad_norm": 2.640625,
+      "learning_rate": 3.829024827233171e-05,
+      "loss": 0.5518,
+      "step": 2746
+    },
+    {
+      "epoch": 0.703232,
+      "grad_norm": 2.703125,
+      "learning_rate": 3.828598242470779e-05,
+      "loss": 0.547,
+      "step": 2747
+    },
+    {
+      "epoch": 0.703488,
+      "grad_norm": 2.375,
+      "learning_rate": 3.828171657708387e-05,
+      "loss": 0.5131,
+      "step": 2748
+    },
+    {
+      "epoch": 0.703744,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.827745072945995e-05,
+      "loss": 0.5441,
+      "step": 2749
+    },
+    {
+      "epoch": 0.704,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.827318488183602e-05,
+      "loss": 0.5444,
+      "step": 2750
+    },
+    {
+      "epoch": 0.704256,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.8268919034212095e-05,
+      "loss": 0.5522,
+      "step": 2751
+    },
+    {
+      "epoch": 0.704512,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.8264653186588176e-05,
+      "loss": 0.5749,
+      "step": 2752
+    },
+    {
+      "epoch": 0.704768,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.826038733896426e-05,
+      "loss": 0.5497,
+      "step": 2753
+    },
+    {
+      "epoch": 0.705024,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.825612149134033e-05,
+      "loss": 0.5373,
+      "step": 2754
+    },
+    {
+      "epoch": 0.70528,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.8251855643716406e-05,
+      "loss": 0.5279,
+      "step": 2755
+    },
+    {
+      "epoch": 0.705536,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.824758979609249e-05,
+      "loss": 0.5375,
+      "step": 2756
+    },
+    {
+      "epoch": 0.705792,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.824332394846856e-05,
+      "loss": 0.5497,
+      "step": 2757
+    },
+    {
+      "epoch": 0.706048,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.8239058100844635e-05,
+      "loss": 0.5788,
+      "step": 2758
+    },
+    {
+      "epoch": 0.706304,
+      "grad_norm": 2.78125,
+      "learning_rate": 3.8234792253220716e-05,
+      "loss": 0.5263,
+      "step": 2759
+    },
+    {
+      "epoch": 0.70656,
+      "grad_norm": 2.6875,
+      "learning_rate": 3.82305264055968e-05,
+      "loss": 0.5083,
+      "step": 2760
+    },
+    {
+      "epoch": 0.706816,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.822626055797287e-05,
+      "loss": 0.556,
+      "step": 2761
+    },
+    {
+      "epoch": 0.707072,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.8221994710348946e-05,
+      "loss": 0.5248,
+      "step": 2762
+    },
+    {
+      "epoch": 0.707328,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.821772886272503e-05,
+      "loss": 0.556,
+      "step": 2763
+    },
+    {
+      "epoch": 0.707584,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.82134630151011e-05,
+      "loss": 0.5838,
+      "step": 2764
+    },
+    {
+      "epoch": 0.70784,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.8209197167477176e-05,
+      "loss": 0.5365,
+      "step": 2765
+    },
+    {
+      "epoch": 0.708096,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.820493131985326e-05,
+      "loss": 0.5787,
+      "step": 2766
+    },
+    {
+      "epoch": 0.708352,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.820066547222934e-05,
+      "loss": 0.5392,
+      "step": 2767
+    },
+    {
+      "epoch": 0.708608,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.819639962460541e-05,
+      "loss": 0.5517,
+      "step": 2768
+    },
+    {
+      "epoch": 0.708864,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.8192133776981486e-05,
+      "loss": 0.5327,
+      "step": 2769
+    },
+    {
+      "epoch": 0.70912,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.818786792935756e-05,
+      "loss": 0.5308,
+      "step": 2770
+    },
+    {
+      "epoch": 0.709376,
+      "grad_norm": 2.5625,
+      "learning_rate": 3.818360208173364e-05,
+      "loss": 0.5433,
+      "step": 2771
+    },
+    {
+      "epoch": 0.709632,
+      "grad_norm": 2.375,
+      "learning_rate": 3.817933623410972e-05,
+      "loss": 0.544,
+      "step": 2772
+    },
+    {
+      "epoch": 0.709888,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.81750703864858e-05,
+      "loss": 0.5806,
+      "step": 2773
+    },
+    {
+      "epoch": 0.710144,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.817080453886188e-05,
+      "loss": 0.5313,
+      "step": 2774
+    },
+    {
+      "epoch": 0.7104,
+      "grad_norm": 2.65625,
+      "learning_rate": 3.816653869123795e-05,
+      "loss": 0.5508,
+      "step": 2775
+    },
+    {
+      "epoch": 0.710656,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.8162272843614026e-05,
+      "loss": 0.5521,
+      "step": 2776
+    },
+    {
+      "epoch": 0.710912,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.81580069959901e-05,
+      "loss": 0.544,
+      "step": 2777
+    },
+    {
+      "epoch": 0.711168,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.815374114836618e-05,
+      "loss": 0.5604,
+      "step": 2778
+    },
+    {
+      "epoch": 0.711424,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.814947530074226e-05,
+      "loss": 0.5566,
+      "step": 2779
+    },
+    {
+      "epoch": 0.71168,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.814520945311834e-05,
+      "loss": 0.5348,
+      "step": 2780
+    },
+    {
+      "epoch": 0.711936,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.814094360549441e-05,
+      "loss": 0.5254,
+      "step": 2781
+    },
+    {
+      "epoch": 0.712192,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.813667775787049e-05,
+      "loss": 0.5451,
+      "step": 2782
+    },
+    {
+      "epoch": 0.712448,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.8132411910246567e-05,
+      "loss": 0.5517,
+      "step": 2783
+    },
+    {
+      "epoch": 0.712704,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.812814606262264e-05,
+      "loss": 0.5487,
+      "step": 2784
+    },
+    {
+      "epoch": 0.71296,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.812388021499872e-05,
+      "loss": 0.555,
+      "step": 2785
+    },
+    {
+      "epoch": 0.713216,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.81196143673748e-05,
+      "loss": 0.5357,
+      "step": 2786
+    },
+    {
+      "epoch": 0.713472,
+      "grad_norm": 2.625,
+      "learning_rate": 3.811534851975088e-05,
+      "loss": 0.5552,
+      "step": 2787
+    },
+    {
+      "epoch": 0.713728,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.811108267212695e-05,
+      "loss": 0.5615,
+      "step": 2788
+    },
+    {
+      "epoch": 0.713984,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.810681682450303e-05,
+      "loss": 0.5716,
+      "step": 2789
+    },
+    {
+      "epoch": 0.71424,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.810255097687911e-05,
+      "loss": 0.5502,
+      "step": 2790
+    },
+    {
+      "epoch": 0.714496,
+      "grad_norm": 2.375,
+      "learning_rate": 3.809828512925519e-05,
+      "loss": 0.519,
+      "step": 2791
+    },
+    {
+      "epoch": 0.714752,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.809401928163126e-05,
+      "loss": 0.5742,
+      "step": 2792
+    },
+    {
+      "epoch": 0.715008,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.808975343400734e-05,
+      "loss": 0.5288,
+      "step": 2793
+    },
+    {
+      "epoch": 0.715264,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.808548758638342e-05,
+      "loss": 0.5553,
+      "step": 2794
+    },
+    {
+      "epoch": 0.71552,
+      "grad_norm": 2.5625,
+      "learning_rate": 3.808122173875949e-05,
+      "loss": 0.5276,
+      "step": 2795
+    },
+    {
+      "epoch": 0.715776,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.8076955891135566e-05,
+      "loss": 0.5395,
+      "step": 2796
+    },
+    {
+      "epoch": 0.716032,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.807269004351165e-05,
+      "loss": 0.5869,
+      "step": 2797
+    },
+    {
+      "epoch": 0.716288,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.806842419588773e-05,
+      "loss": 0.5613,
+      "step": 2798
+    },
+    {
+      "epoch": 0.716544,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.80641583482638e-05,
+      "loss": 0.5399,
+      "step": 2799
+    },
+    {
+      "epoch": 0.7168,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.8059892500639877e-05,
+      "loss": 0.4918,
+      "step": 2800
+    },
+    {
+      "epoch": 0.717056,
+      "grad_norm": 2.125,
+      "learning_rate": 3.805562665301596e-05,
+      "loss": 0.537,
+      "step": 2801
+    },
+    {
+      "epoch": 0.717312,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.805136080539203e-05,
+      "loss": 0.5521,
+      "step": 2802
+    },
+    {
+      "epoch": 0.717568,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.8047094957768106e-05,
+      "loss": 0.5276,
+      "step": 2803
+    },
+    {
+      "epoch": 0.717824,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.804282911014419e-05,
+      "loss": 0.5303,
+      "step": 2804
+    },
+    {
+      "epoch": 0.71808,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.803856326252027e-05,
+      "loss": 0.5887,
+      "step": 2805
+    },
+    {
+      "epoch": 0.718336,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.803429741489634e-05,
+      "loss": 0.5731,
+      "step": 2806
+    },
+    {
+      "epoch": 0.718592,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.803003156727242e-05,
+      "loss": 0.5135,
+      "step": 2807
+    },
+    {
+      "epoch": 0.718848,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.80257657196485e-05,
+      "loss": 0.5421,
+      "step": 2808
+    },
+    {
+      "epoch": 0.719104,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.802149987202457e-05,
+      "loss": 0.5319,
+      "step": 2809
+    },
+    {
+      "epoch": 0.71936,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.801723402440065e-05,
+      "loss": 0.5386,
+      "step": 2810
+    },
+    {
+      "epoch": 0.719616,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.801296817677673e-05,
+      "loss": 0.5302,
+      "step": 2811
+    },
+    {
+      "epoch": 0.719872,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.800870232915281e-05,
+      "loss": 0.4974,
+      "step": 2812
+    },
+    {
+      "epoch": 0.720128,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.800443648152888e-05,
+      "loss": 0.5309,
+      "step": 2813
+    },
+    {
+      "epoch": 0.720384,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.800017063390496e-05,
+      "loss": 0.5337,
+      "step": 2814
+    },
+    {
+      "epoch": 0.72064,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.799590478628103e-05,
+      "loss": 0.5139,
+      "step": 2815
+    },
+    {
+      "epoch": 0.720896,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.799163893865711e-05,
+      "loss": 0.5569,
+      "step": 2816
+    },
+    {
+      "epoch": 0.721152,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.798737309103319e-05,
+      "loss": 0.5555,
+      "step": 2817
+    },
+    {
+      "epoch": 0.721408,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.798310724340927e-05,
+      "loss": 0.5091,
+      "step": 2818
+    },
+    {
+      "epoch": 0.721664,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.797884139578534e-05,
+      "loss": 0.5807,
+      "step": 2819
+    },
+    {
+      "epoch": 0.72192,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.797457554816142e-05,
+      "loss": 0.5854,
+      "step": 2820
+    },
+    {
+      "epoch": 0.722176,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.79703097005375e-05,
+      "loss": 0.5496,
+      "step": 2821
+    },
+    {
+      "epoch": 0.722432,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.796604385291357e-05,
+      "loss": 0.5323,
+      "step": 2822
+    },
+    {
+      "epoch": 0.722688,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.796177800528965e-05,
+      "loss": 0.5157,
+      "step": 2823
+    },
+    {
+      "epoch": 0.722944,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.7957512157665733e-05,
+      "loss": 0.5436,
+      "step": 2824
+    },
+    {
+      "epoch": 0.7232,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.795324631004181e-05,
+      "loss": 0.5656,
+      "step": 2825
+    },
+    {
+      "epoch": 0.723456,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.794898046241788e-05,
+      "loss": 0.5624,
+      "step": 2826
+    },
+    {
+      "epoch": 0.723712,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.794471461479396e-05,
+      "loss": 0.5492,
+      "step": 2827
+    },
+    {
+      "epoch": 0.723968,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.794044876717004e-05,
+      "loss": 0.5251,
+      "step": 2828
+    },
+    {
+      "epoch": 0.724224,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.793618291954611e-05,
+      "loss": 0.5601,
+      "step": 2829
+    },
+    {
+      "epoch": 0.72448,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.793191707192219e-05,
+      "loss": 0.547,
+      "step": 2830
+    },
+    {
+      "epoch": 0.724736,
+      "grad_norm": 2.59375,
+      "learning_rate": 3.7927651224298274e-05,
+      "loss": 0.5384,
+      "step": 2831
+    },
+    {
+      "epoch": 0.724992,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.792338537667435e-05,
+      "loss": 0.5652,
+      "step": 2832
+    },
+    {
+      "epoch": 0.725248,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.791911952905042e-05,
+      "loss": 0.5355,
+      "step": 2833
+    },
+    {
+      "epoch": 0.725504,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.7914853681426496e-05,
+      "loss": 0.537,
+      "step": 2834
+    },
+    {
+      "epoch": 0.72576,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.791058783380258e-05,
+      "loss": 0.529,
+      "step": 2835
+    },
+    {
+      "epoch": 0.726016,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.790632198617866e-05,
+      "loss": 0.5521,
+      "step": 2836
+    },
+    {
+      "epoch": 0.726272,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.790205613855473e-05,
+      "loss": 0.5354,
+      "step": 2837
+    },
+    {
+      "epoch": 0.726528,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.7897790290930814e-05,
+      "loss": 0.5463,
+      "step": 2838
+    },
+    {
+      "epoch": 0.726784,
+      "grad_norm": 2.65625,
+      "learning_rate": 3.789352444330689e-05,
+      "loss": 0.5452,
+      "step": 2839
+    },
+    {
+      "epoch": 0.72704,
+      "grad_norm": 2.828125,
+      "learning_rate": 3.788925859568296e-05,
+      "loss": 0.5452,
+      "step": 2840
+    },
+    {
+      "epoch": 0.727296,
+      "grad_norm": 2.65625,
+      "learning_rate": 3.788499274805904e-05,
+      "loss": 0.5456,
+      "step": 2841
+    },
+    {
+      "epoch": 0.727552,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.788072690043512e-05,
+      "loss": 0.5454,
+      "step": 2842
+    },
+    {
+      "epoch": 0.727808,
+      "grad_norm": 2.375,
+      "learning_rate": 3.78764610528112e-05,
+      "loss": 0.5162,
+      "step": 2843
+    },
+    {
+      "epoch": 0.728064,
+      "grad_norm": 2.5625,
+      "learning_rate": 3.787219520518727e-05,
+      "loss": 0.5379,
+      "step": 2844
+    },
+    {
+      "epoch": 0.72832,
+      "grad_norm": 2.75,
+      "learning_rate": 3.786792935756335e-05,
+      "loss": 0.537,
+      "step": 2845
+    },
+    {
+      "epoch": 0.728576,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.786366350993943e-05,
+      "loss": 0.5504,
+      "step": 2846
+    },
+    {
+      "epoch": 0.728832,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.78593976623155e-05,
+      "loss": 0.5682,
+      "step": 2847
+    },
+    {
+      "epoch": 0.729088,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.785513181469158e-05,
+      "loss": 0.5518,
+      "step": 2848
+    },
+    {
+      "epoch": 0.729344,
+      "grad_norm": 2.5625,
+      "learning_rate": 3.785086596706766e-05,
+      "loss": 0.5748,
+      "step": 2849
+    },
+    {
+      "epoch": 0.7296,
+      "grad_norm": 2.578125,
+      "learning_rate": 3.784660011944374e-05,
+      "loss": 0.5506,
+      "step": 2850
+    },
+    {
+      "epoch": 0.729856,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.784233427181981e-05,
+      "loss": 0.5639,
+      "step": 2851
+    },
+    {
+      "epoch": 0.730112,
+      "grad_norm": 2.375,
+      "learning_rate": 3.783806842419589e-05,
+      "loss": 0.4812,
+      "step": 2852
+    },
+    {
+      "epoch": 0.730368,
+      "grad_norm": 2.609375,
+      "learning_rate": 3.783380257657196e-05,
+      "loss": 0.5118,
+      "step": 2853
+    },
+    {
+      "epoch": 0.730624,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.782953672894804e-05,
+      "loss": 0.564,
+      "step": 2854
+    },
+    {
+      "epoch": 0.73088,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.7825270881324124e-05,
+      "loss": 0.5574,
+      "step": 2855
+    },
+    {
+      "epoch": 0.731136,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.78210050337002e-05,
+      "loss": 0.5323,
+      "step": 2856
+    },
+    {
+      "epoch": 0.731392,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.781673918607628e-05,
+      "loss": 0.5811,
+      "step": 2857
+    },
+    {
+      "epoch": 0.731648,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.7812473338452353e-05,
+      "loss": 0.5664,
+      "step": 2858
+    },
+    {
+      "epoch": 0.731904,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.780820749082843e-05,
+      "loss": 0.5163,
+      "step": 2859
+    },
+    {
+      "epoch": 0.73216,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.78039416432045e-05,
+      "loss": 0.5567,
+      "step": 2860
+    },
+    {
+      "epoch": 0.732416,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.779967579558058e-05,
+      "loss": 0.5378,
+      "step": 2861
+    },
+    {
+      "epoch": 0.732672,
+      "grad_norm": 2.25,
+      "learning_rate": 3.7795409947956664e-05,
+      "loss": 0.5559,
+      "step": 2862
+    },
+    {
+      "epoch": 0.732928,
+      "grad_norm": 2.125,
+      "learning_rate": 3.779114410033274e-05,
+      "loss": 0.5302,
+      "step": 2863
+    },
+    {
+      "epoch": 0.733184,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.778687825270881e-05,
+      "loss": 0.5214,
+      "step": 2864
+    },
+    {
+      "epoch": 0.73344,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.7782612405084894e-05,
+      "loss": 0.5374,
+      "step": 2865
+    },
+    {
+      "epoch": 0.733696,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.777834655746097e-05,
+      "loss": 0.5434,
+      "step": 2866
+    },
+    {
+      "epoch": 0.733952,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.777408070983704e-05,
+      "loss": 0.5168,
+      "step": 2867
+    },
+    {
+      "epoch": 0.734208,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.776981486221312e-05,
+      "loss": 0.5244,
+      "step": 2868
+    },
+    {
+      "epoch": 0.734464,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.7765549014589204e-05,
+      "loss": 0.5419,
+      "step": 2869
+    },
+    {
+      "epoch": 0.73472,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.776128316696528e-05,
+      "loss": 0.5174,
+      "step": 2870
+    },
+    {
+      "epoch": 0.734976,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.775701731934135e-05,
+      "loss": 0.5447,
+      "step": 2871
+    },
+    {
+      "epoch": 0.735232,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.7752751471717434e-05,
+      "loss": 0.5517,
+      "step": 2872
+    },
+    {
+      "epoch": 0.735488,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.774848562409351e-05,
+      "loss": 0.5788,
+      "step": 2873
+    },
+    {
+      "epoch": 0.735744,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.774421977646959e-05,
+      "loss": 0.5313,
+      "step": 2874
+    },
+    {
+      "epoch": 0.736,
+      "grad_norm": 2.609375,
+      "learning_rate": 3.773995392884566e-05,
+      "loss": 0.5324,
+      "step": 2875
+    },
+    {
+      "epoch": 0.736256,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.7735688081221744e-05,
+      "loss": 0.5509,
+      "step": 2876
+    },
+    {
+      "epoch": 0.736512,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.773142223359782e-05,
+      "loss": 0.546,
+      "step": 2877
+    },
+    {
+      "epoch": 0.736768,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.772715638597389e-05,
+      "loss": 0.5728,
+      "step": 2878
+    },
+    {
+      "epoch": 0.737024,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.772289053834997e-05,
+      "loss": 0.531,
+      "step": 2879
+    },
+    {
+      "epoch": 0.73728,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.771862469072605e-05,
+      "loss": 0.5591,
+      "step": 2880
+    },
+    {
+      "epoch": 0.737536,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.771435884310213e-05,
+      "loss": 0.4895,
+      "step": 2881
+    },
+    {
+      "epoch": 0.737792,
+      "grad_norm": 2.5,
+      "learning_rate": 3.7710092995478204e-05,
+      "loss": 0.5357,
+      "step": 2882
+    },
+    {
+      "epoch": 0.738048,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.770582714785428e-05,
+      "loss": 0.5246,
+      "step": 2883
+    },
+    {
+      "epoch": 0.738304,
+      "grad_norm": 2.703125,
+      "learning_rate": 3.770156130023036e-05,
+      "loss": 0.5432,
+      "step": 2884
+    },
+    {
+      "epoch": 0.73856,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.769729545260643e-05,
+      "loss": 0.5834,
+      "step": 2885
+    },
+    {
+      "epoch": 0.738816,
+      "grad_norm": 2.75,
+      "learning_rate": 3.769302960498251e-05,
+      "loss": 0.5425,
+      "step": 2886
+    },
+    {
+      "epoch": 0.739072,
+      "grad_norm": 2.59375,
+      "learning_rate": 3.768876375735859e-05,
+      "loss": 0.515,
+      "step": 2887
+    },
+    {
+      "epoch": 0.739328,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.768449790973467e-05,
+      "loss": 0.5616,
+      "step": 2888
+    },
+    {
+      "epoch": 0.739584,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.7680232062110744e-05,
+      "loss": 0.5092,
+      "step": 2889
+    },
+    {
+      "epoch": 0.73984,
+      "grad_norm": 2.71875,
+      "learning_rate": 3.767596621448682e-05,
+      "loss": 0.5422,
+      "step": 2890
+    },
+    {
+      "epoch": 0.740096,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.76717003668629e-05,
+      "loss": 0.5383,
+      "step": 2891
+    },
+    {
+      "epoch": 0.740352,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.766743451923897e-05,
+      "loss": 0.5411,
+      "step": 2892
+    },
+    {
+      "epoch": 0.740608,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.7663168671615054e-05,
+      "loss": 0.5138,
+      "step": 2893
+    },
+    {
+      "epoch": 0.740864,
+      "grad_norm": 2.578125,
+      "learning_rate": 3.765890282399113e-05,
+      "loss": 0.5178,
+      "step": 2894
+    },
+    {
+      "epoch": 0.74112,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.765463697636721e-05,
+      "loss": 0.5458,
+      "step": 2895
+    },
+    {
+      "epoch": 0.741376,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.7650371128743284e-05,
+      "loss": 0.5176,
+      "step": 2896
+    },
+    {
+      "epoch": 0.741632,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.764610528111936e-05,
+      "loss": 0.5755,
+      "step": 2897
+    },
+    {
+      "epoch": 0.741888,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.764183943349543e-05,
+      "loss": 0.4827,
+      "step": 2898
+    },
+    {
+      "epoch": 0.742144,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.7637573585871514e-05,
+      "loss": 0.4934,
+      "step": 2899
+    },
+    {
+      "epoch": 0.7424,
+      "grad_norm": 2.6875,
+      "learning_rate": 3.7633307738247595e-05,
+      "loss": 0.5073,
+      "step": 2900
+    },
+    {
+      "epoch": 0.742656,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.762904189062367e-05,
+      "loss": 0.471,
+      "step": 2901
+    },
+    {
+      "epoch": 0.742912,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.762477604299974e-05,
+      "loss": 0.4995,
+      "step": 2902
+    },
+    {
+      "epoch": 0.743168,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.7620510195375824e-05,
+      "loss": 0.5658,
+      "step": 2903
+    },
+    {
+      "epoch": 0.743424,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.76162443477519e-05,
+      "loss": 0.4659,
+      "step": 2904
+    },
+    {
+      "epoch": 0.74368,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.761197850012797e-05,
+      "loss": 0.5344,
+      "step": 2905
+    },
+    {
+      "epoch": 0.743936,
+      "grad_norm": 2.25,
+      "learning_rate": 3.7607712652504054e-05,
+      "loss": 0.4914,
+      "step": 2906
+    },
+    {
+      "epoch": 0.744192,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.7603446804880135e-05,
+      "loss": 0.5313,
+      "step": 2907
+    },
+    {
+      "epoch": 0.744448,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.759918095725621e-05,
+      "loss": 0.5504,
+      "step": 2908
+    },
+    {
+      "epoch": 0.744704,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.759491510963228e-05,
+      "loss": 0.5501,
+      "step": 2909
+    },
+    {
+      "epoch": 0.74496,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.7590649262008364e-05,
+      "loss": 0.5284,
+      "step": 2910
+    },
+    {
+      "epoch": 0.745216,
+      "grad_norm": 2.5,
+      "learning_rate": 3.758638341438444e-05,
+      "loss": 0.5348,
+      "step": 2911
+    },
+    {
+      "epoch": 0.745472,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.758211756676052e-05,
+      "loss": 0.5159,
+      "step": 2912
+    },
+    {
+      "epoch": 0.745728,
+      "grad_norm": 2.5,
+      "learning_rate": 3.7577851719136594e-05,
+      "loss": 0.5285,
+      "step": 2913
+    },
+    {
+      "epoch": 0.745984,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.7573585871512675e-05,
+      "loss": 0.5383,
+      "step": 2914
+    },
+    {
+      "epoch": 0.74624,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.756932002388875e-05,
+      "loss": 0.5424,
+      "step": 2915
+    },
+    {
+      "epoch": 0.746496,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.7565054176264823e-05,
+      "loss": 0.5018,
+      "step": 2916
+    },
+    {
+      "epoch": 0.746752,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.75607883286409e-05,
+      "loss": 0.5241,
+      "step": 2917
+    },
+    {
+      "epoch": 0.747008,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.755652248101698e-05,
+      "loss": 0.5353,
+      "step": 2918
+    },
+    {
+      "epoch": 0.747264,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.755225663339306e-05,
+      "loss": 0.5323,
+      "step": 2919
+    },
+    {
+      "epoch": 0.74752,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.7547990785769134e-05,
+      "loss": 0.5541,
+      "step": 2920
+    },
+    {
+      "epoch": 0.747776,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.7543724938145215e-05,
+      "loss": 0.5071,
+      "step": 2921
+    },
+    {
+      "epoch": 0.748032,
+      "grad_norm": 2.5625,
+      "learning_rate": 3.753945909052129e-05,
+      "loss": 0.5269,
+      "step": 2922
+    },
+    {
+      "epoch": 0.748288,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.7535193242897364e-05,
+      "loss": 0.5644,
+      "step": 2923
+    },
+    {
+      "epoch": 0.748544,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.753092739527344e-05,
+      "loss": 0.54,
+      "step": 2924
+    },
+    {
+      "epoch": 0.7488,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.752666154764952e-05,
+      "loss": 0.5532,
+      "step": 2925
+    },
+    {
+      "epoch": 0.749056,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.75223957000256e-05,
+      "loss": 0.5521,
+      "step": 2926
+    },
+    {
+      "epoch": 0.749312,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.7518129852401674e-05,
+      "loss": 0.5328,
+      "step": 2927
+    },
+    {
+      "epoch": 0.749568,
+      "grad_norm": 2.25,
+      "learning_rate": 3.751386400477775e-05,
+      "loss": 0.5205,
+      "step": 2928
+    },
+    {
+      "epoch": 0.749824,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.750959815715383e-05,
+      "loss": 0.5782,
+      "step": 2929
+    },
+    {
+      "epoch": 0.75008,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.7505332309529904e-05,
+      "loss": 0.5072,
+      "step": 2930
+    },
+    {
+      "epoch": 0.750336,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.750106646190598e-05,
+      "loss": 0.5062,
+      "step": 2931
+    },
+    {
+      "epoch": 0.750592,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.749680061428206e-05,
+      "loss": 0.4968,
+      "step": 2932
+    },
+    {
+      "epoch": 0.750848,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.749253476665814e-05,
+      "loss": 0.5192,
+      "step": 2933
+    },
+    {
+      "epoch": 0.751104,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.7488268919034214e-05,
+      "loss": 0.5496,
+      "step": 2934
+    },
+    {
+      "epoch": 0.75136,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.748400307141029e-05,
+      "loss": 0.5517,
+      "step": 2935
+    },
+    {
+      "epoch": 0.751616,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.747973722378637e-05,
+      "loss": 0.5062,
+      "step": 2936
+    },
+    {
+      "epoch": 0.751872,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.7475471376162444e-05,
+      "loss": 0.5396,
+      "step": 2937
+    },
+    {
+      "epoch": 0.752128,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.7471205528538525e-05,
+      "loss": 0.4957,
+      "step": 2938
+    },
+    {
+      "epoch": 0.752384,
+      "grad_norm": 2.375,
+      "learning_rate": 3.74669396809146e-05,
+      "loss": 0.5613,
+      "step": 2939
+    },
+    {
+      "epoch": 0.75264,
+      "grad_norm": 2.25,
+      "learning_rate": 3.746267383329068e-05,
+      "loss": 0.5278,
+      "step": 2940
+    },
+    {
+      "epoch": 0.752896,
+      "grad_norm": 2.375,
+      "learning_rate": 3.7458407985666755e-05,
+      "loss": 0.5645,
+      "step": 2941
+    },
+    {
+      "epoch": 0.753152,
+      "grad_norm": 2.71875,
+      "learning_rate": 3.745414213804283e-05,
+      "loss": 0.5337,
+      "step": 2942
+    },
+    {
+      "epoch": 0.753408,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.74498762904189e-05,
+      "loss": 0.5116,
+      "step": 2943
+    },
+    {
+      "epoch": 0.753664,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.7445610442794984e-05,
+      "loss": 0.5092,
+      "step": 2944
+    },
+    {
+      "epoch": 0.75392,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.7441344595171065e-05,
+      "loss": 0.5412,
+      "step": 2945
+    },
+    {
+      "epoch": 0.754176,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.743707874754714e-05,
+      "loss": 0.5157,
+      "step": 2946
+    },
+    {
+      "epoch": 0.754432,
+      "grad_norm": 2.625,
+      "learning_rate": 3.7432812899923214e-05,
+      "loss": 0.5158,
+      "step": 2947
+    },
+    {
+      "epoch": 0.754688,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.7428547052299295e-05,
+      "loss": 0.5337,
+      "step": 2948
+    },
+    {
+      "epoch": 0.754944,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.742428120467537e-05,
+      "loss": 0.5151,
+      "step": 2949
+    },
+    {
+      "epoch": 0.7552,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.7420015357051443e-05,
+      "loss": 0.5141,
+      "step": 2950
+    },
+    {
+      "epoch": 0.755456,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.7415749509427524e-05,
+      "loss": 0.5102,
+      "step": 2951
+    },
+    {
+      "epoch": 0.755712,
+      "grad_norm": 2.671875,
+      "learning_rate": 3.7411483661803606e-05,
+      "loss": 0.5338,
+      "step": 2952
+    },
+    {
+      "epoch": 0.755968,
+      "grad_norm": 2.734375,
+      "learning_rate": 3.740721781417968e-05,
+      "loss": 0.5477,
+      "step": 2953
+    },
+    {
+      "epoch": 0.756224,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.7402951966555754e-05,
+      "loss": 0.5578,
+      "step": 2954
+    },
+    {
+      "epoch": 0.75648,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.7398686118931835e-05,
+      "loss": 0.5085,
+      "step": 2955
+    },
+    {
+      "epoch": 0.756736,
+      "grad_norm": 2.578125,
+      "learning_rate": 3.739442027130791e-05,
+      "loss": 0.61,
+      "step": 2956
+    },
+    {
+      "epoch": 0.756992,
+      "grad_norm": 2.375,
+      "learning_rate": 3.739015442368399e-05,
+      "loss": 0.5011,
+      "step": 2957
+    },
+    {
+      "epoch": 0.757248,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.7385888576060065e-05,
+      "loss": 0.5368,
+      "step": 2958
+    },
+    {
+      "epoch": 0.757504,
+      "grad_norm": 2.9375,
+      "learning_rate": 3.7381622728436146e-05,
+      "loss": 0.5501,
+      "step": 2959
+    },
+    {
+      "epoch": 0.75776,
+      "grad_norm": 2.71875,
+      "learning_rate": 3.737735688081222e-05,
+      "loss": 0.5794,
+      "step": 2960
+    },
+    {
+      "epoch": 0.758016,
+      "grad_norm": 2.59375,
+      "learning_rate": 3.7373091033188294e-05,
+      "loss": 0.561,
+      "step": 2961
+    },
+    {
+      "epoch": 0.758272,
+      "grad_norm": 2.5625,
+      "learning_rate": 3.736882518556437e-05,
+      "loss": 0.5714,
+      "step": 2962
+    },
+    {
+      "epoch": 0.758528,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.736455933794045e-05,
+      "loss": 0.5144,
+      "step": 2963
+    },
+    {
+      "epoch": 0.758784,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.736029349031653e-05,
+      "loss": 0.5546,
+      "step": 2964
+    },
+    {
+      "epoch": 0.75904,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.7356027642692605e-05,
+      "loss": 0.5475,
+      "step": 2965
+    },
+    {
+      "epoch": 0.759296,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.735176179506868e-05,
+      "loss": 0.5825,
+      "step": 2966
+    },
+    {
+      "epoch": 0.759552,
+      "grad_norm": 2.59375,
+      "learning_rate": 3.734749594744476e-05,
+      "loss": 0.5201,
+      "step": 2967
+    },
+    {
+      "epoch": 0.759808,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.7343230099820834e-05,
+      "loss": 0.518,
+      "step": 2968
+    },
+    {
+      "epoch": 0.760064,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.733896425219691e-05,
+      "loss": 0.5632,
+      "step": 2969
+    },
+    {
+      "epoch": 0.76032,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.733469840457299e-05,
+      "loss": 0.5396,
+      "step": 2970
+    },
+    {
+      "epoch": 0.760576,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.733043255694907e-05,
+      "loss": 0.5063,
+      "step": 2971
+    },
+    {
+      "epoch": 0.760832,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.7326166709325145e-05,
+      "loss": 0.5549,
+      "step": 2972
+    },
+    {
+      "epoch": 0.761088,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.732190086170122e-05,
+      "loss": 0.5556,
+      "step": 2973
+    },
+    {
+      "epoch": 0.761344,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.73176350140773e-05,
+      "loss": 0.5381,
+      "step": 2974
+    },
+    {
+      "epoch": 0.7616,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.7313369166453375e-05,
+      "loss": 0.5301,
+      "step": 2975
+    },
+    {
+      "epoch": 0.761856,
+      "grad_norm": 2.6875,
+      "learning_rate": 3.7309103318829456e-05,
+      "loss": 0.5562,
+      "step": 2976
+    },
+    {
+      "epoch": 0.762112,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.730483747120553e-05,
+      "loss": 0.541,
+      "step": 2977
+    },
+    {
+      "epoch": 0.762368,
+      "grad_norm": 2.65625,
+      "learning_rate": 3.730057162358161e-05,
+      "loss": 0.5483,
+      "step": 2978
+    },
+    {
+      "epoch": 0.762624,
+      "grad_norm": 2.5625,
+      "learning_rate": 3.7296305775957685e-05,
+      "loss": 0.566,
+      "step": 2979
+    },
+    {
+      "epoch": 0.76288,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.729203992833376e-05,
+      "loss": 0.5233,
+      "step": 2980
+    },
+    {
+      "epoch": 0.763136,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.7287774080709834e-05,
+      "loss": 0.5305,
+      "step": 2981
+    },
+    {
+      "epoch": 0.763392,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.7283508233085915e-05,
+      "loss": 0.5209,
+      "step": 2982
+    },
+    {
+      "epoch": 0.763648,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.7279242385461996e-05,
+      "loss": 0.5226,
+      "step": 2983
+    },
+    {
+      "epoch": 0.763904,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.727497653783807e-05,
+      "loss": 0.5336,
+      "step": 2984
+    },
+    {
+      "epoch": 0.76416,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.727071069021415e-05,
+      "loss": 0.5442,
+      "step": 2985
+    },
+    {
+      "epoch": 0.764416,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.7266444842590225e-05,
+      "loss": 0.5363,
+      "step": 2986
+    },
+    {
+      "epoch": 0.764672,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.72621789949663e-05,
+      "loss": 0.5225,
+      "step": 2987
+    },
+    {
+      "epoch": 0.764928,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.7257913147342374e-05,
+      "loss": 0.5034,
+      "step": 2988
+    },
+    {
+      "epoch": 0.765184,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.7253647299718455e-05,
+      "loss": 0.5171,
+      "step": 2989
+    },
+    {
+      "epoch": 0.76544,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.7249381452094536e-05,
+      "loss": 0.5512,
+      "step": 2990
+    },
+    {
+      "epoch": 0.765696,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.724511560447061e-05,
+      "loss": 0.5298,
+      "step": 2991
+    },
+    {
+      "epoch": 0.765952,
+      "grad_norm": 2.59375,
+      "learning_rate": 3.7240849756846685e-05,
+      "loss": 0.5619,
+      "step": 2992
+    },
+    {
+      "epoch": 0.766208,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.7236583909222766e-05,
+      "loss": 0.5227,
+      "step": 2993
+    },
+    {
+      "epoch": 0.766464,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.723231806159884e-05,
+      "loss": 0.5063,
+      "step": 2994
+    },
+    {
+      "epoch": 0.76672,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.722805221397492e-05,
+      "loss": 0.5323,
+      "step": 2995
+    },
+    {
+      "epoch": 0.766976,
+      "grad_norm": 2.5,
+      "learning_rate": 3.7223786366350995e-05,
+      "loss": 0.5344,
+      "step": 2996
+    },
+    {
+      "epoch": 0.767232,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.7219520518727076e-05,
+      "loss": 0.5536,
+      "step": 2997
+    },
+    {
+      "epoch": 0.767488,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.721525467110315e-05,
+      "loss": 0.5457,
+      "step": 2998
+    },
+    {
+      "epoch": 0.767744,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.7210988823479225e-05,
+      "loss": 0.5666,
+      "step": 2999
+    },
+    {
+      "epoch": 0.768,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.72067229758553e-05,
+      "loss": 0.5344,
+      "step": 3000
+    },
+    {
+      "epoch": 0.768256,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.720245712823138e-05,
+      "loss": 0.4925,
+      "step": 3001
+    },
+    {
+      "epoch": 0.768512,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.719819128060746e-05,
+      "loss": 0.4982,
+      "step": 3002
+    },
+    {
+      "epoch": 0.768768,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.7193925432983535e-05,
+      "loss": 0.5292,
+      "step": 3003
+    },
+    {
+      "epoch": 0.769024,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.7189659585359616e-05,
+      "loss": 0.5397,
+      "step": 3004
+    },
+    {
+      "epoch": 0.76928,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.718539373773569e-05,
+      "loss": 0.5474,
+      "step": 3005
+    },
+    {
+      "epoch": 0.769536,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.7181127890111765e-05,
+      "loss": 0.5721,
+      "step": 3006
+    },
+    {
+      "epoch": 0.769792,
+      "grad_norm": 2.25,
+      "learning_rate": 3.717686204248784e-05,
+      "loss": 0.5616,
+      "step": 3007
+    },
+    {
+      "epoch": 0.770048,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.717259619486392e-05,
+      "loss": 0.5374,
+      "step": 3008
+    },
+    {
+      "epoch": 0.770304,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.716833034724e-05,
+      "loss": 0.5552,
+      "step": 3009
+    },
+    {
+      "epoch": 0.77056,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.7164064499616076e-05,
+      "loss": 0.5405,
+      "step": 3010
+    },
+    {
+      "epoch": 0.770816,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.715979865199215e-05,
+      "loss": 0.5579,
+      "step": 3011
+    },
+    {
+      "epoch": 0.771072,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.715553280436823e-05,
+      "loss": 0.5503,
+      "step": 3012
+    },
+    {
+      "epoch": 0.771328,
+      "grad_norm": 2.703125,
+      "learning_rate": 3.7151266956744305e-05,
+      "loss": 0.535,
+      "step": 3013
+    },
+    {
+      "epoch": 0.771584,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.7147001109120386e-05,
+      "loss": 0.5133,
+      "step": 3014
+    },
+    {
+      "epoch": 0.77184,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.714273526149646e-05,
+      "loss": 0.5158,
+      "step": 3015
+    },
+    {
+      "epoch": 0.772096,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.713846941387254e-05,
+      "loss": 0.499,
+      "step": 3016
+    },
+    {
+      "epoch": 0.772352,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.7134203566248616e-05,
+      "loss": 0.5143,
+      "step": 3017
+    },
+    {
+      "epoch": 0.772608,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.712993771862469e-05,
+      "loss": 0.5513,
+      "step": 3018
+    },
+    {
+      "epoch": 0.772864,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.712567187100077e-05,
+      "loss": 0.5255,
+      "step": 3019
+    },
+    {
+      "epoch": 0.77312,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.7121406023376845e-05,
+      "loss": 0.5089,
+      "step": 3020
+    },
+    {
+      "epoch": 0.773376,
+      "grad_norm": 2.59375,
+      "learning_rate": 3.7117140175752926e-05,
+      "loss": 0.5586,
+      "step": 3021
+    },
+    {
+      "epoch": 0.773632,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.7112874328129e-05,
+      "loss": 0.5278,
+      "step": 3022
+    },
+    {
+      "epoch": 0.773888,
+      "grad_norm": 2.578125,
+      "learning_rate": 3.710860848050508e-05,
+      "loss": 0.5193,
+      "step": 3023
+    },
+    {
+      "epoch": 0.774144,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.7104342632881156e-05,
+      "loss": 0.5213,
+      "step": 3024
+    },
+    {
+      "epoch": 0.7744,
+      "grad_norm": 2.671875,
+      "learning_rate": 3.710007678525723e-05,
+      "loss": 0.5165,
+      "step": 3025
+    },
+    {
+      "epoch": 0.774656,
+      "grad_norm": 2.375,
+      "learning_rate": 3.7095810937633304e-05,
+      "loss": 0.5044,
+      "step": 3026
+    },
+    {
+      "epoch": 0.774912,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.7091545090009386e-05,
+      "loss": 0.5625,
+      "step": 3027
+    },
+    {
+      "epoch": 0.775168,
+      "grad_norm": 2.25,
+      "learning_rate": 3.7087279242385467e-05,
+      "loss": 0.5367,
+      "step": 3028
+    },
+    {
+      "epoch": 0.775424,
+      "grad_norm": 2.625,
+      "learning_rate": 3.708301339476154e-05,
+      "loss": 0.5435,
+      "step": 3029
+    },
+    {
+      "epoch": 0.77568,
+      "grad_norm": 2.75,
+      "learning_rate": 3.7078747547137615e-05,
+      "loss": 0.5228,
+      "step": 3030
+    },
+    {
+      "epoch": 0.775936,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.7074481699513696e-05,
+      "loss": 0.5156,
+      "step": 3031
+    },
+    {
+      "epoch": 0.776192,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.707021585188977e-05,
+      "loss": 0.5861,
+      "step": 3032
+    },
+    {
+      "epoch": 0.776448,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.7065950004265845e-05,
+      "loss": 0.5742,
+      "step": 3033
+    },
+    {
+      "epoch": 0.776704,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.706168415664193e-05,
+      "loss": 0.5489,
+      "step": 3034
+    },
+    {
+      "epoch": 0.77696,
+      "grad_norm": 2.609375,
+      "learning_rate": 3.705741830901801e-05,
+      "loss": 0.538,
+      "step": 3035
+    },
+    {
+      "epoch": 0.777216,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.705315246139408e-05,
+      "loss": 0.5495,
+      "step": 3036
+    },
+    {
+      "epoch": 0.777472,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.7048886613770155e-05,
+      "loss": 0.542,
+      "step": 3037
+    },
+    {
+      "epoch": 0.777728,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.7044620766146236e-05,
+      "loss": 0.5385,
+      "step": 3038
+    },
+    {
+      "epoch": 0.777984,
+      "grad_norm": 2.5,
+      "learning_rate": 3.704035491852231e-05,
+      "loss": 0.5691,
+      "step": 3039
+    },
+    {
+      "epoch": 0.77824,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.703608907089839e-05,
+      "loss": 0.5186,
+      "step": 3040
+    },
+    {
+      "epoch": 0.778496,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.7031823223274466e-05,
+      "loss": 0.5552,
+      "step": 3041
+    },
+    {
+      "epoch": 0.778752,
+      "grad_norm": 2.25,
+      "learning_rate": 3.702755737565055e-05,
+      "loss": 0.5314,
+      "step": 3042
+    },
+    {
+      "epoch": 0.779008,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.702329152802662e-05,
+      "loss": 0.4712,
+      "step": 3043
+    },
+    {
+      "epoch": 0.779264,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.7019025680402696e-05,
+      "loss": 0.4886,
+      "step": 3044
+    },
+    {
+      "epoch": 0.77952,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.701475983277877e-05,
+      "loss": 0.5254,
+      "step": 3045
+    },
+    {
+      "epoch": 0.779776,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.701049398515485e-05,
+      "loss": 0.4959,
+      "step": 3046
+    },
+    {
+      "epoch": 0.780032,
+      "grad_norm": 2.5,
+      "learning_rate": 3.700622813753093e-05,
+      "loss": 0.5091,
+      "step": 3047
+    },
+    {
+      "epoch": 0.780288,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.7001962289907006e-05,
+      "loss": 0.4703,
+      "step": 3048
+    },
+    {
+      "epoch": 0.780544,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.699769644228308e-05,
+      "loss": 0.5388,
+      "step": 3049
+    },
+    {
+      "epoch": 0.7808,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.699343059465916e-05,
+      "loss": 0.5241,
+      "step": 3050
+    },
+    {
+      "epoch": 0.781056,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.6989164747035236e-05,
+      "loss": 0.5328,
+      "step": 3051
+    },
+    {
+      "epoch": 0.781312,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.698489889941131e-05,
+      "loss": 0.4998,
+      "step": 3052
+    },
+    {
+      "epoch": 0.781568,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.69806330517874e-05,
+      "loss": 0.5023,
+      "step": 3053
+    },
+    {
+      "epoch": 0.781824,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.697636720416347e-05,
+      "loss": 0.5295,
+      "step": 3054
+    },
+    {
+      "epoch": 0.78208,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.6972101356539546e-05,
+      "loss": 0.5629,
+      "step": 3055
+    },
+    {
+      "epoch": 0.782336,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.696783550891562e-05,
+      "loss": 0.5331,
+      "step": 3056
+    },
+    {
+      "epoch": 0.782592,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.69635696612917e-05,
+      "loss": 0.5342,
+      "step": 3057
+    },
+    {
+      "epoch": 0.782848,
+      "grad_norm": 2.25,
+      "learning_rate": 3.6959303813667776e-05,
+      "loss": 0.5812,
+      "step": 3058
+    },
+    {
+      "epoch": 0.783104,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.695503796604386e-05,
+      "loss": 0.5166,
+      "step": 3059
+    },
+    {
+      "epoch": 0.78336,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.695077211841993e-05,
+      "loss": 0.526,
+      "step": 3060
+    },
+    {
+      "epoch": 0.783616,
+      "grad_norm": 2.375,
+      "learning_rate": 3.694650627079601e-05,
+      "loss": 0.526,
+      "step": 3061
+    },
+    {
+      "epoch": 0.783872,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.6942240423172087e-05,
+      "loss": 0.5242,
+      "step": 3062
+    },
+    {
+      "epoch": 0.784128,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.693797457554816e-05,
+      "loss": 0.5494,
+      "step": 3063
+    },
+    {
+      "epoch": 0.784384,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.6933708727924235e-05,
+      "loss": 0.5128,
+      "step": 3064
+    },
+    {
+      "epoch": 0.78464,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.6929442880300316e-05,
+      "loss": 0.5177,
+      "step": 3065
+    },
+    {
+      "epoch": 0.784896,
+      "grad_norm": 2.59375,
+      "learning_rate": 3.69251770326764e-05,
+      "loss": 0.5656,
+      "step": 3066
+    },
+    {
+      "epoch": 0.785152,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.692091118505247e-05,
+      "loss": 0.5562,
+      "step": 3067
+    },
+    {
+      "epoch": 0.785408,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.691664533742855e-05,
+      "loss": 0.5143,
+      "step": 3068
+    },
+    {
+      "epoch": 0.785664,
+      "grad_norm": 2.703125,
+      "learning_rate": 3.691237948980463e-05,
+      "loss": 0.5163,
+      "step": 3069
+    },
+    {
+      "epoch": 0.78592,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.69081136421807e-05,
+      "loss": 0.5244,
+      "step": 3070
+    },
+    {
+      "epoch": 0.786176,
+      "grad_norm": 2.5625,
+      "learning_rate": 3.6903847794556775e-05,
+      "loss": 0.5577,
+      "step": 3071
+    },
+    {
+      "epoch": 0.786432,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.6899581946932856e-05,
+      "loss": 0.5188,
+      "step": 3072
+    },
+    {
+      "epoch": 0.786688,
+      "grad_norm": 2.5,
+      "learning_rate": 3.689531609930894e-05,
+      "loss": 0.5137,
+      "step": 3073
+    },
+    {
+      "epoch": 0.786944,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.689105025168501e-05,
+      "loss": 0.5191,
+      "step": 3074
+    },
+    {
+      "epoch": 0.7872,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.6886784404061086e-05,
+      "loss": 0.5663,
+      "step": 3075
+    },
+    {
+      "epoch": 0.787456,
+      "grad_norm": 2.625,
+      "learning_rate": 3.688251855643717e-05,
+      "loss": 0.567,
+      "step": 3076
+    },
+    {
+      "epoch": 0.787712,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.687825270881324e-05,
+      "loss": 0.5453,
+      "step": 3077
+    },
+    {
+      "epoch": 0.787968,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.687398686118932e-05,
+      "loss": 0.5114,
+      "step": 3078
+    },
+    {
+      "epoch": 0.788224,
+      "grad_norm": 2.375,
+      "learning_rate": 3.6869721013565396e-05,
+      "loss": 0.5343,
+      "step": 3079
+    },
+    {
+      "epoch": 0.78848,
+      "grad_norm": 2.609375,
+      "learning_rate": 3.686545516594148e-05,
+      "loss": 0.5537,
+      "step": 3080
+    },
+    {
+      "epoch": 0.788736,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.686118931831755e-05,
+      "loss": 0.552,
+      "step": 3081
+    },
+    {
+      "epoch": 0.788992,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.6856923470693626e-05,
+      "loss": 0.4963,
+      "step": 3082
+    },
+    {
+      "epoch": 0.789248,
+      "grad_norm": 2.5625,
+      "learning_rate": 3.685265762306971e-05,
+      "loss": 0.5494,
+      "step": 3083
+    },
+    {
+      "epoch": 0.789504,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.684839177544578e-05,
+      "loss": 0.5183,
+      "step": 3084
+    },
+    {
+      "epoch": 0.78976,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.684412592782186e-05,
+      "loss": 0.4899,
+      "step": 3085
+    },
+    {
+      "epoch": 0.790016,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.683986008019794e-05,
+      "loss": 0.545,
+      "step": 3086
+    },
+    {
+      "epoch": 0.790272,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.683559423257402e-05,
+      "loss": 0.5656,
+      "step": 3087
+    },
+    {
+      "epoch": 0.790528,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.683132838495009e-05,
+      "loss": 0.5002,
+      "step": 3088
+    },
+    {
+      "epoch": 0.790784,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.6827062537326166e-05,
+      "loss": 0.5374,
+      "step": 3089
+    },
+    {
+      "epoch": 0.79104,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.682279668970224e-05,
+      "loss": 0.4902,
+      "step": 3090
+    },
+    {
+      "epoch": 0.791296,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.681853084207832e-05,
+      "loss": 0.5234,
+      "step": 3091
+    },
+    {
+      "epoch": 0.791552,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.68142649944544e-05,
+      "loss": 0.5003,
+      "step": 3092
+    },
+    {
+      "epoch": 0.791808,
+      "grad_norm": 2.125,
+      "learning_rate": 3.680999914683048e-05,
+      "loss": 0.5273,
+      "step": 3093
+    },
+    {
+      "epoch": 0.792064,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.680573329920655e-05,
+      "loss": 0.5342,
+      "step": 3094
+    },
+    {
+      "epoch": 0.79232,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.680146745158263e-05,
+      "loss": 0.5125,
+      "step": 3095
+    },
+    {
+      "epoch": 0.792576,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.6797201603958706e-05,
+      "loss": 0.5615,
+      "step": 3096
+    },
+    {
+      "epoch": 0.792832,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.679293575633479e-05,
+      "loss": 0.4868,
+      "step": 3097
+    },
+    {
+      "epoch": 0.793088,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.678866990871086e-05,
+      "loss": 0.5128,
+      "step": 3098
+    },
+    {
+      "epoch": 0.793344,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.678440406108694e-05,
+      "loss": 0.5118,
+      "step": 3099
+    },
+    {
+      "epoch": 0.7936,
+      "grad_norm": 2.65625,
+      "learning_rate": 3.678013821346302e-05,
+      "loss": 0.561,
+      "step": 3100
+    },
+    {
+      "epoch": 0.793856,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.677587236583909e-05,
+      "loss": 0.4941,
+      "step": 3101
+    },
+    {
+      "epoch": 0.794112,
+      "grad_norm": 2.703125,
+      "learning_rate": 3.677160651821517e-05,
+      "loss": 0.5454,
+      "step": 3102
+    },
+    {
+      "epoch": 0.794368,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.676734067059125e-05,
+      "loss": 0.505,
+      "step": 3103
+    },
+    {
+      "epoch": 0.794624,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.676307482296733e-05,
+      "loss": 0.5039,
+      "step": 3104
+    },
+    {
+      "epoch": 0.79488,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.67588089753434e-05,
+      "loss": 0.5656,
+      "step": 3105
+    },
+    {
+      "epoch": 0.795136,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.675454312771948e-05,
+      "loss": 0.4813,
+      "step": 3106
+    },
+    {
+      "epoch": 0.795392,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.675027728009556e-05,
+      "loss": 0.471,
+      "step": 3107
+    },
+    {
+      "epoch": 0.795648,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.674601143247163e-05,
+      "loss": 0.5699,
+      "step": 3108
+    },
+    {
+      "epoch": 0.795904,
+      "grad_norm": 2.375,
+      "learning_rate": 3.6741745584847706e-05,
+      "loss": 0.5182,
+      "step": 3109
+    },
+    {
+      "epoch": 0.79616,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.673747973722379e-05,
+      "loss": 0.5166,
+      "step": 3110
+    },
+    {
+      "epoch": 0.796416,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.673321388959987e-05,
+      "loss": 0.5042,
+      "step": 3111
+    },
+    {
+      "epoch": 0.796672,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.672894804197594e-05,
+      "loss": 0.537,
+      "step": 3112
+    },
+    {
+      "epoch": 0.796928,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.6724682194352016e-05,
+      "loss": 0.5008,
+      "step": 3113
+    },
+    {
+      "epoch": 0.797184,
+      "grad_norm": 2.5625,
+      "learning_rate": 3.67204163467281e-05,
+      "loss": 0.5311,
+      "step": 3114
+    },
+    {
+      "epoch": 0.79744,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.671615049910417e-05,
+      "loss": 0.5458,
+      "step": 3115
+    },
+    {
+      "epoch": 0.797696,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.671188465148025e-05,
+      "loss": 0.5197,
+      "step": 3116
+    },
+    {
+      "epoch": 0.797952,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.6707618803856334e-05,
+      "loss": 0.4992,
+      "step": 3117
+    },
+    {
+      "epoch": 0.798208,
+      "grad_norm": 2.671875,
+      "learning_rate": 3.670335295623241e-05,
+      "loss": 0.5448,
+      "step": 3118
+    },
+    {
+      "epoch": 0.798464,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.669908710860848e-05,
+      "loss": 0.4939,
+      "step": 3119
+    },
+    {
+      "epoch": 0.79872,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.6694821260984557e-05,
+      "loss": 0.5039,
+      "step": 3120
+    },
+    {
+      "epoch": 0.798976,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.669055541336064e-05,
+      "loss": 0.5228,
+      "step": 3121
+    },
+    {
+      "epoch": 0.799232,
+      "grad_norm": 2.578125,
+      "learning_rate": 3.668628956573671e-05,
+      "loss": 0.5093,
+      "step": 3122
+    },
+    {
+      "epoch": 0.799488,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.668202371811279e-05,
+      "loss": 0.5227,
+      "step": 3123
+    },
+    {
+      "epoch": 0.799744,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.667775787048887e-05,
+      "loss": 0.5115,
+      "step": 3124
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.667349202286495e-05,
+      "loss": 0.5152,
+      "step": 3125
+    },
+    {
+      "epoch": 0.800256,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.666922617524102e-05,
+      "loss": 0.5024,
+      "step": 3126
+    },
+    {
+      "epoch": 0.800512,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.66649603276171e-05,
+      "loss": 0.508,
+      "step": 3127
+    },
+    {
+      "epoch": 0.800768,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.666069447999317e-05,
+      "loss": 0.508,
+      "step": 3128
+    },
+    {
+      "epoch": 0.801024,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.665642863236925e-05,
+      "loss": 0.5357,
+      "step": 3129
+    },
+    {
+      "epoch": 0.80128,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.665216278474533e-05,
+      "loss": 0.5192,
+      "step": 3130
+    },
+    {
+      "epoch": 0.801536,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.664789693712141e-05,
+      "loss": 0.5078,
+      "step": 3131
+    },
+    {
+      "epoch": 0.801792,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.664363108949748e-05,
+      "loss": 0.5018,
+      "step": 3132
+    },
+    {
+      "epoch": 0.802048,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.663936524187356e-05,
+      "loss": 0.4882,
+      "step": 3133
+    },
+    {
+      "epoch": 0.802304,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.663509939424964e-05,
+      "loss": 0.5413,
+      "step": 3134
+    },
+    {
+      "epoch": 0.80256,
+      "grad_norm": 2.828125,
+      "learning_rate": 3.663083354662572e-05,
+      "loss": 0.5564,
+      "step": 3135
+    },
+    {
+      "epoch": 0.802816,
+      "grad_norm": 2.5625,
+      "learning_rate": 3.66265676990018e-05,
+      "loss": 0.5226,
+      "step": 3136
+    },
+    {
+      "epoch": 0.803072,
+      "grad_norm": 2.59375,
+      "learning_rate": 3.662230185137787e-05,
+      "loss": 0.5545,
+      "step": 3137
+    },
+    {
+      "epoch": 0.803328,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.661803600375395e-05,
+      "loss": 0.4932,
+      "step": 3138
+    },
+    {
+      "epoch": 0.803584,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.661377015613002e-05,
+      "loss": 0.514,
+      "step": 3139
+    },
+    {
+      "epoch": 0.80384,
+      "grad_norm": 2.5,
+      "learning_rate": 3.66095043085061e-05,
+      "loss": 0.5145,
+      "step": 3140
+    },
+    {
+      "epoch": 0.804096,
+      "grad_norm": 2.5625,
+      "learning_rate": 3.660523846088218e-05,
+      "loss": 0.5436,
+      "step": 3141
+    },
+    {
+      "epoch": 0.804352,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.660097261325826e-05,
+      "loss": 0.5173,
+      "step": 3142
+    },
+    {
+      "epoch": 0.804608,
+      "grad_norm": 2.578125,
+      "learning_rate": 3.659670676563433e-05,
+      "loss": 0.4887,
+      "step": 3143
+    },
+    {
+      "epoch": 0.804864,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.6592440918010414e-05,
+      "loss": 0.5351,
+      "step": 3144
+    },
+    {
+      "epoch": 0.80512,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.658817507038649e-05,
+      "loss": 0.5322,
+      "step": 3145
+    },
+    {
+      "epoch": 0.805376,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.658390922276256e-05,
+      "loss": 0.572,
+      "step": 3146
+    },
+    {
+      "epoch": 0.805632,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.6579643375138636e-05,
+      "loss": 0.5419,
+      "step": 3147
+    },
+    {
+      "epoch": 0.805888,
+      "grad_norm": 2.125,
+      "learning_rate": 3.657537752751472e-05,
+      "loss": 0.5155,
+      "step": 3148
+    },
+    {
+      "epoch": 0.806144,
+      "grad_norm": 2.609375,
+      "learning_rate": 3.65711116798908e-05,
+      "loss": 0.5285,
+      "step": 3149
+    },
+    {
+      "epoch": 0.8064,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.656684583226687e-05,
+      "loss": 0.5409,
+      "step": 3150
+    },
+    {
+      "epoch": 0.806656,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.6562579984642954e-05,
+      "loss": 0.5419,
+      "step": 3151
+    },
+    {
+      "epoch": 0.806912,
+      "grad_norm": 2.5,
+      "learning_rate": 3.655831413701903e-05,
+      "loss": 0.4725,
+      "step": 3152
+    },
+    {
+      "epoch": 0.807168,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.65540482893951e-05,
+      "loss": 0.4857,
+      "step": 3153
+    },
+    {
+      "epoch": 0.807424,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.6549782441771177e-05,
+      "loss": 0.5326,
+      "step": 3154
+    },
+    {
+      "epoch": 0.80768,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.6545516594147264e-05,
+      "loss": 0.5191,
+      "step": 3155
+    },
+    {
+      "epoch": 0.807936,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.654125074652334e-05,
+      "loss": 0.5251,
+      "step": 3156
+    },
+    {
+      "epoch": 0.808192,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.653698489889941e-05,
+      "loss": 0.5262,
+      "step": 3157
+    },
+    {
+      "epoch": 0.808448,
+      "grad_norm": 2.625,
+      "learning_rate": 3.653271905127549e-05,
+      "loss": 0.5574,
+      "step": 3158
+    },
+    {
+      "epoch": 0.808704,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.652845320365157e-05,
+      "loss": 0.5646,
+      "step": 3159
+    },
+    {
+      "epoch": 0.80896,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.652418735602764e-05,
+      "loss": 0.5281,
+      "step": 3160
+    },
+    {
+      "epoch": 0.809216,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.6519921508403723e-05,
+      "loss": 0.5052,
+      "step": 3161
+    },
+    {
+      "epoch": 0.809472,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.65156556607798e-05,
+      "loss": 0.549,
+      "step": 3162
+    },
+    {
+      "epoch": 0.809728,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.651138981315588e-05,
+      "loss": 0.5274,
+      "step": 3163
+    },
+    {
+      "epoch": 0.809984,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.650712396553195e-05,
+      "loss": 0.534,
+      "step": 3164
+    },
+    {
+      "epoch": 0.81024,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.650285811790803e-05,
+      "loss": 0.5166,
+      "step": 3165
+    },
+    {
+      "epoch": 0.810496,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.649859227028411e-05,
+      "loss": 0.5334,
+      "step": 3166
+    },
+    {
+      "epoch": 0.810752,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.649432642266018e-05,
+      "loss": 0.4908,
+      "step": 3167
+    },
+    {
+      "epoch": 0.811008,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.6490060575036264e-05,
+      "loss": 0.4974,
+      "step": 3168
+    },
+    {
+      "epoch": 0.811264,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.648579472741234e-05,
+      "loss": 0.556,
+      "step": 3169
+    },
+    {
+      "epoch": 0.81152,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.648152887978842e-05,
+      "loss": 0.5199,
+      "step": 3170
+    },
+    {
+      "epoch": 0.811776,
+      "grad_norm": 2.375,
+      "learning_rate": 3.647726303216449e-05,
+      "loss": 0.5424,
+      "step": 3171
+    },
+    {
+      "epoch": 0.812032,
+      "grad_norm": 2.375,
+      "learning_rate": 3.647299718454057e-05,
+      "loss": 0.5,
+      "step": 3172
+    },
+    {
+      "epoch": 0.812288,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.646873133691664e-05,
+      "loss": 0.5596,
+      "step": 3173
+    },
+    {
+      "epoch": 0.812544,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.646446548929273e-05,
+      "loss": 0.5542,
+      "step": 3174
+    },
+    {
+      "epoch": 0.8128,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.6460199641668804e-05,
+      "loss": 0.4932,
+      "step": 3175
+    },
+    {
+      "epoch": 0.813056,
+      "grad_norm": 2.625,
+      "learning_rate": 3.645593379404488e-05,
+      "loss": 0.4987,
+      "step": 3176
+    },
+    {
+      "epoch": 0.813312,
+      "grad_norm": 2.578125,
+      "learning_rate": 3.645166794642095e-05,
+      "loss": 0.5472,
+      "step": 3177
+    },
+    {
+      "epoch": 0.813568,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.6447402098797033e-05,
+      "loss": 0.519,
+      "step": 3178
+    },
+    {
+      "epoch": 0.813824,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.644313625117311e-05,
+      "loss": 0.5376,
+      "step": 3179
+    },
+    {
+      "epoch": 0.81408,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.643887040354919e-05,
+      "loss": 0.5448,
+      "step": 3180
+    },
+    {
+      "epoch": 0.814336,
+      "grad_norm": 2.59375,
+      "learning_rate": 3.643460455592526e-05,
+      "loss": 0.5016,
+      "step": 3181
+    },
+    {
+      "epoch": 0.814592,
+      "grad_norm": 2.734375,
+      "learning_rate": 3.6430338708301344e-05,
+      "loss": 0.5084,
+      "step": 3182
+    },
+    {
+      "epoch": 0.814848,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.642607286067742e-05,
+      "loss": 0.5076,
+      "step": 3183
+    },
+    {
+      "epoch": 0.815104,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.642180701305349e-05,
+      "loss": 0.5318,
+      "step": 3184
+    },
+    {
+      "epoch": 0.81536,
+      "grad_norm": 2.578125,
+      "learning_rate": 3.6417541165429574e-05,
+      "loss": 0.4926,
+      "step": 3185
+    },
+    {
+      "epoch": 0.815616,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.641327531780565e-05,
+      "loss": 0.5302,
+      "step": 3186
+    },
+    {
+      "epoch": 0.815872,
+      "grad_norm": 2.375,
+      "learning_rate": 3.640900947018173e-05,
+      "loss": 0.5269,
+      "step": 3187
+    },
+    {
+      "epoch": 0.816128,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.64047436225578e-05,
+      "loss": 0.521,
+      "step": 3188
+    },
+    {
+      "epoch": 0.816384,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.6400477774933884e-05,
+      "loss": 0.4853,
+      "step": 3189
+    },
+    {
+      "epoch": 0.81664,
+      "grad_norm": 2.5,
+      "learning_rate": 3.639621192730996e-05,
+      "loss": 0.5001,
+      "step": 3190
+    },
+    {
+      "epoch": 0.816896,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.639194607968603e-05,
+      "loss": 0.5155,
+      "step": 3191
+    },
+    {
+      "epoch": 0.817152,
+      "grad_norm": 2.5625,
+      "learning_rate": 3.638768023206211e-05,
+      "loss": 0.4789,
+      "step": 3192
+    },
+    {
+      "epoch": 0.817408,
+      "grad_norm": 2.609375,
+      "learning_rate": 3.638341438443819e-05,
+      "loss": 0.4888,
+      "step": 3193
+    },
+    {
+      "epoch": 0.817664,
+      "grad_norm": 2.5,
+      "learning_rate": 3.637914853681427e-05,
+      "loss": 0.508,
+      "step": 3194
+    },
+    {
+      "epoch": 0.81792,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.6374882689190343e-05,
+      "loss": 0.512,
+      "step": 3195
+    },
+    {
+      "epoch": 0.818176,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.637061684156642e-05,
+      "loss": 0.5128,
+      "step": 3196
+    },
+    {
+      "epoch": 0.818432,
+      "grad_norm": 2.765625,
+      "learning_rate": 3.63663509939425e-05,
+      "loss": 0.5033,
+      "step": 3197
+    },
+    {
+      "epoch": 0.818688,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.636208514631857e-05,
+      "loss": 0.5383,
+      "step": 3198
+    },
+    {
+      "epoch": 0.818944,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.6357819298694654e-05,
+      "loss": 0.4588,
+      "step": 3199
+    },
+    {
+      "epoch": 0.8192,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.6353553451070735e-05,
+      "loss": 0.5546,
+      "step": 3200
+    },
+    {
+      "epoch": 0.819456,
+      "grad_norm": 2.25,
+      "learning_rate": 3.634928760344681e-05,
+      "loss": 0.5386,
+      "step": 3201
+    },
+    {
+      "epoch": 0.819712,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.6345021755822884e-05,
+      "loss": 0.5208,
+      "step": 3202
+    },
+    {
+      "epoch": 0.819968,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.634075590819896e-05,
+      "loss": 0.4903,
+      "step": 3203
+    },
+    {
+      "epoch": 0.820224,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.633649006057504e-05,
+      "loss": 0.5255,
+      "step": 3204
+    },
+    {
+      "epoch": 0.82048,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.633222421295111e-05,
+      "loss": 0.5464,
+      "step": 3205
+    },
+    {
+      "epoch": 0.820736,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.6327958365327194e-05,
+      "loss": 0.5459,
+      "step": 3206
+    },
+    {
+      "epoch": 0.820992,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.632369251770327e-05,
+      "loss": 0.498,
+      "step": 3207
+    },
+    {
+      "epoch": 0.821248,
+      "grad_norm": 2.25,
+      "learning_rate": 3.631942667007935e-05,
+      "loss": 0.5221,
+      "step": 3208
+    },
+    {
+      "epoch": 0.821504,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.6315160822455424e-05,
+      "loss": 0.5172,
+      "step": 3209
+    },
+    {
+      "epoch": 0.82176,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.63108949748315e-05,
+      "loss": 0.5504,
+      "step": 3210
+    },
+    {
+      "epoch": 0.822016,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.630662912720757e-05,
+      "loss": 0.5149,
+      "step": 3211
+    },
+    {
+      "epoch": 0.822272,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.630236327958365e-05,
+      "loss": 0.5661,
+      "step": 3212
+    },
+    {
+      "epoch": 0.822528,
+      "grad_norm": 2.25,
+      "learning_rate": 3.6298097431959734e-05,
+      "loss": 0.502,
+      "step": 3213
+    },
+    {
+      "epoch": 0.822784,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.629383158433581e-05,
+      "loss": 0.5271,
+      "step": 3214
+    },
+    {
+      "epoch": 0.82304,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.628956573671189e-05,
+      "loss": 0.5333,
+      "step": 3215
+    },
+    {
+      "epoch": 0.823296,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.6285299889087964e-05,
+      "loss": 0.5079,
+      "step": 3216
+    },
+    {
+      "epoch": 0.823552,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.628103404146404e-05,
+      "loss": 0.5382,
+      "step": 3217
+    },
+    {
+      "epoch": 0.823808,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.627676819384012e-05,
+      "loss": 0.5225,
+      "step": 3218
+    },
+    {
+      "epoch": 0.824064,
+      "grad_norm": 2.578125,
+      "learning_rate": 3.62725023462162e-05,
+      "loss": 0.5213,
+      "step": 3219
+    },
+    {
+      "epoch": 0.82432,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.6268236498592275e-05,
+      "loss": 0.5091,
+      "step": 3220
+    },
+    {
+      "epoch": 0.824576,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.626397065096835e-05,
+      "loss": 0.5115,
+      "step": 3221
+    },
+    {
+      "epoch": 0.824832,
+      "grad_norm": 2.5,
+      "learning_rate": 3.625970480334442e-05,
+      "loss": 0.5762,
+      "step": 3222
+    },
+    {
+      "epoch": 0.825088,
+      "grad_norm": 2.5625,
+      "learning_rate": 3.6255438955720504e-05,
+      "loss": 0.5312,
+      "step": 3223
+    },
+    {
+      "epoch": 0.825344,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.625117310809658e-05,
+      "loss": 0.5396,
+      "step": 3224
+    },
+    {
+      "epoch": 0.8256,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.624690726047266e-05,
+      "loss": 0.5059,
+      "step": 3225
+    },
+    {
+      "epoch": 0.825856,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.6242641412848734e-05,
+      "loss": 0.5365,
+      "step": 3226
+    },
+    {
+      "epoch": 0.826112,
+      "grad_norm": 2.609375,
+      "learning_rate": 3.6238375565224815e-05,
+      "loss": 0.5622,
+      "step": 3227
+    },
+    {
+      "epoch": 0.826368,
+      "grad_norm": 2.578125,
+      "learning_rate": 3.623410971760089e-05,
+      "loss": 0.4901,
+      "step": 3228
+    },
+    {
+      "epoch": 0.826624,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.622984386997696e-05,
+      "loss": 0.5287,
+      "step": 3229
+    },
+    {
+      "epoch": 0.82688,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.622557802235304e-05,
+      "loss": 0.5562,
+      "step": 3230
+    },
+    {
+      "epoch": 0.827136,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.622131217472912e-05,
+      "loss": 0.5282,
+      "step": 3231
+    },
+    {
+      "epoch": 0.827392,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.62170463271052e-05,
+      "loss": 0.5203,
+      "step": 3232
+    },
+    {
+      "epoch": 0.827648,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.6212780479481274e-05,
+      "loss": 0.5188,
+      "step": 3233
+    },
+    {
+      "epoch": 0.827904,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.6208514631857355e-05,
+      "loss": 0.5045,
+      "step": 3234
+    },
+    {
+      "epoch": 0.82816,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.620424878423343e-05,
+      "loss": 0.4993,
+      "step": 3235
+    },
+    {
+      "epoch": 0.828416,
+      "grad_norm": 2.59375,
+      "learning_rate": 3.6199982936609504e-05,
+      "loss": 0.5,
+      "step": 3236
+    },
+    {
+      "epoch": 0.828672,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.6195717088985585e-05,
+      "loss": 0.5153,
+      "step": 3237
+    },
+    {
+      "epoch": 0.828928,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.6191451241361666e-05,
+      "loss": 0.5303,
+      "step": 3238
+    },
+    {
+      "epoch": 0.829184,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.618718539373774e-05,
+      "loss": 0.4991,
+      "step": 3239
+    },
+    {
+      "epoch": 0.82944,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.6182919546113814e-05,
+      "loss": 0.5162,
+      "step": 3240
+    },
+    {
+      "epoch": 0.829696,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.617865369848989e-05,
+      "loss": 0.5003,
+      "step": 3241
+    },
+    {
+      "epoch": 0.829952,
+      "grad_norm": 2.609375,
+      "learning_rate": 3.617438785086597e-05,
+      "loss": 0.5033,
+      "step": 3242
+    },
+    {
+      "epoch": 0.830208,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.6170122003242044e-05,
+      "loss": 0.514,
+      "step": 3243
+    },
+    {
+      "epoch": 0.830464,
+      "grad_norm": 2.5,
+      "learning_rate": 3.6165856155618125e-05,
+      "loss": 0.519,
+      "step": 3244
+    },
+    {
+      "epoch": 0.83072,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.61615903079942e-05,
+      "loss": 0.5061,
+      "step": 3245
+    },
+    {
+      "epoch": 0.830976,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.615732446037028e-05,
+      "loss": 0.4705,
+      "step": 3246
+    },
+    {
+      "epoch": 0.831232,
+      "grad_norm": 2.5,
+      "learning_rate": 3.6153058612746354e-05,
+      "loss": 0.5185,
+      "step": 3247
+    },
+    {
+      "epoch": 0.831488,
+      "grad_norm": 2.375,
+      "learning_rate": 3.614879276512243e-05,
+      "loss": 0.5235,
+      "step": 3248
+    },
+    {
+      "epoch": 0.831744,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.614452691749851e-05,
+      "loss": 0.4797,
+      "step": 3249
+    },
+    {
+      "epoch": 0.832,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.6140261069874584e-05,
+      "loss": 0.4912,
+      "step": 3250
+    },
+    {
+      "epoch": 0.832256,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.6135995222250665e-05,
+      "loss": 0.5157,
+      "step": 3251
+    },
+    {
+      "epoch": 0.832512,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.613172937462674e-05,
+      "loss": 0.4849,
+      "step": 3252
+    },
+    {
+      "epoch": 0.832768,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.612746352700282e-05,
+      "loss": 0.5299,
+      "step": 3253
+    },
+    {
+      "epoch": 0.833024,
+      "grad_norm": 2.609375,
+      "learning_rate": 3.6123197679378895e-05,
+      "loss": 0.5633,
+      "step": 3254
+    },
+    {
+      "epoch": 0.83328,
+      "grad_norm": 2.734375,
+      "learning_rate": 3.611893183175497e-05,
+      "loss": 0.5237,
+      "step": 3255
+    },
+    {
+      "epoch": 0.833536,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.611466598413104e-05,
+      "loss": 0.5038,
+      "step": 3256
+    },
+    {
+      "epoch": 0.833792,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.611040013650713e-05,
+      "loss": 0.4861,
+      "step": 3257
+    },
+    {
+      "epoch": 0.834048,
+      "grad_norm": 2.5625,
+      "learning_rate": 3.6106134288883205e-05,
+      "loss": 0.5098,
+      "step": 3258
+    },
+    {
+      "epoch": 0.834304,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.610186844125928e-05,
+      "loss": 0.5353,
+      "step": 3259
+    },
+    {
+      "epoch": 0.83456,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.6097602593635354e-05,
+      "loss": 0.5144,
+      "step": 3260
+    },
+    {
+      "epoch": 0.834816,
+      "grad_norm": 2.5,
+      "learning_rate": 3.6093336746011435e-05,
+      "loss": 0.4892,
+      "step": 3261
+    },
+    {
+      "epoch": 0.835072,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.608907089838751e-05,
+      "loss": 0.4935,
+      "step": 3262
+    },
+    {
+      "epoch": 0.835328,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.608480505076359e-05,
+      "loss": 0.5022,
+      "step": 3263
+    },
+    {
+      "epoch": 0.835584,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.608053920313967e-05,
+      "loss": 0.5367,
+      "step": 3264
+    },
+    {
+      "epoch": 0.83584,
+      "grad_norm": 2.5,
+      "learning_rate": 3.6076273355515745e-05,
+      "loss": 0.5409,
+      "step": 3265
+    },
+    {
+      "epoch": 0.836096,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.607200750789182e-05,
+      "loss": 0.49,
+      "step": 3266
+    },
+    {
+      "epoch": 0.836352,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.6067741660267894e-05,
+      "loss": 0.4932,
+      "step": 3267
+    },
+    {
+      "epoch": 0.836608,
+      "grad_norm": 2.25,
+      "learning_rate": 3.6063475812643975e-05,
+      "loss": 0.5147,
+      "step": 3268
+    },
+    {
+      "epoch": 0.836864,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.605920996502005e-05,
+      "loss": 0.5222,
+      "step": 3269
+    },
+    {
+      "epoch": 0.83712,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.605494411739613e-05,
+      "loss": 0.4957,
+      "step": 3270
+    },
+    {
+      "epoch": 0.837376,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.6050678269772204e-05,
+      "loss": 0.5026,
+      "step": 3271
+    },
+    {
+      "epoch": 0.837632,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.6046412422148286e-05,
+      "loss": 0.5184,
+      "step": 3272
+    },
+    {
+      "epoch": 0.837888,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.604214657452436e-05,
+      "loss": 0.5764,
+      "step": 3273
+    },
+    {
+      "epoch": 0.838144,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.6037880726900434e-05,
+      "loss": 0.5068,
+      "step": 3274
+    },
+    {
+      "epoch": 0.8384,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.603361487927651e-05,
+      "loss": 0.5143,
+      "step": 3275
+    },
+    {
+      "epoch": 0.838656,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.6029349031652596e-05,
+      "loss": 0.527,
+      "step": 3276
+    },
+    {
+      "epoch": 0.838912,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.602508318402867e-05,
+      "loss": 0.5135,
+      "step": 3277
+    },
+    {
+      "epoch": 0.839168,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.6020817336404745e-05,
+      "loss": 0.4837,
+      "step": 3278
+    },
+    {
+      "epoch": 0.839424,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.601655148878082e-05,
+      "loss": 0.5135,
+      "step": 3279
+    },
+    {
+      "epoch": 0.83968,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.60122856411569e-05,
+      "loss": 0.5367,
+      "step": 3280
+    },
+    {
+      "epoch": 0.839936,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.6008019793532974e-05,
+      "loss": 0.4927,
+      "step": 3281
+    },
+    {
+      "epoch": 0.840192,
+      "grad_norm": 2.375,
+      "learning_rate": 3.6003753945909055e-05,
+      "loss": 0.5461,
+      "step": 3282
+    },
+    {
+      "epoch": 0.840448,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.5999488098285136e-05,
+      "loss": 0.4898,
+      "step": 3283
+    },
+    {
+      "epoch": 0.840704,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.599522225066121e-05,
+      "loss": 0.5079,
+      "step": 3284
+    },
+    {
+      "epoch": 0.84096,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.5990956403037285e-05,
+      "loss": 0.5093,
+      "step": 3285
+    },
+    {
+      "epoch": 0.841216,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.598669055541336e-05,
+      "loss": 0.5102,
+      "step": 3286
+    },
+    {
+      "epoch": 0.841472,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.598242470778944e-05,
+      "loss": 0.5092,
+      "step": 3287
+    },
+    {
+      "epoch": 0.841728,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.5978158860165514e-05,
+      "loss": 0.5381,
+      "step": 3288
+    },
+    {
+      "epoch": 0.841984,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.5973893012541595e-05,
+      "loss": 0.5142,
+      "step": 3289
+    },
+    {
+      "epoch": 0.84224,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.596962716491767e-05,
+      "loss": 0.523,
+      "step": 3290
+    },
+    {
+      "epoch": 0.842496,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.596536131729375e-05,
+      "loss": 0.4927,
+      "step": 3291
+    },
+    {
+      "epoch": 0.842752,
+      "grad_norm": 2.671875,
+      "learning_rate": 3.5961095469669825e-05,
+      "loss": 0.5102,
+      "step": 3292
+    },
+    {
+      "epoch": 0.843008,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.59568296220459e-05,
+      "loss": 0.4875,
+      "step": 3293
+    },
+    {
+      "epoch": 0.843264,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.5952563774421974e-05,
+      "loss": 0.5278,
+      "step": 3294
+    },
+    {
+      "epoch": 0.84352,
+      "grad_norm": 2.25,
+      "learning_rate": 3.5948297926798055e-05,
+      "loss": 0.5241,
+      "step": 3295
+    },
+    {
+      "epoch": 0.843776,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.5944032079174136e-05,
+      "loss": 0.5122,
+      "step": 3296
+    },
+    {
+      "epoch": 0.844032,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.593976623155021e-05,
+      "loss": 0.5034,
+      "step": 3297
+    },
+    {
+      "epoch": 0.844288,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.593550038392629e-05,
+      "loss": 0.5121,
+      "step": 3298
+    },
+    {
+      "epoch": 0.844544,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.5931234536302365e-05,
+      "loss": 0.5445,
+      "step": 3299
+    },
+    {
+      "epoch": 0.8448,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.592696868867844e-05,
+      "loss": 0.5264,
+      "step": 3300
+    },
+    {
+      "epoch": 0.845056,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.592270284105452e-05,
+      "loss": 0.5317,
+      "step": 3301
+    },
+    {
+      "epoch": 0.845312,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.59184369934306e-05,
+      "loss": 0.5318,
+      "step": 3302
+    },
+    {
+      "epoch": 0.845568,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.5914171145806676e-05,
+      "loss": 0.5466,
+      "step": 3303
+    },
+    {
+      "epoch": 0.845824,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.590990529818275e-05,
+      "loss": 0.5303,
+      "step": 3304
+    },
+    {
+      "epoch": 0.84608,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.5905639450558824e-05,
+      "loss": 0.4983,
+      "step": 3305
+    },
+    {
+      "epoch": 0.846336,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.5901373602934905e-05,
+      "loss": 0.4969,
+      "step": 3306
+    },
+    {
+      "epoch": 0.846592,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.589710775531098e-05,
+      "loss": 0.47,
+      "step": 3307
+    },
+    {
+      "epoch": 0.846848,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.589284190768706e-05,
+      "loss": 0.5243,
+      "step": 3308
+    },
+    {
+      "epoch": 0.847104,
+      "grad_norm": 2.5,
+      "learning_rate": 3.5888576060063135e-05,
+      "loss": 0.5509,
+      "step": 3309
+    },
+    {
+      "epoch": 0.84736,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.5884310212439216e-05,
+      "loss": 0.5238,
+      "step": 3310
+    },
+    {
+      "epoch": 0.847616,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.588004436481529e-05,
+      "loss": 0.5024,
+      "step": 3311
+    },
+    {
+      "epoch": 0.847872,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.5875778517191365e-05,
+      "loss": 0.502,
+      "step": 3312
+    },
+    {
+      "epoch": 0.848128,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.5871512669567446e-05,
+      "loss": 0.4912,
+      "step": 3313
+    },
+    {
+      "epoch": 0.848384,
+      "grad_norm": 2.5,
+      "learning_rate": 3.586724682194352e-05,
+      "loss": 0.4914,
+      "step": 3314
+    },
+    {
+      "epoch": 0.84864,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.58629809743196e-05,
+      "loss": 0.5037,
+      "step": 3315
+    },
+    {
+      "epoch": 0.848896,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.5858715126695675e-05,
+      "loss": 0.4905,
+      "step": 3316
+    },
+    {
+      "epoch": 0.849152,
+      "grad_norm": 2.65625,
+      "learning_rate": 3.5854449279071756e-05,
+      "loss": 0.5128,
+      "step": 3317
+    },
+    {
+      "epoch": 0.849408,
+      "grad_norm": 2.609375,
+      "learning_rate": 3.585018343144783e-05,
+      "loss": 0.4801,
+      "step": 3318
+    },
+    {
+      "epoch": 0.849664,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.5845917583823905e-05,
+      "loss": 0.5155,
+      "step": 3319
+    },
+    {
+      "epoch": 0.84992,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.5841651736199986e-05,
+      "loss": 0.4985,
+      "step": 3320
+    },
+    {
+      "epoch": 0.850176,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.583738588857607e-05,
+      "loss": 0.5153,
+      "step": 3321
+    },
+    {
+      "epoch": 0.850432,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.583312004095214e-05,
+      "loss": 0.5437,
+      "step": 3322
+    },
+    {
+      "epoch": 0.850688,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.5828854193328215e-05,
+      "loss": 0.5327,
+      "step": 3323
+    },
+    {
+      "epoch": 0.850944,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.582458834570429e-05,
+      "loss": 0.4993,
+      "step": 3324
+    },
+    {
+      "epoch": 0.8512,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.582032249808037e-05,
+      "loss": 0.556,
+      "step": 3325
+    },
+    {
+      "epoch": 0.851456,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.5816056650456445e-05,
+      "loss": 0.5039,
+      "step": 3326
+    },
+    {
+      "epoch": 0.851712,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.5811790802832526e-05,
+      "loss": 0.5073,
+      "step": 3327
+    },
+    {
+      "epoch": 0.851968,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.58075249552086e-05,
+      "loss": 0.5459,
+      "step": 3328
+    },
+    {
+      "epoch": 0.852224,
+      "grad_norm": 2.578125,
+      "learning_rate": 3.580325910758468e-05,
+      "loss": 0.5352,
+      "step": 3329
+    },
+    {
+      "epoch": 0.85248,
+      "grad_norm": 2.25,
+      "learning_rate": 3.5798993259960756e-05,
+      "loss": 0.5384,
+      "step": 3330
+    },
+    {
+      "epoch": 0.852736,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.579472741233683e-05,
+      "loss": 0.4982,
+      "step": 3331
+    },
+    {
+      "epoch": 0.852992,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.579046156471291e-05,
+      "loss": 0.5063,
+      "step": 3332
+    },
+    {
+      "epoch": 0.853248,
+      "grad_norm": 2.25,
+      "learning_rate": 3.5786195717088985e-05,
+      "loss": 0.5499,
+      "step": 3333
+    },
+    {
+      "epoch": 0.853504,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.5781929869465066e-05,
+      "loss": 0.5559,
+      "step": 3334
+    },
+    {
+      "epoch": 0.85376,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.577766402184114e-05,
+      "loss": 0.5355,
+      "step": 3335
+    },
+    {
+      "epoch": 0.854016,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.577339817421722e-05,
+      "loss": 0.5167,
+      "step": 3336
+    },
+    {
+      "epoch": 0.854272,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.5769132326593296e-05,
+      "loss": 0.5239,
+      "step": 3337
+    },
+    {
+      "epoch": 0.854528,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.576486647896937e-05,
+      "loss": 0.5052,
+      "step": 3338
+    },
+    {
+      "epoch": 0.854784,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.576060063134545e-05,
+      "loss": 0.4815,
+      "step": 3339
+    },
+    {
+      "epoch": 0.85504,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.575633478372153e-05,
+      "loss": 0.5443,
+      "step": 3340
+    },
+    {
+      "epoch": 0.855296,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.5752068936097606e-05,
+      "loss": 0.5319,
+      "step": 3341
+    },
+    {
+      "epoch": 0.855552,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.574780308847368e-05,
+      "loss": 0.5075,
+      "step": 3342
+    },
+    {
+      "epoch": 0.855808,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.5743537240849755e-05,
+      "loss": 0.4833,
+      "step": 3343
+    },
+    {
+      "epoch": 0.856064,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.5739271393225836e-05,
+      "loss": 0.5142,
+      "step": 3344
+    },
+    {
+      "epoch": 0.85632,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.573500554560191e-05,
+      "loss": 0.5145,
+      "step": 3345
+    },
+    {
+      "epoch": 0.856576,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.573073969797799e-05,
+      "loss": 0.4633,
+      "step": 3346
+    },
+    {
+      "epoch": 0.856832,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.572647385035407e-05,
+      "loss": 0.4943,
+      "step": 3347
+    },
+    {
+      "epoch": 0.857088,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.5722208002730147e-05,
+      "loss": 0.5307,
+      "step": 3348
+    },
+    {
+      "epoch": 0.857344,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.571794215510622e-05,
+      "loss": 0.5196,
+      "step": 3349
+    },
+    {
+      "epoch": 0.8576,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.5713676307482295e-05,
+      "loss": 0.5142,
+      "step": 3350
+    },
+    {
+      "epoch": 0.857856,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.5709410459858376e-05,
+      "loss": 0.5348,
+      "step": 3351
+    },
+    {
+      "epoch": 0.858112,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.570514461223445e-05,
+      "loss": 0.565,
+      "step": 3352
+    },
+    {
+      "epoch": 0.858368,
+      "grad_norm": 2.25,
+      "learning_rate": 3.570087876461053e-05,
+      "loss": 0.5038,
+      "step": 3353
+    },
+    {
+      "epoch": 0.858624,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.5696612916986606e-05,
+      "loss": 0.514,
+      "step": 3354
+    },
+    {
+      "epoch": 0.85888,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.569234706936269e-05,
+      "loss": 0.5199,
+      "step": 3355
+    },
+    {
+      "epoch": 0.859136,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.568808122173876e-05,
+      "loss": 0.5344,
+      "step": 3356
+    },
+    {
+      "epoch": 0.859392,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.5683815374114835e-05,
+      "loss": 0.5108,
+      "step": 3357
+    },
+    {
+      "epoch": 0.859648,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.5679549526490916e-05,
+      "loss": 0.4906,
+      "step": 3358
+    },
+    {
+      "epoch": 0.859904,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.5675283678867e-05,
+      "loss": 0.5166,
+      "step": 3359
+    },
+    {
+      "epoch": 0.86016,
+      "grad_norm": 2.5,
+      "learning_rate": 3.567101783124307e-05,
+      "loss": 0.5089,
+      "step": 3360
+    },
+    {
+      "epoch": 0.860416,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.5666751983619146e-05,
+      "loss": 0.513,
+      "step": 3361
+    },
+    {
+      "epoch": 0.860672,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.566248613599523e-05,
+      "loss": 0.5296,
+      "step": 3362
+    },
+    {
+      "epoch": 0.860928,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.56582202883713e-05,
+      "loss": 0.4885,
+      "step": 3363
+    },
+    {
+      "epoch": 0.861184,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.5653954440747376e-05,
+      "loss": 0.4651,
+      "step": 3364
+    },
+    {
+      "epoch": 0.86144,
+      "grad_norm": 2.59375,
+      "learning_rate": 3.5649688593123457e-05,
+      "loss": 0.5342,
+      "step": 3365
+    },
+    {
+      "epoch": 0.861696,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.564542274549954e-05,
+      "loss": 0.5042,
+      "step": 3366
+    },
+    {
+      "epoch": 0.861952,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.564115689787561e-05,
+      "loss": 0.5077,
+      "step": 3367
+    },
+    {
+      "epoch": 0.862208,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.5636891050251686e-05,
+      "loss": 0.5179,
+      "step": 3368
+    },
+    {
+      "epoch": 0.862464,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.563262520262776e-05,
+      "loss": 0.487,
+      "step": 3369
+    },
+    {
+      "epoch": 0.86272,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.562835935500384e-05,
+      "loss": 0.5098,
+      "step": 3370
+    },
+    {
+      "epoch": 0.862976,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.5624093507379916e-05,
+      "loss": 0.4889,
+      "step": 3371
+    },
+    {
+      "epoch": 0.863232,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.5619827659756e-05,
+      "loss": 0.5185,
+      "step": 3372
+    },
+    {
+      "epoch": 0.863488,
+      "grad_norm": 2.5,
+      "learning_rate": 3.561556181213207e-05,
+      "loss": 0.4828,
+      "step": 3373
+    },
+    {
+      "epoch": 0.863744,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.561129596450815e-05,
+      "loss": 0.5199,
+      "step": 3374
+    },
+    {
+      "epoch": 0.864,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.5607030116884226e-05,
+      "loss": 0.5039,
+      "step": 3375
+    },
+    {
+      "epoch": 0.864256,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.56027642692603e-05,
+      "loss": 0.5116,
+      "step": 3376
+    },
+    {
+      "epoch": 0.864512,
+      "grad_norm": 2.578125,
+      "learning_rate": 3.5598498421636375e-05,
+      "loss": 0.5025,
+      "step": 3377
+    },
+    {
+      "epoch": 0.864768,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.559423257401246e-05,
+      "loss": 0.5112,
+      "step": 3378
+    },
+    {
+      "epoch": 0.865024,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.558996672638854e-05,
+      "loss": 0.4943,
+      "step": 3379
+    },
+    {
+      "epoch": 0.86528,
+      "grad_norm": 2.5,
+      "learning_rate": 3.558570087876461e-05,
+      "loss": 0.5472,
+      "step": 3380
+    },
+    {
+      "epoch": 0.865536,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.558143503114069e-05,
+      "loss": 0.5125,
+      "step": 3381
+    },
+    {
+      "epoch": 0.865792,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.5577169183516767e-05,
+      "loss": 0.4918,
+      "step": 3382
+    },
+    {
+      "epoch": 0.866048,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.557290333589284e-05,
+      "loss": 0.5152,
+      "step": 3383
+    },
+    {
+      "epoch": 0.866304,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.556863748826892e-05,
+      "loss": 0.5087,
+      "step": 3384
+    },
+    {
+      "epoch": 0.86656,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.5564371640645e-05,
+      "loss": 0.5232,
+      "step": 3385
+    },
+    {
+      "epoch": 0.866816,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.556010579302108e-05,
+      "loss": 0.5326,
+      "step": 3386
+    },
+    {
+      "epoch": 0.867072,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.555583994539715e-05,
+      "loss": 0.478,
+      "step": 3387
+    },
+    {
+      "epoch": 0.867328,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.5551574097773226e-05,
+      "loss": 0.5488,
+      "step": 3388
+    },
+    {
+      "epoch": 0.867584,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.554730825014931e-05,
+      "loss": 0.5024,
+      "step": 3389
+    },
+    {
+      "epoch": 0.86784,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.554304240252538e-05,
+      "loss": 0.5054,
+      "step": 3390
+    },
+    {
+      "epoch": 0.868096,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.553877655490146e-05,
+      "loss": 0.4764,
+      "step": 3391
+    },
+    {
+      "epoch": 0.868352,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.5534510707277536e-05,
+      "loss": 0.493,
+      "step": 3392
+    },
+    {
+      "epoch": 0.868608,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.553024485965362e-05,
+      "loss": 0.4957,
+      "step": 3393
+    },
+    {
+      "epoch": 0.868864,
+      "grad_norm": 2.625,
+      "learning_rate": 3.552597901202969e-05,
+      "loss": 0.5027,
+      "step": 3394
+    },
+    {
+      "epoch": 0.86912,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.5521713164405766e-05,
+      "loss": 0.5012,
+      "step": 3395
+    },
+    {
+      "epoch": 0.869376,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.551744731678185e-05,
+      "loss": 0.5347,
+      "step": 3396
+    },
+    {
+      "epoch": 0.869632,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.551318146915792e-05,
+      "loss": 0.5038,
+      "step": 3397
+    },
+    {
+      "epoch": 0.869888,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.5508915621534e-05,
+      "loss": 0.4912,
+      "step": 3398
+    },
+    {
+      "epoch": 0.870144,
+      "grad_norm": 2.25,
+      "learning_rate": 3.5504649773910076e-05,
+      "loss": 0.5134,
+      "step": 3399
+    },
+    {
+      "epoch": 0.8704,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.550038392628616e-05,
+      "loss": 0.4929,
+      "step": 3400
+    },
+    {
+      "epoch": 0.870656,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.549611807866223e-05,
+      "loss": 0.4853,
+      "step": 3401
+    },
+    {
+      "epoch": 0.870912,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.5491852231038306e-05,
+      "loss": 0.5235,
+      "step": 3402
+    },
+    {
+      "epoch": 0.871168,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.548758638341439e-05,
+      "loss": 0.509,
+      "step": 3403
+    },
+    {
+      "epoch": 0.871424,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.548332053579047e-05,
+      "loss": 0.5,
+      "step": 3404
+    },
+    {
+      "epoch": 0.87168,
+      "grad_norm": 2.625,
+      "learning_rate": 3.547905468816654e-05,
+      "loss": 0.5116,
+      "step": 3405
+    },
+    {
+      "epoch": 0.871936,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.547478884054262e-05,
+      "loss": 0.4935,
+      "step": 3406
+    },
+    {
+      "epoch": 0.872192,
+      "grad_norm": 2.375,
+      "learning_rate": 3.547052299291869e-05,
+      "loss": 0.493,
+      "step": 3407
+    },
+    {
+      "epoch": 0.872448,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.546625714529477e-05,
+      "loss": 0.5176,
+      "step": 3408
+    },
+    {
+      "epoch": 0.872704,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.5461991297670846e-05,
+      "loss": 0.5174,
+      "step": 3409
+    },
+    {
+      "epoch": 0.87296,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.545772545004693e-05,
+      "loss": 0.5196,
+      "step": 3410
+    },
+    {
+      "epoch": 0.873216,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.545345960242301e-05,
+      "loss": 0.461,
+      "step": 3411
+    },
+    {
+      "epoch": 0.873472,
+      "grad_norm": 2.25,
+      "learning_rate": 3.544919375479908e-05,
+      "loss": 0.5081,
+      "step": 3412
+    },
+    {
+      "epoch": 0.873728,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.544492790717516e-05,
+      "loss": 0.5064,
+      "step": 3413
+    },
+    {
+      "epoch": 0.873984,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.544066205955123e-05,
+      "loss": 0.5103,
+      "step": 3414
+    },
+    {
+      "epoch": 0.87424,
+      "grad_norm": 2.125,
+      "learning_rate": 3.543639621192731e-05,
+      "loss": 0.5048,
+      "step": 3415
+    },
+    {
+      "epoch": 0.874496,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.5432130364303386e-05,
+      "loss": 0.505,
+      "step": 3416
+    },
+    {
+      "epoch": 0.874752,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.542786451667947e-05,
+      "loss": 0.4976,
+      "step": 3417
+    },
+    {
+      "epoch": 0.875008,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.542359866905554e-05,
+      "loss": 0.5096,
+      "step": 3418
+    },
+    {
+      "epoch": 0.875264,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.541933282143162e-05,
+      "loss": 0.5211,
+      "step": 3419
+    },
+    {
+      "epoch": 0.87552,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.54150669738077e-05,
+      "loss": 0.5245,
+      "step": 3420
+    },
+    {
+      "epoch": 0.875776,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.541080112618377e-05,
+      "loss": 0.5057,
+      "step": 3421
+    },
+    {
+      "epoch": 0.876032,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.540653527855985e-05,
+      "loss": 0.5401,
+      "step": 3422
+    },
+    {
+      "epoch": 0.876288,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.5402269430935933e-05,
+      "loss": 0.5147,
+      "step": 3423
+    },
+    {
+      "epoch": 0.876544,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.539800358331201e-05,
+      "loss": 0.5517,
+      "step": 3424
+    },
+    {
+      "epoch": 0.8768,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.539373773568808e-05,
+      "loss": 0.5105,
+      "step": 3425
+    },
+    {
+      "epoch": 0.877056,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.5389471888064156e-05,
+      "loss": 0.5538,
+      "step": 3426
+    },
+    {
+      "epoch": 0.877312,
+      "grad_norm": 2.25,
+      "learning_rate": 3.538520604044024e-05,
+      "loss": 0.5351,
+      "step": 3427
+    },
+    {
+      "epoch": 0.877568,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.538094019281631e-05,
+      "loss": 0.4895,
+      "step": 3428
+    },
+    {
+      "epoch": 0.877824,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.537667434519239e-05,
+      "loss": 0.488,
+      "step": 3429
+    },
+    {
+      "epoch": 0.87808,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.5372408497568474e-05,
+      "loss": 0.5034,
+      "step": 3430
+    },
+    {
+      "epoch": 0.878336,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.536814264994455e-05,
+      "loss": 0.4956,
+      "step": 3431
+    },
+    {
+      "epoch": 0.878592,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.536387680232062e-05,
+      "loss": 0.5163,
+      "step": 3432
+    },
+    {
+      "epoch": 0.878848,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.5359610954696696e-05,
+      "loss": 0.5002,
+      "step": 3433
+    },
+    {
+      "epoch": 0.879104,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.535534510707278e-05,
+      "loss": 0.4556,
+      "step": 3434
+    },
+    {
+      "epoch": 0.87936,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.535107925944885e-05,
+      "loss": 0.4779,
+      "step": 3435
+    },
+    {
+      "epoch": 0.879616,
+      "grad_norm": 2.5625,
+      "learning_rate": 3.534681341182493e-05,
+      "loss": 0.5304,
+      "step": 3436
+    },
+    {
+      "epoch": 0.879872,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.534254756420101e-05,
+      "loss": 0.5131,
+      "step": 3437
+    },
+    {
+      "epoch": 0.880128,
+      "grad_norm": 2.125,
+      "learning_rate": 3.533828171657709e-05,
+      "loss": 0.5224,
+      "step": 3438
+    },
+    {
+      "epoch": 0.880384,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.533401586895316e-05,
+      "loss": 0.5187,
+      "step": 3439
+    },
+    {
+      "epoch": 0.88064,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.5329750021329237e-05,
+      "loss": 0.4939,
+      "step": 3440
+    },
+    {
+      "epoch": 0.880896,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.532548417370532e-05,
+      "loss": 0.4712,
+      "step": 3441
+    },
+    {
+      "epoch": 0.881152,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.53212183260814e-05,
+      "loss": 0.5077,
+      "step": 3442
+    },
+    {
+      "epoch": 0.881408,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.531695247845747e-05,
+      "loss": 0.5006,
+      "step": 3443
+    },
+    {
+      "epoch": 0.881664,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.531268663083355e-05,
+      "loss": 0.4813,
+      "step": 3444
+    },
+    {
+      "epoch": 0.88192,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.530842078320963e-05,
+      "loss": 0.5313,
+      "step": 3445
+    },
+    {
+      "epoch": 0.882176,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.53041549355857e-05,
+      "loss": 0.5049,
+      "step": 3446
+    },
+    {
+      "epoch": 0.882432,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.529988908796178e-05,
+      "loss": 0.5168,
+      "step": 3447
+    },
+    {
+      "epoch": 0.882688,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.529562324033786e-05,
+      "loss": 0.453,
+      "step": 3448
+    },
+    {
+      "epoch": 0.882944,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.529135739271394e-05,
+      "loss": 0.5364,
+      "step": 3449
+    },
+    {
+      "epoch": 0.8832,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.528709154509001e-05,
+      "loss": 0.5163,
+      "step": 3450
+    },
+    {
+      "epoch": 0.883456,
+      "grad_norm": 2.71875,
+      "learning_rate": 3.528282569746609e-05,
+      "loss": 0.4985,
+      "step": 3451
+    },
+    {
+      "epoch": 0.883712,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.527855984984216e-05,
+      "loss": 0.4846,
+      "step": 3452
+    },
+    {
+      "epoch": 0.883968,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.527429400221824e-05,
+      "loss": 0.4843,
+      "step": 3453
+    },
+    {
+      "epoch": 0.884224,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.527002815459432e-05,
+      "loss": 0.515,
+      "step": 3454
+    },
+    {
+      "epoch": 0.88448,
+      "grad_norm": 2.125,
+      "learning_rate": 3.52657623069704e-05,
+      "loss": 0.5321,
+      "step": 3455
+    },
+    {
+      "epoch": 0.884736,
+      "grad_norm": 2.125,
+      "learning_rate": 3.526149645934647e-05,
+      "loss": 0.5458,
+      "step": 3456
+    },
+    {
+      "epoch": 0.884992,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.525723061172255e-05,
+      "loss": 0.4718,
+      "step": 3457
+    },
+    {
+      "epoch": 0.885248,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.525296476409863e-05,
+      "loss": 0.5068,
+      "step": 3458
+    },
+    {
+      "epoch": 0.885504,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.52486989164747e-05,
+      "loss": 0.5109,
+      "step": 3459
+    },
+    {
+      "epoch": 0.88576,
+      "grad_norm": 2.125,
+      "learning_rate": 3.524443306885078e-05,
+      "loss": 0.5281,
+      "step": 3460
+    },
+    {
+      "epoch": 0.886016,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.5240167221226864e-05,
+      "loss": 0.4799,
+      "step": 3461
+    },
+    {
+      "epoch": 0.886272,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.523590137360294e-05,
+      "loss": 0.4821,
+      "step": 3462
+    },
+    {
+      "epoch": 0.886528,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.523163552597901e-05,
+      "loss": 0.5092,
+      "step": 3463
+    },
+    {
+      "epoch": 0.886784,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.5227369678355094e-05,
+      "loss": 0.5185,
+      "step": 3464
+    },
+    {
+      "epoch": 0.88704,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.522310383073117e-05,
+      "loss": 0.5013,
+      "step": 3465
+    },
+    {
+      "epoch": 0.887296,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.521883798310724e-05,
+      "loss": 0.5407,
+      "step": 3466
+    },
+    {
+      "epoch": 0.887552,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.521457213548332e-05,
+      "loss": 0.4836,
+      "step": 3467
+    },
+    {
+      "epoch": 0.887808,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.5210306287859404e-05,
+      "loss": 0.5127,
+      "step": 3468
+    },
+    {
+      "epoch": 0.888064,
+      "grad_norm": 2.25,
+      "learning_rate": 3.520604044023548e-05,
+      "loss": 0.5314,
+      "step": 3469
+    },
+    {
+      "epoch": 0.88832,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.520177459261155e-05,
+      "loss": 0.4894,
+      "step": 3470
+    },
+    {
+      "epoch": 0.888576,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.519750874498763e-05,
+      "loss": 0.5126,
+      "step": 3471
+    },
+    {
+      "epoch": 0.888832,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.519324289736371e-05,
+      "loss": 0.4856,
+      "step": 3472
+    },
+    {
+      "epoch": 0.889088,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.518897704973978e-05,
+      "loss": 0.4952,
+      "step": 3473
+    },
+    {
+      "epoch": 0.889344,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.518471120211586e-05,
+      "loss": 0.4937,
+      "step": 3474
+    },
+    {
+      "epoch": 0.8896,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.518044535449194e-05,
+      "loss": 0.529,
+      "step": 3475
+    },
+    {
+      "epoch": 0.889856,
+      "grad_norm": 2.25,
+      "learning_rate": 3.517617950686802e-05,
+      "loss": 0.4997,
+      "step": 3476
+    },
+    {
+      "epoch": 0.890112,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.517191365924409e-05,
+      "loss": 0.4899,
+      "step": 3477
+    },
+    {
+      "epoch": 0.890368,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.516764781162017e-05,
+      "loss": 0.5299,
+      "step": 3478
+    },
+    {
+      "epoch": 0.890624,
+      "grad_norm": 2.125,
+      "learning_rate": 3.516338196399625e-05,
+      "loss": 0.4887,
+      "step": 3479
+    },
+    {
+      "epoch": 0.89088,
+      "grad_norm": 2.25,
+      "learning_rate": 3.515911611637233e-05,
+      "loss": 0.49,
+      "step": 3480
+    },
+    {
+      "epoch": 0.891136,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.5154850268748403e-05,
+      "loss": 0.4917,
+      "step": 3481
+    },
+    {
+      "epoch": 0.891392,
+      "grad_norm": 2.375,
+      "learning_rate": 3.515058442112448e-05,
+      "loss": 0.508,
+      "step": 3482
+    },
+    {
+      "epoch": 0.891648,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.514631857350056e-05,
+      "loss": 0.4666,
+      "step": 3483
+    },
+    {
+      "epoch": 0.891904,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.514205272587663e-05,
+      "loss": 0.5461,
+      "step": 3484
+    },
+    {
+      "epoch": 0.89216,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.513778687825271e-05,
+      "loss": 0.4856,
+      "step": 3485
+    },
+    {
+      "epoch": 0.892416,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.513352103062879e-05,
+      "loss": 0.5328,
+      "step": 3486
+    },
+    {
+      "epoch": 0.892672,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.512925518300487e-05,
+      "loss": 0.5115,
+      "step": 3487
+    },
+    {
+      "epoch": 0.892928,
+      "grad_norm": 2.375,
+      "learning_rate": 3.5124989335380944e-05,
+      "loss": 0.4932,
+      "step": 3488
+    },
+    {
+      "epoch": 0.893184,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.512072348775702e-05,
+      "loss": 0.5346,
+      "step": 3489
+    },
+    {
+      "epoch": 0.89344,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.511645764013309e-05,
+      "loss": 0.5135,
+      "step": 3490
+    },
+    {
+      "epoch": 0.893696,
+      "grad_norm": 2.25,
+      "learning_rate": 3.511219179250917e-05,
+      "loss": 0.5084,
+      "step": 3491
+    },
+    {
+      "epoch": 0.893952,
+      "grad_norm": 2.65625,
+      "learning_rate": 3.510792594488525e-05,
+      "loss": 0.5306,
+      "step": 3492
+    },
+    {
+      "epoch": 0.894208,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.510366009726133e-05,
+      "loss": 0.4438,
+      "step": 3493
+    },
+    {
+      "epoch": 0.894464,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.509939424963741e-05,
+      "loss": 0.5106,
+      "step": 3494
+    },
+    {
+      "epoch": 0.89472,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.5095128402013484e-05,
+      "loss": 0.4777,
+      "step": 3495
+    },
+    {
+      "epoch": 0.894976,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.509086255438956e-05,
+      "loss": 0.5131,
+      "step": 3496
+    },
+    {
+      "epoch": 0.895232,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.508659670676563e-05,
+      "loss": 0.5152,
+      "step": 3497
+    },
+    {
+      "epoch": 0.895488,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.5082330859141713e-05,
+      "loss": 0.4853,
+      "step": 3498
+    },
+    {
+      "epoch": 0.895744,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.5078065011517794e-05,
+      "loss": 0.4825,
+      "step": 3499
+    },
+    {
+      "epoch": 0.896,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.507379916389387e-05,
+      "loss": 0.5303,
+      "step": 3500
+    },
+    {
+      "epoch": 0.896256,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.506953331626994e-05,
+      "loss": 0.4981,
+      "step": 3501
+    },
+    {
+      "epoch": 0.896512,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.5065267468646024e-05,
+      "loss": 0.452,
+      "step": 3502
+    },
+    {
+      "epoch": 0.896768,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.50610016210221e-05,
+      "loss": 0.5593,
+      "step": 3503
+    },
+    {
+      "epoch": 0.897024,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.505673577339817e-05,
+      "loss": 0.5591,
+      "step": 3504
+    },
+    {
+      "epoch": 0.89728,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.5052469925774254e-05,
+      "loss": 0.4895,
+      "step": 3505
+    },
+    {
+      "epoch": 0.897536,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.5048204078150335e-05,
+      "loss": 0.5089,
+      "step": 3506
+    },
+    {
+      "epoch": 0.897792,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.504393823052641e-05,
+      "loss": 0.5247,
+      "step": 3507
+    },
+    {
+      "epoch": 0.898048,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.503967238290248e-05,
+      "loss": 0.5075,
+      "step": 3508
+    },
+    {
+      "epoch": 0.898304,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.503540653527856e-05,
+      "loss": 0.5109,
+      "step": 3509
+    },
+    {
+      "epoch": 0.89856,
+      "grad_norm": 2.375,
+      "learning_rate": 3.503114068765464e-05,
+      "loss": 0.5009,
+      "step": 3510
+    },
+    {
+      "epoch": 0.898816,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.502687484003071e-05,
+      "loss": 0.5123,
+      "step": 3511
+    },
+    {
+      "epoch": 0.899072,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.5022608992406794e-05,
+      "loss": 0.5001,
+      "step": 3512
+    },
+    {
+      "epoch": 0.899328,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.5018343144782875e-05,
+      "loss": 0.5143,
+      "step": 3513
+    },
+    {
+      "epoch": 0.899584,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.501407729715895e-05,
+      "loss": 0.4777,
+      "step": 3514
+    },
+    {
+      "epoch": 0.89984,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.5009811449535023e-05,
+      "loss": 0.5153,
+      "step": 3515
+    },
+    {
+      "epoch": 0.900096,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.50055456019111e-05,
+      "loss": 0.4968,
+      "step": 3516
+    },
+    {
+      "epoch": 0.900352,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.500127975428718e-05,
+      "loss": 0.5158,
+      "step": 3517
+    },
+    {
+      "epoch": 0.900608,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.499701390666325e-05,
+      "loss": 0.4786,
+      "step": 3518
+    },
+    {
+      "epoch": 0.900864,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.4992748059039334e-05,
+      "loss": 0.5266,
+      "step": 3519
+    },
+    {
+      "epoch": 0.90112,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.498848221141541e-05,
+      "loss": 0.5144,
+      "step": 3520
+    },
+    {
+      "epoch": 0.901376,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.498421636379149e-05,
+      "loss": 0.476,
+      "step": 3521
+    },
+    {
+      "epoch": 0.901632,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.4979950516167564e-05,
+      "loss": 0.5082,
+      "step": 3522
+    },
+    {
+      "epoch": 0.901888,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.497568466854364e-05,
+      "loss": 0.4818,
+      "step": 3523
+    },
+    {
+      "epoch": 0.902144,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.497141882091972e-05,
+      "loss": 0.5127,
+      "step": 3524
+    },
+    {
+      "epoch": 0.9024,
+      "grad_norm": 2.25,
+      "learning_rate": 3.49671529732958e-05,
+      "loss": 0.5039,
+      "step": 3525
+    },
+    {
+      "epoch": 0.902656,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.4962887125671874e-05,
+      "loss": 0.5278,
+      "step": 3526
+    },
+    {
+      "epoch": 0.902912,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.495862127804795e-05,
+      "loss": 0.4951,
+      "step": 3527
+    },
+    {
+      "epoch": 0.903168,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.495435543042403e-05,
+      "loss": 0.5159,
+      "step": 3528
+    },
+    {
+      "epoch": 0.903424,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.4950089582800104e-05,
+      "loss": 0.4937,
+      "step": 3529
+    },
+    {
+      "epoch": 0.90368,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.494582373517618e-05,
+      "loss": 0.5193,
+      "step": 3530
+    },
+    {
+      "epoch": 0.903936,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.494155788755226e-05,
+      "loss": 0.5178,
+      "step": 3531
+    },
+    {
+      "epoch": 0.904192,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.493729203992834e-05,
+      "loss": 0.4887,
+      "step": 3532
+    },
+    {
+      "epoch": 0.904448,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.4933026192304414e-05,
+      "loss": 0.4621,
+      "step": 3533
+    },
+    {
+      "epoch": 0.904704,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.492876034468049e-05,
+      "loss": 0.5005,
+      "step": 3534
+    },
+    {
+      "epoch": 0.90496,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.492449449705656e-05,
+      "loss": 0.4948,
+      "step": 3535
+    },
+    {
+      "epoch": 0.905216,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.4920228649432644e-05,
+      "loss": 0.5031,
+      "step": 3536
+    },
+    {
+      "epoch": 0.905472,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.491596280180872e-05,
+      "loss": 0.4871,
+      "step": 3537
+    },
+    {
+      "epoch": 0.905728,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.49116969541848e-05,
+      "loss": 0.4817,
+      "step": 3538
+    },
+    {
+      "epoch": 0.905984,
+      "grad_norm": 2.5625,
+      "learning_rate": 3.4907431106560874e-05,
+      "loss": 0.4975,
+      "step": 3539
+    },
+    {
+      "epoch": 0.90624,
+      "grad_norm": 2.125,
+      "learning_rate": 3.4903165258936955e-05,
+      "loss": 0.5033,
+      "step": 3540
+    },
+    {
+      "epoch": 0.906496,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.489889941131303e-05,
+      "loss": 0.5082,
+      "step": 3541
+    },
+    {
+      "epoch": 0.906752,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.48946335636891e-05,
+      "loss": 0.5466,
+      "step": 3542
+    },
+    {
+      "epoch": 0.907008,
+      "grad_norm": 2.375,
+      "learning_rate": 3.4890367716065184e-05,
+      "loss": 0.4767,
+      "step": 3543
+    },
+    {
+      "epoch": 0.907264,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.4886101868441265e-05,
+      "loss": 0.4652,
+      "step": 3544
+    },
+    {
+      "epoch": 0.90752,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.488183602081734e-05,
+      "loss": 0.5006,
+      "step": 3545
+    },
+    {
+      "epoch": 0.907776,
+      "grad_norm": 2.578125,
+      "learning_rate": 3.4877570173193414e-05,
+      "loss": 0.4746,
+      "step": 3546
+    },
+    {
+      "epoch": 0.908032,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.4873304325569495e-05,
+      "loss": 0.5322,
+      "step": 3547
+    },
+    {
+      "epoch": 0.908288,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.486903847794557e-05,
+      "loss": 0.494,
+      "step": 3548
+    },
+    {
+      "epoch": 0.908544,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.486477263032164e-05,
+      "loss": 0.5383,
+      "step": 3549
+    },
+    {
+      "epoch": 0.9088,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.4860506782697724e-05,
+      "loss": 0.4723,
+      "step": 3550
+    },
+    {
+      "epoch": 0.909056,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.4856240935073805e-05,
+      "loss": 0.4684,
+      "step": 3551
+    },
+    {
+      "epoch": 0.909312,
+      "grad_norm": 2.375,
+      "learning_rate": 3.485197508744988e-05,
+      "loss": 0.5073,
+      "step": 3552
+    },
+    {
+      "epoch": 0.909568,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.4847709239825954e-05,
+      "loss": 0.445,
+      "step": 3553
+    },
+    {
+      "epoch": 0.909824,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.484344339220203e-05,
+      "loss": 0.5473,
+      "step": 3554
+    },
+    {
+      "epoch": 0.91008,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.483917754457811e-05,
+      "loss": 0.5229,
+      "step": 3555
+    },
+    {
+      "epoch": 0.910336,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.4834911696954184e-05,
+      "loss": 0.5255,
+      "step": 3556
+    },
+    {
+      "epoch": 0.910592,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.4830645849330265e-05,
+      "loss": 0.5059,
+      "step": 3557
+    },
+    {
+      "epoch": 0.910848,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.482638000170634e-05,
+      "loss": 0.5462,
+      "step": 3558
+    },
+    {
+      "epoch": 0.911104,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.482211415408242e-05,
+      "loss": 0.5255,
+      "step": 3559
+    },
+    {
+      "epoch": 0.91136,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.4817848306458494e-05,
+      "loss": 0.4848,
+      "step": 3560
+    },
+    {
+      "epoch": 0.911616,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.481358245883457e-05,
+      "loss": 0.4867,
+      "step": 3561
+    },
+    {
+      "epoch": 0.911872,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.480931661121065e-05,
+      "loss": 0.45,
+      "step": 3562
+    },
+    {
+      "epoch": 0.912128,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.480505076358673e-05,
+      "loss": 0.4817,
+      "step": 3563
+    },
+    {
+      "epoch": 0.912384,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.4800784915962805e-05,
+      "loss": 0.5104,
+      "step": 3564
+    },
+    {
+      "epoch": 0.91264,
+      "grad_norm": 2.25,
+      "learning_rate": 3.479651906833888e-05,
+      "loss": 0.4734,
+      "step": 3565
+    },
+    {
+      "epoch": 0.912896,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.479225322071496e-05,
+      "loss": 0.4985,
+      "step": 3566
+    },
+    {
+      "epoch": 0.913152,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.4787987373091034e-05,
+      "loss": 0.4824,
+      "step": 3567
+    },
+    {
+      "epoch": 0.913408,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.478372152546711e-05,
+      "loss": 0.5089,
+      "step": 3568
+    },
+    {
+      "epoch": 0.913664,
+      "grad_norm": 2.125,
+      "learning_rate": 3.477945567784319e-05,
+      "loss": 0.4981,
+      "step": 3569
+    },
+    {
+      "epoch": 0.91392,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.477518983021927e-05,
+      "loss": 0.4932,
+      "step": 3570
+    },
+    {
+      "epoch": 0.914176,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.4770923982595345e-05,
+      "loss": 0.4855,
+      "step": 3571
+    },
+    {
+      "epoch": 0.914432,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.476665813497142e-05,
+      "loss": 0.5074,
+      "step": 3572
+    },
+    {
+      "epoch": 0.914688,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.4762392287347493e-05,
+      "loss": 0.474,
+      "step": 3573
+    },
+    {
+      "epoch": 0.914944,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.4758126439723575e-05,
+      "loss": 0.5042,
+      "step": 3574
+    },
+    {
+      "epoch": 0.9152,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.475386059209965e-05,
+      "loss": 0.5244,
+      "step": 3575
+    },
+    {
+      "epoch": 0.915456,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.474959474447573e-05,
+      "loss": 0.5337,
+      "step": 3576
+    },
+    {
+      "epoch": 0.915712,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.474532889685181e-05,
+      "loss": 0.525,
+      "step": 3577
+    },
+    {
+      "epoch": 0.915968,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.4741063049227885e-05,
+      "loss": 0.4813,
+      "step": 3578
+    },
+    {
+      "epoch": 0.916224,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.473679720160396e-05,
+      "loss": 0.5057,
+      "step": 3579
+    },
+    {
+      "epoch": 0.91648,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.4732531353980034e-05,
+      "loss": 0.5055,
+      "step": 3580
+    },
+    {
+      "epoch": 0.916736,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.4728265506356115e-05,
+      "loss": 0.488,
+      "step": 3581
+    },
+    {
+      "epoch": 0.916992,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.4723999658732196e-05,
+      "loss": 0.512,
+      "step": 3582
+    },
+    {
+      "epoch": 0.917248,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.471973381110827e-05,
+      "loss": 0.4863,
+      "step": 3583
+    },
+    {
+      "epoch": 0.917504,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.4715467963484344e-05,
+      "loss": 0.5124,
+      "step": 3584
+    },
+    {
+      "epoch": 0.91776,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.4711202115860425e-05,
+      "loss": 0.4948,
+      "step": 3585
+    },
+    {
+      "epoch": 0.918016,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.47069362682365e-05,
+      "loss": 0.5251,
+      "step": 3586
+    },
+    {
+      "epoch": 0.918272,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.4702670420612574e-05,
+      "loss": 0.4943,
+      "step": 3587
+    },
+    {
+      "epoch": 0.918528,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.4698404572988655e-05,
+      "loss": 0.5118,
+      "step": 3588
+    },
+    {
+      "epoch": 0.918784,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.4694138725364736e-05,
+      "loss": 0.5342,
+      "step": 3589
+    },
+    {
+      "epoch": 0.91904,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.468987287774081e-05,
+      "loss": 0.4604,
+      "step": 3590
+    },
+    {
+      "epoch": 0.919296,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.4685607030116884e-05,
+      "loss": 0.5024,
+      "step": 3591
+    },
+    {
+      "epoch": 0.919552,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.4681341182492966e-05,
+      "loss": 0.4688,
+      "step": 3592
+    },
+    {
+      "epoch": 0.919808,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.467707533486904e-05,
+      "loss": 0.5047,
+      "step": 3593
+    },
+    {
+      "epoch": 0.920064,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.4672809487245114e-05,
+      "loss": 0.4865,
+      "step": 3594
+    },
+    {
+      "epoch": 0.92032,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.4668543639621195e-05,
+      "loss": 0.5005,
+      "step": 3595
+    },
+    {
+      "epoch": 0.920576,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.4664277791997276e-05,
+      "loss": 0.5003,
+      "step": 3596
+    },
+    {
+      "epoch": 0.920832,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.466001194437335e-05,
+      "loss": 0.4525,
+      "step": 3597
+    },
+    {
+      "epoch": 0.921088,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.4655746096749425e-05,
+      "loss": 0.449,
+      "step": 3598
+    },
+    {
+      "epoch": 0.921344,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.46514802491255e-05,
+      "loss": 0.4993,
+      "step": 3599
+    },
+    {
+      "epoch": 0.9216,
+      "grad_norm": 2.375,
+      "learning_rate": 3.464721440150158e-05,
+      "loss": 0.4945,
+      "step": 3600
+    },
+    {
+      "epoch": 0.921856,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.464294855387766e-05,
+      "loss": 0.5001,
+      "step": 3601
+    },
+    {
+      "epoch": 0.922112,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.4638682706253735e-05,
+      "loss": 0.5268,
+      "step": 3602
+    },
+    {
+      "epoch": 0.922368,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.463441685862981e-05,
+      "loss": 0.5231,
+      "step": 3603
+    },
+    {
+      "epoch": 0.922624,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.463015101100589e-05,
+      "loss": 0.5194,
+      "step": 3604
+    },
+    {
+      "epoch": 0.92288,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.4625885163381965e-05,
+      "loss": 0.5162,
+      "step": 3605
+    },
+    {
+      "epoch": 0.923136,
+      "grad_norm": 2.375,
+      "learning_rate": 3.462161931575804e-05,
+      "loss": 0.4952,
+      "step": 3606
+    },
+    {
+      "epoch": 0.923392,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.461735346813412e-05,
+      "loss": 0.479,
+      "step": 3607
+    },
+    {
+      "epoch": 0.923648,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.46130876205102e-05,
+      "loss": 0.4776,
+      "step": 3608
+    },
+    {
+      "epoch": 0.923904,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.4608821772886275e-05,
+      "loss": 0.489,
+      "step": 3609
+    },
+    {
+      "epoch": 0.92416,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.460455592526235e-05,
+      "loss": 0.5242,
+      "step": 3610
+    },
+    {
+      "epoch": 0.924416,
+      "grad_norm": 2.375,
+      "learning_rate": 3.460029007763843e-05,
+      "loss": 0.4934,
+      "step": 3611
+    },
+    {
+      "epoch": 0.924672,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.4596024230014505e-05,
+      "loss": 0.49,
+      "step": 3612
+    },
+    {
+      "epoch": 0.924928,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.459175838239058e-05,
+      "loss": 0.5136,
+      "step": 3613
+    },
+    {
+      "epoch": 0.925184,
+      "grad_norm": 2.25,
+      "learning_rate": 3.458749253476666e-05,
+      "loss": 0.5069,
+      "step": 3614
+    },
+    {
+      "epoch": 0.92544,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.458322668714274e-05,
+      "loss": 0.5054,
+      "step": 3615
+    },
+    {
+      "epoch": 0.925696,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.4578960839518816e-05,
+      "loss": 0.4994,
+      "step": 3616
+    },
+    {
+      "epoch": 0.925952,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.457469499189489e-05,
+      "loss": 0.4913,
+      "step": 3617
+    },
+    {
+      "epoch": 0.926208,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.4570429144270964e-05,
+      "loss": 0.4965,
+      "step": 3618
+    },
+    {
+      "epoch": 0.926464,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.4566163296647045e-05,
+      "loss": 0.479,
+      "step": 3619
+    },
+    {
+      "epoch": 0.92672,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.456189744902312e-05,
+      "loss": 0.4736,
+      "step": 3620
+    },
+    {
+      "epoch": 0.926976,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.45576316013992e-05,
+      "loss": 0.4534,
+      "step": 3621
+    },
+    {
+      "epoch": 0.927232,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.4553365753775275e-05,
+      "loss": 0.492,
+      "step": 3622
+    },
+    {
+      "epoch": 0.927488,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.4549099906151356e-05,
+      "loss": 0.4713,
+      "step": 3623
+    },
+    {
+      "epoch": 0.927744,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.454483405852743e-05,
+      "loss": 0.4956,
+      "step": 3624
+    },
+    {
+      "epoch": 0.928,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.4540568210903504e-05,
+      "loss": 0.4796,
+      "step": 3625
+    },
+    {
+      "epoch": 0.928256,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.4536302363279585e-05,
+      "loss": 0.4788,
+      "step": 3626
+    },
+    {
+      "epoch": 0.928512,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.4532036515655666e-05,
+      "loss": 0.5308,
+      "step": 3627
+    },
+    {
+      "epoch": 0.928768,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.452777066803174e-05,
+      "loss": 0.4743,
+      "step": 3628
+    },
+    {
+      "epoch": 0.929024,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.4523504820407815e-05,
+      "loss": 0.4816,
+      "step": 3629
+    },
+    {
+      "epoch": 0.92928,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.4519238972783896e-05,
+      "loss": 0.4578,
+      "step": 3630
+    },
+    {
+      "epoch": 0.929536,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.451497312515997e-05,
+      "loss": 0.4856,
+      "step": 3631
+    },
+    {
+      "epoch": 0.929792,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.4510707277536045e-05,
+      "loss": 0.4729,
+      "step": 3632
+    },
+    {
+      "epoch": 0.930048,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.4506441429912126e-05,
+      "loss": 0.5034,
+      "step": 3633
+    },
+    {
+      "epoch": 0.930304,
+      "grad_norm": 2.25,
+      "learning_rate": 3.450217558228821e-05,
+      "loss": 0.4911,
+      "step": 3634
+    },
+    {
+      "epoch": 0.93056,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.449790973466428e-05,
+      "loss": 0.4776,
+      "step": 3635
+    },
+    {
+      "epoch": 0.930816,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.4493643887040355e-05,
+      "loss": 0.4856,
+      "step": 3636
+    },
+    {
+      "epoch": 0.931072,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.448937803941643e-05,
+      "loss": 0.5012,
+      "step": 3637
+    },
+    {
+      "epoch": 0.931328,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.448511219179251e-05,
+      "loss": 0.4843,
+      "step": 3638
+    },
+    {
+      "epoch": 0.931584,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.4480846344168585e-05,
+      "loss": 0.5073,
+      "step": 3639
+    },
+    {
+      "epoch": 0.93184,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.4476580496544666e-05,
+      "loss": 0.5194,
+      "step": 3640
+    },
+    {
+      "epoch": 0.932096,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.447231464892075e-05,
+      "loss": 0.5273,
+      "step": 3641
+    },
+    {
+      "epoch": 0.932352,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.446804880129682e-05,
+      "loss": 0.5061,
+      "step": 3642
+    },
+    {
+      "epoch": 0.932608,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.4463782953672895e-05,
+      "loss": 0.5012,
+      "step": 3643
+    },
+    {
+      "epoch": 0.932864,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.445951710604897e-05,
+      "loss": 0.5347,
+      "step": 3644
+    },
+    {
+      "epoch": 0.93312,
+      "grad_norm": 2.5,
+      "learning_rate": 3.445525125842505e-05,
+      "loss": 0.4966,
+      "step": 3645
+    },
+    {
+      "epoch": 0.933376,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.445098541080113e-05,
+      "loss": 0.5068,
+      "step": 3646
+    },
+    {
+      "epoch": 0.933632,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.4446719563177206e-05,
+      "loss": 0.5329,
+      "step": 3647
+    },
+    {
+      "epoch": 0.933888,
+      "grad_norm": 2.25,
+      "learning_rate": 3.444245371555328e-05,
+      "loss": 0.5054,
+      "step": 3648
+    },
+    {
+      "epoch": 0.934144,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.443818786792936e-05,
+      "loss": 0.5273,
+      "step": 3649
+    },
+    {
+      "epoch": 0.9344,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.4433922020305436e-05,
+      "loss": 0.5032,
+      "step": 3650
+    },
+    {
+      "epoch": 0.934656,
+      "grad_norm": 2.25,
+      "learning_rate": 3.442965617268151e-05,
+      "loss": 0.536,
+      "step": 3651
+    },
+    {
+      "epoch": 0.934912,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.442539032505759e-05,
+      "loss": 0.4991,
+      "step": 3652
+    },
+    {
+      "epoch": 0.935168,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.442112447743367e-05,
+      "loss": 0.5286,
+      "step": 3653
+    },
+    {
+      "epoch": 0.935424,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.4416858629809746e-05,
+      "loss": 0.504,
+      "step": 3654
+    },
+    {
+      "epoch": 0.93568,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.441259278218582e-05,
+      "loss": 0.5213,
+      "step": 3655
+    },
+    {
+      "epoch": 0.935936,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.4408326934561895e-05,
+      "loss": 0.5401,
+      "step": 3656
+    },
+    {
+      "epoch": 0.936192,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.4404061086937976e-05,
+      "loss": 0.5408,
+      "step": 3657
+    },
+    {
+      "epoch": 0.936448,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.439979523931405e-05,
+      "loss": 0.4846,
+      "step": 3658
+    },
+    {
+      "epoch": 0.936704,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.439552939169013e-05,
+      "loss": 0.4782,
+      "step": 3659
+    },
+    {
+      "epoch": 0.93696,
+      "grad_norm": 2.5,
+      "learning_rate": 3.439126354406621e-05,
+      "loss": 0.5549,
+      "step": 3660
+    },
+    {
+      "epoch": 0.937216,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.4386997696442286e-05,
+      "loss": 0.4995,
+      "step": 3661
+    },
+    {
+      "epoch": 0.937472,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.438273184881836e-05,
+      "loss": 0.4724,
+      "step": 3662
+    },
+    {
+      "epoch": 0.937728,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.4378466001194435e-05,
+      "loss": 0.5064,
+      "step": 3663
+    },
+    {
+      "epoch": 0.937984,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.4374200153570516e-05,
+      "loss": 0.5121,
+      "step": 3664
+    },
+    {
+      "epoch": 0.93824,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.43699343059466e-05,
+      "loss": 0.4872,
+      "step": 3665
+    },
+    {
+      "epoch": 0.938496,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.436566845832267e-05,
+      "loss": 0.5006,
+      "step": 3666
+    },
+    {
+      "epoch": 0.938752,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.4361402610698746e-05,
+      "loss": 0.4882,
+      "step": 3667
+    },
+    {
+      "epoch": 0.939008,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.435713676307483e-05,
+      "loss": 0.4967,
+      "step": 3668
+    },
+    {
+      "epoch": 0.939264,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.43528709154509e-05,
+      "loss": 0.4667,
+      "step": 3669
+    },
+    {
+      "epoch": 0.93952,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.4348605067826975e-05,
+      "loss": 0.4785,
+      "step": 3670
+    },
+    {
+      "epoch": 0.939776,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.4344339220203056e-05,
+      "loss": 0.5218,
+      "step": 3671
+    },
+    {
+      "epoch": 0.940032,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.434007337257914e-05,
+      "loss": 0.5074,
+      "step": 3672
+    },
+    {
+      "epoch": 0.940288,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.433580752495521e-05,
+      "loss": 0.4702,
+      "step": 3673
+    },
+    {
+      "epoch": 0.940544,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.4331541677331286e-05,
+      "loss": 0.4992,
+      "step": 3674
+    },
+    {
+      "epoch": 0.9408,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.432727582970737e-05,
+      "loss": 0.4896,
+      "step": 3675
+    },
+    {
+      "epoch": 0.941056,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.432300998208344e-05,
+      "loss": 0.503,
+      "step": 3676
+    },
+    {
+      "epoch": 0.941312,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.4318744134459515e-05,
+      "loss": 0.4892,
+      "step": 3677
+    },
+    {
+      "epoch": 0.941568,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.4314478286835596e-05,
+      "loss": 0.5493,
+      "step": 3678
+    },
+    {
+      "epoch": 0.941824,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.431021243921168e-05,
+      "loss": 0.5078,
+      "step": 3679
+    },
+    {
+      "epoch": 0.94208,
+      "grad_norm": 2.25,
+      "learning_rate": 3.430594659158775e-05,
+      "loss": 0.471,
+      "step": 3680
+    },
+    {
+      "epoch": 0.942336,
+      "grad_norm": 2.5625,
+      "learning_rate": 3.4301680743963826e-05,
+      "loss": 0.5019,
+      "step": 3681
+    },
+    {
+      "epoch": 0.942592,
+      "grad_norm": 2.625,
+      "learning_rate": 3.42974148963399e-05,
+      "loss": 0.4878,
+      "step": 3682
+    },
+    {
+      "epoch": 0.942848,
+      "grad_norm": 2.375,
+      "learning_rate": 3.429314904871598e-05,
+      "loss": 0.5249,
+      "step": 3683
+    },
+    {
+      "epoch": 0.943104,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.428888320109206e-05,
+      "loss": 0.5024,
+      "step": 3684
+    },
+    {
+      "epoch": 0.94336,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.4284617353468137e-05,
+      "loss": 0.5189,
+      "step": 3685
+    },
+    {
+      "epoch": 0.943616,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.428035150584421e-05,
+      "loss": 0.4771,
+      "step": 3686
+    },
+    {
+      "epoch": 0.943872,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.427608565822029e-05,
+      "loss": 0.4424,
+      "step": 3687
+    },
+    {
+      "epoch": 0.944128,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.4271819810596366e-05,
+      "loss": 0.508,
+      "step": 3688
+    },
+    {
+      "epoch": 0.944384,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.426755396297244e-05,
+      "loss": 0.4929,
+      "step": 3689
+    },
+    {
+      "epoch": 0.94464,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.426328811534852e-05,
+      "loss": 0.4641,
+      "step": 3690
+    },
+    {
+      "epoch": 0.944896,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.42590222677246e-05,
+      "loss": 0.5017,
+      "step": 3691
+    },
+    {
+      "epoch": 0.945152,
+      "grad_norm": 2.375,
+      "learning_rate": 3.425475642010068e-05,
+      "loss": 0.5205,
+      "step": 3692
+    },
+    {
+      "epoch": 0.945408,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.425049057247675e-05,
+      "loss": 0.5188,
+      "step": 3693
+    },
+    {
+      "epoch": 0.945664,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.424622472485283e-05,
+      "loss": 0.4933,
+      "step": 3694
+    },
+    {
+      "epoch": 0.94592,
+      "grad_norm": 2.25,
+      "learning_rate": 3.4241958877228906e-05,
+      "loss": 0.4986,
+      "step": 3695
+    },
+    {
+      "epoch": 0.946176,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.423769302960498e-05,
+      "loss": 0.5263,
+      "step": 3696
+    },
+    {
+      "epoch": 0.946432,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.423342718198106e-05,
+      "loss": 0.5318,
+      "step": 3697
+    },
+    {
+      "epoch": 0.946688,
+      "grad_norm": 2.125,
+      "learning_rate": 3.422916133435714e-05,
+      "loss": 0.543,
+      "step": 3698
+    },
+    {
+      "epoch": 0.946944,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.422489548673322e-05,
+      "loss": 0.4567,
+      "step": 3699
+    },
+    {
+      "epoch": 0.9472,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.422062963910929e-05,
+      "loss": 0.5252,
+      "step": 3700
+    },
+    {
+      "epoch": 0.947456,
+      "grad_norm": 2.609375,
+      "learning_rate": 3.4216363791485365e-05,
+      "loss": 0.5084,
+      "step": 3701
+    },
+    {
+      "epoch": 0.947712,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.4212097943861447e-05,
+      "loss": 0.4985,
+      "step": 3702
+    },
+    {
+      "epoch": 0.947968,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.420783209623753e-05,
+      "loss": 0.4821,
+      "step": 3703
+    },
+    {
+      "epoch": 0.948224,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.42035662486136e-05,
+      "loss": 0.5087,
+      "step": 3704
+    },
+    {
+      "epoch": 0.94848,
+      "grad_norm": 2.25,
+      "learning_rate": 3.4199300400989676e-05,
+      "loss": 0.4975,
+      "step": 3705
+    },
+    {
+      "epoch": 0.948736,
+      "grad_norm": 2.25,
+      "learning_rate": 3.419503455336576e-05,
+      "loss": 0.4862,
+      "step": 3706
+    },
+    {
+      "epoch": 0.948992,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.419076870574183e-05,
+      "loss": 0.5231,
+      "step": 3707
+    },
+    {
+      "epoch": 0.949248,
+      "grad_norm": 2.375,
+      "learning_rate": 3.4186502858117906e-05,
+      "loss": 0.4936,
+      "step": 3708
+    },
+    {
+      "epoch": 0.949504,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.418223701049399e-05,
+      "loss": 0.4622,
+      "step": 3709
+    },
+    {
+      "epoch": 0.94976,
+      "grad_norm": 2.25,
+      "learning_rate": 3.417797116287007e-05,
+      "loss": 0.4737,
+      "step": 3710
+    },
+    {
+      "epoch": 0.950016,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.417370531524614e-05,
+      "loss": 0.4479,
+      "step": 3711
+    },
+    {
+      "epoch": 0.950272,
+      "grad_norm": 2.5,
+      "learning_rate": 3.4169439467622216e-05,
+      "loss": 0.4685,
+      "step": 3712
+    },
+    {
+      "epoch": 0.950528,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.41651736199983e-05,
+      "loss": 0.5163,
+      "step": 3713
+    },
+    {
+      "epoch": 0.950784,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.416090777237437e-05,
+      "loss": 0.4898,
+      "step": 3714
+    },
+    {
+      "epoch": 0.95104,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.4156641924750446e-05,
+      "loss": 0.4809,
+      "step": 3715
+    },
+    {
+      "epoch": 0.951296,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.415237607712653e-05,
+      "loss": 0.4931,
+      "step": 3716
+    },
+    {
+      "epoch": 0.951552,
+      "grad_norm": 2.25,
+      "learning_rate": 3.414811022950261e-05,
+      "loss": 0.464,
+      "step": 3717
+    },
+    {
+      "epoch": 0.951808,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.414384438187868e-05,
+      "loss": 0.4884,
+      "step": 3718
+    },
+    {
+      "epoch": 0.952064,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.4139578534254756e-05,
+      "loss": 0.4813,
+      "step": 3719
+    },
+    {
+      "epoch": 0.95232,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.413531268663083e-05,
+      "loss": 0.4804,
+      "step": 3720
+    },
+    {
+      "epoch": 0.952576,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.413104683900691e-05,
+      "loss": 0.541,
+      "step": 3721
+    },
+    {
+      "epoch": 0.952832,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.4126780991382986e-05,
+      "loss": 0.5281,
+      "step": 3722
+    },
+    {
+      "epoch": 0.953088,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.412251514375907e-05,
+      "loss": 0.5242,
+      "step": 3723
+    },
+    {
+      "epoch": 0.953344,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.411824929613515e-05,
+      "loss": 0.4404,
+      "step": 3724
+    },
+    {
+      "epoch": 0.9536,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.411398344851122e-05,
+      "loss": 0.4791,
+      "step": 3725
+    },
+    {
+      "epoch": 0.953856,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.41097176008873e-05,
+      "loss": 0.4851,
+      "step": 3726
+    },
+    {
+      "epoch": 0.954112,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.410545175326337e-05,
+      "loss": 0.4631,
+      "step": 3727
+    },
+    {
+      "epoch": 0.954368,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.410118590563945e-05,
+      "loss": 0.5158,
+      "step": 3728
+    },
+    {
+      "epoch": 0.954624,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.409692005801553e-05,
+      "loss": 0.5201,
+      "step": 3729
+    },
+    {
+      "epoch": 0.95488,
+      "grad_norm": 2.609375,
+      "learning_rate": 3.409265421039161e-05,
+      "loss": 0.5134,
+      "step": 3730
+    },
+    {
+      "epoch": 0.955136,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.408838836276768e-05,
+      "loss": 0.4847,
+      "step": 3731
+    },
+    {
+      "epoch": 0.955392,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.408412251514376e-05,
+      "loss": 0.469,
+      "step": 3732
+    },
+    {
+      "epoch": 0.955648,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.407985666751984e-05,
+      "loss": 0.5211,
+      "step": 3733
+    },
+    {
+      "epoch": 0.955904,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.407559081989591e-05,
+      "loss": 0.5048,
+      "step": 3734
+    },
+    {
+      "epoch": 0.95616,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.407132497227199e-05,
+      "loss": 0.5183,
+      "step": 3735
+    },
+    {
+      "epoch": 0.956416,
+      "grad_norm": 2.375,
+      "learning_rate": 3.406705912464807e-05,
+      "loss": 0.4729,
+      "step": 3736
+    },
+    {
+      "epoch": 0.956672,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.406279327702415e-05,
+      "loss": 0.5088,
+      "step": 3737
+    },
+    {
+      "epoch": 0.956928,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.405852742940022e-05,
+      "loss": 0.4899,
+      "step": 3738
+    },
+    {
+      "epoch": 0.957184,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.40542615817763e-05,
+      "loss": 0.491,
+      "step": 3739
+    },
+    {
+      "epoch": 0.95744,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.404999573415238e-05,
+      "loss": 0.4751,
+      "step": 3740
+    },
+    {
+      "epoch": 0.957696,
+      "grad_norm": 2.5625,
+      "learning_rate": 3.404572988652845e-05,
+      "loss": 0.5244,
+      "step": 3741
+    },
+    {
+      "epoch": 0.957952,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.404146403890453e-05,
+      "loss": 0.4982,
+      "step": 3742
+    },
+    {
+      "epoch": 0.958208,
+      "grad_norm": 2.5625,
+      "learning_rate": 3.4037198191280613e-05,
+      "loss": 0.5011,
+      "step": 3743
+    },
+    {
+      "epoch": 0.958464,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.403293234365669e-05,
+      "loss": 0.4931,
+      "step": 3744
+    },
+    {
+      "epoch": 0.95872,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.402866649603276e-05,
+      "loss": 0.4854,
+      "step": 3745
+    },
+    {
+      "epoch": 0.958976,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.4024400648408836e-05,
+      "loss": 0.5265,
+      "step": 3746
+    },
+    {
+      "epoch": 0.959232,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.402013480078492e-05,
+      "loss": 0.5248,
+      "step": 3747
+    },
+    {
+      "epoch": 0.959488,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.4015868953161e-05,
+      "loss": 0.4575,
+      "step": 3748
+    },
+    {
+      "epoch": 0.959744,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.401160310553707e-05,
+      "loss": 0.4426,
+      "step": 3749
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.400733725791315e-05,
+      "loss": 0.4648,
+      "step": 3750
+    },
+    {
+      "epoch": 0.960256,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.400307141028923e-05,
+      "loss": 0.4572,
+      "step": 3751
+    },
+    {
+      "epoch": 0.960512,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.39988055626653e-05,
+      "loss": 0.491,
+      "step": 3752
+    },
+    {
+      "epoch": 0.960768,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.3994539715041376e-05,
+      "loss": 0.5035,
+      "step": 3753
+    },
+    {
+      "epoch": 0.961024,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.399027386741746e-05,
+      "loss": 0.4903,
+      "step": 3754
+    },
+    {
+      "epoch": 0.96128,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.398600801979354e-05,
+      "loss": 0.5152,
+      "step": 3755
+    },
+    {
+      "epoch": 0.961536,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.398174217216961e-05,
+      "loss": 0.482,
+      "step": 3756
+    },
+    {
+      "epoch": 0.961792,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.397747632454569e-05,
+      "loss": 0.5078,
+      "step": 3757
+    },
+    {
+      "epoch": 0.962048,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.397321047692177e-05,
+      "loss": 0.4853,
+      "step": 3758
+    },
+    {
+      "epoch": 0.962304,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.396894462929784e-05,
+      "loss": 0.475,
+      "step": 3759
+    },
+    {
+      "epoch": 0.96256,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.396467878167392e-05,
+      "loss": 0.5144,
+      "step": 3760
+    },
+    {
+      "epoch": 0.962816,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.396041293405e-05,
+      "loss": 0.4641,
+      "step": 3761
+    },
+    {
+      "epoch": 0.963072,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.395614708642608e-05,
+      "loss": 0.4806,
+      "step": 3762
+    },
+    {
+      "epoch": 0.963328,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.395188123880215e-05,
+      "loss": 0.4823,
+      "step": 3763
+    },
+    {
+      "epoch": 0.963584,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.394761539117823e-05,
+      "loss": 0.5207,
+      "step": 3764
+    },
+    {
+      "epoch": 0.96384,
+      "grad_norm": 2.578125,
+      "learning_rate": 3.39433495435543e-05,
+      "loss": 0.4859,
+      "step": 3765
+    },
+    {
+      "epoch": 0.964096,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.393908369593038e-05,
+      "loss": 0.4846,
+      "step": 3766
+    },
+    {
+      "epoch": 0.964352,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.3934817848306464e-05,
+      "loss": 0.4818,
+      "step": 3767
+    },
+    {
+      "epoch": 0.964608,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.393055200068254e-05,
+      "loss": 0.465,
+      "step": 3768
+    },
+    {
+      "epoch": 0.964864,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.392628615305861e-05,
+      "loss": 0.512,
+      "step": 3769
+    },
+    {
+      "epoch": 0.96512,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.392202030543469e-05,
+      "loss": 0.505,
+      "step": 3770
+    },
+    {
+      "epoch": 0.965376,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.391775445781077e-05,
+      "loss": 0.4837,
+      "step": 3771
+    },
+    {
+      "epoch": 0.965632,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.391348861018684e-05,
+      "loss": 0.4932,
+      "step": 3772
+    },
+    {
+      "epoch": 0.965888,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.390922276256292e-05,
+      "loss": 0.4995,
+      "step": 3773
+    },
+    {
+      "epoch": 0.966144,
+      "grad_norm": 2.25,
+      "learning_rate": 3.3904956914939004e-05,
+      "loss": 0.5137,
+      "step": 3774
+    },
+    {
+      "epoch": 0.9664,
+      "grad_norm": 2.25,
+      "learning_rate": 3.390069106731508e-05,
+      "loss": 0.5027,
+      "step": 3775
+    },
+    {
+      "epoch": 0.966656,
+      "grad_norm": 2.125,
+      "learning_rate": 3.389642521969115e-05,
+      "loss": 0.4823,
+      "step": 3776
+    },
+    {
+      "epoch": 0.966912,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.389215937206723e-05,
+      "loss": 0.5013,
+      "step": 3777
+    },
+    {
+      "epoch": 0.967168,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.388789352444331e-05,
+      "loss": 0.4811,
+      "step": 3778
+    },
+    {
+      "epoch": 0.967424,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.388362767681938e-05,
+      "loss": 0.503,
+      "step": 3779
+    },
+    {
+      "epoch": 0.96768,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.387936182919546e-05,
+      "loss": 0.4709,
+      "step": 3780
+    },
+    {
+      "epoch": 0.967936,
+      "grad_norm": 2.25,
+      "learning_rate": 3.3875095981571544e-05,
+      "loss": 0.5062,
+      "step": 3781
+    },
+    {
+      "epoch": 0.968192,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.387083013394762e-05,
+      "loss": 0.5032,
+      "step": 3782
+    },
+    {
+      "epoch": 0.968448,
+      "grad_norm": 2.125,
+      "learning_rate": 3.386656428632369e-05,
+      "loss": 0.5072,
+      "step": 3783
+    },
+    {
+      "epoch": 0.968704,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.386229843869977e-05,
+      "loss": 0.528,
+      "step": 3784
+    },
+    {
+      "epoch": 0.96896,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.385803259107585e-05,
+      "loss": 0.4822,
+      "step": 3785
+    },
+    {
+      "epoch": 0.969216,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.385376674345193e-05,
+      "loss": 0.5497,
+      "step": 3786
+    },
+    {
+      "epoch": 0.969472,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.3849500895828e-05,
+      "loss": 0.5051,
+      "step": 3787
+    },
+    {
+      "epoch": 0.969728,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.3845235048204084e-05,
+      "loss": 0.4726,
+      "step": 3788
+    },
+    {
+      "epoch": 0.969984,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.384096920058016e-05,
+      "loss": 0.4972,
+      "step": 3789
+    },
+    {
+      "epoch": 0.97024,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.383670335295623e-05,
+      "loss": 0.4859,
+      "step": 3790
+    },
+    {
+      "epoch": 0.970496,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.383243750533231e-05,
+      "loss": 0.4885,
+      "step": 3791
+    },
+    {
+      "epoch": 0.970752,
+      "grad_norm": 2.25,
+      "learning_rate": 3.382817165770839e-05,
+      "loss": 0.5295,
+      "step": 3792
+    },
+    {
+      "epoch": 0.971008,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.382390581008447e-05,
+      "loss": 0.4624,
+      "step": 3793
+    },
+    {
+      "epoch": 0.971264,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.381963996246054e-05,
+      "loss": 0.4893,
+      "step": 3794
+    },
+    {
+      "epoch": 0.97152,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.381537411483662e-05,
+      "loss": 0.4772,
+      "step": 3795
+    },
+    {
+      "epoch": 0.971776,
+      "grad_norm": 2.25,
+      "learning_rate": 3.38111082672127e-05,
+      "loss": 0.4792,
+      "step": 3796
+    },
+    {
+      "epoch": 0.972032,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.380684241958877e-05,
+      "loss": 0.4879,
+      "step": 3797
+    },
+    {
+      "epoch": 0.972288,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.380257657196485e-05,
+      "loss": 0.5034,
+      "step": 3798
+    },
+    {
+      "epoch": 0.972544,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.379831072434093e-05,
+      "loss": 0.4929,
+      "step": 3799
+    },
+    {
+      "epoch": 0.9728,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.379404487671701e-05,
+      "loss": 0.4642,
+      "step": 3800
+    },
+    {
+      "epoch": 0.973056,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.3789779029093084e-05,
+      "loss": 0.4636,
+      "step": 3801
+    },
+    {
+      "epoch": 0.973312,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.378551318146916e-05,
+      "loss": 0.4853,
+      "step": 3802
+    },
+    {
+      "epoch": 0.973568,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.378124733384523e-05,
+      "loss": 0.4819,
+      "step": 3803
+    },
+    {
+      "epoch": 0.973824,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.377698148622131e-05,
+      "loss": 0.517,
+      "step": 3804
+    },
+    {
+      "epoch": 0.97408,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.3772715638597394e-05,
+      "loss": 0.5027,
+      "step": 3805
+    },
+    {
+      "epoch": 0.974336,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.376844979097347e-05,
+      "loss": 0.4886,
+      "step": 3806
+    },
+    {
+      "epoch": 0.974592,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.376418394334955e-05,
+      "loss": 0.5149,
+      "step": 3807
+    },
+    {
+      "epoch": 0.974848,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.3759918095725624e-05,
+      "loss": 0.4779,
+      "step": 3808
+    },
+    {
+      "epoch": 0.975104,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.37556522481017e-05,
+      "loss": 0.4909,
+      "step": 3809
+    },
+    {
+      "epoch": 0.97536,
+      "grad_norm": 2.25,
+      "learning_rate": 3.375138640047777e-05,
+      "loss": 0.4432,
+      "step": 3810
+    },
+    {
+      "epoch": 0.975616,
+      "grad_norm": 2.375,
+      "learning_rate": 3.374712055285385e-05,
+      "loss": 0.5014,
+      "step": 3811
+    },
+    {
+      "epoch": 0.975872,
+      "grad_norm": 2.578125,
+      "learning_rate": 3.3742854705229934e-05,
+      "loss": 0.5047,
+      "step": 3812
+    },
+    {
+      "epoch": 0.976128,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.373858885760601e-05,
+      "loss": 0.4991,
+      "step": 3813
+    },
+    {
+      "epoch": 0.976384,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.373432300998208e-05,
+      "loss": 0.5027,
+      "step": 3814
+    },
+    {
+      "epoch": 0.97664,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.3730057162358164e-05,
+      "loss": 0.4926,
+      "step": 3815
+    },
+    {
+      "epoch": 0.976896,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.372579131473424e-05,
+      "loss": 0.5128,
+      "step": 3816
+    },
+    {
+      "epoch": 0.977152,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.372152546711031e-05,
+      "loss": 0.4604,
+      "step": 3817
+    },
+    {
+      "epoch": 0.977408,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.3717259619486393e-05,
+      "loss": 0.5114,
+      "step": 3818
+    },
+    {
+      "epoch": 0.977664,
+      "grad_norm": 2.5,
+      "learning_rate": 3.3712993771862475e-05,
+      "loss": 0.4743,
+      "step": 3819
+    },
+    {
+      "epoch": 0.97792,
+      "grad_norm": 2.65625,
+      "learning_rate": 3.370872792423855e-05,
+      "loss": 0.5018,
+      "step": 3820
+    },
+    {
+      "epoch": 0.978176,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.370446207661462e-05,
+      "loss": 0.4654,
+      "step": 3821
+    },
+    {
+      "epoch": 0.978432,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.3700196228990704e-05,
+      "loss": 0.4847,
+      "step": 3822
+    },
+    {
+      "epoch": 0.978688,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.369593038136678e-05,
+      "loss": 0.4516,
+      "step": 3823
+    },
+    {
+      "epoch": 0.978944,
+      "grad_norm": 2.671875,
+      "learning_rate": 3.369166453374286e-05,
+      "loss": 0.4869,
+      "step": 3824
+    },
+    {
+      "epoch": 0.9792,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.3687398686118934e-05,
+      "loss": 0.4876,
+      "step": 3825
+    },
+    {
+      "epoch": 0.979456,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.3683132838495015e-05,
+      "loss": 0.4621,
+      "step": 3826
+    },
+    {
+      "epoch": 0.979712,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.367886699087109e-05,
+      "loss": 0.4587,
+      "step": 3827
+    },
+    {
+      "epoch": 0.979968,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.367460114324716e-05,
+      "loss": 0.4467,
+      "step": 3828
+    },
+    {
+      "epoch": 0.980224,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.367033529562324e-05,
+      "loss": 0.5063,
+      "step": 3829
+    },
+    {
+      "epoch": 0.98048,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.366606944799932e-05,
+      "loss": 0.4589,
+      "step": 3830
+    },
+    {
+      "epoch": 0.980736,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.36618036003754e-05,
+      "loss": 0.5155,
+      "step": 3831
+    },
+    {
+      "epoch": 0.980992,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.3657537752751474e-05,
+      "loss": 0.4778,
+      "step": 3832
+    },
+    {
+      "epoch": 0.981248,
+      "grad_norm": 2.25,
+      "learning_rate": 3.365327190512755e-05,
+      "loss": 0.5174,
+      "step": 3833
+    },
+    {
+      "epoch": 0.981504,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.364900605750363e-05,
+      "loss": 0.4958,
+      "step": 3834
+    },
+    {
+      "epoch": 0.98176,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.3644740209879703e-05,
+      "loss": 0.5073,
+      "step": 3835
+    },
+    {
+      "epoch": 0.982016,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.364047436225578e-05,
+      "loss": 0.4638,
+      "step": 3836
+    },
+    {
+      "epoch": 0.982272,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.363620851463186e-05,
+      "loss": 0.4525,
+      "step": 3837
+    },
+    {
+      "epoch": 0.982528,
+      "grad_norm": 2.25,
+      "learning_rate": 3.363194266700794e-05,
+      "loss": 0.5202,
+      "step": 3838
+    },
+    {
+      "epoch": 0.982784,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.3627676819384014e-05,
+      "loss": 0.4964,
+      "step": 3839
+    },
+    {
+      "epoch": 0.98304,
+      "grad_norm": 2.375,
+      "learning_rate": 3.362341097176009e-05,
+      "loss": 0.5015,
+      "step": 3840
+    },
+    {
+      "epoch": 0.983296,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.361914512413617e-05,
+      "loss": 0.488,
+      "step": 3841
+    },
+    {
+      "epoch": 0.983552,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.3614879276512244e-05,
+      "loss": 0.4964,
+      "step": 3842
+    },
+    {
+      "epoch": 0.983808,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.361061342888832e-05,
+      "loss": 0.4957,
+      "step": 3843
+    },
+    {
+      "epoch": 0.984064,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.36063475812644e-05,
+      "loss": 0.5118,
+      "step": 3844
+    },
+    {
+      "epoch": 0.98432,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.360208173364048e-05,
+      "loss": 0.4672,
+      "step": 3845
+    },
+    {
+      "epoch": 0.984576,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.3597815886016554e-05,
+      "loss": 0.4896,
+      "step": 3846
+    },
+    {
+      "epoch": 0.984832,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.359355003839263e-05,
+      "loss": 0.5131,
+      "step": 3847
+    },
+    {
+      "epoch": 0.985088,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.35892841907687e-05,
+      "loss": 0.5008,
+      "step": 3848
+    },
+    {
+      "epoch": 0.985344,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.3585018343144784e-05,
+      "loss": 0.4959,
+      "step": 3849
+    },
+    {
+      "epoch": 0.9856,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.3580752495520865e-05,
+      "loss": 0.48,
+      "step": 3850
+    },
+    {
+      "epoch": 0.985856,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.357648664789694e-05,
+      "loss": 0.5099,
+      "step": 3851
+    },
+    {
+      "epoch": 0.986112,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.357222080027301e-05,
+      "loss": 0.4735,
+      "step": 3852
+    },
+    {
+      "epoch": 0.986368,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.3567954952649094e-05,
+      "loss": 0.4815,
+      "step": 3853
+    },
+    {
+      "epoch": 0.986624,
+      "grad_norm": 2.609375,
+      "learning_rate": 3.356368910502517e-05,
+      "loss": 0.4867,
+      "step": 3854
+    },
+    {
+      "epoch": 0.98688,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.355942325740124e-05,
+      "loss": 0.5005,
+      "step": 3855
+    },
+    {
+      "epoch": 0.987136,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.3555157409777324e-05,
+      "loss": 0.5434,
+      "step": 3856
+    },
+    {
+      "epoch": 0.987392,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.3550891562153405e-05,
+      "loss": 0.4612,
+      "step": 3857
+    },
+    {
+      "epoch": 0.987648,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.354662571452948e-05,
+      "loss": 0.4804,
+      "step": 3858
+    },
+    {
+      "epoch": 0.987904,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.3542359866905554e-05,
+      "loss": 0.485,
+      "step": 3859
+    },
+    {
+      "epoch": 0.98816,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.3538094019281635e-05,
+      "loss": 0.4789,
+      "step": 3860
+    },
+    {
+      "epoch": 0.988416,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.353382817165771e-05,
+      "loss": 0.4731,
+      "step": 3861
+    },
+    {
+      "epoch": 0.988672,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.352956232403378e-05,
+      "loss": 0.4701,
+      "step": 3862
+    },
+    {
+      "epoch": 0.988928,
+      "grad_norm": 2.125,
+      "learning_rate": 3.3525296476409864e-05,
+      "loss": 0.4775,
+      "step": 3863
+    },
+    {
+      "epoch": 0.989184,
+      "grad_norm": 2.125,
+      "learning_rate": 3.3521030628785945e-05,
+      "loss": 0.4985,
+      "step": 3864
+    },
+    {
+      "epoch": 0.98944,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.351676478116202e-05,
+      "loss": 0.4636,
+      "step": 3865
+    },
+    {
+      "epoch": 0.989696,
+      "grad_norm": 1.984375,
+      "learning_rate": 3.3512498933538094e-05,
+      "loss": 0.5203,
+      "step": 3866
+    },
+    {
+      "epoch": 0.989952,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.350823308591417e-05,
+      "loss": 0.4858,
+      "step": 3867
+    },
+    {
+      "epoch": 0.990208,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.350396723829025e-05,
+      "loss": 0.5072,
+      "step": 3868
+    },
+    {
+      "epoch": 0.990464,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.349970139066633e-05,
+      "loss": 0.4969,
+      "step": 3869
+    },
+    {
+      "epoch": 0.99072,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.3495435543042404e-05,
+      "loss": 0.4636,
+      "step": 3870
+    },
+    {
+      "epoch": 0.990976,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.3491169695418485e-05,
+      "loss": 0.4846,
+      "step": 3871
+    },
+    {
+      "epoch": 0.991232,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.348690384779456e-05,
+      "loss": 0.4855,
+      "step": 3872
+    },
+    {
+      "epoch": 0.991488,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.3482638000170634e-05,
+      "loss": 0.5011,
+      "step": 3873
+    },
+    {
+      "epoch": 0.991744,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.347837215254671e-05,
+      "loss": 0.4817,
+      "step": 3874
+    },
+    {
+      "epoch": 0.992,
+      "grad_norm": 2.375,
+      "learning_rate": 3.347410630492279e-05,
+      "loss": 0.4965,
+      "step": 3875
+    },
+    {
+      "epoch": 0.992256,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.346984045729887e-05,
+      "loss": 0.4626,
+      "step": 3876
+    },
+    {
+      "epoch": 0.992512,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.3465574609674945e-05,
+      "loss": 0.4998,
+      "step": 3877
+    },
+    {
+      "epoch": 0.992768,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.346130876205102e-05,
+      "loss": 0.5193,
+      "step": 3878
+    },
+    {
+      "epoch": 0.993024,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.34570429144271e-05,
+      "loss": 0.506,
+      "step": 3879
+    },
+    {
+      "epoch": 0.99328,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.3452777066803174e-05,
+      "loss": 0.5095,
+      "step": 3880
+    },
+    {
+      "epoch": 0.993536,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.344851121917925e-05,
+      "loss": 0.5032,
+      "step": 3881
+    },
+    {
+      "epoch": 0.993792,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.344424537155533e-05,
+      "loss": 0.4633,
+      "step": 3882
+    },
+    {
+      "epoch": 0.994048,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.343997952393141e-05,
+      "loss": 0.47,
+      "step": 3883
+    },
+    {
+      "epoch": 0.994304,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.3435713676307485e-05,
+      "loss": 0.4759,
+      "step": 3884
+    },
+    {
+      "epoch": 0.99456,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.343144782868356e-05,
+      "loss": 0.4414,
+      "step": 3885
+    },
+    {
+      "epoch": 0.994816,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.342718198105963e-05,
+      "loss": 0.4538,
+      "step": 3886
+    },
+    {
+      "epoch": 0.995072,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.3422916133435714e-05,
+      "loss": 0.47,
+      "step": 3887
+    },
+    {
+      "epoch": 0.995328,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.3418650285811795e-05,
+      "loss": 0.482,
+      "step": 3888
+    },
+    {
+      "epoch": 0.995584,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.341438443818787e-05,
+      "loss": 0.4839,
+      "step": 3889
+    },
+    {
+      "epoch": 0.99584,
+      "grad_norm": 2.578125,
+      "learning_rate": 3.341011859056395e-05,
+      "loss": 0.4781,
+      "step": 3890
+    },
+    {
+      "epoch": 0.996096,
+      "grad_norm": 2.5,
+      "learning_rate": 3.3405852742940025e-05,
+      "loss": 0.4726,
+      "step": 3891
+    },
+    {
+      "epoch": 0.996352,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.34015868953161e-05,
+      "loss": 0.4316,
+      "step": 3892
+    },
+    {
+      "epoch": 0.996608,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.3397321047692174e-05,
+      "loss": 0.4763,
+      "step": 3893
+    },
+    {
+      "epoch": 0.996864,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.3393055200068255e-05,
+      "loss": 0.5042,
+      "step": 3894
+    },
+    {
+      "epoch": 0.99712,
+      "grad_norm": 2.609375,
+      "learning_rate": 3.3388789352444336e-05,
+      "loss": 0.5116,
+      "step": 3895
+    },
+    {
+      "epoch": 0.997376,
+      "grad_norm": 2.375,
+      "learning_rate": 3.338452350482041e-05,
+      "loss": 0.4846,
+      "step": 3896
+    },
+    {
+      "epoch": 0.997632,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.3380257657196484e-05,
+      "loss": 0.4934,
+      "step": 3897
+    },
+    {
+      "epoch": 0.997888,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.3375991809572565e-05,
+      "loss": 0.4689,
+      "step": 3898
+    },
+    {
+      "epoch": 0.998144,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.337172596194864e-05,
+      "loss": 0.5356,
+      "step": 3899
+    },
+    {
+      "epoch": 0.9984,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.3367460114324714e-05,
+      "loss": 0.4998,
+      "step": 3900
+    },
+    {
+      "epoch": 0.998656,
+      "grad_norm": 1.859375,
+      "learning_rate": 3.3363194266700795e-05,
+      "loss": 0.4831,
+      "step": 3901
+    },
+    {
+      "epoch": 0.998912,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.3358928419076876e-05,
+      "loss": 0.4674,
+      "step": 3902
+    },
+    {
+      "epoch": 0.999168,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.335466257145295e-05,
+      "loss": 0.4598,
+      "step": 3903
+    },
+    {
+      "epoch": 0.999424,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.3350396723829024e-05,
+      "loss": 0.4828,
+      "step": 3904
+    },
+    {
+      "epoch": 0.99968,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.3346130876205105e-05,
+      "loss": 0.4691,
+      "step": 3905
+    },
+    {
+      "epoch": 0.999936,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.334186502858118e-05,
+      "loss": 0.4696,
+      "step": 3906
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.796875,
+      "learning_rate": 3.333759918095726e-05,
+      "loss": 0.5325,
+      "step": 3907
+    },
+    {
+      "epoch": 1.000256,
+      "grad_norm": 2.6875,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.4657,
+      "step": 3908
+    },
+    {
+      "epoch": 1.000512,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.3329067485709416e-05,
+      "loss": 0.4815,
+      "step": 3909
+    },
+    {
+      "epoch": 1.000768,
+      "grad_norm": 2.125,
+      "learning_rate": 3.332480163808549e-05,
+      "loss": 0.4406,
+      "step": 3910
+    },
+    {
+      "epoch": 1.001024,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.3320535790461565e-05,
+      "loss": 0.4588,
+      "step": 3911
+    },
+    {
+      "epoch": 1.00128,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.331626994283764e-05,
+      "loss": 0.4907,
+      "step": 3912
+    },
+    {
+      "epoch": 1.001536,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.331200409521372e-05,
+      "loss": 0.4992,
+      "step": 3913
+    },
+    {
+      "epoch": 1.001792,
+      "grad_norm": 2.375,
+      "learning_rate": 3.33077382475898e-05,
+      "loss": 0.4587,
+      "step": 3914
+    },
+    {
+      "epoch": 1.002048,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.3303472399965875e-05,
+      "loss": 0.4451,
+      "step": 3915
+    },
+    {
+      "epoch": 1.002304,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.329920655234195e-05,
+      "loss": 0.5152,
+      "step": 3916
+    },
+    {
+      "epoch": 1.00256,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.329494070471803e-05,
+      "loss": 0.462,
+      "step": 3917
+    },
+    {
+      "epoch": 1.002816,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.3290674857094105e-05,
+      "loss": 0.4751,
+      "step": 3918
+    },
+    {
+      "epoch": 1.003072,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.328640900947018e-05,
+      "loss": 0.5172,
+      "step": 3919
+    },
+    {
+      "epoch": 1.003328,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.328214316184626e-05,
+      "loss": 0.4762,
+      "step": 3920
+    },
+    {
+      "epoch": 1.003584,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.327787731422234e-05,
+      "loss": 0.4734,
+      "step": 3921
+    },
+    {
+      "epoch": 1.00384,
+      "grad_norm": 2.125,
+      "learning_rate": 3.3273611466598415e-05,
+      "loss": 0.4871,
+      "step": 3922
+    },
+    {
+      "epoch": 1.004096,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.326934561897449e-05,
+      "loss": 0.4794,
+      "step": 3923
+    },
+    {
+      "epoch": 1.004352,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.326507977135057e-05,
+      "loss": 0.4742,
+      "step": 3924
+    },
+    {
+      "epoch": 1.004608,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.3260813923726645e-05,
+      "loss": 0.484,
+      "step": 3925
+    },
+    {
+      "epoch": 1.004864,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.3256548076102726e-05,
+      "loss": 0.4534,
+      "step": 3926
+    },
+    {
+      "epoch": 1.00512,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.32522822284788e-05,
+      "loss": 0.4463,
+      "step": 3927
+    },
+    {
+      "epoch": 1.005376,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.324801638085488e-05,
+      "loss": 0.4457,
+      "step": 3928
+    },
+    {
+      "epoch": 1.005632,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.3243750533230956e-05,
+      "loss": 0.4717,
+      "step": 3929
+    },
+    {
+      "epoch": 1.005888,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.323948468560703e-05,
+      "loss": 0.5274,
+      "step": 3930
+    },
+    {
+      "epoch": 1.006144,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.3235218837983104e-05,
+      "loss": 0.4778,
+      "step": 3931
+    },
+    {
+      "epoch": 1.0064,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.3230952990359185e-05,
+      "loss": 0.4759,
+      "step": 3932
+    },
+    {
+      "epoch": 1.006656,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.3226687142735266e-05,
+      "loss": 0.4742,
+      "step": 3933
+    },
+    {
+      "epoch": 1.006912,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.322242129511134e-05,
+      "loss": 0.4716,
+      "step": 3934
+    },
+    {
+      "epoch": 1.007168,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.3218155447487415e-05,
+      "loss": 0.4383,
+      "step": 3935
+    },
+    {
+      "epoch": 1.007424,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.3213889599863496e-05,
+      "loss": 0.4831,
+      "step": 3936
+    },
+    {
+      "epoch": 1.00768,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.320962375223957e-05,
+      "loss": 0.4801,
+      "step": 3937
+    },
+    {
+      "epoch": 1.007936,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.3205357904615644e-05,
+      "loss": 0.4888,
+      "step": 3938
+    },
+    {
+      "epoch": 1.008192,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.3201092056991725e-05,
+      "loss": 0.4626,
+      "step": 3939
+    },
+    {
+      "epoch": 1.008448,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.3196826209367806e-05,
+      "loss": 0.4484,
+      "step": 3940
+    },
+    {
+      "epoch": 1.008704,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.319256036174388e-05,
+      "loss": 0.4841,
+      "step": 3941
+    },
+    {
+      "epoch": 1.00896,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.3188294514119955e-05,
+      "loss": 0.4817,
+      "step": 3942
+    },
+    {
+      "epoch": 1.009216,
+      "grad_norm": 2.5,
+      "learning_rate": 3.3184028666496036e-05,
+      "loss": 0.4823,
+      "step": 3943
+    },
+    {
+      "epoch": 1.009472,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.317976281887211e-05,
+      "loss": 0.4619,
+      "step": 3944
+    },
+    {
+      "epoch": 1.009728,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.3175496971248184e-05,
+      "loss": 0.4719,
+      "step": 3945
+    },
+    {
+      "epoch": 1.009984,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.3171231123624265e-05,
+      "loss": 0.5123,
+      "step": 3946
+    },
+    {
+      "epoch": 1.01024,
+      "grad_norm": 2.375,
+      "learning_rate": 3.3166965276000347e-05,
+      "loss": 0.4368,
+      "step": 3947
+    },
+    {
+      "epoch": 1.010496,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.316269942837642e-05,
+      "loss": 0.4847,
+      "step": 3948
+    },
+    {
+      "epoch": 1.010752,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.3158433580752495e-05,
+      "loss": 0.4461,
+      "step": 3949
+    },
+    {
+      "epoch": 1.011008,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.315416773312857e-05,
+      "loss": 0.4814,
+      "step": 3950
+    },
+    {
+      "epoch": 1.011264,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.314990188550465e-05,
+      "loss": 0.4989,
+      "step": 3951
+    },
+    {
+      "epoch": 1.01152,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.314563603788073e-05,
+      "loss": 0.4682,
+      "step": 3952
+    },
+    {
+      "epoch": 1.011776,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.3141370190256806e-05,
+      "loss": 0.4824,
+      "step": 3953
+    },
+    {
+      "epoch": 1.012032,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.313710434263289e-05,
+      "loss": 0.5049,
+      "step": 3954
+    },
+    {
+      "epoch": 1.012288,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.313283849500896e-05,
+      "loss": 0.4902,
+      "step": 3955
+    },
+    {
+      "epoch": 1.012544,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.3128572647385035e-05,
+      "loss": 0.493,
+      "step": 3956
+    },
+    {
+      "epoch": 1.0128,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.312430679976111e-05,
+      "loss": 0.4996,
+      "step": 3957
+    },
+    {
+      "epoch": 1.013056,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.312004095213719e-05,
+      "loss": 0.5052,
+      "step": 3958
+    },
+    {
+      "epoch": 1.013312,
+      "grad_norm": 2.125,
+      "learning_rate": 3.311577510451327e-05,
+      "loss": 0.4567,
+      "step": 3959
+    },
+    {
+      "epoch": 1.013568,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.3111509256889346e-05,
+      "loss": 0.4948,
+      "step": 3960
+    },
+    {
+      "epoch": 1.013824,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.310724340926542e-05,
+      "loss": 0.4617,
+      "step": 3961
+    },
+    {
+      "epoch": 1.01408,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.31029775616415e-05,
+      "loss": 0.4001,
+      "step": 3962
+    },
+    {
+      "epoch": 1.014336,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.3098711714017575e-05,
+      "loss": 0.4776,
+      "step": 3963
+    },
+    {
+      "epoch": 1.014592,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.309444586639365e-05,
+      "loss": 0.4942,
+      "step": 3964
+    },
+    {
+      "epoch": 1.014848,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.309018001876973e-05,
+      "loss": 0.4559,
+      "step": 3965
+    },
+    {
+      "epoch": 1.015104,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.308591417114581e-05,
+      "loss": 0.507,
+      "step": 3966
+    },
+    {
+      "epoch": 1.01536,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.3081648323521886e-05,
+      "loss": 0.4754,
+      "step": 3967
+    },
+    {
+      "epoch": 1.015616,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.307738247589796e-05,
+      "loss": 0.4926,
+      "step": 3968
+    },
+    {
+      "epoch": 1.015872,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.307311662827404e-05,
+      "loss": 0.4771,
+      "step": 3969
+    },
+    {
+      "epoch": 1.016128,
+      "grad_norm": 2.375,
+      "learning_rate": 3.3068850780650116e-05,
+      "loss": 0.4628,
+      "step": 3970
+    },
+    {
+      "epoch": 1.016384,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.30645849330262e-05,
+      "loss": 0.4804,
+      "step": 3971
+    },
+    {
+      "epoch": 1.01664,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.306031908540227e-05,
+      "loss": 0.4672,
+      "step": 3972
+    },
+    {
+      "epoch": 1.016896,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.305605323777835e-05,
+      "loss": 0.4484,
+      "step": 3973
+    },
+    {
+      "epoch": 1.017152,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.3051787390154426e-05,
+      "loss": 0.502,
+      "step": 3974
+    },
+    {
+      "epoch": 1.017408,
+      "grad_norm": 2.375,
+      "learning_rate": 3.30475215425305e-05,
+      "loss": 0.4717,
+      "step": 3975
+    },
+    {
+      "epoch": 1.017664,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.3043255694906575e-05,
+      "loss": 0.5019,
+      "step": 3976
+    },
+    {
+      "epoch": 1.01792,
+      "grad_norm": 2.640625,
+      "learning_rate": 3.3038989847282656e-05,
+      "loss": 0.4799,
+      "step": 3977
+    },
+    {
+      "epoch": 1.018176,
+      "grad_norm": 2.578125,
+      "learning_rate": 3.303472399965874e-05,
+      "loss": 0.5038,
+      "step": 3978
+    },
+    {
+      "epoch": 1.018432,
+      "grad_norm": 2.125,
+      "learning_rate": 3.303045815203481e-05,
+      "loss": 0.4771,
+      "step": 3979
+    },
+    {
+      "epoch": 1.018688,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.3026192304410885e-05,
+      "loss": 0.4914,
+      "step": 3980
+    },
+    {
+      "epoch": 1.018944,
+      "grad_norm": 2.625,
+      "learning_rate": 3.3021926456786966e-05,
+      "loss": 0.4941,
+      "step": 3981
+    },
+    {
+      "epoch": 1.0192,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.301766060916304e-05,
+      "loss": 0.5168,
+      "step": 3982
+    },
+    {
+      "epoch": 1.019456,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.3013394761539115e-05,
+      "loss": 0.444,
+      "step": 3983
+    },
+    {
+      "epoch": 1.019712,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.3009128913915196e-05,
+      "loss": 0.4627,
+      "step": 3984
+    },
+    {
+      "epoch": 1.019968,
+      "grad_norm": 2.578125,
+      "learning_rate": 3.300486306629128e-05,
+      "loss": 0.4529,
+      "step": 3985
+    },
+    {
+      "epoch": 1.020224,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.300059721866735e-05,
+      "loss": 0.4949,
+      "step": 3986
+    },
+    {
+      "epoch": 1.02048,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.2996331371043426e-05,
+      "loss": 0.4889,
+      "step": 3987
+    },
+    {
+      "epoch": 1.020736,
+      "grad_norm": 2.125,
+      "learning_rate": 3.299206552341951e-05,
+      "loss": 0.5128,
+      "step": 3988
+    },
+    {
+      "epoch": 1.020992,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.298779967579558e-05,
+      "loss": 0.4913,
+      "step": 3989
+    },
+    {
+      "epoch": 1.021248,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.298353382817166e-05,
+      "loss": 0.4654,
+      "step": 3990
+    },
+    {
+      "epoch": 1.021504,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.2979267980547736e-05,
+      "loss": 0.4611,
+      "step": 3991
+    },
+    {
+      "epoch": 1.02176,
+      "grad_norm": 2.25,
+      "learning_rate": 3.297500213292382e-05,
+      "loss": 0.454,
+      "step": 3992
+    },
+    {
+      "epoch": 1.022016,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.297073628529989e-05,
+      "loss": 0.4539,
+      "step": 3993
+    },
+    {
+      "epoch": 1.022272,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.2966470437675966e-05,
+      "loss": 0.4989,
+      "step": 3994
+    },
+    {
+      "epoch": 1.022528,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.296220459005204e-05,
+      "loss": 0.4841,
+      "step": 3995
+    },
+    {
+      "epoch": 1.022784,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.295793874242812e-05,
+      "loss": 0.4903,
+      "step": 3996
+    },
+    {
+      "epoch": 1.02304,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.29536728948042e-05,
+      "loss": 0.4932,
+      "step": 3997
+    },
+    {
+      "epoch": 1.023296,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.2949407047180276e-05,
+      "loss": 0.4366,
+      "step": 3998
+    },
+    {
+      "epoch": 1.023552,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.294514119955635e-05,
+      "loss": 0.5043,
+      "step": 3999
+    },
+    {
+      "epoch": 1.023808,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.294087535193243e-05,
+      "loss": 0.4894,
+      "step": 4000
+    },
+    {
+      "epoch": 1.024064,
+      "grad_norm": 2.25,
+      "learning_rate": 3.2936609504308506e-05,
+      "loss": 0.5011,
+      "step": 4001
+    },
+    {
+      "epoch": 1.02432,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.293234365668458e-05,
+      "loss": 0.5308,
+      "step": 4002
+    },
+    {
+      "epoch": 1.024576,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.292807780906066e-05,
+      "loss": 0.4527,
+      "step": 4003
+    },
+    {
+      "epoch": 1.024832,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.292381196143674e-05,
+      "loss": 0.4964,
+      "step": 4004
+    },
+    {
+      "epoch": 1.025088,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.2919546113812817e-05,
+      "loss": 0.5044,
+      "step": 4005
+    },
+    {
+      "epoch": 1.025344,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.291528026618889e-05,
+      "loss": 0.4851,
+      "step": 4006
+    },
+    {
+      "epoch": 1.0256,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.291101441856497e-05,
+      "loss": 0.4811,
+      "step": 4007
+    },
+    {
+      "epoch": 1.025856,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.2906748570941046e-05,
+      "loss": 0.4755,
+      "step": 4008
+    },
+    {
+      "epoch": 1.026112,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.290248272331713e-05,
+      "loss": 0.4608,
+      "step": 4009
+    },
+    {
+      "epoch": 1.026368,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.28982168756932e-05,
+      "loss": 0.5113,
+      "step": 4010
+    },
+    {
+      "epoch": 1.026624,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.289395102806928e-05,
+      "loss": 0.4266,
+      "step": 4011
+    },
+    {
+      "epoch": 1.02688,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.288968518044536e-05,
+      "loss": 0.4647,
+      "step": 4012
+    },
+    {
+      "epoch": 1.027136,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.288541933282143e-05,
+      "loss": 0.4685,
+      "step": 4013
+    },
+    {
+      "epoch": 1.027392,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.2881153485197505e-05,
+      "loss": 0.4699,
+      "step": 4014
+    },
+    {
+      "epoch": 1.027648,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.2876887637573586e-05,
+      "loss": 0.4855,
+      "step": 4015
+    },
+    {
+      "epoch": 1.027904,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.287262178994967e-05,
+      "loss": 0.5065,
+      "step": 4016
+    },
+    {
+      "epoch": 1.02816,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.286835594232574e-05,
+      "loss": 0.4565,
+      "step": 4017
+    },
+    {
+      "epoch": 1.028416,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.286409009470182e-05,
+      "loss": 0.4741,
+      "step": 4018
+    },
+    {
+      "epoch": 1.028672,
+      "grad_norm": 2.25,
+      "learning_rate": 3.28598242470779e-05,
+      "loss": 0.456,
+      "step": 4019
+    },
+    {
+      "epoch": 1.028928,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.285555839945397e-05,
+      "loss": 0.4745,
+      "step": 4020
+    },
+    {
+      "epoch": 1.029184,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.2851292551830046e-05,
+      "loss": 0.5134,
+      "step": 4021
+    },
+    {
+      "epoch": 1.02944,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.2847026704206127e-05,
+      "loss": 0.4863,
+      "step": 4022
+    },
+    {
+      "epoch": 1.029696,
+      "grad_norm": 1.984375,
+      "learning_rate": 3.284276085658221e-05,
+      "loss": 0.4665,
+      "step": 4023
+    },
+    {
+      "epoch": 1.029952,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.283849500895828e-05,
+      "loss": 0.4677,
+      "step": 4024
+    },
+    {
+      "epoch": 1.030208,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.2834229161334356e-05,
+      "loss": 0.5026,
+      "step": 4025
+    },
+    {
+      "epoch": 1.030464,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.282996331371044e-05,
+      "loss": 0.4866,
+      "step": 4026
+    },
+    {
+      "epoch": 1.03072,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.282569746608651e-05,
+      "loss": 0.4832,
+      "step": 4027
+    },
+    {
+      "epoch": 1.030976,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.282143161846259e-05,
+      "loss": 0.4981,
+      "step": 4028
+    },
+    {
+      "epoch": 1.031232,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.281716577083867e-05,
+      "loss": 0.4963,
+      "step": 4029
+    },
+    {
+      "epoch": 1.031488,
+      "grad_norm": 2.375,
+      "learning_rate": 3.281289992321475e-05,
+      "loss": 0.4653,
+      "step": 4030
+    },
+    {
+      "epoch": 1.031744,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.280863407559082e-05,
+      "loss": 0.4768,
+      "step": 4031
+    },
+    {
+      "epoch": 1.032,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.2804368227966896e-05,
+      "loss": 0.4767,
+      "step": 4032
+    },
+    {
+      "epoch": 1.032256,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.280010238034297e-05,
+      "loss": 0.4891,
+      "step": 4033
+    },
+    {
+      "epoch": 1.032512,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.279583653271905e-05,
+      "loss": 0.4471,
+      "step": 4034
+    },
+    {
+      "epoch": 1.032768,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.279157068509513e-05,
+      "loss": 0.4918,
+      "step": 4035
+    },
+    {
+      "epoch": 1.033024,
+      "grad_norm": 1.9921875,
+      "learning_rate": 3.278730483747121e-05,
+      "loss": 0.4994,
+      "step": 4036
+    },
+    {
+      "epoch": 1.03328,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.278303898984729e-05,
+      "loss": 0.4551,
+      "step": 4037
+    },
+    {
+      "epoch": 1.033536,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.277877314222336e-05,
+      "loss": 0.4842,
+      "step": 4038
+    },
+    {
+      "epoch": 1.033792,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.2774507294599437e-05,
+      "loss": 0.4921,
+      "step": 4039
+    },
+    {
+      "epoch": 1.034048,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.277024144697551e-05,
+      "loss": 0.4605,
+      "step": 4040
+    },
+    {
+      "epoch": 1.0343040000000001,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.276597559935159e-05,
+      "loss": 0.4954,
+      "step": 4041
+    },
+    {
+      "epoch": 1.03456,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.276170975172767e-05,
+      "loss": 0.4659,
+      "step": 4042
+    },
+    {
+      "epoch": 1.034816,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.275744390410375e-05,
+      "loss": 0.4622,
+      "step": 4043
+    },
+    {
+      "epoch": 1.035072,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.275317805647982e-05,
+      "loss": 0.4688,
+      "step": 4044
+    },
+    {
+      "epoch": 1.035328,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.27489122088559e-05,
+      "loss": 0.504,
+      "step": 4045
+    },
+    {
+      "epoch": 1.035584,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.274464636123198e-05,
+      "loss": 0.4886,
+      "step": 4046
+    },
+    {
+      "epoch": 1.03584,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.274038051360806e-05,
+      "loss": 0.4798,
+      "step": 4047
+    },
+    {
+      "epoch": 1.036096,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.273611466598413e-05,
+      "loss": 0.4838,
+      "step": 4048
+    },
+    {
+      "epoch": 1.036352,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.273184881836021e-05,
+      "loss": 0.4894,
+      "step": 4049
+    },
+    {
+      "epoch": 1.036608,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.272758297073629e-05,
+      "loss": 0.5009,
+      "step": 4050
+    },
+    {
+      "epoch": 1.036864,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.272331712311236e-05,
+      "loss": 0.4502,
+      "step": 4051
+    },
+    {
+      "epoch": 1.03712,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.271905127548844e-05,
+      "loss": 0.4493,
+      "step": 4052
+    },
+    {
+      "epoch": 1.037376,
+      "grad_norm": 2.125,
+      "learning_rate": 3.271478542786452e-05,
+      "loss": 0.4677,
+      "step": 4053
+    },
+    {
+      "epoch": 1.0376319999999999,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.27105195802406e-05,
+      "loss": 0.4414,
+      "step": 4054
+    },
+    {
+      "epoch": 1.037888,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.270625373261667e-05,
+      "loss": 0.4813,
+      "step": 4055
+    },
+    {
+      "epoch": 1.038144,
+      "grad_norm": 2.375,
+      "learning_rate": 3.270198788499275e-05,
+      "loss": 0.4728,
+      "step": 4056
+    },
+    {
+      "epoch": 1.0384,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.269772203736883e-05,
+      "loss": 0.457,
+      "step": 4057
+    },
+    {
+      "epoch": 1.038656,
+      "grad_norm": 2.25,
+      "learning_rate": 3.26934561897449e-05,
+      "loss": 0.4593,
+      "step": 4058
+    },
+    {
+      "epoch": 1.038912,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.2689190342120976e-05,
+      "loss": 0.4917,
+      "step": 4059
+    },
+    {
+      "epoch": 1.039168,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.268492449449706e-05,
+      "loss": 0.4853,
+      "step": 4060
+    },
+    {
+      "epoch": 1.039424,
+      "grad_norm": 2.125,
+      "learning_rate": 3.268065864687314e-05,
+      "loss": 0.4259,
+      "step": 4061
+    },
+    {
+      "epoch": 1.03968,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.267639279924921e-05,
+      "loss": 0.458,
+      "step": 4062
+    },
+    {
+      "epoch": 1.039936,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.267212695162529e-05,
+      "loss": 0.4926,
+      "step": 4063
+    },
+    {
+      "epoch": 1.040192,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.266786110400137e-05,
+      "loss": 0.4774,
+      "step": 4064
+    },
+    {
+      "epoch": 1.040448,
+      "grad_norm": 2.25,
+      "learning_rate": 3.266359525637744e-05,
+      "loss": 0.4664,
+      "step": 4065
+    },
+    {
+      "epoch": 1.040704,
+      "grad_norm": 2.5,
+      "learning_rate": 3.2659329408753516e-05,
+      "loss": 0.417,
+      "step": 4066
+    },
+    {
+      "epoch": 1.04096,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.2655063561129604e-05,
+      "loss": 0.4771,
+      "step": 4067
+    },
+    {
+      "epoch": 1.041216,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.265079771350568e-05,
+      "loss": 0.4275,
+      "step": 4068
+    },
+    {
+      "epoch": 1.041472,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.264653186588175e-05,
+      "loss": 0.4745,
+      "step": 4069
+    },
+    {
+      "epoch": 1.041728,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.264226601825783e-05,
+      "loss": 0.4669,
+      "step": 4070
+    },
+    {
+      "epoch": 1.041984,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.263800017063391e-05,
+      "loss": 0.4821,
+      "step": 4071
+    },
+    {
+      "epoch": 1.04224,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.263373432300998e-05,
+      "loss": 0.441,
+      "step": 4072
+    },
+    {
+      "epoch": 1.042496,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.262946847538606e-05,
+      "loss": 0.527,
+      "step": 4073
+    },
+    {
+      "epoch": 1.042752,
+      "grad_norm": 2.125,
+      "learning_rate": 3.262520262776214e-05,
+      "loss": 0.4939,
+      "step": 4074
+    },
+    {
+      "epoch": 1.043008,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.262093678013822e-05,
+      "loss": 0.4754,
+      "step": 4075
+    },
+    {
+      "epoch": 1.043264,
+      "grad_norm": 2.25,
+      "learning_rate": 3.261667093251429e-05,
+      "loss": 0.4763,
+      "step": 4076
+    },
+    {
+      "epoch": 1.04352,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.261240508489037e-05,
+      "loss": 0.4743,
+      "step": 4077
+    },
+    {
+      "epoch": 1.043776,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.260813923726644e-05,
+      "loss": 0.4596,
+      "step": 4078
+    },
+    {
+      "epoch": 1.044032,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.260387338964252e-05,
+      "loss": 0.4653,
+      "step": 4079
+    },
+    {
+      "epoch": 1.044288,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.2599607542018603e-05,
+      "loss": 0.4975,
+      "step": 4080
+    },
+    {
+      "epoch": 1.044544,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.259534169439468e-05,
+      "loss": 0.4456,
+      "step": 4081
+    },
+    {
+      "epoch": 1.0448,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.259107584677075e-05,
+      "loss": 0.4485,
+      "step": 4082
+    },
+    {
+      "epoch": 1.045056,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.258680999914683e-05,
+      "loss": 0.4417,
+      "step": 4083
+    },
+    {
+      "epoch": 1.045312,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.258254415152291e-05,
+      "loss": 0.4892,
+      "step": 4084
+    },
+    {
+      "epoch": 1.045568,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.257827830389898e-05,
+      "loss": 0.4617,
+      "step": 4085
+    },
+    {
+      "epoch": 1.045824,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.257401245627506e-05,
+      "loss": 0.4492,
+      "step": 4086
+    },
+    {
+      "epoch": 1.04608,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.2569746608651144e-05,
+      "loss": 0.4652,
+      "step": 4087
+    },
+    {
+      "epoch": 1.046336,
+      "grad_norm": 2.5,
+      "learning_rate": 3.256548076102722e-05,
+      "loss": 0.472,
+      "step": 4088
+    },
+    {
+      "epoch": 1.046592,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.256121491340329e-05,
+      "loss": 0.5045,
+      "step": 4089
+    },
+    {
+      "epoch": 1.046848,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.255694906577937e-05,
+      "loss": 0.4921,
+      "step": 4090
+    },
+    {
+      "epoch": 1.047104,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.255268321815545e-05,
+      "loss": 0.4685,
+      "step": 4091
+    },
+    {
+      "epoch": 1.04736,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.254841737053153e-05,
+      "loss": 0.4904,
+      "step": 4092
+    },
+    {
+      "epoch": 1.047616,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.25441515229076e-05,
+      "loss": 0.503,
+      "step": 4093
+    },
+    {
+      "epoch": 1.047872,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.2539885675283684e-05,
+      "loss": 0.4577,
+      "step": 4094
+    },
+    {
+      "epoch": 1.048128,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.253561982765976e-05,
+      "loss": 0.4567,
+      "step": 4095
+    },
+    {
+      "epoch": 1.048384,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.253135398003583e-05,
+      "loss": 0.4622,
+      "step": 4096
+    },
+    {
+      "epoch": 1.04864,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.2527088132411907e-05,
+      "loss": 0.4745,
+      "step": 4097
+    },
+    {
+      "epoch": 1.048896,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.252282228478799e-05,
+      "loss": 0.5326,
+      "step": 4098
+    },
+    {
+      "epoch": 1.049152,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.251855643716407e-05,
+      "loss": 0.4751,
+      "step": 4099
+    },
+    {
+      "epoch": 1.049408,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.251429058954014e-05,
+      "loss": 0.478,
+      "step": 4100
+    },
+    {
+      "epoch": 1.049664,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.2510024741916224e-05,
+      "loss": 0.4804,
+      "step": 4101
+    },
+    {
+      "epoch": 1.04992,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.25057588942923e-05,
+      "loss": 0.4936,
+      "step": 4102
+    },
+    {
+      "epoch": 1.050176,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.250149304666837e-05,
+      "loss": 0.5166,
+      "step": 4103
+    },
+    {
+      "epoch": 1.050432,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.249722719904445e-05,
+      "loss": 0.4789,
+      "step": 4104
+    },
+    {
+      "epoch": 1.050688,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.249296135142053e-05,
+      "loss": 0.4907,
+      "step": 4105
+    },
+    {
+      "epoch": 1.050944,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.248869550379661e-05,
+      "loss": 0.4461,
+      "step": 4106
+    },
+    {
+      "epoch": 1.0512,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.248442965617268e-05,
+      "loss": 0.4863,
+      "step": 4107
+    },
+    {
+      "epoch": 1.051456,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.248016380854876e-05,
+      "loss": 0.4612,
+      "step": 4108
+    },
+    {
+      "epoch": 1.051712,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.247589796092484e-05,
+      "loss": 0.468,
+      "step": 4109
+    },
+    {
+      "epoch": 1.051968,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.247163211330091e-05,
+      "loss": 0.4787,
+      "step": 4110
+    },
+    {
+      "epoch": 1.052224,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.2467366265676994e-05,
+      "loss": 0.4613,
+      "step": 4111
+    },
+    {
+      "epoch": 1.05248,
+      "grad_norm": 2.375,
+      "learning_rate": 3.246310041805307e-05,
+      "loss": 0.4828,
+      "step": 4112
+    },
+    {
+      "epoch": 1.052736,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.245883457042915e-05,
+      "loss": 0.4854,
+      "step": 4113
+    },
+    {
+      "epoch": 1.052992,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.245456872280522e-05,
+      "loss": 0.4717,
+      "step": 4114
+    },
+    {
+      "epoch": 1.053248,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.24503028751813e-05,
+      "loss": 0.4833,
+      "step": 4115
+    },
+    {
+      "epoch": 1.053504,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.244603702755738e-05,
+      "loss": 0.45,
+      "step": 4116
+    },
+    {
+      "epoch": 1.05376,
+      "grad_norm": 2.125,
+      "learning_rate": 3.244177117993345e-05,
+      "loss": 0.4764,
+      "step": 4117
+    },
+    {
+      "epoch": 1.054016,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.2437505332309534e-05,
+      "loss": 0.4217,
+      "step": 4118
+    },
+    {
+      "epoch": 1.054272,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.243323948468561e-05,
+      "loss": 0.4664,
+      "step": 4119
+    },
+    {
+      "epoch": 1.054528,
+      "grad_norm": 2.25,
+      "learning_rate": 3.242897363706169e-05,
+      "loss": 0.4817,
+      "step": 4120
+    },
+    {
+      "epoch": 1.054784,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.2424707789437764e-05,
+      "loss": 0.4788,
+      "step": 4121
+    },
+    {
+      "epoch": 1.05504,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.242044194181384e-05,
+      "loss": 0.4592,
+      "step": 4122
+    },
+    {
+      "epoch": 1.055296,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.241617609418991e-05,
+      "loss": 0.4856,
+      "step": 4123
+    },
+    {
+      "epoch": 1.055552,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.241191024656599e-05,
+      "loss": 0.4621,
+      "step": 4124
+    },
+    {
+      "epoch": 1.055808,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.2407644398942074e-05,
+      "loss": 0.437,
+      "step": 4125
+    },
+    {
+      "epoch": 1.0560640000000001,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.240337855131815e-05,
+      "loss": 0.4724,
+      "step": 4126
+    },
+    {
+      "epoch": 1.05632,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.239911270369422e-05,
+      "loss": 0.4417,
+      "step": 4127
+    },
+    {
+      "epoch": 1.056576,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.2394846856070304e-05,
+      "loss": 0.4609,
+      "step": 4128
+    },
+    {
+      "epoch": 1.056832,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.239058100844638e-05,
+      "loss": 0.4591,
+      "step": 4129
+    },
+    {
+      "epoch": 1.057088,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.238631516082246e-05,
+      "loss": 0.4419,
+      "step": 4130
+    },
+    {
+      "epoch": 1.057344,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.238204931319853e-05,
+      "loss": 0.494,
+      "step": 4131
+    },
+    {
+      "epoch": 1.0576,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.2377783465574614e-05,
+      "loss": 0.4616,
+      "step": 4132
+    },
+    {
+      "epoch": 1.057856,
+      "grad_norm": 2.125,
+      "learning_rate": 3.237351761795069e-05,
+      "loss": 0.4669,
+      "step": 4133
+    },
+    {
+      "epoch": 1.058112,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.236925177032676e-05,
+      "loss": 0.48,
+      "step": 4134
+    },
+    {
+      "epoch": 1.058368,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.2364985922702844e-05,
+      "loss": 0.4898,
+      "step": 4135
+    },
+    {
+      "epoch": 1.058624,
+      "grad_norm": 2.375,
+      "learning_rate": 3.236072007507892e-05,
+      "loss": 0.4835,
+      "step": 4136
+    },
+    {
+      "epoch": 1.05888,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.2356454227455e-05,
+      "loss": 0.4857,
+      "step": 4137
+    },
+    {
+      "epoch": 1.059136,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.2352188379831073e-05,
+      "loss": 0.4665,
+      "step": 4138
+    },
+    {
+      "epoch": 1.059392,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.2347922532207155e-05,
+      "loss": 0.4922,
+      "step": 4139
+    },
+    {
+      "epoch": 1.059648,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.234365668458323e-05,
+      "loss": 0.4852,
+      "step": 4140
+    },
+    {
+      "epoch": 1.059904,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.23393908369593e-05,
+      "loss": 0.4885,
+      "step": 4141
+    },
+    {
+      "epoch": 1.06016,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.233512498933538e-05,
+      "loss": 0.4283,
+      "step": 4142
+    },
+    {
+      "epoch": 1.060416,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.233085914171146e-05,
+      "loss": 0.4794,
+      "step": 4143
+    },
+    {
+      "epoch": 1.060672,
+      "grad_norm": 2.25,
+      "learning_rate": 3.232659329408754e-05,
+      "loss": 0.4792,
+      "step": 4144
+    },
+    {
+      "epoch": 1.060928,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.2322327446463614e-05,
+      "loss": 0.4805,
+      "step": 4145
+    },
+    {
+      "epoch": 1.061184,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.231806159883969e-05,
+      "loss": 0.4682,
+      "step": 4146
+    },
+    {
+      "epoch": 1.06144,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.231379575121577e-05,
+      "loss": 0.464,
+      "step": 4147
+    },
+    {
+      "epoch": 1.061696,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.230952990359184e-05,
+      "loss": 0.4891,
+      "step": 4148
+    },
+    {
+      "epoch": 1.061952,
+      "grad_norm": 2.65625,
+      "learning_rate": 3.2305264055967924e-05,
+      "loss": 0.4387,
+      "step": 4149
+    },
+    {
+      "epoch": 1.062208,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.2300998208344005e-05,
+      "loss": 0.461,
+      "step": 4150
+    },
+    {
+      "epoch": 1.062464,
+      "grad_norm": 2.5,
+      "learning_rate": 3.229673236072008e-05,
+      "loss": 0.4525,
+      "step": 4151
+    },
+    {
+      "epoch": 1.06272,
+      "grad_norm": 2.375,
+      "learning_rate": 3.2292466513096154e-05,
+      "loss": 0.469,
+      "step": 4152
+    },
+    {
+      "epoch": 1.062976,
+      "grad_norm": 2.59375,
+      "learning_rate": 3.228820066547223e-05,
+      "loss": 0.4901,
+      "step": 4153
+    },
+    {
+      "epoch": 1.063232,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.228393481784831e-05,
+      "loss": 0.4456,
+      "step": 4154
+    },
+    {
+      "epoch": 1.063488,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.2279668970224383e-05,
+      "loss": 0.4602,
+      "step": 4155
+    },
+    {
+      "epoch": 1.063744,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.2275403122600464e-05,
+      "loss": 0.4443,
+      "step": 4156
+    },
+    {
+      "epoch": 1.064,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.227113727497654e-05,
+      "loss": 0.5215,
+      "step": 4157
+    },
+    {
+      "epoch": 1.064256,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.226687142735262e-05,
+      "loss": 0.4627,
+      "step": 4158
+    },
+    {
+      "epoch": 1.064512,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.2262605579728694e-05,
+      "loss": 0.461,
+      "step": 4159
+    },
+    {
+      "epoch": 1.064768,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.225833973210477e-05,
+      "loss": 0.5169,
+      "step": 4160
+    },
+    {
+      "epoch": 1.065024,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.225407388448084e-05,
+      "loss": 0.4384,
+      "step": 4161
+    },
+    {
+      "epoch": 1.06528,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.2249808036856924e-05,
+      "loss": 0.4521,
+      "step": 4162
+    },
+    {
+      "epoch": 1.065536,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.2245542189233005e-05,
+      "loss": 0.472,
+      "step": 4163
+    },
+    {
+      "epoch": 1.065792,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.224127634160908e-05,
+      "loss": 0.4727,
+      "step": 4164
+    },
+    {
+      "epoch": 1.0660479999999999,
+      "grad_norm": 2.25,
+      "learning_rate": 3.223701049398516e-05,
+      "loss": 0.4607,
+      "step": 4165
+    },
+    {
+      "epoch": 1.066304,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.2232744646361234e-05,
+      "loss": 0.4853,
+      "step": 4166
+    },
+    {
+      "epoch": 1.06656,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.222847879873731e-05,
+      "loss": 0.4787,
+      "step": 4167
+    },
+    {
+      "epoch": 1.066816,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.222421295111338e-05,
+      "loss": 0.4833,
+      "step": 4168
+    },
+    {
+      "epoch": 1.067072,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.221994710348947e-05,
+      "loss": 0.5178,
+      "step": 4169
+    },
+    {
+      "epoch": 1.067328,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.2215681255865545e-05,
+      "loss": 0.5278,
+      "step": 4170
+    },
+    {
+      "epoch": 1.067584,
+      "grad_norm": 2.125,
+      "learning_rate": 3.221141540824162e-05,
+      "loss": 0.4675,
+      "step": 4171
+    },
+    {
+      "epoch": 1.06784,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.2207149560617693e-05,
+      "loss": 0.4596,
+      "step": 4172
+    },
+    {
+      "epoch": 1.068096,
+      "grad_norm": 2.25,
+      "learning_rate": 3.2202883712993774e-05,
+      "loss": 0.4874,
+      "step": 4173
+    },
+    {
+      "epoch": 1.068352,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.219861786536985e-05,
+      "loss": 0.4614,
+      "step": 4174
+    },
+    {
+      "epoch": 1.068608,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.219435201774593e-05,
+      "loss": 0.5008,
+      "step": 4175
+    },
+    {
+      "epoch": 1.068864,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.2190086170122004e-05,
+      "loss": 0.4767,
+      "step": 4176
+    },
+    {
+      "epoch": 1.06912,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.2185820322498085e-05,
+      "loss": 0.4714,
+      "step": 4177
+    },
+    {
+      "epoch": 1.069376,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.218155447487416e-05,
+      "loss": 0.4604,
+      "step": 4178
+    },
+    {
+      "epoch": 1.069632,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.2177288627250234e-05,
+      "loss": 0.4942,
+      "step": 4179
+    },
+    {
+      "epoch": 1.069888,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.217302277962631e-05,
+      "loss": 0.4548,
+      "step": 4180
+    },
+    {
+      "epoch": 1.070144,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.216875693200239e-05,
+      "loss": 0.4481,
+      "step": 4181
+    },
+    {
+      "epoch": 1.0704,
+      "grad_norm": 2.25,
+      "learning_rate": 3.216449108437847e-05,
+      "loss": 0.5416,
+      "step": 4182
+    },
+    {
+      "epoch": 1.070656,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.2160225236754544e-05,
+      "loss": 0.4813,
+      "step": 4183
+    },
+    {
+      "epoch": 1.070912,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.2155959389130625e-05,
+      "loss": 0.4796,
+      "step": 4184
+    },
+    {
+      "epoch": 1.071168,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.21516935415067e-05,
+      "loss": 0.4799,
+      "step": 4185
+    },
+    {
+      "epoch": 1.071424,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.2147427693882774e-05,
+      "loss": 0.4323,
+      "step": 4186
+    },
+    {
+      "epoch": 1.07168,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.214316184625885e-05,
+      "loss": 0.4568,
+      "step": 4187
+    },
+    {
+      "epoch": 1.071936,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.2138895998634936e-05,
+      "loss": 0.4761,
+      "step": 4188
+    },
+    {
+      "epoch": 1.072192,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.213463015101101e-05,
+      "loss": 0.4539,
+      "step": 4189
+    },
+    {
+      "epoch": 1.072448,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.2130364303387084e-05,
+      "loss": 0.4867,
+      "step": 4190
+    },
+    {
+      "epoch": 1.072704,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.212609845576316e-05,
+      "loss": 0.4543,
+      "step": 4191
+    },
+    {
+      "epoch": 1.07296,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.212183260813924e-05,
+      "loss": 0.4349,
+      "step": 4192
+    },
+    {
+      "epoch": 1.073216,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.2117566760515314e-05,
+      "loss": 0.4691,
+      "step": 4193
+    },
+    {
+      "epoch": 1.073472,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.2113300912891395e-05,
+      "loss": 0.475,
+      "step": 4194
+    },
+    {
+      "epoch": 1.073728,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.210903506526747e-05,
+      "loss": 0.4838,
+      "step": 4195
+    },
+    {
+      "epoch": 1.073984,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.210476921764355e-05,
+      "loss": 0.5128,
+      "step": 4196
+    },
+    {
+      "epoch": 1.07424,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.2100503370019625e-05,
+      "loss": 0.4628,
+      "step": 4197
+    },
+    {
+      "epoch": 1.074496,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.20962375223957e-05,
+      "loss": 0.475,
+      "step": 4198
+    },
+    {
+      "epoch": 1.074752,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.209197167477178e-05,
+      "loss": 0.4751,
+      "step": 4199
+    },
+    {
+      "epoch": 1.075008,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.2087705827147854e-05,
+      "loss": 0.4671,
+      "step": 4200
+    },
+    {
+      "epoch": 1.075264,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.2083439979523935e-05,
+      "loss": 0.4526,
+      "step": 4201
+    },
+    {
+      "epoch": 1.07552,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.207917413190001e-05,
+      "loss": 0.4491,
+      "step": 4202
+    },
+    {
+      "epoch": 1.075776,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.207490828427609e-05,
+      "loss": 0.4656,
+      "step": 4203
+    },
+    {
+      "epoch": 1.076032,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.2070642436652165e-05,
+      "loss": 0.4679,
+      "step": 4204
+    },
+    {
+      "epoch": 1.076288,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.206637658902824e-05,
+      "loss": 0.4728,
+      "step": 4205
+    },
+    {
+      "epoch": 1.076544,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.206211074140431e-05,
+      "loss": 0.4752,
+      "step": 4206
+    },
+    {
+      "epoch": 1.0768,
+      "grad_norm": 2.5,
+      "learning_rate": 3.2057844893780394e-05,
+      "loss": 0.4712,
+      "step": 4207
+    },
+    {
+      "epoch": 1.077056,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.2053579046156475e-05,
+      "loss": 0.5074,
+      "step": 4208
+    },
+    {
+      "epoch": 1.077312,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.204931319853255e-05,
+      "loss": 0.4351,
+      "step": 4209
+    },
+    {
+      "epoch": 1.077568,
+      "grad_norm": 2.25,
+      "learning_rate": 3.2045047350908624e-05,
+      "loss": 0.4419,
+      "step": 4210
+    },
+    {
+      "epoch": 1.0778240000000001,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.2040781503284705e-05,
+      "loss": 0.4909,
+      "step": 4211
+    },
+    {
+      "epoch": 1.07808,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.203651565566078e-05,
+      "loss": 0.4753,
+      "step": 4212
+    },
+    {
+      "epoch": 1.078336,
+      "grad_norm": 2.25,
+      "learning_rate": 3.203224980803686e-05,
+      "loss": 0.4695,
+      "step": 4213
+    },
+    {
+      "epoch": 1.078592,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.2027983960412935e-05,
+      "loss": 0.4329,
+      "step": 4214
+    },
+    {
+      "epoch": 1.078848,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.2023718112789016e-05,
+      "loss": 0.4976,
+      "step": 4215
+    },
+    {
+      "epoch": 1.079104,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.201945226516509e-05,
+      "loss": 0.4189,
+      "step": 4216
+    },
+    {
+      "epoch": 1.07936,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.2015186417541164e-05,
+      "loss": 0.4573,
+      "step": 4217
+    },
+    {
+      "epoch": 1.079616,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.2010920569917245e-05,
+      "loss": 0.4687,
+      "step": 4218
+    },
+    {
+      "epoch": 1.079872,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.200665472229332e-05,
+      "loss": 0.4412,
+      "step": 4219
+    },
+    {
+      "epoch": 1.080128,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.20023888746694e-05,
+      "loss": 0.5149,
+      "step": 4220
+    },
+    {
+      "epoch": 1.080384,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.1998123027045475e-05,
+      "loss": 0.4784,
+      "step": 4221
+    },
+    {
+      "epoch": 1.08064,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.1993857179421556e-05,
+      "loss": 0.4583,
+      "step": 4222
+    },
+    {
+      "epoch": 1.080896,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.198959133179763e-05,
+      "loss": 0.4492,
+      "step": 4223
+    },
+    {
+      "epoch": 1.081152,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.1985325484173704e-05,
+      "loss": 0.4408,
+      "step": 4224
+    },
+    {
+      "epoch": 1.081408,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.198105963654978e-05,
+      "loss": 0.4686,
+      "step": 4225
+    },
+    {
+      "epoch": 1.081664,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.197679378892586e-05,
+      "loss": 0.4738,
+      "step": 4226
+    },
+    {
+      "epoch": 1.08192,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.197252794130194e-05,
+      "loss": 0.4627,
+      "step": 4227
+    },
+    {
+      "epoch": 1.082176,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.1968262093678015e-05,
+      "loss": 0.4687,
+      "step": 4228
+    },
+    {
+      "epoch": 1.082432,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.196399624605409e-05,
+      "loss": 0.4723,
+      "step": 4229
+    },
+    {
+      "epoch": 1.082688,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.195973039843017e-05,
+      "loss": 0.511,
+      "step": 4230
+    },
+    {
+      "epoch": 1.082944,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.1955464550806245e-05,
+      "loss": 0.4524,
+      "step": 4231
+    },
+    {
+      "epoch": 1.0832,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.1951198703182326e-05,
+      "loss": 0.4537,
+      "step": 4232
+    },
+    {
+      "epoch": 1.083456,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.1946932855558407e-05,
+      "loss": 0.5108,
+      "step": 4233
+    },
+    {
+      "epoch": 1.083712,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.194266700793448e-05,
+      "loss": 0.4641,
+      "step": 4234
+    },
+    {
+      "epoch": 1.083968,
+      "grad_norm": 2.375,
+      "learning_rate": 3.1938401160310555e-05,
+      "loss": 0.4694,
+      "step": 4235
+    },
+    {
+      "epoch": 1.084224,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.193413531268663e-05,
+      "loss": 0.4891,
+      "step": 4236
+    },
+    {
+      "epoch": 1.08448,
+      "grad_norm": 1.9921875,
+      "learning_rate": 3.192986946506271e-05,
+      "loss": 0.4677,
+      "step": 4237
+    },
+    {
+      "epoch": 1.084736,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.1925603617438785e-05,
+      "loss": 0.4673,
+      "step": 4238
+    },
+    {
+      "epoch": 1.084992,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.1921337769814866e-05,
+      "loss": 0.4518,
+      "step": 4239
+    },
+    {
+      "epoch": 1.085248,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.191707192219094e-05,
+      "loss": 0.4772,
+      "step": 4240
+    },
+    {
+      "epoch": 1.085504,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.191280607456702e-05,
+      "loss": 0.485,
+      "step": 4241
+    },
+    {
+      "epoch": 1.08576,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.1908540226943095e-05,
+      "loss": 0.4637,
+      "step": 4242
+    },
+    {
+      "epoch": 1.086016,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.190427437931917e-05,
+      "loss": 0.4584,
+      "step": 4243
+    },
+    {
+      "epoch": 1.086272,
+      "grad_norm": 2.375,
+      "learning_rate": 3.1900008531695244e-05,
+      "loss": 0.4553,
+      "step": 4244
+    },
+    {
+      "epoch": 1.086528,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.1895742684071325e-05,
+      "loss": 0.4875,
+      "step": 4245
+    },
+    {
+      "epoch": 1.086784,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.1891476836447406e-05,
+      "loss": 0.4827,
+      "step": 4246
+    },
+    {
+      "epoch": 1.08704,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.188721098882348e-05,
+      "loss": 0.5046,
+      "step": 4247
+    },
+    {
+      "epoch": 1.087296,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.188294514119956e-05,
+      "loss": 0.4274,
+      "step": 4248
+    },
+    {
+      "epoch": 1.087552,
+      "grad_norm": 2.25,
+      "learning_rate": 3.1878679293575636e-05,
+      "loss": 0.4393,
+      "step": 4249
+    },
+    {
+      "epoch": 1.0878079999999999,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.187441344595171e-05,
+      "loss": 0.4712,
+      "step": 4250
+    },
+    {
+      "epoch": 1.088064,
+      "grad_norm": 2.25,
+      "learning_rate": 3.187014759832779e-05,
+      "loss": 0.4933,
+      "step": 4251
+    },
+    {
+      "epoch": 1.08832,
+      "grad_norm": 2.25,
+      "learning_rate": 3.186588175070387e-05,
+      "loss": 0.4637,
+      "step": 4252
+    },
+    {
+      "epoch": 1.088576,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.1861615903079946e-05,
+      "loss": 0.4786,
+      "step": 4253
+    },
+    {
+      "epoch": 1.088832,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.185735005545602e-05,
+      "loss": 0.4716,
+      "step": 4254
+    },
+    {
+      "epoch": 1.089088,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.1853084207832095e-05,
+      "loss": 0.4601,
+      "step": 4255
+    },
+    {
+      "epoch": 1.089344,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.1848818360208176e-05,
+      "loss": 0.4939,
+      "step": 4256
+    },
+    {
+      "epoch": 1.0896,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.184455251258425e-05,
+      "loss": 0.4422,
+      "step": 4257
+    },
+    {
+      "epoch": 1.089856,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.184028666496033e-05,
+      "loss": 0.4615,
+      "step": 4258
+    },
+    {
+      "epoch": 1.090112,
+      "grad_norm": 2.125,
+      "learning_rate": 3.1836020817336405e-05,
+      "loss": 0.4571,
+      "step": 4259
+    },
+    {
+      "epoch": 1.090368,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.1831754969712486e-05,
+      "loss": 0.4641,
+      "step": 4260
+    },
+    {
+      "epoch": 1.090624,
+      "grad_norm": 2.25,
+      "learning_rate": 3.182748912208856e-05,
+      "loss": 0.45,
+      "step": 4261
+    },
+    {
+      "epoch": 1.09088,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.1823223274464635e-05,
+      "loss": 0.4549,
+      "step": 4262
+    },
+    {
+      "epoch": 1.091136,
+      "grad_norm": 2.125,
+      "learning_rate": 3.181895742684071e-05,
+      "loss": 0.4566,
+      "step": 4263
+    },
+    {
+      "epoch": 1.091392,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.181469157921679e-05,
+      "loss": 0.4602,
+      "step": 4264
+    },
+    {
+      "epoch": 1.091648,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.181042573159287e-05,
+      "loss": 0.4649,
+      "step": 4265
+    },
+    {
+      "epoch": 1.091904,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.1806159883968945e-05,
+      "loss": 0.4094,
+      "step": 4266
+    },
+    {
+      "epoch": 1.09216,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.1801894036345027e-05,
+      "loss": 0.4848,
+      "step": 4267
+    },
+    {
+      "epoch": 1.092416,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.17976281887211e-05,
+      "loss": 0.4548,
+      "step": 4268
+    },
+    {
+      "epoch": 1.092672,
+      "grad_norm": 2.125,
+      "learning_rate": 3.1793362341097175e-05,
+      "loss": 0.4451,
+      "step": 4269
+    },
+    {
+      "epoch": 1.092928,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.178909649347325e-05,
+      "loss": 0.4672,
+      "step": 4270
+    },
+    {
+      "epoch": 1.093184,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.178483064584934e-05,
+      "loss": 0.4735,
+      "step": 4271
+    },
+    {
+      "epoch": 1.09344,
+      "grad_norm": 2.25,
+      "learning_rate": 3.178056479822541e-05,
+      "loss": 0.4947,
+      "step": 4272
+    },
+    {
+      "epoch": 1.093696,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.1776298950601486e-05,
+      "loss": 0.4889,
+      "step": 4273
+    },
+    {
+      "epoch": 1.093952,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.177203310297756e-05,
+      "loss": 0.484,
+      "step": 4274
+    },
+    {
+      "epoch": 1.094208,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.176776725535364e-05,
+      "loss": 0.4961,
+      "step": 4275
+    },
+    {
+      "epoch": 1.094464,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.1763501407729715e-05,
+      "loss": 0.463,
+      "step": 4276
+    },
+    {
+      "epoch": 1.09472,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.1759235560105796e-05,
+      "loss": 0.4225,
+      "step": 4277
+    },
+    {
+      "epoch": 1.094976,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.175496971248187e-05,
+      "loss": 0.4742,
+      "step": 4278
+    },
+    {
+      "epoch": 1.095232,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.175070386485795e-05,
+      "loss": 0.5012,
+      "step": 4279
+    },
+    {
+      "epoch": 1.095488,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.1746438017234026e-05,
+      "loss": 0.5103,
+      "step": 4280
+    },
+    {
+      "epoch": 1.095744,
+      "grad_norm": 1.9921875,
+      "learning_rate": 3.17421721696101e-05,
+      "loss": 0.5072,
+      "step": 4281
+    },
+    {
+      "epoch": 1.096,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.173790632198618e-05,
+      "loss": 0.4483,
+      "step": 4282
+    },
+    {
+      "epoch": 1.096256,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.1733640474362255e-05,
+      "loss": 0.4582,
+      "step": 4283
+    },
+    {
+      "epoch": 1.096512,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.1729374626738336e-05,
+      "loss": 0.4546,
+      "step": 4284
+    },
+    {
+      "epoch": 1.096768,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.172510877911441e-05,
+      "loss": 0.4844,
+      "step": 4285
+    },
+    {
+      "epoch": 1.097024,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.172084293149049e-05,
+      "loss": 0.4554,
+      "step": 4286
+    },
+    {
+      "epoch": 1.09728,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.1716577083866566e-05,
+      "loss": 0.4561,
+      "step": 4287
+    },
+    {
+      "epoch": 1.097536,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.171231123624264e-05,
+      "loss": 0.4675,
+      "step": 4288
+    },
+    {
+      "epoch": 1.097792,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.1708045388618715e-05,
+      "loss": 0.4804,
+      "step": 4289
+    },
+    {
+      "epoch": 1.098048,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.17037795409948e-05,
+      "loss": 0.4585,
+      "step": 4290
+    },
+    {
+      "epoch": 1.098304,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.169951369337088e-05,
+      "loss": 0.4624,
+      "step": 4291
+    },
+    {
+      "epoch": 1.09856,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.169524784574695e-05,
+      "loss": 0.4816,
+      "step": 4292
+    },
+    {
+      "epoch": 1.098816,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.1690981998123025e-05,
+      "loss": 0.4676,
+      "step": 4293
+    },
+    {
+      "epoch": 1.099072,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.1686716150499106e-05,
+      "loss": 0.4963,
+      "step": 4294
+    },
+    {
+      "epoch": 1.099328,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.168245030287518e-05,
+      "loss": 0.4704,
+      "step": 4295
+    },
+    {
+      "epoch": 1.0995840000000001,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.167818445525126e-05,
+      "loss": 0.4913,
+      "step": 4296
+    },
+    {
+      "epoch": 1.09984,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.167391860762734e-05,
+      "loss": 0.4622,
+      "step": 4297
+    },
+    {
+      "epoch": 1.100096,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.166965276000342e-05,
+      "loss": 0.5046,
+      "step": 4298
+    },
+    {
+      "epoch": 1.100352,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.166538691237949e-05,
+      "loss": 0.5073,
+      "step": 4299
+    },
+    {
+      "epoch": 1.100608,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.1661121064755565e-05,
+      "loss": 0.474,
+      "step": 4300
+    },
+    {
+      "epoch": 1.100864,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.1656855217131646e-05,
+      "loss": 0.4778,
+      "step": 4301
+    },
+    {
+      "epoch": 1.10112,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.165258936950772e-05,
+      "loss": 0.4814,
+      "step": 4302
+    },
+    {
+      "epoch": 1.101376,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.16483235218838e-05,
+      "loss": 0.4886,
+      "step": 4303
+    },
+    {
+      "epoch": 1.101632,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.1644057674259876e-05,
+      "loss": 0.4774,
+      "step": 4304
+    },
+    {
+      "epoch": 1.101888,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.163979182663596e-05,
+      "loss": 0.4804,
+      "step": 4305
+    },
+    {
+      "epoch": 1.102144,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.163552597901203e-05,
+      "loss": 0.4774,
+      "step": 4306
+    },
+    {
+      "epoch": 1.1024,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.1631260131388106e-05,
+      "loss": 0.4906,
+      "step": 4307
+    },
+    {
+      "epoch": 1.102656,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.162699428376418e-05,
+      "loss": 0.4852,
+      "step": 4308
+    },
+    {
+      "epoch": 1.102912,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.162272843614026e-05,
+      "loss": 0.4468,
+      "step": 4309
+    },
+    {
+      "epoch": 1.103168,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.161846258851634e-05,
+      "loss": 0.4981,
+      "step": 4310
+    },
+    {
+      "epoch": 1.103424,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.1614196740892416e-05,
+      "loss": 0.4611,
+      "step": 4311
+    },
+    {
+      "epoch": 1.10368,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.160993089326849e-05,
+      "loss": 0.4813,
+      "step": 4312
+    },
+    {
+      "epoch": 1.103936,
+      "grad_norm": 2.125,
+      "learning_rate": 3.160566504564457e-05,
+      "loss": 0.4761,
+      "step": 4313
+    },
+    {
+      "epoch": 1.104192,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.1601399198020646e-05,
+      "loss": 0.4767,
+      "step": 4314
+    },
+    {
+      "epoch": 1.104448,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.159713335039673e-05,
+      "loss": 0.4624,
+      "step": 4315
+    },
+    {
+      "epoch": 1.104704,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.159286750277281e-05,
+      "loss": 0.4536,
+      "step": 4316
+    },
+    {
+      "epoch": 1.10496,
+      "grad_norm": 2.25,
+      "learning_rate": 3.158860165514888e-05,
+      "loss": 0.4736,
+      "step": 4317
+    },
+    {
+      "epoch": 1.105216,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.1584335807524956e-05,
+      "loss": 0.4432,
+      "step": 4318
+    },
+    {
+      "epoch": 1.105472,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.158006995990103e-05,
+      "loss": 0.4852,
+      "step": 4319
+    },
+    {
+      "epoch": 1.105728,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.157580411227711e-05,
+      "loss": 0.4486,
+      "step": 4320
+    },
+    {
+      "epoch": 1.105984,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.1571538264653186e-05,
+      "loss": 0.477,
+      "step": 4321
+    },
+    {
+      "epoch": 1.1062400000000001,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.156727241702927e-05,
+      "loss": 0.462,
+      "step": 4322
+    },
+    {
+      "epoch": 1.106496,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.156300656940534e-05,
+      "loss": 0.4486,
+      "step": 4323
+    },
+    {
+      "epoch": 1.106752,
+      "grad_norm": 2.0,
+      "learning_rate": 3.155874072178142e-05,
+      "loss": 0.4911,
+      "step": 4324
+    },
+    {
+      "epoch": 1.107008,
+      "grad_norm": 2.125,
+      "learning_rate": 3.1554474874157497e-05,
+      "loss": 0.466,
+      "step": 4325
+    },
+    {
+      "epoch": 1.107264,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.155020902653357e-05,
+      "loss": 0.4585,
+      "step": 4326
+    },
+    {
+      "epoch": 1.10752,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.1545943178909645e-05,
+      "loss": 0.483,
+      "step": 4327
+    },
+    {
+      "epoch": 1.107776,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.1541677331285726e-05,
+      "loss": 0.468,
+      "step": 4328
+    },
+    {
+      "epoch": 1.108032,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.153741148366181e-05,
+      "loss": 0.4925,
+      "step": 4329
+    },
+    {
+      "epoch": 1.108288,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.153314563603788e-05,
+      "loss": 0.4122,
+      "step": 4330
+    },
+    {
+      "epoch": 1.108544,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.152887978841396e-05,
+      "loss": 0.4767,
+      "step": 4331
+    },
+    {
+      "epoch": 1.1088,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.152461394079004e-05,
+      "loss": 0.4656,
+      "step": 4332
+    },
+    {
+      "epoch": 1.109056,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.152034809316611e-05,
+      "loss": 0.4784,
+      "step": 4333
+    },
+    {
+      "epoch": 1.109312,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.151608224554219e-05,
+      "loss": 0.4802,
+      "step": 4334
+    },
+    {
+      "epoch": 1.1095679999999999,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.151181639791827e-05,
+      "loss": 0.4488,
+      "step": 4335
+    },
+    {
+      "epoch": 1.109824,
+      "grad_norm": 1.953125,
+      "learning_rate": 3.150755055029435e-05,
+      "loss": 0.4568,
+      "step": 4336
+    },
+    {
+      "epoch": 1.11008,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.150328470267042e-05,
+      "loss": 0.4496,
+      "step": 4337
+    },
+    {
+      "epoch": 1.110336,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.1499018855046496e-05,
+      "loss": 0.4653,
+      "step": 4338
+    },
+    {
+      "epoch": 1.110592,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.149475300742258e-05,
+      "loss": 0.4398,
+      "step": 4339
+    },
+    {
+      "epoch": 1.110848,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.149048715979865e-05,
+      "loss": 0.4572,
+      "step": 4340
+    },
+    {
+      "epoch": 1.111104,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.148622131217473e-05,
+      "loss": 0.4928,
+      "step": 4341
+    },
+    {
+      "epoch": 1.11136,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.1481955464550807e-05,
+      "loss": 0.4545,
+      "step": 4342
+    },
+    {
+      "epoch": 1.111616,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.147768961692689e-05,
+      "loss": 0.4307,
+      "step": 4343
+    },
+    {
+      "epoch": 1.111872,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.147342376930296e-05,
+      "loss": 0.4713,
+      "step": 4344
+    },
+    {
+      "epoch": 1.112128,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.1469157921679036e-05,
+      "loss": 0.4859,
+      "step": 4345
+    },
+    {
+      "epoch": 1.112384,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.146489207405512e-05,
+      "loss": 0.4606,
+      "step": 4346
+    },
+    {
+      "epoch": 1.11264,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.146062622643119e-05,
+      "loss": 0.4831,
+      "step": 4347
+    },
+    {
+      "epoch": 1.112896,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.145636037880727e-05,
+      "loss": 0.4902,
+      "step": 4348
+    },
+    {
+      "epoch": 1.113152,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.145209453118335e-05,
+      "loss": 0.4766,
+      "step": 4349
+    },
+    {
+      "epoch": 1.113408,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.144782868355943e-05,
+      "loss": 0.4901,
+      "step": 4350
+    },
+    {
+      "epoch": 1.113664,
+      "grad_norm": 2.375,
+      "learning_rate": 3.14435628359355e-05,
+      "loss": 0.4792,
+      "step": 4351
+    },
+    {
+      "epoch": 1.11392,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.1439296988311576e-05,
+      "loss": 0.4549,
+      "step": 4352
+    },
+    {
+      "epoch": 1.114176,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.143503114068766e-05,
+      "loss": 0.4845,
+      "step": 4353
+    },
+    {
+      "epoch": 1.114432,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.143076529306374e-05,
+      "loss": 0.4901,
+      "step": 4354
+    },
+    {
+      "epoch": 1.114688,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.142649944543981e-05,
+      "loss": 0.5236,
+      "step": 4355
+    },
+    {
+      "epoch": 1.114944,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.142223359781589e-05,
+      "loss": 0.4772,
+      "step": 4356
+    },
+    {
+      "epoch": 1.1152,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.141796775019196e-05,
+      "loss": 0.4566,
+      "step": 4357
+    },
+    {
+      "epoch": 1.115456,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.141370190256804e-05,
+      "loss": 0.4886,
+      "step": 4358
+    },
+    {
+      "epoch": 1.115712,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.1409436054944117e-05,
+      "loss": 0.4342,
+      "step": 4359
+    },
+    {
+      "epoch": 1.115968,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.14051702073202e-05,
+      "loss": 0.468,
+      "step": 4360
+    },
+    {
+      "epoch": 1.1162239999999999,
+      "grad_norm": 2.125,
+      "learning_rate": 3.140090435969627e-05,
+      "loss": 0.4659,
+      "step": 4361
+    },
+    {
+      "epoch": 1.11648,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.139663851207235e-05,
+      "loss": 0.4976,
+      "step": 4362
+    },
+    {
+      "epoch": 1.116736,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.139237266444843e-05,
+      "loss": 0.4343,
+      "step": 4363
+    },
+    {
+      "epoch": 1.116992,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.13881068168245e-05,
+      "loss": 0.4639,
+      "step": 4364
+    },
+    {
+      "epoch": 1.117248,
+      "grad_norm": 1.9453125,
+      "learning_rate": 3.138384096920058e-05,
+      "loss": 0.4853,
+      "step": 4365
+    },
+    {
+      "epoch": 1.117504,
+      "grad_norm": 1.984375,
+      "learning_rate": 3.137957512157666e-05,
+      "loss": 0.4872,
+      "step": 4366
+    },
+    {
+      "epoch": 1.11776,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.137530927395274e-05,
+      "loss": 0.4804,
+      "step": 4367
+    },
+    {
+      "epoch": 1.118016,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.137104342632881e-05,
+      "loss": 0.4205,
+      "step": 4368
+    },
+    {
+      "epoch": 1.118272,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.136677757870489e-05,
+      "loss": 0.4787,
+      "step": 4369
+    },
+    {
+      "epoch": 1.118528,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.136251173108097e-05,
+      "loss": 0.4607,
+      "step": 4370
+    },
+    {
+      "epoch": 1.118784,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.135824588345704e-05,
+      "loss": 0.472,
+      "step": 4371
+    },
+    {
+      "epoch": 1.11904,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.135398003583312e-05,
+      "loss": 0.4908,
+      "step": 4372
+    },
+    {
+      "epoch": 1.119296,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.1349714188209204e-05,
+      "loss": 0.4236,
+      "step": 4373
+    },
+    {
+      "epoch": 1.119552,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.134544834058528e-05,
+      "loss": 0.4884,
+      "step": 4374
+    },
+    {
+      "epoch": 1.119808,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.134118249296135e-05,
+      "loss": 0.4718,
+      "step": 4375
+    },
+    {
+      "epoch": 1.120064,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.1336916645337426e-05,
+      "loss": 0.4439,
+      "step": 4376
+    },
+    {
+      "epoch": 1.12032,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.133265079771351e-05,
+      "loss": 0.4755,
+      "step": 4377
+    },
+    {
+      "epoch": 1.120576,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.132838495008958e-05,
+      "loss": 0.4479,
+      "step": 4378
+    },
+    {
+      "epoch": 1.120832,
+      "grad_norm": 2.125,
+      "learning_rate": 3.132411910246566e-05,
+      "loss": 0.4315,
+      "step": 4379
+    },
+    {
+      "epoch": 1.121088,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.1319853254841744e-05,
+      "loss": 0.4514,
+      "step": 4380
+    },
+    {
+      "epoch": 1.121344,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.131558740721782e-05,
+      "loss": 0.4446,
+      "step": 4381
+    },
+    {
+      "epoch": 1.1216,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.131132155959389e-05,
+      "loss": 0.4524,
+      "step": 4382
+    },
+    {
+      "epoch": 1.121856,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.130705571196997e-05,
+      "loss": 0.4688,
+      "step": 4383
+    },
+    {
+      "epoch": 1.122112,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.130278986434605e-05,
+      "loss": 0.4999,
+      "step": 4384
+    },
+    {
+      "epoch": 1.122368,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.129852401672212e-05,
+      "loss": 0.4722,
+      "step": 4385
+    },
+    {
+      "epoch": 1.122624,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.12942581690982e-05,
+      "loss": 0.5092,
+      "step": 4386
+    },
+    {
+      "epoch": 1.12288,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.128999232147428e-05,
+      "loss": 0.4275,
+      "step": 4387
+    },
+    {
+      "epoch": 1.123136,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.128572647385036e-05,
+      "loss": 0.4908,
+      "step": 4388
+    },
+    {
+      "epoch": 1.123392,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.128146062622643e-05,
+      "loss": 0.4801,
+      "step": 4389
+    },
+    {
+      "epoch": 1.123648,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.127719477860251e-05,
+      "loss": 0.4604,
+      "step": 4390
+    },
+    {
+      "epoch": 1.123904,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.127292893097858e-05,
+      "loss": 0.4504,
+      "step": 4391
+    },
+    {
+      "epoch": 1.12416,
+      "grad_norm": 2.5625,
+      "learning_rate": 3.126866308335467e-05,
+      "loss": 0.4551,
+      "step": 4392
+    },
+    {
+      "epoch": 1.124416,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.126439723573074e-05,
+      "loss": 0.4785,
+      "step": 4393
+    },
+    {
+      "epoch": 1.124672,
+      "grad_norm": 2.25,
+      "learning_rate": 3.126013138810682e-05,
+      "loss": 0.4422,
+      "step": 4394
+    },
+    {
+      "epoch": 1.124928,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.12558655404829e-05,
+      "loss": 0.4463,
+      "step": 4395
+    },
+    {
+      "epoch": 1.125184,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.125159969285897e-05,
+      "loss": 0.4675,
+      "step": 4396
+    },
+    {
+      "epoch": 1.12544,
+      "grad_norm": 2.25,
+      "learning_rate": 3.124733384523505e-05,
+      "loss": 0.4971,
+      "step": 4397
+    },
+    {
+      "epoch": 1.125696,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.124306799761113e-05,
+      "loss": 0.4431,
+      "step": 4398
+    },
+    {
+      "epoch": 1.125952,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.123880214998721e-05,
+      "loss": 0.4362,
+      "step": 4399
+    },
+    {
+      "epoch": 1.126208,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.1234536302363283e-05,
+      "loss": 0.4729,
+      "step": 4400
+    },
+    {
+      "epoch": 1.126464,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.123027045473936e-05,
+      "loss": 0.4969,
+      "step": 4401
+    },
+    {
+      "epoch": 1.12672,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.122600460711543e-05,
+      "loss": 0.4772,
+      "step": 4402
+    },
+    {
+      "epoch": 1.126976,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.122173875949151e-05,
+      "loss": 0.4606,
+      "step": 4403
+    },
+    {
+      "epoch": 1.127232,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.121747291186759e-05,
+      "loss": 0.5,
+      "step": 4404
+    },
+    {
+      "epoch": 1.127488,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.121320706424367e-05,
+      "loss": 0.4449,
+      "step": 4405
+    },
+    {
+      "epoch": 1.127744,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.120894121661974e-05,
+      "loss": 0.4533,
+      "step": 4406
+    },
+    {
+      "epoch": 1.1280000000000001,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.1204675368995824e-05,
+      "loss": 0.481,
+      "step": 4407
+    },
+    {
+      "epoch": 1.128256,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.12004095213719e-05,
+      "loss": 0.464,
+      "step": 4408
+    },
+    {
+      "epoch": 1.128512,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.119614367374797e-05,
+      "loss": 0.4797,
+      "step": 4409
+    },
+    {
+      "epoch": 1.128768,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.1191877826124046e-05,
+      "loss": 0.4863,
+      "step": 4410
+    },
+    {
+      "epoch": 1.129024,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.118761197850013e-05,
+      "loss": 0.4611,
+      "step": 4411
+    },
+    {
+      "epoch": 1.12928,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.118334613087621e-05,
+      "loss": 0.4706,
+      "step": 4412
+    },
+    {
+      "epoch": 1.129536,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.117908028325228e-05,
+      "loss": 0.4496,
+      "step": 4413
+    },
+    {
+      "epoch": 1.129792,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.1174814435628364e-05,
+      "loss": 0.501,
+      "step": 4414
+    },
+    {
+      "epoch": 1.130048,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.117054858800444e-05,
+      "loss": 0.5118,
+      "step": 4415
+    },
+    {
+      "epoch": 1.130304,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.116628274038051e-05,
+      "loss": 0.4615,
+      "step": 4416
+    },
+    {
+      "epoch": 1.13056,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.116201689275659e-05,
+      "loss": 0.4772,
+      "step": 4417
+    },
+    {
+      "epoch": 1.130816,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.1157751045132674e-05,
+      "loss": 0.4635,
+      "step": 4418
+    },
+    {
+      "epoch": 1.131072,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.115348519750875e-05,
+      "loss": 0.4705,
+      "step": 4419
+    },
+    {
+      "epoch": 1.131328,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.114921934988482e-05,
+      "loss": 0.4837,
+      "step": 4420
+    },
+    {
+      "epoch": 1.131584,
+      "grad_norm": 1.90625,
+      "learning_rate": 3.11449535022609e-05,
+      "loss": 0.4308,
+      "step": 4421
+    },
+    {
+      "epoch": 1.13184,
+      "grad_norm": 1.9765625,
+      "learning_rate": 3.114068765463698e-05,
+      "loss": 0.4826,
+      "step": 4422
+    },
+    {
+      "epoch": 1.132096,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.113642180701305e-05,
+      "loss": 0.4152,
+      "step": 4423
+    },
+    {
+      "epoch": 1.132352,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.1132155959389134e-05,
+      "loss": 0.4455,
+      "step": 4424
+    },
+    {
+      "epoch": 1.132608,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.112789011176521e-05,
+      "loss": 0.4595,
+      "step": 4425
+    },
+    {
+      "epoch": 1.132864,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.112362426414129e-05,
+      "loss": 0.5098,
+      "step": 4426
+    },
+    {
+      "epoch": 1.13312,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.111935841651736e-05,
+      "loss": 0.4671,
+      "step": 4427
+    },
+    {
+      "epoch": 1.133376,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.111509256889344e-05,
+      "loss": 0.4547,
+      "step": 4428
+    },
+    {
+      "epoch": 1.133632,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.111082672126952e-05,
+      "loss": 0.4584,
+      "step": 4429
+    },
+    {
+      "epoch": 1.133888,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.110656087364559e-05,
+      "loss": 0.4848,
+      "step": 4430
+    },
+    {
+      "epoch": 1.134144,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.1102295026021674e-05,
+      "loss": 0.4746,
+      "step": 4431
+    },
+    {
+      "epoch": 1.1344,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.109802917839775e-05,
+      "loss": 0.4676,
+      "step": 4432
+    },
+    {
+      "epoch": 1.134656,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.109376333077383e-05,
+      "loss": 0.4498,
+      "step": 4433
+    },
+    {
+      "epoch": 1.134912,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.10894974831499e-05,
+      "loss": 0.4809,
+      "step": 4434
+    },
+    {
+      "epoch": 1.135168,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.108523163552598e-05,
+      "loss": 0.4753,
+      "step": 4435
+    },
+    {
+      "epoch": 1.135424,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.108096578790206e-05,
+      "loss": 0.4481,
+      "step": 4436
+    },
+    {
+      "epoch": 1.13568,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.107669994027814e-05,
+      "loss": 0.4371,
+      "step": 4437
+    },
+    {
+      "epoch": 1.135936,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.1072434092654214e-05,
+      "loss": 0.454,
+      "step": 4438
+    },
+    {
+      "epoch": 1.136192,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.106816824503029e-05,
+      "loss": 0.4905,
+      "step": 4439
+    },
+    {
+      "epoch": 1.1364480000000001,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.106390239740636e-05,
+      "loss": 0.499,
+      "step": 4440
+    },
+    {
+      "epoch": 1.136704,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.1059636549782444e-05,
+      "loss": 0.474,
+      "step": 4441
+    },
+    {
+      "epoch": 1.13696,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.105537070215852e-05,
+      "loss": 0.4463,
+      "step": 4442
+    },
+    {
+      "epoch": 1.137216,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.10511048545346e-05,
+      "loss": 0.4742,
+      "step": 4443
+    },
+    {
+      "epoch": 1.137472,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.104683900691068e-05,
+      "loss": 0.4471,
+      "step": 4444
+    },
+    {
+      "epoch": 1.137728,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.1042573159286754e-05,
+      "loss": 0.4748,
+      "step": 4445
+    },
+    {
+      "epoch": 1.1379839999999999,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.103830731166283e-05,
+      "loss": 0.4565,
+      "step": 4446
+    },
+    {
+      "epoch": 1.13824,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.10340414640389e-05,
+      "loss": 0.4648,
+      "step": 4447
+    },
+    {
+      "epoch": 1.138496,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.1029775616414984e-05,
+      "loss": 0.4101,
+      "step": 4448
+    },
+    {
+      "epoch": 1.138752,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.102550976879106e-05,
+      "loss": 0.5034,
+      "step": 4449
+    },
+    {
+      "epoch": 1.139008,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.102124392116714e-05,
+      "loss": 0.4063,
+      "step": 4450
+    },
+    {
+      "epoch": 1.139264,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.101697807354321e-05,
+      "loss": 0.4768,
+      "step": 4451
+    },
+    {
+      "epoch": 1.13952,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.1012712225919294e-05,
+      "loss": 0.4316,
+      "step": 4452
+    },
+    {
+      "epoch": 1.139776,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.100844637829537e-05,
+      "loss": 0.4523,
+      "step": 4453
+    },
+    {
+      "epoch": 1.140032,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.100418053067144e-05,
+      "loss": 0.4971,
+      "step": 4454
+    },
+    {
+      "epoch": 1.140288,
+      "grad_norm": 2.625,
+      "learning_rate": 3.0999914683047524e-05,
+      "loss": 0.5111,
+      "step": 4455
+    },
+    {
+      "epoch": 1.140544,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.0995648835423605e-05,
+      "loss": 0.452,
+      "step": 4456
+    },
+    {
+      "epoch": 1.1408,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.099138298779968e-05,
+      "loss": 0.4453,
+      "step": 4457
+    },
+    {
+      "epoch": 1.141056,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.0987117140175753e-05,
+      "loss": 0.4367,
+      "step": 4458
+    },
+    {
+      "epoch": 1.141312,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.098285129255183e-05,
+      "loss": 0.4662,
+      "step": 4459
+    },
+    {
+      "epoch": 1.141568,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.097858544492791e-05,
+      "loss": 0.4868,
+      "step": 4460
+    },
+    {
+      "epoch": 1.141824,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.097431959730398e-05,
+      "loss": 0.423,
+      "step": 4461
+    },
+    {
+      "epoch": 1.14208,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.0970053749680064e-05,
+      "loss": 0.4656,
+      "step": 4462
+    },
+    {
+      "epoch": 1.142336,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.0965787902056145e-05,
+      "loss": 0.4644,
+      "step": 4463
+    },
+    {
+      "epoch": 1.142592,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.096152205443222e-05,
+      "loss": 0.4353,
+      "step": 4464
+    },
+    {
+      "epoch": 1.142848,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.0957256206808294e-05,
+      "loss": 0.4897,
+      "step": 4465
+    },
+    {
+      "epoch": 1.1431040000000001,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.095299035918437e-05,
+      "loss": 0.4678,
+      "step": 4466
+    },
+    {
+      "epoch": 1.14336,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.094872451156045e-05,
+      "loss": 0.4473,
+      "step": 4467
+    },
+    {
+      "epoch": 1.143616,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.094445866393652e-05,
+      "loss": 0.4449,
+      "step": 4468
+    },
+    {
+      "epoch": 1.143872,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.0940192816312604e-05,
+      "loss": 0.4799,
+      "step": 4469
+    },
+    {
+      "epoch": 1.144128,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.093592696868868e-05,
+      "loss": 0.4396,
+      "step": 4470
+    },
+    {
+      "epoch": 1.144384,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.093166112106476e-05,
+      "loss": 0.4358,
+      "step": 4471
+    },
+    {
+      "epoch": 1.1446399999999999,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.0927395273440834e-05,
+      "loss": 0.4551,
+      "step": 4472
+    },
+    {
+      "epoch": 1.144896,
+      "grad_norm": 2.25,
+      "learning_rate": 3.092312942581691e-05,
+      "loss": 0.4656,
+      "step": 4473
+    },
+    {
+      "epoch": 1.145152,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.091886357819299e-05,
+      "loss": 0.4823,
+      "step": 4474
+    },
+    {
+      "epoch": 1.145408,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.091459773056907e-05,
+      "loss": 0.4313,
+      "step": 4475
+    },
+    {
+      "epoch": 1.145664,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.0910331882945144e-05,
+      "loss": 0.4691,
+      "step": 4476
+    },
+    {
+      "epoch": 1.14592,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.090606603532122e-05,
+      "loss": 0.453,
+      "step": 4477
+    },
+    {
+      "epoch": 1.146176,
+      "grad_norm": 1.953125,
+      "learning_rate": 3.09018001876973e-05,
+      "loss": 0.4595,
+      "step": 4478
+    },
+    {
+      "epoch": 1.146432,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.0897534340073374e-05,
+      "loss": 0.4531,
+      "step": 4479
+    },
+    {
+      "epoch": 1.146688,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.089326849244945e-05,
+      "loss": 0.4819,
+      "step": 4480
+    },
+    {
+      "epoch": 1.146944,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.088900264482553e-05,
+      "loss": 0.5024,
+      "step": 4481
+    },
+    {
+      "epoch": 1.1472,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.088473679720161e-05,
+      "loss": 0.4645,
+      "step": 4482
+    },
+    {
+      "epoch": 1.147456,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.0880470949577685e-05,
+      "loss": 0.441,
+      "step": 4483
+    },
+    {
+      "epoch": 1.147712,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.087620510195376e-05,
+      "loss": 0.4625,
+      "step": 4484
+    },
+    {
+      "epoch": 1.147968,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.087193925432983e-05,
+      "loss": 0.4517,
+      "step": 4485
+    },
+    {
+      "epoch": 1.148224,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.0867673406705914e-05,
+      "loss": 0.4667,
+      "step": 4486
+    },
+    {
+      "epoch": 1.14848,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.086340755908199e-05,
+      "loss": 0.459,
+      "step": 4487
+    },
+    {
+      "epoch": 1.148736,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.085914171145807e-05,
+      "loss": 0.4826,
+      "step": 4488
+    },
+    {
+      "epoch": 1.148992,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.0854875863834144e-05,
+      "loss": 0.4218,
+      "step": 4489
+    },
+    {
+      "epoch": 1.149248,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.0850610016210225e-05,
+      "loss": 0.4532,
+      "step": 4490
+    },
+    {
+      "epoch": 1.149504,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.08463441685863e-05,
+      "loss": 0.4739,
+      "step": 4491
+    },
+    {
+      "epoch": 1.1497600000000001,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.0842078320962373e-05,
+      "loss": 0.4584,
+      "step": 4492
+    },
+    {
+      "epoch": 1.150016,
+      "grad_norm": 2.25,
+      "learning_rate": 3.0837812473338454e-05,
+      "loss": 0.4543,
+      "step": 4493
+    },
+    {
+      "epoch": 1.150272,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.0833546625714536e-05,
+      "loss": 0.4469,
+      "step": 4494
+    },
+    {
+      "epoch": 1.150528,
+      "grad_norm": 2.375,
+      "learning_rate": 3.082928077809061e-05,
+      "loss": 0.4538,
+      "step": 4495
+    },
+    {
+      "epoch": 1.150784,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.0825014930466684e-05,
+      "loss": 0.4673,
+      "step": 4496
+    },
+    {
+      "epoch": 1.15104,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.0820749082842765e-05,
+      "loss": 0.4671,
+      "step": 4497
+    },
+    {
+      "epoch": 1.151296,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.081648323521884e-05,
+      "loss": 0.4292,
+      "step": 4498
+    },
+    {
+      "epoch": 1.151552,
+      "grad_norm": 2.125,
+      "learning_rate": 3.0812217387594914e-05,
+      "loss": 0.4694,
+      "step": 4499
+    },
+    {
+      "epoch": 1.151808,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.0807951539970995e-05,
+      "loss": 0.4519,
+      "step": 4500
+    },
+    {
+      "epoch": 1.152064,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.0803685692347076e-05,
+      "loss": 0.4417,
+      "step": 4501
+    },
+    {
+      "epoch": 1.15232,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.079941984472315e-05,
+      "loss": 0.5055,
+      "step": 4502
+    },
+    {
+      "epoch": 1.152576,
+      "grad_norm": 2.25,
+      "learning_rate": 3.0795153997099224e-05,
+      "loss": 0.5238,
+      "step": 4503
+    },
+    {
+      "epoch": 1.152832,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.07908881494753e-05,
+      "loss": 0.4613,
+      "step": 4504
+    },
+    {
+      "epoch": 1.153088,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.078662230185138e-05,
+      "loss": 0.4503,
+      "step": 4505
+    },
+    {
+      "epoch": 1.153344,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.0782356454227454e-05,
+      "loss": 0.4578,
+      "step": 4506
+    },
+    {
+      "epoch": 1.1536,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.0778090606603535e-05,
+      "loss": 0.4857,
+      "step": 4507
+    },
+    {
+      "epoch": 1.153856,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.077382475897961e-05,
+      "loss": 0.4692,
+      "step": 4508
+    },
+    {
+      "epoch": 1.154112,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.076955891135569e-05,
+      "loss": 0.4599,
+      "step": 4509
+    },
+    {
+      "epoch": 1.154368,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.0765293063731764e-05,
+      "loss": 0.4288,
+      "step": 4510
+    },
+    {
+      "epoch": 1.154624,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.076102721610784e-05,
+      "loss": 0.402,
+      "step": 4511
+    },
+    {
+      "epoch": 1.15488,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.075676136848392e-05,
+      "loss": 0.4684,
+      "step": 4512
+    },
+    {
+      "epoch": 1.155136,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.075249552086e-05,
+      "loss": 0.4498,
+      "step": 4513
+    },
+    {
+      "epoch": 1.155392,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.0748229673236075e-05,
+      "loss": 0.4898,
+      "step": 4514
+    },
+    {
+      "epoch": 1.155648,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.074396382561215e-05,
+      "loss": 0.4683,
+      "step": 4515
+    },
+    {
+      "epoch": 1.155904,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.073969797798823e-05,
+      "loss": 0.4857,
+      "step": 4516
+    },
+    {
+      "epoch": 1.15616,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.0735432130364305e-05,
+      "loss": 0.5124,
+      "step": 4517
+    },
+    {
+      "epoch": 1.156416,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.073116628274038e-05,
+      "loss": 0.4675,
+      "step": 4518
+    },
+    {
+      "epoch": 1.156672,
+      "grad_norm": 2.25,
+      "learning_rate": 3.072690043511646e-05,
+      "loss": 0.4837,
+      "step": 4519
+    },
+    {
+      "epoch": 1.156928,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.072263458749254e-05,
+      "loss": 0.4882,
+      "step": 4520
+    },
+    {
+      "epoch": 1.157184,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.0718368739868615e-05,
+      "loss": 0.4578,
+      "step": 4521
+    },
+    {
+      "epoch": 1.15744,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.071410289224469e-05,
+      "loss": 0.4682,
+      "step": 4522
+    },
+    {
+      "epoch": 1.157696,
+      "grad_norm": 2.125,
+      "learning_rate": 3.0709837044620764e-05,
+      "loss": 0.476,
+      "step": 4523
+    },
+    {
+      "epoch": 1.157952,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.0705571196996845e-05,
+      "loss": 0.4604,
+      "step": 4524
+    },
+    {
+      "epoch": 1.158208,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.070130534937292e-05,
+      "loss": 0.468,
+      "step": 4525
+    },
+    {
+      "epoch": 1.158464,
+      "grad_norm": 2.125,
+      "learning_rate": 3.0697039501749e-05,
+      "loss": 0.4454,
+      "step": 4526
+    },
+    {
+      "epoch": 1.15872,
+      "grad_norm": 2.125,
+      "learning_rate": 3.069277365412508e-05,
+      "loss": 0.4538,
+      "step": 4527
+    },
+    {
+      "epoch": 1.158976,
+      "grad_norm": 2.125,
+      "learning_rate": 3.0688507806501155e-05,
+      "loss": 0.4874,
+      "step": 4528
+    },
+    {
+      "epoch": 1.159232,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.068424195887723e-05,
+      "loss": 0.4462,
+      "step": 4529
+    },
+    {
+      "epoch": 1.159488,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.0679976111253304e-05,
+      "loss": 0.4676,
+      "step": 4530
+    },
+    {
+      "epoch": 1.1597439999999999,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.0675710263629385e-05,
+      "loss": 0.4619,
+      "step": 4531
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.067144441600546e-05,
+      "loss": 0.4442,
+      "step": 4532
+    },
+    {
+      "epoch": 1.160256,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.066717856838154e-05,
+      "loss": 0.4736,
+      "step": 4533
+    },
+    {
+      "epoch": 1.160512,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.0662912720757615e-05,
+      "loss": 0.4522,
+      "step": 4534
+    },
+    {
+      "epoch": 1.160768,
+      "grad_norm": 2.375,
+      "learning_rate": 3.0658646873133696e-05,
+      "loss": 0.4575,
+      "step": 4535
+    },
+    {
+      "epoch": 1.161024,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.065438102550977e-05,
+      "loss": 0.4764,
+      "step": 4536
+    },
+    {
+      "epoch": 1.16128,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.0650115177885844e-05,
+      "loss": 0.4675,
+      "step": 4537
+    },
+    {
+      "epoch": 1.161536,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.0645849330261925e-05,
+      "loss": 0.4804,
+      "step": 4538
+    },
+    {
+      "epoch": 1.161792,
+      "grad_norm": 1.9765625,
+      "learning_rate": 3.0641583482638006e-05,
+      "loss": 0.4497,
+      "step": 4539
+    },
+    {
+      "epoch": 1.162048,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.063731763501408e-05,
+      "loss": 0.4982,
+      "step": 4540
+    },
+    {
+      "epoch": 1.162304,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.0633051787390155e-05,
+      "loss": 0.4859,
+      "step": 4541
+    },
+    {
+      "epoch": 1.16256,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.0628785939766236e-05,
+      "loss": 0.482,
+      "step": 4542
+    },
+    {
+      "epoch": 1.162816,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.062452009214231e-05,
+      "loss": 0.4295,
+      "step": 4543
+    },
+    {
+      "epoch": 1.163072,
+      "grad_norm": 2.125,
+      "learning_rate": 3.0620254244518384e-05,
+      "loss": 0.4657,
+      "step": 4544
+    },
+    {
+      "epoch": 1.163328,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.0615988396894465e-05,
+      "loss": 0.4672,
+      "step": 4545
+    },
+    {
+      "epoch": 1.163584,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.0611722549270546e-05,
+      "loss": 0.4402,
+      "step": 4546
+    },
+    {
+      "epoch": 1.16384,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.060745670164662e-05,
+      "loss": 0.4579,
+      "step": 4547
+    },
+    {
+      "epoch": 1.164096,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.0603190854022695e-05,
+      "loss": 0.4705,
+      "step": 4548
+    },
+    {
+      "epoch": 1.164352,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.059892500639877e-05,
+      "loss": 0.4415,
+      "step": 4549
+    },
+    {
+      "epoch": 1.164608,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.059465915877485e-05,
+      "loss": 0.4504,
+      "step": 4550
+    },
+    {
+      "epoch": 1.1648640000000001,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.0590393311150925e-05,
+      "loss": 0.4621,
+      "step": 4551
+    },
+    {
+      "epoch": 1.16512,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.0586127463527006e-05,
+      "loss": 0.4391,
+      "step": 4552
+    },
+    {
+      "epoch": 1.165376,
+      "grad_norm": 2.0,
+      "learning_rate": 3.058186161590308e-05,
+      "loss": 0.5252,
+      "step": 4553
+    },
+    {
+      "epoch": 1.165632,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.057759576827916e-05,
+      "loss": 0.4772,
+      "step": 4554
+    },
+    {
+      "epoch": 1.165888,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.0573329920655235e-05,
+      "loss": 0.4485,
+      "step": 4555
+    },
+    {
+      "epoch": 1.166144,
+      "grad_norm": 1.9921875,
+      "learning_rate": 3.056906407303131e-05,
+      "loss": 0.4832,
+      "step": 4556
+    },
+    {
+      "epoch": 1.1663999999999999,
+      "grad_norm": 2.0,
+      "learning_rate": 3.056479822540739e-05,
+      "loss": 0.4706,
+      "step": 4557
+    },
+    {
+      "epoch": 1.166656,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.056053237778347e-05,
+      "loss": 0.4295,
+      "step": 4558
+    },
+    {
+      "epoch": 1.166912,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.0556266530159546e-05,
+      "loss": 0.446,
+      "step": 4559
+    },
+    {
+      "epoch": 1.167168,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.055200068253562e-05,
+      "loss": 0.4769,
+      "step": 4560
+    },
+    {
+      "epoch": 1.167424,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.05477348349117e-05,
+      "loss": 0.4736,
+      "step": 4561
+    },
+    {
+      "epoch": 1.16768,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.0543468987287775e-05,
+      "loss": 0.4748,
+      "step": 4562
+    },
+    {
+      "epoch": 1.167936,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.053920313966385e-05,
+      "loss": 0.4741,
+      "step": 4563
+    },
+    {
+      "epoch": 1.168192,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.053493729203993e-05,
+      "loss": 0.5136,
+      "step": 4564
+    },
+    {
+      "epoch": 1.168448,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.053067144441601e-05,
+      "loss": 0.499,
+      "step": 4565
+    },
+    {
+      "epoch": 1.168704,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.0526405596792086e-05,
+      "loss": 0.4792,
+      "step": 4566
+    },
+    {
+      "epoch": 1.16896,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.052213974916816e-05,
+      "loss": 0.4732,
+      "step": 4567
+    },
+    {
+      "epoch": 1.169216,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.0517873901544234e-05,
+      "loss": 0.4445,
+      "step": 4568
+    },
+    {
+      "epoch": 1.169472,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.051360805392032e-05,
+      "loss": 0.4265,
+      "step": 4569
+    },
+    {
+      "epoch": 1.169728,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.0509342206296393e-05,
+      "loss": 0.5065,
+      "step": 4570
+    },
+    {
+      "epoch": 1.169984,
+      "grad_norm": 2.0,
+      "learning_rate": 3.0505076358672467e-05,
+      "loss": 0.4873,
+      "step": 4571
+    },
+    {
+      "epoch": 1.17024,
+      "grad_norm": 1.9609375,
+      "learning_rate": 3.0500810511048545e-05,
+      "loss": 0.482,
+      "step": 4572
+    },
+    {
+      "epoch": 1.170496,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.0496544663424626e-05,
+      "loss": 0.5186,
+      "step": 4573
+    },
+    {
+      "epoch": 1.170752,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.04922788158007e-05,
+      "loss": 0.4418,
+      "step": 4574
+    },
+    {
+      "epoch": 1.171008,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.0488012968176778e-05,
+      "loss": 0.481,
+      "step": 4575
+    },
+    {
+      "epoch": 1.171264,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.048374712055286e-05,
+      "loss": 0.4925,
+      "step": 4576
+    },
+    {
+      "epoch": 1.1715200000000001,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.0479481272928933e-05,
+      "loss": 0.4854,
+      "step": 4577
+    },
+    {
+      "epoch": 1.171776,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.0475215425305008e-05,
+      "loss": 0.4283,
+      "step": 4578
+    },
+    {
+      "epoch": 1.172032,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.0470949577681085e-05,
+      "loss": 0.4792,
+      "step": 4579
+    },
+    {
+      "epoch": 1.172288,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.0466683730057166e-05,
+      "loss": 0.4668,
+      "step": 4580
+    },
+    {
+      "epoch": 1.172544,
+      "grad_norm": 1.9921875,
+      "learning_rate": 3.046241788243324e-05,
+      "loss": 0.5002,
+      "step": 4581
+    },
+    {
+      "epoch": 1.1728,
+      "grad_norm": 1.9921875,
+      "learning_rate": 3.0458152034809318e-05,
+      "loss": 0.4999,
+      "step": 4582
+    },
+    {
+      "epoch": 1.1730559999999999,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.0453886187185393e-05,
+      "loss": 0.5085,
+      "step": 4583
+    },
+    {
+      "epoch": 1.173312,
+      "grad_norm": 2.25,
+      "learning_rate": 3.0449620339561474e-05,
+      "loss": 0.4799,
+      "step": 4584
+    },
+    {
+      "epoch": 1.173568,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.044535449193755e-05,
+      "loss": 0.4654,
+      "step": 4585
+    },
+    {
+      "epoch": 1.173824,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.0441088644313626e-05,
+      "loss": 0.4726,
+      "step": 4586
+    },
+    {
+      "epoch": 1.17408,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.04368227966897e-05,
+      "loss": 0.4482,
+      "step": 4587
+    },
+    {
+      "epoch": 1.174336,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.043255694906578e-05,
+      "loss": 0.4454,
+      "step": 4588
+    },
+    {
+      "epoch": 1.174592,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.042829110144186e-05,
+      "loss": 0.4608,
+      "step": 4589
+    },
+    {
+      "epoch": 1.174848,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.0424025253817933e-05,
+      "loss": 0.4575,
+      "step": 4590
+    },
+    {
+      "epoch": 1.175104,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.041975940619401e-05,
+      "loss": 0.5161,
+      "step": 4591
+    },
+    {
+      "epoch": 1.17536,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.041549355857009e-05,
+      "loss": 0.4451,
+      "step": 4592
+    },
+    {
+      "epoch": 1.175616,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.0411227710946166e-05,
+      "loss": 0.4653,
+      "step": 4593
+    },
+    {
+      "epoch": 1.175872,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.0406961863322243e-05,
+      "loss": 0.4301,
+      "step": 4594
+    },
+    {
+      "epoch": 1.176128,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.0402696015698324e-05,
+      "loss": 0.4761,
+      "step": 4595
+    },
+    {
+      "epoch": 1.176384,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.03984301680744e-05,
+      "loss": 0.4678,
+      "step": 4596
+    },
+    {
+      "epoch": 1.17664,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.0394164320450473e-05,
+      "loss": 0.4602,
+      "step": 4597
+    },
+    {
+      "epoch": 1.176896,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.038989847282655e-05,
+      "loss": 0.4667,
+      "step": 4598
+    },
+    {
+      "epoch": 1.177152,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.038563262520263e-05,
+      "loss": 0.4587,
+      "step": 4599
+    },
+    {
+      "epoch": 1.177408,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.0381366777578706e-05,
+      "loss": 0.4459,
+      "step": 4600
+    },
+    {
+      "epoch": 1.177664,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.0377100929954784e-05,
+      "loss": 0.4826,
+      "step": 4601
+    },
+    {
+      "epoch": 1.17792,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.0372835082330858e-05,
+      "loss": 0.3995,
+      "step": 4602
+    },
+    {
+      "epoch": 1.1781760000000001,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.036856923470694e-05,
+      "loss": 0.4437,
+      "step": 4603
+    },
+    {
+      "epoch": 1.178432,
+      "grad_norm": 2.25,
+      "learning_rate": 3.0364303387083017e-05,
+      "loss": 0.4635,
+      "step": 4604
+    },
+    {
+      "epoch": 1.178688,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.036003753945909e-05,
+      "loss": 0.4479,
+      "step": 4605
+    },
+    {
+      "epoch": 1.178944,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.0355771691835165e-05,
+      "loss": 0.4501,
+      "step": 4606
+    },
+    {
+      "epoch": 1.1792,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.0351505844211246e-05,
+      "loss": 0.4371,
+      "step": 4607
+    },
+    {
+      "epoch": 1.179456,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.0347239996587324e-05,
+      "loss": 0.4381,
+      "step": 4608
+    },
+    {
+      "epoch": 1.179712,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.0342974148963398e-05,
+      "loss": 0.4749,
+      "step": 4609
+    },
+    {
+      "epoch": 1.179968,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.033870830133948e-05,
+      "loss": 0.4741,
+      "step": 4610
+    },
+    {
+      "epoch": 1.180224,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.0334442453715557e-05,
+      "loss": 0.4801,
+      "step": 4611
+    },
+    {
+      "epoch": 1.18048,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.033017660609163e-05,
+      "loss": 0.4516,
+      "step": 4612
+    },
+    {
+      "epoch": 1.180736,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.032591075846771e-05,
+      "loss": 0.466,
+      "step": 4613
+    },
+    {
+      "epoch": 1.180992,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.032164491084379e-05,
+      "loss": 0.4795,
+      "step": 4614
+    },
+    {
+      "epoch": 1.181248,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.0317379063219864e-05,
+      "loss": 0.4161,
+      "step": 4615
+    },
+    {
+      "epoch": 1.1815039999999999,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.0313113215595938e-05,
+      "loss": 0.4578,
+      "step": 4616
+    },
+    {
+      "epoch": 1.18176,
+      "grad_norm": 1.9609375,
+      "learning_rate": 3.0308847367972016e-05,
+      "loss": 0.4527,
+      "step": 4617
+    },
+    {
+      "epoch": 1.182016,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.0304581520348097e-05,
+      "loss": 0.4299,
+      "step": 4618
+    },
+    {
+      "epoch": 1.182272,
+      "grad_norm": 1.984375,
+      "learning_rate": 3.030031567272417e-05,
+      "loss": 0.4678,
+      "step": 4619
+    },
+    {
+      "epoch": 1.182528,
+      "grad_norm": 1.921875,
+      "learning_rate": 3.029604982510025e-05,
+      "loss": 0.4725,
+      "step": 4620
+    },
+    {
+      "epoch": 1.182784,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.0291783977476323e-05,
+      "loss": 0.4761,
+      "step": 4621
+    },
+    {
+      "epoch": 1.18304,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.0287518129852404e-05,
+      "loss": 0.4731,
+      "step": 4622
+    },
+    {
+      "epoch": 1.183296,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.0283252282228482e-05,
+      "loss": 0.467,
+      "step": 4623
+    },
+    {
+      "epoch": 1.183552,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.0278986434604556e-05,
+      "loss": 0.4819,
+      "step": 4624
+    },
+    {
+      "epoch": 1.183808,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.0274720586980637e-05,
+      "loss": 0.4577,
+      "step": 4625
+    },
+    {
+      "epoch": 1.184064,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.027045473935671e-05,
+      "loss": 0.4767,
+      "step": 4626
+    },
+    {
+      "epoch": 1.18432,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.026618889173279e-05,
+      "loss": 0.4774,
+      "step": 4627
+    },
+    {
+      "epoch": 1.184576,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.0261923044108863e-05,
+      "loss": 0.476,
+      "step": 4628
+    },
+    {
+      "epoch": 1.184832,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.0257657196484944e-05,
+      "loss": 0.5066,
+      "step": 4629
+    },
+    {
+      "epoch": 1.185088,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.0253391348861022e-05,
+      "loss": 0.4793,
+      "step": 4630
+    },
+    {
+      "epoch": 1.185344,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.0249125501237096e-05,
+      "loss": 0.4725,
+      "step": 4631
+    },
+    {
+      "epoch": 1.1856,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.0244859653613174e-05,
+      "loss": 0.5027,
+      "step": 4632
+    },
+    {
+      "epoch": 1.185856,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.0240593805989255e-05,
+      "loss": 0.4561,
+      "step": 4633
+    },
+    {
+      "epoch": 1.186112,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.023632795836533e-05,
+      "loss": 0.4559,
+      "step": 4634
+    },
+    {
+      "epoch": 1.186368,
+      "grad_norm": 2.0,
+      "learning_rate": 3.0232062110741403e-05,
+      "loss": 0.4443,
+      "step": 4635
+    },
+    {
+      "epoch": 1.1866240000000001,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.022779626311748e-05,
+      "loss": 0.4431,
+      "step": 4636
+    },
+    {
+      "epoch": 1.18688,
+      "grad_norm": 2.125,
+      "learning_rate": 3.0223530415493562e-05,
+      "loss": 0.471,
+      "step": 4637
+    },
+    {
+      "epoch": 1.187136,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.0219264567869636e-05,
+      "loss": 0.4754,
+      "step": 4638
+    },
+    {
+      "epoch": 1.187392,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.0214998720245714e-05,
+      "loss": 0.4267,
+      "step": 4639
+    },
+    {
+      "epoch": 1.187648,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.021073287262179e-05,
+      "loss": 0.4704,
+      "step": 4640
+    },
+    {
+      "epoch": 1.187904,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.020646702499787e-05,
+      "loss": 0.4671,
+      "step": 4641
+    },
+    {
+      "epoch": 1.1881599999999999,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.0202201177373947e-05,
+      "loss": 0.4465,
+      "step": 4642
+    },
+    {
+      "epoch": 1.188416,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.019793532975002e-05,
+      "loss": 0.4463,
+      "step": 4643
+    },
+    {
+      "epoch": 1.188672,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.0193669482126102e-05,
+      "loss": 0.4259,
+      "step": 4644
+    },
+    {
+      "epoch": 1.188928,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.0189403634502177e-05,
+      "loss": 0.4798,
+      "step": 4645
+    },
+    {
+      "epoch": 1.189184,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.0185137786878254e-05,
+      "loss": 0.4263,
+      "step": 4646
+    },
+    {
+      "epoch": 1.18944,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.018087193925433e-05,
+      "loss": 0.4608,
+      "step": 4647
+    },
+    {
+      "epoch": 1.189696,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.017660609163041e-05,
+      "loss": 0.4329,
+      "step": 4648
+    },
+    {
+      "epoch": 1.189952,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.0172340244006487e-05,
+      "loss": 0.4984,
+      "step": 4649
+    },
+    {
+      "epoch": 1.190208,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.016807439638256e-05,
+      "loss": 0.4654,
+      "step": 4650
+    },
+    {
+      "epoch": 1.190464,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.016380854875864e-05,
+      "loss": 0.456,
+      "step": 4651
+    },
+    {
+      "epoch": 1.19072,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.015954270113472e-05,
+      "loss": 0.4896,
+      "step": 4652
+    },
+    {
+      "epoch": 1.190976,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.0155276853510794e-05,
+      "loss": 0.4857,
+      "step": 4653
+    },
+    {
+      "epoch": 1.191232,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.015101100588687e-05,
+      "loss": 0.4679,
+      "step": 4654
+    },
+    {
+      "epoch": 1.191488,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.0146745158262946e-05,
+      "loss": 0.4847,
+      "step": 4655
+    },
+    {
+      "epoch": 1.191744,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.0142479310639027e-05,
+      "loss": 0.4762,
+      "step": 4656
+    },
+    {
+      "epoch": 1.192,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.0138213463015102e-05,
+      "loss": 0.4202,
+      "step": 4657
+    },
+    {
+      "epoch": 1.192256,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.013394761539118e-05,
+      "loss": 0.4186,
+      "step": 4658
+    },
+    {
+      "epoch": 1.192512,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.012968176776726e-05,
+      "loss": 0.4607,
+      "step": 4659
+    },
+    {
+      "epoch": 1.192768,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.0125415920143335e-05,
+      "loss": 0.4498,
+      "step": 4660
+    },
+    {
+      "epoch": 1.193024,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.0121150072519412e-05,
+      "loss": 0.4955,
+      "step": 4661
+    },
+    {
+      "epoch": 1.1932800000000001,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.0116884224895487e-05,
+      "loss": 0.4611,
+      "step": 4662
+    },
+    {
+      "epoch": 1.193536,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.0112618377271568e-05,
+      "loss": 0.4334,
+      "step": 4663
+    },
+    {
+      "epoch": 1.193792,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.0108352529647642e-05,
+      "loss": 0.456,
+      "step": 4664
+    },
+    {
+      "epoch": 1.194048,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.010408668202372e-05,
+      "loss": 0.4636,
+      "step": 4665
+    },
+    {
+      "epoch": 1.194304,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.0099820834399794e-05,
+      "loss": 0.4753,
+      "step": 4666
+    },
+    {
+      "epoch": 1.19456,
+      "grad_norm": 2.25,
+      "learning_rate": 3.0095554986775875e-05,
+      "loss": 0.4647,
+      "step": 4667
+    },
+    {
+      "epoch": 1.1948159999999999,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.0091289139151953e-05,
+      "loss": 0.4314,
+      "step": 4668
+    },
+    {
+      "epoch": 1.195072,
+      "grad_norm": 1.9609375,
+      "learning_rate": 3.0087023291528027e-05,
+      "loss": 0.4687,
+      "step": 4669
+    },
+    {
+      "epoch": 1.195328,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.00827574439041e-05,
+      "loss": 0.4289,
+      "step": 4670
+    },
+    {
+      "epoch": 1.195584,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.0078491596280185e-05,
+      "loss": 0.4406,
+      "step": 4671
+    },
+    {
+      "epoch": 1.19584,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.007422574865626e-05,
+      "loss": 0.4896,
+      "step": 4672
+    },
+    {
+      "epoch": 1.196096,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.0069959901032334e-05,
+      "loss": 0.4505,
+      "step": 4673
+    },
+    {
+      "epoch": 1.196352,
+      "grad_norm": 2.25,
+      "learning_rate": 3.0065694053408415e-05,
+      "loss": 0.4598,
+      "step": 4674
+    },
+    {
+      "epoch": 1.196608,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.0061428205784493e-05,
+      "loss": 0.4257,
+      "step": 4675
+    },
+    {
+      "epoch": 1.196864,
+      "grad_norm": 1.9921875,
+      "learning_rate": 3.0057162358160567e-05,
+      "loss": 0.4857,
+      "step": 4676
+    },
+    {
+      "epoch": 1.19712,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.0052896510536645e-05,
+      "loss": 0.4161,
+      "step": 4677
+    },
+    {
+      "epoch": 1.197376,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.0048630662912726e-05,
+      "loss": 0.4241,
+      "step": 4678
+    },
+    {
+      "epoch": 1.197632,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.00443648152888e-05,
+      "loss": 0.4789,
+      "step": 4679
+    },
+    {
+      "epoch": 1.197888,
+      "grad_norm": 1.9765625,
+      "learning_rate": 3.0040098967664874e-05,
+      "loss": 0.4597,
+      "step": 4680
+    },
+    {
+      "epoch": 1.198144,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.0035833120040952e-05,
+      "loss": 0.476,
+      "step": 4681
+    },
+    {
+      "epoch": 1.1984,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.0031567272417033e-05,
+      "loss": 0.4789,
+      "step": 4682
+    },
+    {
+      "epoch": 1.198656,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.0027301424793107e-05,
+      "loss": 0.4481,
+      "step": 4683
+    },
+    {
+      "epoch": 1.198912,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.0023035577169185e-05,
+      "loss": 0.491,
+      "step": 4684
+    },
+    {
+      "epoch": 1.199168,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.001876972954526e-05,
+      "loss": 0.4427,
+      "step": 4685
+    },
+    {
+      "epoch": 1.199424,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.001450388192134e-05,
+      "loss": 0.4521,
+      "step": 4686
+    },
+    {
+      "epoch": 1.19968,
+      "grad_norm": 2.375,
+      "learning_rate": 3.0010238034297418e-05,
+      "loss": 0.4826,
+      "step": 4687
+    },
+    {
+      "epoch": 1.1999360000000001,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.0005972186673492e-05,
+      "loss": 0.4503,
+      "step": 4688
+    },
+    {
+      "epoch": 1.200192,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.0001706339049566e-05,
+      "loss": 0.4591,
+      "step": 4689
+    },
+    {
+      "epoch": 1.200448,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.999744049142565e-05,
+      "loss": 0.4714,
+      "step": 4690
+    },
+    {
+      "epoch": 1.200704,
+      "grad_norm": 2.390625,
+      "learning_rate": 2.9993174643801725e-05,
+      "loss": 0.4688,
+      "step": 4691
+    },
+    {
+      "epoch": 1.20096,
+      "grad_norm": 2.21875,
+      "learning_rate": 2.99889087961778e-05,
+      "loss": 0.4449,
+      "step": 4692
+    },
+    {
+      "epoch": 1.201216,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.998464294855388e-05,
+      "loss": 0.4897,
+      "step": 4693
+    },
+    {
+      "epoch": 1.201472,
+      "grad_norm": 2.34375,
+      "learning_rate": 2.9980377100929958e-05,
+      "loss": 0.4622,
+      "step": 4694
+    },
+    {
+      "epoch": 1.201728,
+      "grad_norm": 2.296875,
+      "learning_rate": 2.9976111253306032e-05,
+      "loss": 0.4744,
+      "step": 4695
+    },
+    {
+      "epoch": 1.201984,
+      "grad_norm": 2.421875,
+      "learning_rate": 2.997184540568211e-05,
+      "loss": 0.4409,
+      "step": 4696
+    },
+    {
+      "epoch": 1.20224,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.996757955805819e-05,
+      "loss": 0.4737,
+      "step": 4697
+    },
+    {
+      "epoch": 1.202496,
+      "grad_norm": 2.28125,
+      "learning_rate": 2.9963313710434265e-05,
+      "loss": 0.4615,
+      "step": 4698
+    },
+    {
+      "epoch": 1.202752,
+      "grad_norm": 2.3125,
+      "learning_rate": 2.995904786281034e-05,
+      "loss": 0.4454,
+      "step": 4699
+    },
+    {
+      "epoch": 1.203008,
+      "grad_norm": 2.34375,
+      "learning_rate": 2.9954782015186417e-05,
+      "loss": 0.4756,
+      "step": 4700
+    },
+    {
+      "epoch": 1.203264,
+      "grad_norm": 2.078125,
+      "learning_rate": 2.9950516167562498e-05,
+      "loss": 0.4477,
+      "step": 4701
+    },
+    {
+      "epoch": 1.20352,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.9946250319938572e-05,
+      "loss": 0.4959,
+      "step": 4702
+    },
+    {
+      "epoch": 1.203776,
+      "grad_norm": 2.34375,
+      "learning_rate": 2.994198447231465e-05,
+      "loss": 0.469,
+      "step": 4703
+    },
+    {
+      "epoch": 1.204032,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.9937718624690724e-05,
+      "loss": 0.4216,
+      "step": 4704
+    },
+    {
+      "epoch": 1.204288,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.9933452777066805e-05,
+      "loss": 0.4475,
+      "step": 4705
+    },
+    {
+      "epoch": 1.204544,
+      "grad_norm": 2.3125,
+      "learning_rate": 2.9929186929442883e-05,
+      "loss": 0.4809,
+      "step": 4706
+    },
+    {
+      "epoch": 1.2048,
+      "grad_norm": 2.21875,
+      "learning_rate": 2.9924921081818957e-05,
+      "loss": 0.4802,
+      "step": 4707
+    },
+    {
+      "epoch": 1.205056,
+      "grad_norm": 2.03125,
+      "learning_rate": 2.992065523419504e-05,
+      "loss": 0.4683,
+      "step": 4708
+    },
+    {
+      "epoch": 1.205312,
+      "grad_norm": 2.078125,
+      "learning_rate": 2.9916389386571113e-05,
+      "loss": 0.4861,
+      "step": 4709
+    },
+    {
+      "epoch": 1.205568,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.991212353894719e-05,
+      "loss": 0.4673,
+      "step": 4710
+    },
+    {
+      "epoch": 1.205824,
+      "grad_norm": 1.828125,
+      "learning_rate": 2.9907857691323265e-05,
+      "loss": 0.4619,
+      "step": 4711
+    },
+    {
+      "epoch": 1.20608,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.9903591843699346e-05,
+      "loss": 0.4481,
+      "step": 4712
+    },
+    {
+      "epoch": 1.206336,
+      "grad_norm": 1.953125,
+      "learning_rate": 2.9899325996075423e-05,
+      "loss": 0.4566,
+      "step": 4713
+    },
+    {
+      "epoch": 1.206592,
+      "grad_norm": 2.078125,
+      "learning_rate": 2.9895060148451498e-05,
+      "loss": 0.4613,
+      "step": 4714
+    },
+    {
+      "epoch": 1.206848,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.9890794300827575e-05,
+      "loss": 0.4808,
+      "step": 4715
+    },
+    {
+      "epoch": 1.207104,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.9886528453203656e-05,
+      "loss": 0.4753,
+      "step": 4716
+    },
+    {
+      "epoch": 1.20736,
+      "grad_norm": 2.03125,
+      "learning_rate": 2.988226260557973e-05,
+      "loss": 0.4613,
+      "step": 4717
+    },
+    {
+      "epoch": 1.207616,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.9877996757955805e-05,
+      "loss": 0.4183,
+      "step": 4718
+    },
+    {
+      "epoch": 1.207872,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.9873730910331882e-05,
+      "loss": 0.4493,
+      "step": 4719
+    },
+    {
+      "epoch": 1.208128,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.9869465062707963e-05,
+      "loss": 0.4341,
+      "step": 4720
+    },
+    {
+      "epoch": 1.2083840000000001,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.9865199215084038e-05,
+      "loss": 0.4377,
+      "step": 4721
+    },
+    {
+      "epoch": 1.20864,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.9860933367460115e-05,
+      "loss": 0.5218,
+      "step": 4722
+    },
+    {
+      "epoch": 1.208896,
+      "grad_norm": 2.125,
+      "learning_rate": 2.9856667519836196e-05,
+      "loss": 0.4269,
+      "step": 4723
+    },
+    {
+      "epoch": 1.209152,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.985240167221227e-05,
+      "loss": 0.4421,
+      "step": 4724
+    },
+    {
+      "epoch": 1.209408,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.984813582458835e-05,
+      "loss": 0.4814,
+      "step": 4725
+    },
+    {
+      "epoch": 1.209664,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.9843869976964423e-05,
+      "loss": 0.4216,
+      "step": 4726
+    },
+    {
+      "epoch": 1.2099199999999999,
+      "grad_norm": 2.28125,
+      "learning_rate": 2.9839604129340504e-05,
+      "loss": 0.4397,
+      "step": 4727
+    },
+    {
+      "epoch": 1.210176,
+      "grad_norm": 2.359375,
+      "learning_rate": 2.9835338281716578e-05,
+      "loss": 0.4471,
+      "step": 4728
+    },
+    {
+      "epoch": 1.210432,
+      "grad_norm": 2.296875,
+      "learning_rate": 2.9831072434092656e-05,
+      "loss": 0.459,
+      "step": 4729
+    },
+    {
+      "epoch": 1.210688,
+      "grad_norm": 2.25,
+      "learning_rate": 2.982680658646873e-05,
+      "loss": 0.4364,
+      "step": 4730
+    },
+    {
+      "epoch": 1.210944,
+      "grad_norm": 2.125,
+      "learning_rate": 2.982254073884481e-05,
+      "loss": 0.472,
+      "step": 4731
+    },
+    {
+      "epoch": 1.2112,
+      "grad_norm": 2.25,
+      "learning_rate": 2.981827489122089e-05,
+      "loss": 0.4309,
+      "step": 4732
+    },
+    {
+      "epoch": 1.211456,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.9814009043596963e-05,
+      "loss": 0.4547,
+      "step": 4733
+    },
+    {
+      "epoch": 1.211712,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.980974319597304e-05,
+      "loss": 0.4541,
+      "step": 4734
+    },
+    {
+      "epoch": 1.211968,
+      "grad_norm": 2.078125,
+      "learning_rate": 2.980547734834912e-05,
+      "loss": 0.4612,
+      "step": 4735
+    },
+    {
+      "epoch": 1.212224,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.9801211500725196e-05,
+      "loss": 0.4292,
+      "step": 4736
+    },
+    {
+      "epoch": 1.21248,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.979694565310127e-05,
+      "loss": 0.4794,
+      "step": 4737
+    },
+    {
+      "epoch": 1.212736,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.9792679805477348e-05,
+      "loss": 0.4912,
+      "step": 4738
+    },
+    {
+      "epoch": 1.212992,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.978841395785343e-05,
+      "loss": 0.4537,
+      "step": 4739
+    },
+    {
+      "epoch": 1.213248,
+      "grad_norm": 2.125,
+      "learning_rate": 2.9784148110229503e-05,
+      "loss": 0.4765,
+      "step": 4740
+    },
+    {
+      "epoch": 1.213504,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.977988226260558e-05,
+      "loss": 0.5052,
+      "step": 4741
+    },
+    {
+      "epoch": 1.21376,
+      "grad_norm": 2.078125,
+      "learning_rate": 2.977561641498166e-05,
+      "loss": 0.4686,
+      "step": 4742
+    },
+    {
+      "epoch": 1.214016,
+      "grad_norm": 2.03125,
+      "learning_rate": 2.9771350567357736e-05,
+      "loss": 0.4423,
+      "step": 4743
+    },
+    {
+      "epoch": 1.214272,
+      "grad_norm": 2.03125,
+      "learning_rate": 2.9767084719733814e-05,
+      "loss": 0.4155,
+      "step": 4744
+    },
+    {
+      "epoch": 1.214528,
+      "grad_norm": 2.25,
+      "learning_rate": 2.9762818872109888e-05,
+      "loss": 0.4607,
+      "step": 4745
+    },
+    {
+      "epoch": 1.214784,
+      "grad_norm": 2.25,
+      "learning_rate": 2.975855302448597e-05,
+      "loss": 0.4877,
+      "step": 4746
+    },
+    {
+      "epoch": 1.2150400000000001,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.9754287176862043e-05,
+      "loss": 0.4836,
+      "step": 4747
+    },
+    {
+      "epoch": 1.215296,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.975002132923812e-05,
+      "loss": 0.449,
+      "step": 4748
+    },
+    {
+      "epoch": 1.215552,
+      "grad_norm": 2.125,
+      "learning_rate": 2.9745755481614195e-05,
+      "loss": 0.4641,
+      "step": 4749
+    },
+    {
+      "epoch": 1.215808,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.9741489633990276e-05,
+      "loss": 0.4996,
+      "step": 4750
+    },
+    {
+      "epoch": 1.216064,
+      "grad_norm": 1.96875,
+      "learning_rate": 2.9737223786366354e-05,
+      "loss": 0.4644,
+      "step": 4751
+    },
+    {
+      "epoch": 1.21632,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.9732957938742428e-05,
+      "loss": 0.4845,
+      "step": 4752
+    },
+    {
+      "epoch": 1.2165759999999999,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.9728692091118506e-05,
+      "loss": 0.4866,
+      "step": 4753
+    },
+    {
+      "epoch": 1.216832,
+      "grad_norm": 2.03125,
+      "learning_rate": 2.9724426243494587e-05,
+      "loss": 0.4271,
+      "step": 4754
+    },
+    {
+      "epoch": 1.217088,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.972016039587066e-05,
+      "loss": 0.4583,
+      "step": 4755
+    },
+    {
+      "epoch": 1.217344,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.9715894548246735e-05,
+      "loss": 0.5022,
+      "step": 4756
+    },
+    {
+      "epoch": 1.2176,
+      "grad_norm": 2.078125,
+      "learning_rate": 2.9711628700622816e-05,
+      "loss": 0.4357,
+      "step": 4757
+    },
+    {
+      "epoch": 1.217856,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.9707362852998894e-05,
+      "loss": 0.44,
+      "step": 4758
+    },
+    {
+      "epoch": 1.218112,
+      "grad_norm": 2.375,
+      "learning_rate": 2.9703097005374968e-05,
+      "loss": 0.4694,
+      "step": 4759
+    },
+    {
+      "epoch": 1.218368,
+      "grad_norm": 2.28125,
+      "learning_rate": 2.9698831157751046e-05,
+      "loss": 0.4589,
+      "step": 4760
+    },
+    {
+      "epoch": 1.218624,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.9694565310127127e-05,
+      "loss": 0.5157,
+      "step": 4761
+    },
+    {
+      "epoch": 1.21888,
+      "grad_norm": 2.015625,
+      "learning_rate": 2.96902994625032e-05,
+      "loss": 0.4576,
+      "step": 4762
+    },
+    {
+      "epoch": 1.219136,
+      "grad_norm": 1.9921875,
+      "learning_rate": 2.968603361487928e-05,
+      "loss": 0.4269,
+      "step": 4763
+    },
+    {
+      "epoch": 1.219392,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.9681767767255353e-05,
+      "loss": 0.4527,
+      "step": 4764
+    },
+    {
+      "epoch": 1.219648,
+      "grad_norm": 2.03125,
+      "learning_rate": 2.9677501919631434e-05,
+      "loss": 0.4686,
+      "step": 4765
+    },
+    {
+      "epoch": 1.219904,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.967323607200751e-05,
+      "loss": 0.4544,
+      "step": 4766
+    },
+    {
+      "epoch": 1.22016,
+      "grad_norm": 2.03125,
+      "learning_rate": 2.9668970224383586e-05,
+      "loss": 0.4468,
+      "step": 4767
+    },
+    {
+      "epoch": 1.220416,
+      "grad_norm": 2.28125,
+      "learning_rate": 2.966470437675966e-05,
+      "loss": 0.4407,
+      "step": 4768
+    },
+    {
+      "epoch": 1.220672,
+      "grad_norm": 2.015625,
+      "learning_rate": 2.966043852913574e-05,
+      "loss": 0.4622,
+      "step": 4769
+    },
+    {
+      "epoch": 1.220928,
+      "grad_norm": 2.375,
+      "learning_rate": 2.965617268151182e-05,
+      "loss": 0.4663,
+      "step": 4770
+    },
+    {
+      "epoch": 1.221184,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.9651906833887893e-05,
+      "loss": 0.4784,
+      "step": 4771
+    },
+    {
+      "epoch": 1.22144,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.9647640986263974e-05,
+      "loss": 0.4365,
+      "step": 4772
+    },
+    {
+      "epoch": 1.2216960000000001,
+      "grad_norm": 2.359375,
+      "learning_rate": 2.9643375138640052e-05,
+      "loss": 0.4734,
+      "step": 4773
+    },
+    {
+      "epoch": 1.221952,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.9639109291016126e-05,
+      "loss": 0.4382,
+      "step": 4774
+    },
+    {
+      "epoch": 1.222208,
+      "grad_norm": 2.015625,
+      "learning_rate": 2.96348434433922e-05,
+      "loss": 0.4366,
+      "step": 4775
+    },
+    {
+      "epoch": 1.222464,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.963057759576828e-05,
+      "loss": 0.4856,
+      "step": 4776
+    },
+    {
+      "epoch": 1.22272,
+      "grad_norm": 2.28125,
+      "learning_rate": 2.962631174814436e-05,
+      "loss": 0.4613,
+      "step": 4777
+    },
+    {
+      "epoch": 1.222976,
+      "grad_norm": 2.359375,
+      "learning_rate": 2.9622045900520434e-05,
+      "loss": 0.4536,
+      "step": 4778
+    },
+    {
+      "epoch": 1.223232,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.961778005289651e-05,
+      "loss": 0.4447,
+      "step": 4779
+    },
+    {
+      "epoch": 1.223488,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.9613514205272592e-05,
+      "loss": 0.4245,
+      "step": 4780
+    },
+    {
+      "epoch": 1.223744,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.9609248357648666e-05,
+      "loss": 0.483,
+      "step": 4781
+    },
+    {
+      "epoch": 1.224,
+      "grad_norm": 2.390625,
+      "learning_rate": 2.9604982510024744e-05,
+      "loss": 0.4334,
+      "step": 4782
+    },
+    {
+      "epoch": 1.224256,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.960071666240082e-05,
+      "loss": 0.4401,
+      "step": 4783
+    },
+    {
+      "epoch": 1.224512,
+      "grad_norm": 2.25,
+      "learning_rate": 2.95964508147769e-05,
+      "loss": 0.4561,
+      "step": 4784
+    },
+    {
+      "epoch": 1.224768,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.9592184967152974e-05,
+      "loss": 0.4179,
+      "step": 4785
+    },
+    {
+      "epoch": 1.225024,
+      "grad_norm": 2.03125,
+      "learning_rate": 2.958791911952905e-05,
+      "loss": 0.4821,
+      "step": 4786
+    },
+    {
+      "epoch": 1.22528,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.9583653271905126e-05,
+      "loss": 0.4464,
+      "step": 4787
+    },
+    {
+      "epoch": 1.225536,
+      "grad_norm": 2.0625,
+      "learning_rate": 2.9579387424281207e-05,
+      "loss": 0.4542,
+      "step": 4788
+    },
+    {
+      "epoch": 1.225792,
+      "grad_norm": 2.390625,
+      "learning_rate": 2.9575121576657284e-05,
+      "loss": 0.4371,
+      "step": 4789
+    },
+    {
+      "epoch": 1.226048,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.957085572903336e-05,
+      "loss": 0.4835,
+      "step": 4790
+    },
+    {
+      "epoch": 1.226304,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.956658988140944e-05,
+      "loss": 0.4506,
+      "step": 4791
+    },
+    {
+      "epoch": 1.22656,
+      "grad_norm": 2.015625,
+      "learning_rate": 2.9562324033785517e-05,
+      "loss": 0.492,
+      "step": 4792
+    },
+    {
+      "epoch": 1.226816,
+      "grad_norm": 2.0,
+      "learning_rate": 2.955805818616159e-05,
+      "loss": 0.4553,
+      "step": 4793
+    },
+    {
+      "epoch": 1.227072,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.9553792338537666e-05,
+      "loss": 0.4591,
+      "step": 4794
+    },
+    {
+      "epoch": 1.227328,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.9549526490913747e-05,
+      "loss": 0.4531,
+      "step": 4795
+    },
+    {
+      "epoch": 1.227584,
+      "grad_norm": 2.125,
+      "learning_rate": 2.9545260643289825e-05,
+      "loss": 0.4488,
+      "step": 4796
+    },
+    {
+      "epoch": 1.22784,
+      "grad_norm": 2.03125,
+      "learning_rate": 2.95409947956659e-05,
+      "loss": 0.4322,
+      "step": 4797
+    },
+    {
+      "epoch": 1.228096,
+      "grad_norm": 2.25,
+      "learning_rate": 2.9536728948041976e-05,
+      "loss": 0.46,
+      "step": 4798
+    },
+    {
+      "epoch": 1.228352,
+      "grad_norm": 1.875,
+      "learning_rate": 2.9532463100418057e-05,
+      "loss": 0.4684,
+      "step": 4799
+    },
+    {
+      "epoch": 1.228608,
+      "grad_norm": 2.125,
+      "learning_rate": 2.9528197252794132e-05,
+      "loss": 0.4682,
+      "step": 4800
+    },
+    {
+      "epoch": 1.228864,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.9523931405170206e-05,
+      "loss": 0.4107,
+      "step": 4801
+    },
+    {
+      "epoch": 1.22912,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.9519665557546284e-05,
+      "loss": 0.4478,
+      "step": 4802
+    },
+    {
+      "epoch": 1.229376,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.9515399709922365e-05,
+      "loss": 0.4691,
+      "step": 4803
+    },
+    {
+      "epoch": 1.229632,
+      "grad_norm": 2.0625,
+      "learning_rate": 2.951113386229844e-05,
+      "loss": 0.432,
+      "step": 4804
+    },
+    {
+      "epoch": 1.229888,
+      "grad_norm": 2.375,
+      "learning_rate": 2.9506868014674517e-05,
+      "loss": 0.4668,
+      "step": 4805
+    },
+    {
+      "epoch": 1.230144,
+      "grad_norm": 2.03125,
+      "learning_rate": 2.9502602167050598e-05,
+      "loss": 0.456,
+      "step": 4806
+    },
+    {
+      "epoch": 1.2304,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.9498336319426672e-05,
+      "loss": 0.4524,
+      "step": 4807
+    },
+    {
+      "epoch": 1.230656,
+      "grad_norm": 2.25,
+      "learning_rate": 2.949407047180275e-05,
+      "loss": 0.4786,
+      "step": 4808
+    },
+    {
+      "epoch": 1.230912,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.9489804624178824e-05,
+      "loss": 0.4654,
+      "step": 4809
+    },
+    {
+      "epoch": 1.231168,
+      "grad_norm": 2.21875,
+      "learning_rate": 2.9485538776554905e-05,
+      "loss": 0.4558,
+      "step": 4810
+    },
+    {
+      "epoch": 1.231424,
+      "grad_norm": 2.078125,
+      "learning_rate": 2.948127292893098e-05,
+      "loss": 0.4303,
+      "step": 4811
+    },
+    {
+      "epoch": 1.2316799999999999,
+      "grad_norm": 2.015625,
+      "learning_rate": 2.9477007081307057e-05,
+      "loss": 0.487,
+      "step": 4812
+    },
+    {
+      "epoch": 1.231936,
+      "grad_norm": 2.3125,
+      "learning_rate": 2.947274123368313e-05,
+      "loss": 0.4581,
+      "step": 4813
+    },
+    {
+      "epoch": 1.232192,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.9468475386059212e-05,
+      "loss": 0.4603,
+      "step": 4814
+    },
+    {
+      "epoch": 1.232448,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.946420953843529e-05,
+      "loss": 0.4919,
+      "step": 4815
+    },
+    {
+      "epoch": 1.232704,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.9459943690811364e-05,
+      "loss": 0.4479,
+      "step": 4816
+    },
+    {
+      "epoch": 1.23296,
+      "grad_norm": 2.28125,
+      "learning_rate": 2.9455677843187442e-05,
+      "loss": 0.4529,
+      "step": 4817
+    },
+    {
+      "epoch": 1.233216,
+      "grad_norm": 2.0,
+      "learning_rate": 2.9451411995563523e-05,
+      "loss": 0.5019,
+      "step": 4818
+    },
+    {
+      "epoch": 1.233472,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.9447146147939597e-05,
+      "loss": 0.4711,
+      "step": 4819
+    },
+    {
+      "epoch": 1.233728,
+      "grad_norm": 2.265625,
+      "learning_rate": 2.944288030031567e-05,
+      "loss": 0.4423,
+      "step": 4820
+    },
+    {
+      "epoch": 1.233984,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.9438614452691752e-05,
+      "loss": 0.5155,
+      "step": 4821
+    },
+    {
+      "epoch": 1.23424,
+      "grad_norm": 2.0625,
+      "learning_rate": 2.943434860506783e-05,
+      "loss": 0.4493,
+      "step": 4822
+    },
+    {
+      "epoch": 1.234496,
+      "grad_norm": 2.078125,
+      "learning_rate": 2.9430082757443904e-05,
+      "loss": 0.4703,
+      "step": 4823
+    },
+    {
+      "epoch": 1.234752,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.9425816909819982e-05,
+      "loss": 0.4432,
+      "step": 4824
+    },
+    {
+      "epoch": 1.235008,
+      "grad_norm": 2.125,
+      "learning_rate": 2.9421551062196063e-05,
+      "loss": 0.4603,
+      "step": 4825
+    },
+    {
+      "epoch": 1.235264,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.9417285214572137e-05,
+      "loss": 0.464,
+      "step": 4826
+    },
+    {
+      "epoch": 1.23552,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.9413019366948215e-05,
+      "loss": 0.4624,
+      "step": 4827
+    },
+    {
+      "epoch": 1.235776,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.940875351932429e-05,
+      "loss": 0.4464,
+      "step": 4828
+    },
+    {
+      "epoch": 1.236032,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.940448767170037e-05,
+      "loss": 0.4304,
+      "step": 4829
+    },
+    {
+      "epoch": 1.236288,
+      "grad_norm": 2.0625,
+      "learning_rate": 2.9400221824076444e-05,
+      "loss": 0.4647,
+      "step": 4830
+    },
+    {
+      "epoch": 1.236544,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.9395955976452522e-05,
+      "loss": 0.4497,
+      "step": 4831
+    },
+    {
+      "epoch": 1.2368000000000001,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.9391690128828596e-05,
+      "loss": 0.4336,
+      "step": 4832
+    },
+    {
+      "epoch": 1.237056,
+      "grad_norm": 1.9921875,
+      "learning_rate": 2.9387424281204677e-05,
+      "loss": 0.4935,
+      "step": 4833
+    },
+    {
+      "epoch": 1.237312,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.9383158433580755e-05,
+      "loss": 0.4624,
+      "step": 4834
+    },
+    {
+      "epoch": 1.237568,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.937889258595683e-05,
+      "loss": 0.4787,
+      "step": 4835
+    },
+    {
+      "epoch": 1.237824,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.9374626738332907e-05,
+      "loss": 0.4311,
+      "step": 4836
+    },
+    {
+      "epoch": 1.23808,
+      "grad_norm": 2.25,
+      "learning_rate": 2.9370360890708988e-05,
+      "loss": 0.4855,
+      "step": 4837
+    },
+    {
+      "epoch": 1.2383359999999999,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.9366095043085062e-05,
+      "loss": 0.436,
+      "step": 4838
+    },
+    {
+      "epoch": 1.238592,
+      "grad_norm": 2.125,
+      "learning_rate": 2.9361829195461137e-05,
+      "loss": 0.4563,
+      "step": 4839
+    },
+    {
+      "epoch": 1.238848,
+      "grad_norm": 2.25,
+      "learning_rate": 2.9357563347837218e-05,
+      "loss": 0.4903,
+      "step": 4840
+    },
+    {
+      "epoch": 1.239104,
+      "grad_norm": 2.21875,
+      "learning_rate": 2.9353297500213295e-05,
+      "loss": 0.4745,
+      "step": 4841
+    },
+    {
+      "epoch": 1.23936,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.934903165258937e-05,
+      "loss": 0.482,
+      "step": 4842
+    },
+    {
+      "epoch": 1.239616,
+      "grad_norm": 2.28125,
+      "learning_rate": 2.9344765804965447e-05,
+      "loss": 0.4392,
+      "step": 4843
+    },
+    {
+      "epoch": 1.239872,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.9340499957341528e-05,
+      "loss": 0.4886,
+      "step": 4844
+    },
+    {
+      "epoch": 1.240128,
+      "grad_norm": 2.28125,
+      "learning_rate": 2.9336234109717602e-05,
+      "loss": 0.4782,
+      "step": 4845
+    },
+    {
+      "epoch": 1.240384,
+      "grad_norm": 2.03125,
+      "learning_rate": 2.933196826209368e-05,
+      "loss": 0.4379,
+      "step": 4846
+    },
+    {
+      "epoch": 1.24064,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.9327702414469754e-05,
+      "loss": 0.4593,
+      "step": 4847
+    },
+    {
+      "epoch": 1.240896,
+      "grad_norm": 2.421875,
+      "learning_rate": 2.9323436566845835e-05,
+      "loss": 0.4733,
+      "step": 4848
+    },
+    {
+      "epoch": 1.241152,
+      "grad_norm": 2.0,
+      "learning_rate": 2.931917071922191e-05,
+      "loss": 0.4479,
+      "step": 4849
+    },
+    {
+      "epoch": 1.241408,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.9314904871597987e-05,
+      "loss": 0.5144,
+      "step": 4850
+    },
+    {
+      "epoch": 1.241664,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.931063902397406e-05,
+      "loss": 0.4681,
+      "step": 4851
+    },
+    {
+      "epoch": 1.24192,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.9306373176350143e-05,
+      "loss": 0.4681,
+      "step": 4852
+    },
+    {
+      "epoch": 1.242176,
+      "grad_norm": 1.9765625,
+      "learning_rate": 2.930210732872622e-05,
+      "loss": 0.4387,
+      "step": 4853
+    },
+    {
+      "epoch": 1.242432,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.9297841481102295e-05,
+      "loss": 0.4442,
+      "step": 4854
+    },
+    {
+      "epoch": 1.242688,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.9293575633478376e-05,
+      "loss": 0.4461,
+      "step": 4855
+    },
+    {
+      "epoch": 1.242944,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.9289309785854453e-05,
+      "loss": 0.4767,
+      "step": 4856
+    },
+    {
+      "epoch": 1.2432,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.9285043938230528e-05,
+      "loss": 0.4674,
+      "step": 4857
+    },
+    {
+      "epoch": 1.2434560000000001,
+      "grad_norm": 1.9765625,
+      "learning_rate": 2.9280778090606602e-05,
+      "loss": 0.4803,
+      "step": 4858
+    },
+    {
+      "epoch": 1.243712,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.9276512242982683e-05,
+      "loss": 0.4677,
+      "step": 4859
+    },
+    {
+      "epoch": 1.243968,
+      "grad_norm": 2.265625,
+      "learning_rate": 2.927224639535876e-05,
+      "loss": 0.4906,
+      "step": 4860
+    },
+    {
+      "epoch": 1.244224,
+      "grad_norm": 2.328125,
+      "learning_rate": 2.9267980547734835e-05,
+      "loss": 0.4491,
+      "step": 4861
+    },
+    {
+      "epoch": 1.24448,
+      "grad_norm": 2.421875,
+      "learning_rate": 2.9263714700110912e-05,
+      "loss": 0.4675,
+      "step": 4862
+    },
+    {
+      "epoch": 1.244736,
+      "grad_norm": 2.359375,
+      "learning_rate": 2.9259448852486993e-05,
+      "loss": 0.374,
+      "step": 4863
+    },
+    {
+      "epoch": 1.2449919999999999,
+      "grad_norm": 2.03125,
+      "learning_rate": 2.9255183004863068e-05,
+      "loss": 0.4874,
+      "step": 4864
+    },
+    {
+      "epoch": 1.245248,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.9250917157239145e-05,
+      "loss": 0.4231,
+      "step": 4865
+    },
+    {
+      "epoch": 1.245504,
+      "grad_norm": 2.359375,
+      "learning_rate": 2.924665130961522e-05,
+      "loss": 0.4353,
+      "step": 4866
+    },
+    {
+      "epoch": 1.24576,
+      "grad_norm": 2.375,
+      "learning_rate": 2.92423854619913e-05,
+      "loss": 0.4292,
+      "step": 4867
+    },
+    {
+      "epoch": 1.246016,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.9238119614367375e-05,
+      "loss": 0.4695,
+      "step": 4868
+    },
+    {
+      "epoch": 1.246272,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.9233853766743453e-05,
+      "loss": 0.4673,
+      "step": 4869
+    },
+    {
+      "epoch": 1.246528,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.9229587919119534e-05,
+      "loss": 0.4458,
+      "step": 4870
+    },
+    {
+      "epoch": 1.246784,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.9225322071495608e-05,
+      "loss": 0.4351,
+      "step": 4871
+    },
+    {
+      "epoch": 1.24704,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.9221056223871686e-05,
+      "loss": 0.488,
+      "step": 4872
+    },
+    {
+      "epoch": 1.247296,
+      "grad_norm": 2.21875,
+      "learning_rate": 2.921679037624776e-05,
+      "loss": 0.466,
+      "step": 4873
+    },
+    {
+      "epoch": 1.247552,
+      "grad_norm": 2.25,
+      "learning_rate": 2.921252452862384e-05,
+      "loss": 0.4591,
+      "step": 4874
+    },
+    {
+      "epoch": 1.247808,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.920825868099992e-05,
+      "loss": 0.4663,
+      "step": 4875
+    },
+    {
+      "epoch": 1.248064,
+      "grad_norm": 2.0625,
+      "learning_rate": 2.9203992833375993e-05,
+      "loss": 0.4446,
+      "step": 4876
+    },
+    {
+      "epoch": 1.24832,
+      "grad_norm": 2.078125,
+      "learning_rate": 2.9199726985752067e-05,
+      "loss": 0.4467,
+      "step": 4877
+    },
+    {
+      "epoch": 1.248576,
+      "grad_norm": 2.328125,
+      "learning_rate": 2.9195461138128148e-05,
+      "loss": 0.4314,
+      "step": 4878
+    },
+    {
+      "epoch": 1.248832,
+      "grad_norm": 2.28125,
+      "learning_rate": 2.9191195290504226e-05,
+      "loss": 0.469,
+      "step": 4879
+    },
+    {
+      "epoch": 1.249088,
+      "grad_norm": 2.296875,
+      "learning_rate": 2.91869294428803e-05,
+      "loss": 0.4564,
+      "step": 4880
+    },
+    {
+      "epoch": 1.249344,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.9182663595256378e-05,
+      "loss": 0.4283,
+      "step": 4881
+    },
+    {
+      "epoch": 1.2496,
+      "grad_norm": 2.0625,
+      "learning_rate": 2.917839774763246e-05,
+      "loss": 0.4682,
+      "step": 4882
+    },
+    {
+      "epoch": 1.249856,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.9174131900008533e-05,
+      "loss": 0.4861,
+      "step": 4883
+    },
+    {
+      "epoch": 1.2501120000000001,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.916986605238461e-05,
+      "loss": 0.4028,
+      "step": 4884
+    },
+    {
+      "epoch": 1.250368,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.9165600204760685e-05,
+      "loss": 0.4377,
+      "step": 4885
+    },
+    {
+      "epoch": 1.250624,
+      "grad_norm": 2.0,
+      "learning_rate": 2.9161334357136766e-05,
+      "loss": 0.4436,
+      "step": 4886
+    },
+    {
+      "epoch": 1.25088,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.915706850951284e-05,
+      "loss": 0.4368,
+      "step": 4887
+    },
+    {
+      "epoch": 1.251136,
+      "grad_norm": 1.8984375,
+      "learning_rate": 2.9152802661888918e-05,
+      "loss": 0.4987,
+      "step": 4888
+    },
+    {
+      "epoch": 1.251392,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.9148536814265e-05,
+      "loss": 0.4345,
+      "step": 4889
+    },
+    {
+      "epoch": 1.2516479999999999,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.9144270966641073e-05,
+      "loss": 0.4797,
+      "step": 4890
+    },
+    {
+      "epoch": 1.2519040000000001,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.914000511901715e-05,
+      "loss": 0.4341,
+      "step": 4891
+    },
+    {
+      "epoch": 1.25216,
+      "grad_norm": 2.125,
+      "learning_rate": 2.9135739271393225e-05,
+      "loss": 0.4497,
+      "step": 4892
+    },
+    {
+      "epoch": 1.252416,
+      "grad_norm": 2.34375,
+      "learning_rate": 2.9131473423769306e-05,
+      "loss": 0.4376,
+      "step": 4893
+    },
+    {
+      "epoch": 1.252672,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.9127207576145384e-05,
+      "loss": 0.4703,
+      "step": 4894
+    },
+    {
+      "epoch": 1.252928,
+      "grad_norm": 1.9765625,
+      "learning_rate": 2.9122941728521458e-05,
+      "loss": 0.4764,
+      "step": 4895
+    },
+    {
+      "epoch": 1.253184,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.9118675880897532e-05,
+      "loss": 0.4362,
+      "step": 4896
+    },
+    {
+      "epoch": 1.2534399999999999,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.9114410033273613e-05,
+      "loss": 0.4717,
+      "step": 4897
+    },
+    {
+      "epoch": 1.253696,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.911014418564969e-05,
+      "loss": 0.4642,
+      "step": 4898
+    },
+    {
+      "epoch": 1.253952,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.9105878338025765e-05,
+      "loss": 0.4604,
+      "step": 4899
+    },
+    {
+      "epoch": 1.254208,
+      "grad_norm": 2.25,
+      "learning_rate": 2.9101612490401843e-05,
+      "loss": 0.468,
+      "step": 4900
+    },
+    {
+      "epoch": 1.254464,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.9097346642777924e-05,
+      "loss": 0.4383,
+      "step": 4901
+    },
+    {
+      "epoch": 1.25472,
+      "grad_norm": 1.984375,
+      "learning_rate": 2.9093080795153998e-05,
+      "loss": 0.4621,
+      "step": 4902
+    },
+    {
+      "epoch": 1.254976,
+      "grad_norm": 2.078125,
+      "learning_rate": 2.9088814947530073e-05,
+      "loss": 0.4219,
+      "step": 4903
+    },
+    {
+      "epoch": 1.255232,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.9084549099906157e-05,
+      "loss": 0.4848,
+      "step": 4904
+    },
+    {
+      "epoch": 1.255488,
+      "grad_norm": 2.125,
+      "learning_rate": 2.908028325228223e-05,
+      "loss": 0.4564,
+      "step": 4905
+    },
+    {
+      "epoch": 1.255744,
+      "grad_norm": 2.3125,
+      "learning_rate": 2.9076017404658306e-05,
+      "loss": 0.4199,
+      "step": 4906
+    },
+    {
+      "epoch": 1.256,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.9071751557034383e-05,
+      "loss": 0.4314,
+      "step": 4907
+    },
+    {
+      "epoch": 1.256256,
+      "grad_norm": 2.21875,
+      "learning_rate": 2.9067485709410464e-05,
+      "loss": 0.4826,
+      "step": 4908
+    },
+    {
+      "epoch": 1.256512,
+      "grad_norm": 2.3125,
+      "learning_rate": 2.906321986178654e-05,
+      "loss": 0.4288,
+      "step": 4909
+    },
+    {
+      "epoch": 1.256768,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.9058954014162616e-05,
+      "loss": 0.4464,
+      "step": 4910
+    },
+    {
+      "epoch": 1.257024,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.905468816653869e-05,
+      "loss": 0.4412,
+      "step": 4911
+    },
+    {
+      "epoch": 1.25728,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.905042231891477e-05,
+      "loss": 0.4255,
+      "step": 4912
+    },
+    {
+      "epoch": 1.257536,
+      "grad_norm": 2.015625,
+      "learning_rate": 2.9046156471290846e-05,
+      "loss": 0.4873,
+      "step": 4913
+    },
+    {
+      "epoch": 1.257792,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.9041890623666923e-05,
+      "loss": 0.4731,
+      "step": 4914
+    },
+    {
+      "epoch": 1.258048,
+      "grad_norm": 2.25,
+      "learning_rate": 2.9037624776042998e-05,
+      "loss": 0.4431,
+      "step": 4915
+    },
+    {
+      "epoch": 1.2583039999999999,
+      "grad_norm": 2.25,
+      "learning_rate": 2.903335892841908e-05,
+      "loss": 0.4565,
+      "step": 4916
+    },
+    {
+      "epoch": 1.2585600000000001,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.9029093080795156e-05,
+      "loss": 0.4549,
+      "step": 4917
+    },
+    {
+      "epoch": 1.258816,
+      "grad_norm": 2.0625,
+      "learning_rate": 2.902482723317123e-05,
+      "loss": 0.4506,
+      "step": 4918
+    },
+    {
+      "epoch": 1.259072,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.902056138554731e-05,
+      "loss": 0.4781,
+      "step": 4919
+    },
+    {
+      "epoch": 1.259328,
+      "grad_norm": 2.125,
+      "learning_rate": 2.901629553792339e-05,
+      "loss": 0.4571,
+      "step": 4920
+    },
+    {
+      "epoch": 1.259584,
+      "grad_norm": 2.21875,
+      "learning_rate": 2.9012029690299464e-05,
+      "loss": 0.4315,
+      "step": 4921
+    },
+    {
+      "epoch": 1.25984,
+      "grad_norm": 2.078125,
+      "learning_rate": 2.9007763842675538e-05,
+      "loss": 0.4584,
+      "step": 4922
+    },
+    {
+      "epoch": 1.2600959999999999,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.9003497995051622e-05,
+      "loss": 0.4645,
+      "step": 4923
+    },
+    {
+      "epoch": 1.260352,
+      "grad_norm": 2.375,
+      "learning_rate": 2.8999232147427697e-05,
+      "loss": 0.4505,
+      "step": 4924
+    },
+    {
+      "epoch": 1.260608,
+      "grad_norm": 2.328125,
+      "learning_rate": 2.899496629980377e-05,
+      "loss": 0.4662,
+      "step": 4925
+    },
+    {
+      "epoch": 1.260864,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.899070045217985e-05,
+      "loss": 0.4528,
+      "step": 4926
+    },
+    {
+      "epoch": 1.26112,
+      "grad_norm": 2.03125,
+      "learning_rate": 2.898643460455593e-05,
+      "loss": 0.4743,
+      "step": 4927
+    },
+    {
+      "epoch": 1.261376,
+      "grad_norm": 2.265625,
+      "learning_rate": 2.8982168756932004e-05,
+      "loss": 0.4628,
+      "step": 4928
+    },
+    {
+      "epoch": 1.261632,
+      "grad_norm": 2.421875,
+      "learning_rate": 2.897790290930808e-05,
+      "loss": 0.4664,
+      "step": 4929
+    },
+    {
+      "epoch": 1.261888,
+      "grad_norm": 2.359375,
+      "learning_rate": 2.8973637061684156e-05,
+      "loss": 0.4712,
+      "step": 4930
+    },
+    {
+      "epoch": 1.262144,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.8969371214060237e-05,
+      "loss": 0.4546,
+      "step": 4931
+    },
+    {
+      "epoch": 1.2624,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.896510536643631e-05,
+      "loss": 0.4716,
+      "step": 4932
+    },
+    {
+      "epoch": 1.262656,
+      "grad_norm": 2.03125,
+      "learning_rate": 2.896083951881239e-05,
+      "loss": 0.5083,
+      "step": 4933
+    },
+    {
+      "epoch": 1.262912,
+      "grad_norm": 2.265625,
+      "learning_rate": 2.8956573671188463e-05,
+      "loss": 0.4599,
+      "step": 4934
+    },
+    {
+      "epoch": 1.263168,
+      "grad_norm": 2.421875,
+      "learning_rate": 2.8952307823564544e-05,
+      "loss": 0.4378,
+      "step": 4935
+    },
+    {
+      "epoch": 1.263424,
+      "grad_norm": 1.921875,
+      "learning_rate": 2.894804197594062e-05,
+      "loss": 0.4356,
+      "step": 4936
+    },
+    {
+      "epoch": 1.26368,
+      "grad_norm": 1.9375,
+      "learning_rate": 2.8943776128316696e-05,
+      "loss": 0.4503,
+      "step": 4937
+    },
+    {
+      "epoch": 1.263936,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.8939510280692777e-05,
+      "loss": 0.408,
+      "step": 4938
+    },
+    {
+      "epoch": 1.264192,
+      "grad_norm": 1.9921875,
+      "learning_rate": 2.8935244433068855e-05,
+      "loss": 0.4424,
+      "step": 4939
+    },
+    {
+      "epoch": 1.264448,
+      "grad_norm": 2.015625,
+      "learning_rate": 2.893097858544493e-05,
+      "loss": 0.4708,
+      "step": 4940
+    },
+    {
+      "epoch": 1.264704,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.8926712737821003e-05,
+      "loss": 0.4512,
+      "step": 4941
+    },
+    {
+      "epoch": 1.2649599999999999,
+      "grad_norm": 2.328125,
+      "learning_rate": 2.8922446890197084e-05,
+      "loss": 0.4449,
+      "step": 4942
+    },
+    {
+      "epoch": 1.2652160000000001,
+      "grad_norm": 2.3125,
+      "learning_rate": 2.8918181042573162e-05,
+      "loss": 0.4617,
+      "step": 4943
+    },
+    {
+      "epoch": 1.265472,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.8913915194949236e-05,
+      "loss": 0.4754,
+      "step": 4944
+    },
+    {
+      "epoch": 1.265728,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.8909649347325314e-05,
+      "loss": 0.4765,
+      "step": 4945
+    },
+    {
+      "epoch": 1.265984,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.8905383499701395e-05,
+      "loss": 0.4746,
+      "step": 4946
+    },
+    {
+      "epoch": 1.26624,
+      "grad_norm": 2.28125,
+      "learning_rate": 2.890111765207747e-05,
+      "loss": 0.4477,
+      "step": 4947
+    },
+    {
+      "epoch": 1.266496,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.8896851804453547e-05,
+      "loss": 0.457,
+      "step": 4948
+    },
+    {
+      "epoch": 1.2667519999999999,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.889258595682962e-05,
+      "loss": 0.4515,
+      "step": 4949
+    },
+    {
+      "epoch": 1.2670080000000001,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.8888320109205702e-05,
+      "loss": 0.4255,
+      "step": 4950
+    },
+    {
+      "epoch": 1.267264,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.8884054261581776e-05,
+      "loss": 0.4548,
+      "step": 4951
+    },
+    {
+      "epoch": 1.26752,
+      "grad_norm": 1.8828125,
+      "learning_rate": 2.8879788413957854e-05,
+      "loss": 0.4512,
+      "step": 4952
+    },
+    {
+      "epoch": 1.267776,
+      "grad_norm": 2.078125,
+      "learning_rate": 2.8875522566333935e-05,
+      "loss": 0.4796,
+      "step": 4953
+    },
+    {
+      "epoch": 1.268032,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.887125671871001e-05,
+      "loss": 0.4392,
+      "step": 4954
+    },
+    {
+      "epoch": 1.268288,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.8866990871086087e-05,
+      "loss": 0.4584,
+      "step": 4955
+    },
+    {
+      "epoch": 1.268544,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.886272502346216e-05,
+      "loss": 0.4619,
+      "step": 4956
+    },
+    {
+      "epoch": 1.2688,
+      "grad_norm": 2.015625,
+      "learning_rate": 2.8858459175838242e-05,
+      "loss": 0.5105,
+      "step": 4957
+    },
+    {
+      "epoch": 1.269056,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.885419332821432e-05,
+      "loss": 0.4593,
+      "step": 4958
+    },
+    {
+      "epoch": 1.269312,
+      "grad_norm": 2.125,
+      "learning_rate": 2.8849927480590394e-05,
+      "loss": 0.4586,
+      "step": 4959
+    },
+    {
+      "epoch": 1.269568,
+      "grad_norm": 1.9765625,
+      "learning_rate": 2.884566163296647e-05,
+      "loss": 0.4392,
+      "step": 4960
+    },
+    {
+      "epoch": 1.269824,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.884139578534255e-05,
+      "loss": 0.3788,
+      "step": 4961
+    },
+    {
+      "epoch": 1.27008,
+      "grad_norm": 2.0625,
+      "learning_rate": 2.8837129937718627e-05,
+      "loss": 0.4306,
+      "step": 4962
+    },
+    {
+      "epoch": 1.270336,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.88328640900947e-05,
+      "loss": 0.4625,
+      "step": 4963
+    },
+    {
+      "epoch": 1.270592,
+      "grad_norm": 1.8515625,
+      "learning_rate": 2.882859824247078e-05,
+      "loss": 0.4535,
+      "step": 4964
+    },
+    {
+      "epoch": 1.270848,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.882433239484686e-05,
+      "loss": 0.4549,
+      "step": 4965
+    },
+    {
+      "epoch": 1.271104,
+      "grad_norm": 2.125,
+      "learning_rate": 2.8820066547222934e-05,
+      "loss": 0.4542,
+      "step": 4966
+    },
+    {
+      "epoch": 1.27136,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.8815800699599012e-05,
+      "loss": 0.4742,
+      "step": 4967
+    },
+    {
+      "epoch": 1.271616,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.8811534851975086e-05,
+      "loss": 0.4813,
+      "step": 4968
+    },
+    {
+      "epoch": 1.2718720000000001,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.8807269004351167e-05,
+      "loss": 0.4503,
+      "step": 4969
+    },
+    {
+      "epoch": 1.272128,
+      "grad_norm": 2.125,
+      "learning_rate": 2.880300315672724e-05,
+      "loss": 0.4789,
+      "step": 4970
+    },
+    {
+      "epoch": 1.272384,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.879873730910332e-05,
+      "loss": 0.4339,
+      "step": 4971
+    },
+    {
+      "epoch": 1.27264,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.87944714614794e-05,
+      "loss": 0.4288,
+      "step": 4972
+    },
+    {
+      "epoch": 1.272896,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.8790205613855474e-05,
+      "loss": 0.4364,
+      "step": 4973
+    },
+    {
+      "epoch": 1.273152,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.8785939766231552e-05,
+      "loss": 0.4372,
+      "step": 4974
+    },
+    {
+      "epoch": 1.2734079999999999,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.8781673918607626e-05,
+      "loss": 0.4795,
+      "step": 4975
+    },
+    {
+      "epoch": 1.2736640000000001,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.8777408070983707e-05,
+      "loss": 0.4431,
+      "step": 4976
+    },
+    {
+      "epoch": 1.27392,
+      "grad_norm": 2.078125,
+      "learning_rate": 2.8773142223359785e-05,
+      "loss": 0.4779,
+      "step": 4977
+    },
+    {
+      "epoch": 1.274176,
+      "grad_norm": 2.34375,
+      "learning_rate": 2.876887637573586e-05,
+      "loss": 0.4342,
+      "step": 4978
+    },
+    {
+      "epoch": 1.274432,
+      "grad_norm": 2.21875,
+      "learning_rate": 2.8764610528111934e-05,
+      "loss": 0.4563,
+      "step": 4979
+    },
+    {
+      "epoch": 1.274688,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.8760344680488015e-05,
+      "loss": 0.4607,
+      "step": 4980
+    },
+    {
+      "epoch": 1.274944,
+      "grad_norm": 2.0625,
+      "learning_rate": 2.8756078832864092e-05,
+      "loss": 0.4879,
+      "step": 4981
+    },
+    {
+      "epoch": 1.2752,
+      "grad_norm": 2.3125,
+      "learning_rate": 2.8751812985240167e-05,
+      "loss": 0.4662,
+      "step": 4982
+    },
+    {
+      "epoch": 1.275456,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.8747547137616244e-05,
+      "loss": 0.4606,
+      "step": 4983
+    },
+    {
+      "epoch": 1.275712,
+      "grad_norm": 2.015625,
+      "learning_rate": 2.8743281289992325e-05,
+      "loss": 0.4594,
+      "step": 4984
+    },
+    {
+      "epoch": 1.275968,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.87390154423684e-05,
+      "loss": 0.4264,
+      "step": 4985
+    },
+    {
+      "epoch": 1.276224,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.8734749594744477e-05,
+      "loss": 0.4332,
+      "step": 4986
+    },
+    {
+      "epoch": 1.27648,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.8730483747120558e-05,
+      "loss": 0.4797,
+      "step": 4987
+    },
+    {
+      "epoch": 1.276736,
+      "grad_norm": 2.3125,
+      "learning_rate": 2.8726217899496633e-05,
+      "loss": 0.4523,
+      "step": 4988
+    },
+    {
+      "epoch": 1.276992,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.8721952051872707e-05,
+      "loss": 0.45,
+      "step": 4989
+    },
+    {
+      "epoch": 1.277248,
+      "grad_norm": 1.8671875,
+      "learning_rate": 2.8717686204248784e-05,
+      "loss": 0.4661,
+      "step": 4990
+    },
+    {
+      "epoch": 1.277504,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.8713420356624865e-05,
+      "loss": 0.4456,
+      "step": 4991
+    },
+    {
+      "epoch": 1.27776,
+      "grad_norm": 2.328125,
+      "learning_rate": 2.870915450900094e-05,
+      "loss": 0.482,
+      "step": 4992
+    },
+    {
+      "epoch": 1.278016,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.8704888661377017e-05,
+      "loss": 0.4273,
+      "step": 4993
+    },
+    {
+      "epoch": 1.278272,
+      "grad_norm": 2.25,
+      "learning_rate": 2.870062281375309e-05,
+      "loss": 0.4383,
+      "step": 4994
+    },
+    {
+      "epoch": 1.278528,
+      "grad_norm": 2.265625,
+      "learning_rate": 2.8696356966129173e-05,
+      "loss": 0.4647,
+      "step": 4995
+    },
+    {
+      "epoch": 1.278784,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.869209111850525e-05,
+      "loss": 0.4304,
+      "step": 4996
+    },
+    {
+      "epoch": 1.27904,
+      "grad_norm": 1.984375,
+      "learning_rate": 2.8687825270881325e-05,
+      "loss": 0.4818,
+      "step": 4997
+    },
+    {
+      "epoch": 1.279296,
+      "grad_norm": 2.25,
+      "learning_rate": 2.86835594232574e-05,
+      "loss": 0.4102,
+      "step": 4998
+    },
+    {
+      "epoch": 1.279552,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.867929357563348e-05,
+      "loss": 0.4788,
+      "step": 4999
+    },
+    {
+      "epoch": 1.279808,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.8675027728009558e-05,
+      "loss": 0.4808,
+      "step": 5000
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 11721,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 5000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.2838790556347498e+19,
+  "train_batch_size": 32,
+  "trial_name": null,
+  "trial_params": null
+}